1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 If a signal of stream rises above this level it will affect the gain
396 By default it is 0.125. Range is between 0.00097563 and 1.
399 Set a ratio by which the signal is reduced. 1:2 means that if the level
400 rose 4dB above the threshold, it will be only 2dB above after the reduction.
401 Default is 2. Range is between 1 and 20.
404 Amount of milliseconds the signal has to rise above the threshold before gain
405 reduction starts. Default is 20. Range is between 0.01 and 2000.
408 Amount of milliseconds the signal has to fall below the threshold before
409 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
412 Set the amount by how much signal will be amplified after processing.
413 Default is 1. Range is from 1 to 64.
416 Curve the sharp knee around the threshold to enter gain reduction more softly.
417 Default is 2.82843. Range is between 1 and 8.
420 Choose if the @code{average} level between all channels of input stream
421 or the louder(@code{maximum}) channel of input stream affects the
422 reduction. Default is @code{average}.
425 Should the exact signal be taken in case of @code{peak} or an RMS one in case
426 of @code{rms}. Default is @code{rms} which is mostly smoother.
429 How much to use compressed signal in output. Default is 1.
430 Range is between 0 and 1.
434 Simple audio dynamic range commpression/expansion filter.
436 The filter accepts the following options:
440 Set contrast. Default is 33. Allowed range is between 0 and 100.
445 Copy the input audio source unchanged to the output. This is mainly useful for
450 Apply cross fade from one input audio stream to another input audio stream.
451 The cross fade is applied for specified duration near the end of first stream.
453 The filter accepts the following options:
457 Specify the number of samples for which the cross fade effect has to last.
458 At the end of the cross fade effect the first input audio will be completely
459 silent. Default is 44100.
462 Specify the duration of the cross fade effect. See
463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
464 for the accepted syntax.
465 By default the duration is determined by @var{nb_samples}.
466 If set this option is used instead of @var{nb_samples}.
469 Should first stream end overlap with second stream start. Default is enabled.
472 Set curve for cross fade transition for first stream.
475 Set curve for cross fade transition for second stream.
477 For description of available curve types see @ref{afade} filter description.
484 Cross fade from one input to another:
486 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
490 Cross fade from one input to another but without overlapping:
492 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Split audio stream into several bands.
499 This filter splits audio stream into two or more frequency ranges.
500 Summing all streams back will give flat output.
502 The filter accepts the following options:
506 Set split frequencies. Those must be positive and increasing.
509 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
510 Default is @var{4th}.
515 Reduce audio bit resolution.
517 This filter is bit crusher with enhanced functionality. A bit crusher
518 is used to audibly reduce number of bits an audio signal is sampled
519 with. This doesn't change the bit depth at all, it just produces the
520 effect. Material reduced in bit depth sounds more harsh and "digital".
521 This filter is able to even round to continuous values instead of discrete
523 Additionally it has a D/C offset which results in different crushing of
524 the lower and the upper half of the signal.
525 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
527 Another feature of this filter is the logarithmic mode.
528 This setting switches from linear distances between bits to logarithmic ones.
529 The result is a much more "natural" sounding crusher which doesn't gate low
530 signals for example. The human ear has a logarithmic perception,
531 so this kind of crushing is much more pleasant.
532 Logarithmic crushing is also able to get anti-aliased.
534 The filter accepts the following options:
550 Can be linear: @code{lin} or logarithmic: @code{log}.
559 Set sample reduction.
562 Enable LFO. By default disabled.
573 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
577 Remove impulsive noise from input audio.
579 Samples detected as impulsive noise are replaced by interpolated samples using
580 autoregressive modelling.
584 Set window size, in milliseconds. Allowed range is from @code{10} to
585 @code{100}. Default value is @code{55} milliseconds.
586 This sets size of window which will be processed at once.
589 Set window overlap, in percentage of window size. Allowed range is from
590 @code{50} to @code{95}. Default value is @code{75} percent.
591 Setting this to a very high value increases impulsive noise removal but makes
592 whole process much slower.
595 Set autoregression order, in percentage of window size. Allowed range is from
596 @code{0} to @code{25}. Default value is @code{2} percent. This option also
597 controls quality of interpolated samples using neighbour good samples.
600 Set threshold value. Allowed range is from @code{1} to @code{100}.
601 Default value is @code{2}.
602 This controls the strength of impulsive noise which is going to be removed.
603 The lower value, the more samples will be detected as impulsive noise.
606 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
607 @code{10}. Default value is @code{2}.
608 If any two samples deteced as noise are spaced less than this value then any
609 sample inbetween those two samples will be also detected as noise.
614 It accepts the following values:
617 Select overlap-add method. Even not interpolated samples are slightly
618 changed with this method.
621 Select overlap-save method. Not interpolated samples remain unchanged.
624 Default value is @code{a}.
628 Remove clipped samples from input audio.
630 Samples detected as clipped are replaced by interpolated samples using
631 autoregressive modelling.
635 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
636 Default value is @code{55} milliseconds.
637 This sets size of window which will be processed at once.
640 Set window overlap, in percentage of window size. Allowed range is from @code{50}
641 to @code{95}. Default value is @code{75} percent.
644 Set autoregression order, in percentage of window size. Allowed range is from
645 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
646 quality of interpolated samples using neighbour good samples.
649 Set threshold value. Allowed range is from @code{1} to @code{100}.
650 Default value is @code{10}. Higher values make clip detection less aggressive.
653 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
654 Default value is @code{1000}. Higher values make clip detection less aggressive.
659 It accepts the following values:
662 Select overlap-add method. Even not interpolated samples are slightly changed
666 Select overlap-save method. Not interpolated samples remain unchanged.
669 Default value is @code{a}.
674 Delay one or more audio channels.
676 Samples in delayed channel are filled with silence.
678 The filter accepts the following option:
682 Set list of delays in milliseconds for each channel separated by '|'.
683 Unused delays will be silently ignored. If number of given delays is
684 smaller than number of channels all remaining channels will not be delayed.
685 If you want to delay exact number of samples, append 'S' to number.
692 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
693 the second channel (and any other channels that may be present) unchanged.
699 Delay second channel by 500 samples, the third channel by 700 samples and leave
700 the first channel (and any other channels that may be present) unchanged.
706 @section aderivative, aintegral
708 Compute derivative/integral of audio stream.
710 Applying both filters one after another produces original audio.
714 Apply echoing to the input audio.
716 Echoes are reflected sound and can occur naturally amongst mountains
717 (and sometimes large buildings) when talking or shouting; digital echo
718 effects emulate this behaviour and are often used to help fill out the
719 sound of a single instrument or vocal. The time difference between the
720 original signal and the reflection is the @code{delay}, and the
721 loudness of the reflected signal is the @code{decay}.
722 Multiple echoes can have different delays and decays.
724 A description of the accepted parameters follows.
728 Set input gain of reflected signal. Default is @code{0.6}.
731 Set output gain of reflected signal. Default is @code{0.3}.
734 Set list of time intervals in milliseconds between original signal and reflections
735 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
736 Default is @code{1000}.
739 Set list of loudness of reflected signals separated by '|'.
740 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
741 Default is @code{0.5}.
748 Make it sound as if there are twice as many instruments as are actually playing:
750 aecho=0.8:0.88:60:0.4
754 If delay is very short, then it sound like a (metallic) robot playing music:
760 A longer delay will sound like an open air concert in the mountains:
762 aecho=0.8:0.9:1000:0.3
766 Same as above but with one more mountain:
768 aecho=0.8:0.9:1000|1800:0.3|0.25
773 Audio emphasis filter creates or restores material directly taken from LPs or
774 emphased CDs with different filter curves. E.g. to store music on vinyl the
775 signal has to be altered by a filter first to even out the disadvantages of
776 this recording medium.
777 Once the material is played back the inverse filter has to be applied to
778 restore the distortion of the frequency response.
780 The filter accepts the following options:
790 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
791 use @code{production} mode. Default is @code{reproduction} mode.
794 Set filter type. Selects medium. Can be one of the following:
806 select Compact Disc (CD).
812 select 50µs (FM-KF).
814 select 75µs (FM-KF).
820 Modify an audio signal according to the specified expressions.
822 This filter accepts one or more expressions (one for each channel),
823 which are evaluated and used to modify a corresponding audio signal.
825 It accepts the following parameters:
829 Set the '|'-separated expressions list for each separate channel. If
830 the number of input channels is greater than the number of
831 expressions, the last specified expression is used for the remaining
834 @item channel_layout, c
835 Set output channel layout. If not specified, the channel layout is
836 specified by the number of expressions. If set to @samp{same}, it will
837 use by default the same input channel layout.
840 Each expression in @var{exprs} can contain the following constants and functions:
844 channel number of the current expression
847 number of the evaluated sample, starting from 0
853 time of the evaluated sample expressed in seconds
856 @item nb_out_channels
857 input and output number of channels
860 the value of input channel with number @var{CH}
863 Note: this filter is slow. For faster processing you should use a
872 aeval=val(ch)/2:c=same
876 Invert phase of the second channel:
885 Apply fade-in/out effect to input audio.
887 A description of the accepted parameters follows.
891 Specify the effect type, can be either @code{in} for fade-in, or
892 @code{out} for a fade-out effect. Default is @code{in}.
894 @item start_sample, ss
895 Specify the number of the start sample for starting to apply the fade
896 effect. Default is 0.
899 Specify the number of samples for which the fade effect has to last. At
900 the end of the fade-in effect the output audio will have the same
901 volume as the input audio, at the end of the fade-out transition
902 the output audio will be silence. Default is 44100.
905 Specify the start time of the fade effect. Default is 0.
906 The value must be specified as a time duration; see
907 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
908 for the accepted syntax.
909 If set this option is used instead of @var{start_sample}.
912 Specify the duration of the fade effect. See
913 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
914 for the accepted syntax.
915 At the end of the fade-in effect the output audio will have the same
916 volume as the input audio, at the end of the fade-out transition
917 the output audio will be silence.
918 By default the duration is determined by @var{nb_samples}.
919 If set this option is used instead of @var{nb_samples}.
922 Set curve for fade transition.
924 It accepts the following values:
927 select triangular, linear slope (default)
929 select quarter of sine wave
931 select half of sine wave
933 select exponential sine wave
937 select inverted parabola
951 select inverted quarter of sine wave
953 select inverted half of sine wave
955 select double-exponential seat
957 select double-exponential sigmoid
959 select logistic sigmoid
967 Fade in first 15 seconds of audio:
973 Fade out last 25 seconds of a 900 seconds audio:
975 afade=t=out:st=875:d=25
980 Denoise audio samples with FFT.
982 A description of the accepted parameters follows.
986 Set the noise reduction in dB, allowed range is 0.01 to 97.
987 Default value is 12 dB.
990 Set the noise floor in dB, allowed range is -80 to -20.
991 Default value is -50 dB.
996 It accepts the following values:
1005 Select shellac noise.
1008 Select custom noise, defined in @code{bn} option.
1010 Default value is white noise.
1014 Set custom band noise for every one of 15 bands.
1015 Bands are separated by ' ' or '|'.
1018 Set the residual floor in dB, allowed range is -80 to -20.
1019 Default value is -38 dB.
1022 Enable noise tracking. By default is disabled.
1023 With this enabled, noise floor is automatically adjusted.
1026 Enable residual tracking. By default is disabled.
1029 Set the output mode.
1031 It accepts the following values:
1034 Pass input unchanged.
1037 Pass noise filtered out.
1042 Default value is @var{o}.
1046 @subsection Commands
1048 This filter supports the following commands:
1050 @item sample_noise, sn
1051 Start or stop measuring noise profile.
1052 Syntax for the command is : "start" or "stop" string.
1053 After measuring noise profile is stopped it will be
1054 automatically applied in filtering.
1056 @item noise_reduction, nr
1057 Change noise reduction. Argument is single float number.
1058 Syntax for the command is : "@var{noise_reduction}"
1060 @item noise_floor, nf
1061 Change noise floor. Argument is single float number.
1062 Syntax for the command is : "@var{noise_floor}"
1064 @item output_mode, om
1065 Change output mode operation.
1066 Syntax for the command is : "i", "o" or "n" string.
1070 Apply arbitrary expressions to samples in frequency domain.
1074 Set frequency domain real expression for each separate channel separated
1075 by '|'. Default is "1".
1076 If the number of input channels is greater than the number of
1077 expressions, the last specified expression is used for the remaining
1081 Set frequency domain imaginary expression for each separate channel
1082 separated by '|'. If not set, @var{real} option is used.
1084 Each expression in @var{real} and @var{imag} can contain the following
1092 current frequency bin number
1095 number of available bins
1098 channel number of the current expression
1110 It accepts the following values:
1126 Default is @code{w4096}
1129 Set window function. Default is @code{hann}.
1132 Set window overlap. If set to 1, the recommended overlap for selected
1133 window function will be picked. Default is @code{0.75}.
1136 @subsection Examples
1140 Leave almost only low frequencies in audio:
1142 afftfilt="1-clip((b/nb)*b,0,1)"
1149 Apply an arbitrary Frequency Impulse Response filter.
1151 This filter is designed for applying long FIR filters,
1152 up to 60 seconds long.
1154 It can be used as component for digital crossover filters,
1155 room equalization, cross talk cancellation, wavefield synthesis,
1156 auralization, ambiophonics and ambisonics.
1158 This filter uses second stream as FIR coefficients.
1159 If second stream holds single channel, it will be used
1160 for all input channels in first stream, otherwise
1161 number of channels in second stream must be same as
1162 number of channels in first stream.
1164 It accepts the following parameters:
1168 Set dry gain. This sets input gain.
1171 Set wet gain. This sets final output gain.
1174 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1177 Enable applying gain measured from power of IR.
1179 Set which approach to use for auto gain measurement.
1183 Do not apply any gain.
1186 select peak gain, very conservative approach. This is default value.
1189 select DC gain, limited application.
1192 select gain to noise approach, this is most popular one.
1196 Set gain to be applied to IR coefficients before filtering.
1197 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1200 Set format of IR stream. Can be @code{mono} or @code{input}.
1201 Default is @code{input}.
1204 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1205 Allowed range is 0.1 to 60 seconds.
1208 Show IR frequency reponse, magnitude and phase in additional video stream.
1209 By default it is disabled.
1212 Set for which IR channel to display frequency response. By default is first channel
1213 displayed. This option is used only when @var{response} is enabled.
1216 Set video stream size. This option is used only when @var{response} is enabled.
1219 @subsection Examples
1223 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1225 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1232 Set output format constraints for the input audio. The framework will
1233 negotiate the most appropriate format to minimize conversions.
1235 It accepts the following parameters:
1239 A '|'-separated list of requested sample formats.
1242 A '|'-separated list of requested sample rates.
1244 @item channel_layouts
1245 A '|'-separated list of requested channel layouts.
1247 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1248 for the required syntax.
1251 If a parameter is omitted, all values are allowed.
1253 Force the output to either unsigned 8-bit or signed 16-bit stereo
1255 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1260 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1261 processing reduces disturbing noise between useful signals.
1263 Gating is done by detecting the volume below a chosen level @var{threshold}
1264 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1265 floor is set via @var{range}. Because an exact manipulation of the signal
1266 would cause distortion of the waveform the reduction can be levelled over
1267 time. This is done by setting @var{attack} and @var{release}.
1269 @var{attack} determines how long the signal has to fall below the threshold
1270 before any reduction will occur and @var{release} sets the time the signal
1271 has to rise above the threshold to reduce the reduction again.
1272 Shorter signals than the chosen attack time will be left untouched.
1276 Set input level before filtering.
1277 Default is 1. Allowed range is from 0.015625 to 64.
1280 Set the level of gain reduction when the signal is below the threshold.
1281 Default is 0.06125. Allowed range is from 0 to 1.
1284 If a signal rises above this level the gain reduction is released.
1285 Default is 0.125. Allowed range is from 0 to 1.
1288 Set a ratio by which the signal is reduced.
1289 Default is 2. Allowed range is from 1 to 9000.
1292 Amount of milliseconds the signal has to rise above the threshold before gain
1294 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1297 Amount of milliseconds the signal has to fall below the threshold before the
1298 reduction is increased again. Default is 250 milliseconds.
1299 Allowed range is from 0.01 to 9000.
1302 Set amount of amplification of signal after processing.
1303 Default is 1. Allowed range is from 1 to 64.
1306 Curve the sharp knee around the threshold to enter gain reduction more softly.
1307 Default is 2.828427125. Allowed range is from 1 to 8.
1310 Choose if exact signal should be taken for detection or an RMS like one.
1311 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1314 Choose if the average level between all channels or the louder channel affects
1316 Default is @code{average}. Can be @code{average} or @code{maximum}.
1321 Apply an arbitrary Infinite Impulse Response filter.
1323 It accepts the following parameters:
1327 Set numerator/zeros coefficients.
1330 Set denominator/poles coefficients.
1342 Set coefficients format.
1348 Z-plane zeros/poles, cartesian (default)
1350 Z-plane zeros/poles, polar radians
1352 Z-plane zeros/poles, polar degrees
1356 Set kind of processing.
1357 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1360 Set filtering precision.
1364 double-precision floating-point (default)
1366 single-precision floating-point
1374 Show IR frequency reponse, magnitude and phase in additional video stream.
1375 By default it is disabled.
1378 Set for which IR channel to display frequency response. By default is first channel
1379 displayed. This option is used only when @var{response} is enabled.
1382 Set video stream size. This option is used only when @var{response} is enabled.
1385 Coefficients in @code{tf} format are separated by spaces and are in ascending
1388 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1389 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1392 Different coefficients and gains can be provided for every channel, in such case
1393 use '|' to separate coefficients or gains. Last provided coefficients will be
1394 used for all remaining channels.
1396 @subsection Examples
1400 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1402 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
1406 Same as above but in @code{zp} format:
1408 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
1414 The limiter prevents an input signal from rising over a desired threshold.
1415 This limiter uses lookahead technology to prevent your signal from distorting.
1416 It means that there is a small delay after the signal is processed. Keep in mind
1417 that the delay it produces is the attack time you set.
1419 The filter accepts the following options:
1423 Set input gain. Default is 1.
1426 Set output gain. Default is 1.
1429 Don't let signals above this level pass the limiter. Default is 1.
1432 The limiter will reach its attenuation level in this amount of time in
1433 milliseconds. Default is 5 milliseconds.
1436 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1437 Default is 50 milliseconds.
1440 When gain reduction is always needed ASC takes care of releasing to an
1441 average reduction level rather than reaching a reduction of 0 in the release
1445 Select how much the release time is affected by ASC, 0 means nearly no changes
1446 in release time while 1 produces higher release times.
1449 Auto level output signal. Default is enabled.
1450 This normalizes audio back to 0dB if enabled.
1453 Depending on picked setting it is recommended to upsample input 2x or 4x times
1454 with @ref{aresample} before applying this filter.
1458 Apply a two-pole all-pass filter with central frequency (in Hz)
1459 @var{frequency}, and filter-width @var{width}.
1460 An all-pass filter changes the audio's frequency to phase relationship
1461 without changing its frequency to amplitude relationship.
1463 The filter accepts the following options:
1467 Set frequency in Hz.
1470 Set method to specify band-width of filter.
1485 Specify the band-width of a filter in width_type units.
1488 Specify which channels to filter, by default all available are filtered.
1491 @subsection Commands
1493 This filter supports the following commands:
1496 Change allpass frequency.
1497 Syntax for the command is : "@var{frequency}"
1500 Change allpass width_type.
1501 Syntax for the command is : "@var{width_type}"
1504 Change allpass width.
1505 Syntax for the command is : "@var{width}"
1512 The filter accepts the following options:
1516 Set the number of loops. Setting this value to -1 will result in infinite loops.
1520 Set maximal number of samples. Default is 0.
1523 Set first sample of loop. Default is 0.
1529 Merge two or more audio streams into a single multi-channel stream.
1531 The filter accepts the following options:
1536 Set the number of inputs. Default is 2.
1540 If the channel layouts of the inputs are disjoint, and therefore compatible,
1541 the channel layout of the output will be set accordingly and the channels
1542 will be reordered as necessary. If the channel layouts of the inputs are not
1543 disjoint, the output will have all the channels of the first input then all
1544 the channels of the second input, in that order, and the channel layout of
1545 the output will be the default value corresponding to the total number of
1548 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1549 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1550 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1551 first input, b1 is the first channel of the second input).
1553 On the other hand, if both input are in stereo, the output channels will be
1554 in the default order: a1, a2, b1, b2, and the channel layout will be
1555 arbitrarily set to 4.0, which may or may not be the expected value.
1557 All inputs must have the same sample rate, and format.
1559 If inputs do not have the same duration, the output will stop with the
1562 @subsection Examples
1566 Merge two mono files into a stereo stream:
1568 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1572 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1574 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
1580 Mixes multiple audio inputs into a single output.
1582 Note that this filter only supports float samples (the @var{amerge}
1583 and @var{pan} audio filters support many formats). If the @var{amix}
1584 input has integer samples then @ref{aresample} will be automatically
1585 inserted to perform the conversion to float samples.
1589 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1591 will mix 3 input audio streams to a single output with the same duration as the
1592 first input and a dropout transition time of 3 seconds.
1594 It accepts the following parameters:
1598 The number of inputs. If unspecified, it defaults to 2.
1601 How to determine the end-of-stream.
1605 The duration of the longest input. (default)
1608 The duration of the shortest input.
1611 The duration of the first input.
1615 @item dropout_transition
1616 The transition time, in seconds, for volume renormalization when an input
1617 stream ends. The default value is 2 seconds.
1620 Specify weight of each input audio stream as sequence.
1621 Each weight is separated by space. By default all inputs have same weight.
1626 Multiply first audio stream with second audio stream and store result
1627 in output audio stream. Multiplication is done by multiplying each
1628 sample from first stream with sample at same position from second stream.
1630 With this element-wise multiplication one can create amplitude fades and
1631 amplitude modulations.
1633 @section anequalizer
1635 High-order parametric multiband equalizer for each channel.
1637 It accepts the following parameters:
1641 This option string is in format:
1642 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1643 Each equalizer band is separated by '|'.
1647 Set channel number to which equalization will be applied.
1648 If input doesn't have that channel the entry is ignored.
1651 Set central frequency for band.
1652 If input doesn't have that frequency the entry is ignored.
1655 Set band width in hertz.
1658 Set band gain in dB.
1661 Set filter type for band, optional, can be:
1665 Butterworth, this is default.
1676 With this option activated frequency response of anequalizer is displayed
1680 Set video stream size. Only useful if curves option is activated.
1683 Set max gain that will be displayed. Only useful if curves option is activated.
1684 Setting this to a reasonable value makes it possible to display gain which is derived from
1685 neighbour bands which are too close to each other and thus produce higher gain
1686 when both are activated.
1689 Set frequency scale used to draw frequency response in video output.
1690 Can be linear or logarithmic. Default is logarithmic.
1693 Set color for each channel curve which is going to be displayed in video stream.
1694 This is list of color names separated by space or by '|'.
1695 Unrecognised or missing colors will be replaced by white color.
1698 @subsection Examples
1702 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1703 for first 2 channels using Chebyshev type 1 filter:
1705 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1709 @subsection Commands
1711 This filter supports the following commands:
1714 Alter existing filter parameters.
1715 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1717 @var{fN} is existing filter number, starting from 0, if no such filter is available
1719 @var{freq} set new frequency parameter.
1720 @var{width} set new width parameter in herz.
1721 @var{gain} set new gain parameter in dB.
1723 Full filter invocation with asendcmd may look like this:
1724 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1729 Pass the audio source unchanged to the output.
1733 Pad the end of an audio stream with silence.
1735 This can be used together with @command{ffmpeg} @option{-shortest} to
1736 extend audio streams to the same length as the video stream.
1738 A description of the accepted options follows.
1742 Set silence packet size. Default value is 4096.
1745 Set the number of samples of silence to add to the end. After the
1746 value is reached, the stream is terminated. This option is mutually
1747 exclusive with @option{whole_len}.
1750 Set the minimum total number of samples in the output audio stream. If
1751 the value is longer than the input audio length, silence is added to
1752 the end, until the value is reached. This option is mutually exclusive
1753 with @option{pad_len}.
1756 If neither the @option{pad_len} nor the @option{whole_len} option is
1757 set, the filter will add silence to the end of the input stream
1760 @subsection Examples
1764 Add 1024 samples of silence to the end of the input:
1770 Make sure the audio output will contain at least 10000 samples, pad
1771 the input with silence if required:
1773 apad=whole_len=10000
1777 Use @command{ffmpeg} to pad the audio input with silence, so that the
1778 video stream will always result the shortest and will be converted
1779 until the end in the output file when using the @option{shortest}
1782 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1787 Add a phasing effect to the input audio.
1789 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1790 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1792 A description of the accepted parameters follows.
1796 Set input gain. Default is 0.4.
1799 Set output gain. Default is 0.74
1802 Set delay in milliseconds. Default is 3.0.
1805 Set decay. Default is 0.4.
1808 Set modulation speed in Hz. Default is 0.5.
1811 Set modulation type. Default is triangular.
1813 It accepts the following values:
1822 Audio pulsator is something between an autopanner and a tremolo.
1823 But it can produce funny stereo effects as well. Pulsator changes the volume
1824 of the left and right channel based on a LFO (low frequency oscillator) with
1825 different waveforms and shifted phases.
1826 This filter have the ability to define an offset between left and right
1827 channel. An offset of 0 means that both LFO shapes match each other.
1828 The left and right channel are altered equally - a conventional tremolo.
1829 An offset of 50% means that the shape of the right channel is exactly shifted
1830 in phase (or moved backwards about half of the frequency) - pulsator acts as
1831 an autopanner. At 1 both curves match again. Every setting in between moves the
1832 phase shift gapless between all stages and produces some "bypassing" sounds with
1833 sine and triangle waveforms. The more you set the offset near 1 (starting from
1834 the 0.5) the faster the signal passes from the left to the right speaker.
1836 The filter accepts the following options:
1840 Set input gain. By default it is 1. Range is [0.015625 - 64].
1843 Set output gain. By default it is 1. Range is [0.015625 - 64].
1846 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1847 sawup or sawdown. Default is sine.
1850 Set modulation. Define how much of original signal is affected by the LFO.
1853 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1856 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1859 Set pulse width. Default is 1. Allowed range is [0 - 2].
1862 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1865 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1869 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1873 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1874 if timing is set to hz.
1880 Resample the input audio to the specified parameters, using the
1881 libswresample library. If none are specified then the filter will
1882 automatically convert between its input and output.
1884 This filter is also able to stretch/squeeze the audio data to make it match
1885 the timestamps or to inject silence / cut out audio to make it match the
1886 timestamps, do a combination of both or do neither.
1888 The filter accepts the syntax
1889 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1890 expresses a sample rate and @var{resampler_options} is a list of
1891 @var{key}=@var{value} pairs, separated by ":". See the
1892 @ref{Resampler Options,,"Resampler Options" section in the
1893 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1894 for the complete list of supported options.
1896 @subsection Examples
1900 Resample the input audio to 44100Hz:
1906 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1907 samples per second compensation:
1909 aresample=async=1000
1915 Reverse an audio clip.
1917 Warning: This filter requires memory to buffer the entire clip, so trimming
1920 @subsection Examples
1924 Take the first 5 seconds of a clip, and reverse it.
1926 atrim=end=5,areverse
1930 @section asetnsamples
1932 Set the number of samples per each output audio frame.
1934 The last output packet may contain a different number of samples, as
1935 the filter will flush all the remaining samples when the input audio
1938 The filter accepts the following options:
1942 @item nb_out_samples, n
1943 Set the number of frames per each output audio frame. The number is
1944 intended as the number of samples @emph{per each channel}.
1945 Default value is 1024.
1948 If set to 1, the filter will pad the last audio frame with zeroes, so
1949 that the last frame will contain the same number of samples as the
1950 previous ones. Default value is 1.
1953 For example, to set the number of per-frame samples to 1234 and
1954 disable padding for the last frame, use:
1956 asetnsamples=n=1234:p=0
1961 Set the sample rate without altering the PCM data.
1962 This will result in a change of speed and pitch.
1964 The filter accepts the following options:
1967 @item sample_rate, r
1968 Set the output sample rate. Default is 44100 Hz.
1973 Show a line containing various information for each input audio frame.
1974 The input audio is not modified.
1976 The shown line contains a sequence of key/value pairs of the form
1977 @var{key}:@var{value}.
1979 The following values are shown in the output:
1983 The (sequential) number of the input frame, starting from 0.
1986 The presentation timestamp of the input frame, in time base units; the time base
1987 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1990 The presentation timestamp of the input frame in seconds.
1993 position of the frame in the input stream, -1 if this information in
1994 unavailable and/or meaningless (for example in case of synthetic audio)
2003 The sample rate for the audio frame.
2006 The number of samples (per channel) in the frame.
2009 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2010 audio, the data is treated as if all the planes were concatenated.
2012 @item plane_checksums
2013 A list of Adler-32 checksums for each data plane.
2019 Display time domain statistical information about the audio channels.
2020 Statistics are calculated and displayed for each audio channel and,
2021 where applicable, an overall figure is also given.
2023 It accepts the following option:
2026 Short window length in seconds, used for peak and trough RMS measurement.
2027 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2031 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2032 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2035 Available keys for each channel are:
2071 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2072 this @code{lavfi.astats.Overall.Peak_count}.
2074 For description what each key means read below.
2077 Set number of frame after which stats are going to be recalculated.
2078 Default is disabled.
2081 A description of each shown parameter follows:
2085 Mean amplitude displacement from zero.
2088 Minimal sample level.
2091 Maximal sample level.
2093 @item Min difference
2094 Minimal difference between two consecutive samples.
2096 @item Max difference
2097 Maximal difference between two consecutive samples.
2099 @item Mean difference
2100 Mean difference between two consecutive samples.
2101 The average of each difference between two consecutive samples.
2103 @item RMS difference
2104 Root Mean Square difference between two consecutive samples.
2108 Standard peak and RMS level measured in dBFS.
2112 Peak and trough values for RMS level measured over a short window.
2115 Standard ratio of peak to RMS level (note: not in dB).
2118 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2119 (i.e. either @var{Min level} or @var{Max level}).
2122 Number of occasions (not the number of samples) that the signal attained either
2123 @var{Min level} or @var{Max level}.
2126 Overall bit depth of audio. Number of bits used for each sample.
2129 Measured dynamic range of audio in dB.
2131 @item Zero crossings
2132 Number of points where the waveform crosses the zero level axis.
2134 @item Zero crossings rate
2135 Rate of Zero crossings and number of audio samples.
2142 The filter accepts exactly one parameter, the audio tempo. If not
2143 specified then the filter will assume nominal 1.0 tempo. Tempo must
2144 be in the [0.5, 100.0] range.
2146 Note that tempo greater than 2 will skip some samples rather than
2147 blend them in. If for any reason this is a concern it is always
2148 possible to daisy-chain several instances of atempo to achieve the
2149 desired product tempo.
2151 @subsection Examples
2155 Slow down audio to 80% tempo:
2161 To speed up audio to 300% tempo:
2167 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2169 atempo=sqrt(3),atempo=sqrt(3)
2175 Trim the input so that the output contains one continuous subpart of the input.
2177 It accepts the following parameters:
2180 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2181 sample with the timestamp @var{start} will be the first sample in the output.
2184 Specify time of the first audio sample that will be dropped, i.e. the
2185 audio sample immediately preceding the one with the timestamp @var{end} will be
2186 the last sample in the output.
2189 Same as @var{start}, except this option sets the start timestamp in samples
2193 Same as @var{end}, except this option sets the end timestamp in samples instead
2197 The maximum duration of the output in seconds.
2200 The number of the first sample that should be output.
2203 The number of the first sample that should be dropped.
2206 @option{start}, @option{end}, and @option{duration} are expressed as time
2207 duration specifications; see
2208 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2210 Note that the first two sets of the start/end options and the @option{duration}
2211 option look at the frame timestamp, while the _sample options simply count the
2212 samples that pass through the filter. So start/end_pts and start/end_sample will
2213 give different results when the timestamps are wrong, inexact or do not start at
2214 zero. Also note that this filter does not modify the timestamps. If you wish
2215 to have the output timestamps start at zero, insert the asetpts filter after the
2218 If multiple start or end options are set, this filter tries to be greedy and
2219 keep all samples that match at least one of the specified constraints. To keep
2220 only the part that matches all the constraints at once, chain multiple atrim
2223 The defaults are such that all the input is kept. So it is possible to set e.g.
2224 just the end values to keep everything before the specified time.
2229 Drop everything except the second minute of input:
2231 ffmpeg -i INPUT -af atrim=60:120
2235 Keep only the first 1000 samples:
2237 ffmpeg -i INPUT -af atrim=end_sample=1000
2244 Apply a two-pole Butterworth band-pass filter with central
2245 frequency @var{frequency}, and (3dB-point) band-width width.
2246 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2247 instead of the default: constant 0dB peak gain.
2248 The filter roll off at 6dB per octave (20dB per decade).
2250 The filter accepts the following options:
2254 Set the filter's central frequency. Default is @code{3000}.
2257 Constant skirt gain if set to 1. Defaults to 0.
2260 Set method to specify band-width of filter.
2275 Specify the band-width of a filter in width_type units.
2278 Specify which channels to filter, by default all available are filtered.
2281 @subsection Commands
2283 This filter supports the following commands:
2286 Change bandpass frequency.
2287 Syntax for the command is : "@var{frequency}"
2290 Change bandpass width_type.
2291 Syntax for the command is : "@var{width_type}"
2294 Change bandpass width.
2295 Syntax for the command is : "@var{width}"
2300 Apply a two-pole Butterworth band-reject filter with central
2301 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2302 The filter roll off at 6dB per octave (20dB per decade).
2304 The filter accepts the following options:
2308 Set the filter's central frequency. Default is @code{3000}.
2311 Set method to specify band-width of filter.
2326 Specify the band-width of a filter in width_type units.
2329 Specify which channels to filter, by default all available are filtered.
2332 @subsection Commands
2334 This filter supports the following commands:
2337 Change bandreject frequency.
2338 Syntax for the command is : "@var{frequency}"
2341 Change bandreject width_type.
2342 Syntax for the command is : "@var{width_type}"
2345 Change bandreject width.
2346 Syntax for the command is : "@var{width}"
2349 @section bass, lowshelf
2351 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2352 shelving filter with a response similar to that of a standard
2353 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2355 The filter accepts the following options:
2359 Give the gain at 0 Hz. Its useful range is about -20
2360 (for a large cut) to +20 (for a large boost).
2361 Beware of clipping when using a positive gain.
2364 Set the filter's central frequency and so can be used
2365 to extend or reduce the frequency range to be boosted or cut.
2366 The default value is @code{100} Hz.
2369 Set method to specify band-width of filter.
2384 Determine how steep is the filter's shelf transition.
2387 Specify which channels to filter, by default all available are filtered.
2390 @subsection Commands
2392 This filter supports the following commands:
2395 Change bass frequency.
2396 Syntax for the command is : "@var{frequency}"
2399 Change bass width_type.
2400 Syntax for the command is : "@var{width_type}"
2404 Syntax for the command is : "@var{width}"
2408 Syntax for the command is : "@var{gain}"
2413 Apply a biquad IIR filter with the given coefficients.
2414 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2415 are the numerator and denominator coefficients respectively.
2416 and @var{channels}, @var{c} specify which channels to filter, by default all
2417 available are filtered.
2419 @subsection Commands
2421 This filter supports the following commands:
2429 Change biquad parameter.
2430 Syntax for the command is : "@var{value}"
2434 Bauer stereo to binaural transformation, which improves headphone listening of
2435 stereo audio records.
2437 To enable compilation of this filter you need to configure FFmpeg with
2438 @code{--enable-libbs2b}.
2440 It accepts the following parameters:
2444 Pre-defined crossfeed level.
2448 Default level (fcut=700, feed=50).
2451 Chu Moy circuit (fcut=700, feed=60).
2454 Jan Meier circuit (fcut=650, feed=95).
2459 Cut frequency (in Hz).
2468 Remap input channels to new locations.
2470 It accepts the following parameters:
2473 Map channels from input to output. The argument is a '|'-separated list of
2474 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2475 @var{in_channel} form. @var{in_channel} can be either the name of the input
2476 channel (e.g. FL for front left) or its index in the input channel layout.
2477 @var{out_channel} is the name of the output channel or its index in the output
2478 channel layout. If @var{out_channel} is not given then it is implicitly an
2479 index, starting with zero and increasing by one for each mapping.
2481 @item channel_layout
2482 The channel layout of the output stream.
2485 If no mapping is present, the filter will implicitly map input channels to
2486 output channels, preserving indices.
2488 @subsection Examples
2492 For example, assuming a 5.1+downmix input MOV file,
2494 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2496 will create an output WAV file tagged as stereo from the downmix channels of
2500 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2502 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2506 @section channelsplit
2508 Split each channel from an input audio stream into a separate output stream.
2510 It accepts the following parameters:
2512 @item channel_layout
2513 The channel layout of the input stream. The default is "stereo".
2515 A channel layout describing the channels to be extracted as separate output streams
2516 or "all" to extract each input channel as a separate stream. The default is "all".
2518 Choosing channels not present in channel layout in the input will result in an error.
2521 @subsection Examples
2525 For example, assuming a stereo input MP3 file,
2527 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2529 will create an output Matroska file with two audio streams, one containing only
2530 the left channel and the other the right channel.
2533 Split a 5.1 WAV file into per-channel files:
2535 ffmpeg -i in.wav -filter_complex
2536 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2537 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2538 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2543 Extract only LFE from a 5.1 WAV file:
2545 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2546 -map '[LFE]' lfe.wav
2551 Add a chorus effect to the audio.
2553 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2555 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2556 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2557 The modulation depth defines the range the modulated delay is played before or after
2558 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2559 sound tuned around the original one, like in a chorus where some vocals are slightly
2562 It accepts the following parameters:
2565 Set input gain. Default is 0.4.
2568 Set output gain. Default is 0.4.
2571 Set delays. A typical delay is around 40ms to 60ms.
2583 @subsection Examples
2589 chorus=0.7:0.9:55:0.4:0.25:2
2595 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2599 Fuller sounding chorus with three delays:
2601 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
2606 Compress or expand the audio's dynamic range.
2608 It accepts the following parameters:
2614 A list of times in seconds for each channel over which the instantaneous level
2615 of the input signal is averaged to determine its volume. @var{attacks} refers to
2616 increase of volume and @var{decays} refers to decrease of volume. For most
2617 situations, the attack time (response to the audio getting louder) should be
2618 shorter than the decay time, because the human ear is more sensitive to sudden
2619 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2620 a typical value for decay is 0.8 seconds.
2621 If specified number of attacks & decays is lower than number of channels, the last
2622 set attack/decay will be used for all remaining channels.
2625 A list of points for the transfer function, specified in dB relative to the
2626 maximum possible signal amplitude. Each key points list must be defined using
2627 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2628 @code{x0/y0 x1/y1 x2/y2 ....}
2630 The input values must be in strictly increasing order but the transfer function
2631 does not have to be monotonically rising. The point @code{0/0} is assumed but
2632 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2633 function are @code{-70/-70|-60/-20|1/0}.
2636 Set the curve radius in dB for all joints. It defaults to 0.01.
2639 Set the additional gain in dB to be applied at all points on the transfer
2640 function. This allows for easy adjustment of the overall gain.
2644 Set an initial volume, in dB, to be assumed for each channel when filtering
2645 starts. This permits the user to supply a nominal level initially, so that, for
2646 example, a very large gain is not applied to initial signal levels before the
2647 companding has begun to operate. A typical value for audio which is initially
2648 quiet is -90 dB. It defaults to 0.
2651 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2652 delayed before being fed to the volume adjuster. Specifying a delay
2653 approximately equal to the attack/decay times allows the filter to effectively
2654 operate in predictive rather than reactive mode. It defaults to 0.
2658 @subsection Examples
2662 Make music with both quiet and loud passages suitable for listening to in a
2665 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2668 Another example for audio with whisper and explosion parts:
2670 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2674 A noise gate for when the noise is at a lower level than the signal:
2676 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2680 Here is another noise gate, this time for when the noise is at a higher level
2681 than the signal (making it, in some ways, similar to squelch):
2683 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2687 2:1 compression starting at -6dB:
2689 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2693 2:1 compression starting at -9dB:
2695 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2699 2:1 compression starting at -12dB:
2701 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2705 2:1 compression starting at -18dB:
2707 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2711 3:1 compression starting at -15dB:
2713 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2719 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2725 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
2729 Hard limiter at -6dB:
2731 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2735 Hard limiter at -12dB:
2737 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2741 Hard noise gate at -35 dB:
2743 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2749 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2753 @section compensationdelay
2755 Compensation Delay Line is a metric based delay to compensate differing
2756 positions of microphones or speakers.
2758 For example, you have recorded guitar with two microphones placed in
2759 different location. Because the front of sound wave has fixed speed in
2760 normal conditions, the phasing of microphones can vary and depends on
2761 their location and interposition. The best sound mix can be achieved when
2762 these microphones are in phase (synchronized). Note that distance of
2763 ~30 cm between microphones makes one microphone to capture signal in
2764 antiphase to another microphone. That makes the final mix sounding moody.
2765 This filter helps to solve phasing problems by adding different delays
2766 to each microphone track and make them synchronized.
2768 The best result can be reached when you take one track as base and
2769 synchronize other tracks one by one with it.
2770 Remember that synchronization/delay tolerance depends on sample rate, too.
2771 Higher sample rates will give more tolerance.
2773 It accepts the following parameters:
2777 Set millimeters distance. This is compensation distance for fine tuning.
2781 Set cm distance. This is compensation distance for tightening distance setup.
2785 Set meters distance. This is compensation distance for hard distance setup.
2789 Set dry amount. Amount of unprocessed (dry) signal.
2793 Set wet amount. Amount of processed (wet) signal.
2797 Set temperature degree in Celsius. This is the temperature of the environment.
2802 Apply headphone crossfeed filter.
2804 Crossfeed is the process of blending the left and right channels of stereo
2806 It is mainly used to reduce extreme stereo separation of low frequencies.
2808 The intent is to produce more speaker like sound to the listener.
2810 The filter accepts the following options:
2814 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2815 This sets gain of low shelf filter for side part of stereo image.
2816 Default is -6dB. Max allowed is -30db when strength is set to 1.
2819 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2820 This sets cut off frequency of low shelf filter. Default is cut off near
2821 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2824 Set input gain. Default is 0.9.
2827 Set output gain. Default is 1.
2830 @section crystalizer
2831 Simple algorithm to expand audio dynamic range.
2833 The filter accepts the following options:
2837 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2838 (unchanged sound) to 10.0 (maximum effect).
2841 Enable clipping. By default is enabled.
2845 Apply a DC shift to the audio.
2847 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2848 in the recording chain) from the audio. The effect of a DC offset is reduced
2849 headroom and hence volume. The @ref{astats} filter can be used to determine if
2850 a signal has a DC offset.
2854 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2858 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2859 used to prevent clipping.
2863 Measure audio dynamic range.
2865 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2866 is found in transition material. And anything less that 8 have very poor dynamics
2867 and is very compressed.
2869 The filter accepts the following options:
2873 Set window length in seconds used to split audio into segments of equal length.
2874 Default is 3 seconds.
2878 Dynamic Audio Normalizer.
2880 This filter applies a certain amount of gain to the input audio in order
2881 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2882 contrast to more "simple" normalization algorithms, the Dynamic Audio
2883 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2884 This allows for applying extra gain to the "quiet" sections of the audio
2885 while avoiding distortions or clipping the "loud" sections. In other words:
2886 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2887 sections, in the sense that the volume of each section is brought to the
2888 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2889 this goal *without* applying "dynamic range compressing". It will retain 100%
2890 of the dynamic range *within* each section of the audio file.
2894 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2895 Default is 500 milliseconds.
2896 The Dynamic Audio Normalizer processes the input audio in small chunks,
2897 referred to as frames. This is required, because a peak magnitude has no
2898 meaning for just a single sample value. Instead, we need to determine the
2899 peak magnitude for a contiguous sequence of sample values. While a "standard"
2900 normalizer would simply use the peak magnitude of the complete file, the
2901 Dynamic Audio Normalizer determines the peak magnitude individually for each
2902 frame. The length of a frame is specified in milliseconds. By default, the
2903 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2904 been found to give good results with most files.
2905 Note that the exact frame length, in number of samples, will be determined
2906 automatically, based on the sampling rate of the individual input audio file.
2909 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2910 number. Default is 31.
2911 Probably the most important parameter of the Dynamic Audio Normalizer is the
2912 @code{window size} of the Gaussian smoothing filter. The filter's window size
2913 is specified in frames, centered around the current frame. For the sake of
2914 simplicity, this must be an odd number. Consequently, the default value of 31
2915 takes into account the current frame, as well as the 15 preceding frames and
2916 the 15 subsequent frames. Using a larger window results in a stronger
2917 smoothing effect and thus in less gain variation, i.e. slower gain
2918 adaptation. Conversely, using a smaller window results in a weaker smoothing
2919 effect and thus in more gain variation, i.e. faster gain adaptation.
2920 In other words, the more you increase this value, the more the Dynamic Audio
2921 Normalizer will behave like a "traditional" normalization filter. On the
2922 contrary, the more you decrease this value, the more the Dynamic Audio
2923 Normalizer will behave like a dynamic range compressor.
2926 Set the target peak value. This specifies the highest permissible magnitude
2927 level for the normalized audio input. This filter will try to approach the
2928 target peak magnitude as closely as possible, but at the same time it also
2929 makes sure that the normalized signal will never exceed the peak magnitude.
2930 A frame's maximum local gain factor is imposed directly by the target peak
2931 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2932 It is not recommended to go above this value.
2935 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2936 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2937 factor for each input frame, i.e. the maximum gain factor that does not
2938 result in clipping or distortion. The maximum gain factor is determined by
2939 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2940 additionally bounds the frame's maximum gain factor by a predetermined
2941 (global) maximum gain factor. This is done in order to avoid excessive gain
2942 factors in "silent" or almost silent frames. By default, the maximum gain
2943 factor is 10.0, For most inputs the default value should be sufficient and
2944 it usually is not recommended to increase this value. Though, for input
2945 with an extremely low overall volume level, it may be necessary to allow even
2946 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2947 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2948 Instead, a "sigmoid" threshold function will be applied. This way, the
2949 gain factors will smoothly approach the threshold value, but never exceed that
2953 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2954 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2955 This means that the maximum local gain factor for each frame is defined
2956 (only) by the frame's highest magnitude sample. This way, the samples can
2957 be amplified as much as possible without exceeding the maximum signal
2958 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2959 Normalizer can also take into account the frame's root mean square,
2960 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2961 determine the power of a time-varying signal. It is therefore considered
2962 that the RMS is a better approximation of the "perceived loudness" than
2963 just looking at the signal's peak magnitude. Consequently, by adjusting all
2964 frames to a constant RMS value, a uniform "perceived loudness" can be
2965 established. If a target RMS value has been specified, a frame's local gain
2966 factor is defined as the factor that would result in exactly that RMS value.
2967 Note, however, that the maximum local gain factor is still restricted by the
2968 frame's highest magnitude sample, in order to prevent clipping.
2971 Enable channels coupling. By default is enabled.
2972 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2973 amount. This means the same gain factor will be applied to all channels, i.e.
2974 the maximum possible gain factor is determined by the "loudest" channel.
2975 However, in some recordings, it may happen that the volume of the different
2976 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2977 In this case, this option can be used to disable the channel coupling. This way,
2978 the gain factor will be determined independently for each channel, depending
2979 only on the individual channel's highest magnitude sample. This allows for
2980 harmonizing the volume of the different channels.
2983 Enable DC bias correction. By default is disabled.
2984 An audio signal (in the time domain) is a sequence of sample values.
2985 In the Dynamic Audio Normalizer these sample values are represented in the
2986 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2987 audio signal, or "waveform", should be centered around the zero point.
2988 That means if we calculate the mean value of all samples in a file, or in a
2989 single frame, then the result should be 0.0 or at least very close to that
2990 value. If, however, there is a significant deviation of the mean value from
2991 0.0, in either positive or negative direction, this is referred to as a
2992 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2993 Audio Normalizer provides optional DC bias correction.
2994 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2995 the mean value, or "DC correction" offset, of each input frame and subtract
2996 that value from all of the frame's sample values which ensures those samples
2997 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2998 boundaries, the DC correction offset values will be interpolated smoothly
2999 between neighbouring frames.
3002 Enable alternative boundary mode. By default is disabled.
3003 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3004 around each frame. This includes the preceding frames as well as the
3005 subsequent frames. However, for the "boundary" frames, located at the very
3006 beginning and at the very end of the audio file, not all neighbouring
3007 frames are available. In particular, for the first few frames in the audio
3008 file, the preceding frames are not known. And, similarly, for the last few
3009 frames in the audio file, the subsequent frames are not known. Thus, the
3010 question arises which gain factors should be assumed for the missing frames
3011 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3012 to deal with this situation. The default boundary mode assumes a gain factor
3013 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3014 "fade out" at the beginning and at the end of the input, respectively.
3017 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3018 By default, the Dynamic Audio Normalizer does not apply "traditional"
3019 compression. This means that signal peaks will not be pruned and thus the
3020 full dynamic range will be retained within each local neighbourhood. However,
3021 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3022 normalization algorithm with a more "traditional" compression.
3023 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3024 (thresholding) function. If (and only if) the compression feature is enabled,
3025 all input frames will be processed by a soft knee thresholding function prior
3026 to the actual normalization process. Put simply, the thresholding function is
3027 going to prune all samples whose magnitude exceeds a certain threshold value.
3028 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3029 value. Instead, the threshold value will be adjusted for each individual
3031 In general, smaller parameters result in stronger compression, and vice versa.
3032 Values below 3.0 are not recommended, because audible distortion may appear.
3037 Make audio easier to listen to on headphones.
3039 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3040 so that when listened to on headphones the stereo image is moved from
3041 inside your head (standard for headphones) to outside and in front of
3042 the listener (standard for speakers).
3048 Apply a two-pole peaking equalisation (EQ) filter. With this
3049 filter, the signal-level at and around a selected frequency can
3050 be increased or decreased, whilst (unlike bandpass and bandreject
3051 filters) that at all other frequencies is unchanged.
3053 In order to produce complex equalisation curves, this filter can
3054 be given several times, each with a different central frequency.
3056 The filter accepts the following options:
3060 Set the filter's central frequency in Hz.
3063 Set method to specify band-width of filter.
3078 Specify the band-width of a filter in width_type units.
3081 Set the required gain or attenuation in dB.
3082 Beware of clipping when using a positive gain.
3085 Specify which channels to filter, by default all available are filtered.
3088 @subsection Examples
3091 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3093 equalizer=f=1000:t=h:width=200:g=-10
3097 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3099 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3103 @subsection Commands
3105 This filter supports the following commands:
3108 Change equalizer frequency.
3109 Syntax for the command is : "@var{frequency}"
3112 Change equalizer width_type.
3113 Syntax for the command is : "@var{width_type}"
3116 Change equalizer width.
3117 Syntax for the command is : "@var{width}"
3120 Change equalizer gain.
3121 Syntax for the command is : "@var{gain}"
3124 @section extrastereo
3126 Linearly increases the difference between left and right channels which
3127 adds some sort of "live" effect to playback.
3129 The filter accepts the following options:
3133 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3134 (average of both channels), with 1.0 sound will be unchanged, with
3135 -1.0 left and right channels will be swapped.
3138 Enable clipping. By default is enabled.
3141 @section firequalizer
3142 Apply FIR Equalization using arbitrary frequency response.
3144 The filter accepts the following option:
3148 Set gain curve equation (in dB). The expression can contain variables:
3151 the evaluated frequency
3155 channel number, set to 0 when multichannels evaluation is disabled
3157 channel id, see libavutil/channel_layout.h, set to the first channel id when
3158 multichannels evaluation is disabled
3162 channel_layout, see libavutil/channel_layout.h
3167 @item gain_interpolate(f)
3168 interpolate gain on frequency f based on gain_entry
3169 @item cubic_interpolate(f)
3170 same as gain_interpolate, but smoother
3172 This option is also available as command. Default is @code{gain_interpolate(f)}.
3175 Set gain entry for gain_interpolate function. The expression can
3179 store gain entry at frequency f with value g
3181 This option is also available as command.
3184 Set filter delay in seconds. Higher value means more accurate.
3185 Default is @code{0.01}.
3188 Set filter accuracy in Hz. Lower value means more accurate.
3189 Default is @code{5}.
3192 Set window function. Acceptable values are:
3195 rectangular window, useful when gain curve is already smooth
3197 hann window (default)
3203 3-terms continuous 1st derivative nuttall window
3205 minimum 3-terms discontinuous nuttall window
3207 4-terms continuous 1st derivative nuttall window
3209 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3211 blackman-harris window
3217 If enabled, use fixed number of audio samples. This improves speed when
3218 filtering with large delay. Default is disabled.
3221 Enable multichannels evaluation on gain. Default is disabled.
3224 Enable zero phase mode by subtracting timestamp to compensate delay.
3225 Default is disabled.
3228 Set scale used by gain. Acceptable values are:
3231 linear frequency, linear gain
3233 linear frequency, logarithmic (in dB) gain (default)
3235 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3237 logarithmic frequency, logarithmic gain
3241 Set file for dumping, suitable for gnuplot.
3244 Set scale for dumpfile. Acceptable values are same with scale option.
3248 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3249 Default is disabled.
3252 Enable minimum phase impulse response. Default is disabled.
3255 @subsection Examples
3260 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3263 lowpass at 1000 Hz with gain_entry:
3265 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3268 custom equalization:
3270 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3273 higher delay with zero phase to compensate delay:
3275 firequalizer=delay=0.1:fixed=on:zero_phase=on
3278 lowpass on left channel, highpass on right channel:
3280 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3281 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3286 Apply a flanging effect to the audio.
3288 The filter accepts the following options:
3292 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3295 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3298 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3302 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3303 Default value is 71.
3306 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3309 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3310 Default value is @var{sinusoidal}.
3313 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3314 Default value is 25.
3317 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3318 Default is @var{linear}.
3322 Apply Haas effect to audio.
3324 Note that this makes most sense to apply on mono signals.
3325 With this filter applied to mono signals it give some directionality and
3326 stretches its stereo image.
3328 The filter accepts the following options:
3332 Set input level. By default is @var{1}, or 0dB
3335 Set output level. By default is @var{1}, or 0dB.
3338 Set gain applied to side part of signal. By default is @var{1}.
3341 Set kind of middle source. Can be one of the following:
3351 Pick middle part signal of stereo image.
3354 Pick side part signal of stereo image.
3358 Change middle phase. By default is disabled.
3361 Set left channel delay. By default is @var{2.05} milliseconds.
3364 Set left channel balance. By default is @var{-1}.
3367 Set left channel gain. By default is @var{1}.
3370 Change left phase. By default is disabled.
3373 Set right channel delay. By defaults is @var{2.12} milliseconds.
3376 Set right channel balance. By default is @var{1}.
3379 Set right channel gain. By default is @var{1}.
3382 Change right phase. By default is enabled.
3387 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3388 embedded HDCD codes is expanded into a 20-bit PCM stream.
3390 The filter supports the Peak Extend and Low-level Gain Adjustment features
3391 of HDCD, and detects the Transient Filter flag.
3394 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3397 When using the filter with wav, note the default encoding for wav is 16-bit,
3398 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3399 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3401 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3402 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3405 The filter accepts the following options:
3408 @item disable_autoconvert
3409 Disable any automatic format conversion or resampling in the filter graph.
3411 @item process_stereo
3412 Process the stereo channels together. If target_gain does not match between
3413 channels, consider it invalid and use the last valid target_gain.
3416 Set the code detect timer period in ms.
3419 Always extend peaks above -3dBFS even if PE isn't signaled.
3422 Replace audio with a solid tone and adjust the amplitude to signal some
3423 specific aspect of the decoding process. The output file can be loaded in
3424 an audio editor alongside the original to aid analysis.
3426 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3433 Gain adjustment level at each sample
3435 Samples where peak extend occurs
3437 Samples where the code detect timer is active
3439 Samples where the target gain does not match between channels
3445 Apply head-related transfer functions (HRTFs) to create virtual
3446 loudspeakers around the user for binaural listening via headphones.
3447 The HRIRs are provided via additional streams, for each channel
3448 one stereo input stream is needed.
3450 The filter accepts the following options:
3454 Set mapping of input streams for convolution.
3455 The argument is a '|'-separated list of channel names in order as they
3456 are given as additional stream inputs for filter.
3457 This also specify number of input streams. Number of input streams
3458 must be not less than number of channels in first stream plus one.
3461 Set gain applied to audio. Value is in dB. Default is 0.
3464 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3465 processing audio in time domain which is slow.
3466 @var{freq} is processing audio in frequency domain which is fast.
3467 Default is @var{freq}.
3470 Set custom gain for LFE channels. Value is in dB. Default is 0.
3473 Set size of frame in number of samples which will be processed at once.
3474 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3477 Set format of hrir stream.
3478 Default value is @var{stereo}. Alternative value is @var{multich}.
3479 If value is set to @var{stereo}, number of additional streams should
3480 be greater or equal to number of input channels in first input stream.
3481 Also each additional stream should have stereo number of channels.
3482 If value is set to @var{multich}, number of additional streams should
3483 be exactly one. Also number of input channels of additional stream
3484 should be equal or greater than twice number of channels of first input
3488 @subsection Examples
3492 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3493 each amovie filter use stereo file with IR coefficients as input.
3494 The files give coefficients for each position of virtual loudspeaker:
3496 ffmpeg -i input.wav -lavfi-complex "amovie=azi_270_ele_0_DFC.wav[sr],amovie=azi_90_ele_0_DFC.wav[sl],amovie=azi_225_ele_0_DFC.wav[br],amovie=azi_135_ele_0_DFC.wav[bl],amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe],amovie=azi_35_ele_0_DFC.wav[fl],amovie=azi_325_ele_0_DFC.wav[fr],[a:0][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3501 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3502 but now in @var{multich} @var{hrir} format.
3504 ffmpeg -i input.wav -lavfi-complex "amovie=minp.wav[hrirs],[a:0][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3511 Apply a high-pass filter with 3dB point frequency.
3512 The filter can be either single-pole, or double-pole (the default).
3513 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3515 The filter accepts the following options:
3519 Set frequency in Hz. Default is 3000.
3522 Set number of poles. Default is 2.
3525 Set method to specify band-width of filter.
3540 Specify the band-width of a filter in width_type units.
3541 Applies only to double-pole filter.
3542 The default is 0.707q and gives a Butterworth response.
3545 Specify which channels to filter, by default all available are filtered.
3548 @subsection Commands
3550 This filter supports the following commands:
3553 Change highpass frequency.
3554 Syntax for the command is : "@var{frequency}"
3557 Change highpass width_type.
3558 Syntax for the command is : "@var{width_type}"
3561 Change highpass width.
3562 Syntax for the command is : "@var{width}"
3567 Join multiple input streams into one multi-channel stream.
3569 It accepts the following parameters:
3573 The number of input streams. It defaults to 2.
3575 @item channel_layout
3576 The desired output channel layout. It defaults to stereo.
3579 Map channels from inputs to output. The argument is a '|'-separated list of
3580 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3581 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3582 can be either the name of the input channel (e.g. FL for front left) or its
3583 index in the specified input stream. @var{out_channel} is the name of the output
3587 The filter will attempt to guess the mappings when they are not specified
3588 explicitly. It does so by first trying to find an unused matching input channel
3589 and if that fails it picks the first unused input channel.
3591 Join 3 inputs (with properly set channel layouts):
3593 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3596 Build a 5.1 output from 6 single-channel streams:
3598 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3599 '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'
3605 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3607 To enable compilation of this filter you need to configure FFmpeg with
3608 @code{--enable-ladspa}.
3612 Specifies the name of LADSPA plugin library to load. If the environment
3613 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3614 each one of the directories specified by the colon separated list in
3615 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3616 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3617 @file{/usr/lib/ladspa/}.
3620 Specifies the plugin within the library. Some libraries contain only
3621 one plugin, but others contain many of them. If this is not set filter
3622 will list all available plugins within the specified library.
3625 Set the '|' separated list of controls which are zero or more floating point
3626 values that determine the behavior of the loaded plugin (for example delay,
3628 Controls need to be defined using the following syntax:
3629 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3630 @var{valuei} is the value set on the @var{i}-th control.
3631 Alternatively they can be also defined using the following syntax:
3632 @var{value0}|@var{value1}|@var{value2}|..., where
3633 @var{valuei} is the value set on the @var{i}-th control.
3634 If @option{controls} is set to @code{help}, all available controls and
3635 their valid ranges are printed.
3637 @item sample_rate, s
3638 Specify the sample rate, default to 44100. Only used if plugin have
3642 Set the number of samples per channel per each output frame, default
3643 is 1024. Only used if plugin have zero inputs.
3646 Set the minimum duration of the sourced audio. See
3647 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3648 for the accepted syntax.
3649 Note that the resulting duration may be greater than the specified duration,
3650 as the generated audio is always cut at the end of a complete frame.
3651 If not specified, or the expressed duration is negative, the audio is
3652 supposed to be generated forever.
3653 Only used if plugin have zero inputs.
3657 @subsection Examples
3661 List all available plugins within amp (LADSPA example plugin) library:
3667 List all available controls and their valid ranges for @code{vcf_notch}
3668 plugin from @code{VCF} library:
3670 ladspa=f=vcf:p=vcf_notch:c=help
3674 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3677 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3681 Add reverberation to the audio using TAP-plugins
3682 (Tom's Audio Processing plugins):
3684 ladspa=file=tap_reverb:tap_reverb
3688 Generate white noise, with 0.2 amplitude:
3690 ladspa=file=cmt:noise_source_white:c=c0=.2
3694 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3695 @code{C* Audio Plugin Suite} (CAPS) library:
3697 ladspa=file=caps:Click:c=c1=20'
3701 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3703 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3707 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3708 @code{SWH Plugins} collection:
3710 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3714 Attenuate low frequencies using Multiband EQ from Steve Harris
3715 @code{SWH Plugins} collection:
3717 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3721 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3724 ladspa=caps:Narrower
3728 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3730 ladspa=caps:White:.2
3734 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3736 ladspa=caps:Fractal:c=c1=1
3740 Dynamic volume normalization using @code{VLevel} plugin:
3742 ladspa=vlevel-ladspa:vlevel_mono
3746 @subsection Commands
3748 This filter supports the following commands:
3751 Modify the @var{N}-th control value.
3753 If the specified value is not valid, it is ignored and prior one is kept.
3758 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3759 Support for both single pass (livestreams, files) and double pass (files) modes.
3760 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3761 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3762 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3764 The filter accepts the following options:
3768 Set integrated loudness target.
3769 Range is -70.0 - -5.0. Default value is -24.0.
3772 Set loudness range target.
3773 Range is 1.0 - 20.0. Default value is 7.0.
3776 Set maximum true peak.
3777 Range is -9.0 - +0.0. Default value is -2.0.
3779 @item measured_I, measured_i
3780 Measured IL of input file.
3781 Range is -99.0 - +0.0.
3783 @item measured_LRA, measured_lra
3784 Measured LRA of input file.
3785 Range is 0.0 - 99.0.
3787 @item measured_TP, measured_tp
3788 Measured true peak of input file.
3789 Range is -99.0 - +99.0.
3791 @item measured_thresh
3792 Measured threshold of input file.
3793 Range is -99.0 - +0.0.
3796 Set offset gain. Gain is applied before the true-peak limiter.
3797 Range is -99.0 - +99.0. Default is +0.0.
3800 Normalize linearly if possible.
3801 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3802 to be specified in order to use this mode.
3803 Options are true or false. Default is true.
3806 Treat mono input files as "dual-mono". If a mono file is intended for playback
3807 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3808 If set to @code{true}, this option will compensate for this effect.
3809 Multi-channel input files are not affected by this option.
3810 Options are true or false. Default is false.
3813 Set print format for stats. Options are summary, json, or none.
3814 Default value is none.
3819 Apply a low-pass filter with 3dB point frequency.
3820 The filter can be either single-pole or double-pole (the default).
3821 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3823 The filter accepts the following options:
3827 Set frequency in Hz. Default is 500.
3830 Set number of poles. Default is 2.
3833 Set method to specify band-width of filter.
3848 Specify the band-width of a filter in width_type units.
3849 Applies only to double-pole filter.
3850 The default is 0.707q and gives a Butterworth response.
3853 Specify which channels to filter, by default all available are filtered.
3856 @subsection Examples
3859 Lowpass only LFE channel, it LFE is not present it does nothing:
3865 @subsection Commands
3867 This filter supports the following commands:
3870 Change lowpass frequency.
3871 Syntax for the command is : "@var{frequency}"
3874 Change lowpass width_type.
3875 Syntax for the command is : "@var{width_type}"
3878 Change lowpass width.
3879 Syntax for the command is : "@var{width}"
3884 Load a LV2 (LADSPA Version 2) plugin.
3886 To enable compilation of this filter you need to configure FFmpeg with
3887 @code{--enable-lv2}.
3891 Specifies the plugin URI. You may need to escape ':'.
3894 Set the '|' separated list of controls which are zero or more floating point
3895 values that determine the behavior of the loaded plugin (for example delay,
3897 If @option{controls} is set to @code{help}, all available controls and
3898 their valid ranges are printed.
3900 @item sample_rate, s
3901 Specify the sample rate, default to 44100. Only used if plugin have
3905 Set the number of samples per channel per each output frame, default
3906 is 1024. Only used if plugin have zero inputs.
3909 Set the minimum duration of the sourced audio. See
3910 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3911 for the accepted syntax.
3912 Note that the resulting duration may be greater than the specified duration,
3913 as the generated audio is always cut at the end of a complete frame.
3914 If not specified, or the expressed duration is negative, the audio is
3915 supposed to be generated forever.
3916 Only used if plugin have zero inputs.
3919 @subsection Examples
3923 Apply bass enhancer plugin from Calf:
3925 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3929 Apply vinyl plugin from Calf:
3931 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3935 Apply bit crusher plugin from ArtyFX:
3937 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3942 Multiband Compress or expand the audio's dynamic range.
3944 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3945 This is akin to the crossover of a loudspeaker, and results in flat frequency
3946 response when absent compander action.
3948 It accepts the following parameters:
3952 This option syntax is:
3953 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3954 For explanation of each item refer to compand filter documentation.
3960 Mix channels with specific gain levels. The filter accepts the output
3961 channel layout followed by a set of channels definitions.
3963 This filter is also designed to efficiently remap the channels of an audio
3966 The filter accepts parameters of the form:
3967 "@var{l}|@var{outdef}|@var{outdef}|..."
3971 output channel layout or number of channels
3974 output channel specification, of the form:
3975 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3978 output channel to define, either a channel name (FL, FR, etc.) or a channel
3979 number (c0, c1, etc.)
3982 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3985 input channel to use, see out_name for details; it is not possible to mix
3986 named and numbered input channels
3989 If the `=' in a channel specification is replaced by `<', then the gains for
3990 that specification will be renormalized so that the total is 1, thus
3991 avoiding clipping noise.
3993 @subsection Mixing examples
3995 For example, if you want to down-mix from stereo to mono, but with a bigger
3996 factor for the left channel:
3998 pan=1c|c0=0.9*c0+0.1*c1
4001 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4002 7-channels surround:
4004 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4007 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4008 that should be preferred (see "-ac" option) unless you have very specific
4011 @subsection Remapping examples
4013 The channel remapping will be effective if, and only if:
4016 @item gain coefficients are zeroes or ones,
4017 @item only one input per channel output,
4020 If all these conditions are satisfied, the filter will notify the user ("Pure
4021 channel mapping detected"), and use an optimized and lossless method to do the
4024 For example, if you have a 5.1 source and want a stereo audio stream by
4025 dropping the extra channels:
4027 pan="stereo| c0=FL | c1=FR"
4030 Given the same source, you can also switch front left and front right channels
4031 and keep the input channel layout:
4033 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4036 If the input is a stereo audio stream, you can mute the front left channel (and
4037 still keep the stereo channel layout) with:
4042 Still with a stereo audio stream input, you can copy the right channel in both
4043 front left and right:
4045 pan="stereo| c0=FR | c1=FR"
4050 ReplayGain scanner filter. This filter takes an audio stream as an input and
4051 outputs it unchanged.
4052 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4056 Convert the audio sample format, sample rate and channel layout. It is
4057 not meant to be used directly.
4060 Apply time-stretching and pitch-shifting with librubberband.
4062 To enable compilation of this filter, you need to configure FFmpeg with
4063 @code{--enable-librubberband}.
4065 The filter accepts the following options:
4069 Set tempo scale factor.
4072 Set pitch scale factor.
4075 Set transients detector.
4076 Possible values are:
4085 Possible values are:
4094 Possible values are:
4101 Set processing window size.
4102 Possible values are:
4111 Possible values are:
4118 Enable formant preservation when shift pitching.
4119 Possible values are:
4127 Possible values are:
4136 Possible values are:
4143 @section sidechaincompress
4145 This filter acts like normal compressor but has the ability to compress
4146 detected signal using second input signal.
4147 It needs two input streams and returns one output stream.
4148 First input stream will be processed depending on second stream signal.
4149 The filtered signal then can be filtered with other filters in later stages of
4150 processing. See @ref{pan} and @ref{amerge} filter.
4152 The filter accepts the following options:
4156 Set input gain. Default is 1. Range is between 0.015625 and 64.
4159 If a signal of second stream raises above this level it will affect the gain
4160 reduction of first stream.
4161 By default is 0.125. Range is between 0.00097563 and 1.
4164 Set a ratio about which the signal is reduced. 1:2 means that if the level
4165 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4166 Default is 2. Range is between 1 and 20.
4169 Amount of milliseconds the signal has to rise above the threshold before gain
4170 reduction starts. Default is 20. Range is between 0.01 and 2000.
4173 Amount of milliseconds the signal has to fall below the threshold before
4174 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4177 Set the amount by how much signal will be amplified after processing.
4178 Default is 1. Range is from 1 to 64.
4181 Curve the sharp knee around the threshold to enter gain reduction more softly.
4182 Default is 2.82843. Range is between 1 and 8.
4185 Choose if the @code{average} level between all channels of side-chain stream
4186 or the louder(@code{maximum}) channel of side-chain stream affects the
4187 reduction. Default is @code{average}.
4190 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4191 of @code{rms}. Default is @code{rms} which is mainly smoother.
4194 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4197 How much to use compressed signal in output. Default is 1.
4198 Range is between 0 and 1.
4201 @subsection Examples
4205 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4206 depending on the signal of 2nd input and later compressed signal to be
4207 merged with 2nd input:
4209 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4213 @section sidechaingate
4215 A sidechain gate acts like a normal (wideband) gate but has the ability to
4216 filter the detected signal before sending it to the gain reduction stage.
4217 Normally a gate uses the full range signal to detect a level above the
4219 For example: If you cut all lower frequencies from your sidechain signal
4220 the gate will decrease the volume of your track only if not enough highs
4221 appear. With this technique you are able to reduce the resonation of a
4222 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4224 It needs two input streams and returns one output stream.
4225 First input stream will be processed depending on second stream signal.
4227 The filter accepts the following options:
4231 Set input level before filtering.
4232 Default is 1. Allowed range is from 0.015625 to 64.
4235 Set the level of gain reduction when the signal is below the threshold.
4236 Default is 0.06125. Allowed range is from 0 to 1.
4239 If a signal rises above this level the gain reduction is released.
4240 Default is 0.125. Allowed range is from 0 to 1.
4243 Set a ratio about which the signal is reduced.
4244 Default is 2. Allowed range is from 1 to 9000.
4247 Amount of milliseconds the signal has to rise above the threshold before gain
4249 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4252 Amount of milliseconds the signal has to fall below the threshold before the
4253 reduction is increased again. Default is 250 milliseconds.
4254 Allowed range is from 0.01 to 9000.
4257 Set amount of amplification of signal after processing.
4258 Default is 1. Allowed range is from 1 to 64.
4261 Curve the sharp knee around the threshold to enter gain reduction more softly.
4262 Default is 2.828427125. Allowed range is from 1 to 8.
4265 Choose if exact signal should be taken for detection or an RMS like one.
4266 Default is rms. Can be peak or rms.
4269 Choose if the average level between all channels or the louder channel affects
4271 Default is average. Can be average or maximum.
4274 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4277 @section silencedetect
4279 Detect silence in an audio stream.
4281 This filter logs a message when it detects that the input audio volume is less
4282 or equal to a noise tolerance value for a duration greater or equal to the
4283 minimum detected noise duration.
4285 The printed times and duration are expressed in seconds.
4287 The filter accepts the following options:
4291 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4292 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4295 Set silence duration until notification (default is 2 seconds).
4298 Process each channel separately, instead of combined. By default is disabled.
4301 @subsection Examples
4305 Detect 5 seconds of silence with -50dB noise tolerance:
4307 silencedetect=n=-50dB:d=5
4311 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4312 tolerance in @file{silence.mp3}:
4314 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4318 @section silenceremove
4320 Remove silence from the beginning, middle or end of the audio.
4322 The filter accepts the following options:
4326 This value is used to indicate if audio should be trimmed at beginning of
4327 the audio. A value of zero indicates no silence should be trimmed from the
4328 beginning. When specifying a non-zero value, it trims audio up until it
4329 finds non-silence. Normally, when trimming silence from beginning of audio
4330 the @var{start_periods} will be @code{1} but it can be increased to higher
4331 values to trim all audio up to specific count of non-silence periods.
4332 Default value is @code{0}.
4334 @item start_duration
4335 Specify the amount of time that non-silence must be detected before it stops
4336 trimming audio. By increasing the duration, bursts of noises can be treated
4337 as silence and trimmed off. Default value is @code{0}.
4339 @item start_threshold
4340 This indicates what sample value should be treated as silence. For digital
4341 audio, a value of @code{0} may be fine but for audio recorded from analog,
4342 you may wish to increase the value to account for background noise.
4343 Can be specified in dB (in case "dB" is appended to the specified value)
4344 or amplitude ratio. Default value is @code{0}.
4347 Set the count for trimming silence from the end of audio.
4348 To remove silence from the middle of a file, specify a @var{stop_periods}
4349 that is negative. This value is then treated as a positive value and is
4350 used to indicate the effect should restart processing as specified by
4351 @var{start_periods}, making it suitable for removing periods of silence
4352 in the middle of the audio.
4353 Default value is @code{0}.
4356 Specify a duration of silence that must exist before audio is not copied any
4357 more. By specifying a higher duration, silence that is wanted can be left in
4359 Default value is @code{0}.
4361 @item stop_threshold
4362 This is the same as @option{start_threshold} but for trimming silence from
4364 Can be specified in dB (in case "dB" is appended to the specified value)
4365 or amplitude ratio. Default value is @code{0}.
4368 This indicates that @var{stop_duration} length of audio should be left intact
4369 at the beginning of each period of silence.
4370 For example, if you want to remove long pauses between words but do not want
4371 to remove the pauses completely. Default value is @code{0}.
4374 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4375 and works better with digital silence which is exactly 0.
4376 Default value is @code{rms}.
4379 Set ratio used to calculate size of window for detecting silence.
4380 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4383 @subsection Examples
4387 The following example shows how this filter can be used to start a recording
4388 that does not contain the delay at the start which usually occurs between
4389 pressing the record button and the start of the performance:
4391 silenceremove=1:5:0.02
4395 Trim all silence encountered from beginning to end where there is more than 1
4396 second of silence in audio:
4398 silenceremove=0:0:0:-1:1:-90dB
4404 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4405 loudspeakers around the user for binaural listening via headphones (audio
4406 formats up to 9 channels supported).
4407 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4408 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4409 Austrian Academy of Sciences.
4411 To enable compilation of this filter you need to configure FFmpeg with
4412 @code{--enable-libmysofa}.
4414 The filter accepts the following options:
4418 Set the SOFA file used for rendering.
4421 Set gain applied to audio. Value is in dB. Default is 0.
4424 Set rotation of virtual loudspeakers in deg. Default is 0.
4427 Set elevation of virtual speakers in deg. Default is 0.
4430 Set distance in meters between loudspeakers and the listener with near-field
4431 HRTFs. Default is 1.
4434 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4435 processing audio in time domain which is slow.
4436 @var{freq} is processing audio in frequency domain which is fast.
4437 Default is @var{freq}.
4440 Set custom positions of virtual loudspeakers. Syntax for this option is:
4441 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4442 Each virtual loudspeaker is described with short channel name following with
4443 azimuth and elevation in degrees.
4444 Each virtual loudspeaker description is separated by '|'.
4445 For example to override front left and front right channel positions use:
4446 'speakers=FL 45 15|FR 345 15'.
4447 Descriptions with unrecognised channel names are ignored.
4450 Set custom gain for LFE channels. Value is in dB. Default is 0.
4453 @subsection Examples
4457 Using ClubFritz6 sofa file:
4459 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4463 Using ClubFritz12 sofa file and bigger radius with small rotation:
4465 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4469 Similar as above but with custom speaker positions for front left, front right, back left and back right
4470 and also with custom gain:
4472 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4476 @section stereotools
4478 This filter has some handy utilities to manage stereo signals, for converting
4479 M/S stereo recordings to L/R signal while having control over the parameters
4480 or spreading the stereo image of master track.
4482 The filter accepts the following options:
4486 Set input level before filtering for both channels. Defaults is 1.
4487 Allowed range is from 0.015625 to 64.
4490 Set output level after filtering for both channels. Defaults is 1.
4491 Allowed range is from 0.015625 to 64.
4494 Set input balance between both channels. Default is 0.
4495 Allowed range is from -1 to 1.
4498 Set output balance between both channels. Default is 0.
4499 Allowed range is from -1 to 1.
4502 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4503 clipping. Disabled by default.
4506 Mute the left channel. Disabled by default.
4509 Mute the right channel. Disabled by default.
4512 Change the phase of the left channel. Disabled by default.
4515 Change the phase of the right channel. Disabled by default.
4518 Set stereo mode. Available values are:
4522 Left/Right to Left/Right, this is default.
4525 Left/Right to Mid/Side.
4528 Mid/Side to Left/Right.
4531 Left/Right to Left/Left.
4534 Left/Right to Right/Right.
4537 Left/Right to Left + Right.
4540 Left/Right to Right/Left.
4543 Mid/Side to Left/Left.
4546 Mid/Side to Right/Right.
4550 Set level of side signal. Default is 1.
4551 Allowed range is from 0.015625 to 64.
4554 Set balance of side signal. Default is 0.
4555 Allowed range is from -1 to 1.
4558 Set level of the middle signal. Default is 1.
4559 Allowed range is from 0.015625 to 64.
4562 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4565 Set stereo base between mono and inversed channels. Default is 0.
4566 Allowed range is from -1 to 1.
4569 Set delay in milliseconds how much to delay left from right channel and
4570 vice versa. Default is 0. Allowed range is from -20 to 20.
4573 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4576 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4578 @item bmode_in, bmode_out
4579 Set balance mode for balance_in/balance_out option.
4581 Can be one of the following:
4585 Classic balance mode. Attenuate one channel at time.
4586 Gain is raised up to 1.
4589 Similar as classic mode above but gain is raised up to 2.
4592 Equal power distribution, from -6dB to +6dB range.
4596 @subsection Examples
4600 Apply karaoke like effect:
4602 stereotools=mlev=0.015625
4606 Convert M/S signal to L/R:
4608 "stereotools=mode=ms>lr"
4612 @section stereowiden
4614 This filter enhance the stereo effect by suppressing signal common to both
4615 channels and by delaying the signal of left into right and vice versa,
4616 thereby widening the stereo effect.
4618 The filter accepts the following options:
4622 Time in milliseconds of the delay of left signal into right and vice versa.
4623 Default is 20 milliseconds.
4626 Amount of gain in delayed signal into right and vice versa. Gives a delay
4627 effect of left signal in right output and vice versa which gives widening
4628 effect. Default is 0.3.
4631 Cross feed of left into right with inverted phase. This helps in suppressing
4632 the mono. If the value is 1 it will cancel all the signal common to both
4633 channels. Default is 0.3.
4636 Set level of input signal of original channel. Default is 0.8.
4639 @section superequalizer
4640 Apply 18 band equalizer.
4642 The filter accepts the following options:
4649 Set 131Hz band gain.
4651 Set 185Hz band gain.
4653 Set 262Hz band gain.
4655 Set 370Hz band gain.
4657 Set 523Hz band gain.
4659 Set 740Hz band gain.
4661 Set 1047Hz band gain.
4663 Set 1480Hz band gain.
4665 Set 2093Hz band gain.
4667 Set 2960Hz band gain.
4669 Set 4186Hz band gain.
4671 Set 5920Hz band gain.
4673 Set 8372Hz band gain.
4675 Set 11840Hz band gain.
4677 Set 16744Hz band gain.
4679 Set 20000Hz band gain.
4683 Apply audio surround upmix filter.
4685 This filter allows to produce multichannel output from audio stream.
4687 The filter accepts the following options:
4691 Set output channel layout. By default, this is @var{5.1}.
4693 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4694 for the required syntax.
4697 Set input channel layout. By default, this is @var{stereo}.
4699 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4700 for the required syntax.
4703 Set input volume level. By default, this is @var{1}.
4706 Set output volume level. By default, this is @var{1}.
4709 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4712 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4715 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4718 Set front center input volume. By default, this is @var{1}.
4721 Set front center output volume. By default, this is @var{1}.
4724 Set LFE input volume. By default, this is @var{1}.
4727 Set LFE output volume. By default, this is @var{1}.
4730 @section treble, highshelf
4732 Boost or cut treble (upper) frequencies of the audio using a two-pole
4733 shelving filter with a response similar to that of a standard
4734 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4736 The filter accepts the following options:
4740 Give the gain at whichever is the lower of ~22 kHz and the
4741 Nyquist frequency. Its useful range is about -20 (for a large cut)
4742 to +20 (for a large boost). Beware of clipping when using a positive gain.
4745 Set the filter's central frequency and so can be used
4746 to extend or reduce the frequency range to be boosted or cut.
4747 The default value is @code{3000} Hz.
4750 Set method to specify band-width of filter.
4765 Determine how steep is the filter's shelf transition.
4768 Specify which channels to filter, by default all available are filtered.
4771 @subsection Commands
4773 This filter supports the following commands:
4776 Change treble frequency.
4777 Syntax for the command is : "@var{frequency}"
4780 Change treble width_type.
4781 Syntax for the command is : "@var{width_type}"
4784 Change treble width.
4785 Syntax for the command is : "@var{width}"
4789 Syntax for the command is : "@var{gain}"
4794 Sinusoidal amplitude modulation.
4796 The filter accepts the following options:
4800 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4801 (20 Hz or lower) will result in a tremolo effect.
4802 This filter may also be used as a ring modulator by specifying
4803 a modulation frequency higher than 20 Hz.
4804 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4807 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4808 Default value is 0.5.
4813 Sinusoidal phase modulation.
4815 The filter accepts the following options:
4819 Modulation frequency in Hertz.
4820 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4823 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4824 Default value is 0.5.
4829 Adjust the input audio volume.
4831 It accepts the following parameters:
4835 Set audio volume expression.
4837 Output values are clipped to the maximum value.
4839 The output audio volume is given by the relation:
4841 @var{output_volume} = @var{volume} * @var{input_volume}
4844 The default value for @var{volume} is "1.0".
4847 This parameter represents the mathematical precision.
4849 It determines which input sample formats will be allowed, which affects the
4850 precision of the volume scaling.
4854 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4856 32-bit floating-point; this limits input sample format to FLT. (default)
4858 64-bit floating-point; this limits input sample format to DBL.
4862 Choose the behaviour on encountering ReplayGain side data in input frames.
4866 Remove ReplayGain side data, ignoring its contents (the default).
4869 Ignore ReplayGain side data, but leave it in the frame.
4872 Prefer the track gain, if present.
4875 Prefer the album gain, if present.
4878 @item replaygain_preamp
4879 Pre-amplification gain in dB to apply to the selected replaygain gain.
4881 Default value for @var{replaygain_preamp} is 0.0.
4884 Set when the volume expression is evaluated.
4886 It accepts the following values:
4889 only evaluate expression once during the filter initialization, or
4890 when the @samp{volume} command is sent
4893 evaluate expression for each incoming frame
4896 Default value is @samp{once}.
4899 The volume expression can contain the following parameters.
4903 frame number (starting at zero)
4906 @item nb_consumed_samples
4907 number of samples consumed by the filter
4909 number of samples in the current frame
4911 original frame position in the file
4917 PTS at start of stream
4919 time at start of stream
4925 last set volume value
4928 Note that when @option{eval} is set to @samp{once} only the
4929 @var{sample_rate} and @var{tb} variables are available, all other
4930 variables will evaluate to NAN.
4932 @subsection Commands
4934 This filter supports the following commands:
4937 Modify the volume expression.
4938 The command accepts the same syntax of the corresponding option.
4940 If the specified expression is not valid, it is kept at its current
4942 @item replaygain_noclip
4943 Prevent clipping by limiting the gain applied.
4945 Default value for @var{replaygain_noclip} is 1.
4949 @subsection Examples
4953 Halve the input audio volume:
4957 volume=volume=-6.0206dB
4960 In all the above example the named key for @option{volume} can be
4961 omitted, for example like in:
4967 Increase input audio power by 6 decibels using fixed-point precision:
4969 volume=volume=6dB:precision=fixed
4973 Fade volume after time 10 with an annihilation period of 5 seconds:
4975 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4979 @section volumedetect
4981 Detect the volume of the input video.
4983 The filter has no parameters. The input is not modified. Statistics about
4984 the volume will be printed in the log when the input stream end is reached.
4986 In particular it will show the mean volume (root mean square), maximum
4987 volume (on a per-sample basis), and the beginning of a histogram of the
4988 registered volume values (from the maximum value to a cumulated 1/1000 of
4991 All volumes are in decibels relative to the maximum PCM value.
4993 @subsection Examples
4995 Here is an excerpt of the output:
4997 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4998 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4999 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5000 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5001 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5002 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5003 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5004 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5005 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5011 The mean square energy is approximately -27 dB, or 10^-2.7.
5013 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5015 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5018 In other words, raising the volume by +4 dB does not cause any clipping,
5019 raising it by +5 dB causes clipping for 6 samples, etc.
5021 @c man end AUDIO FILTERS
5023 @chapter Audio Sources
5024 @c man begin AUDIO SOURCES
5026 Below is a description of the currently available audio sources.
5030 Buffer audio frames, and make them available to the filter chain.
5032 This source is mainly intended for a programmatic use, in particular
5033 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5035 It accepts the following parameters:
5039 The timebase which will be used for timestamps of submitted frames. It must be
5040 either a floating-point number or in @var{numerator}/@var{denominator} form.
5043 The sample rate of the incoming audio buffers.
5046 The sample format of the incoming audio buffers.
5047 Either a sample format name or its corresponding integer representation from
5048 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5050 @item channel_layout
5051 The channel layout of the incoming audio buffers.
5052 Either a channel layout name from channel_layout_map in
5053 @file{libavutil/channel_layout.c} or its corresponding integer representation
5054 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5057 The number of channels of the incoming audio buffers.
5058 If both @var{channels} and @var{channel_layout} are specified, then they
5063 @subsection Examples
5066 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5069 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5070 Since the sample format with name "s16p" corresponds to the number
5071 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5074 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5079 Generate an audio signal specified by an expression.
5081 This source accepts in input one or more expressions (one for each
5082 channel), which are evaluated and used to generate a corresponding
5085 This source accepts the following options:
5089 Set the '|'-separated expressions list for each separate channel. In case the
5090 @option{channel_layout} option is not specified, the selected channel layout
5091 depends on the number of provided expressions. Otherwise the last
5092 specified expression is applied to the remaining output channels.
5094 @item channel_layout, c
5095 Set the channel layout. The number of channels in the specified layout
5096 must be equal to the number of specified expressions.
5099 Set the minimum duration of the sourced audio. See
5100 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5101 for the accepted syntax.
5102 Note that the resulting duration may be greater than the specified
5103 duration, as the generated audio is always cut at the end of a
5106 If not specified, or the expressed duration is negative, the audio is
5107 supposed to be generated forever.
5110 Set the number of samples per channel per each output frame,
5113 @item sample_rate, s
5114 Specify the sample rate, default to 44100.
5117 Each expression in @var{exprs} can contain the following constants:
5121 number of the evaluated sample, starting from 0
5124 time of the evaluated sample expressed in seconds, starting from 0
5131 @subsection Examples
5141 Generate a sin signal with frequency of 440 Hz, set sample rate to
5144 aevalsrc="sin(440*2*PI*t):s=8000"
5148 Generate a two channels signal, specify the channel layout (Front
5149 Center + Back Center) explicitly:
5151 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5155 Generate white noise:
5157 aevalsrc="-2+random(0)"
5161 Generate an amplitude modulated signal:
5163 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5167 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5169 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5176 The null audio source, return unprocessed audio frames. It is mainly useful
5177 as a template and to be employed in analysis / debugging tools, or as
5178 the source for filters which ignore the input data (for example the sox
5181 This source accepts the following options:
5185 @item channel_layout, cl
5187 Specifies the channel layout, and can be either an integer or a string
5188 representing a channel layout. The default value of @var{channel_layout}
5191 Check the channel_layout_map definition in
5192 @file{libavutil/channel_layout.c} for the mapping between strings and
5193 channel layout values.
5195 @item sample_rate, r
5196 Specifies the sample rate, and defaults to 44100.
5199 Set the number of samples per requested frames.
5203 @subsection Examples
5207 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5209 anullsrc=r=48000:cl=4
5213 Do the same operation with a more obvious syntax:
5215 anullsrc=r=48000:cl=mono
5219 All the parameters need to be explicitly defined.
5223 Synthesize a voice utterance using the libflite library.
5225 To enable compilation of this filter you need to configure FFmpeg with
5226 @code{--enable-libflite}.
5228 Note that versions of the flite library prior to 2.0 are not thread-safe.
5230 The filter accepts the following options:
5235 If set to 1, list the names of the available voices and exit
5236 immediately. Default value is 0.
5239 Set the maximum number of samples per frame. Default value is 512.
5242 Set the filename containing the text to speak.
5245 Set the text to speak.
5248 Set the voice to use for the speech synthesis. Default value is
5249 @code{kal}. See also the @var{list_voices} option.
5252 @subsection Examples
5256 Read from file @file{speech.txt}, and synthesize the text using the
5257 standard flite voice:
5259 flite=textfile=speech.txt
5263 Read the specified text selecting the @code{slt} voice:
5265 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5269 Input text to ffmpeg:
5271 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5275 Make @file{ffplay} speak the specified text, using @code{flite} and
5276 the @code{lavfi} device:
5278 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5282 For more information about libflite, check:
5283 @url{http://www.festvox.org/flite/}
5287 Generate a noise audio signal.
5289 The filter accepts the following options:
5292 @item sample_rate, r
5293 Specify the sample rate. Default value is 48000 Hz.
5296 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5300 Specify the duration of the generated audio stream. Not specifying this option
5301 results in noise with an infinite length.
5303 @item color, colour, c
5304 Specify the color of noise. Available noise colors are white, pink, brown,
5305 blue and violet. Default color is white.
5308 Specify a value used to seed the PRNG.
5311 Set the number of samples per each output frame, default is 1024.
5314 @subsection Examples
5319 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5321 anoisesrc=d=60:c=pink:r=44100:a=0.5
5327 Generate odd-tap Hilbert transform FIR coefficients.
5329 The resulting stream can be used with @ref{afir} filter for phase-shifting
5330 the signal by 90 degrees.
5332 This is used in many matrix coding schemes and for analytic signal generation.
5333 The process is often written as a multiplication by i (or j), the imaginary unit.
5335 The filter accepts the following options:
5339 @item sample_rate, s
5340 Set sample rate, default is 44100.
5343 Set length of FIR filter, default is 22051.
5346 Set number of samples per each frame.
5349 Set window function to be used when generating FIR coefficients.
5354 Generate an audio signal made of a sine wave with amplitude 1/8.
5356 The audio signal is bit-exact.
5358 The filter accepts the following options:
5363 Set the carrier frequency. Default is 440 Hz.
5365 @item beep_factor, b
5366 Enable a periodic beep every second with frequency @var{beep_factor} times
5367 the carrier frequency. Default is 0, meaning the beep is disabled.
5369 @item sample_rate, r
5370 Specify the sample rate, default is 44100.
5373 Specify the duration of the generated audio stream.
5375 @item samples_per_frame
5376 Set the number of samples per output frame.
5378 The expression can contain the following constants:
5382 The (sequential) number of the output audio frame, starting from 0.
5385 The PTS (Presentation TimeStamp) of the output audio frame,
5386 expressed in @var{TB} units.
5389 The PTS of the output audio frame, expressed in seconds.
5392 The timebase of the output audio frames.
5395 Default is @code{1024}.
5398 @subsection Examples
5403 Generate a simple 440 Hz sine wave:
5409 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5413 sine=frequency=220:beep_factor=4:duration=5
5417 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5420 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5424 @c man end AUDIO SOURCES
5426 @chapter Audio Sinks
5427 @c man begin AUDIO SINKS
5429 Below is a description of the currently available audio sinks.
5431 @section abuffersink
5433 Buffer audio frames, and make them available to the end of filter chain.
5435 This sink is mainly intended for programmatic use, in particular
5436 through the interface defined in @file{libavfilter/buffersink.h}
5437 or the options system.
5439 It accepts a pointer to an AVABufferSinkContext structure, which
5440 defines the incoming buffers' formats, to be passed as the opaque
5441 parameter to @code{avfilter_init_filter} for initialization.
5444 Null audio sink; do absolutely nothing with the input audio. It is
5445 mainly useful as a template and for use in analysis / debugging
5448 @c man end AUDIO SINKS
5450 @chapter Video Filters
5451 @c man begin VIDEO FILTERS
5453 When you configure your FFmpeg build, you can disable any of the
5454 existing filters using @code{--disable-filters}.
5455 The configure output will show the video filters included in your
5458 Below is a description of the currently available video filters.
5460 @section alphaextract
5462 Extract the alpha component from the input as a grayscale video. This
5463 is especially useful with the @var{alphamerge} filter.
5467 Add or replace the alpha component of the primary input with the
5468 grayscale value of a second input. This is intended for use with
5469 @var{alphaextract} to allow the transmission or storage of frame
5470 sequences that have alpha in a format that doesn't support an alpha
5473 For example, to reconstruct full frames from a normal YUV-encoded video
5474 and a separate video created with @var{alphaextract}, you might use:
5476 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5479 Since this filter is designed for reconstruction, it operates on frame
5480 sequences without considering timestamps, and terminates when either
5481 input reaches end of stream. This will cause problems if your encoding
5482 pipeline drops frames. If you're trying to apply an image as an
5483 overlay to a video stream, consider the @var{overlay} filter instead.
5487 Amplify differences between current pixel and pixels of adjacent frames in
5488 same pixel location.
5490 This filter accepts the following options:
5494 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5495 For example radius of 3 will instruct filter to calculate average of 7 frames.
5498 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5501 Set threshold for difference amplification. Any differrence greater or equal to
5502 this value will not alter source pixel. Default is 10.
5503 Allowed range is from 0 to 65535.
5506 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5507 This option controls maximum possible value that will decrease source pixel value.
5510 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5511 This option controls maximum possible value that will increase source pixel value.
5514 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5519 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5520 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5521 Substation Alpha) subtitles files.
5523 This filter accepts the following option in addition to the common options from
5524 the @ref{subtitles} filter:
5528 Set the shaping engine
5530 Available values are:
5533 The default libass shaping engine, which is the best available.
5535 Fast, font-agnostic shaper that can do only substitutions
5537 Slower shaper using OpenType for substitutions and positioning
5540 The default is @code{auto}.
5544 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5546 The filter accepts the following options:
5550 Set threshold A for 1st plane. Default is 0.02.
5551 Valid range is 0 to 0.3.
5554 Set threshold B for 1st plane. Default is 0.04.
5555 Valid range is 0 to 5.
5558 Set threshold A for 2nd plane. Default is 0.02.
5559 Valid range is 0 to 0.3.
5562 Set threshold B for 2nd plane. Default is 0.04.
5563 Valid range is 0 to 5.
5566 Set threshold A for 3rd plane. Default is 0.02.
5567 Valid range is 0 to 0.3.
5570 Set threshold B for 3rd plane. Default is 0.04.
5571 Valid range is 0 to 5.
5573 Threshold A is designed to react on abrupt changes in the input signal and
5574 threshold B is designed to react on continuous changes in the input signal.
5577 Set number of frames filter will use for averaging. Default is 9. Must be odd
5578 number in range [5, 129].
5581 Set what planes of frame filter will use for averaging. Default is all.
5586 Apply average blur filter.
5588 The filter accepts the following options:
5592 Set horizontal radius size.
5595 Set which planes to filter. By default all planes are filtered.
5598 Set vertical radius size, if zero it will be same as @code{sizeX}.
5599 Default is @code{0}.
5604 Compute the bounding box for the non-black pixels in the input frame
5607 This filter computes the bounding box containing all the pixels with a
5608 luminance value greater than the minimum allowed value.
5609 The parameters describing the bounding box are printed on the filter
5612 The filter accepts the following option:
5616 Set the minimal luminance value. Default is @code{16}.
5619 @section bitplanenoise
5621 Show and measure bit plane noise.
5623 The filter accepts the following options:
5627 Set which plane to analyze. Default is @code{1}.
5630 Filter out noisy pixels from @code{bitplane} set above.
5631 Default is disabled.
5634 @section blackdetect
5636 Detect video intervals that are (almost) completely black. Can be
5637 useful to detect chapter transitions, commercials, or invalid
5638 recordings. Output lines contains the time for the start, end and
5639 duration of the detected black interval expressed in seconds.
5641 In order to display the output lines, you need to set the loglevel at
5642 least to the AV_LOG_INFO value.
5644 The filter accepts the following options:
5647 @item black_min_duration, d
5648 Set the minimum detected black duration expressed in seconds. It must
5649 be a non-negative floating point number.
5651 Default value is 2.0.
5653 @item picture_black_ratio_th, pic_th
5654 Set the threshold for considering a picture "black".
5655 Express the minimum value for the ratio:
5657 @var{nb_black_pixels} / @var{nb_pixels}
5660 for which a picture is considered black.
5661 Default value is 0.98.
5663 @item pixel_black_th, pix_th
5664 Set the threshold for considering a pixel "black".
5666 The threshold expresses the maximum pixel luminance value for which a
5667 pixel is considered "black". The provided value is scaled according to
5668 the following equation:
5670 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5673 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5674 the input video format, the range is [0-255] for YUV full-range
5675 formats and [16-235] for YUV non full-range formats.
5677 Default value is 0.10.
5680 The following example sets the maximum pixel threshold to the minimum
5681 value, and detects only black intervals of 2 or more seconds:
5683 blackdetect=d=2:pix_th=0.00
5688 Detect frames that are (almost) completely black. Can be useful to
5689 detect chapter transitions or commercials. Output lines consist of
5690 the frame number of the detected frame, the percentage of blackness,
5691 the position in the file if known or -1 and the timestamp in seconds.
5693 In order to display the output lines, you need to set the loglevel at
5694 least to the AV_LOG_INFO value.
5696 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5697 The value represents the percentage of pixels in the picture that
5698 are below the threshold value.
5700 It accepts the following parameters:
5705 The percentage of the pixels that have to be below the threshold; it defaults to
5708 @item threshold, thresh
5709 The threshold below which a pixel value is considered black; it defaults to
5714 @section blend, tblend
5716 Blend two video frames into each other.
5718 The @code{blend} filter takes two input streams and outputs one
5719 stream, the first input is the "top" layer and second input is
5720 "bottom" layer. By default, the output terminates when the longest input terminates.
5722 The @code{tblend} (time blend) filter takes two consecutive frames
5723 from one single stream, and outputs the result obtained by blending
5724 the new frame on top of the old frame.
5726 A description of the accepted options follows.
5734 Set blend mode for specific pixel component or all pixel components in case
5735 of @var{all_mode}. Default value is @code{normal}.
5737 Available values for component modes are:
5779 Set blend opacity for specific pixel component or all pixel components in case
5780 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5787 Set blend expression for specific pixel component or all pixel components in case
5788 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5790 The expressions can use the following variables:
5794 The sequential number of the filtered frame, starting from @code{0}.
5798 the coordinates of the current sample
5802 the width and height of currently filtered plane
5806 Width and height scale for the plane being filtered. It is the
5807 ratio between the dimensions of the current plane to the luma plane,
5808 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
5809 the luma plane and @code{0.5,0.5} for the chroma planes.
5812 Time of the current frame, expressed in seconds.
5815 Value of pixel component at current location for first video frame (top layer).
5818 Value of pixel component at current location for second video frame (bottom layer).
5822 The @code{blend} filter also supports the @ref{framesync} options.
5824 @subsection Examples
5828 Apply transition from bottom layer to top layer in first 10 seconds:
5830 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5834 Apply linear horizontal transition from top layer to bottom layer:
5836 blend=all_expr='A*(X/W)+B*(1-X/W)'
5840 Apply 1x1 checkerboard effect:
5842 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5846 Apply uncover left effect:
5848 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5852 Apply uncover down effect:
5854 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5858 Apply uncover up-left effect:
5860 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5864 Split diagonally video and shows top and bottom layer on each side:
5866 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5870 Display differences between the current and the previous frame:
5872 tblend=all_mode=grainextract
5878 Denoise frames using Block-Matching 3D algorithm.
5880 The filter accepts the following options.
5884 Set denoising strength. Default value is 1.
5885 Allowed range is from 0 to 999.9.
5886 The denoising algorith is very sensitive to sigma, so adjust it
5887 according to the source.
5890 Set local patch size. This sets dimensions in 2D.
5893 Set sliding step for processing blocks. Default value is 4.
5894 Allowed range is from 1 to 64.
5895 Smaller values allows processing more reference blocks and is slower.
5898 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
5899 When set to 1, no block matching is done. Larger values allows more blocks
5901 Allowed range is from 1 to 256.
5904 Set radius for search block matching. Default is 9.
5905 Allowed range is from 1 to INT32_MAX.
5908 Set step between two search locations for block matching. Default is 1.
5909 Allowed range is from 1 to 64. Smaller is slower.
5912 Set threshold of mean square error for block matching. Valid range is 0 to
5916 Set thresholding parameter for hard thresholding in 3D transformed domain.
5917 Larger values results in stronger hard-thresholding filtering in frequency
5921 Set filtering estimation mode. Can be @code{basic} or @code{final}.
5922 Default is @code{basic}.
5925 If enabled, filter will use 2nd stream for block matching.
5926 Default is disabled for @code{basic} value of @var{estim} option,
5927 and always enabled if value of @var{estim} is @code{final}.
5930 Set planes to filter. Default is all available except alpha.
5933 @subsection Examples
5937 Basic filtering with bm3d:
5939 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
5943 Same as above, but filtering only luma:
5945 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
5949 Same as above, but with both estimation modes:
5951 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
5955 Same as above, but prefilter with @ref{nlmeans} filter instead:
5957 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
5963 Apply a boxblur algorithm to the input video.
5965 It accepts the following parameters:
5969 @item luma_radius, lr
5970 @item luma_power, lp
5971 @item chroma_radius, cr
5972 @item chroma_power, cp
5973 @item alpha_radius, ar
5974 @item alpha_power, ap
5978 A description of the accepted options follows.
5981 @item luma_radius, lr
5982 @item chroma_radius, cr
5983 @item alpha_radius, ar
5984 Set an expression for the box radius in pixels used for blurring the
5985 corresponding input plane.
5987 The radius value must be a non-negative number, and must not be
5988 greater than the value of the expression @code{min(w,h)/2} for the
5989 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5992 Default value for @option{luma_radius} is "2". If not specified,
5993 @option{chroma_radius} and @option{alpha_radius} default to the
5994 corresponding value set for @option{luma_radius}.
5996 The expressions can contain the following constants:
6000 The input width and height in pixels.
6004 The input chroma image width and height in pixels.
6008 The horizontal and vertical chroma subsample values. For example, for the
6009 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6012 @item luma_power, lp
6013 @item chroma_power, cp
6014 @item alpha_power, ap
6015 Specify how many times the boxblur filter is applied to the
6016 corresponding plane.
6018 Default value for @option{luma_power} is 2. If not specified,
6019 @option{chroma_power} and @option{alpha_power} default to the
6020 corresponding value set for @option{luma_power}.
6022 A value of 0 will disable the effect.
6025 @subsection Examples
6029 Apply a boxblur filter with the luma, chroma, and alpha radii
6032 boxblur=luma_radius=2:luma_power=1
6037 Set the luma radius to 2, and alpha and chroma radius to 0:
6039 boxblur=2:1:cr=0:ar=0
6043 Set the luma and chroma radii to a fraction of the video dimension:
6045 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6051 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6052 Deinterlacing Filter").
6054 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6055 interpolation algorithms.
6056 It accepts the following parameters:
6060 The interlacing mode to adopt. It accepts one of the following values:
6064 Output one frame for each frame.
6066 Output one frame for each field.
6069 The default value is @code{send_field}.
6072 The picture field parity assumed for the input interlaced video. It accepts one
6073 of the following values:
6077 Assume the top field is first.
6079 Assume the bottom field is first.
6081 Enable automatic detection of field parity.
6084 The default value is @code{auto}.
6085 If the interlacing is unknown or the decoder does not export this information,
6086 top field first will be assumed.
6089 Specify which frames to deinterlace. Accept one of the following
6094 Deinterlace all frames.
6096 Only deinterlace frames marked as interlaced.
6099 The default value is @code{all}.
6103 YUV colorspace color/chroma keying.
6105 The filter accepts the following options:
6109 The color which will be replaced with transparency.
6112 Similarity percentage with the key color.
6114 0.01 matches only the exact key color, while 1.0 matches everything.
6119 0.0 makes pixels either fully transparent, or not transparent at all.
6121 Higher values result in semi-transparent pixels, with a higher transparency
6122 the more similar the pixels color is to the key color.
6125 Signals that the color passed is already in YUV instead of RGB.
6127 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6128 This can be used to pass exact YUV values as hexadecimal numbers.
6131 @subsection Examples
6135 Make every green pixel in the input image transparent:
6137 ffmpeg -i input.png -vf chromakey=green out.png
6141 Overlay a greenscreen-video on top of a static black background.
6143 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
6149 Display CIE color diagram with pixels overlaid onto it.
6151 The filter accepts the following options:
6166 @item uhdtv, rec2020
6179 Set what gamuts to draw.
6181 See @code{system} option for available values.
6184 Set ciescope size, by default set to 512.
6187 Set intensity used to map input pixel values to CIE diagram.
6190 Set contrast used to draw tongue colors that are out of active color system gamut.
6193 Correct gamma displayed on scope, by default enabled.
6196 Show white point on CIE diagram, by default disabled.
6199 Set input gamma. Used only with XYZ input color space.
6204 Visualize information exported by some codecs.
6206 Some codecs can export information through frames using side-data or other
6207 means. For example, some MPEG based codecs export motion vectors through the
6208 @var{export_mvs} flag in the codec @option{flags2} option.
6210 The filter accepts the following option:
6214 Set motion vectors to visualize.
6216 Available flags for @var{mv} are:
6220 forward predicted MVs of P-frames
6222 forward predicted MVs of B-frames
6224 backward predicted MVs of B-frames
6228 Display quantization parameters using the chroma planes.
6231 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6233 Available flags for @var{mv_type} are:
6237 forward predicted MVs
6239 backward predicted MVs
6242 @item frame_type, ft
6243 Set frame type to visualize motion vectors of.
6245 Available flags for @var{frame_type} are:
6249 intra-coded frames (I-frames)
6251 predicted frames (P-frames)
6253 bi-directionally predicted frames (B-frames)
6257 @subsection Examples
6261 Visualize forward predicted MVs of all frames using @command{ffplay}:
6263 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6267 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6269 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6273 @section colorbalance
6274 Modify intensity of primary colors (red, green and blue) of input frames.
6276 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6277 regions for the red-cyan, green-magenta or blue-yellow balance.
6279 A positive adjustment value shifts the balance towards the primary color, a negative
6280 value towards the complementary color.
6282 The filter accepts the following options:
6288 Adjust red, green and blue shadows (darkest pixels).
6293 Adjust red, green and blue midtones (medium pixels).
6298 Adjust red, green and blue highlights (brightest pixels).
6300 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6303 @subsection Examples
6307 Add red color cast to shadows:
6314 RGB colorspace color keying.
6316 The filter accepts the following options:
6320 The color which will be replaced with transparency.
6323 Similarity percentage with the key color.
6325 0.01 matches only the exact key color, while 1.0 matches everything.
6330 0.0 makes pixels either fully transparent, or not transparent at all.
6332 Higher values result in semi-transparent pixels, with a higher transparency
6333 the more similar the pixels color is to the key color.
6336 @subsection Examples
6340 Make every green pixel in the input image transparent:
6342 ffmpeg -i input.png -vf colorkey=green out.png
6346 Overlay a greenscreen-video on top of a static background image.
6348 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
6352 @section colorlevels
6354 Adjust video input frames using levels.
6356 The filter accepts the following options:
6363 Adjust red, green, blue and alpha input black point.
6364 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6370 Adjust red, green, blue and alpha input white point.
6371 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6373 Input levels are used to lighten highlights (bright tones), darken shadows
6374 (dark tones), change the balance of bright and dark tones.
6380 Adjust red, green, blue and alpha output black point.
6381 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6387 Adjust red, green, blue and alpha output white point.
6388 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6390 Output levels allows manual selection of a constrained output level range.
6393 @subsection Examples
6397 Make video output darker:
6399 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6405 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6409 Make video output lighter:
6411 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6415 Increase brightness:
6417 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6421 @section colorchannelmixer
6423 Adjust video input frames by re-mixing color channels.
6425 This filter modifies a color channel by adding the values associated to
6426 the other channels of the same pixels. For example if the value to
6427 modify is red, the output value will be:
6429 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6432 The filter accepts the following options:
6439 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6440 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6446 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6447 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6453 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6454 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6460 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6461 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6463 Allowed ranges for options are @code{[-2.0, 2.0]}.
6466 @subsection Examples
6470 Convert source to grayscale:
6472 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6475 Simulate sepia tones:
6477 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6481 @section colormatrix
6483 Convert color matrix.
6485 The filter accepts the following options:
6490 Specify the source and destination color matrix. Both values must be
6493 The accepted values are:
6521 For example to convert from BT.601 to SMPTE-240M, use the command:
6523 colormatrix=bt601:smpte240m
6528 Convert colorspace, transfer characteristics or color primaries.
6529 Input video needs to have an even size.
6531 The filter accepts the following options:
6536 Specify all color properties at once.
6538 The accepted values are:
6568 Specify output colorspace.
6570 The accepted values are:
6579 BT.470BG or BT.601-6 625
6582 SMPTE-170M or BT.601-6 525
6591 BT.2020 with non-constant luminance
6597 Specify output transfer characteristics.
6599 The accepted values are:
6611 Constant gamma of 2.2
6614 Constant gamma of 2.8
6617 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6635 BT.2020 for 10-bits content
6638 BT.2020 for 12-bits content
6644 Specify output color primaries.
6646 The accepted values are:
6655 BT.470BG or BT.601-6 625
6658 SMPTE-170M or BT.601-6 525
6682 Specify output color range.
6684 The accepted values are:
6687 TV (restricted) range
6690 MPEG (restricted) range
6701 Specify output color format.
6703 The accepted values are:
6706 YUV 4:2:0 planar 8-bits
6709 YUV 4:2:0 planar 10-bits
6712 YUV 4:2:0 planar 12-bits
6715 YUV 4:2:2 planar 8-bits
6718 YUV 4:2:2 planar 10-bits
6721 YUV 4:2:2 planar 12-bits
6724 YUV 4:4:4 planar 8-bits
6727 YUV 4:4:4 planar 10-bits
6730 YUV 4:4:4 planar 12-bits
6735 Do a fast conversion, which skips gamma/primary correction. This will take
6736 significantly less CPU, but will be mathematically incorrect. To get output
6737 compatible with that produced by the colormatrix filter, use fast=1.
6740 Specify dithering mode.
6742 The accepted values are:
6748 Floyd-Steinberg dithering
6752 Whitepoint adaptation mode.
6754 The accepted values are:
6757 Bradford whitepoint adaptation
6760 von Kries whitepoint adaptation
6763 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6767 Override all input properties at once. Same accepted values as @ref{all}.
6770 Override input colorspace. Same accepted values as @ref{space}.
6773 Override input color primaries. Same accepted values as @ref{primaries}.
6776 Override input transfer characteristics. Same accepted values as @ref{trc}.
6779 Override input color range. Same accepted values as @ref{range}.
6783 The filter converts the transfer characteristics, color space and color
6784 primaries to the specified user values. The output value, if not specified,
6785 is set to a default value based on the "all" property. If that property is
6786 also not specified, the filter will log an error. The output color range and
6787 format default to the same value as the input color range and format. The
6788 input transfer characteristics, color space, color primaries and color range
6789 should be set on the input data. If any of these are missing, the filter will
6790 log an error and no conversion will take place.
6792 For example to convert the input to SMPTE-240M, use the command:
6794 colorspace=smpte240m
6797 @section convolution
6799 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
6801 The filter accepts the following options:
6808 Set matrix for each plane.
6809 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
6810 and from 1 to 49 odd number of signed integers in @var{row} mode.
6816 Set multiplier for calculated value for each plane.
6817 If unset or 0, it will be sum of all matrix elements.
6823 Set bias for each plane. This value is added to the result of the multiplication.
6824 Useful for making the overall image brighter or darker. Default is 0.0.
6830 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
6831 Default is @var{square}.
6834 @subsection Examples
6840 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"
6846 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"
6852 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"
6858 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"
6862 Apply laplacian edge detector which includes diagonals:
6864 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"
6870 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"
6876 Apply 2D convolution of video stream in frequency domain using second stream
6879 The filter accepts the following options:
6883 Set which planes to process.
6886 Set which impulse video frames will be processed, can be @var{first}
6887 or @var{all}. Default is @var{all}.
6890 The @code{convolve} filter also supports the @ref{framesync} options.
6894 Copy the input video source unchanged to the output. This is mainly useful for
6899 Video filtering on GPU using Apple's CoreImage API on OSX.
6901 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6902 processed by video hardware. However, software-based OpenGL implementations
6903 exist which means there is no guarantee for hardware processing. It depends on
6906 There are many filters and image generators provided by Apple that come with a
6907 large variety of options. The filter has to be referenced by its name along
6910 The coreimage filter accepts the following options:
6913 List all available filters and generators along with all their respective
6914 options as well as possible minimum and maximum values along with the default
6921 Specify all filters by their respective name and options.
6922 Use @var{list_filters} to determine all valid filter names and options.
6923 Numerical options are specified by a float value and are automatically clamped
6924 to their respective value range. Vector and color options have to be specified
6925 by a list of space separated float values. Character escaping has to be done.
6926 A special option name @code{default} is available to use default options for a
6929 It is required to specify either @code{default} or at least one of the filter options.
6930 All omitted options are used with their default values.
6931 The syntax of the filter string is as follows:
6933 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6937 Specify a rectangle where the output of the filter chain is copied into the
6938 input image. It is given by a list of space separated float values:
6940 output_rect=x\ y\ width\ height
6942 If not given, the output rectangle equals the dimensions of the input image.
6943 The output rectangle is automatically cropped at the borders of the input
6944 image. Negative values are valid for each component.
6946 output_rect=25\ 25\ 100\ 100
6950 Several filters can be chained for successive processing without GPU-HOST
6951 transfers allowing for fast processing of complex filter chains.
6952 Currently, only filters with zero (generators) or exactly one (filters) input
6953 image and one output image are supported. Also, transition filters are not yet
6956 Some filters generate output images with additional padding depending on the
6957 respective filter kernel. The padding is automatically removed to ensure the
6958 filter output has the same size as the input image.
6960 For image generators, the size of the output image is determined by the
6961 previous output image of the filter chain or the input image of the whole
6962 filterchain, respectively. The generators do not use the pixel information of
6963 this image to generate their output. However, the generated output is
6964 blended onto this image, resulting in partial or complete coverage of the
6967 The @ref{coreimagesrc} video source can be used for generating input images
6968 which are directly fed into the filter chain. By using it, providing input
6969 images by another video source or an input video is not required.
6971 @subsection Examples
6976 List all filters available:
6978 coreimage=list_filters=true
6982 Use the CIBoxBlur filter with default options to blur an image:
6984 coreimage=filter=CIBoxBlur@@default
6988 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6989 its center at 100x100 and a radius of 50 pixels:
6991 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6995 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6996 given as complete and escaped command-line for Apple's standard bash shell:
6998 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7004 Crop the input video to given dimensions.
7006 It accepts the following parameters:
7010 The width of the output video. It defaults to @code{iw}.
7011 This expression is evaluated only once during the filter
7012 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7015 The height of the output video. It defaults to @code{ih}.
7016 This expression is evaluated only once during the filter
7017 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7020 The horizontal position, in the input video, of the left edge of the output
7021 video. It defaults to @code{(in_w-out_w)/2}.
7022 This expression is evaluated per-frame.
7025 The vertical position, in the input video, of the top edge of the output video.
7026 It defaults to @code{(in_h-out_h)/2}.
7027 This expression is evaluated per-frame.
7030 If set to 1 will force the output display aspect ratio
7031 to be the same of the input, by changing the output sample aspect
7032 ratio. It defaults to 0.
7035 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7036 width/height/x/y as specified and will not be rounded to nearest smaller value.
7040 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7041 expressions containing the following constants:
7046 The computed values for @var{x} and @var{y}. They are evaluated for
7051 The input width and height.
7055 These are the same as @var{in_w} and @var{in_h}.
7059 The output (cropped) width and height.
7063 These are the same as @var{out_w} and @var{out_h}.
7066 same as @var{iw} / @var{ih}
7069 input sample aspect ratio
7072 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7076 horizontal and vertical chroma subsample values. For example for the
7077 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7080 The number of the input frame, starting from 0.
7083 the position in the file of the input frame, NAN if unknown
7086 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7090 The expression for @var{out_w} may depend on the value of @var{out_h},
7091 and the expression for @var{out_h} may depend on @var{out_w}, but they
7092 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7093 evaluated after @var{out_w} and @var{out_h}.
7095 The @var{x} and @var{y} parameters specify the expressions for the
7096 position of the top-left corner of the output (non-cropped) area. They
7097 are evaluated for each frame. If the evaluated value is not valid, it
7098 is approximated to the nearest valid value.
7100 The expression for @var{x} may depend on @var{y}, and the expression
7101 for @var{y} may depend on @var{x}.
7103 @subsection Examples
7107 Crop area with size 100x100 at position (12,34).
7112 Using named options, the example above becomes:
7114 crop=w=100:h=100:x=12:y=34
7118 Crop the central input area with size 100x100:
7124 Crop the central input area with size 2/3 of the input video:
7126 crop=2/3*in_w:2/3*in_h
7130 Crop the input video central square:
7137 Delimit the rectangle with the top-left corner placed at position
7138 100:100 and the right-bottom corner corresponding to the right-bottom
7139 corner of the input image.
7141 crop=in_w-100:in_h-100:100:100
7145 Crop 10 pixels from the left and right borders, and 20 pixels from
7146 the top and bottom borders
7148 crop=in_w-2*10:in_h-2*20
7152 Keep only the bottom right quarter of the input image:
7154 crop=in_w/2:in_h/2:in_w/2:in_h/2
7158 Crop height for getting Greek harmony:
7160 crop=in_w:1/PHI*in_w
7164 Apply trembling effect:
7166 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)
7170 Apply erratic camera effect depending on timestamp:
7172 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)"
7176 Set x depending on the value of y:
7178 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7182 @subsection Commands
7184 This filter supports the following commands:
7190 Set width/height of the output video and the horizontal/vertical position
7192 The command accepts the same syntax of the corresponding option.
7194 If the specified expression is not valid, it is kept at its current
7200 Auto-detect the crop size.
7202 It calculates the necessary cropping parameters and prints the
7203 recommended parameters via the logging system. The detected dimensions
7204 correspond to the non-black area of the input video.
7206 It accepts the following parameters:
7211 Set higher black value threshold, which can be optionally specified
7212 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7213 value greater to the set value is considered non-black. It defaults to 24.
7214 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7215 on the bitdepth of the pixel format.
7218 The value which the width/height should be divisible by. It defaults to
7219 16. The offset is automatically adjusted to center the video. Use 2 to
7220 get only even dimensions (needed for 4:2:2 video). 16 is best when
7221 encoding to most video codecs.
7223 @item reset_count, reset
7224 Set the counter that determines after how many frames cropdetect will
7225 reset the previously detected largest video area and start over to
7226 detect the current optimal crop area. Default value is 0.
7228 This can be useful when channel logos distort the video area. 0
7229 indicates 'never reset', and returns the largest area encountered during
7236 Delay video filtering until a given wallclock timestamp. The filter first
7237 passes on @option{preroll} amount of frames, then it buffers at most
7238 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7239 it forwards the buffered frames and also any subsequent frames coming in its
7242 The filter can be used synchronize the output of multiple ffmpeg processes for
7243 realtime output devices like decklink. By putting the delay in the filtering
7244 chain and pre-buffering frames the process can pass on data to output almost
7245 immediately after the target wallclock timestamp is reached.
7247 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7253 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7256 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7259 The maximum duration of content to buffer before waiting for the cue expressed
7260 in seconds. Default is 0.
7267 Apply color adjustments using curves.
7269 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7270 component (red, green and blue) has its values defined by @var{N} key points
7271 tied from each other using a smooth curve. The x-axis represents the pixel
7272 values from the input frame, and the y-axis the new pixel values to be set for
7275 By default, a component curve is defined by the two points @var{(0;0)} and
7276 @var{(1;1)}. This creates a straight line where each original pixel value is
7277 "adjusted" to its own value, which means no change to the image.
7279 The filter allows you to redefine these two points and add some more. A new
7280 curve (using a natural cubic spline interpolation) will be define to pass
7281 smoothly through all these new coordinates. The new defined points needs to be
7282 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7283 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7284 the vector spaces, the values will be clipped accordingly.
7286 The filter accepts the following options:
7290 Select one of the available color presets. This option can be used in addition
7291 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7292 options takes priority on the preset values.
7293 Available presets are:
7296 @item color_negative
7299 @item increase_contrast
7301 @item linear_contrast
7302 @item medium_contrast
7304 @item strong_contrast
7307 Default is @code{none}.
7309 Set the master key points. These points will define a second pass mapping. It
7310 is sometimes called a "luminance" or "value" mapping. It can be used with
7311 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7312 post-processing LUT.
7314 Set the key points for the red component.
7316 Set the key points for the green component.
7318 Set the key points for the blue component.
7320 Set the key points for all components (not including master).
7321 Can be used in addition to the other key points component
7322 options. In this case, the unset component(s) will fallback on this
7323 @option{all} setting.
7325 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7327 Save Gnuplot script of the curves in specified file.
7330 To avoid some filtergraph syntax conflicts, each key points list need to be
7331 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7333 @subsection Examples
7337 Increase slightly the middle level of blue:
7339 curves=blue='0/0 0.5/0.58 1/1'
7345 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'
7347 Here we obtain the following coordinates for each components:
7350 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7352 @code{(0;0) (0.50;0.48) (1;1)}
7354 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7358 The previous example can also be achieved with the associated built-in preset:
7360 curves=preset=vintage
7370 Use a Photoshop preset and redefine the points of the green component:
7372 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7376 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7377 and @command{gnuplot}:
7379 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7380 gnuplot -p /tmp/curves.plt
7386 Video data analysis filter.
7388 This filter shows hexadecimal pixel values of part of video.
7390 The filter accepts the following options:
7394 Set output video size.
7397 Set x offset from where to pick pixels.
7400 Set y offset from where to pick pixels.
7403 Set scope mode, can be one of the following:
7406 Draw hexadecimal pixel values with white color on black background.
7409 Draw hexadecimal pixel values with input video pixel color on black
7413 Draw hexadecimal pixel values on color background picked from input video,
7414 the text color is picked in such way so its always visible.
7418 Draw rows and columns numbers on left and top of video.
7421 Set background opacity.
7426 Denoise frames using 2D DCT (frequency domain filtering).
7428 This filter is not designed for real time.
7430 The filter accepts the following options:
7434 Set the noise sigma constant.
7436 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7437 coefficient (absolute value) below this threshold with be dropped.
7439 If you need a more advanced filtering, see @option{expr}.
7441 Default is @code{0}.
7444 Set number overlapping pixels for each block. Since the filter can be slow, you
7445 may want to reduce this value, at the cost of a less effective filter and the
7446 risk of various artefacts.
7448 If the overlapping value doesn't permit processing the whole input width or
7449 height, a warning will be displayed and according borders won't be denoised.
7451 Default value is @var{blocksize}-1, which is the best possible setting.
7454 Set the coefficient factor expression.
7456 For each coefficient of a DCT block, this expression will be evaluated as a
7457 multiplier value for the coefficient.
7459 If this is option is set, the @option{sigma} option will be ignored.
7461 The absolute value of the coefficient can be accessed through the @var{c}
7465 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7466 @var{blocksize}, which is the width and height of the processed blocks.
7468 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7469 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7470 on the speed processing. Also, a larger block size does not necessarily means a
7474 @subsection Examples
7476 Apply a denoise with a @option{sigma} of @code{4.5}:
7481 The same operation can be achieved using the expression system:
7483 dctdnoiz=e='gte(c, 4.5*3)'
7486 Violent denoise using a block size of @code{16x16}:
7493 Remove banding artifacts from input video.
7494 It works by replacing banded pixels with average value of referenced pixels.
7496 The filter accepts the following options:
7503 Set banding detection threshold for each plane. Default is 0.02.
7504 Valid range is 0.00003 to 0.5.
7505 If difference between current pixel and reference pixel is less than threshold,
7506 it will be considered as banded.
7509 Banding detection range in pixels. Default is 16. If positive, random number
7510 in range 0 to set value will be used. If negative, exact absolute value
7512 The range defines square of four pixels around current pixel.
7515 Set direction in radians from which four pixel will be compared. If positive,
7516 random direction from 0 to set direction will be picked. If negative, exact of
7517 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7518 will pick only pixels on same row and -PI/2 will pick only pixels on same
7522 If enabled, current pixel is compared with average value of all four
7523 surrounding pixels. The default is enabled. If disabled current pixel is
7524 compared with all four surrounding pixels. The pixel is considered banded
7525 if only all four differences with surrounding pixels are less than threshold.
7528 If enabled, current pixel is changed if and only if all pixel components are banded,
7529 e.g. banding detection threshold is triggered for all color components.
7530 The default is disabled.
7535 Remove blocking artifacts from input video.
7537 The filter accepts the following options:
7541 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7542 This controls what kind of deblocking is applied.
7545 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7551 Set blocking detection thresholds. Allowed range is 0 to 1.
7552 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7553 Using higher threshold gives more deblocking strength.
7554 Setting @var{alpha} controls threshold detection at exact edge of block.
7555 Remaining options controls threshold detection near the edge. Each one for
7556 below/above or left/right. Setting any of those to @var{0} disables
7560 Set planes to filter. Default is to filter all available planes.
7563 @subsection Examples
7567 Deblock using weak filter and block size of 4 pixels.
7569 deblock=filter=weak:block=4
7573 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7574 deblocking more edges.
7576 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7580 Similar as above, but filter only first plane.
7582 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7586 Similar as above, but filter only second and third plane.
7588 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7595 Drop duplicated frames at regular intervals.
7597 The filter accepts the following options:
7601 Set the number of frames from which one will be dropped. Setting this to
7602 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7603 Default is @code{5}.
7606 Set the threshold for duplicate detection. If the difference metric for a frame
7607 is less than or equal to this value, then it is declared as duplicate. Default
7611 Set scene change threshold. Default is @code{15}.
7615 Set the size of the x and y-axis blocks used during metric calculations.
7616 Larger blocks give better noise suppression, but also give worse detection of
7617 small movements. Must be a power of two. Default is @code{32}.
7620 Mark main input as a pre-processed input and activate clean source input
7621 stream. This allows the input to be pre-processed with various filters to help
7622 the metrics calculation while keeping the frame selection lossless. When set to
7623 @code{1}, the first stream is for the pre-processed input, and the second
7624 stream is the clean source from where the kept frames are chosen. Default is
7628 Set whether or not chroma is considered in the metric calculations. Default is
7634 Apply 2D deconvolution of video stream in frequency domain using second stream
7637 The filter accepts the following options:
7641 Set which planes to process.
7644 Set which impulse video frames will be processed, can be @var{first}
7645 or @var{all}. Default is @var{all}.
7648 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7649 and height are not same and not power of 2 or if stream prior to convolving
7653 The @code{deconvolve} filter also supports the @ref{framesync} options.
7657 Apply deflate effect to the video.
7659 This filter replaces the pixel by the local(3x3) average by taking into account
7660 only values lower than the pixel.
7662 It accepts the following options:
7669 Limit the maximum change for each plane, default is 65535.
7670 If 0, plane will remain unchanged.
7675 Remove temporal frame luminance variations.
7677 It accepts the following options:
7681 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7684 Set averaging mode to smooth temporal luminance variations.
7686 Available values are:
7711 Do not actually modify frame. Useful when one only wants metadata.
7716 Remove judder produced by partially interlaced telecined content.
7718 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7719 source was partially telecined content then the output of @code{pullup,dejudder}
7720 will have a variable frame rate. May change the recorded frame rate of the
7721 container. Aside from that change, this filter will not affect constant frame
7724 The option available in this filter is:
7728 Specify the length of the window over which the judder repeats.
7730 Accepts any integer greater than 1. Useful values are:
7734 If the original was telecined from 24 to 30 fps (Film to NTSC).
7737 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7740 If a mixture of the two.
7743 The default is @samp{4}.
7748 Suppress a TV station logo by a simple interpolation of the surrounding
7749 pixels. Just set a rectangle covering the logo and watch it disappear
7750 (and sometimes something even uglier appear - your mileage may vary).
7752 It accepts the following parameters:
7757 Specify the top left corner coordinates of the logo. They must be
7762 Specify the width and height of the logo to clear. They must be
7766 Specify the thickness of the fuzzy edge of the rectangle (added to
7767 @var{w} and @var{h}). The default value is 1. This option is
7768 deprecated, setting higher values should no longer be necessary and
7772 When set to 1, a green rectangle is drawn on the screen to simplify
7773 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7774 The default value is 0.
7776 The rectangle is drawn on the outermost pixels which will be (partly)
7777 replaced with interpolated values. The values of the next pixels
7778 immediately outside this rectangle in each direction will be used to
7779 compute the interpolated pixel values inside the rectangle.
7783 @subsection Examples
7787 Set a rectangle covering the area with top left corner coordinates 0,0
7788 and size 100x77, and a band of size 10:
7790 delogo=x=0:y=0:w=100:h=77:band=10
7797 Attempt to fix small changes in horizontal and/or vertical shift. This
7798 filter helps remove camera shake from hand-holding a camera, bumping a
7799 tripod, moving on a vehicle, etc.
7801 The filter accepts the following options:
7809 Specify a rectangular area where to limit the search for motion
7811 If desired the search for motion vectors can be limited to a
7812 rectangular area of the frame defined by its top left corner, width
7813 and height. These parameters have the same meaning as the drawbox
7814 filter which can be used to visualise the position of the bounding
7817 This is useful when simultaneous movement of subjects within the frame
7818 might be confused for camera motion by the motion vector search.
7820 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7821 then the full frame is used. This allows later options to be set
7822 without specifying the bounding box for the motion vector search.
7824 Default - search the whole frame.
7828 Specify the maximum extent of movement in x and y directions in the
7829 range 0-64 pixels. Default 16.
7832 Specify how to generate pixels to fill blanks at the edge of the
7833 frame. Available values are:
7836 Fill zeroes at blank locations
7838 Original image at blank locations
7840 Extruded edge value at blank locations
7842 Mirrored edge at blank locations
7844 Default value is @samp{mirror}.
7847 Specify the blocksize to use for motion search. Range 4-128 pixels,
7851 Specify the contrast threshold for blocks. Only blocks with more than
7852 the specified contrast (difference between darkest and lightest
7853 pixels) will be considered. Range 1-255, default 125.
7856 Specify the search strategy. Available values are:
7859 Set exhaustive search
7861 Set less exhaustive search.
7863 Default value is @samp{exhaustive}.
7866 If set then a detailed log of the motion search is written to the
7873 Remove unwanted contamination of foreground colors, caused by reflected color of
7874 greenscreen or bluescreen.
7876 This filter accepts the following options:
7880 Set what type of despill to use.
7883 Set how spillmap will be generated.
7886 Set how much to get rid of still remaining spill.
7889 Controls amount of red in spill area.
7892 Controls amount of green in spill area.
7893 Should be -1 for greenscreen.
7896 Controls amount of blue in spill area.
7897 Should be -1 for bluescreen.
7900 Controls brightness of spill area, preserving colors.
7903 Modify alpha from generated spillmap.
7908 Apply an exact inverse of the telecine operation. It requires a predefined
7909 pattern specified using the pattern option which must be the same as that passed
7910 to the telecine filter.
7912 This filter accepts the following options:
7921 The default value is @code{top}.
7925 A string of numbers representing the pulldown pattern you wish to apply.
7926 The default value is @code{23}.
7929 A number representing position of the first frame with respect to the telecine
7930 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7935 Apply dilation effect to the video.
7937 This filter replaces the pixel by the local(3x3) maximum.
7939 It accepts the following options:
7946 Limit the maximum change for each plane, default is 65535.
7947 If 0, plane will remain unchanged.
7950 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7953 Flags to local 3x3 coordinates maps like this:
7962 Displace pixels as indicated by second and third input stream.
7964 It takes three input streams and outputs one stream, the first input is the
7965 source, and second and third input are displacement maps.
7967 The second input specifies how much to displace pixels along the
7968 x-axis, while the third input specifies how much to displace pixels
7970 If one of displacement map streams terminates, last frame from that
7971 displacement map will be used.
7973 Note that once generated, displacements maps can be reused over and over again.
7975 A description of the accepted options follows.
7979 Set displace behavior for pixels that are out of range.
7981 Available values are:
7984 Missing pixels are replaced by black pixels.
7987 Adjacent pixels will spread out to replace missing pixels.
7990 Out of range pixels are wrapped so they point to pixels of other side.
7993 Out of range pixels will be replaced with mirrored pixels.
7995 Default is @samp{smear}.
7999 @subsection Examples
8003 Add ripple effect to rgb input of video size hd720:
8005 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
8009 Add wave effect to rgb input of video size hd720:
8011 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
8017 Draw a colored box on the input image.
8019 It accepts the following parameters:
8024 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8028 The expressions which specify the width and height of the box; if 0 they are interpreted as
8029 the input width and height. It defaults to 0.
8032 Specify the color of the box to write. For the general syntax of this option,
8033 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8034 value @code{invert} is used, the box edge color is the same as the
8035 video with inverted luma.
8038 The expression which sets the thickness of the box edge.
8039 A value of @code{fill} will create a filled box. Default value is @code{3}.
8041 See below for the list of accepted constants.
8044 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8045 will overwrite the video's color and alpha pixels.
8046 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8049 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8050 following constants:
8054 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8058 horizontal and vertical chroma subsample values. For example for the
8059 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8063 The input width and height.
8066 The input sample aspect ratio.
8070 The x and y offset coordinates where the box is drawn.
8074 The width and height of the drawn box.
8077 The thickness of the drawn box.
8079 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8080 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8084 @subsection Examples
8088 Draw a black box around the edge of the input image:
8094 Draw a box with color red and an opacity of 50%:
8096 drawbox=10:20:200:60:red@@0.5
8099 The previous example can be specified as:
8101 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8105 Fill the box with pink color:
8107 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8111 Draw a 2-pixel red 2.40:1 mask:
8113 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
8119 Draw a grid on the input image.
8121 It accepts the following parameters:
8126 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8130 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8131 input width and height, respectively, minus @code{thickness}, so image gets
8132 framed. Default to 0.
8135 Specify the color of the grid. For the general syntax of this option,
8136 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8137 value @code{invert} is used, the grid color is the same as the
8138 video with inverted luma.
8141 The expression which sets the thickness of the grid line. Default value is @code{1}.
8143 See below for the list of accepted constants.
8146 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8147 will overwrite the video's color and alpha pixels.
8148 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8151 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8152 following constants:
8156 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8160 horizontal and vertical chroma subsample values. For example for the
8161 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8165 The input grid cell width and height.
8168 The input sample aspect ratio.
8172 The x and y coordinates of some point of grid intersection (meant to configure offset).
8176 The width and height of the drawn cell.
8179 The thickness of the drawn cell.
8181 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8182 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8186 @subsection Examples
8190 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8192 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8196 Draw a white 3x3 grid with an opacity of 50%:
8198 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8205 Draw a text string or text from a specified file on top of a video, using the
8206 libfreetype library.
8208 To enable compilation of this filter, you need to configure FFmpeg with
8209 @code{--enable-libfreetype}.
8210 To enable default font fallback and the @var{font} option you need to
8211 configure FFmpeg with @code{--enable-libfontconfig}.
8212 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8213 @code{--enable-libfribidi}.
8217 It accepts the following parameters:
8222 Used to draw a box around text using the background color.
8223 The value must be either 1 (enable) or 0 (disable).
8224 The default value of @var{box} is 0.
8227 Set the width of the border to be drawn around the box using @var{boxcolor}.
8228 The default value of @var{boxborderw} is 0.
8231 The color to be used for drawing box around text. For the syntax of this
8232 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8234 The default value of @var{boxcolor} is "white".
8237 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8238 The default value of @var{line_spacing} is 0.
8241 Set the width of the border to be drawn around the text using @var{bordercolor}.
8242 The default value of @var{borderw} is 0.
8245 Set the color to be used for drawing border around text. For the syntax of this
8246 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8248 The default value of @var{bordercolor} is "black".
8251 Select how the @var{text} is expanded. Can be either @code{none},
8252 @code{strftime} (deprecated) or
8253 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8257 Set a start time for the count. Value is in microseconds. Only applied
8258 in the deprecated strftime expansion mode. To emulate in normal expansion
8259 mode use the @code{pts} function, supplying the start time (in seconds)
8260 as the second argument.
8263 If true, check and fix text coords to avoid clipping.
8266 The color to be used for drawing fonts. For the syntax of this option, check
8267 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8269 The default value of @var{fontcolor} is "black".
8271 @item fontcolor_expr
8272 String which is expanded the same way as @var{text} to obtain dynamic
8273 @var{fontcolor} value. By default this option has empty value and is not
8274 processed. When this option is set, it overrides @var{fontcolor} option.
8277 The font family to be used for drawing text. By default Sans.
8280 The font file to be used for drawing text. The path must be included.
8281 This parameter is mandatory if the fontconfig support is disabled.
8284 Draw the text applying alpha blending. The value can
8285 be a number between 0.0 and 1.0.
8286 The expression accepts the same variables @var{x, y} as well.
8287 The default value is 1.
8288 Please see @var{fontcolor_expr}.
8291 The font size to be used for drawing text.
8292 The default value of @var{fontsize} is 16.
8295 If set to 1, attempt to shape the text (for example, reverse the order of
8296 right-to-left text and join Arabic characters) before drawing it.
8297 Otherwise, just draw the text exactly as given.
8298 By default 1 (if supported).
8301 The flags to be used for loading the fonts.
8303 The flags map the corresponding flags supported by libfreetype, and are
8304 a combination of the following values:
8311 @item vertical_layout
8312 @item force_autohint
8315 @item ignore_global_advance_width
8317 @item ignore_transform
8323 Default value is "default".
8325 For more information consult the documentation for the FT_LOAD_*
8329 The color to be used for drawing a shadow behind the drawn text. For the
8330 syntax of this option, check the @ref{color syntax,,"Color" section in the
8331 ffmpeg-utils manual,ffmpeg-utils}.
8333 The default value of @var{shadowcolor} is "black".
8337 The x and y offsets for the text shadow position with respect to the
8338 position of the text. They can be either positive or negative
8339 values. The default value for both is "0".
8342 The starting frame number for the n/frame_num variable. The default value
8346 The size in number of spaces to use for rendering the tab.
8350 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8351 format. It can be used with or without text parameter. @var{timecode_rate}
8352 option must be specified.
8354 @item timecode_rate, rate, r
8355 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8356 integer. Minimum value is "1".
8357 Drop-frame timecode is supported for frame rates 30 & 60.
8360 If set to 1, the output of the timecode option will wrap around at 24 hours.
8361 Default is 0 (disabled).
8364 The text string to be drawn. The text must be a sequence of UTF-8
8366 This parameter is mandatory if no file is specified with the parameter
8370 A text file containing text to be drawn. The text must be a sequence
8371 of UTF-8 encoded characters.
8373 This parameter is mandatory if no text string is specified with the
8374 parameter @var{text}.
8376 If both @var{text} and @var{textfile} are specified, an error is thrown.
8379 If set to 1, the @var{textfile} will be reloaded before each frame.
8380 Be sure to update it atomically, or it may be read partially, or even fail.
8384 The expressions which specify the offsets where text will be drawn
8385 within the video frame. They are relative to the top/left border of the
8388 The default value of @var{x} and @var{y} is "0".
8390 See below for the list of accepted constants and functions.
8393 The parameters for @var{x} and @var{y} are expressions containing the
8394 following constants and functions:
8398 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8402 horizontal and vertical chroma subsample values. For example for the
8403 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8406 the height of each text line
8414 @item max_glyph_a, ascent
8415 the maximum distance from the baseline to the highest/upper grid
8416 coordinate used to place a glyph outline point, for all the rendered
8418 It is a positive value, due to the grid's orientation with the Y axis
8421 @item max_glyph_d, descent
8422 the maximum distance from the baseline to the lowest grid coordinate
8423 used to place a glyph outline point, for all the rendered glyphs.
8424 This is a negative value, due to the grid's orientation, with the Y axis
8428 maximum glyph height, that is the maximum height for all the glyphs
8429 contained in the rendered text, it is equivalent to @var{ascent} -
8433 maximum glyph width, that is the maximum width for all the glyphs
8434 contained in the rendered text
8437 the number of input frame, starting from 0
8439 @item rand(min, max)
8440 return a random number included between @var{min} and @var{max}
8443 The input sample aspect ratio.
8446 timestamp expressed in seconds, NAN if the input timestamp is unknown
8449 the height of the rendered text
8452 the width of the rendered text
8456 the x and y offset coordinates where the text is drawn.
8458 These parameters allow the @var{x} and @var{y} expressions to refer
8459 each other, so you can for example specify @code{y=x/dar}.
8462 @anchor{drawtext_expansion}
8463 @subsection Text expansion
8465 If @option{expansion} is set to @code{strftime},
8466 the filter recognizes strftime() sequences in the provided text and
8467 expands them accordingly. Check the documentation of strftime(). This
8468 feature is deprecated.
8470 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8472 If @option{expansion} is set to @code{normal} (which is the default),
8473 the following expansion mechanism is used.
8475 The backslash character @samp{\}, followed by any character, always expands to
8476 the second character.
8478 Sequences of the form @code{%@{...@}} are expanded. The text between the
8479 braces is a function name, possibly followed by arguments separated by ':'.
8480 If the arguments contain special characters or delimiters (':' or '@}'),
8481 they should be escaped.
8483 Note that they probably must also be escaped as the value for the
8484 @option{text} option in the filter argument string and as the filter
8485 argument in the filtergraph description, and possibly also for the shell,
8486 that makes up to four levels of escaping; using a text file avoids these
8489 The following functions are available:
8494 The expression evaluation result.
8496 It must take one argument specifying the expression to be evaluated,
8497 which accepts the same constants and functions as the @var{x} and
8498 @var{y} values. Note that not all constants should be used, for
8499 example the text size is not known when evaluating the expression, so
8500 the constants @var{text_w} and @var{text_h} will have an undefined
8503 @item expr_int_format, eif
8504 Evaluate the expression's value and output as formatted integer.
8506 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8507 The second argument specifies the output format. Allowed values are @samp{x},
8508 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8509 @code{printf} function.
8510 The third parameter is optional and sets the number of positions taken by the output.
8511 It can be used to add padding with zeros from the left.
8514 The time at which the filter is running, expressed in UTC.
8515 It can accept an argument: a strftime() format string.
8518 The time at which the filter is running, expressed in the local time zone.
8519 It can accept an argument: a strftime() format string.
8522 Frame metadata. Takes one or two arguments.
8524 The first argument is mandatory and specifies the metadata key.
8526 The second argument is optional and specifies a default value, used when the
8527 metadata key is not found or empty.
8530 The frame number, starting from 0.
8533 A 1 character description of the current picture type.
8536 The timestamp of the current frame.
8537 It can take up to three arguments.
8539 The first argument is the format of the timestamp; it defaults to @code{flt}
8540 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8541 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8542 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8543 @code{localtime} stands for the timestamp of the frame formatted as
8544 local time zone time.
8546 The second argument is an offset added to the timestamp.
8548 If the format is set to @code{hms}, a third argument @code{24HH} may be
8549 supplied to present the hour part of the formatted timestamp in 24h format
8552 If the format is set to @code{localtime} or @code{gmtime},
8553 a third argument may be supplied: a strftime() format string.
8554 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8557 @subsection Examples
8561 Draw "Test Text" with font FreeSerif, using the default values for the
8562 optional parameters.
8565 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8569 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8570 and y=50 (counting from the top-left corner of the screen), text is
8571 yellow with a red box around it. Both the text and the box have an
8575 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8576 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8579 Note that the double quotes are not necessary if spaces are not used
8580 within the parameter list.
8583 Show the text at the center of the video frame:
8585 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8589 Show the text at a random position, switching to a new position every 30 seconds:
8591 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)"
8595 Show a text line sliding from right to left in the last row of the video
8596 frame. The file @file{LONG_LINE} is assumed to contain a single line
8599 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8603 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8605 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8609 Draw a single green letter "g", at the center of the input video.
8610 The glyph baseline is placed at half screen height.
8612 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8616 Show text for 1 second every 3 seconds:
8618 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8622 Use fontconfig to set the font. Note that the colons need to be escaped.
8624 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8628 Print the date of a real-time encoding (see strftime(3)):
8630 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8634 Show text fading in and out (appearing/disappearing):
8637 DS=1.0 # display start
8638 DE=10.0 # display end
8639 FID=1.5 # fade in duration
8640 FOD=5 # fade out duration
8641 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 @}"
8645 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8646 and the @option{fontsize} value are included in the @option{y} offset.
8648 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8649 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8654 For more information about libfreetype, check:
8655 @url{http://www.freetype.org/}.
8657 For more information about fontconfig, check:
8658 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8660 For more information about libfribidi, check:
8661 @url{http://fribidi.org/}.
8665 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8667 The filter accepts the following options:
8672 Set low and high threshold values used by the Canny thresholding
8675 The high threshold selects the "strong" edge pixels, which are then
8676 connected through 8-connectivity with the "weak" edge pixels selected
8677 by the low threshold.
8679 @var{low} and @var{high} threshold values must be chosen in the range
8680 [0,1], and @var{low} should be lesser or equal to @var{high}.
8682 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8686 Define the drawing mode.
8690 Draw white/gray wires on black background.
8693 Mix the colors to create a paint/cartoon effect.
8696 Apply Canny edge detector on all selected planes.
8698 Default value is @var{wires}.
8701 Select planes for filtering. By default all available planes are filtered.
8704 @subsection Examples
8708 Standard edge detection with custom values for the hysteresis thresholding:
8710 edgedetect=low=0.1:high=0.4
8714 Painting effect without thresholding:
8716 edgedetect=mode=colormix:high=0
8721 Set brightness, contrast, saturation and approximate gamma adjustment.
8723 The filter accepts the following options:
8727 Set the contrast expression. The value must be a float value in range
8728 @code{-2.0} to @code{2.0}. The default value is "1".
8731 Set the brightness expression. The value must be a float value in
8732 range @code{-1.0} to @code{1.0}. The default value is "0".
8735 Set the saturation expression. The value must be a float in
8736 range @code{0.0} to @code{3.0}. The default value is "1".
8739 Set the gamma expression. The value must be a float in range
8740 @code{0.1} to @code{10.0}. The default value is "1".
8743 Set the gamma expression for red. The value must be a float in
8744 range @code{0.1} to @code{10.0}. The default value is "1".
8747 Set the gamma expression for green. The value must be a float in range
8748 @code{0.1} to @code{10.0}. The default value is "1".
8751 Set the gamma expression for blue. The value must be a float in range
8752 @code{0.1} to @code{10.0}. The default value is "1".
8755 Set the gamma weight expression. It can be used to reduce the effect
8756 of a high gamma value on bright image areas, e.g. keep them from
8757 getting overamplified and just plain white. The value must be a float
8758 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8759 gamma correction all the way down while @code{1.0} leaves it at its
8760 full strength. Default is "1".
8763 Set when the expressions for brightness, contrast, saturation and
8764 gamma expressions are evaluated.
8766 It accepts the following values:
8769 only evaluate expressions once during the filter initialization or
8770 when a command is processed
8773 evaluate expressions for each incoming frame
8776 Default value is @samp{init}.
8779 The expressions accept the following parameters:
8782 frame count of the input frame starting from 0
8785 byte position of the corresponding packet in the input file, NAN if
8789 frame rate of the input video, NAN if the input frame rate is unknown
8792 timestamp expressed in seconds, NAN if the input timestamp is unknown
8795 @subsection Commands
8796 The filter supports the following commands:
8800 Set the contrast expression.
8803 Set the brightness expression.
8806 Set the saturation expression.
8809 Set the gamma expression.
8812 Set the gamma_r expression.
8815 Set gamma_g expression.
8818 Set gamma_b expression.
8821 Set gamma_weight expression.
8823 The command accepts the same syntax of the corresponding option.
8825 If the specified expression is not valid, it is kept at its current
8832 Apply erosion effect to the video.
8834 This filter replaces the pixel by the local(3x3) minimum.
8836 It accepts the following options:
8843 Limit the maximum change for each plane, default is 65535.
8844 If 0, plane will remain unchanged.
8847 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8850 Flags to local 3x3 coordinates maps like this:
8857 @section extractplanes
8859 Extract color channel components from input video stream into
8860 separate grayscale video streams.
8862 The filter accepts the following option:
8866 Set plane(s) to extract.
8868 Available values for planes are:
8879 Choosing planes not available in the input will result in an error.
8880 That means you cannot select @code{r}, @code{g}, @code{b} planes
8881 with @code{y}, @code{u}, @code{v} planes at same time.
8884 @subsection Examples
8888 Extract luma, u and v color channel component from input video frame
8889 into 3 grayscale outputs:
8891 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
8897 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8899 For each input image, the filter will compute the optimal mapping from
8900 the input to the output given the codebook length, that is the number
8901 of distinct output colors.
8903 This filter accepts the following options.
8906 @item codebook_length, l
8907 Set codebook length. The value must be a positive integer, and
8908 represents the number of distinct output colors. Default value is 256.
8911 Set the maximum number of iterations to apply for computing the optimal
8912 mapping. The higher the value the better the result and the higher the
8913 computation time. Default value is 1.
8916 Set a random seed, must be an integer included between 0 and
8917 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8918 will try to use a good random seed on a best effort basis.
8921 Set pal8 output pixel format. This option does not work with codebook
8922 length greater than 256.
8927 Measure graylevel entropy in histogram of color channels of video frames.
8929 It accepts the following parameters:
8933 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8935 @var{diff} mode measures entropy of histogram delta values, absolute differences
8936 between neighbour histogram values.
8941 Apply a fade-in/out effect to the input video.
8943 It accepts the following parameters:
8947 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8949 Default is @code{in}.
8951 @item start_frame, s
8952 Specify the number of the frame to start applying the fade
8953 effect at. Default is 0.
8956 The number of frames that the fade effect lasts. At the end of the
8957 fade-in effect, the output video will have the same intensity as the input video.
8958 At the end of the fade-out transition, the output video will be filled with the
8959 selected @option{color}.
8963 If set to 1, fade only alpha channel, if one exists on the input.
8966 @item start_time, st
8967 Specify the timestamp (in seconds) of the frame to start to apply the fade
8968 effect. If both start_frame and start_time are specified, the fade will start at
8969 whichever comes last. Default is 0.
8972 The number of seconds for which the fade effect has to last. At the end of the
8973 fade-in effect the output video will have the same intensity as the input video,
8974 at the end of the fade-out transition the output video will be filled with the
8975 selected @option{color}.
8976 If both duration and nb_frames are specified, duration is used. Default is 0
8977 (nb_frames is used by default).
8980 Specify the color of the fade. Default is "black".
8983 @subsection Examples
8987 Fade in the first 30 frames of video:
8992 The command above is equivalent to:
8998 Fade out the last 45 frames of a 200-frame video:
9001 fade=type=out:start_frame=155:nb_frames=45
9005 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9007 fade=in:0:25, fade=out:975:25
9011 Make the first 5 frames yellow, then fade in from frame 5-24:
9013 fade=in:5:20:color=yellow
9017 Fade in alpha over first 25 frames of video:
9019 fade=in:0:25:alpha=1
9023 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9025 fade=t=in:st=5.5:d=0.5
9031 Apply arbitrary expressions to samples in frequency domain
9035 Adjust the dc value (gain) of the luma plane of the image. The filter
9036 accepts an integer value in range @code{0} to @code{1000}. The default
9037 value is set to @code{0}.
9040 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9041 filter accepts an integer value in range @code{0} to @code{1000}. The
9042 default value is set to @code{0}.
9045 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9046 filter accepts an integer value in range @code{0} to @code{1000}. The
9047 default value is set to @code{0}.
9050 Set the frequency domain weight expression for the luma plane.
9053 Set the frequency domain weight expression for the 1st chroma plane.
9056 Set the frequency domain weight expression for the 2nd chroma plane.
9059 Set when the expressions are evaluated.
9061 It accepts the following values:
9064 Only evaluate expressions once during the filter initialization.
9067 Evaluate expressions for each incoming frame.
9070 Default value is @samp{init}.
9072 The filter accepts the following variables:
9075 The coordinates of the current sample.
9079 The width and height of the image.
9082 The number of input frame, starting from 0.
9085 @subsection Examples
9091 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9097 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9103 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9109 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9115 Denoise frames using 3D FFT (frequency domain filtering).
9117 The filter accepts the following options:
9121 Set the noise sigma constant. This sets denoising strength.
9122 Default value is 1. Allowed range is from 0 to 30.
9123 Using very high sigma with low overlap may give blocking artifacts.
9126 Set amount of denoising. By default all detected noise is reduced.
9127 Default value is 1. Allowed range is from 0 to 1.
9130 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9131 Actual size of block in pixels is 2 to power of @var{block}, so by default
9132 block size in pixels is 2^4 which is 16.
9135 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9138 Set number of previous frames to use for denoising. By default is set to 0.
9141 Set number of next frames to to use for denoising. By default is set to 0.
9144 Set planes which will be filtered, by default are all available filtered
9150 Extract a single field from an interlaced image using stride
9151 arithmetic to avoid wasting CPU time. The output frames are marked as
9154 The filter accepts the following options:
9158 Specify whether to extract the top (if the value is @code{0} or
9159 @code{top}) or the bottom field (if the value is @code{1} or
9165 Create new frames by copying the top and bottom fields from surrounding frames
9166 supplied as numbers by the hint file.
9170 Set file containing hints: absolute/relative frame numbers.
9172 There must be one line for each frame in a clip. Each line must contain two
9173 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9174 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9175 is current frame number for @code{absolute} mode or out of [-1, 1] range
9176 for @code{relative} mode. First number tells from which frame to pick up top
9177 field and second number tells from which frame to pick up bottom field.
9179 If optionally followed by @code{+} output frame will be marked as interlaced,
9180 else if followed by @code{-} output frame will be marked as progressive, else
9181 it will be marked same as input frame.
9182 If line starts with @code{#} or @code{;} that line is skipped.
9185 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9188 Example of first several lines of @code{hint} file for @code{relative} mode:
9191 1,0 - # second frame, use third's frame top field and second's frame bottom field
9192 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9209 Field matching filter for inverse telecine. It is meant to reconstruct the
9210 progressive frames from a telecined stream. The filter does not drop duplicated
9211 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9212 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9214 The separation of the field matching and the decimation is notably motivated by
9215 the possibility of inserting a de-interlacing filter fallback between the two.
9216 If the source has mixed telecined and real interlaced content,
9217 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9218 But these remaining combed frames will be marked as interlaced, and thus can be
9219 de-interlaced by a later filter such as @ref{yadif} before decimation.
9221 In addition to the various configuration options, @code{fieldmatch} can take an
9222 optional second stream, activated through the @option{ppsrc} option. If
9223 enabled, the frames reconstruction will be based on the fields and frames from
9224 this second stream. This allows the first input to be pre-processed in order to
9225 help the various algorithms of the filter, while keeping the output lossless
9226 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9227 or brightness/contrast adjustments can help.
9229 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9230 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9231 which @code{fieldmatch} is based on. While the semantic and usage are very
9232 close, some behaviour and options names can differ.
9234 The @ref{decimate} filter currently only works for constant frame rate input.
9235 If your input has mixed telecined (30fps) and progressive content with a lower
9236 framerate like 24fps use the following filterchain to produce the necessary cfr
9237 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9239 The filter accepts the following options:
9243 Specify the assumed field order of the input stream. Available values are:
9247 Auto detect parity (use FFmpeg's internal parity value).
9249 Assume bottom field first.
9251 Assume top field first.
9254 Note that it is sometimes recommended not to trust the parity announced by the
9257 Default value is @var{auto}.
9260 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9261 sense that it won't risk creating jerkiness due to duplicate frames when
9262 possible, but if there are bad edits or blended fields it will end up
9263 outputting combed frames when a good match might actually exist. On the other
9264 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9265 but will almost always find a good frame if there is one. The other values are
9266 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9267 jerkiness and creating duplicate frames versus finding good matches in sections
9268 with bad edits, orphaned fields, blended fields, etc.
9270 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9272 Available values are:
9276 2-way matching (p/c)
9278 2-way matching, and trying 3rd match if still combed (p/c + n)
9280 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9282 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9283 still combed (p/c + n + u/b)
9285 3-way matching (p/c/n)
9287 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9288 detected as combed (p/c/n + u/b)
9291 The parenthesis at the end indicate the matches that would be used for that
9292 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9295 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9298 Default value is @var{pc_n}.
9301 Mark the main input stream as a pre-processed input, and enable the secondary
9302 input stream as the clean source to pick the fields from. See the filter
9303 introduction for more details. It is similar to the @option{clip2} feature from
9306 Default value is @code{0} (disabled).
9309 Set the field to match from. It is recommended to set this to the same value as
9310 @option{order} unless you experience matching failures with that setting. In
9311 certain circumstances changing the field that is used to match from can have a
9312 large impact on matching performance. Available values are:
9316 Automatic (same value as @option{order}).
9318 Match from the bottom field.
9320 Match from the top field.
9323 Default value is @var{auto}.
9326 Set whether or not chroma is included during the match comparisons. In most
9327 cases it is recommended to leave this enabled. You should set this to @code{0}
9328 only if your clip has bad chroma problems such as heavy rainbowing or other
9329 artifacts. Setting this to @code{0} could also be used to speed things up at
9330 the cost of some accuracy.
9332 Default value is @code{1}.
9336 These define an exclusion band which excludes the lines between @option{y0} and
9337 @option{y1} from being included in the field matching decision. An exclusion
9338 band can be used to ignore subtitles, a logo, or other things that may
9339 interfere with the matching. @option{y0} sets the starting scan line and
9340 @option{y1} sets the ending line; all lines in between @option{y0} and
9341 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9342 @option{y0} and @option{y1} to the same value will disable the feature.
9343 @option{y0} and @option{y1} defaults to @code{0}.
9346 Set the scene change detection threshold as a percentage of maximum change on
9347 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9348 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9349 @option{scthresh} is @code{[0.0, 100.0]}.
9351 Default value is @code{12.0}.
9354 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9355 account the combed scores of matches when deciding what match to use as the
9356 final match. Available values are:
9360 No final matching based on combed scores.
9362 Combed scores are only used when a scene change is detected.
9364 Use combed scores all the time.
9367 Default is @var{sc}.
9370 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9371 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9372 Available values are:
9376 No forced calculation.
9378 Force p/c/n calculations.
9380 Force p/c/n/u/b calculations.
9383 Default value is @var{none}.
9386 This is the area combing threshold used for combed frame detection. This
9387 essentially controls how "strong" or "visible" combing must be to be detected.
9388 Larger values mean combing must be more visible and smaller values mean combing
9389 can be less visible or strong and still be detected. Valid settings are from
9390 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9391 be detected as combed). This is basically a pixel difference value. A good
9392 range is @code{[8, 12]}.
9394 Default value is @code{9}.
9397 Sets whether or not chroma is considered in the combed frame decision. Only
9398 disable this if your source has chroma problems (rainbowing, etc.) that are
9399 causing problems for the combed frame detection with chroma enabled. Actually,
9400 using @option{chroma}=@var{0} is usually more reliable, except for the case
9401 where there is chroma only combing in the source.
9403 Default value is @code{0}.
9407 Respectively set the x-axis and y-axis size of the window used during combed
9408 frame detection. This has to do with the size of the area in which
9409 @option{combpel} pixels are required to be detected as combed for a frame to be
9410 declared combed. See the @option{combpel} parameter description for more info.
9411 Possible values are any number that is a power of 2 starting at 4 and going up
9414 Default value is @code{16}.
9417 The number of combed pixels inside any of the @option{blocky} by
9418 @option{blockx} size blocks on the frame for the frame to be detected as
9419 combed. While @option{cthresh} controls how "visible" the combing must be, this
9420 setting controls "how much" combing there must be in any localized area (a
9421 window defined by the @option{blockx} and @option{blocky} settings) on the
9422 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9423 which point no frames will ever be detected as combed). This setting is known
9424 as @option{MI} in TFM/VFM vocabulary.
9426 Default value is @code{80}.
9429 @anchor{p/c/n/u/b meaning}
9430 @subsection p/c/n/u/b meaning
9432 @subsubsection p/c/n
9434 We assume the following telecined stream:
9437 Top fields: 1 2 2 3 4
9438 Bottom fields: 1 2 3 4 4
9441 The numbers correspond to the progressive frame the fields relate to. Here, the
9442 first two frames are progressive, the 3rd and 4th are combed, and so on.
9444 When @code{fieldmatch} is configured to run a matching from bottom
9445 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9450 B 1 2 3 4 4 <-- matching reference
9459 As a result of the field matching, we can see that some frames get duplicated.
9460 To perform a complete inverse telecine, you need to rely on a decimation filter
9461 after this operation. See for instance the @ref{decimate} filter.
9463 The same operation now matching from top fields (@option{field}=@var{top})
9468 T 1 2 2 3 4 <-- matching reference
9478 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9479 basically, they refer to the frame and field of the opposite parity:
9482 @item @var{p} matches the field of the opposite parity in the previous frame
9483 @item @var{c} matches the field of the opposite parity in the current frame
9484 @item @var{n} matches the field of the opposite parity in the next frame
9489 The @var{u} and @var{b} matching are a bit special in the sense that they match
9490 from the opposite parity flag. In the following examples, we assume that we are
9491 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9492 'x' is placed above and below each matched fields.
9494 With bottom matching (@option{field}=@var{bottom}):
9499 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9500 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9508 With top matching (@option{field}=@var{top}):
9513 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9514 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9522 @subsection Examples
9524 Simple IVTC of a top field first telecined stream:
9526 fieldmatch=order=tff:combmatch=none, decimate
9529 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9531 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9536 Transform the field order of the input video.
9538 It accepts the following parameters:
9543 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9544 for bottom field first.
9547 The default value is @samp{tff}.
9549 The transformation is done by shifting the picture content up or down
9550 by one line, and filling the remaining line with appropriate picture content.
9551 This method is consistent with most broadcast field order converters.
9553 If the input video is not flagged as being interlaced, or it is already
9554 flagged as being of the required output field order, then this filter does
9555 not alter the incoming video.
9557 It is very useful when converting to or from PAL DV material,
9558 which is bottom field first.
9562 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9565 @section fifo, afifo
9567 Buffer input images and send them when they are requested.
9569 It is mainly useful when auto-inserted by the libavfilter
9572 It does not take parameters.
9574 @section fillborders
9576 Fill borders of the input video, without changing video stream dimensions.
9577 Sometimes video can have garbage at the four edges and you may not want to
9578 crop video input to keep size multiple of some number.
9580 This filter accepts the following options:
9584 Number of pixels to fill from left border.
9587 Number of pixels to fill from right border.
9590 Number of pixels to fill from top border.
9593 Number of pixels to fill from bottom border.
9598 It accepts the following values:
9601 fill pixels using outermost pixels
9604 fill pixels using mirroring
9607 fill pixels with constant value
9610 Default is @var{smear}.
9613 Set color for pixels in fixed mode. Default is @var{black}.
9618 Find a rectangular object
9620 It accepts the following options:
9624 Filepath of the object image, needs to be in gray8.
9627 Detection threshold, default is 0.5.
9630 Number of mipmaps, default is 3.
9632 @item xmin, ymin, xmax, ymax
9633 Specifies the rectangle in which to search.
9636 @subsection Examples
9640 Generate a representative palette of a given video using @command{ffmpeg}:
9642 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9648 Cover a rectangular object
9650 It accepts the following options:
9654 Filepath of the optional cover image, needs to be in yuv420.
9659 It accepts the following values:
9662 cover it by the supplied image
9664 cover it by interpolating the surrounding pixels
9667 Default value is @var{blur}.
9670 @subsection Examples
9674 Generate a representative palette of a given video using @command{ffmpeg}:
9676 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9682 Flood area with values of same pixel components with another values.
9684 It accepts the following options:
9687 Set pixel x coordinate.
9690 Set pixel y coordinate.
9693 Set source #0 component value.
9696 Set source #1 component value.
9699 Set source #2 component value.
9702 Set source #3 component value.
9705 Set destination #0 component value.
9708 Set destination #1 component value.
9711 Set destination #2 component value.
9714 Set destination #3 component value.
9720 Convert the input video to one of the specified pixel formats.
9721 Libavfilter will try to pick one that is suitable as input to
9724 It accepts the following parameters:
9728 A '|'-separated list of pixel format names, such as
9729 "pix_fmts=yuv420p|monow|rgb24".
9733 @subsection Examples
9737 Convert the input video to the @var{yuv420p} format
9739 format=pix_fmts=yuv420p
9742 Convert the input video to any of the formats in the list
9744 format=pix_fmts=yuv420p|yuv444p|yuv410p
9751 Convert the video to specified constant frame rate by duplicating or dropping
9752 frames as necessary.
9754 It accepts the following parameters:
9758 The desired output frame rate. The default is @code{25}.
9761 Assume the first PTS should be the given value, in seconds. This allows for
9762 padding/trimming at the start of stream. By default, no assumption is made
9763 about the first frame's expected PTS, so no padding or trimming is done.
9764 For example, this could be set to 0 to pad the beginning with duplicates of
9765 the first frame if a video stream starts after the audio stream or to trim any
9766 frames with a negative PTS.
9769 Timestamp (PTS) rounding method.
9771 Possible values are:
9778 round towards -infinity
9780 round towards +infinity
9784 The default is @code{near}.
9787 Action performed when reading the last frame.
9789 Possible values are:
9792 Use same timestamp rounding method as used for other frames.
9794 Pass through last frame if input duration has not been reached yet.
9796 The default is @code{round}.
9800 Alternatively, the options can be specified as a flat string:
9801 @var{fps}[:@var{start_time}[:@var{round}]].
9803 See also the @ref{setpts} filter.
9805 @subsection Examples
9809 A typical usage in order to set the fps to 25:
9815 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9817 fps=fps=film:round=near
9823 Pack two different video streams into a stereoscopic video, setting proper
9824 metadata on supported codecs. The two views should have the same size and
9825 framerate and processing will stop when the shorter video ends. Please note
9826 that you may conveniently adjust view properties with the @ref{scale} and
9829 It accepts the following parameters:
9833 The desired packing format. Supported values are:
9838 The views are next to each other (default).
9841 The views are on top of each other.
9844 The views are packed by line.
9847 The views are packed by column.
9850 The views are temporally interleaved.
9859 # Convert left and right views into a frame-sequential video
9860 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9862 # Convert views into a side-by-side video with the same output resolution as the input
9863 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
9868 Change the frame rate by interpolating new video output frames from the source
9871 This filter is not designed to function correctly with interlaced media. If
9872 you wish to change the frame rate of interlaced media then you are required
9873 to deinterlace before this filter and re-interlace after this filter.
9875 A description of the accepted options follows.
9879 Specify the output frames per second. This option can also be specified
9880 as a value alone. The default is @code{50}.
9883 Specify the start of a range where the output frame will be created as a
9884 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9885 the default is @code{15}.
9888 Specify the end of a range where the output frame will be created as a
9889 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9890 the default is @code{240}.
9893 Specify the level at which a scene change is detected as a value between
9894 0 and 100 to indicate a new scene; a low value reflects a low
9895 probability for the current frame to introduce a new scene, while a higher
9896 value means the current frame is more likely to be one.
9897 The default is @code{8.2}.
9900 Specify flags influencing the filter process.
9902 Available value for @var{flags} is:
9905 @item scene_change_detect, scd
9906 Enable scene change detection using the value of the option @var{scene}.
9907 This flag is enabled by default.
9913 Select one frame every N-th frame.
9915 This filter accepts the following option:
9918 Select frame after every @code{step} frames.
9919 Allowed values are positive integers higher than 0. Default value is @code{1}.
9925 Apply a frei0r effect to the input video.
9927 To enable the compilation of this filter, you need to install the frei0r
9928 header and configure FFmpeg with @code{--enable-frei0r}.
9930 It accepts the following parameters:
9935 The name of the frei0r effect to load. If the environment variable
9936 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9937 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9938 Otherwise, the standard frei0r paths are searched, in this order:
9939 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9940 @file{/usr/lib/frei0r-1/}.
9943 A '|'-separated list of parameters to pass to the frei0r effect.
9947 A frei0r effect parameter can be a boolean (its value is either
9948 "y" or "n"), a double, a color (specified as
9949 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9950 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9951 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9952 a position (specified as @var{X}/@var{Y}, where
9953 @var{X} and @var{Y} are floating point numbers) and/or a string.
9955 The number and types of parameters depend on the loaded effect. If an
9956 effect parameter is not specified, the default value is set.
9958 @subsection Examples
9962 Apply the distort0r effect, setting the first two double parameters:
9964 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9968 Apply the colordistance effect, taking a color as the first parameter:
9970 frei0r=colordistance:0.2/0.3/0.4
9971 frei0r=colordistance:violet
9972 frei0r=colordistance:0x112233
9976 Apply the perspective effect, specifying the top left and top right image
9979 frei0r=perspective:0.2/0.2|0.8/0.2
9983 For more information, see
9984 @url{http://frei0r.dyne.org}
9988 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9990 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9991 processing filter, one of them is performed once per block, not per pixel.
9992 This allows for much higher speed.
9994 The filter accepts the following options:
9998 Set quality. This option defines the number of levels for averaging. It accepts
9999 an integer in the range 4-5. Default value is @code{4}.
10002 Force a constant quantization parameter. It accepts an integer in range 0-63.
10003 If not set, the filter will use the QP from the video stream (if available).
10006 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10007 more details but also more artifacts, while higher values make the image smoother
10008 but also blurrier. Default value is @code{0} − PSNR optimal.
10010 @item use_bframe_qp
10011 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10012 option may cause flicker since the B-Frames have often larger QP. Default is
10013 @code{0} (not enabled).
10019 Apply Gaussian blur filter.
10021 The filter accepts the following options:
10025 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10028 Set number of steps for Gaussian approximation. Defauls is @code{1}.
10031 Set which planes to filter. By default all planes are filtered.
10034 Set vertical sigma, if negative it will be same as @code{sigma}.
10035 Default is @code{-1}.
10040 The filter accepts the following options:
10043 @item lum_expr, lum
10044 Set the luminance expression.
10046 Set the chrominance blue expression.
10048 Set the chrominance red expression.
10049 @item alpha_expr, a
10050 Set the alpha expression.
10052 Set the red expression.
10053 @item green_expr, g
10054 Set the green expression.
10056 Set the blue expression.
10059 The colorspace is selected according to the specified options. If one
10060 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10061 options is specified, the filter will automatically select a YCbCr
10062 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10063 @option{blue_expr} options is specified, it will select an RGB
10066 If one of the chrominance expression is not defined, it falls back on the other
10067 one. If no alpha expression is specified it will evaluate to opaque value.
10068 If none of chrominance expressions are specified, they will evaluate
10069 to the luminance expression.
10071 The expressions can use the following variables and functions:
10075 The sequential number of the filtered frame, starting from @code{0}.
10079 The coordinates of the current sample.
10083 The width and height of the image.
10087 Width and height scale depending on the currently filtered plane. It is the
10088 ratio between the corresponding luma plane number of pixels and the current
10089 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10090 @code{0.5,0.5} for chroma planes.
10093 Time of the current frame, expressed in seconds.
10096 Return the value of the pixel at location (@var{x},@var{y}) of the current
10100 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10104 Return the value of the pixel at location (@var{x},@var{y}) of the
10105 blue-difference chroma plane. Return 0 if there is no such plane.
10108 Return the value of the pixel at location (@var{x},@var{y}) of the
10109 red-difference chroma plane. Return 0 if there is no such plane.
10114 Return the value of the pixel at location (@var{x},@var{y}) of the
10115 red/green/blue component. Return 0 if there is no such component.
10118 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10119 plane. Return 0 if there is no such plane.
10122 For functions, if @var{x} and @var{y} are outside the area, the value will be
10123 automatically clipped to the closer edge.
10125 @subsection Examples
10129 Flip the image horizontally:
10135 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10136 wavelength of 100 pixels:
10138 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10142 Generate a fancy enigmatic moving light:
10144 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
10148 Generate a quick emboss effect:
10150 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10154 Modify RGB components depending on pixel position:
10156 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10160 Create a radial gradient that is the same size as the input (also see
10161 the @ref{vignette} filter):
10163 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10169 Fix the banding artifacts that are sometimes introduced into nearly flat
10170 regions by truncation to 8-bit color depth.
10171 Interpolate the gradients that should go where the bands are, and
10174 It is designed for playback only. Do not use it prior to
10175 lossy compression, because compression tends to lose the dither and
10176 bring back the bands.
10178 It accepts the following parameters:
10183 The maximum amount by which the filter will change any one pixel. This is also
10184 the threshold for detecting nearly flat regions. Acceptable values range from
10185 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10189 The neighborhood to fit the gradient to. A larger radius makes for smoother
10190 gradients, but also prevents the filter from modifying the pixels near detailed
10191 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10192 values will be clipped to the valid range.
10196 Alternatively, the options can be specified as a flat string:
10197 @var{strength}[:@var{radius}]
10199 @subsection Examples
10203 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10209 Specify radius, omitting the strength (which will fall-back to the default
10218 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10219 and corrects the scene colors accordingly.
10221 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10223 The filter accepts the following options:
10227 The order of differentiation to be applied on the scene. Must be chosen in the range
10228 [0,2] and default value is 1.
10231 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10232 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10233 max value instead of calculating Minkowski distance.
10236 The standard deviation of Gaussian blur to be applied on the scene. Must be
10237 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10238 can't be euqal to 0 if @var{difford} is greater than 0.
10241 @subsection Examples
10247 greyedge=difford=1:minknorm=5:sigma=2
10253 greyedge=difford=1:minknorm=0:sigma=2
10261 Apply a Hald CLUT to a video stream.
10263 First input is the video stream to process, and second one is the Hald CLUT.
10264 The Hald CLUT input can be a simple picture or a complete video stream.
10266 The filter accepts the following options:
10270 Force termination when the shortest input terminates. Default is @code{0}.
10272 Continue applying the last CLUT after the end of the stream. A value of
10273 @code{0} disable the filter after the last frame of the CLUT is reached.
10274 Default is @code{1}.
10277 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10278 filters share the same internals).
10280 More information about the Hald CLUT can be found on Eskil Steenberg's website
10281 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10283 @subsection Workflow examples
10285 @subsubsection Hald CLUT video stream
10287 Generate an identity Hald CLUT stream altered with various effects:
10289 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
10292 Note: make sure you use a lossless codec.
10294 Then use it with @code{haldclut} to apply it on some random stream:
10296 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10299 The Hald CLUT will be applied to the 10 first seconds (duration of
10300 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10301 to the remaining frames of the @code{mandelbrot} stream.
10303 @subsubsection Hald CLUT with preview
10305 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10306 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10307 biggest possible square starting at the top left of the picture. The remaining
10308 padding pixels (bottom or right) will be ignored. This area can be used to add
10309 a preview of the Hald CLUT.
10311 Typically, the following generated Hald CLUT will be supported by the
10312 @code{haldclut} filter:
10315 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10316 pad=iw+320 [padded_clut];
10317 smptebars=s=320x256, split [a][b];
10318 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10319 [main][b] overlay=W-320" -frames:v 1 clut.png
10322 It contains the original and a preview of the effect of the CLUT: SMPTE color
10323 bars are displayed on the right-top, and below the same color bars processed by
10326 Then, the effect of this Hald CLUT can be visualized with:
10328 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10333 Flip the input video horizontally.
10335 For example, to horizontally flip the input video with @command{ffmpeg}:
10337 ffmpeg -i in.avi -vf "hflip" out.avi
10341 This filter applies a global color histogram equalization on a
10344 It can be used to correct video that has a compressed range of pixel
10345 intensities. The filter redistributes the pixel intensities to
10346 equalize their distribution across the intensity range. It may be
10347 viewed as an "automatically adjusting contrast filter". This filter is
10348 useful only for correcting degraded or poorly captured source
10351 The filter accepts the following options:
10355 Determine the amount of equalization to be applied. As the strength
10356 is reduced, the distribution of pixel intensities more-and-more
10357 approaches that of the input frame. The value must be a float number
10358 in the range [0,1] and defaults to 0.200.
10361 Set the maximum intensity that can generated and scale the output
10362 values appropriately. The strength should be set as desired and then
10363 the intensity can be limited if needed to avoid washing-out. The value
10364 must be a float number in the range [0,1] and defaults to 0.210.
10367 Set the antibanding level. If enabled the filter will randomly vary
10368 the luminance of output pixels by a small amount to avoid banding of
10369 the histogram. Possible values are @code{none}, @code{weak} or
10370 @code{strong}. It defaults to @code{none}.
10375 Compute and draw a color distribution histogram for the input video.
10377 The computed histogram is a representation of the color component
10378 distribution in an image.
10380 Standard histogram displays the color components distribution in an image.
10381 Displays color graph for each color component. Shows distribution of
10382 the Y, U, V, A or R, G, B components, depending on input format, in the
10383 current frame. Below each graph a color component scale meter is shown.
10385 The filter accepts the following options:
10389 Set height of level. Default value is @code{200}.
10390 Allowed range is [50, 2048].
10393 Set height of color scale. Default value is @code{12}.
10394 Allowed range is [0, 40].
10398 It accepts the following values:
10401 Per color component graphs are placed below each other.
10404 Per color component graphs are placed side by side.
10407 Presents information identical to that in the @code{parade}, except
10408 that the graphs representing color components are superimposed directly
10411 Default is @code{stack}.
10414 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10415 Default is @code{linear}.
10418 Set what color components to display.
10419 Default is @code{7}.
10422 Set foreground opacity. Default is @code{0.7}.
10425 Set background opacity. Default is @code{0.5}.
10428 @subsection Examples
10433 Calculate and draw histogram:
10435 ffplay -i input -vf histogram
10443 This is a high precision/quality 3d denoise filter. It aims to reduce
10444 image noise, producing smooth images and making still images really
10445 still. It should enhance compressibility.
10447 It accepts the following optional parameters:
10451 A non-negative floating point number which specifies spatial luma strength.
10452 It defaults to 4.0.
10454 @item chroma_spatial
10455 A non-negative floating point number which specifies spatial chroma strength.
10456 It defaults to 3.0*@var{luma_spatial}/4.0.
10459 A floating point number which specifies luma temporal strength. It defaults to
10460 6.0*@var{luma_spatial}/4.0.
10463 A floating point number which specifies chroma temporal strength. It defaults to
10464 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10467 @section hwdownload
10469 Download hardware frames to system memory.
10471 The input must be in hardware frames, and the output a non-hardware format.
10472 Not all formats will be supported on the output - it may be necessary to insert
10473 an additional @option{format} filter immediately following in the graph to get
10474 the output in a supported format.
10478 Map hardware frames to system memory or to another device.
10480 This filter has several different modes of operation; which one is used depends
10481 on the input and output formats:
10484 Hardware frame input, normal frame output
10486 Map the input frames to system memory and pass them to the output. If the
10487 original hardware frame is later required (for example, after overlaying
10488 something else on part of it), the @option{hwmap} filter can be used again
10489 in the next mode to retrieve it.
10491 Normal frame input, hardware frame output
10493 If the input is actually a software-mapped hardware frame, then unmap it -
10494 that is, return the original hardware frame.
10496 Otherwise, a device must be provided. Create new hardware surfaces on that
10497 device for the output, then map them back to the software format at the input
10498 and give those frames to the preceding filter. This will then act like the
10499 @option{hwupload} filter, but may be able to avoid an additional copy when
10500 the input is already in a compatible format.
10502 Hardware frame input and output
10504 A device must be supplied for the output, either directly or with the
10505 @option{derive_device} option. The input and output devices must be of
10506 different types and compatible - the exact meaning of this is
10507 system-dependent, but typically it means that they must refer to the same
10508 underlying hardware context (for example, refer to the same graphics card).
10510 If the input frames were originally created on the output device, then unmap
10511 to retrieve the original frames.
10513 Otherwise, map the frames to the output device - create new hardware frames
10514 on the output corresponding to the frames on the input.
10517 The following additional parameters are accepted:
10521 Set the frame mapping mode. Some combination of:
10524 The mapped frame should be readable.
10526 The mapped frame should be writeable.
10528 The mapping will always overwrite the entire frame.
10530 This may improve performance in some cases, as the original contents of the
10531 frame need not be loaded.
10533 The mapping must not involve any copying.
10535 Indirect mappings to copies of frames are created in some cases where either
10536 direct mapping is not possible or it would have unexpected properties.
10537 Setting this flag ensures that the mapping is direct and will fail if that is
10540 Defaults to @var{read+write} if not specified.
10542 @item derive_device @var{type}
10543 Rather than using the device supplied at initialisation, instead derive a new
10544 device of type @var{type} from the device the input frames exist on.
10547 In a hardware to hardware mapping, map in reverse - create frames in the sink
10548 and map them back to the source. This may be necessary in some cases where
10549 a mapping in one direction is required but only the opposite direction is
10550 supported by the devices being used.
10552 This option is dangerous - it may break the preceding filter in undefined
10553 ways if there are any additional constraints on that filter's output.
10554 Do not use it without fully understanding the implications of its use.
10559 Upload system memory frames to hardware surfaces.
10561 The device to upload to must be supplied when the filter is initialised. If
10562 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10565 @anchor{hwupload_cuda}
10566 @section hwupload_cuda
10568 Upload system memory frames to a CUDA device.
10570 It accepts the following optional parameters:
10574 The number of the CUDA device to use
10579 Apply a high-quality magnification filter designed for pixel art. This filter
10580 was originally created by Maxim Stepin.
10582 It accepts the following option:
10586 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10587 @code{hq3x} and @code{4} for @code{hq4x}.
10588 Default is @code{3}.
10592 Stack input videos horizontally.
10594 All streams must be of same pixel format and of same height.
10596 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10597 to create same output.
10599 The filter accept the following option:
10603 Set number of input streams. Default is 2.
10606 If set to 1, force the output to terminate when the shortest input
10607 terminates. Default value is 0.
10612 Modify the hue and/or the saturation of the input.
10614 It accepts the following parameters:
10618 Specify the hue angle as a number of degrees. It accepts an expression,
10619 and defaults to "0".
10622 Specify the saturation in the [-10,10] range. It accepts an expression and
10626 Specify the hue angle as a number of radians. It accepts an
10627 expression, and defaults to "0".
10630 Specify the brightness in the [-10,10] range. It accepts an expression and
10634 @option{h} and @option{H} are mutually exclusive, and can't be
10635 specified at the same time.
10637 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10638 expressions containing the following constants:
10642 frame count of the input frame starting from 0
10645 presentation timestamp of the input frame expressed in time base units
10648 frame rate of the input video, NAN if the input frame rate is unknown
10651 timestamp expressed in seconds, NAN if the input timestamp is unknown
10654 time base of the input video
10657 @subsection Examples
10661 Set the hue to 90 degrees and the saturation to 1.0:
10667 Same command but expressing the hue in radians:
10673 Rotate hue and make the saturation swing between 0
10674 and 2 over a period of 1 second:
10676 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10680 Apply a 3 seconds saturation fade-in effect starting at 0:
10682 hue="s=min(t/3\,1)"
10685 The general fade-in expression can be written as:
10687 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10691 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10693 hue="s=max(0\, min(1\, (8-t)/3))"
10696 The general fade-out expression can be written as:
10698 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10703 @subsection Commands
10705 This filter supports the following commands:
10711 Modify the hue and/or the saturation and/or brightness of the input video.
10712 The command accepts the same syntax of the corresponding option.
10714 If the specified expression is not valid, it is kept at its current
10718 @section hysteresis
10720 Grow first stream into second stream by connecting components.
10721 This makes it possible to build more robust edge masks.
10723 This filter accepts the following options:
10727 Set which planes will be processed as bitmap, unprocessed planes will be
10728 copied from first stream.
10729 By default value 0xf, all planes will be processed.
10732 Set threshold which is used in filtering. If pixel component value is higher than
10733 this value filter algorithm for connecting components is activated.
10734 By default value is 0.
10739 Detect video interlacing type.
10741 This filter tries to detect if the input frames are interlaced, progressive,
10742 top or bottom field first. It will also try to detect fields that are
10743 repeated between adjacent frames (a sign of telecine).
10745 Single frame detection considers only immediately adjacent frames when classifying each frame.
10746 Multiple frame detection incorporates the classification history of previous frames.
10748 The filter will log these metadata values:
10751 @item single.current_frame
10752 Detected type of current frame using single-frame detection. One of:
10753 ``tff'' (top field first), ``bff'' (bottom field first),
10754 ``progressive'', or ``undetermined''
10757 Cumulative number of frames detected as top field first using single-frame detection.
10760 Cumulative number of frames detected as top field first using multiple-frame detection.
10763 Cumulative number of frames detected as bottom field first using single-frame detection.
10765 @item multiple.current_frame
10766 Detected type of current frame using multiple-frame detection. One of:
10767 ``tff'' (top field first), ``bff'' (bottom field first),
10768 ``progressive'', or ``undetermined''
10771 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10773 @item single.progressive
10774 Cumulative number of frames detected as progressive using single-frame detection.
10776 @item multiple.progressive
10777 Cumulative number of frames detected as progressive using multiple-frame detection.
10779 @item single.undetermined
10780 Cumulative number of frames that could not be classified using single-frame detection.
10782 @item multiple.undetermined
10783 Cumulative number of frames that could not be classified using multiple-frame detection.
10785 @item repeated.current_frame
10786 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10788 @item repeated.neither
10789 Cumulative number of frames with no repeated field.
10792 Cumulative number of frames with the top field repeated from the previous frame's top field.
10794 @item repeated.bottom
10795 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10798 The filter accepts the following options:
10802 Set interlacing threshold.
10804 Set progressive threshold.
10806 Threshold for repeated field detection.
10808 Number of frames after which a given frame's contribution to the
10809 statistics is halved (i.e., it contributes only 0.5 to its
10810 classification). The default of 0 means that all frames seen are given
10811 full weight of 1.0 forever.
10812 @item analyze_interlaced_flag
10813 When this is not 0 then idet will use the specified number of frames to determine
10814 if the interlaced flag is accurate, it will not count undetermined frames.
10815 If the flag is found to be accurate it will be used without any further
10816 computations, if it is found to be inaccurate it will be cleared without any
10817 further computations. This allows inserting the idet filter as a low computational
10818 method to clean up the interlaced flag
10823 Deinterleave or interleave fields.
10825 This filter allows one to process interlaced images fields without
10826 deinterlacing them. Deinterleaving splits the input frame into 2
10827 fields (so called half pictures). Odd lines are moved to the top
10828 half of the output image, even lines to the bottom half.
10829 You can process (filter) them independently and then re-interleave them.
10831 The filter accepts the following options:
10835 @item chroma_mode, c
10836 @item alpha_mode, a
10837 Available values for @var{luma_mode}, @var{chroma_mode} and
10838 @var{alpha_mode} are:
10844 @item deinterleave, d
10845 Deinterleave fields, placing one above the other.
10847 @item interleave, i
10848 Interleave fields. Reverse the effect of deinterleaving.
10850 Default value is @code{none}.
10852 @item luma_swap, ls
10853 @item chroma_swap, cs
10854 @item alpha_swap, as
10855 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10860 Apply inflate effect to the video.
10862 This filter replaces the pixel by the local(3x3) average by taking into account
10863 only values higher than the pixel.
10865 It accepts the following options:
10872 Limit the maximum change for each plane, default is 65535.
10873 If 0, plane will remain unchanged.
10878 Simple interlacing filter from progressive contents. This interleaves upper (or
10879 lower) lines from odd frames with lower (or upper) lines from even frames,
10880 halving the frame rate and preserving image height.
10883 Original Original New Frame
10884 Frame 'j' Frame 'j+1' (tff)
10885 ========== =========== ==================
10886 Line 0 --------------------> Frame 'j' Line 0
10887 Line 1 Line 1 ----> Frame 'j+1' Line 1
10888 Line 2 ---------------------> Frame 'j' Line 2
10889 Line 3 Line 3 ----> Frame 'j+1' Line 3
10891 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10894 It accepts the following optional parameters:
10898 This determines whether the interlaced frame is taken from the even
10899 (tff - default) or odd (bff) lines of the progressive frame.
10902 Vertical lowpass filter to avoid twitter interlacing and
10903 reduce moire patterns.
10907 Disable vertical lowpass filter
10910 Enable linear filter (default)
10913 Enable complex filter. This will slightly less reduce twitter and moire
10914 but better retain detail and subjective sharpness impression.
10921 Deinterlace input video by applying Donald Graft's adaptive kernel
10922 deinterling. Work on interlaced parts of a video to produce
10923 progressive frames.
10925 The description of the accepted parameters follows.
10929 Set the threshold which affects the filter's tolerance when
10930 determining if a pixel line must be processed. It must be an integer
10931 in the range [0,255] and defaults to 10. A value of 0 will result in
10932 applying the process on every pixels.
10935 Paint pixels exceeding the threshold value to white if set to 1.
10939 Set the fields order. Swap fields if set to 1, leave fields alone if
10943 Enable additional sharpening if set to 1. Default is 0.
10946 Enable twoway sharpening if set to 1. Default is 0.
10949 @subsection Examples
10953 Apply default values:
10955 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10959 Enable additional sharpening:
10965 Paint processed pixels in white:
10971 @section lenscorrection
10973 Correct radial lens distortion
10975 This filter can be used to correct for radial distortion as can result from the use
10976 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10977 one can use tools available for example as part of opencv or simply trial-and-error.
10978 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10979 and extract the k1 and k2 coefficients from the resulting matrix.
10981 Note that effectively the same filter is available in the open-source tools Krita and
10982 Digikam from the KDE project.
10984 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10985 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10986 brightness distribution, so you may want to use both filters together in certain
10987 cases, though you will have to take care of ordering, i.e. whether vignetting should
10988 be applied before or after lens correction.
10990 @subsection Options
10992 The filter accepts the following options:
10996 Relative x-coordinate of the focal point of the image, and thereby the center of the
10997 distortion. This value has a range [0,1] and is expressed as fractions of the image
10998 width. Default is 0.5.
11000 Relative y-coordinate of the focal point of the image, and thereby the center of the
11001 distortion. This value has a range [0,1] and is expressed as fractions of the image
11002 height. Default is 0.5.
11004 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11005 no correction. Default is 0.
11007 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11008 0 means no correction. Default is 0.
11011 The formula that generates the correction is:
11013 @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)
11015 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11016 distances from the focal point in the source and target images, respectively.
11020 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11022 The @code{lensfun} filter requires the camera make, camera model, and lens model
11023 to apply the lens correction. The filter will load the lensfun database and
11024 query it to find the corresponding camera and lens entries in the database. As
11025 long as these entries can be found with the given options, the filter can
11026 perform corrections on frames. Note that incomplete strings will result in the
11027 filter choosing the best match with the given options, and the filter will
11028 output the chosen camera and lens models (logged with level "info"). You must
11029 provide the make, camera model, and lens model as they are required.
11031 The filter accepts the following options:
11035 The make of the camera (for example, "Canon"). This option is required.
11038 The model of the camera (for example, "Canon EOS 100D"). This option is
11042 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11043 option is required.
11046 The type of correction to apply. The following values are valid options:
11050 Enables fixing lens vignetting.
11053 Enables fixing lens geometry. This is the default.
11056 Enables fixing chromatic aberrations.
11059 Enables fixing lens vignetting and lens geometry.
11062 Enables fixing lens vignetting and chromatic aberrations.
11065 Enables fixing both lens geometry and chromatic aberrations.
11068 Enables all possible corrections.
11072 The focal length of the image/video (zoom; expected constant for video). For
11073 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11074 range should be chosen when using that lens. Default 18.
11077 The aperture of the image/video (expected constant for video). Note that
11078 aperture is only used for vignetting correction. Default 3.5.
11080 @item focus_distance
11081 The focus distance of the image/video (expected constant for video). Note that
11082 focus distance is only used for vignetting and only slightly affects the
11083 vignetting correction process. If unknown, leave it at the default value (which
11086 @item target_geometry
11087 The target geometry of the output image/video. The following values are valid
11091 @item rectilinear (default)
11094 @item equirectangular
11095 @item fisheye_orthographic
11096 @item fisheye_stereographic
11097 @item fisheye_equisolid
11098 @item fisheye_thoby
11101 Apply the reverse of image correction (instead of correcting distortion, apply
11104 @item interpolation
11105 The type of interpolation used when correcting distortion. The following values
11110 @item linear (default)
11115 @subsection Examples
11119 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11120 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11124 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
11128 Apply the same as before, but only for the first 5 seconds of video.
11131 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
11138 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11139 score between two input videos.
11141 The obtained VMAF score is printed through the logging system.
11143 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11144 After installing the library it can be enabled using:
11145 @code{./configure --enable-libvmaf --enable-version3}.
11146 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11148 The filter has following options:
11152 Set the model path which is to be used for SVM.
11153 Default value: @code{"vmaf_v0.6.1.pkl"}
11156 Set the file path to be used to store logs.
11159 Set the format of the log file (xml or json).
11161 @item enable_transform
11162 Enables transform for computing vmaf.
11165 Invokes the phone model which will generate VMAF scores higher than in the
11166 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11169 Enables computing psnr along with vmaf.
11172 Enables computing ssim along with vmaf.
11175 Enables computing ms_ssim along with vmaf.
11178 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11181 Set number of threads to be used when computing vmaf.
11184 Set interval for frame subsampling used when computing vmaf.
11186 @item enable_conf_interval
11187 Enables confidence interval.
11190 This filter also supports the @ref{framesync} options.
11192 On the below examples the input file @file{main.mpg} being processed is
11193 compared with the reference file @file{ref.mpg}.
11196 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11199 Example with options:
11201 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
11206 Limits the pixel components values to the specified range [min, max].
11208 The filter accepts the following options:
11212 Lower bound. Defaults to the lowest allowed value for the input.
11215 Upper bound. Defaults to the highest allowed value for the input.
11218 Specify which planes will be processed. Defaults to all available.
11225 The filter accepts the following options:
11229 Set the number of loops. Setting this value to -1 will result in infinite loops.
11233 Set maximal size in number of frames. Default is 0.
11236 Set first frame of loop. Default is 0.
11241 Apply a 1D LUT to an input video.
11243 The filter accepts the following options:
11247 Set the 1D LUT file name.
11249 Currently supported formats:
11256 Select interpolation mode.
11258 Available values are:
11262 Use values from the nearest defined point.
11264 Interpolate values using the linear interpolation.
11266 Interpolate values using the cubic interpolation.
11273 Apply a 3D LUT to an input video.
11275 The filter accepts the following options:
11279 Set the 3D LUT file name.
11281 Currently supported formats:
11293 Select interpolation mode.
11295 Available values are:
11299 Use values from the nearest defined point.
11301 Interpolate values using the 8 points defining a cube.
11303 Interpolate values using a tetrahedron.
11307 This filter also supports the @ref{framesync} options.
11311 Turn certain luma values into transparency.
11313 The filter accepts the following options:
11317 Set the luma which will be used as base for transparency.
11318 Default value is @code{0}.
11321 Set the range of luma values to be keyed out.
11322 Default value is @code{0}.
11325 Set the range of softness. Default value is @code{0}.
11326 Use this to control gradual transition from zero to full transparency.
11329 @section lut, lutrgb, lutyuv
11331 Compute a look-up table for binding each pixel component input value
11332 to an output value, and apply it to the input video.
11334 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11335 to an RGB input video.
11337 These filters accept the following parameters:
11340 set first pixel component expression
11342 set second pixel component expression
11344 set third pixel component expression
11346 set fourth pixel component expression, corresponds to the alpha component
11349 set red component expression
11351 set green component expression
11353 set blue component expression
11355 alpha component expression
11358 set Y/luminance component expression
11360 set U/Cb component expression
11362 set V/Cr component expression
11365 Each of them specifies the expression to use for computing the lookup table for
11366 the corresponding pixel component values.
11368 The exact component associated to each of the @var{c*} options depends on the
11371 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11372 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11374 The expressions can contain the following constants and functions:
11379 The input width and height.
11382 The input value for the pixel component.
11385 The input value, clipped to the @var{minval}-@var{maxval} range.
11388 The maximum value for the pixel component.
11391 The minimum value for the pixel component.
11394 The negated value for the pixel component value, clipped to the
11395 @var{minval}-@var{maxval} range; it corresponds to the expression
11396 "maxval-clipval+minval".
11399 The computed value in @var{val}, clipped to the
11400 @var{minval}-@var{maxval} range.
11402 @item gammaval(gamma)
11403 The computed gamma correction value of the pixel component value,
11404 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11406 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11410 All expressions default to "val".
11412 @subsection Examples
11416 Negate input video:
11418 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11419 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11422 The above is the same as:
11424 lutrgb="r=negval:g=negval:b=negval"
11425 lutyuv="y=negval:u=negval:v=negval"
11435 Remove chroma components, turning the video into a graytone image:
11437 lutyuv="u=128:v=128"
11441 Apply a luma burning effect:
11447 Remove green and blue components:
11453 Set a constant alpha channel value on input:
11455 format=rgba,lutrgb=a="maxval-minval/2"
11459 Correct luminance gamma by a factor of 0.5:
11461 lutyuv=y=gammaval(0.5)
11465 Discard least significant bits of luma:
11467 lutyuv=y='bitand(val, 128+64+32)'
11471 Technicolor like effect:
11473 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11477 @section lut2, tlut2
11479 The @code{lut2} filter takes two input streams and outputs one
11482 The @code{tlut2} (time lut2) filter takes two consecutive frames
11483 from one single stream.
11485 This filter accepts the following parameters:
11488 set first pixel component expression
11490 set second pixel component expression
11492 set third pixel component expression
11494 set fourth pixel component expression, corresponds to the alpha component
11497 Each of them specifies the expression to use for computing the lookup table for
11498 the corresponding pixel component values.
11500 The exact component associated to each of the @var{c*} options depends on the
11503 The expressions can contain the following constants:
11508 The input width and height.
11511 The first input value for the pixel component.
11514 The second input value for the pixel component.
11517 The first input video bit depth.
11520 The second input video bit depth.
11523 All expressions default to "x".
11525 @subsection Examples
11529 Highlight differences between two RGB video streams:
11531 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)'
11535 Highlight differences between two YUV video streams:
11537 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)'
11541 Show max difference between two video streams:
11543 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)))'
11547 @section maskedclamp
11549 Clamp the first input stream with the second input and third input stream.
11551 Returns the value of first stream to be between second input
11552 stream - @code{undershoot} and third input stream + @code{overshoot}.
11554 This filter accepts the following options:
11557 Default value is @code{0}.
11560 Default value is @code{0}.
11563 Set which planes will be processed as bitmap, unprocessed planes will be
11564 copied from first stream.
11565 By default value 0xf, all planes will be processed.
11568 @section maskedmerge
11570 Merge the first input stream with the second input stream using per pixel
11571 weights in the third input stream.
11573 A value of 0 in the third stream pixel component means that pixel component
11574 from first stream is returned unchanged, while maximum value (eg. 255 for
11575 8-bit videos) means that pixel component from second stream is returned
11576 unchanged. Intermediate values define the amount of merging between both
11577 input stream's pixel components.
11579 This filter accepts the following options:
11582 Set which planes will be processed as bitmap, unprocessed planes will be
11583 copied from first stream.
11584 By default value 0xf, all planes will be processed.
11589 Apply motion-compensation deinterlacing.
11591 It needs one field per frame as input and must thus be used together
11592 with yadif=1/3 or equivalent.
11594 This filter accepts the following options:
11597 Set the deinterlacing mode.
11599 It accepts one of the following values:
11604 use iterative motion estimation
11606 like @samp{slow}, but use multiple reference frames.
11608 Default value is @samp{fast}.
11611 Set the picture field parity assumed for the input video. It must be
11612 one of the following values:
11616 assume top field first
11618 assume bottom field first
11621 Default value is @samp{bff}.
11624 Set per-block quantization parameter (QP) used by the internal
11627 Higher values should result in a smoother motion vector field but less
11628 optimal individual vectors. Default value is 1.
11631 @section mergeplanes
11633 Merge color channel components from several video streams.
11635 The filter accepts up to 4 input streams, and merge selected input
11636 planes to the output video.
11638 This filter accepts the following options:
11641 Set input to output plane mapping. Default is @code{0}.
11643 The mappings is specified as a bitmap. It should be specified as a
11644 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
11645 mapping for the first plane of the output stream. 'A' sets the number of
11646 the input stream to use (from 0 to 3), and 'a' the plane number of the
11647 corresponding input to use (from 0 to 3). The rest of the mappings is
11648 similar, 'Bb' describes the mapping for the output stream second
11649 plane, 'Cc' describes the mapping for the output stream third plane and
11650 'Dd' describes the mapping for the output stream fourth plane.
11653 Set output pixel format. Default is @code{yuva444p}.
11656 @subsection Examples
11660 Merge three gray video streams of same width and height into single video stream:
11662 [a0][a1][a2]mergeplanes=0x001020:yuv444p
11666 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
11668 [a0][a1]mergeplanes=0x00010210:yuva444p
11672 Swap Y and A plane in yuva444p stream:
11674 format=yuva444p,mergeplanes=0x03010200:yuva444p
11678 Swap U and V plane in yuv420p stream:
11680 format=yuv420p,mergeplanes=0x000201:yuv420p
11684 Cast a rgb24 clip to yuv444p:
11686 format=rgb24,mergeplanes=0x000102:yuv444p
11692 Estimate and export motion vectors using block matching algorithms.
11693 Motion vectors are stored in frame side data to be used by other filters.
11695 This filter accepts the following options:
11698 Specify the motion estimation method. Accepts one of the following values:
11702 Exhaustive search algorithm.
11704 Three step search algorithm.
11706 Two dimensional logarithmic search algorithm.
11708 New three step search algorithm.
11710 Four step search algorithm.
11712 Diamond search algorithm.
11714 Hexagon-based search algorithm.
11716 Enhanced predictive zonal search algorithm.
11718 Uneven multi-hexagon search algorithm.
11720 Default value is @samp{esa}.
11723 Macroblock size. Default @code{16}.
11726 Search parameter. Default @code{7}.
11729 @section midequalizer
11731 Apply Midway Image Equalization effect using two video streams.
11733 Midway Image Equalization adjusts a pair of images to have the same
11734 histogram, while maintaining their dynamics as much as possible. It's
11735 useful for e.g. matching exposures from a pair of stereo cameras.
11737 This filter has two inputs and one output, which must be of same pixel format, but
11738 may be of different sizes. The output of filter is first input adjusted with
11739 midway histogram of both inputs.
11741 This filter accepts the following option:
11745 Set which planes to process. Default is @code{15}, which is all available planes.
11748 @section minterpolate
11750 Convert the video to specified frame rate using motion interpolation.
11752 This filter accepts the following options:
11755 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}.
11758 Motion interpolation mode. Following values are accepted:
11761 Duplicate previous or next frame for interpolating new ones.
11763 Blend source frames. Interpolated frame is mean of previous and next frames.
11765 Motion compensated interpolation. Following options are effective when this mode is selected:
11769 Motion compensation mode. Following values are accepted:
11772 Overlapped block motion compensation.
11774 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
11776 Default mode is @samp{obmc}.
11779 Motion estimation mode. Following values are accepted:
11782 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
11784 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11786 Default mode is @samp{bilat}.
11789 The algorithm to be used for motion estimation. Following values are accepted:
11792 Exhaustive search algorithm.
11794 Three step search algorithm.
11796 Two dimensional logarithmic search algorithm.
11798 New three step search algorithm.
11800 Four step search algorithm.
11802 Diamond search algorithm.
11804 Hexagon-based search algorithm.
11806 Enhanced predictive zonal search algorithm.
11808 Uneven multi-hexagon search algorithm.
11810 Default algorithm is @samp{epzs}.
11813 Macroblock size. Default @code{16}.
11816 Motion estimation search parameter. Default @code{32}.
11819 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).
11824 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:
11827 Disable scene change detection.
11829 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11831 Default method is @samp{fdiff}.
11833 @item scd_threshold
11834 Scene change detection threshold. Default is @code{5.0}.
11839 Mix several video input streams into one video stream.
11841 A description of the accepted options follows.
11845 The number of inputs. If unspecified, it defaults to 2.
11848 Specify weight of each input video stream as sequence.
11849 Each weight is separated by space. If number of weights
11850 is smaller than number of @var{frames} last specified
11851 weight will be used for all remaining unset weights.
11854 Specify scale, if it is set it will be multiplied with sum
11855 of each weight multiplied with pixel values to give final destination
11856 pixel value. By default @var{scale} is auto scaled to sum of weights.
11859 Specify how end of stream is determined.
11862 The duration of the longest input. (default)
11865 The duration of the shortest input.
11868 The duration of the first input.
11872 @section mpdecimate
11874 Drop frames that do not differ greatly from the previous frame in
11875 order to reduce frame rate.
11877 The main use of this filter is for very-low-bitrate encoding
11878 (e.g. streaming over dialup modem), but it could in theory be used for
11879 fixing movies that were inverse-telecined incorrectly.
11881 A description of the accepted options follows.
11885 Set the maximum number of consecutive frames which can be dropped (if
11886 positive), or the minimum interval between dropped frames (if
11887 negative). If the value is 0, the frame is dropped disregarding the
11888 number of previous sequentially dropped frames.
11890 Default value is 0.
11895 Set the dropping threshold values.
11897 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11898 represent actual pixel value differences, so a threshold of 64
11899 corresponds to 1 unit of difference for each pixel, or the same spread
11900 out differently over the block.
11902 A frame is a candidate for dropping if no 8x8 blocks differ by more
11903 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11904 meaning the whole image) differ by more than a threshold of @option{lo}.
11906 Default value for @option{hi} is 64*12, default value for @option{lo} is
11907 64*5, and default value for @option{frac} is 0.33.
11913 Negate (invert) the input video.
11915 It accepts the following option:
11920 With value 1, it negates the alpha component, if present. Default value is 0.
11926 Denoise frames using Non-Local Means algorithm.
11928 Each pixel is adjusted by looking for other pixels with similar contexts. This
11929 context similarity is defined by comparing their surrounding patches of size
11930 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11933 Note that the research area defines centers for patches, which means some
11934 patches will be made of pixels outside that research area.
11936 The filter accepts the following options.
11940 Set denoising strength.
11946 Same as @option{p} but for chroma planes.
11948 The default value is @var{0} and means automatic.
11954 Same as @option{r} but for chroma planes.
11956 The default value is @var{0} and means automatic.
11961 Deinterlace video using neural network edge directed interpolation.
11963 This filter accepts the following options:
11967 Mandatory option, without binary file filter can not work.
11968 Currently file can be found here:
11969 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11972 Set which frames to deinterlace, by default it is @code{all}.
11973 Can be @code{all} or @code{interlaced}.
11976 Set mode of operation.
11978 Can be one of the following:
11982 Use frame flags, both fields.
11984 Use frame flags, single field.
11986 Use top field only.
11988 Use bottom field only.
11990 Use both fields, top first.
11992 Use both fields, bottom first.
11996 Set which planes to process, by default filter process all frames.
11999 Set size of local neighborhood around each pixel, used by the predictor neural
12002 Can be one of the following:
12015 Set the number of neurons in predictor neural network.
12016 Can be one of the following:
12027 Controls the number of different neural network predictions that are blended
12028 together to compute the final output value. Can be @code{fast}, default or
12032 Set which set of weights to use in the predictor.
12033 Can be one of the following:
12037 weights trained to minimize absolute error
12039 weights trained to minimize squared error
12043 Controls whether or not the prescreener neural network is used to decide
12044 which pixels should be processed by the predictor neural network and which
12045 can be handled by simple cubic interpolation.
12046 The prescreener is trained to know whether cubic interpolation will be
12047 sufficient for a pixel or whether it should be predicted by the predictor nn.
12048 The computational complexity of the prescreener nn is much less than that of
12049 the predictor nn. Since most pixels can be handled by cubic interpolation,
12050 using the prescreener generally results in much faster processing.
12051 The prescreener is pretty accurate, so the difference between using it and not
12052 using it is almost always unnoticeable.
12054 Can be one of the following:
12062 Default is @code{new}.
12065 Set various debugging flags.
12070 Force libavfilter not to use any of the specified pixel formats for the
12071 input to the next filter.
12073 It accepts the following parameters:
12077 A '|'-separated list of pixel format names, such as
12078 pix_fmts=yuv420p|monow|rgb24".
12082 @subsection Examples
12086 Force libavfilter to use a format different from @var{yuv420p} for the
12087 input to the vflip filter:
12089 noformat=pix_fmts=yuv420p,vflip
12093 Convert the input video to any of the formats not contained in the list:
12095 noformat=yuv420p|yuv444p|yuv410p
12101 Add noise on video input frame.
12103 The filter accepts the following options:
12111 Set noise seed for specific pixel component or all pixel components in case
12112 of @var{all_seed}. Default value is @code{123457}.
12114 @item all_strength, alls
12115 @item c0_strength, c0s
12116 @item c1_strength, c1s
12117 @item c2_strength, c2s
12118 @item c3_strength, c3s
12119 Set noise strength for specific pixel component or all pixel components in case
12120 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12122 @item all_flags, allf
12123 @item c0_flags, c0f
12124 @item c1_flags, c1f
12125 @item c2_flags, c2f
12126 @item c3_flags, c3f
12127 Set pixel component flags or set flags for all components if @var{all_flags}.
12128 Available values for component flags are:
12131 averaged temporal noise (smoother)
12133 mix random noise with a (semi)regular pattern
12135 temporal noise (noise pattern changes between frames)
12137 uniform noise (gaussian otherwise)
12141 @subsection Examples
12143 Add temporal and uniform noise to input video:
12145 noise=alls=20:allf=t+u
12150 Normalize RGB video (aka histogram stretching, contrast stretching).
12151 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12153 For each channel of each frame, the filter computes the input range and maps
12154 it linearly to the user-specified output range. The output range defaults
12155 to the full dynamic range from pure black to pure white.
12157 Temporal smoothing can be used on the input range to reduce flickering (rapid
12158 changes in brightness) caused when small dark or bright objects enter or leave
12159 the scene. This is similar to the auto-exposure (automatic gain control) on a
12160 video camera, and, like a video camera, it may cause a period of over- or
12161 under-exposure of the video.
12163 The R,G,B channels can be normalized independently, which may cause some
12164 color shifting, or linked together as a single channel, which prevents
12165 color shifting. Linked normalization preserves hue. Independent normalization
12166 does not, so it can be used to remove some color casts. Independent and linked
12167 normalization can be combined in any ratio.
12169 The normalize filter accepts the following options:
12174 Colors which define the output range. The minimum input value is mapped to
12175 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12176 The defaults are black and white respectively. Specifying white for
12177 @var{blackpt} and black for @var{whitept} will give color-inverted,
12178 normalized video. Shades of grey can be used to reduce the dynamic range
12179 (contrast). Specifying saturated colors here can create some interesting
12183 The number of previous frames to use for temporal smoothing. The input range
12184 of each channel is smoothed using a rolling average over the current frame
12185 and the @var{smoothing} previous frames. The default is 0 (no temporal
12189 Controls the ratio of independent (color shifting) channel normalization to
12190 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12191 independent. Defaults to 1.0 (fully independent).
12194 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12195 expensive no-op. Defaults to 1.0 (full strength).
12199 @subsection Examples
12201 Stretch video contrast to use the full dynamic range, with no temporal
12202 smoothing; may flicker depending on the source content:
12204 normalize=blackpt=black:whitept=white:smoothing=0
12207 As above, but with 50 frames of temporal smoothing; flicker should be
12208 reduced, depending on the source content:
12210 normalize=blackpt=black:whitept=white:smoothing=50
12213 As above, but with hue-preserving linked channel normalization:
12215 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12218 As above, but with half strength:
12220 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12223 Map the darkest input color to red, the brightest input color to cyan:
12225 normalize=blackpt=red:whitept=cyan
12230 Pass the video source unchanged to the output.
12233 Optical Character Recognition
12235 This filter uses Tesseract for optical character recognition. To enable
12236 compilation of this filter, you need to configure FFmpeg with
12237 @code{--enable-libtesseract}.
12239 It accepts the following options:
12243 Set datapath to tesseract data. Default is to use whatever was
12244 set at installation.
12247 Set language, default is "eng".
12250 Set character whitelist.
12253 Set character blacklist.
12256 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12260 Apply a video transform using libopencv.
12262 To enable this filter, install the libopencv library and headers and
12263 configure FFmpeg with @code{--enable-libopencv}.
12265 It accepts the following parameters:
12270 The name of the libopencv filter to apply.
12272 @item filter_params
12273 The parameters to pass to the libopencv filter. If not specified, the default
12274 values are assumed.
12278 Refer to the official libopencv documentation for more precise
12280 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12282 Several libopencv filters are supported; see the following subsections.
12287 Dilate an image by using a specific structuring element.
12288 It corresponds to the libopencv function @code{cvDilate}.
12290 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12292 @var{struct_el} represents a structuring element, and has the syntax:
12293 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12295 @var{cols} and @var{rows} represent the number of columns and rows of
12296 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12297 point, and @var{shape} the shape for the structuring element. @var{shape}
12298 must be "rect", "cross", "ellipse", or "custom".
12300 If the value for @var{shape} is "custom", it must be followed by a
12301 string of the form "=@var{filename}". The file with name
12302 @var{filename} is assumed to represent a binary image, with each
12303 printable character corresponding to a bright pixel. When a custom
12304 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12305 or columns and rows of the read file are assumed instead.
12307 The default value for @var{struct_el} is "3x3+0x0/rect".
12309 @var{nb_iterations} specifies the number of times the transform is
12310 applied to the image, and defaults to 1.
12314 # Use the default values
12317 # Dilate using a structuring element with a 5x5 cross, iterating two times
12318 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12320 # Read the shape from the file diamond.shape, iterating two times.
12321 # The file diamond.shape may contain a pattern of characters like this
12327 # The specified columns and rows are ignored
12328 # but the anchor point coordinates are not
12329 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12334 Erode an image by using a specific structuring element.
12335 It corresponds to the libopencv function @code{cvErode}.
12337 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12338 with the same syntax and semantics as the @ref{dilate} filter.
12342 Smooth the input video.
12344 The filter takes the following parameters:
12345 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12347 @var{type} is the type of smooth filter to apply, and must be one of
12348 the following values: "blur", "blur_no_scale", "median", "gaussian",
12349 or "bilateral". The default value is "gaussian".
12351 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12352 depend on the smooth type. @var{param1} and
12353 @var{param2} accept integer positive values or 0. @var{param3} and
12354 @var{param4} accept floating point values.
12356 The default value for @var{param1} is 3. The default value for the
12357 other parameters is 0.
12359 These parameters correspond to the parameters assigned to the
12360 libopencv function @code{cvSmooth}.
12362 @section oscilloscope
12364 2D Video Oscilloscope.
12366 Useful to measure spatial impulse, step responses, chroma delays, etc.
12368 It accepts the following parameters:
12372 Set scope center x position.
12375 Set scope center y position.
12378 Set scope size, relative to frame diagonal.
12381 Set scope tilt/rotation.
12387 Set trace center x position.
12390 Set trace center y position.
12393 Set trace width, relative to width of frame.
12396 Set trace height, relative to height of frame.
12399 Set which components to trace. By default it traces first three components.
12402 Draw trace grid. By default is enabled.
12405 Draw some statistics. By default is enabled.
12408 Draw scope. By default is enabled.
12411 @subsection Examples
12415 Inspect full first row of video frame.
12417 oscilloscope=x=0.5:y=0:s=1
12421 Inspect full last row of video frame.
12423 oscilloscope=x=0.5:y=1:s=1
12427 Inspect full 5th line of video frame of height 1080.
12429 oscilloscope=x=0.5:y=5/1080:s=1
12433 Inspect full last column of video frame.
12435 oscilloscope=x=1:y=0.5:s=1:t=1
12443 Overlay one video on top of another.
12445 It takes two inputs and has one output. The first input is the "main"
12446 video on which the second input is overlaid.
12448 It accepts the following parameters:
12450 A description of the accepted options follows.
12455 Set the expression for the x and y coordinates of the overlaid video
12456 on the main video. Default value is "0" for both expressions. In case
12457 the expression is invalid, it is set to a huge value (meaning that the
12458 overlay will not be displayed within the output visible area).
12461 See @ref{framesync}.
12464 Set when the expressions for @option{x}, and @option{y} are evaluated.
12466 It accepts the following values:
12469 only evaluate expressions once during the filter initialization or
12470 when a command is processed
12473 evaluate expressions for each incoming frame
12476 Default value is @samp{frame}.
12479 See @ref{framesync}.
12482 Set the format for the output video.
12484 It accepts the following values:
12487 force YUV420 output
12490 force YUV422 output
12493 force YUV444 output
12496 force packed RGB output
12499 force planar RGB output
12502 automatically pick format
12505 Default value is @samp{yuv420}.
12508 See @ref{framesync}.
12511 Set format of alpha of the overlaid video, it can be @var{straight} or
12512 @var{premultiplied}. Default is @var{straight}.
12515 The @option{x}, and @option{y} expressions can contain the following
12521 The main input width and height.
12525 The overlay input width and height.
12529 The computed values for @var{x} and @var{y}. They are evaluated for
12534 horizontal and vertical chroma subsample values of the output
12535 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12539 the number of input frame, starting from 0
12542 the position in the file of the input frame, NAN if unknown
12545 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12549 This filter also supports the @ref{framesync} options.
12551 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12552 when evaluation is done @emph{per frame}, and will evaluate to NAN
12553 when @option{eval} is set to @samp{init}.
12555 Be aware that frames are taken from each input video in timestamp
12556 order, hence, if their initial timestamps differ, it is a good idea
12557 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12558 have them begin in the same zero timestamp, as the example for
12559 the @var{movie} filter does.
12561 You can chain together more overlays but you should test the
12562 efficiency of such approach.
12564 @subsection Commands
12566 This filter supports the following commands:
12570 Modify the x and y of the overlay input.
12571 The command accepts the same syntax of the corresponding option.
12573 If the specified expression is not valid, it is kept at its current
12577 @subsection Examples
12581 Draw the overlay at 10 pixels from the bottom right corner of the main
12584 overlay=main_w-overlay_w-10:main_h-overlay_h-10
12587 Using named options the example above becomes:
12589 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
12593 Insert a transparent PNG logo in the bottom left corner of the input,
12594 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12596 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12600 Insert 2 different transparent PNG logos (second logo on bottom
12601 right corner) using the @command{ffmpeg} tool:
12603 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
12607 Add a transparent color layer on top of the main video; @code{WxH}
12608 must specify the size of the main input to the overlay filter:
12610 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
12614 Play an original video and a filtered version (here with the deshake
12615 filter) side by side using the @command{ffplay} tool:
12617 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
12620 The above command is the same as:
12622 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
12626 Make a sliding overlay appearing from the left to the right top part of the
12627 screen starting since time 2:
12629 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
12633 Compose output by putting two input videos side to side:
12635 ffmpeg -i left.avi -i right.avi -filter_complex "
12636 nullsrc=size=200x100 [background];
12637 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
12638 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
12639 [background][left] overlay=shortest=1 [background+left];
12640 [background+left][right] overlay=shortest=1:x=100 [left+right]
12645 Mask 10-20 seconds of a video by applying the delogo filter to a section
12647 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
12648 -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]'
12653 Chain several overlays in cascade:
12655 nullsrc=s=200x200 [bg];
12656 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
12657 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
12658 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
12659 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
12660 [in3] null, [mid2] overlay=100:100 [out0]
12667 Apply Overcomplete Wavelet denoiser.
12669 The filter accepts the following options:
12675 Larger depth values will denoise lower frequency components more, but
12676 slow down filtering.
12678 Must be an int in the range 8-16, default is @code{8}.
12680 @item luma_strength, ls
12683 Must be a double value in the range 0-1000, default is @code{1.0}.
12685 @item chroma_strength, cs
12686 Set chroma strength.
12688 Must be a double value in the range 0-1000, default is @code{1.0}.
12694 Add paddings to the input image, and place the original input at the
12695 provided @var{x}, @var{y} coordinates.
12697 It accepts the following parameters:
12702 Specify an expression for the size of the output image with the
12703 paddings added. If the value for @var{width} or @var{height} is 0, the
12704 corresponding input size is used for the output.
12706 The @var{width} expression can reference the value set by the
12707 @var{height} expression, and vice versa.
12709 The default value of @var{width} and @var{height} is 0.
12713 Specify the offsets to place the input image at within the padded area,
12714 with respect to the top/left border of the output image.
12716 The @var{x} expression can reference the value set by the @var{y}
12717 expression, and vice versa.
12719 The default value of @var{x} and @var{y} is 0.
12721 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
12722 so the input image is centered on the padded area.
12725 Specify the color of the padded area. For the syntax of this option,
12726 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
12727 manual,ffmpeg-utils}.
12729 The default value of @var{color} is "black".
12732 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
12734 It accepts the following values:
12738 Only evaluate expressions once during the filter initialization or when
12739 a command is processed.
12742 Evaluate expressions for each incoming frame.
12746 Default value is @samp{init}.
12749 Pad to aspect instead to a resolution.
12753 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
12754 options are expressions containing the following constants:
12759 The input video width and height.
12763 These are the same as @var{in_w} and @var{in_h}.
12767 The output width and height (the size of the padded area), as
12768 specified by the @var{width} and @var{height} expressions.
12772 These are the same as @var{out_w} and @var{out_h}.
12776 The x and y offsets as specified by the @var{x} and @var{y}
12777 expressions, or NAN if not yet specified.
12780 same as @var{iw} / @var{ih}
12783 input sample aspect ratio
12786 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
12790 The horizontal and vertical chroma subsample values. For example for the
12791 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12794 @subsection Examples
12798 Add paddings with the color "violet" to the input video. The output video
12799 size is 640x480, and the top-left corner of the input video is placed at
12802 pad=640:480:0:40:violet
12805 The example above is equivalent to the following command:
12807 pad=width=640:height=480:x=0:y=40:color=violet
12811 Pad the input to get an output with dimensions increased by 3/2,
12812 and put the input video at the center of the padded area:
12814 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12818 Pad the input to get a squared output with size equal to the maximum
12819 value between the input width and height, and put the input video at
12820 the center of the padded area:
12822 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12826 Pad the input to get a final w/h ratio of 16:9:
12828 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12832 In case of anamorphic video, in order to set the output display aspect
12833 correctly, it is necessary to use @var{sar} in the expression,
12834 according to the relation:
12836 (ih * X / ih) * sar = output_dar
12837 X = output_dar / sar
12840 Thus the previous example needs to be modified to:
12842 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12846 Double the output size and put the input video in the bottom-right
12847 corner of the output padded area:
12849 pad="2*iw:2*ih:ow-iw:oh-ih"
12853 @anchor{palettegen}
12854 @section palettegen
12856 Generate one palette for a whole video stream.
12858 It accepts the following options:
12862 Set the maximum number of colors to quantize in the palette.
12863 Note: the palette will still contain 256 colors; the unused palette entries
12866 @item reserve_transparent
12867 Create a palette of 255 colors maximum and reserve the last one for
12868 transparency. Reserving the transparency color is useful for GIF optimization.
12869 If not set, the maximum of colors in the palette will be 256. You probably want
12870 to disable this option for a standalone image.
12873 @item transparency_color
12874 Set the color that will be used as background for transparency.
12877 Set statistics mode.
12879 It accepts the following values:
12882 Compute full frame histograms.
12884 Compute histograms only for the part that differs from previous frame. This
12885 might be relevant to give more importance to the moving part of your input if
12886 the background is static.
12888 Compute new histogram for each frame.
12891 Default value is @var{full}.
12894 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12895 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12896 color quantization of the palette. This information is also visible at
12897 @var{info} logging level.
12899 @subsection Examples
12903 Generate a representative palette of a given video using @command{ffmpeg}:
12905 ffmpeg -i input.mkv -vf palettegen palette.png
12909 @section paletteuse
12911 Use a palette to downsample an input video stream.
12913 The filter takes two inputs: one video stream and a palette. The palette must
12914 be a 256 pixels image.
12916 It accepts the following options:
12920 Select dithering mode. Available algorithms are:
12923 Ordered 8x8 bayer dithering (deterministic)
12925 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12926 Note: this dithering is sometimes considered "wrong" and is included as a
12928 @item floyd_steinberg
12929 Floyd and Steingberg dithering (error diffusion)
12931 Frankie Sierra dithering v2 (error diffusion)
12933 Frankie Sierra dithering v2 "Lite" (error diffusion)
12936 Default is @var{sierra2_4a}.
12939 When @var{bayer} dithering is selected, this option defines the scale of the
12940 pattern (how much the crosshatch pattern is visible). A low value means more
12941 visible pattern for less banding, and higher value means less visible pattern
12942 at the cost of more banding.
12944 The option must be an integer value in the range [0,5]. Default is @var{2}.
12947 If set, define the zone to process
12951 Only the changing rectangle will be reprocessed. This is similar to GIF
12952 cropping/offsetting compression mechanism. This option can be useful for speed
12953 if only a part of the image is changing, and has use cases such as limiting the
12954 scope of the error diffusal @option{dither} to the rectangle that bounds the
12955 moving scene (it leads to more deterministic output if the scene doesn't change
12956 much, and as a result less moving noise and better GIF compression).
12959 Default is @var{none}.
12962 Take new palette for each output frame.
12964 @item alpha_threshold
12965 Sets the alpha threshold for transparency. Alpha values above this threshold
12966 will be treated as completely opaque, and values below this threshold will be
12967 treated as completely transparent.
12969 The option must be an integer value in the range [0,255]. Default is @var{128}.
12972 @subsection Examples
12976 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12977 using @command{ffmpeg}:
12979 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12983 @section perspective
12985 Correct perspective of video not recorded perpendicular to the screen.
12987 A description of the accepted parameters follows.
12998 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12999 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13000 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13001 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13002 then the corners of the source will be sent to the specified coordinates.
13004 The expressions can use the following variables:
13009 the width and height of video frame.
13013 Output frame count.
13016 @item interpolation
13017 Set interpolation for perspective correction.
13019 It accepts the following values:
13025 Default value is @samp{linear}.
13028 Set interpretation of coordinate options.
13030 It accepts the following values:
13034 Send point in the source specified by the given coordinates to
13035 the corners of the destination.
13037 @item 1, destination
13039 Send the corners of the source to the point in the destination specified
13040 by the given coordinates.
13042 Default value is @samp{source}.
13046 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13048 It accepts the following values:
13051 only evaluate expressions once during the filter initialization or
13052 when a command is processed
13055 evaluate expressions for each incoming frame
13058 Default value is @samp{init}.
13063 Delay interlaced video by one field time so that the field order changes.
13065 The intended use is to fix PAL movies that have been captured with the
13066 opposite field order to the film-to-video transfer.
13068 A description of the accepted parameters follows.
13074 It accepts the following values:
13077 Capture field order top-first, transfer bottom-first.
13078 Filter will delay the bottom field.
13081 Capture field order bottom-first, transfer top-first.
13082 Filter will delay the top field.
13085 Capture and transfer with the same field order. This mode only exists
13086 for the documentation of the other options to refer to, but if you
13087 actually select it, the filter will faithfully do nothing.
13090 Capture field order determined automatically by field flags, transfer
13092 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13093 basis using field flags. If no field information is available,
13094 then this works just like @samp{u}.
13097 Capture unknown or varying, transfer opposite.
13098 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13099 analyzing the images and selecting the alternative that produces best
13100 match between the fields.
13103 Capture top-first, transfer unknown or varying.
13104 Filter selects among @samp{t} and @samp{p} using image analysis.
13107 Capture bottom-first, transfer unknown or varying.
13108 Filter selects among @samp{b} and @samp{p} using image analysis.
13111 Capture determined by field flags, transfer unknown or varying.
13112 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13113 image analysis. If no field information is available, then this works just
13114 like @samp{U}. This is the default mode.
13117 Both capture and transfer unknown or varying.
13118 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13122 @section pixdesctest
13124 Pixel format descriptor test filter, mainly useful for internal
13125 testing. The output video should be equal to the input video.
13129 format=monow, pixdesctest
13132 can be used to test the monowhite pixel format descriptor definition.
13136 Display sample values of color channels. Mainly useful for checking color
13137 and levels. Minimum supported resolution is 640x480.
13139 The filters accept the following options:
13143 Set scope X position, relative offset on X axis.
13146 Set scope Y position, relative offset on Y axis.
13155 Set window opacity. This window also holds statistics about pixel area.
13158 Set window X position, relative offset on X axis.
13161 Set window Y position, relative offset on Y axis.
13166 Enable the specified chain of postprocessing subfilters using libpostproc. This
13167 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13168 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13169 Each subfilter and some options have a short and a long name that can be used
13170 interchangeably, i.e. dr/dering are the same.
13172 The filters accept the following options:
13176 Set postprocessing subfilters string.
13179 All subfilters share common options to determine their scope:
13183 Honor the quality commands for this subfilter.
13186 Do chrominance filtering, too (default).
13189 Do luminance filtering only (no chrominance).
13192 Do chrominance filtering only (no luminance).
13195 These options can be appended after the subfilter name, separated by a '|'.
13197 Available subfilters are:
13200 @item hb/hdeblock[|difference[|flatness]]
13201 Horizontal deblocking filter
13204 Difference factor where higher values mean more deblocking (default: @code{32}).
13206 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13209 @item vb/vdeblock[|difference[|flatness]]
13210 Vertical deblocking filter
13213 Difference factor where higher values mean more deblocking (default: @code{32}).
13215 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13218 @item ha/hadeblock[|difference[|flatness]]
13219 Accurate horizontal deblocking filter
13222 Difference factor where higher values mean more deblocking (default: @code{32}).
13224 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13227 @item va/vadeblock[|difference[|flatness]]
13228 Accurate vertical deblocking filter
13231 Difference factor where higher values mean more deblocking (default: @code{32}).
13233 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13237 The horizontal and vertical deblocking filters share the difference and
13238 flatness values so you cannot set different horizontal and vertical
13242 @item h1/x1hdeblock
13243 Experimental horizontal deblocking filter
13245 @item v1/x1vdeblock
13246 Experimental vertical deblocking filter
13251 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13254 larger -> stronger filtering
13256 larger -> stronger filtering
13258 larger -> stronger filtering
13261 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13264 Stretch luminance to @code{0-255}.
13267 @item lb/linblenddeint
13268 Linear blend deinterlacing filter that deinterlaces the given block by
13269 filtering all lines with a @code{(1 2 1)} filter.
13271 @item li/linipoldeint
13272 Linear interpolating deinterlacing filter that deinterlaces the given block by
13273 linearly interpolating every second line.
13275 @item ci/cubicipoldeint
13276 Cubic interpolating deinterlacing filter deinterlaces the given block by
13277 cubically interpolating every second line.
13279 @item md/mediandeint
13280 Median deinterlacing filter that deinterlaces the given block by applying a
13281 median filter to every second line.
13283 @item fd/ffmpegdeint
13284 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13285 second line with a @code{(-1 4 2 4 -1)} filter.
13288 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13289 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13291 @item fq/forceQuant[|quantizer]
13292 Overrides the quantizer table from the input with the constant quantizer you
13300 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13303 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13306 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13309 @subsection Examples
13313 Apply horizontal and vertical deblocking, deringing and automatic
13314 brightness/contrast:
13320 Apply default filters without brightness/contrast correction:
13326 Apply default filters and temporal denoiser:
13328 pp=default/tmpnoise|1|2|3
13332 Apply deblocking on luminance only, and switch vertical deblocking on or off
13333 automatically depending on available CPU time:
13340 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13341 similar to spp = 6 with 7 point DCT, where only the center sample is
13344 The filter accepts the following options:
13348 Force a constant quantization parameter. It accepts an integer in range
13349 0 to 63. If not set, the filter will use the QP from the video stream
13353 Set thresholding mode. Available modes are:
13357 Set hard thresholding.
13359 Set soft thresholding (better de-ringing effect, but likely blurrier).
13361 Set medium thresholding (good results, default).
13365 @section premultiply
13366 Apply alpha premultiply effect to input video stream using first plane
13367 of second stream as alpha.
13369 Both streams must have same dimensions and same pixel format.
13371 The filter accepts the following option:
13375 Set which planes will be processed, unprocessed planes will be copied.
13376 By default value 0xf, all planes will be processed.
13379 Do not require 2nd input for processing, instead use alpha plane from input stream.
13383 Apply prewitt operator to input video stream.
13385 The filter accepts the following option:
13389 Set which planes will be processed, unprocessed planes will be copied.
13390 By default value 0xf, all planes will be processed.
13393 Set value which will be multiplied with filtered result.
13396 Set value which will be added to filtered result.
13399 @anchor{program_opencl}
13400 @section program_opencl
13402 Filter video using an OpenCL program.
13407 OpenCL program source file.
13410 Kernel name in program.
13413 Number of inputs to the filter. Defaults to 1.
13416 Size of output frames. Defaults to the same as the first input.
13420 The program source file must contain a kernel function with the given name,
13421 which will be run once for each plane of the output. Each run on a plane
13422 gets enqueued as a separate 2D global NDRange with one work-item for each
13423 pixel to be generated. The global ID offset for each work-item is therefore
13424 the coordinates of a pixel in the destination image.
13426 The kernel function needs to take the following arguments:
13429 Destination image, @var{__write_only image2d_t}.
13431 This image will become the output; the kernel should write all of it.
13433 Frame index, @var{unsigned int}.
13435 This is a counter starting from zero and increasing by one for each frame.
13437 Source images, @var{__read_only image2d_t}.
13439 These are the most recent images on each input. The kernel may read from
13440 them to generate the output, but they can't be written to.
13447 Copy the input to the output (output must be the same size as the input).
13449 __kernel void copy(__write_only image2d_t destination,
13450 unsigned int index,
13451 __read_only image2d_t source)
13453 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13455 int2 location = (int2)(get_global_id(0), get_global_id(1));
13457 float4 value = read_imagef(source, sampler, location);
13459 write_imagef(destination, location, value);
13464 Apply a simple transformation, rotating the input by an amount increasing
13465 with the index counter. Pixel values are linearly interpolated by the
13466 sampler, and the output need not have the same dimensions as the input.
13468 __kernel void rotate_image(__write_only image2d_t dst,
13469 unsigned int index,
13470 __read_only image2d_t src)
13472 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13473 CLK_FILTER_LINEAR);
13475 float angle = (float)index / 100.0f;
13477 float2 dst_dim = convert_float2(get_image_dim(dst));
13478 float2 src_dim = convert_float2(get_image_dim(src));
13480 float2 dst_cen = dst_dim / 2.0f;
13481 float2 src_cen = src_dim / 2.0f;
13483 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13485 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13487 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13488 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13490 src_pos = src_pos * src_dim / dst_dim;
13492 float2 src_loc = src_pos + src_cen;
13494 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13495 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13496 write_imagef(dst, dst_loc, 0.5f);
13498 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13503 Blend two inputs together, with the amount of each input used varying
13504 with the index counter.
13506 __kernel void blend_images(__write_only image2d_t dst,
13507 unsigned int index,
13508 __read_only image2d_t src1,
13509 __read_only image2d_t src2)
13511 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13512 CLK_FILTER_LINEAR);
13514 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13516 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13517 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13518 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13520 float4 val1 = read_imagef(src1, sampler, src1_loc);
13521 float4 val2 = read_imagef(src2, sampler, src2_loc);
13523 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13529 @section pseudocolor
13531 Alter frame colors in video with pseudocolors.
13533 This filter accept the following options:
13537 set pixel first component expression
13540 set pixel second component expression
13543 set pixel third component expression
13546 set pixel fourth component expression, corresponds to the alpha component
13549 set component to use as base for altering colors
13552 Each of them specifies the expression to use for computing the lookup table for
13553 the corresponding pixel component values.
13555 The expressions can contain the following constants and functions:
13560 The input width and height.
13563 The input value for the pixel component.
13565 @item ymin, umin, vmin, amin
13566 The minimum allowed component value.
13568 @item ymax, umax, vmax, amax
13569 The maximum allowed component value.
13572 All expressions default to "val".
13574 @subsection Examples
13578 Change too high luma values to gradient:
13580 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'"
13586 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
13587 Ratio) between two input videos.
13589 This filter takes in input two input videos, the first input is
13590 considered the "main" source and is passed unchanged to the
13591 output. The second input is used as a "reference" video for computing
13594 Both video inputs must have the same resolution and pixel format for
13595 this filter to work correctly. Also it assumes that both inputs
13596 have the same number of frames, which are compared one by one.
13598 The obtained average PSNR is printed through the logging system.
13600 The filter stores the accumulated MSE (mean squared error) of each
13601 frame, and at the end of the processing it is averaged across all frames
13602 equally, and the following formula is applied to obtain the PSNR:
13605 PSNR = 10*log10(MAX^2/MSE)
13608 Where MAX is the average of the maximum values of each component of the
13611 The description of the accepted parameters follows.
13614 @item stats_file, f
13615 If specified the filter will use the named file to save the PSNR of
13616 each individual frame. When filename equals "-" the data is sent to
13619 @item stats_version
13620 Specifies which version of the stats file format to use. Details of
13621 each format are written below.
13622 Default value is 1.
13624 @item stats_add_max
13625 Determines whether the max value is output to the stats log.
13626 Default value is 0.
13627 Requires stats_version >= 2. If this is set and stats_version < 2,
13628 the filter will return an error.
13631 This filter also supports the @ref{framesync} options.
13633 The file printed if @var{stats_file} is selected, contains a sequence of
13634 key/value pairs of the form @var{key}:@var{value} for each compared
13637 If a @var{stats_version} greater than 1 is specified, a header line precedes
13638 the list of per-frame-pair stats, with key value pairs following the frame
13639 format with the following parameters:
13642 @item psnr_log_version
13643 The version of the log file format. Will match @var{stats_version}.
13646 A comma separated list of the per-frame-pair parameters included in
13650 A description of each shown per-frame-pair parameter follows:
13654 sequential number of the input frame, starting from 1
13657 Mean Square Error pixel-by-pixel average difference of the compared
13658 frames, averaged over all the image components.
13660 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
13661 Mean Square Error pixel-by-pixel average difference of the compared
13662 frames for the component specified by the suffix.
13664 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
13665 Peak Signal to Noise ratio of the compared frames for the component
13666 specified by the suffix.
13668 @item max_avg, max_y, max_u, max_v
13669 Maximum allowed value for each channel, and average over all
13675 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13676 [main][ref] psnr="stats_file=stats.log" [out]
13679 On this example the input file being processed is compared with the
13680 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
13681 is stored in @file{stats.log}.
13686 Pulldown reversal (inverse telecine) filter, capable of handling mixed
13687 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
13690 The pullup filter is designed to take advantage of future context in making
13691 its decisions. This filter is stateless in the sense that it does not lock
13692 onto a pattern to follow, but it instead looks forward to the following
13693 fields in order to identify matches and rebuild progressive frames.
13695 To produce content with an even framerate, insert the fps filter after
13696 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
13697 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
13699 The filter accepts the following options:
13706 These options set the amount of "junk" to ignore at the left, right, top, and
13707 bottom of the image, respectively. Left and right are in units of 8 pixels,
13708 while top and bottom are in units of 2 lines.
13709 The default is 8 pixels on each side.
13712 Set the strict breaks. Setting this option to 1 will reduce the chances of
13713 filter generating an occasional mismatched frame, but it may also cause an
13714 excessive number of frames to be dropped during high motion sequences.
13715 Conversely, setting it to -1 will make filter match fields more easily.
13716 This may help processing of video where there is slight blurring between
13717 the fields, but may also cause there to be interlaced frames in the output.
13718 Default value is @code{0}.
13721 Set the metric plane to use. It accepts the following values:
13727 Use chroma blue plane.
13730 Use chroma red plane.
13733 This option may be set to use chroma plane instead of the default luma plane
13734 for doing filter's computations. This may improve accuracy on very clean
13735 source material, but more likely will decrease accuracy, especially if there
13736 is chroma noise (rainbow effect) or any grayscale video.
13737 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
13738 load and make pullup usable in realtime on slow machines.
13741 For best results (without duplicated frames in the output file) it is
13742 necessary to change the output frame rate. For example, to inverse
13743 telecine NTSC input:
13745 ffmpeg -i input -vf pullup -r 24000/1001 ...
13750 Change video quantization parameters (QP).
13752 The filter accepts the following option:
13756 Set expression for quantization parameter.
13759 The expression is evaluated through the eval API and can contain, among others,
13760 the following constants:
13764 1 if index is not 129, 0 otherwise.
13767 Sequential index starting from -129 to 128.
13770 @subsection Examples
13774 Some equation like:
13782 Flush video frames from internal cache of frames into a random order.
13783 No frame is discarded.
13784 Inspired by @ref{frei0r} nervous filter.
13788 Set size in number of frames of internal cache, in range from @code{2} to
13789 @code{512}. Default is @code{30}.
13792 Set seed for random number generator, must be an integer included between
13793 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
13794 less than @code{0}, the filter will try to use a good random seed on a
13798 @section readeia608
13800 Read closed captioning (EIA-608) information from the top lines of a video frame.
13802 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13803 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13804 with EIA-608 data (starting from 0). A description of each metadata value follows:
13807 @item lavfi.readeia608.X.cc
13808 The two bytes stored as EIA-608 data (printed in hexadecimal).
13810 @item lavfi.readeia608.X.line
13811 The number of the line on which the EIA-608 data was identified and read.
13814 This filter accepts the following options:
13818 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13821 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13824 Set minimal acceptable amplitude change for sync codes detection.
13825 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13828 Set the ratio of width reserved for sync code detection.
13829 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13832 Set the max peaks height difference for sync code detection.
13833 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13836 Set max peaks period difference for sync code detection.
13837 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13840 Set the first two max start code bits differences.
13841 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13844 Set the minimum ratio of bits height compared to 3rd start code bit.
13845 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13848 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13851 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13854 Enable checking the parity bit. In the event of a parity error, the filter will output
13855 @code{0x00} for that character. Default is false.
13858 @subsection Examples
13862 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13864 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
13870 Read vertical interval timecode (VITC) information from the top lines of a
13873 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13874 timecode value, if a valid timecode has been detected. Further metadata key
13875 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13876 timecode data has been found or not.
13878 This filter accepts the following options:
13882 Set the maximum number of lines to scan for VITC data. If the value is set to
13883 @code{-1} the full video frame is scanned. Default is @code{45}.
13886 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13887 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13890 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13891 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13894 @subsection Examples
13898 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13899 draw @code{--:--:--:--} as a placeholder:
13901 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13907 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13909 Destination pixel at position (X, Y) will be picked from source (x, y) position
13910 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13911 value for pixel will be used for destination pixel.
13913 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13914 will have Xmap/Ymap video stream dimensions.
13915 Xmap and Ymap input video streams are 16bit depth, single channel.
13917 @section removegrain
13919 The removegrain filter is a spatial denoiser for progressive video.
13923 Set mode for the first plane.
13926 Set mode for the second plane.
13929 Set mode for the third plane.
13932 Set mode for the fourth plane.
13935 Range of mode is from 0 to 24. Description of each mode follows:
13939 Leave input plane unchanged. Default.
13942 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13945 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13948 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13951 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13952 This is equivalent to a median filter.
13955 Line-sensitive clipping giving the minimal change.
13958 Line-sensitive clipping, intermediate.
13961 Line-sensitive clipping, intermediate.
13964 Line-sensitive clipping, intermediate.
13967 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13970 Replaces the target pixel with the closest neighbour.
13973 [1 2 1] horizontal and vertical kernel blur.
13979 Bob mode, interpolates top field from the line where the neighbours
13980 pixels are the closest.
13983 Bob mode, interpolates bottom field from the line where the neighbours
13984 pixels are the closest.
13987 Bob mode, interpolates top field. Same as 13 but with a more complicated
13988 interpolation formula.
13991 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13992 interpolation formula.
13995 Clips the pixel with the minimum and maximum of respectively the maximum and
13996 minimum of each pair of opposite neighbour pixels.
13999 Line-sensitive clipping using opposite neighbours whose greatest distance from
14000 the current pixel is minimal.
14003 Replaces the pixel with the average of its 8 neighbours.
14006 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14009 Clips pixels using the averages of opposite neighbour.
14012 Same as mode 21 but simpler and faster.
14015 Small edge and halo removal, but reputed useless.
14021 @section removelogo
14023 Suppress a TV station logo, using an image file to determine which
14024 pixels comprise the logo. It works by filling in the pixels that
14025 comprise the logo with neighboring pixels.
14027 The filter accepts the following options:
14031 Set the filter bitmap file, which can be any image format supported by
14032 libavformat. The width and height of the image file must match those of the
14033 video stream being processed.
14036 Pixels in the provided bitmap image with a value of zero are not
14037 considered part of the logo, non-zero pixels are considered part of
14038 the logo. If you use white (255) for the logo and black (0) for the
14039 rest, you will be safe. For making the filter bitmap, it is
14040 recommended to take a screen capture of a black frame with the logo
14041 visible, and then using a threshold filter followed by the erode
14042 filter once or twice.
14044 If needed, little splotches can be fixed manually. Remember that if
14045 logo pixels are not covered, the filter quality will be much
14046 reduced. Marking too many pixels as part of the logo does not hurt as
14047 much, but it will increase the amount of blurring needed to cover over
14048 the image and will destroy more information than necessary, and extra
14049 pixels will slow things down on a large logo.
14051 @section repeatfields
14053 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14054 fields based on its value.
14058 Reverse a video clip.
14060 Warning: This filter requires memory to buffer the entire clip, so trimming
14063 @subsection Examples
14067 Take the first 5 seconds of a clip, and reverse it.
14074 Apply roberts cross operator to input video stream.
14076 The filter accepts the following option:
14080 Set which planes will be processed, unprocessed planes will be copied.
14081 By default value 0xf, all planes will be processed.
14084 Set value which will be multiplied with filtered result.
14087 Set value which will be added to filtered result.
14092 Rotate video by an arbitrary angle expressed in radians.
14094 The filter accepts the following options:
14096 A description of the optional parameters follows.
14099 Set an expression for the angle by which to rotate the input video
14100 clockwise, expressed as a number of radians. A negative value will
14101 result in a counter-clockwise rotation. By default it is set to "0".
14103 This expression is evaluated for each frame.
14106 Set the output width expression, default value is "iw".
14107 This expression is evaluated just once during configuration.
14110 Set the output height expression, default value is "ih".
14111 This expression is evaluated just once during configuration.
14114 Enable bilinear interpolation if set to 1, a value of 0 disables
14115 it. Default value is 1.
14118 Set the color used to fill the output area not covered by the rotated
14119 image. For the general syntax of this option, check the
14120 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14121 If the special value "none" is selected then no
14122 background is printed (useful for example if the background is never shown).
14124 Default value is "black".
14127 The expressions for the angle and the output size can contain the
14128 following constants and functions:
14132 sequential number of the input frame, starting from 0. It is always NAN
14133 before the first frame is filtered.
14136 time in seconds of the input frame, it is set to 0 when the filter is
14137 configured. It is always NAN before the first frame is filtered.
14141 horizontal and vertical chroma subsample values. For example for the
14142 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14146 the input video width and height
14150 the output width and height, that is the size of the padded area as
14151 specified by the @var{width} and @var{height} expressions
14155 the minimal width/height required for completely containing the input
14156 video rotated by @var{a} radians.
14158 These are only available when computing the @option{out_w} and
14159 @option{out_h} expressions.
14162 @subsection Examples
14166 Rotate the input by PI/6 radians clockwise:
14172 Rotate the input by PI/6 radians counter-clockwise:
14178 Rotate the input by 45 degrees clockwise:
14184 Apply a constant rotation with period T, starting from an angle of PI/3:
14186 rotate=PI/3+2*PI*t/T
14190 Make the input video rotation oscillating with a period of T
14191 seconds and an amplitude of A radians:
14193 rotate=A*sin(2*PI/T*t)
14197 Rotate the video, output size is chosen so that the whole rotating
14198 input video is always completely contained in the output:
14200 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14204 Rotate the video, reduce the output size so that no background is ever
14207 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14211 @subsection Commands
14213 The filter supports the following commands:
14217 Set the angle expression.
14218 The command accepts the same syntax of the corresponding option.
14220 If the specified expression is not valid, it is kept at its current
14226 Apply Shape Adaptive Blur.
14228 The filter accepts the following options:
14231 @item luma_radius, lr
14232 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14233 value is 1.0. A greater value will result in a more blurred image, and
14234 in slower processing.
14236 @item luma_pre_filter_radius, lpfr
14237 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14240 @item luma_strength, ls
14241 Set luma maximum difference between pixels to still be considered, must
14242 be a value in the 0.1-100.0 range, default value is 1.0.
14244 @item chroma_radius, cr
14245 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14246 greater value will result in a more blurred image, and in slower
14249 @item chroma_pre_filter_radius, cpfr
14250 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14252 @item chroma_strength, cs
14253 Set chroma maximum difference between pixels to still be considered,
14254 must be a value in the -0.9-100.0 range.
14257 Each chroma option value, if not explicitly specified, is set to the
14258 corresponding luma option value.
14263 Scale (resize) the input video, using the libswscale library.
14265 The scale filter forces the output display aspect ratio to be the same
14266 of the input, by changing the output sample aspect ratio.
14268 If the input image format is different from the format requested by
14269 the next filter, the scale filter will convert the input to the
14272 @subsection Options
14273 The filter accepts the following options, or any of the options
14274 supported by the libswscale scaler.
14276 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14277 the complete list of scaler options.
14282 Set the output video dimension expression. Default value is the input
14285 If the @var{width} or @var{w} value is 0, the input width is used for
14286 the output. If the @var{height} or @var{h} value is 0, the input height
14287 is used for the output.
14289 If one and only one of the values is -n with n >= 1, the scale filter
14290 will use a value that maintains the aspect ratio of the input image,
14291 calculated from the other specified dimension. After that it will,
14292 however, make sure that the calculated dimension is divisible by n and
14293 adjust the value if necessary.
14295 If both values are -n with n >= 1, the behavior will be identical to
14296 both values being set to 0 as previously detailed.
14298 See below for the list of accepted constants for use in the dimension
14302 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14306 Only evaluate expressions once during the filter initialization or when a command is processed.
14309 Evaluate expressions for each incoming frame.
14313 Default value is @samp{init}.
14317 Set the interlacing mode. It accepts the following values:
14321 Force interlaced aware scaling.
14324 Do not apply interlaced scaling.
14327 Select interlaced aware scaling depending on whether the source frames
14328 are flagged as interlaced or not.
14331 Default value is @samp{0}.
14334 Set libswscale scaling flags. See
14335 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14336 complete list of values. If not explicitly specified the filter applies
14340 @item param0, param1
14341 Set libswscale input parameters for scaling algorithms that need them. See
14342 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14343 complete documentation. If not explicitly specified the filter applies
14349 Set the video size. For the syntax of this option, check the
14350 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14352 @item in_color_matrix
14353 @item out_color_matrix
14354 Set in/output YCbCr color space type.
14356 This allows the autodetected value to be overridden as well as allows forcing
14357 a specific value used for the output and encoder.
14359 If not specified, the color space type depends on the pixel format.
14365 Choose automatically.
14368 Format conforming to International Telecommunication Union (ITU)
14369 Recommendation BT.709.
14372 Set color space conforming to the United States Federal Communications
14373 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14376 Set color space conforming to:
14380 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14383 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14386 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14391 Set color space conforming to SMPTE ST 240:1999.
14396 Set in/output YCbCr sample range.
14398 This allows the autodetected value to be overridden as well as allows forcing
14399 a specific value used for the output and encoder. If not specified, the
14400 range depends on the pixel format. Possible values:
14404 Choose automatically.
14407 Set full range (0-255 in case of 8-bit luma).
14409 @item mpeg/limited/tv
14410 Set "MPEG" range (16-235 in case of 8-bit luma).
14413 @item force_original_aspect_ratio
14414 Enable decreasing or increasing output video width or height if necessary to
14415 keep the original aspect ratio. Possible values:
14419 Scale the video as specified and disable this feature.
14422 The output video dimensions will automatically be decreased if needed.
14425 The output video dimensions will automatically be increased if needed.
14429 One useful instance of this option is that when you know a specific device's
14430 maximum allowed resolution, you can use this to limit the output video to
14431 that, while retaining the aspect ratio. For example, device A allows
14432 1280x720 playback, and your video is 1920x800. Using this option (set it to
14433 decrease) and specifying 1280x720 to the command line makes the output
14436 Please note that this is a different thing than specifying -1 for @option{w}
14437 or @option{h}, you still need to specify the output resolution for this option
14442 The values of the @option{w} and @option{h} options are expressions
14443 containing the following constants:
14448 The input width and height
14452 These are the same as @var{in_w} and @var{in_h}.
14456 The output (scaled) width and height
14460 These are the same as @var{out_w} and @var{out_h}
14463 The same as @var{iw} / @var{ih}
14466 input sample aspect ratio
14469 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14473 horizontal and vertical input chroma subsample values. For example for the
14474 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14478 horizontal and vertical output chroma subsample values. For example for the
14479 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14482 @subsection Examples
14486 Scale the input video to a size of 200x100
14491 This is equivalent to:
14502 Specify a size abbreviation for the output size:
14507 which can also be written as:
14513 Scale the input to 2x:
14515 scale=w=2*iw:h=2*ih
14519 The above is the same as:
14521 scale=2*in_w:2*in_h
14525 Scale the input to 2x with forced interlaced scaling:
14527 scale=2*iw:2*ih:interl=1
14531 Scale the input to half size:
14533 scale=w=iw/2:h=ih/2
14537 Increase the width, and set the height to the same size:
14543 Seek Greek harmony:
14550 Increase the height, and set the width to 3/2 of the height:
14552 scale=w=3/2*oh:h=3/5*ih
14556 Increase the size, making the size a multiple of the chroma
14559 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
14563 Increase the width to a maximum of 500 pixels,
14564 keeping the same aspect ratio as the input:
14566 scale=w='min(500\, iw*3/2):h=-1'
14570 Make pixels square by combining scale and setsar:
14572 scale='trunc(ih*dar):ih',setsar=1/1
14576 Make pixels square by combining scale and setsar,
14577 making sure the resulting resolution is even (required by some codecs):
14579 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
14583 @subsection Commands
14585 This filter supports the following commands:
14589 Set the output video dimension expression.
14590 The command accepts the same syntax of the corresponding option.
14592 If the specified expression is not valid, it is kept at its current
14598 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
14599 format conversion on CUDA video frames. Setting the output width and height
14600 works in the same way as for the @var{scale} filter.
14602 The following additional options are accepted:
14605 The pixel format of the output CUDA frames. If set to the string "same" (the
14606 default), the input format will be kept. Note that automatic format negotiation
14607 and conversion is not yet supported for hardware frames
14610 The interpolation algorithm used for resizing. One of the following:
14617 @item cubic2p_bspline
14618 2-parameter cubic (B=1, C=0)
14620 @item cubic2p_catmullrom
14621 2-parameter cubic (B=0, C=1/2)
14623 @item cubic2p_b05c03
14624 2-parameter cubic (B=1/2, C=3/10)
14636 Scale (resize) the input video, based on a reference video.
14638 See the scale filter for available options, scale2ref supports the same but
14639 uses the reference video instead of the main input as basis. scale2ref also
14640 supports the following additional constants for the @option{w} and
14641 @option{h} options:
14646 The main input video's width and height
14649 The same as @var{main_w} / @var{main_h}
14652 The main input video's sample aspect ratio
14654 @item main_dar, mdar
14655 The main input video's display aspect ratio. Calculated from
14656 @code{(main_w / main_h) * main_sar}.
14660 The main input video's horizontal and vertical chroma subsample values.
14661 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
14665 @subsection Examples
14669 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
14671 'scale2ref[b][a];[a][b]overlay'
14675 @anchor{selectivecolor}
14676 @section selectivecolor
14678 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
14679 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
14680 by the "purity" of the color (that is, how saturated it already is).
14682 This filter is similar to the Adobe Photoshop Selective Color tool.
14684 The filter accepts the following options:
14687 @item correction_method
14688 Select color correction method.
14690 Available values are:
14693 Specified adjustments are applied "as-is" (added/subtracted to original pixel
14696 Specified adjustments are relative to the original component value.
14698 Default is @code{absolute}.
14700 Adjustments for red pixels (pixels where the red component is the maximum)
14702 Adjustments for yellow pixels (pixels where the blue component is the minimum)
14704 Adjustments for green pixels (pixels where the green component is the maximum)
14706 Adjustments for cyan pixels (pixels where the red component is the minimum)
14708 Adjustments for blue pixels (pixels where the blue component is the maximum)
14710 Adjustments for magenta pixels (pixels where the green component is the minimum)
14712 Adjustments for white pixels (pixels where all components are greater than 128)
14714 Adjustments for all pixels except pure black and pure white
14716 Adjustments for black pixels (pixels where all components are lesser than 128)
14718 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
14721 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
14722 4 space separated floating point adjustment values in the [-1,1] range,
14723 respectively to adjust the amount of cyan, magenta, yellow and black for the
14724 pixels of its range.
14726 @subsection Examples
14730 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
14731 increase magenta by 27% in blue areas:
14733 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
14737 Use a Photoshop selective color preset:
14739 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
14743 @anchor{separatefields}
14744 @section separatefields
14746 The @code{separatefields} takes a frame-based video input and splits
14747 each frame into its components fields, producing a new half height clip
14748 with twice the frame rate and twice the frame count.
14750 This filter use field-dominance information in frame to decide which
14751 of each pair of fields to place first in the output.
14752 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
14754 @section setdar, setsar
14756 The @code{setdar} filter sets the Display Aspect Ratio for the filter
14759 This is done by changing the specified Sample (aka Pixel) Aspect
14760 Ratio, according to the following equation:
14762 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
14765 Keep in mind that the @code{setdar} filter does not modify the pixel
14766 dimensions of the video frame. Also, the display aspect ratio set by
14767 this filter may be changed by later filters in the filterchain,
14768 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
14771 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
14772 the filter output video.
14774 Note that as a consequence of the application of this filter, the
14775 output display aspect ratio will change according to the equation
14778 Keep in mind that the sample aspect ratio set by the @code{setsar}
14779 filter may be changed by later filters in the filterchain, e.g. if
14780 another "setsar" or a "setdar" filter is applied.
14782 It accepts the following parameters:
14785 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
14786 Set the aspect ratio used by the filter.
14788 The parameter can be a floating point number string, an expression, or
14789 a string of the form @var{num}:@var{den}, where @var{num} and
14790 @var{den} are the numerator and denominator of the aspect ratio. If
14791 the parameter is not specified, it is assumed the value "0".
14792 In case the form "@var{num}:@var{den}" is used, the @code{:} character
14796 Set the maximum integer value to use for expressing numerator and
14797 denominator when reducing the expressed aspect ratio to a rational.
14798 Default value is @code{100}.
14802 The parameter @var{sar} is an expression containing
14803 the following constants:
14807 These are approximated values for the mathematical constants e
14808 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14811 The input width and height.
14814 These are the same as @var{w} / @var{h}.
14817 The input sample aspect ratio.
14820 The input display aspect ratio. It is the same as
14821 (@var{w} / @var{h}) * @var{sar}.
14824 Horizontal and vertical chroma subsample values. For example, for the
14825 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14828 @subsection Examples
14833 To change the display aspect ratio to 16:9, specify one of the following:
14840 To change the sample aspect ratio to 10:11, specify:
14846 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14847 1000 in the aspect ratio reduction, use the command:
14849 setdar=ratio=16/9:max=1000
14857 Force field for the output video frame.
14859 The @code{setfield} filter marks the interlace type field for the
14860 output frames. It does not change the input frame, but only sets the
14861 corresponding property, which affects how the frame is treated by
14862 following filters (e.g. @code{fieldorder} or @code{yadif}).
14864 The filter accepts the following options:
14869 Available values are:
14873 Keep the same field property.
14876 Mark the frame as bottom-field-first.
14879 Mark the frame as top-field-first.
14882 Mark the frame as progressive.
14888 Show a line containing various information for each input video frame.
14889 The input video is not modified.
14891 The shown line contains a sequence of key/value pairs of the form
14892 @var{key}:@var{value}.
14894 The following values are shown in the output:
14898 The (sequential) number of the input frame, starting from 0.
14901 The Presentation TimeStamp of the input frame, expressed as a number of
14902 time base units. The time base unit depends on the filter input pad.
14905 The Presentation TimeStamp of the input frame, expressed as a number of
14909 The position of the frame in the input stream, or -1 if this information is
14910 unavailable and/or meaningless (for example in case of synthetic video).
14913 The pixel format name.
14916 The sample aspect ratio of the input frame, expressed in the form
14917 @var{num}/@var{den}.
14920 The size of the input frame. For the syntax of this option, check the
14921 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14924 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14925 for bottom field first).
14928 This is 1 if the frame is a key frame, 0 otherwise.
14931 The picture type of the input frame ("I" for an I-frame, "P" for a
14932 P-frame, "B" for a B-frame, or "?" for an unknown type).
14933 Also refer to the documentation of the @code{AVPictureType} enum and of
14934 the @code{av_get_picture_type_char} function defined in
14935 @file{libavutil/avutil.h}.
14938 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14940 @item plane_checksum
14941 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14942 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14945 @section showpalette
14947 Displays the 256 colors palette of each frame. This filter is only relevant for
14948 @var{pal8} pixel format frames.
14950 It accepts the following option:
14954 Set the size of the box used to represent one palette color entry. Default is
14955 @code{30} (for a @code{30x30} pixel box).
14958 @section shuffleframes
14960 Reorder and/or duplicate and/or drop video frames.
14962 It accepts the following parameters:
14966 Set the destination indexes of input frames.
14967 This is space or '|' separated list of indexes that maps input frames to output
14968 frames. Number of indexes also sets maximal value that each index may have.
14969 '-1' index have special meaning and that is to drop frame.
14972 The first frame has the index 0. The default is to keep the input unchanged.
14974 @subsection Examples
14978 Swap second and third frame of every three frames of the input:
14980 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14984 Swap 10th and 1st frame of every ten frames of the input:
14986 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14990 @section shuffleplanes
14992 Reorder and/or duplicate video planes.
14994 It accepts the following parameters:
14999 The index of the input plane to be used as the first output plane.
15002 The index of the input plane to be used as the second output plane.
15005 The index of the input plane to be used as the third output plane.
15008 The index of the input plane to be used as the fourth output plane.
15012 The first plane has the index 0. The default is to keep the input unchanged.
15014 @subsection Examples
15018 Swap the second and third planes of the input:
15020 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15024 @anchor{signalstats}
15025 @section signalstats
15026 Evaluate various visual metrics that assist in determining issues associated
15027 with the digitization of analog video media.
15029 By default the filter will log these metadata values:
15033 Display the minimal Y value contained within the input frame. Expressed in
15037 Display the Y value at the 10% percentile within the input frame. Expressed in
15041 Display the average Y value within the input frame. Expressed in range of
15045 Display the Y value at the 90% percentile within the input frame. Expressed in
15049 Display the maximum Y value contained within the input frame. Expressed in
15053 Display the minimal U value contained within the input frame. Expressed in
15057 Display the U value at the 10% percentile within the input frame. Expressed in
15061 Display the average U value within the input frame. Expressed in range of
15065 Display the U value at the 90% percentile within the input frame. Expressed in
15069 Display the maximum U value contained within the input frame. Expressed in
15073 Display the minimal V value contained within the input frame. Expressed in
15077 Display the V value at the 10% percentile within the input frame. Expressed in
15081 Display the average V value within the input frame. Expressed in range of
15085 Display the V value at the 90% percentile within the input frame. Expressed in
15089 Display the maximum V value contained within the input frame. Expressed in
15093 Display the minimal saturation value contained within the input frame.
15094 Expressed in range of [0-~181.02].
15097 Display the saturation value at the 10% percentile within the input frame.
15098 Expressed in range of [0-~181.02].
15101 Display the average saturation value within the input frame. Expressed in range
15105 Display the saturation value at the 90% percentile within the input frame.
15106 Expressed in range of [0-~181.02].
15109 Display the maximum saturation value contained within the input frame.
15110 Expressed in range of [0-~181.02].
15113 Display the median value for hue within the input frame. Expressed in range of
15117 Display the average value for hue within the input frame. Expressed in range of
15121 Display the average of sample value difference between all values of the Y
15122 plane in the current frame and corresponding values of the previous input frame.
15123 Expressed in range of [0-255].
15126 Display the average of sample value difference between all values of the U
15127 plane in the current frame and corresponding values of the previous input frame.
15128 Expressed in range of [0-255].
15131 Display the average of sample value difference between all values of the V
15132 plane in the current frame and corresponding values of the previous input frame.
15133 Expressed in range of [0-255].
15136 Display bit depth of Y plane in current frame.
15137 Expressed in range of [0-16].
15140 Display bit depth of U plane in current frame.
15141 Expressed in range of [0-16].
15144 Display bit depth of V plane in current frame.
15145 Expressed in range of [0-16].
15148 The filter accepts the following options:
15154 @option{stat} specify an additional form of image analysis.
15155 @option{out} output video with the specified type of pixel highlighted.
15157 Both options accept the following values:
15161 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15162 unlike the neighboring pixels of the same field. Examples of temporal outliers
15163 include the results of video dropouts, head clogs, or tape tracking issues.
15166 Identify @var{vertical line repetition}. Vertical line repetition includes
15167 similar rows of pixels within a frame. In born-digital video vertical line
15168 repetition is common, but this pattern is uncommon in video digitized from an
15169 analog source. When it occurs in video that results from the digitization of an
15170 analog source it can indicate concealment from a dropout compensator.
15173 Identify pixels that fall outside of legal broadcast range.
15177 Set the highlight color for the @option{out} option. The default color is
15181 @subsection Examples
15185 Output data of various video metrics:
15187 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15191 Output specific data about the minimum and maximum values of the Y plane per frame:
15193 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15197 Playback video while highlighting pixels that are outside of broadcast range in red.
15199 ffplay example.mov -vf signalstats="out=brng:color=red"
15203 Playback video with signalstats metadata drawn over the frame.
15205 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15208 The contents of signalstat_drawtext.txt used in the command are:
15211 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15212 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15213 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15214 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15222 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15223 input. In this case the matching between the inputs can be calculated additionally.
15224 The filter always passes through the first input. The signature of each stream can
15225 be written into a file.
15227 It accepts the following options:
15231 Enable or disable the matching process.
15233 Available values are:
15237 Disable the calculation of a matching (default).
15239 Calculate the matching for the whole video and output whether the whole video
15240 matches or only parts.
15242 Calculate only until a matching is found or the video ends. Should be faster in
15247 Set the number of inputs. The option value must be a non negative integer.
15248 Default value is 1.
15251 Set the path to which the output is written. If there is more than one input,
15252 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15253 integer), that will be replaced with the input number. If no filename is
15254 specified, no output will be written. This is the default.
15257 Choose the output format.
15259 Available values are:
15263 Use the specified binary representation (default).
15265 Use the specified xml representation.
15269 Set threshold to detect one word as similar. The option value must be an integer
15270 greater than zero. The default value is 9000.
15273 Set threshold to detect all words as similar. The option value must be an integer
15274 greater than zero. The default value is 60000.
15277 Set threshold to detect frames as similar. The option value must be an integer
15278 greater than zero. The default value is 116.
15281 Set the minimum length of a sequence in frames to recognize it as matching
15282 sequence. The option value must be a non negative integer value.
15283 The default value is 0.
15286 Set the minimum relation, that matching frames to all frames must have.
15287 The option value must be a double value between 0 and 1. The default value is 0.5.
15290 @subsection Examples
15294 To calculate the signature of an input video and store it in signature.bin:
15296 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15300 To detect whether two videos match and store the signatures in XML format in
15301 signature0.xml and signature1.xml:
15303 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 -
15311 Blur the input video without impacting the outlines.
15313 It accepts the following options:
15316 @item luma_radius, lr
15317 Set the luma radius. The option value must be a float number in
15318 the range [0.1,5.0] that specifies the variance of the gaussian filter
15319 used to blur the image (slower if larger). Default value is 1.0.
15321 @item luma_strength, ls
15322 Set the luma strength. The option value must be a float number
15323 in the range [-1.0,1.0] that configures the blurring. A value included
15324 in [0.0,1.0] will blur the image whereas a value included in
15325 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15327 @item luma_threshold, lt
15328 Set the luma threshold used as a coefficient to determine
15329 whether a pixel should be blurred or not. The option value must be an
15330 integer in the range [-30,30]. A value of 0 will filter all the image,
15331 a value included in [0,30] will filter flat areas and a value included
15332 in [-30,0] will filter edges. Default value is 0.
15334 @item chroma_radius, cr
15335 Set the chroma radius. The option value must be a float number in
15336 the range [0.1,5.0] that specifies the variance of the gaussian filter
15337 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15339 @item chroma_strength, cs
15340 Set the chroma strength. The option value must be a float number
15341 in the range [-1.0,1.0] that configures the blurring. A value included
15342 in [0.0,1.0] will blur the image whereas a value included in
15343 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15345 @item chroma_threshold, ct
15346 Set the chroma threshold used as a coefficient to determine
15347 whether a pixel should be blurred or not. The option value must be an
15348 integer in the range [-30,30]. A value of 0 will filter all the image,
15349 a value included in [0,30] will filter flat areas and a value included
15350 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15353 If a chroma option is not explicitly set, the corresponding luma value
15358 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15360 This filter takes in input two input videos, the first input is
15361 considered the "main" source and is passed unchanged to the
15362 output. The second input is used as a "reference" video for computing
15365 Both video inputs must have the same resolution and pixel format for
15366 this filter to work correctly. Also it assumes that both inputs
15367 have the same number of frames, which are compared one by one.
15369 The filter stores the calculated SSIM of each frame.
15371 The description of the accepted parameters follows.
15374 @item stats_file, f
15375 If specified the filter will use the named file to save the SSIM of
15376 each individual frame. When filename equals "-" the data is sent to
15380 The file printed if @var{stats_file} is selected, contains a sequence of
15381 key/value pairs of the form @var{key}:@var{value} for each compared
15384 A description of each shown parameter follows:
15388 sequential number of the input frame, starting from 1
15390 @item Y, U, V, R, G, B
15391 SSIM of the compared frames for the component specified by the suffix.
15394 SSIM of the compared frames for the whole frame.
15397 Same as above but in dB representation.
15400 This filter also supports the @ref{framesync} options.
15404 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15405 [main][ref] ssim="stats_file=stats.log" [out]
15408 On this example the input file being processed is compared with the
15409 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15410 is stored in @file{stats.log}.
15412 Another example with both psnr and ssim at same time:
15414 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
15419 Convert between different stereoscopic image formats.
15421 The filters accept the following options:
15425 Set stereoscopic image format of input.
15427 Available values for input image formats are:
15430 side by side parallel (left eye left, right eye right)
15433 side by side crosseye (right eye left, left eye right)
15436 side by side parallel with half width resolution
15437 (left eye left, right eye right)
15440 side by side crosseye with half width resolution
15441 (right eye left, left eye right)
15444 above-below (left eye above, right eye below)
15447 above-below (right eye above, left eye below)
15450 above-below with half height resolution
15451 (left eye above, right eye below)
15454 above-below with half height resolution
15455 (right eye above, left eye below)
15458 alternating frames (left eye first, right eye second)
15461 alternating frames (right eye first, left eye second)
15464 interleaved rows (left eye has top row, right eye starts on next row)
15467 interleaved rows (right eye has top row, left eye starts on next row)
15470 interleaved columns, left eye first
15473 interleaved columns, right eye first
15475 Default value is @samp{sbsl}.
15479 Set stereoscopic image format of output.
15483 side by side parallel (left eye left, right eye right)
15486 side by side crosseye (right eye left, left eye right)
15489 side by side parallel with half width resolution
15490 (left eye left, right eye right)
15493 side by side crosseye with half width resolution
15494 (right eye left, left eye right)
15497 above-below (left eye above, right eye below)
15500 above-below (right eye above, left eye below)
15503 above-below with half height resolution
15504 (left eye above, right eye below)
15507 above-below with half height resolution
15508 (right eye above, left eye below)
15511 alternating frames (left eye first, right eye second)
15514 alternating frames (right eye first, left eye second)
15517 interleaved rows (left eye has top row, right eye starts on next row)
15520 interleaved rows (right eye has top row, left eye starts on next row)
15523 anaglyph red/blue gray
15524 (red filter on left eye, blue filter on right eye)
15527 anaglyph red/green gray
15528 (red filter on left eye, green filter on right eye)
15531 anaglyph red/cyan gray
15532 (red filter on left eye, cyan filter on right eye)
15535 anaglyph red/cyan half colored
15536 (red filter on left eye, cyan filter on right eye)
15539 anaglyph red/cyan color
15540 (red filter on left eye, cyan filter on right eye)
15543 anaglyph red/cyan color optimized with the least squares projection of dubois
15544 (red filter on left eye, cyan filter on right eye)
15547 anaglyph green/magenta gray
15548 (green filter on left eye, magenta filter on right eye)
15551 anaglyph green/magenta half colored
15552 (green filter on left eye, magenta filter on right eye)
15555 anaglyph green/magenta colored
15556 (green filter on left eye, magenta filter on right eye)
15559 anaglyph green/magenta color optimized with the least squares projection of dubois
15560 (green filter on left eye, magenta filter on right eye)
15563 anaglyph yellow/blue gray
15564 (yellow filter on left eye, blue filter on right eye)
15567 anaglyph yellow/blue half colored
15568 (yellow filter on left eye, blue filter on right eye)
15571 anaglyph yellow/blue colored
15572 (yellow filter on left eye, blue filter on right eye)
15575 anaglyph yellow/blue color optimized with the least squares projection of dubois
15576 (yellow filter on left eye, blue filter on right eye)
15579 mono output (left eye only)
15582 mono output (right eye only)
15585 checkerboard, left eye first
15588 checkerboard, right eye first
15591 interleaved columns, left eye first
15594 interleaved columns, right eye first
15600 Default value is @samp{arcd}.
15603 @subsection Examples
15607 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
15613 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
15619 @section streamselect, astreamselect
15620 Select video or audio streams.
15622 The filter accepts the following options:
15626 Set number of inputs. Default is 2.
15629 Set input indexes to remap to outputs.
15632 @subsection Commands
15634 The @code{streamselect} and @code{astreamselect} filter supports the following
15639 Set input indexes to remap to outputs.
15642 @subsection Examples
15646 Select first 5 seconds 1st stream and rest of time 2nd stream:
15648 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
15652 Same as above, but for audio:
15654 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
15659 Apply sobel operator to input video stream.
15661 The filter accepts the following option:
15665 Set which planes will be processed, unprocessed planes will be copied.
15666 By default value 0xf, all planes will be processed.
15669 Set value which will be multiplied with filtered result.
15672 Set value which will be added to filtered result.
15678 Apply a simple postprocessing filter that compresses and decompresses the image
15679 at several (or - in the case of @option{quality} level @code{6} - all) shifts
15680 and average the results.
15682 The filter accepts the following options:
15686 Set quality. This option defines the number of levels for averaging. It accepts
15687 an integer in the range 0-6. If set to @code{0}, the filter will have no
15688 effect. A value of @code{6} means the higher quality. For each increment of
15689 that value the speed drops by a factor of approximately 2. Default value is
15693 Force a constant quantization parameter. If not set, the filter will use the QP
15694 from the video stream (if available).
15697 Set thresholding mode. Available modes are:
15701 Set hard thresholding (default).
15703 Set soft thresholding (better de-ringing effect, but likely blurrier).
15706 @item use_bframe_qp
15707 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
15708 option may cause flicker since the B-Frames have often larger QP. Default is
15709 @code{0} (not enabled).
15714 Scale the input by applying one of the super-resolution methods based on
15715 convolutional neural networks. Supported models:
15719 Super-Resolution Convolutional Neural Network model (SRCNN).
15720 See @url{https://arxiv.org/abs/1501.00092}.
15723 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
15724 See @url{https://arxiv.org/abs/1609.05158}.
15727 Training scripts as well as scripts for model generation are provided in
15728 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
15730 The filter accepts the following options:
15734 Specify which DNN backend to use for model loading and execution. This option accepts
15735 the following values:
15739 Native implementation of DNN loading and execution.
15742 TensorFlow backend. To enable this backend you
15743 need to install the TensorFlow for C library (see
15744 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
15745 @code{--enable-libtensorflow}
15748 Default value is @samp{native}.
15751 Set path to model file specifying network architecture and its parameters.
15752 Note that different backends use different file formats. TensorFlow backend
15753 can load files for both formats, while native backend can load files for only
15757 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
15758 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
15759 input upscaled using bicubic upscaling with proper scale factor.
15765 Draw subtitles on top of input video using the libass library.
15767 To enable compilation of this filter you need to configure FFmpeg with
15768 @code{--enable-libass}. This filter also requires a build with libavcodec and
15769 libavformat to convert the passed subtitles file to ASS (Advanced Substation
15770 Alpha) subtitles format.
15772 The filter accepts the following options:
15776 Set the filename of the subtitle file to read. It must be specified.
15778 @item original_size
15779 Specify the size of the original video, the video for which the ASS file
15780 was composed. For the syntax of this option, check the
15781 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15782 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
15783 correctly scale the fonts if the aspect ratio has been changed.
15786 Set a directory path containing fonts that can be used by the filter.
15787 These fonts will be used in addition to whatever the font provider uses.
15790 Process alpha channel, by default alpha channel is untouched.
15793 Set subtitles input character encoding. @code{subtitles} filter only. Only
15794 useful if not UTF-8.
15796 @item stream_index, si
15797 Set subtitles stream index. @code{subtitles} filter only.
15800 Override default style or script info parameters of the subtitles. It accepts a
15801 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
15804 If the first key is not specified, it is assumed that the first value
15805 specifies the @option{filename}.
15807 For example, to render the file @file{sub.srt} on top of the input
15808 video, use the command:
15813 which is equivalent to:
15815 subtitles=filename=sub.srt
15818 To render the default subtitles stream from file @file{video.mkv}, use:
15820 subtitles=video.mkv
15823 To render the second subtitles stream from that file, use:
15825 subtitles=video.mkv:si=1
15828 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
15829 @code{DejaVu Serif}, use:
15831 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
15834 @section super2xsai
15836 Scale the input by 2x and smooth using the Super2xSaI (Scale and
15837 Interpolate) pixel art scaling algorithm.
15839 Useful for enlarging pixel art images without reducing sharpness.
15843 Swap two rectangular objects in video.
15845 This filter accepts the following options:
15855 Set 1st rect x coordinate.
15858 Set 1st rect y coordinate.
15861 Set 2nd rect x coordinate.
15864 Set 2nd rect y coordinate.
15866 All expressions are evaluated once for each frame.
15869 The all options are expressions containing the following constants:
15874 The input width and height.
15877 same as @var{w} / @var{h}
15880 input sample aspect ratio
15883 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15886 The number of the input frame, starting from 0.
15889 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15892 the position in the file of the input frame, NAN if unknown
15900 Apply telecine process to the video.
15902 This filter accepts the following options:
15911 The default value is @code{top}.
15915 A string of numbers representing the pulldown pattern you wish to apply.
15916 The default value is @code{23}.
15920 Some typical patterns:
15925 24p: 2332 (preferred)
15932 24p: 222222222223 ("Euro pulldown")
15939 Apply threshold effect to video stream.
15941 This filter needs four video streams to perform thresholding.
15942 First stream is stream we are filtering.
15943 Second stream is holding threshold values, third stream is holding min values,
15944 and last, fourth stream is holding max values.
15946 The filter accepts the following option:
15950 Set which planes will be processed, unprocessed planes will be copied.
15951 By default value 0xf, all planes will be processed.
15954 For example if first stream pixel's component value is less then threshold value
15955 of pixel component from 2nd threshold stream, third stream value will picked,
15956 otherwise fourth stream pixel component value will be picked.
15958 Using color source filter one can perform various types of thresholding:
15960 @subsection Examples
15964 Binary threshold, using gray color as threshold:
15966 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15970 Inverted binary threshold, using gray color as threshold:
15972 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15976 Truncate binary threshold, using gray color as threshold:
15978 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15982 Threshold to zero, using gray color as threshold:
15984 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15988 Inverted threshold to zero, using gray color as threshold:
15990 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15995 Select the most representative frame in a given sequence of consecutive frames.
15997 The filter accepts the following options:
16001 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16002 will pick one of them, and then handle the next batch of @var{n} frames until
16003 the end. Default is @code{100}.
16006 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16007 value will result in a higher memory usage, so a high value is not recommended.
16009 @subsection Examples
16013 Extract one picture each 50 frames:
16019 Complete example of a thumbnail creation with @command{ffmpeg}:
16021 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16027 Tile several successive frames together.
16029 The filter accepts the following options:
16034 Set the grid size (i.e. the number of lines and columns). For the syntax of
16035 this option, check the
16036 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16039 Set the maximum number of frames to render in the given area. It must be less
16040 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16041 the area will be used.
16044 Set the outer border margin in pixels.
16047 Set the inner border thickness (i.e. the number of pixels between frames). For
16048 more advanced padding options (such as having different values for the edges),
16049 refer to the pad video filter.
16052 Specify the color of the unused area. For the syntax of this option, check the
16053 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16054 The default value of @var{color} is "black".
16057 Set the number of frames to overlap when tiling several successive frames together.
16058 The value must be between @code{0} and @var{nb_frames - 1}.
16061 Set the number of frames to initially be empty before displaying first output frame.
16062 This controls how soon will one get first output frame.
16063 The value must be between @code{0} and @var{nb_frames - 1}.
16066 @subsection Examples
16070 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16072 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16074 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16075 duplicating each output frame to accommodate the originally detected frame
16079 Display @code{5} pictures in an area of @code{3x2} frames,
16080 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16081 mixed flat and named options:
16083 tile=3x2:nb_frames=5:padding=7:margin=2
16087 @section tinterlace
16089 Perform various types of temporal field interlacing.
16091 Frames are counted starting from 1, so the first input frame is
16094 The filter accepts the following options:
16099 Specify the mode of the interlacing. This option can also be specified
16100 as a value alone. See below for a list of values for this option.
16102 Available values are:
16106 Move odd frames into the upper field, even into the lower field,
16107 generating a double height frame at half frame rate.
16111 Frame 1 Frame 2 Frame 3 Frame 4
16113 11111 22222 33333 44444
16114 11111 22222 33333 44444
16115 11111 22222 33333 44444
16116 11111 22222 33333 44444
16130 Only output odd frames, even frames are dropped, generating a frame with
16131 unchanged height at half frame rate.
16136 Frame 1 Frame 2 Frame 3 Frame 4
16138 11111 22222 33333 44444
16139 11111 22222 33333 44444
16140 11111 22222 33333 44444
16141 11111 22222 33333 44444
16151 Only output even frames, odd frames are dropped, generating a frame with
16152 unchanged height at half frame rate.
16157 Frame 1 Frame 2 Frame 3 Frame 4
16159 11111 22222 33333 44444
16160 11111 22222 33333 44444
16161 11111 22222 33333 44444
16162 11111 22222 33333 44444
16172 Expand each frame to full height, but pad alternate lines with black,
16173 generating a frame with double height at the same input frame rate.
16178 Frame 1 Frame 2 Frame 3 Frame 4
16180 11111 22222 33333 44444
16181 11111 22222 33333 44444
16182 11111 22222 33333 44444
16183 11111 22222 33333 44444
16186 11111 ..... 33333 .....
16187 ..... 22222 ..... 44444
16188 11111 ..... 33333 .....
16189 ..... 22222 ..... 44444
16190 11111 ..... 33333 .....
16191 ..... 22222 ..... 44444
16192 11111 ..... 33333 .....
16193 ..... 22222 ..... 44444
16197 @item interleave_top, 4
16198 Interleave the upper field from odd frames with the lower field from
16199 even frames, generating a frame with unchanged height at half frame rate.
16204 Frame 1 Frame 2 Frame 3 Frame 4
16206 11111<- 22222 33333<- 44444
16207 11111 22222<- 33333 44444<-
16208 11111<- 22222 33333<- 44444
16209 11111 22222<- 33333 44444<-
16219 @item interleave_bottom, 5
16220 Interleave the lower field from odd frames with the upper field from
16221 even frames, generating a frame with unchanged height at half frame rate.
16226 Frame 1 Frame 2 Frame 3 Frame 4
16228 11111 22222<- 33333 44444<-
16229 11111<- 22222 33333<- 44444
16230 11111 22222<- 33333 44444<-
16231 11111<- 22222 33333<- 44444
16241 @item interlacex2, 6
16242 Double frame rate with unchanged height. Frames are inserted each
16243 containing the second temporal field from the previous input frame and
16244 the first temporal field from the next input frame. This mode relies on
16245 the top_field_first flag. Useful for interlaced video displays with no
16246 field synchronisation.
16251 Frame 1 Frame 2 Frame 3 Frame 4
16253 11111 22222 33333 44444
16254 11111 22222 33333 44444
16255 11111 22222 33333 44444
16256 11111 22222 33333 44444
16259 11111 22222 22222 33333 33333 44444 44444
16260 11111 11111 22222 22222 33333 33333 44444
16261 11111 22222 22222 33333 33333 44444 44444
16262 11111 11111 22222 22222 33333 33333 44444
16267 Move odd frames into the upper field, even into the lower field,
16268 generating a double height frame at same frame rate.
16273 Frame 1 Frame 2 Frame 3 Frame 4
16275 11111 22222 33333 44444
16276 11111 22222 33333 44444
16277 11111 22222 33333 44444
16278 11111 22222 33333 44444
16281 11111 33333 33333 55555
16282 22222 22222 44444 44444
16283 11111 33333 33333 55555
16284 22222 22222 44444 44444
16285 11111 33333 33333 55555
16286 22222 22222 44444 44444
16287 11111 33333 33333 55555
16288 22222 22222 44444 44444
16293 Numeric values are deprecated but are accepted for backward
16294 compatibility reasons.
16296 Default mode is @code{merge}.
16299 Specify flags influencing the filter process.
16301 Available value for @var{flags} is:
16304 @item low_pass_filter, vlfp
16305 Enable linear vertical low-pass filtering in the filter.
16306 Vertical low-pass filtering is required when creating an interlaced
16307 destination from a progressive source which contains high-frequency
16308 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16311 @item complex_filter, cvlfp
16312 Enable complex vertical low-pass filtering.
16313 This will slightly less reduce interlace 'twitter' and Moire
16314 patterning but better retain detail and subjective sharpness impression.
16318 Vertical low-pass filtering can only be enabled for @option{mode}
16319 @var{interleave_top} and @var{interleave_bottom}.
16325 Mix successive video frames.
16327 A description of the accepted options follows.
16331 The number of successive frames to mix. If unspecified, it defaults to 3.
16334 Specify weight of each input video frame.
16335 Each weight is separated by space. If number of weights is smaller than
16336 number of @var{frames} last specified weight will be used for all remaining
16340 Specify scale, if it is set it will be multiplied with sum
16341 of each weight multiplied with pixel values to give final destination
16342 pixel value. By default @var{scale} is auto scaled to sum of weights.
16345 @subsection Examples
16349 Average 7 successive frames:
16351 tmix=frames=7:weights="1 1 1 1 1 1 1"
16355 Apply simple temporal convolution:
16357 tmix=frames=3:weights="-1 3 -1"
16361 Similar as above but only showing temporal differences:
16363 tmix=frames=3:weights="-1 2 -1":scale=1
16368 Tone map colors from different dynamic ranges.
16370 This filter expects data in single precision floating point, as it needs to
16371 operate on (and can output) out-of-range values. Another filter, such as
16372 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16374 The tonemapping algorithms implemented only work on linear light, so input
16375 data should be linearized beforehand (and possibly correctly tagged).
16378 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16381 @subsection Options
16382 The filter accepts the following options.
16386 Set the tone map algorithm to use.
16388 Possible values are:
16391 Do not apply any tone map, only desaturate overbright pixels.
16394 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16395 in-range values, while distorting out-of-range values.
16398 Stretch the entire reference gamut to a linear multiple of the display.
16401 Fit a logarithmic transfer between the tone curves.
16404 Preserve overall image brightness with a simple curve, using nonlinear
16405 contrast, which results in flattening details and degrading color accuracy.
16408 Preserve both dark and bright details better than @var{reinhard}, at the cost
16409 of slightly darkening everything. Use it when detail preservation is more
16410 important than color and brightness accuracy.
16413 Smoothly map out-of-range values, while retaining contrast and colors for
16414 in-range material as much as possible. Use it when color accuracy is more
16415 important than detail preservation.
16421 Tune the tone mapping algorithm.
16423 This affects the following algorithms:
16429 Specifies the scale factor to use while stretching.
16433 Specifies the exponent of the function.
16437 Specify an extra linear coefficient to multiply into the signal before clipping.
16441 Specify the local contrast coefficient at the display peak.
16442 Default to 0.5, which means that in-gamut values will be about half as bright
16449 Specify the transition point from linear to mobius transform. Every value
16450 below this point is guaranteed to be mapped 1:1. The higher the value, the
16451 more accurate the result will be, at the cost of losing bright details.
16452 Default to 0.3, which due to the steep initial slope still preserves in-range
16453 colors fairly accurately.
16457 Apply desaturation for highlights that exceed this level of brightness. The
16458 higher the parameter, the more color information will be preserved. This
16459 setting helps prevent unnaturally blown-out colors for super-highlights, by
16460 (smoothly) turning into white instead. This makes images feel more natural,
16461 at the cost of reducing information about out-of-range colors.
16463 The default of 2.0 is somewhat conservative and will mostly just apply to
16464 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
16466 This option works only if the input frame has a supported color tag.
16469 Override signal/nominal/reference peak with this value. Useful when the
16470 embedded peak information in display metadata is not reliable or when tone
16471 mapping from a lower range to a higher range.
16477 Transpose rows with columns in the input video and optionally flip it.
16479 It accepts the following parameters:
16484 Specify the transposition direction.
16486 Can assume the following values:
16488 @item 0, 4, cclock_flip
16489 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
16497 Rotate by 90 degrees clockwise, that is:
16505 Rotate by 90 degrees counterclockwise, that is:
16512 @item 3, 7, clock_flip
16513 Rotate by 90 degrees clockwise and vertically flip, that is:
16521 For values between 4-7, the transposition is only done if the input
16522 video geometry is portrait and not landscape. These values are
16523 deprecated, the @code{passthrough} option should be used instead.
16525 Numerical values are deprecated, and should be dropped in favor of
16526 symbolic constants.
16529 Do not apply the transposition if the input geometry matches the one
16530 specified by the specified value. It accepts the following values:
16533 Always apply transposition.
16535 Preserve portrait geometry (when @var{height} >= @var{width}).
16537 Preserve landscape geometry (when @var{width} >= @var{height}).
16540 Default value is @code{none}.
16543 For example to rotate by 90 degrees clockwise and preserve portrait
16546 transpose=dir=1:passthrough=portrait
16549 The command above can also be specified as:
16551 transpose=1:portrait
16554 @section transpose_npp
16556 Transpose rows with columns in the input video and optionally flip it.
16557 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
16559 It accepts the following parameters:
16564 Specify the transposition direction.
16566 Can assume the following values:
16569 Rotate by 90 degrees counterclockwise and vertically flip. (default)
16572 Rotate by 90 degrees clockwise.
16575 Rotate by 90 degrees counterclockwise.
16578 Rotate by 90 degrees clockwise and vertically flip.
16582 Do not apply the transposition if the input geometry matches the one
16583 specified by the specified value. It accepts the following values:
16586 Always apply transposition. (default)
16588 Preserve portrait geometry (when @var{height} >= @var{width}).
16590 Preserve landscape geometry (when @var{width} >= @var{height}).
16596 Trim the input so that the output contains one continuous subpart of the input.
16598 It accepts the following parameters:
16601 Specify the time of the start of the kept section, i.e. the frame with the
16602 timestamp @var{start} will be the first frame in the output.
16605 Specify the time of the first frame that will be dropped, i.e. the frame
16606 immediately preceding the one with the timestamp @var{end} will be the last
16607 frame in the output.
16610 This is the same as @var{start}, except this option sets the start timestamp
16611 in timebase units instead of seconds.
16614 This is the same as @var{end}, except this option sets the end timestamp
16615 in timebase units instead of seconds.
16618 The maximum duration of the output in seconds.
16621 The number of the first frame that should be passed to the output.
16624 The number of the first frame that should be dropped.
16627 @option{start}, @option{end}, and @option{duration} are expressed as time
16628 duration specifications; see
16629 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16630 for the accepted syntax.
16632 Note that the first two sets of the start/end options and the @option{duration}
16633 option look at the frame timestamp, while the _frame variants simply count the
16634 frames that pass through the filter. Also note that this filter does not modify
16635 the timestamps. If you wish for the output timestamps to start at zero, insert a
16636 setpts filter after the trim filter.
16638 If multiple start or end options are set, this filter tries to be greedy and
16639 keep all the frames that match at least one of the specified constraints. To keep
16640 only the part that matches all the constraints at once, chain multiple trim
16643 The defaults are such that all the input is kept. So it is possible to set e.g.
16644 just the end values to keep everything before the specified time.
16649 Drop everything except the second minute of input:
16651 ffmpeg -i INPUT -vf trim=60:120
16655 Keep only the first second:
16657 ffmpeg -i INPUT -vf trim=duration=1
16662 @section unpremultiply
16663 Apply alpha unpremultiply effect to input video stream using first plane
16664 of second stream as alpha.
16666 Both streams must have same dimensions and same pixel format.
16668 The filter accepts the following option:
16672 Set which planes will be processed, unprocessed planes will be copied.
16673 By default value 0xf, all planes will be processed.
16675 If the format has 1 or 2 components, then luma is bit 0.
16676 If the format has 3 or 4 components:
16677 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
16678 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
16679 If present, the alpha channel is always the last bit.
16682 Do not require 2nd input for processing, instead use alpha plane from input stream.
16688 Sharpen or blur the input video.
16690 It accepts the following parameters:
16693 @item luma_msize_x, lx
16694 Set the luma matrix horizontal size. It must be an odd integer between
16695 3 and 23. The default value is 5.
16697 @item luma_msize_y, ly
16698 Set the luma matrix vertical size. It must be an odd integer between 3
16699 and 23. The default value is 5.
16701 @item luma_amount, la
16702 Set the luma effect strength. It must be a floating point number, reasonable
16703 values lay between -1.5 and 1.5.
16705 Negative values will blur the input video, while positive values will
16706 sharpen it, a value of zero will disable the effect.
16708 Default value is 1.0.
16710 @item chroma_msize_x, cx
16711 Set the chroma matrix horizontal size. It must be an odd integer
16712 between 3 and 23. The default value is 5.
16714 @item chroma_msize_y, cy
16715 Set the chroma matrix vertical size. It must be an odd integer
16716 between 3 and 23. The default value is 5.
16718 @item chroma_amount, ca
16719 Set the chroma effect strength. It must be a floating point number, reasonable
16720 values lay between -1.5 and 1.5.
16722 Negative values will blur the input video, while positive values will
16723 sharpen it, a value of zero will disable the effect.
16725 Default value is 0.0.
16729 All parameters are optional and default to the equivalent of the
16730 string '5:5:1.0:5:5:0.0'.
16732 @subsection Examples
16736 Apply strong luma sharpen effect:
16738 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
16742 Apply a strong blur of both luma and chroma parameters:
16744 unsharp=7:7:-2:7:7:-2
16750 Apply ultra slow/simple postprocessing filter that compresses and decompresses
16751 the image at several (or - in the case of @option{quality} level @code{8} - all)
16752 shifts and average the results.
16754 The way this differs from the behavior of spp is that uspp actually encodes &
16755 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
16756 DCT similar to MJPEG.
16758 The filter accepts the following options:
16762 Set quality. This option defines the number of levels for averaging. It accepts
16763 an integer in the range 0-8. If set to @code{0}, the filter will have no
16764 effect. A value of @code{8} means the higher quality. For each increment of
16765 that value the speed drops by a factor of approximately 2. Default value is
16769 Force a constant quantization parameter. If not set, the filter will use the QP
16770 from the video stream (if available).
16773 @section vaguedenoiser
16775 Apply a wavelet based denoiser.
16777 It transforms each frame from the video input into the wavelet domain,
16778 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
16779 the obtained coefficients. It does an inverse wavelet transform after.
16780 Due to wavelet properties, it should give a nice smoothed result, and
16781 reduced noise, without blurring picture features.
16783 This filter accepts the following options:
16787 The filtering strength. The higher, the more filtered the video will be.
16788 Hard thresholding can use a higher threshold than soft thresholding
16789 before the video looks overfiltered. Default value is 2.
16792 The filtering method the filter will use.
16794 It accepts the following values:
16797 All values under the threshold will be zeroed.
16800 All values under the threshold will be zeroed. All values above will be
16801 reduced by the threshold.
16804 Scales or nullifies coefficients - intermediary between (more) soft and
16805 (less) hard thresholding.
16808 Default is garrote.
16811 Number of times, the wavelet will decompose the picture. Picture can't
16812 be decomposed beyond a particular point (typically, 8 for a 640x480
16813 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
16816 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
16819 A list of the planes to process. By default all planes are processed.
16822 @section vectorscope
16824 Display 2 color component values in the two dimensional graph (which is called
16827 This filter accepts the following options:
16831 Set vectorscope mode.
16833 It accepts the following values:
16836 Gray values are displayed on graph, higher brightness means more pixels have
16837 same component color value on location in graph. This is the default mode.
16840 Gray values are displayed on graph. Surrounding pixels values which are not
16841 present in video frame are drawn in gradient of 2 color components which are
16842 set by option @code{x} and @code{y}. The 3rd color component is static.
16845 Actual color components values present in video frame are displayed on graph.
16848 Similar as color2 but higher frequency of same values @code{x} and @code{y}
16849 on graph increases value of another color component, which is luminance by
16850 default values of @code{x} and @code{y}.
16853 Actual colors present in video frame are displayed on graph. If two different
16854 colors map to same position on graph then color with higher value of component
16855 not present in graph is picked.
16858 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
16859 component picked from radial gradient.
16863 Set which color component will be represented on X-axis. Default is @code{1}.
16866 Set which color component will be represented on Y-axis. Default is @code{2}.
16869 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
16870 of color component which represents frequency of (X, Y) location in graph.
16875 No envelope, this is default.
16878 Instant envelope, even darkest single pixel will be clearly highlighted.
16881 Hold maximum and minimum values presented in graph over time. This way you
16882 can still spot out of range values without constantly looking at vectorscope.
16885 Peak and instant envelope combined together.
16889 Set what kind of graticule to draw.
16897 Set graticule opacity.
16900 Set graticule flags.
16904 Draw graticule for white point.
16907 Draw graticule for black point.
16910 Draw color points short names.
16914 Set background opacity.
16916 @item lthreshold, l
16917 Set low threshold for color component not represented on X or Y axis.
16918 Values lower than this value will be ignored. Default is 0.
16919 Note this value is multiplied with actual max possible value one pixel component
16920 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
16923 @item hthreshold, h
16924 Set high threshold for color component not represented on X or Y axis.
16925 Values higher than this value will be ignored. Default is 1.
16926 Note this value is multiplied with actual max possible value one pixel component
16927 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
16928 is 0.9 * 255 = 230.
16930 @item colorspace, c
16931 Set what kind of colorspace to use when drawing graticule.
16940 @anchor{vidstabdetect}
16941 @section vidstabdetect
16943 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16944 @ref{vidstabtransform} for pass 2.
16946 This filter generates a file with relative translation and rotation
16947 transform information about subsequent frames, which is then used by
16948 the @ref{vidstabtransform} filter.
16950 To enable compilation of this filter you need to configure FFmpeg with
16951 @code{--enable-libvidstab}.
16953 This filter accepts the following options:
16957 Set the path to the file used to write the transforms information.
16958 Default value is @file{transforms.trf}.
16961 Set how shaky the video is and how quick the camera is. It accepts an
16962 integer in the range 1-10, a value of 1 means little shakiness, a
16963 value of 10 means strong shakiness. Default value is 5.
16966 Set the accuracy of the detection process. It must be a value in the
16967 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16968 accuracy. Default value is 15.
16971 Set stepsize of the search process. The region around minimum is
16972 scanned with 1 pixel resolution. Default value is 6.
16975 Set minimum contrast. Below this value a local measurement field is
16976 discarded. Must be a floating point value in the range 0-1. Default
16980 Set reference frame number for tripod mode.
16982 If enabled, the motion of the frames is compared to a reference frame
16983 in the filtered stream, identified by the specified number. The idea
16984 is to compensate all movements in a more-or-less static scene and keep
16985 the camera view absolutely still.
16987 If set to 0, it is disabled. The frames are counted starting from 1.
16990 Show fields and transforms in the resulting frames. It accepts an
16991 integer in the range 0-2. Default value is 0, which disables any
16995 @subsection Examples
16999 Use default values:
17005 Analyze strongly shaky movie and put the results in file
17006 @file{mytransforms.trf}:
17008 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17012 Visualize the result of internal transformations in the resulting
17015 vidstabdetect=show=1
17019 Analyze a video with medium shakiness using @command{ffmpeg}:
17021 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17025 @anchor{vidstabtransform}
17026 @section vidstabtransform
17028 Video stabilization/deshaking: pass 2 of 2,
17029 see @ref{vidstabdetect} for pass 1.
17031 Read a file with transform information for each frame and
17032 apply/compensate them. Together with the @ref{vidstabdetect}
17033 filter this can be used to deshake videos. See also
17034 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17035 the @ref{unsharp} filter, see below.
17037 To enable compilation of this filter you need to configure FFmpeg with
17038 @code{--enable-libvidstab}.
17040 @subsection Options
17044 Set path to the file used to read the transforms. Default value is
17045 @file{transforms.trf}.
17048 Set the number of frames (value*2 + 1) used for lowpass filtering the
17049 camera movements. Default value is 10.
17051 For example a number of 10 means that 21 frames are used (10 in the
17052 past and 10 in the future) to smoothen the motion in the video. A
17053 larger value leads to a smoother video, but limits the acceleration of
17054 the camera (pan/tilt movements). 0 is a special case where a static
17055 camera is simulated.
17058 Set the camera path optimization algorithm.
17060 Accepted values are:
17063 gaussian kernel low-pass filter on camera motion (default)
17065 averaging on transformations
17069 Set maximal number of pixels to translate frames. Default value is -1,
17073 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17074 value is -1, meaning no limit.
17077 Specify how to deal with borders that may be visible due to movement
17080 Available values are:
17083 keep image information from previous frame (default)
17085 fill the border black
17089 Invert transforms if set to 1. Default value is 0.
17092 Consider transforms as relative to previous frame if set to 1,
17093 absolute if set to 0. Default value is 0.
17096 Set percentage to zoom. A positive value will result in a zoom-in
17097 effect, a negative value in a zoom-out effect. Default value is 0 (no
17101 Set optimal zooming to avoid borders.
17103 Accepted values are:
17108 optimal static zoom value is determined (only very strong movements
17109 will lead to visible borders) (default)
17111 optimal adaptive zoom value is determined (no borders will be
17112 visible), see @option{zoomspeed}
17115 Note that the value given at zoom is added to the one calculated here.
17118 Set percent to zoom maximally each frame (enabled when
17119 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17123 Specify type of interpolation.
17125 Available values are:
17130 linear only horizontal
17132 linear in both directions (default)
17134 cubic in both directions (slow)
17138 Enable virtual tripod mode if set to 1, which is equivalent to
17139 @code{relative=0:smoothing=0}. Default value is 0.
17141 Use also @code{tripod} option of @ref{vidstabdetect}.
17144 Increase log verbosity if set to 1. Also the detected global motions
17145 are written to the temporary file @file{global_motions.trf}. Default
17149 @subsection Examples
17153 Use @command{ffmpeg} for a typical stabilization with default values:
17155 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17158 Note the use of the @ref{unsharp} filter which is always recommended.
17161 Zoom in a bit more and load transform data from a given file:
17163 vidstabtransform=zoom=5:input="mytransforms.trf"
17167 Smoothen the video even more:
17169 vidstabtransform=smoothing=30
17175 Flip the input video vertically.
17177 For example, to vertically flip a video with @command{ffmpeg}:
17179 ffmpeg -i in.avi -vf "vflip" out.avi
17184 Detect variable frame rate video.
17186 This filter tries to detect if the input is variable or constant frame rate.
17188 At end it will output number of frames detected as having variable delta pts,
17189 and ones with constant delta pts.
17190 If there was frames with variable delta, than it will also show min and max delta
17196 Make or reverse a natural vignetting effect.
17198 The filter accepts the following options:
17202 Set lens angle expression as a number of radians.
17204 The value is clipped in the @code{[0,PI/2]} range.
17206 Default value: @code{"PI/5"}
17210 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17214 Set forward/backward mode.
17216 Available modes are:
17219 The larger the distance from the central point, the darker the image becomes.
17222 The larger the distance from the central point, the brighter the image becomes.
17223 This can be used to reverse a vignette effect, though there is no automatic
17224 detection to extract the lens @option{angle} and other settings (yet). It can
17225 also be used to create a burning effect.
17228 Default value is @samp{forward}.
17231 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17233 It accepts the following values:
17236 Evaluate expressions only once during the filter initialization.
17239 Evaluate expressions for each incoming frame. This is way slower than the
17240 @samp{init} mode since it requires all the scalers to be re-computed, but it
17241 allows advanced dynamic expressions.
17244 Default value is @samp{init}.
17247 Set dithering to reduce the circular banding effects. Default is @code{1}
17251 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17252 Setting this value to the SAR of the input will make a rectangular vignetting
17253 following the dimensions of the video.
17255 Default is @code{1/1}.
17258 @subsection Expressions
17260 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17261 following parameters.
17266 input width and height
17269 the number of input frame, starting from 0
17272 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17273 @var{TB} units, NAN if undefined
17276 frame rate of the input video, NAN if the input frame rate is unknown
17279 the PTS (Presentation TimeStamp) of the filtered video frame,
17280 expressed in seconds, NAN if undefined
17283 time base of the input video
17287 @subsection Examples
17291 Apply simple strong vignetting effect:
17297 Make a flickering vignetting:
17299 vignette='PI/4+random(1)*PI/50':eval=frame
17304 @section vmafmotion
17306 Obtain the average vmaf motion score of a video.
17307 It is one of the component filters of VMAF.
17309 The obtained average motion score is printed through the logging system.
17311 In the below example the input file @file{ref.mpg} is being processed and score
17315 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
17319 Stack input videos vertically.
17321 All streams must be of same pixel format and of same width.
17323 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
17324 to create same output.
17326 The filter accept the following option:
17330 Set number of input streams. Default is 2.
17333 If set to 1, force the output to terminate when the shortest input
17334 terminates. Default value is 0.
17339 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17340 Deinterlacing Filter").
17342 Based on the process described by Martin Weston for BBC R&D, and
17343 implemented based on the de-interlace algorithm written by Jim
17344 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17345 uses filter coefficients calculated by BBC R&D.
17347 There are two sets of filter coefficients, so called "simple":
17348 and "complex". Which set of filter coefficients is used can
17349 be set by passing an optional parameter:
17353 Set the interlacing filter coefficients. Accepts one of the following values:
17357 Simple filter coefficient set.
17359 More-complex filter coefficient set.
17361 Default value is @samp{complex}.
17364 Specify which frames to deinterlace. Accept one of the following values:
17368 Deinterlace all frames,
17370 Only deinterlace frames marked as interlaced.
17373 Default value is @samp{all}.
17377 Video waveform monitor.
17379 The waveform monitor plots color component intensity. By default luminance
17380 only. Each column of the waveform corresponds to a column of pixels in the
17383 It accepts the following options:
17387 Can be either @code{row}, or @code{column}. Default is @code{column}.
17388 In row mode, the graph on the left side represents color component value 0 and
17389 the right side represents value = 255. In column mode, the top side represents
17390 color component value = 0 and bottom side represents value = 255.
17393 Set intensity. Smaller values are useful to find out how many values of the same
17394 luminance are distributed across input rows/columns.
17395 Default value is @code{0.04}. Allowed range is [0, 1].
17398 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
17399 In mirrored mode, higher values will be represented on the left
17400 side for @code{row} mode and at the top for @code{column} mode. Default is
17401 @code{1} (mirrored).
17405 It accepts the following values:
17408 Presents information identical to that in the @code{parade}, except
17409 that the graphs representing color components are superimposed directly
17412 This display mode makes it easier to spot relative differences or similarities
17413 in overlapping areas of the color components that are supposed to be identical,
17414 such as neutral whites, grays, or blacks.
17417 Display separate graph for the color components side by side in
17418 @code{row} mode or one below the other in @code{column} mode.
17421 Display separate graph for the color components side by side in
17422 @code{column} mode or one below the other in @code{row} mode.
17424 Using this display mode makes it easy to spot color casts in the highlights
17425 and shadows of an image, by comparing the contours of the top and the bottom
17426 graphs of each waveform. Since whites, grays, and blacks are characterized
17427 by exactly equal amounts of red, green, and blue, neutral areas of the picture
17428 should display three waveforms of roughly equal width/height. If not, the
17429 correction is easy to perform by making level adjustments the three waveforms.
17431 Default is @code{stack}.
17433 @item components, c
17434 Set which color components to display. Default is 1, which means only luminance
17435 or red color component if input is in RGB colorspace. If is set for example to
17436 7 it will display all 3 (if) available color components.
17441 No envelope, this is default.
17444 Instant envelope, minimum and maximum values presented in graph will be easily
17445 visible even with small @code{step} value.
17448 Hold minimum and maximum values presented in graph across time. This way you
17449 can still spot out of range values without constantly looking at waveforms.
17452 Peak and instant envelope combined together.
17458 No filtering, this is default.
17461 Luma and chroma combined together.
17464 Similar as above, but shows difference between blue and red chroma.
17467 Similar as above, but use different colors.
17470 Displays only chroma.
17473 Displays actual color value on waveform.
17476 Similar as above, but with luma showing frequency of chroma values.
17480 Set which graticule to display.
17484 Do not display graticule.
17487 Display green graticule showing legal broadcast ranges.
17490 Display orange graticule showing legal broadcast ranges.
17494 Set graticule opacity.
17497 Set graticule flags.
17501 Draw numbers above lines. By default enabled.
17504 Draw dots instead of lines.
17508 Set scale used for displaying graticule.
17515 Default is digital.
17518 Set background opacity.
17521 @section weave, doubleweave
17523 The @code{weave} takes a field-based video input and join
17524 each two sequential fields into single frame, producing a new double
17525 height clip with half the frame rate and half the frame count.
17527 The @code{doubleweave} works same as @code{weave} but without
17528 halving frame rate and frame count.
17530 It accepts the following option:
17534 Set first field. Available values are:
17538 Set the frame as top-field-first.
17541 Set the frame as bottom-field-first.
17545 @subsection Examples
17549 Interlace video using @ref{select} and @ref{separatefields} filter:
17551 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
17556 Apply the xBR high-quality magnification filter which is designed for pixel
17557 art. It follows a set of edge-detection rules, see
17558 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
17560 It accepts the following option:
17564 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
17565 @code{3xBR} and @code{4} for @code{4xBR}.
17566 Default is @code{3}.
17572 Deinterlace the input video ("yadif" means "yet another deinterlacing
17575 It accepts the following parameters:
17581 The interlacing mode to adopt. It accepts one of the following values:
17584 @item 0, send_frame
17585 Output one frame for each frame.
17586 @item 1, send_field
17587 Output one frame for each field.
17588 @item 2, send_frame_nospatial
17589 Like @code{send_frame}, but it skips the spatial interlacing check.
17590 @item 3, send_field_nospatial
17591 Like @code{send_field}, but it skips the spatial interlacing check.
17594 The default value is @code{send_frame}.
17597 The picture field parity assumed for the input interlaced video. It accepts one
17598 of the following values:
17602 Assume the top field is first.
17604 Assume the bottom field is first.
17606 Enable automatic detection of field parity.
17609 The default value is @code{auto}.
17610 If the interlacing is unknown or the decoder does not export this information,
17611 top field first will be assumed.
17614 Specify which frames to deinterlace. Accept one of the following
17619 Deinterlace all frames.
17620 @item 1, interlaced
17621 Only deinterlace frames marked as interlaced.
17624 The default value is @code{all}.
17629 Apply Zoom & Pan effect.
17631 This filter accepts the following options:
17635 Set the zoom expression. Default is 1.
17639 Set the x and y expression. Default is 0.
17642 Set the duration expression in number of frames.
17643 This sets for how many number of frames effect will last for
17644 single input image.
17647 Set the output image size, default is 'hd720'.
17650 Set the output frame rate, default is '25'.
17653 Each expression can contain the following constants:
17672 Output frame count.
17676 Last calculated 'x' and 'y' position from 'x' and 'y' expression
17677 for current input frame.
17681 'x' and 'y' of last output frame of previous input frame or 0 when there was
17682 not yet such frame (first input frame).
17685 Last calculated zoom from 'z' expression for current input frame.
17688 Last calculated zoom of last output frame of previous input frame.
17691 Number of output frames for current input frame. Calculated from 'd' expression
17692 for each input frame.
17695 number of output frames created for previous input frame
17698 Rational number: input width / input height
17701 sample aspect ratio
17704 display aspect ratio
17708 @subsection Examples
17712 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
17714 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
17718 Zoom-in up to 1.5 and pan always at center of picture:
17720 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17724 Same as above but without pausing:
17726 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17732 Scale (resize) the input video, using the z.lib library:
17733 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
17734 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
17736 The zscale filter forces the output display aspect ratio to be the same
17737 as the input, by changing the output sample aspect ratio.
17739 If the input image format is different from the format requested by
17740 the next filter, the zscale filter will convert the input to the
17743 @subsection Options
17744 The filter accepts the following options.
17749 Set the output video dimension expression. Default value is the input
17752 If the @var{width} or @var{w} value is 0, the input width is used for
17753 the output. If the @var{height} or @var{h} value is 0, the input height
17754 is used for the output.
17756 If one and only one of the values is -n with n >= 1, the zscale filter
17757 will use a value that maintains the aspect ratio of the input image,
17758 calculated from the other specified dimension. After that it will,
17759 however, make sure that the calculated dimension is divisible by n and
17760 adjust the value if necessary.
17762 If both values are -n with n >= 1, the behavior will be identical to
17763 both values being set to 0 as previously detailed.
17765 See below for the list of accepted constants for use in the dimension
17769 Set the video size. For the syntax of this option, check the
17770 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17773 Set the dither type.
17775 Possible values are:
17780 @item error_diffusion
17786 Set the resize filter type.
17788 Possible values are:
17798 Default is bilinear.
17801 Set the color range.
17803 Possible values are:
17810 Default is same as input.
17813 Set the color primaries.
17815 Possible values are:
17825 Default is same as input.
17828 Set the transfer characteristics.
17830 Possible values are:
17844 Default is same as input.
17847 Set the colorspace matrix.
17849 Possible value are:
17860 Default is same as input.
17863 Set the input color range.
17865 Possible values are:
17872 Default is same as input.
17874 @item primariesin, pin
17875 Set the input color primaries.
17877 Possible values are:
17887 Default is same as input.
17889 @item transferin, tin
17890 Set the input transfer characteristics.
17892 Possible values are:
17903 Default is same as input.
17905 @item matrixin, min
17906 Set the input colorspace matrix.
17908 Possible value are:
17920 Set the output chroma location.
17922 Possible values are:
17933 @item chromalin, cin
17934 Set the input chroma location.
17936 Possible values are:
17948 Set the nominal peak luminance.
17951 The values of the @option{w} and @option{h} options are expressions
17952 containing the following constants:
17957 The input width and height
17961 These are the same as @var{in_w} and @var{in_h}.
17965 The output (scaled) width and height
17969 These are the same as @var{out_w} and @var{out_h}
17972 The same as @var{iw} / @var{ih}
17975 input sample aspect ratio
17978 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17982 horizontal and vertical input chroma subsample values. For example for the
17983 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17987 horizontal and vertical output chroma subsample values. For example for the
17988 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17994 @c man end VIDEO FILTERS
17996 @chapter Video Sources
17997 @c man begin VIDEO SOURCES
17999 Below is a description of the currently available video sources.
18003 Buffer video frames, and make them available to the filter chain.
18005 This source is mainly intended for a programmatic use, in particular
18006 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
18008 It accepts the following parameters:
18013 Specify the size (width and height) of the buffered video frames. For the
18014 syntax of this option, check the
18015 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18018 The input video width.
18021 The input video height.
18024 A string representing the pixel format of the buffered video frames.
18025 It may be a number corresponding to a pixel format, or a pixel format
18029 Specify the timebase assumed by the timestamps of the buffered frames.
18032 Specify the frame rate expected for the video stream.
18034 @item pixel_aspect, sar
18035 The sample (pixel) aspect ratio of the input video.
18038 Specify the optional parameters to be used for the scale filter which
18039 is automatically inserted when an input change is detected in the
18040 input size or format.
18042 @item hw_frames_ctx
18043 When using a hardware pixel format, this should be a reference to an
18044 AVHWFramesContext describing input frames.
18049 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
18052 will instruct the source to accept video frames with size 320x240 and
18053 with format "yuv410p", assuming 1/24 as the timestamps timebase and
18054 square pixels (1:1 sample aspect ratio).
18055 Since the pixel format with name "yuv410p" corresponds to the number 6
18056 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
18057 this example corresponds to:
18059 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
18062 Alternatively, the options can be specified as a flat string, but this
18063 syntax is deprecated:
18065 @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}]
18069 Create a pattern generated by an elementary cellular automaton.
18071 The initial state of the cellular automaton can be defined through the
18072 @option{filename} and @option{pattern} options. If such options are
18073 not specified an initial state is created randomly.
18075 At each new frame a new row in the video is filled with the result of
18076 the cellular automaton next generation. The behavior when the whole
18077 frame is filled is defined by the @option{scroll} option.
18079 This source accepts the following options:
18083 Read the initial cellular automaton state, i.e. the starting row, from
18084 the specified file.
18085 In the file, each non-whitespace character is considered an alive
18086 cell, a newline will terminate the row, and further characters in the
18087 file will be ignored.
18090 Read the initial cellular automaton state, i.e. the starting row, from
18091 the specified string.
18093 Each non-whitespace character in the string is considered an alive
18094 cell, a newline will terminate the row, and further characters in the
18095 string will be ignored.
18098 Set the video rate, that is the number of frames generated per second.
18101 @item random_fill_ratio, ratio
18102 Set the random fill ratio for the initial cellular automaton row. It
18103 is a floating point number value ranging from 0 to 1, defaults to
18106 This option is ignored when a file or a pattern is specified.
18108 @item random_seed, seed
18109 Set the seed for filling randomly the initial row, must be an integer
18110 included between 0 and UINT32_MAX. If not specified, or if explicitly
18111 set to -1, the filter will try to use a good random seed on a best
18115 Set the cellular automaton rule, it is a number ranging from 0 to 255.
18116 Default value is 110.
18119 Set the size of the output video. For the syntax of this option, check the
18120 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18122 If @option{filename} or @option{pattern} is specified, the size is set
18123 by default to the width of the specified initial state row, and the
18124 height is set to @var{width} * PHI.
18126 If @option{size} is set, it must contain the width of the specified
18127 pattern string, and the specified pattern will be centered in the
18130 If a filename or a pattern string is not specified, the size value
18131 defaults to "320x518" (used for a randomly generated initial state).
18134 If set to 1, scroll the output upward when all the rows in the output
18135 have been already filled. If set to 0, the new generated row will be
18136 written over the top row just after the bottom row is filled.
18139 @item start_full, full
18140 If set to 1, completely fill the output with generated rows before
18141 outputting the first frame.
18142 This is the default behavior, for disabling set the value to 0.
18145 If set to 1, stitch the left and right row edges together.
18146 This is the default behavior, for disabling set the value to 0.
18149 @subsection Examples
18153 Read the initial state from @file{pattern}, and specify an output of
18156 cellauto=f=pattern:s=200x400
18160 Generate a random initial row with a width of 200 cells, with a fill
18163 cellauto=ratio=2/3:s=200x200
18167 Create a pattern generated by rule 18 starting by a single alive cell
18168 centered on an initial row with width 100:
18170 cellauto=p=@@:s=100x400:full=0:rule=18
18174 Specify a more elaborated initial pattern:
18176 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
18181 @anchor{coreimagesrc}
18182 @section coreimagesrc
18183 Video source generated on GPU using Apple's CoreImage API on OSX.
18185 This video source is a specialized version of the @ref{coreimage} video filter.
18186 Use a core image generator at the beginning of the applied filterchain to
18187 generate the content.
18189 The coreimagesrc video source accepts the following options:
18191 @item list_generators
18192 List all available generators along with all their respective options as well as
18193 possible minimum and maximum values along with the default values.
18195 list_generators=true
18199 Specify the size of the sourced video. For the syntax of this option, check the
18200 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18201 The default value is @code{320x240}.
18204 Specify the frame rate of the sourced video, as the number of frames
18205 generated per second. It has to be a string in the format
18206 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
18207 number or a valid video frame rate abbreviation. The default value is
18211 Set the sample aspect ratio of the sourced video.
18214 Set the duration of the sourced video. See
18215 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18216 for the accepted syntax.
18218 If not specified, or the expressed duration is negative, the video is
18219 supposed to be generated forever.
18222 Additionally, all options of the @ref{coreimage} video filter are accepted.
18223 A complete filterchain can be used for further processing of the
18224 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
18225 and examples for details.
18227 @subsection Examples
18232 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
18233 given as complete and escaped command-line for Apple's standard bash shell:
18235 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
18237 This example is equivalent to the QRCode example of @ref{coreimage} without the
18238 need for a nullsrc video source.
18242 @section mandelbrot
18244 Generate a Mandelbrot set fractal, and progressively zoom towards the
18245 point specified with @var{start_x} and @var{start_y}.
18247 This source accepts the following options:
18252 Set the terminal pts value. Default value is 400.
18255 Set the terminal scale value.
18256 Must be a floating point value. Default value is 0.3.
18259 Set the inner coloring mode, that is the algorithm used to draw the
18260 Mandelbrot fractal internal region.
18262 It shall assume one of the following values:
18267 Show time until convergence.
18269 Set color based on point closest to the origin of the iterations.
18274 Default value is @var{mincol}.
18277 Set the bailout value. Default value is 10.0.
18280 Set the maximum of iterations performed by the rendering
18281 algorithm. Default value is 7189.
18284 Set outer coloring mode.
18285 It shall assume one of following values:
18287 @item iteration_count
18288 Set iteration cound mode.
18289 @item normalized_iteration_count
18290 set normalized iteration count mode.
18292 Default value is @var{normalized_iteration_count}.
18295 Set frame rate, expressed as number of frames per second. Default
18299 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
18300 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
18303 Set the initial scale value. Default value is 3.0.
18306 Set the initial x position. Must be a floating point value between
18307 -100 and 100. Default value is -0.743643887037158704752191506114774.
18310 Set the initial y position. Must be a floating point value between
18311 -100 and 100. Default value is -0.131825904205311970493132056385139.
18316 Generate various test patterns, as generated by the MPlayer test filter.
18318 The size of the generated video is fixed, and is 256x256.
18319 This source is useful in particular for testing encoding features.
18321 This source accepts the following options:
18326 Specify the frame rate of the sourced video, as the number of frames
18327 generated per second. It has to be a string in the format
18328 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
18329 number or a valid video frame rate abbreviation. The default value is
18333 Set the duration of the sourced video. See
18334 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18335 for the accepted syntax.
18337 If not specified, or the expressed duration is negative, the video is
18338 supposed to be generated forever.
18342 Set the number or the name of the test to perform. Supported tests are:
18358 Default value is "all", which will cycle through the list of all tests.
18363 mptestsrc=t=dc_luma
18366 will generate a "dc_luma" test pattern.
18368 @section frei0r_src
18370 Provide a frei0r source.
18372 To enable compilation of this filter you need to install the frei0r
18373 header and configure FFmpeg with @code{--enable-frei0r}.
18375 This source accepts the following parameters:
18380 The size of the video to generate. For the syntax of this option, check the
18381 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18384 The framerate of the generated video. It may be a string of the form
18385 @var{num}/@var{den} or a frame rate abbreviation.
18388 The name to the frei0r source to load. For more information regarding frei0r and
18389 how to set the parameters, read the @ref{frei0r} section in the video filters
18392 @item filter_params
18393 A '|'-separated list of parameters to pass to the frei0r source.
18397 For example, to generate a frei0r partik0l source with size 200x200
18398 and frame rate 10 which is overlaid on the overlay filter main input:
18400 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
18405 Generate a life pattern.
18407 This source is based on a generalization of John Conway's life game.
18409 The sourced input represents a life grid, each pixel represents a cell
18410 which can be in one of two possible states, alive or dead. Every cell
18411 interacts with its eight neighbours, which are the cells that are
18412 horizontally, vertically, or diagonally adjacent.
18414 At each interaction the grid evolves according to the adopted rule,
18415 which specifies the number of neighbor alive cells which will make a
18416 cell stay alive or born. The @option{rule} option allows one to specify
18419 This source accepts the following options:
18423 Set the file from which to read the initial grid state. In the file,
18424 each non-whitespace character is considered an alive cell, and newline
18425 is used to delimit the end of each row.
18427 If this option is not specified, the initial grid is generated
18431 Set the video rate, that is the number of frames generated per second.
18434 @item random_fill_ratio, ratio
18435 Set the random fill ratio for the initial random grid. It is a
18436 floating point number value ranging from 0 to 1, defaults to 1/PHI.
18437 It is ignored when a file is specified.
18439 @item random_seed, seed
18440 Set the seed for filling the initial random grid, must be an integer
18441 included between 0 and UINT32_MAX. If not specified, or if explicitly
18442 set to -1, the filter will try to use a good random seed on a best
18448 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
18449 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
18450 @var{NS} specifies the number of alive neighbor cells which make a
18451 live cell stay alive, and @var{NB} the number of alive neighbor cells
18452 which make a dead cell to become alive (i.e. to "born").
18453 "s" and "b" can be used in place of "S" and "B", respectively.
18455 Alternatively a rule can be specified by an 18-bits integer. The 9
18456 high order bits are used to encode the next cell state if it is alive
18457 for each number of neighbor alive cells, the low order bits specify
18458 the rule for "borning" new cells. Higher order bits encode for an
18459 higher number of neighbor cells.
18460 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
18461 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
18463 Default value is "S23/B3", which is the original Conway's game of life
18464 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
18465 cells, and will born a new cell if there are three alive cells around
18469 Set the size of the output video. For the syntax of this option, check the
18470 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18472 If @option{filename} is specified, the size is set by default to the
18473 same size of the input file. If @option{size} is set, it must contain
18474 the size specified in the input file, and the initial grid defined in
18475 that file is centered in the larger resulting area.
18477 If a filename is not specified, the size value defaults to "320x240"
18478 (used for a randomly generated initial grid).
18481 If set to 1, stitch the left and right grid edges together, and the
18482 top and bottom edges also. Defaults to 1.
18485 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
18486 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
18487 value from 0 to 255.
18490 Set the color of living (or new born) cells.
18493 Set the color of dead cells. If @option{mold} is set, this is the first color
18494 used to represent a dead cell.
18497 Set mold color, for definitely dead and moldy cells.
18499 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
18500 ffmpeg-utils manual,ffmpeg-utils}.
18503 @subsection Examples
18507 Read a grid from @file{pattern}, and center it on a grid of size
18510 life=f=pattern:s=300x300
18514 Generate a random grid of size 200x200, with a fill ratio of 2/3:
18516 life=ratio=2/3:s=200x200
18520 Specify a custom rule for evolving a randomly generated grid:
18526 Full example with slow death effect (mold) using @command{ffplay}:
18528 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
18535 @anchor{haldclutsrc}
18538 @anchor{pal100bars}
18539 @anchor{rgbtestsrc}
18541 @anchor{smptehdbars}
18544 @anchor{yuvtestsrc}
18545 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
18547 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
18549 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
18551 The @code{color} source provides an uniformly colored input.
18553 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
18554 @ref{haldclut} filter.
18556 The @code{nullsrc} source returns unprocessed video frames. It is
18557 mainly useful to be employed in analysis / debugging tools, or as the
18558 source for filters which ignore the input data.
18560 The @code{pal75bars} source generates a color bars pattern, based on
18561 EBU PAL recommendations with 75% color levels.
18563 The @code{pal100bars} source generates a color bars pattern, based on
18564 EBU PAL recommendations with 100% color levels.
18566 The @code{rgbtestsrc} source generates an RGB test pattern useful for
18567 detecting RGB vs BGR issues. You should see a red, green and blue
18568 stripe from top to bottom.
18570 The @code{smptebars} source generates a color bars pattern, based on
18571 the SMPTE Engineering Guideline EG 1-1990.
18573 The @code{smptehdbars} source generates a color bars pattern, based on
18574 the SMPTE RP 219-2002.
18576 The @code{testsrc} source generates a test video pattern, showing a
18577 color pattern, a scrolling gradient and a timestamp. This is mainly
18578 intended for testing purposes.
18580 The @code{testsrc2} source is similar to testsrc, but supports more
18581 pixel formats instead of just @code{rgb24}. This allows using it as an
18582 input for other tests without requiring a format conversion.
18584 The @code{yuvtestsrc} source generates an YUV test pattern. You should
18585 see a y, cb and cr stripe from top to bottom.
18587 The sources accept the following parameters:
18592 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
18593 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
18594 pixels to be used as identity matrix for 3D lookup tables. Each component is
18595 coded on a @code{1/(N*N)} scale.
18598 Specify the color of the source, only available in the @code{color}
18599 source. For the syntax of this option, check the
18600 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18603 Specify the size of the sourced video. For the syntax of this option, check the
18604 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18605 The default value is @code{320x240}.
18607 This option is not available with the @code{allrgb}, @code{allyuv}, and
18608 @code{haldclutsrc} filters.
18611 Specify the frame rate of the sourced video, as the number of frames
18612 generated per second. It has to be a string in the format
18613 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
18614 number or a valid video frame rate abbreviation. The default value is
18618 Set the duration of the sourced video. See
18619 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18620 for the accepted syntax.
18622 If not specified, or the expressed duration is negative, the video is
18623 supposed to be generated forever.
18626 Set the sample aspect ratio of the sourced video.
18629 Specify the alpha (opacity) of the background, only available in the
18630 @code{testsrc2} source. The value must be between 0 (fully transparent) and
18631 255 (fully opaque, the default).
18634 Set the number of decimals to show in the timestamp, only available in the
18635 @code{testsrc} source.
18637 The displayed timestamp value will correspond to the original
18638 timestamp value multiplied by the power of 10 of the specified
18639 value. Default value is 0.
18642 @subsection Examples
18646 Generate a video with a duration of 5.3 seconds, with size
18647 176x144 and a frame rate of 10 frames per second:
18649 testsrc=duration=5.3:size=qcif:rate=10
18653 The following graph description will generate a red source
18654 with an opacity of 0.2, with size "qcif" and a frame rate of 10
18657 color=c=red@@0.2:s=qcif:r=10
18661 If the input content is to be ignored, @code{nullsrc} can be used. The
18662 following command generates noise in the luminance plane by employing
18663 the @code{geq} filter:
18665 nullsrc=s=256x256, geq=random(1)*255:128:128
18669 @subsection Commands
18671 The @code{color} source supports the following commands:
18675 Set the color of the created image. Accepts the same syntax of the
18676 corresponding @option{color} option.
18681 Generate video using an OpenCL program.
18686 OpenCL program source file.
18689 Kernel name in program.
18692 Size of frames to generate. This must be set.
18695 Pixel format to use for the generated frames. This must be set.
18698 Number of frames generated every second. Default value is '25'.
18702 For details of how the program loading works, see the @ref{program_opencl}
18709 Generate a colour ramp by setting pixel values from the position of the pixel
18710 in the output image. (Note that this will work with all pixel formats, but
18711 the generated output will not be the same.)
18713 __kernel void ramp(__write_only image2d_t dst,
18714 unsigned int index)
18716 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18719 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
18721 write_imagef(dst, loc, val);
18726 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
18728 __kernel void sierpinski_carpet(__write_only image2d_t dst,
18729 unsigned int index)
18731 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18733 float4 value = 0.0f;
18734 int x = loc.x + index;
18735 int y = loc.y + index;
18736 while (x > 0 || y > 0) {
18737 if (x % 3 == 1 && y % 3 == 1) {
18745 write_imagef(dst, loc, value);
18751 @c man end VIDEO SOURCES
18753 @chapter Video Sinks
18754 @c man begin VIDEO SINKS
18756 Below is a description of the currently available video sinks.
18758 @section buffersink
18760 Buffer video frames, and make them available to the end of the filter
18763 This sink is mainly intended for programmatic use, in particular
18764 through the interface defined in @file{libavfilter/buffersink.h}
18765 or the options system.
18767 It accepts a pointer to an AVBufferSinkContext structure, which
18768 defines the incoming buffers' formats, to be passed as the opaque
18769 parameter to @code{avfilter_init_filter} for initialization.
18773 Null video sink: do absolutely nothing with the input video. It is
18774 mainly useful as a template and for use in analysis / debugging
18777 @c man end VIDEO SINKS
18779 @chapter Multimedia Filters
18780 @c man begin MULTIMEDIA FILTERS
18782 Below is a description of the currently available multimedia filters.
18786 Convert input audio to a video output, displaying the audio bit scope.
18788 The filter accepts the following options:
18792 Set frame rate, expressed as number of frames per second. Default
18796 Specify the video size for the output. For the syntax of this option, check the
18797 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18798 Default value is @code{1024x256}.
18801 Specify list of colors separated by space or by '|' which will be used to
18802 draw channels. Unrecognized or missing colors will be replaced
18806 @section ahistogram
18808 Convert input audio to a video output, displaying the volume histogram.
18810 The filter accepts the following options:
18814 Specify how histogram is calculated.
18816 It accepts the following values:
18819 Use single histogram for all channels.
18821 Use separate histogram for each channel.
18823 Default is @code{single}.
18826 Set frame rate, expressed as number of frames per second. Default
18830 Specify the video size for the output. For the syntax of this option, check the
18831 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18832 Default value is @code{hd720}.
18837 It accepts the following values:
18848 reverse logarithmic
18850 Default is @code{log}.
18853 Set amplitude scale.
18855 It accepts the following values:
18862 Default is @code{log}.
18865 Set how much frames to accumulate in histogram.
18866 Defauls is 1. Setting this to -1 accumulates all frames.
18869 Set histogram ratio of window height.
18872 Set sonogram sliding.
18874 It accepts the following values:
18877 replace old rows with new ones.
18879 scroll from top to bottom.
18881 Default is @code{replace}.
18884 @section aphasemeter
18886 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
18887 representing mean phase of current audio frame. A video output can also be produced and is
18888 enabled by default. The audio is passed through as first output.
18890 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
18891 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
18892 and @code{1} means channels are in phase.
18894 The filter accepts the following options, all related to its video output:
18898 Set the output frame rate. Default value is @code{25}.
18901 Set the video size for the output. For the syntax of this option, check the
18902 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18903 Default value is @code{800x400}.
18908 Specify the red, green, blue contrast. Default values are @code{2},
18909 @code{7} and @code{1}.
18910 Allowed range is @code{[0, 255]}.
18913 Set color which will be used for drawing median phase. If color is
18914 @code{none} which is default, no median phase value will be drawn.
18917 Enable video output. Default is enabled.
18920 @section avectorscope
18922 Convert input audio to a video output, representing the audio vector
18925 The filter is used to measure the difference between channels of stereo
18926 audio stream. A monoaural signal, consisting of identical left and right
18927 signal, results in straight vertical line. Any stereo separation is visible
18928 as a deviation from this line, creating a Lissajous figure.
18929 If the straight (or deviation from it) but horizontal line appears this
18930 indicates that the left and right channels are out of phase.
18932 The filter accepts the following options:
18936 Set the vectorscope mode.
18938 Available values are:
18941 Lissajous rotated by 45 degrees.
18944 Same as above but not rotated.
18947 Shape resembling half of circle.
18950 Default value is @samp{lissajous}.
18953 Set the video size for the output. For the syntax of this option, check the
18954 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18955 Default value is @code{400x400}.
18958 Set the output frame rate. Default value is @code{25}.
18964 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18965 @code{160}, @code{80} and @code{255}.
18966 Allowed range is @code{[0, 255]}.
18972 Specify the red, green, blue and alpha fade. Default values are @code{15},
18973 @code{10}, @code{5} and @code{5}.
18974 Allowed range is @code{[0, 255]}.
18977 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18978 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18981 Set the vectorscope drawing mode.
18983 Available values are:
18986 Draw dot for each sample.
18989 Draw line between previous and current sample.
18992 Default value is @samp{dot}.
18995 Specify amplitude scale of audio samples.
18997 Available values are:
19013 Swap left channel axis with right channel axis.
19023 Mirror only x axis.
19026 Mirror only y axis.
19034 @subsection Examples
19038 Complete example using @command{ffplay}:
19040 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19041 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
19045 @section bench, abench
19047 Benchmark part of a filtergraph.
19049 The filter accepts the following options:
19053 Start or stop a timer.
19055 Available values are:
19058 Get the current time, set it as frame metadata (using the key
19059 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
19062 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
19063 the input frame metadata to get the time difference. Time difference, average,
19064 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
19065 @code{min}) are then printed. The timestamps are expressed in seconds.
19069 @subsection Examples
19073 Benchmark @ref{selectivecolor} filter:
19075 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
19081 Concatenate audio and video streams, joining them together one after the
19084 The filter works on segments of synchronized video and audio streams. All
19085 segments must have the same number of streams of each type, and that will
19086 also be the number of streams at output.
19088 The filter accepts the following options:
19093 Set the number of segments. Default is 2.
19096 Set the number of output video streams, that is also the number of video
19097 streams in each segment. Default is 1.
19100 Set the number of output audio streams, that is also the number of audio
19101 streams in each segment. Default is 0.
19104 Activate unsafe mode: do not fail if segments have a different format.
19108 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
19109 @var{a} audio outputs.
19111 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
19112 segment, in the same order as the outputs, then the inputs for the second
19115 Related streams do not always have exactly the same duration, for various
19116 reasons including codec frame size or sloppy authoring. For that reason,
19117 related synchronized streams (e.g. a video and its audio track) should be
19118 concatenated at once. The concat filter will use the duration of the longest
19119 stream in each segment (except the last one), and if necessary pad shorter
19120 audio streams with silence.
19122 For this filter to work correctly, all segments must start at timestamp 0.
19124 All corresponding streams must have the same parameters in all segments; the
19125 filtering system will automatically select a common pixel format for video
19126 streams, and a common sample format, sample rate and channel layout for
19127 audio streams, but other settings, such as resolution, must be converted
19128 explicitly by the user.
19130 Different frame rates are acceptable but will result in variable frame rate
19131 at output; be sure to configure the output file to handle it.
19133 @subsection Examples
19137 Concatenate an opening, an episode and an ending, all in bilingual version
19138 (video in stream 0, audio in streams 1 and 2):
19140 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
19141 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
19142 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
19143 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
19147 Concatenate two parts, handling audio and video separately, using the
19148 (a)movie sources, and adjusting the resolution:
19150 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
19151 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
19152 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
19154 Note that a desync will happen at the stitch if the audio and video streams
19155 do not have exactly the same duration in the first file.
19159 @subsection Commands
19161 This filter supports the following commands:
19164 Close the current segment and step to the next one
19167 @section drawgraph, adrawgraph
19169 Draw a graph using input video or audio metadata.
19171 It accepts the following parameters:
19175 Set 1st frame metadata key from which metadata values will be used to draw a graph.
19178 Set 1st foreground color expression.
19181 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
19184 Set 2nd foreground color expression.
19187 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
19190 Set 3rd foreground color expression.
19193 Set 4th frame metadata key from which metadata values will be used to draw a graph.
19196 Set 4th foreground color expression.
19199 Set minimal value of metadata value.
19202 Set maximal value of metadata value.
19205 Set graph background color. Default is white.
19210 Available values for mode is:
19217 Default is @code{line}.
19222 Available values for slide is:
19225 Draw new frame when right border is reached.
19228 Replace old columns with new ones.
19231 Scroll from right to left.
19234 Scroll from left to right.
19237 Draw single picture.
19240 Default is @code{frame}.
19243 Set size of graph video. For the syntax of this option, check the
19244 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19245 The default value is @code{900x256}.
19247 The foreground color expressions can use the following variables:
19250 Minimal value of metadata value.
19253 Maximal value of metadata value.
19256 Current metadata key value.
19259 The color is defined as 0xAABBGGRR.
19262 Example using metadata from @ref{signalstats} filter:
19264 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
19267 Example using metadata from @ref{ebur128} filter:
19269 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
19275 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
19276 it unchanged. By default, it logs a message at a frequency of 10Hz with the
19277 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
19278 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
19280 The filter also has a video output (see the @var{video} option) with a real
19281 time graph to observe the loudness evolution. The graphic contains the logged
19282 message mentioned above, so it is not printed anymore when this option is set,
19283 unless the verbose logging is set. The main graphing area contains the
19284 short-term loudness (3 seconds of analysis), and the gauge on the right is for
19285 the momentary loudness (400 milliseconds), but can optionally be configured
19286 to instead display short-term loudness (see @var{gauge}).
19288 The green area marks a +/- 1LU target range around the target loudness
19289 (-23LUFS by default, unless modified through @var{target}).
19291 More information about the Loudness Recommendation EBU R128 on
19292 @url{http://tech.ebu.ch/loudness}.
19294 The filter accepts the following options:
19299 Activate the video output. The audio stream is passed unchanged whether this
19300 option is set or no. The video stream will be the first output stream if
19301 activated. Default is @code{0}.
19304 Set the video size. This option is for video only. For the syntax of this
19306 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19307 Default and minimum resolution is @code{640x480}.
19310 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
19311 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
19312 other integer value between this range is allowed.
19315 Set metadata injection. If set to @code{1}, the audio input will be segmented
19316 into 100ms output frames, each of them containing various loudness information
19317 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
19319 Default is @code{0}.
19322 Force the frame logging level.
19324 Available values are:
19327 information logging level
19329 verbose logging level
19332 By default, the logging level is set to @var{info}. If the @option{video} or
19333 the @option{metadata} options are set, it switches to @var{verbose}.
19338 Available modes can be cumulated (the option is a @code{flag} type). Possible
19342 Disable any peak mode (default).
19344 Enable sample-peak mode.
19346 Simple peak mode looking for the higher sample value. It logs a message
19347 for sample-peak (identified by @code{SPK}).
19349 Enable true-peak mode.
19351 If enabled, the peak lookup is done on an over-sampled version of the input
19352 stream for better peak accuracy. It logs a message for true-peak.
19353 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
19354 This mode requires a build with @code{libswresample}.
19358 Treat mono input files as "dual mono". If a mono file is intended for playback
19359 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
19360 If set to @code{true}, this option will compensate for this effect.
19361 Multi-channel input files are not affected by this option.
19364 Set a specific pan law to be used for the measurement of dual mono files.
19365 This parameter is optional, and has a default value of -3.01dB.
19368 Set a specific target level (in LUFS) used as relative zero in the visualization.
19369 This parameter is optional and has a default value of -23LUFS as specified
19370 by EBU R128. However, material published online may prefer a level of -16LUFS
19371 (e.g. for use with podcasts or video platforms).
19374 Set the value displayed by the gauge. Valid values are @code{momentary} and s
19375 @code{shortterm}. By default the momentary value will be used, but in certain
19376 scenarios it may be more useful to observe the short term value instead (e.g.
19380 Sets the display scale for the loudness. Valid parameters are @code{absolute}
19381 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
19382 video output, not the summary or continuous log output.
19385 @subsection Examples
19389 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
19391 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
19395 Run an analysis with @command{ffmpeg}:
19397 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
19401 @section interleave, ainterleave
19403 Temporally interleave frames from several inputs.
19405 @code{interleave} works with video inputs, @code{ainterleave} with audio.
19407 These filters read frames from several inputs and send the oldest
19408 queued frame to the output.
19410 Input streams must have well defined, monotonically increasing frame
19413 In order to submit one frame to output, these filters need to enqueue
19414 at least one frame for each input, so they cannot work in case one
19415 input is not yet terminated and will not receive incoming frames.
19417 For example consider the case when one input is a @code{select} filter
19418 which always drops input frames. The @code{interleave} filter will keep
19419 reading from that input, but it will never be able to send new frames
19420 to output until the input sends an end-of-stream signal.
19422 Also, depending on inputs synchronization, the filters will drop
19423 frames in case one input receives more frames than the other ones, and
19424 the queue is already filled.
19426 These filters accept the following options:
19430 Set the number of different inputs, it is 2 by default.
19433 @subsection Examples
19437 Interleave frames belonging to different streams using @command{ffmpeg}:
19439 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
19443 Add flickering blur effect:
19445 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
19449 @section metadata, ametadata
19451 Manipulate frame metadata.
19453 This filter accepts the following options:
19457 Set mode of operation of the filter.
19459 Can be one of the following:
19463 If both @code{value} and @code{key} is set, select frames
19464 which have such metadata. If only @code{key} is set, select
19465 every frame that has such key in metadata.
19468 Add new metadata @code{key} and @code{value}. If key is already available
19472 Modify value of already present key.
19475 If @code{value} is set, delete only keys that have such value.
19476 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
19480 Print key and its value if metadata was found. If @code{key} is not set print all
19481 metadata values available in frame.
19485 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
19488 Set metadata value which will be used. This option is mandatory for
19489 @code{modify} and @code{add} mode.
19492 Which function to use when comparing metadata value and @code{value}.
19494 Can be one of following:
19498 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
19501 Values are interpreted as strings, returns true if metadata value starts with
19502 the @code{value} option string.
19505 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
19508 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
19511 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
19514 Values are interpreted as floats, returns true if expression from option @code{expr}
19519 Set expression which is used when @code{function} is set to @code{expr}.
19520 The expression is evaluated through the eval API and can contain the following
19525 Float representation of @code{value} from metadata key.
19528 Float representation of @code{value} as supplied by user in @code{value} option.
19532 If specified in @code{print} mode, output is written to the named file. Instead of
19533 plain filename any writable url can be specified. Filename ``-'' is a shorthand
19534 for standard output. If @code{file} option is not set, output is written to the log
19535 with AV_LOG_INFO loglevel.
19539 @subsection Examples
19543 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
19546 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
19549 Print silencedetect output to file @file{metadata.txt}.
19551 silencedetect,ametadata=mode=print:file=metadata.txt
19554 Direct all metadata to a pipe with file descriptor 4.
19556 metadata=mode=print:file='pipe\:4'
19560 @section perms, aperms
19562 Set read/write permissions for the output frames.
19564 These filters are mainly aimed at developers to test direct path in the
19565 following filter in the filtergraph.
19567 The filters accept the following options:
19571 Select the permissions mode.
19573 It accepts the following values:
19576 Do nothing. This is the default.
19578 Set all the output frames read-only.
19580 Set all the output frames directly writable.
19582 Make the frame read-only if writable, and writable if read-only.
19584 Set each output frame read-only or writable randomly.
19588 Set the seed for the @var{random} mode, must be an integer included between
19589 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
19590 @code{-1}, the filter will try to use a good random seed on a best effort
19594 Note: in case of auto-inserted filter between the permission filter and the
19595 following one, the permission might not be received as expected in that
19596 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
19597 perms/aperms filter can avoid this problem.
19599 @section realtime, arealtime
19601 Slow down filtering to match real time approximately.
19603 These filters will pause the filtering for a variable amount of time to
19604 match the output rate with the input timestamps.
19605 They are similar to the @option{re} option to @code{ffmpeg}.
19607 They accept the following options:
19611 Time limit for the pauses. Any pause longer than that will be considered
19612 a timestamp discontinuity and reset the timer. Default is 2 seconds.
19616 @section select, aselect
19618 Select frames to pass in output.
19620 This filter accepts the following options:
19625 Set expression, which is evaluated for each input frame.
19627 If the expression is evaluated to zero, the frame is discarded.
19629 If the evaluation result is negative or NaN, the frame is sent to the
19630 first output; otherwise it is sent to the output with index
19631 @code{ceil(val)-1}, assuming that the input index starts from 0.
19633 For example a value of @code{1.2} corresponds to the output with index
19634 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
19637 Set the number of outputs. The output to which to send the selected
19638 frame is based on the result of the evaluation. Default value is 1.
19641 The expression can contain the following constants:
19645 The (sequential) number of the filtered frame, starting from 0.
19648 The (sequential) number of the selected frame, starting from 0.
19650 @item prev_selected_n
19651 The sequential number of the last selected frame. It's NAN if undefined.
19654 The timebase of the input timestamps.
19657 The PTS (Presentation TimeStamp) of the filtered video frame,
19658 expressed in @var{TB} units. It's NAN if undefined.
19661 The PTS of the filtered video frame,
19662 expressed in seconds. It's NAN if undefined.
19665 The PTS of the previously filtered video frame. It's NAN if undefined.
19667 @item prev_selected_pts
19668 The PTS of the last previously filtered video frame. It's NAN if undefined.
19670 @item prev_selected_t
19671 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
19674 The PTS of the first video frame in the video. It's NAN if undefined.
19677 The time of the first video frame in the video. It's NAN if undefined.
19679 @item pict_type @emph{(video only)}
19680 The type of the filtered frame. It can assume one of the following
19692 @item interlace_type @emph{(video only)}
19693 The frame interlace type. It can assume one of the following values:
19696 The frame is progressive (not interlaced).
19698 The frame is top-field-first.
19700 The frame is bottom-field-first.
19703 @item consumed_sample_n @emph{(audio only)}
19704 the number of selected samples before the current frame
19706 @item samples_n @emph{(audio only)}
19707 the number of samples in the current frame
19709 @item sample_rate @emph{(audio only)}
19710 the input sample rate
19713 This is 1 if the filtered frame is a key-frame, 0 otherwise.
19716 the position in the file of the filtered frame, -1 if the information
19717 is not available (e.g. for synthetic video)
19719 @item scene @emph{(video only)}
19720 value between 0 and 1 to indicate a new scene; a low value reflects a low
19721 probability for the current frame to introduce a new scene, while a higher
19722 value means the current frame is more likely to be one (see the example below)
19724 @item concatdec_select
19725 The concat demuxer can select only part of a concat input file by setting an
19726 inpoint and an outpoint, but the output packets may not be entirely contained
19727 in the selected interval. By using this variable, it is possible to skip frames
19728 generated by the concat demuxer which are not exactly contained in the selected
19731 This works by comparing the frame pts against the @var{lavf.concat.start_time}
19732 and the @var{lavf.concat.duration} packet metadata values which are also
19733 present in the decoded frames.
19735 The @var{concatdec_select} variable is -1 if the frame pts is at least
19736 start_time and either the duration metadata is missing or the frame pts is less
19737 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
19740 That basically means that an input frame is selected if its pts is within the
19741 interval set by the concat demuxer.
19745 The default value of the select expression is "1".
19747 @subsection Examples
19751 Select all frames in input:
19756 The example above is the same as:
19768 Select only I-frames:
19770 select='eq(pict_type\,I)'
19774 Select one frame every 100:
19776 select='not(mod(n\,100))'
19780 Select only frames contained in the 10-20 time interval:
19782 select=between(t\,10\,20)
19786 Select only I-frames contained in the 10-20 time interval:
19788 select=between(t\,10\,20)*eq(pict_type\,I)
19792 Select frames with a minimum distance of 10 seconds:
19794 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
19798 Use aselect to select only audio frames with samples number > 100:
19800 aselect='gt(samples_n\,100)'
19804 Create a mosaic of the first scenes:
19806 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
19809 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
19813 Send even and odd frames to separate outputs, and compose them:
19815 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
19819 Select useful frames from an ffconcat file which is using inpoints and
19820 outpoints but where the source files are not intra frame only.
19822 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
19826 @section sendcmd, asendcmd
19828 Send commands to filters in the filtergraph.
19830 These filters read commands to be sent to other filters in the
19833 @code{sendcmd} must be inserted between two video filters,
19834 @code{asendcmd} must be inserted between two audio filters, but apart
19835 from that they act the same way.
19837 The specification of commands can be provided in the filter arguments
19838 with the @var{commands} option, or in a file specified by the
19839 @var{filename} option.
19841 These filters accept the following options:
19844 Set the commands to be read and sent to the other filters.
19846 Set the filename of the commands to be read and sent to the other
19850 @subsection Commands syntax
19852 A commands description consists of a sequence of interval
19853 specifications, comprising a list of commands to be executed when a
19854 particular event related to that interval occurs. The occurring event
19855 is typically the current frame time entering or leaving a given time
19858 An interval is specified by the following syntax:
19860 @var{START}[-@var{END}] @var{COMMANDS};
19863 The time interval is specified by the @var{START} and @var{END} times.
19864 @var{END} is optional and defaults to the maximum time.
19866 The current frame time is considered within the specified interval if
19867 it is included in the interval [@var{START}, @var{END}), that is when
19868 the time is greater or equal to @var{START} and is lesser than
19871 @var{COMMANDS} consists of a sequence of one or more command
19872 specifications, separated by ",", relating to that interval. The
19873 syntax of a command specification is given by:
19875 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
19878 @var{FLAGS} is optional and specifies the type of events relating to
19879 the time interval which enable sending the specified command, and must
19880 be a non-null sequence of identifier flags separated by "+" or "|" and
19881 enclosed between "[" and "]".
19883 The following flags are recognized:
19886 The command is sent when the current frame timestamp enters the
19887 specified interval. In other words, the command is sent when the
19888 previous frame timestamp was not in the given interval, and the
19892 The command is sent when the current frame timestamp leaves the
19893 specified interval. In other words, the command is sent when the
19894 previous frame timestamp was in the given interval, and the
19898 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
19901 @var{TARGET} specifies the target of the command, usually the name of
19902 the filter class or a specific filter instance name.
19904 @var{COMMAND} specifies the name of the command for the target filter.
19906 @var{ARG} is optional and specifies the optional list of argument for
19907 the given @var{COMMAND}.
19909 Between one interval specification and another, whitespaces, or
19910 sequences of characters starting with @code{#} until the end of line,
19911 are ignored and can be used to annotate comments.
19913 A simplified BNF description of the commands specification syntax
19916 @var{COMMAND_FLAG} ::= "enter" | "leave"
19917 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
19918 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
19919 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
19920 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
19921 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
19924 @subsection Examples
19928 Specify audio tempo change at second 4:
19930 asendcmd=c='4.0 atempo tempo 1.5',atempo
19934 Target a specific filter instance:
19936 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
19940 Specify a list of drawtext and hue commands in a file.
19942 # show text in the interval 5-10
19943 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
19944 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
19946 # desaturate the image in the interval 15-20
19947 15.0-20.0 [enter] hue s 0,
19948 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
19950 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
19952 # apply an exponential saturation fade-out effect, starting from time 25
19953 25 [enter] hue s exp(25-t)
19956 A filtergraph allowing to read and process the above command list
19957 stored in a file @file{test.cmd}, can be specified with:
19959 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
19964 @section setpts, asetpts
19966 Change the PTS (presentation timestamp) of the input frames.
19968 @code{setpts} works on video frames, @code{asetpts} on audio frames.
19970 This filter accepts the following options:
19975 The expression which is evaluated for each frame to construct its timestamp.
19979 The expression is evaluated through the eval API and can contain the following
19983 @item FRAME_RATE, FR
19984 frame rate, only defined for constant frame-rate video
19987 The presentation timestamp in input
19990 The count of the input frame for video or the number of consumed samples,
19991 not including the current frame for audio, starting from 0.
19993 @item NB_CONSUMED_SAMPLES
19994 The number of consumed samples, not including the current frame (only
19997 @item NB_SAMPLES, S
19998 The number of samples in the current frame (only audio)
20000 @item SAMPLE_RATE, SR
20001 The audio sample rate.
20004 The PTS of the first frame.
20007 the time in seconds of the first frame
20010 State whether the current frame is interlaced.
20013 the time in seconds of the current frame
20016 original position in the file of the frame, or undefined if undefined
20017 for the current frame
20020 The previous input PTS.
20023 previous input time in seconds
20026 The previous output PTS.
20029 previous output time in seconds
20032 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
20036 The wallclock (RTC) time at the start of the movie in microseconds.
20039 The timebase of the input timestamps.
20043 @subsection Examples
20047 Start counting PTS from zero
20049 setpts=PTS-STARTPTS
20053 Apply fast motion effect:
20059 Apply slow motion effect:
20065 Set fixed rate of 25 frames per second:
20071 Set fixed rate 25 fps with some jitter:
20073 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
20077 Apply an offset of 10 seconds to the input PTS:
20083 Generate timestamps from a "live source" and rebase onto the current timebase:
20085 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
20089 Generate timestamps by counting samples:
20098 Force color range for the output video frame.
20100 The @code{setrange} filter marks the color range property for the
20101 output frames. It does not change the input frame, but only sets the
20102 corresponding property, which affects how the frame is treated by
20105 The filter accepts the following options:
20110 Available values are:
20114 Keep the same color range property.
20116 @item unspecified, unknown
20117 Set the color range as unspecified.
20119 @item limited, tv, mpeg
20120 Set the color range as limited.
20122 @item full, pc, jpeg
20123 Set the color range as full.
20127 @section settb, asettb
20129 Set the timebase to use for the output frames timestamps.
20130 It is mainly useful for testing timebase configuration.
20132 It accepts the following parameters:
20137 The expression which is evaluated into the output timebase.
20141 The value for @option{tb} is an arithmetic expression representing a
20142 rational. The expression can contain the constants "AVTB" (the default
20143 timebase), "intb" (the input timebase) and "sr" (the sample rate,
20144 audio only). Default value is "intb".
20146 @subsection Examples
20150 Set the timebase to 1/25:
20156 Set the timebase to 1/10:
20162 Set the timebase to 1001/1000:
20168 Set the timebase to 2*intb:
20174 Set the default timebase value:
20181 Convert input audio to a video output representing frequency spectrum
20182 logarithmically using Brown-Puckette constant Q transform algorithm with
20183 direct frequency domain coefficient calculation (but the transform itself
20184 is not really constant Q, instead the Q factor is actually variable/clamped),
20185 with musical tone scale, from E0 to D#10.
20187 The filter accepts the following options:
20191 Specify the video size for the output. It must be even. For the syntax of this option,
20192 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20193 Default value is @code{1920x1080}.
20196 Set the output frame rate. Default value is @code{25}.
20199 Set the bargraph height. It must be even. Default value is @code{-1} which
20200 computes the bargraph height automatically.
20203 Set the axis height. It must be even. Default value is @code{-1} which computes
20204 the axis height automatically.
20207 Set the sonogram height. It must be even. Default value is @code{-1} which
20208 computes the sonogram height automatically.
20211 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
20212 instead. Default value is @code{1}.
20214 @item sono_v, volume
20215 Specify the sonogram volume expression. It can contain variables:
20218 the @var{bar_v} evaluated expression
20219 @item frequency, freq, f
20220 the frequency where it is evaluated
20221 @item timeclamp, tc
20222 the value of @var{timeclamp} option
20226 @item a_weighting(f)
20227 A-weighting of equal loudness
20228 @item b_weighting(f)
20229 B-weighting of equal loudness
20230 @item c_weighting(f)
20231 C-weighting of equal loudness.
20233 Default value is @code{16}.
20235 @item bar_v, volume2
20236 Specify the bargraph volume expression. It can contain variables:
20239 the @var{sono_v} evaluated expression
20240 @item frequency, freq, f
20241 the frequency where it is evaluated
20242 @item timeclamp, tc
20243 the value of @var{timeclamp} option
20247 @item a_weighting(f)
20248 A-weighting of equal loudness
20249 @item b_weighting(f)
20250 B-weighting of equal loudness
20251 @item c_weighting(f)
20252 C-weighting of equal loudness.
20254 Default value is @code{sono_v}.
20256 @item sono_g, gamma
20257 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
20258 higher gamma makes the spectrum having more range. Default value is @code{3}.
20259 Acceptable range is @code{[1, 7]}.
20261 @item bar_g, gamma2
20262 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
20266 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
20267 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
20269 @item timeclamp, tc
20270 Specify the transform timeclamp. At low frequency, there is trade-off between
20271 accuracy in time domain and frequency domain. If timeclamp is lower,
20272 event in time domain is represented more accurately (such as fast bass drum),
20273 otherwise event in frequency domain is represented more accurately
20274 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
20277 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
20278 limits future samples by applying asymmetric windowing in time domain, useful
20279 when low latency is required. Accepted range is @code{[0, 1]}.
20282 Specify the transform base frequency. Default value is @code{20.01523126408007475},
20283 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
20286 Specify the transform end frequency. Default value is @code{20495.59681441799654},
20287 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
20290 This option is deprecated and ignored.
20293 Specify the transform length in time domain. Use this option to control accuracy
20294 trade-off between time domain and frequency domain at every frequency sample.
20295 It can contain variables:
20297 @item frequency, freq, f
20298 the frequency where it is evaluated
20299 @item timeclamp, tc
20300 the value of @var{timeclamp} option.
20302 Default value is @code{384*tc/(384+tc*f)}.
20305 Specify the transform count for every video frame. Default value is @code{6}.
20306 Acceptable range is @code{[1, 30]}.
20309 Specify the transform count for every single pixel. Default value is @code{0},
20310 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
20313 Specify font file for use with freetype to draw the axis. If not specified,
20314 use embedded font. Note that drawing with font file or embedded font is not
20315 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
20319 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
20320 The : in the pattern may be replaced by | to avoid unnecessary escaping.
20323 Specify font color expression. This is arithmetic expression that should return
20324 integer value 0xRRGGBB. It can contain variables:
20326 @item frequency, freq, f
20327 the frequency where it is evaluated
20328 @item timeclamp, tc
20329 the value of @var{timeclamp} option
20334 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
20335 @item r(x), g(x), b(x)
20336 red, green, and blue value of intensity x.
20338 Default value is @code{st(0, (midi(f)-59.5)/12);
20339 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
20340 r(1-ld(1)) + b(ld(1))}.
20343 Specify image file to draw the axis. This option override @var{fontfile} and
20344 @var{fontcolor} option.
20347 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
20348 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
20349 Default value is @code{1}.
20352 Set colorspace. The accepted values are:
20355 Unspecified (default)
20364 BT.470BG or BT.601-6 625
20367 SMPTE-170M or BT.601-6 525
20373 BT.2020 with non-constant luminance
20378 Set spectrogram color scheme. This is list of floating point values with format
20379 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
20380 The default is @code{1|0.5|0|0|0.5|1}.
20384 @subsection Examples
20388 Playing audio while showing the spectrum:
20390 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
20394 Same as above, but with frame rate 30 fps:
20396 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
20400 Playing at 1280x720:
20402 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
20406 Disable sonogram display:
20412 A1 and its harmonics: A1, A2, (near)E3, A3:
20414 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),
20415 asplit[a][out1]; [a] showcqt [out0]'
20419 Same as above, but with more accuracy in frequency domain:
20421 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),
20422 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
20428 bar_v=10:sono_v=bar_v*a_weighting(f)
20432 Custom gamma, now spectrum is linear to the amplitude.
20438 Custom tlength equation:
20440 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)))'
20444 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
20446 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
20450 Custom font using fontconfig:
20452 font='Courier New,Monospace,mono|bold'
20456 Custom frequency range with custom axis using image file:
20458 axisfile=myaxis.png:basefreq=40:endfreq=10000
20464 Convert input audio to video output representing the audio power spectrum.
20465 Audio amplitude is on Y-axis while frequency is on X-axis.
20467 The filter accepts the following options:
20471 Specify size of video. For the syntax of this option, check the
20472 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20473 Default is @code{1024x512}.
20477 This set how each frequency bin will be represented.
20479 It accepts the following values:
20485 Default is @code{bar}.
20488 Set amplitude scale.
20490 It accepts the following values:
20504 Default is @code{log}.
20507 Set frequency scale.
20509 It accepts the following values:
20518 Reverse logarithmic scale.
20520 Default is @code{lin}.
20525 It accepts the following values:
20541 Default is @code{w2048}
20544 Set windowing function.
20546 It accepts the following values:
20568 Default is @code{hanning}.
20571 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20572 which means optimal overlap for selected window function will be picked.
20575 Set time averaging. Setting this to 0 will display current maximal peaks.
20576 Default is @code{1}, which means time averaging is disabled.
20579 Specify list of colors separated by space or by '|' which will be used to
20580 draw channel frequencies. Unrecognized or missing colors will be replaced
20584 Set channel display mode.
20586 It accepts the following values:
20591 Default is @code{combined}.
20594 Set minimum amplitude used in @code{log} amplitude scaler.
20598 @anchor{showspectrum}
20599 @section showspectrum
20601 Convert input audio to a video output, representing the audio frequency
20604 The filter accepts the following options:
20608 Specify the video size for the output. For the syntax of this option, check the
20609 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20610 Default value is @code{640x512}.
20613 Specify how the spectrum should slide along the window.
20615 It accepts the following values:
20618 the samples start again on the left when they reach the right
20620 the samples scroll from right to left
20622 frames are only produced when the samples reach the right
20624 the samples scroll from left to right
20627 Default value is @code{replace}.
20630 Specify display mode.
20632 It accepts the following values:
20635 all channels are displayed in the same row
20637 all channels are displayed in separate rows
20640 Default value is @samp{combined}.
20643 Specify display color mode.
20645 It accepts the following values:
20648 each channel is displayed in a separate color
20650 each channel is displayed using the same color scheme
20652 each channel is displayed using the rainbow color scheme
20654 each channel is displayed using the moreland color scheme
20656 each channel is displayed using the nebulae color scheme
20658 each channel is displayed using the fire color scheme
20660 each channel is displayed using the fiery color scheme
20662 each channel is displayed using the fruit color scheme
20664 each channel is displayed using the cool color scheme
20666 each channel is displayed using the magma color scheme
20668 each channel is displayed using the green color scheme
20671 Default value is @samp{channel}.
20674 Specify scale used for calculating intensity color values.
20676 It accepts the following values:
20681 square root, default
20692 Default value is @samp{sqrt}.
20695 Set saturation modifier for displayed colors. Negative values provide
20696 alternative color scheme. @code{0} is no saturation at all.
20697 Saturation must be in [-10.0, 10.0] range.
20698 Default value is @code{1}.
20701 Set window function.
20703 It accepts the following values:
20727 Default value is @code{hann}.
20730 Set orientation of time vs frequency axis. Can be @code{vertical} or
20731 @code{horizontal}. Default is @code{vertical}.
20734 Set ratio of overlap window. Default value is @code{0}.
20735 When value is @code{1} overlap is set to recommended size for specific
20736 window function currently used.
20739 Set scale gain for calculating intensity color values.
20740 Default value is @code{1}.
20743 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
20746 Set color rotation, must be in [-1.0, 1.0] range.
20747 Default value is @code{0}.
20750 Set start frequency from which to display spectrogram. Default is @code{0}.
20753 Set stop frequency to which to display spectrogram. Default is @code{0}.
20756 Set upper frame rate limit. Default is @code{auto}, unlimited.
20759 Draw time and frequency axes and legends. Default is disabled.
20762 The usage is very similar to the showwaves filter; see the examples in that
20765 @subsection Examples
20769 Large window with logarithmic color scaling:
20771 showspectrum=s=1280x480:scale=log
20775 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
20777 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20778 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
20782 @section showspectrumpic
20784 Convert input audio to a single video frame, representing the audio frequency
20787 The filter accepts the following options:
20791 Specify the video size for the output. For the syntax of this option, check the
20792 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20793 Default value is @code{4096x2048}.
20796 Specify display mode.
20798 It accepts the following values:
20801 all channels are displayed in the same row
20803 all channels are displayed in separate rows
20805 Default value is @samp{combined}.
20808 Specify display color mode.
20810 It accepts the following values:
20813 each channel is displayed in a separate color
20815 each channel is displayed using the same color scheme
20817 each channel is displayed using the rainbow color scheme
20819 each channel is displayed using the moreland color scheme
20821 each channel is displayed using the nebulae color scheme
20823 each channel is displayed using the fire color scheme
20825 each channel is displayed using the fiery color scheme
20827 each channel is displayed using the fruit color scheme
20829 each channel is displayed using the cool color scheme
20831 each channel is displayed using the magma color scheme
20833 each channel is displayed using the green color scheme
20835 Default value is @samp{intensity}.
20838 Specify scale used for calculating intensity color values.
20840 It accepts the following values:
20845 square root, default
20855 Default value is @samp{log}.
20858 Set saturation modifier for displayed colors. Negative values provide
20859 alternative color scheme. @code{0} is no saturation at all.
20860 Saturation must be in [-10.0, 10.0] range.
20861 Default value is @code{1}.
20864 Set window function.
20866 It accepts the following values:
20889 Default value is @code{hann}.
20892 Set orientation of time vs frequency axis. Can be @code{vertical} or
20893 @code{horizontal}. Default is @code{vertical}.
20896 Set scale gain for calculating intensity color values.
20897 Default value is @code{1}.
20900 Draw time and frequency axes and legends. Default is enabled.
20903 Set color rotation, must be in [-1.0, 1.0] range.
20904 Default value is @code{0}.
20907 Set start frequency from which to display spectrogram. Default is @code{0}.
20910 Set stop frequency to which to display spectrogram. Default is @code{0}.
20913 @subsection Examples
20917 Extract an audio spectrogram of a whole audio track
20918 in a 1024x1024 picture using @command{ffmpeg}:
20920 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
20924 @section showvolume
20926 Convert input audio volume to a video output.
20928 The filter accepts the following options:
20935 Set border width, allowed range is [0, 5]. Default is 1.
20938 Set channel width, allowed range is [80, 8192]. Default is 400.
20941 Set channel height, allowed range is [1, 900]. Default is 20.
20944 Set fade, allowed range is [0, 1]. Default is 0.95.
20947 Set volume color expression.
20949 The expression can use the following variables:
20953 Current max volume of channel in dB.
20959 Current channel number, starting from 0.
20963 If set, displays channel names. Default is enabled.
20966 If set, displays volume values. Default is enabled.
20969 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
20970 default is @code{h}.
20973 Set step size, allowed range is [0, 5]. Default is 0, which means
20977 Set background opacity, allowed range is [0, 1]. Default is 0.
20980 Set metering mode, can be peak: @code{p} or rms: @code{r},
20981 default is @code{p}.
20984 Set display scale, can be linear: @code{lin} or log: @code{log},
20985 default is @code{lin}.
20989 If set to > 0., display a line for the max level
20990 in the previous seconds.
20991 default is disabled: @code{0.}
20994 The color of the max line. Use when @code{dm} option is set to > 0.
20995 default is: @code{orange}
21000 Convert input audio to a video output, representing the samples waves.
21002 The filter accepts the following options:
21006 Specify the video size for the output. For the syntax of this option, check the
21007 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21008 Default value is @code{600x240}.
21013 Available values are:
21016 Draw a point for each sample.
21019 Draw a vertical line for each sample.
21022 Draw a point for each sample and a line between them.
21025 Draw a centered vertical line for each sample.
21028 Default value is @code{point}.
21031 Set the number of samples which are printed on the same column. A
21032 larger value will decrease the frame rate. Must be a positive
21033 integer. This option can be set only if the value for @var{rate}
21034 is not explicitly specified.
21037 Set the (approximate) output frame rate. This is done by setting the
21038 option @var{n}. Default value is "25".
21040 @item split_channels
21041 Set if channels should be drawn separately or overlap. Default value is 0.
21044 Set colors separated by '|' which are going to be used for drawing of each channel.
21047 Set amplitude scale.
21049 Available values are:
21067 Set the draw mode. This is mostly useful to set for high @var{n}.
21069 Available values are:
21072 Scale pixel values for each drawn sample.
21075 Draw every sample directly.
21078 Default value is @code{scale}.
21081 @subsection Examples
21085 Output the input file audio and the corresponding video representation
21088 amovie=a.mp3,asplit[out0],showwaves[out1]
21092 Create a synthetic signal and show it with showwaves, forcing a
21093 frame rate of 30 frames per second:
21095 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
21099 @section showwavespic
21101 Convert input audio to a single video frame, representing the samples waves.
21103 The filter accepts the following options:
21107 Specify the video size for the output. For the syntax of this option, check the
21108 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21109 Default value is @code{600x240}.
21111 @item split_channels
21112 Set if channels should be drawn separately or overlap. Default value is 0.
21115 Set colors separated by '|' which are going to be used for drawing of each channel.
21118 Set amplitude scale.
21120 Available values are:
21138 @subsection Examples
21142 Extract a channel split representation of the wave form of a whole audio track
21143 in a 1024x800 picture using @command{ffmpeg}:
21145 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
21149 @section sidedata, asidedata
21151 Delete frame side data, or select frames based on it.
21153 This filter accepts the following options:
21157 Set mode of operation of the filter.
21159 Can be one of the following:
21163 Select every frame with side data of @code{type}.
21166 Delete side data of @code{type}. If @code{type} is not set, delete all side
21172 Set side data type used with all modes. Must be set for @code{select} mode. For
21173 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
21174 in @file{libavutil/frame.h}. For example, to choose
21175 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
21179 @section spectrumsynth
21181 Sythesize audio from 2 input video spectrums, first input stream represents
21182 magnitude across time and second represents phase across time.
21183 The filter will transform from frequency domain as displayed in videos back
21184 to time domain as presented in audio output.
21186 This filter is primarily created for reversing processed @ref{showspectrum}
21187 filter outputs, but can synthesize sound from other spectrograms too.
21188 But in such case results are going to be poor if the phase data is not
21189 available, because in such cases phase data need to be recreated, usually
21190 its just recreated from random noise.
21191 For best results use gray only output (@code{channel} color mode in
21192 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
21193 @code{lin} scale for phase video. To produce phase, for 2nd video, use
21194 @code{data} option. Inputs videos should generally use @code{fullframe}
21195 slide mode as that saves resources needed for decoding video.
21197 The filter accepts the following options:
21201 Specify sample rate of output audio, the sample rate of audio from which
21202 spectrum was generated may differ.
21205 Set number of channels represented in input video spectrums.
21208 Set scale which was used when generating magnitude input spectrum.
21209 Can be @code{lin} or @code{log}. Default is @code{log}.
21212 Set slide which was used when generating inputs spectrums.
21213 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
21214 Default is @code{fullframe}.
21217 Set window function used for resynthesis.
21220 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
21221 which means optimal overlap for selected window function will be picked.
21224 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
21225 Default is @code{vertical}.
21228 @subsection Examples
21232 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
21233 then resynthesize videos back to audio with spectrumsynth:
21235 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
21236 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
21237 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
21241 @section split, asplit
21243 Split input into several identical outputs.
21245 @code{asplit} works with audio input, @code{split} with video.
21247 The filter accepts a single parameter which specifies the number of outputs. If
21248 unspecified, it defaults to 2.
21250 @subsection Examples
21254 Create two separate outputs from the same input:
21256 [in] split [out0][out1]
21260 To create 3 or more outputs, you need to specify the number of
21263 [in] asplit=3 [out0][out1][out2]
21267 Create two separate outputs from the same input, one cropped and
21270 [in] split [splitout1][splitout2];
21271 [splitout1] crop=100:100:0:0 [cropout];
21272 [splitout2] pad=200:200:100:100 [padout];
21276 Create 5 copies of the input audio with @command{ffmpeg}:
21278 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
21284 Receive commands sent through a libzmq client, and forward them to
21285 filters in the filtergraph.
21287 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
21288 must be inserted between two video filters, @code{azmq} between two
21289 audio filters. Both are capable to send messages to any filter type.
21291 To enable these filters you need to install the libzmq library and
21292 headers and configure FFmpeg with @code{--enable-libzmq}.
21294 For more information about libzmq see:
21295 @url{http://www.zeromq.org/}
21297 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
21298 receives messages sent through a network interface defined by the
21299 @option{bind_address} (or the abbreviation "@option{b}") option.
21300 Default value of this option is @file{tcp://localhost:5555}. You may
21301 want to alter this value to your needs, but do not forget to escape any
21302 ':' signs (see @ref{filtergraph escaping}).
21304 The received message must be in the form:
21306 @var{TARGET} @var{COMMAND} [@var{ARG}]
21309 @var{TARGET} specifies the target of the command, usually the name of
21310 the filter class or a specific filter instance name. The default
21311 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
21312 but you can override this by using the @samp{filter_name@@id} syntax
21313 (see @ref{Filtergraph syntax}).
21315 @var{COMMAND} specifies the name of the command for the target filter.
21317 @var{ARG} is optional and specifies the optional argument list for the
21318 given @var{COMMAND}.
21320 Upon reception, the message is processed and the corresponding command
21321 is injected into the filtergraph. Depending on the result, the filter
21322 will send a reply to the client, adopting the format:
21324 @var{ERROR_CODE} @var{ERROR_REASON}
21328 @var{MESSAGE} is optional.
21330 @subsection Examples
21332 Look at @file{tools/zmqsend} for an example of a zmq client which can
21333 be used to send commands processed by these filters.
21335 Consider the following filtergraph generated by @command{ffplay}.
21336 In this example the last overlay filter has an instance name. All other
21337 filters will have default instance names.
21340 ffplay -dumpgraph 1 -f lavfi "
21341 color=s=100x100:c=red [l];
21342 color=s=100x100:c=blue [r];
21343 nullsrc=s=200x100, zmq [bg];
21344 [bg][l] overlay [bg+l];
21345 [bg+l][r] overlay@@my=x=100 "
21348 To change the color of the left side of the video, the following
21349 command can be used:
21351 echo Parsed_color_0 c yellow | tools/zmqsend
21354 To change the right side:
21356 echo Parsed_color_1 c pink | tools/zmqsend
21359 To change the position of the right side:
21361 echo overlay@@my x 150 | tools/zmqsend
21365 @c man end MULTIMEDIA FILTERS
21367 @chapter Multimedia Sources
21368 @c man begin MULTIMEDIA SOURCES
21370 Below is a description of the currently available multimedia sources.
21374 This is the same as @ref{movie} source, except it selects an audio
21380 Read audio and/or video stream(s) from a movie container.
21382 It accepts the following parameters:
21386 The name of the resource to read (not necessarily a file; it can also be a
21387 device or a stream accessed through some protocol).
21389 @item format_name, f
21390 Specifies the format assumed for the movie to read, and can be either
21391 the name of a container or an input device. If not specified, the
21392 format is guessed from @var{movie_name} or by probing.
21394 @item seek_point, sp
21395 Specifies the seek point in seconds. The frames will be output
21396 starting from this seek point. The parameter is evaluated with
21397 @code{av_strtod}, so the numerical value may be suffixed by an IS
21398 postfix. The default value is "0".
21401 Specifies the streams to read. Several streams can be specified,
21402 separated by "+". The source will then have as many outputs, in the
21403 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
21404 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
21405 respectively the default (best suited) video and audio stream. Default
21406 is "dv", or "da" if the filter is called as "amovie".
21408 @item stream_index, si
21409 Specifies the index of the video stream to read. If the value is -1,
21410 the most suitable video stream will be automatically selected. The default
21411 value is "-1". Deprecated. If the filter is called "amovie", it will select
21412 audio instead of video.
21415 Specifies how many times to read the stream in sequence.
21416 If the value is 0, the stream will be looped infinitely.
21417 Default value is "1".
21419 Note that when the movie is looped the source timestamps are not
21420 changed, so it will generate non monotonically increasing timestamps.
21422 @item discontinuity
21423 Specifies the time difference between frames above which the point is
21424 considered a timestamp discontinuity which is removed by adjusting the later
21428 It allows overlaying a second video on top of the main input of
21429 a filtergraph, as shown in this graph:
21431 input -----------> deltapts0 --> overlay --> output
21434 movie --> scale--> deltapts1 -------+
21436 @subsection Examples
21440 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
21441 on top of the input labelled "in":
21443 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
21444 [in] setpts=PTS-STARTPTS [main];
21445 [main][over] overlay=16:16 [out]
21449 Read from a video4linux2 device, and overlay it on top of the input
21452 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
21453 [in] setpts=PTS-STARTPTS [main];
21454 [main][over] overlay=16:16 [out]
21458 Read the first video stream and the audio stream with id 0x81 from
21459 dvd.vob; the video is connected to the pad named "video" and the audio is
21460 connected to the pad named "audio":
21462 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
21466 @subsection Commands
21468 Both movie and amovie support the following commands:
21471 Perform seek using "av_seek_frame".
21472 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
21475 @var{stream_index}: If stream_index is -1, a default
21476 stream is selected, and @var{timestamp} is automatically converted
21477 from AV_TIME_BASE units to the stream specific time_base.
21479 @var{timestamp}: Timestamp in AVStream.time_base units
21480 or, if no stream is specified, in AV_TIME_BASE units.
21482 @var{flags}: Flags which select direction and seeking mode.
21486 Get movie duration in AV_TIME_BASE units.
21490 @c man end MULTIMEDIA SOURCES