1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 If a signal of stream rises above this level it will affect the gain
396 By default it is 0.125. Range is between 0.00097563 and 1.
399 Set a ratio by which the signal is reduced. 1:2 means that if the level
400 rose 4dB above the threshold, it will be only 2dB above after the reduction.
401 Default is 2. Range is between 1 and 20.
404 Amount of milliseconds the signal has to rise above the threshold before gain
405 reduction starts. Default is 20. Range is between 0.01 and 2000.
408 Amount of milliseconds the signal has to fall below the threshold before
409 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
412 Set the amount by how much signal will be amplified after processing.
413 Default is 1. Range is from 1 to 64.
416 Curve the sharp knee around the threshold to enter gain reduction more softly.
417 Default is 2.82843. Range is between 1 and 8.
420 Choose if the @code{average} level between all channels of input stream
421 or the louder(@code{maximum}) channel of input stream affects the
422 reduction. Default is @code{average}.
425 Should the exact signal be taken in case of @code{peak} or an RMS one in case
426 of @code{rms}. Default is @code{rms} which is mostly smoother.
429 How much to use compressed signal in output. Default is 1.
430 Range is between 0 and 1.
434 Simple audio dynamic range compression/expansion filter.
436 The filter accepts the following options:
440 Set contrast. Default is 33. Allowed range is between 0 and 100.
445 Copy the input audio source unchanged to the output. This is mainly useful for
450 Apply cross fade from one input audio stream to another input audio stream.
451 The cross fade is applied for specified duration near the end of first stream.
453 The filter accepts the following options:
457 Specify the number of samples for which the cross fade effect has to last.
458 At the end of the cross fade effect the first input audio will be completely
459 silent. Default is 44100.
462 Specify the duration of the cross fade effect. See
463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
464 for the accepted syntax.
465 By default the duration is determined by @var{nb_samples}.
466 If set this option is used instead of @var{nb_samples}.
469 Should first stream end overlap with second stream start. Default is enabled.
472 Set curve for cross fade transition for first stream.
475 Set curve for cross fade transition for second stream.
477 For description of available curve types see @ref{afade} filter description.
484 Cross fade from one input to another:
486 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
490 Cross fade from one input to another but without overlapping:
492 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Split audio stream into several bands.
499 This filter splits audio stream into two or more frequency ranges.
500 Summing all streams back will give flat output.
502 The filter accepts the following options:
506 Set split frequencies. Those must be positive and increasing.
509 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
510 Default is @var{4th}.
515 Reduce audio bit resolution.
517 This filter is bit crusher with enhanced functionality. A bit crusher
518 is used to audibly reduce number of bits an audio signal is sampled
519 with. This doesn't change the bit depth at all, it just produces the
520 effect. Material reduced in bit depth sounds more harsh and "digital".
521 This filter is able to even round to continuous values instead of discrete
523 Additionally it has a D/C offset which results in different crushing of
524 the lower and the upper half of the signal.
525 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
527 Another feature of this filter is the logarithmic mode.
528 This setting switches from linear distances between bits to logarithmic ones.
529 The result is a much more "natural" sounding crusher which doesn't gate low
530 signals for example. The human ear has a logarithmic perception,
531 so this kind of crushing is much more pleasant.
532 Logarithmic crushing is also able to get anti-aliased.
534 The filter accepts the following options:
550 Can be linear: @code{lin} or logarithmic: @code{log}.
559 Set sample reduction.
562 Enable LFO. By default disabled.
573 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
577 Remove impulsive noise from input audio.
579 Samples detected as impulsive noise are replaced by interpolated samples using
580 autoregressive modelling.
584 Set window size, in milliseconds. Allowed range is from @code{10} to
585 @code{100}. Default value is @code{55} milliseconds.
586 This sets size of window which will be processed at once.
589 Set window overlap, in percentage of window size. Allowed range is from
590 @code{50} to @code{95}. Default value is @code{75} percent.
591 Setting this to a very high value increases impulsive noise removal but makes
592 whole process much slower.
595 Set autoregression order, in percentage of window size. Allowed range is from
596 @code{0} to @code{25}. Default value is @code{2} percent. This option also
597 controls quality of interpolated samples using neighbour good samples.
600 Set threshold value. Allowed range is from @code{1} to @code{100}.
601 Default value is @code{2}.
602 This controls the strength of impulsive noise which is going to be removed.
603 The lower value, the more samples will be detected as impulsive noise.
606 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
607 @code{10}. Default value is @code{2}.
608 If any two samples detected as noise are spaced less than this value then any
609 sample between those two samples will be also detected as noise.
614 It accepts the following values:
617 Select overlap-add method. Even not interpolated samples are slightly
618 changed with this method.
621 Select overlap-save method. Not interpolated samples remain unchanged.
624 Default value is @code{a}.
628 Remove clipped samples from input audio.
630 Samples detected as clipped are replaced by interpolated samples using
631 autoregressive modelling.
635 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
636 Default value is @code{55} milliseconds.
637 This sets size of window which will be processed at once.
640 Set window overlap, in percentage of window size. Allowed range is from @code{50}
641 to @code{95}. Default value is @code{75} percent.
644 Set autoregression order, in percentage of window size. Allowed range is from
645 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
646 quality of interpolated samples using neighbour good samples.
649 Set threshold value. Allowed range is from @code{1} to @code{100}.
650 Default value is @code{10}. Higher values make clip detection less aggressive.
653 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
654 Default value is @code{1000}. Higher values make clip detection less aggressive.
659 It accepts the following values:
662 Select overlap-add method. Even not interpolated samples are slightly changed
666 Select overlap-save method. Not interpolated samples remain unchanged.
669 Default value is @code{a}.
674 Delay one or more audio channels.
676 Samples in delayed channel are filled with silence.
678 The filter accepts the following option:
682 Set list of delays in milliseconds for each channel separated by '|'.
683 Unused delays will be silently ignored. If number of given delays is
684 smaller than number of channels all remaining channels will not be delayed.
685 If you want to delay exact number of samples, append 'S' to number.
686 If you want instead to delay in seconds, append 's' to number.
693 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
694 the second channel (and any other channels that may be present) unchanged.
700 Delay second channel by 500 samples, the third channel by 700 samples and leave
701 the first channel (and any other channels that may be present) unchanged.
707 @section aderivative, aintegral
709 Compute derivative/integral of audio stream.
711 Applying both filters one after another produces original audio.
715 Apply echoing to the input audio.
717 Echoes are reflected sound and can occur naturally amongst mountains
718 (and sometimes large buildings) when talking or shouting; digital echo
719 effects emulate this behaviour and are often used to help fill out the
720 sound of a single instrument or vocal. The time difference between the
721 original signal and the reflection is the @code{delay}, and the
722 loudness of the reflected signal is the @code{decay}.
723 Multiple echoes can have different delays and decays.
725 A description of the accepted parameters follows.
729 Set input gain of reflected signal. Default is @code{0.6}.
732 Set output gain of reflected signal. Default is @code{0.3}.
735 Set list of time intervals in milliseconds between original signal and reflections
736 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
737 Default is @code{1000}.
740 Set list of loudness of reflected signals separated by '|'.
741 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
742 Default is @code{0.5}.
749 Make it sound as if there are twice as many instruments as are actually playing:
751 aecho=0.8:0.88:60:0.4
755 If delay is very short, then it sound like a (metallic) robot playing music:
761 A longer delay will sound like an open air concert in the mountains:
763 aecho=0.8:0.9:1000:0.3
767 Same as above but with one more mountain:
769 aecho=0.8:0.9:1000|1800:0.3|0.25
774 Audio emphasis filter creates or restores material directly taken from LPs or
775 emphased CDs with different filter curves. E.g. to store music on vinyl the
776 signal has to be altered by a filter first to even out the disadvantages of
777 this recording medium.
778 Once the material is played back the inverse filter has to be applied to
779 restore the distortion of the frequency response.
781 The filter accepts the following options:
791 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
792 use @code{production} mode. Default is @code{reproduction} mode.
795 Set filter type. Selects medium. Can be one of the following:
807 select Compact Disc (CD).
813 select 50µs (FM-KF).
815 select 75µs (FM-KF).
821 Modify an audio signal according to the specified expressions.
823 This filter accepts one or more expressions (one for each channel),
824 which are evaluated and used to modify a corresponding audio signal.
826 It accepts the following parameters:
830 Set the '|'-separated expressions list for each separate channel. If
831 the number of input channels is greater than the number of
832 expressions, the last specified expression is used for the remaining
835 @item channel_layout, c
836 Set output channel layout. If not specified, the channel layout is
837 specified by the number of expressions. If set to @samp{same}, it will
838 use by default the same input channel layout.
841 Each expression in @var{exprs} can contain the following constants and functions:
845 channel number of the current expression
848 number of the evaluated sample, starting from 0
854 time of the evaluated sample expressed in seconds
857 @item nb_out_channels
858 input and output number of channels
861 the value of input channel with number @var{CH}
864 Note: this filter is slow. For faster processing you should use a
873 aeval=val(ch)/2:c=same
877 Invert phase of the second channel:
886 Apply fade-in/out effect to input audio.
888 A description of the accepted parameters follows.
892 Specify the effect type, can be either @code{in} for fade-in, or
893 @code{out} for a fade-out effect. Default is @code{in}.
895 @item start_sample, ss
896 Specify the number of the start sample for starting to apply the fade
897 effect. Default is 0.
900 Specify the number of samples for which the fade effect has to last. At
901 the end of the fade-in effect the output audio will have the same
902 volume as the input audio, at the end of the fade-out transition
903 the output audio will be silence. Default is 44100.
906 Specify the start time of the fade effect. Default is 0.
907 The value must be specified as a time duration; see
908 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
909 for the accepted syntax.
910 If set this option is used instead of @var{start_sample}.
913 Specify the duration of the fade effect. See
914 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
915 for the accepted syntax.
916 At the end of the fade-in effect the output audio will have the same
917 volume as the input audio, at the end of the fade-out transition
918 the output audio will be silence.
919 By default the duration is determined by @var{nb_samples}.
920 If set this option is used instead of @var{nb_samples}.
923 Set curve for fade transition.
925 It accepts the following values:
928 select triangular, linear slope (default)
930 select quarter of sine wave
932 select half of sine wave
934 select exponential sine wave
938 select inverted parabola
952 select inverted quarter of sine wave
954 select inverted half of sine wave
956 select double-exponential seat
958 select double-exponential sigmoid
960 select logistic sigmoid
970 Fade in first 15 seconds of audio:
976 Fade out last 25 seconds of a 900 seconds audio:
978 afade=t=out:st=875:d=25
983 Denoise audio samples with FFT.
985 A description of the accepted parameters follows.
989 Set the noise reduction in dB, allowed range is 0.01 to 97.
990 Default value is 12 dB.
993 Set the noise floor in dB, allowed range is -80 to -20.
994 Default value is -50 dB.
999 It accepts the following values:
1008 Select shellac noise.
1011 Select custom noise, defined in @code{bn} option.
1013 Default value is white noise.
1017 Set custom band noise for every one of 15 bands.
1018 Bands are separated by ' ' or '|'.
1021 Set the residual floor in dB, allowed range is -80 to -20.
1022 Default value is -38 dB.
1025 Enable noise tracking. By default is disabled.
1026 With this enabled, noise floor is automatically adjusted.
1029 Enable residual tracking. By default is disabled.
1032 Set the output mode.
1034 It accepts the following values:
1037 Pass input unchanged.
1040 Pass noise filtered out.
1045 Default value is @var{o}.
1049 @subsection Commands
1051 This filter supports the following commands:
1053 @item sample_noise, sn
1054 Start or stop measuring noise profile.
1055 Syntax for the command is : "start" or "stop" string.
1056 After measuring noise profile is stopped it will be
1057 automatically applied in filtering.
1059 @item noise_reduction, nr
1060 Change noise reduction. Argument is single float number.
1061 Syntax for the command is : "@var{noise_reduction}"
1063 @item noise_floor, nf
1064 Change noise floor. Argument is single float number.
1065 Syntax for the command is : "@var{noise_floor}"
1067 @item output_mode, om
1068 Change output mode operation.
1069 Syntax for the command is : "i", "o" or "n" string.
1073 Apply arbitrary expressions to samples in frequency domain.
1077 Set frequency domain real expression for each separate channel separated
1078 by '|'. Default is "re".
1079 If the number of input channels is greater than the number of
1080 expressions, the last specified expression is used for the remaining
1084 Set frequency domain imaginary expression for each separate channel
1085 separated by '|'. Default is "im".
1087 Each expression in @var{real} and @var{imag} can contain the following
1088 constants and functions:
1095 current frequency bin number
1098 number of available bins
1101 channel number of the current expression
1110 current real part of frequency bin of current channel
1113 current imaginary part of frequency bin of current channel
1116 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1119 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1125 It accepts the following values:
1141 Default is @code{w4096}
1144 Set window function. Default is @code{hann}.
1147 Set window overlap. If set to 1, the recommended overlap for selected
1148 window function will be picked. Default is @code{0.75}.
1151 @subsection Examples
1155 Leave almost only low frequencies in audio:
1157 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1164 Apply an arbitrary Frequency Impulse Response filter.
1166 This filter is designed for applying long FIR filters,
1167 up to 60 seconds long.
1169 It can be used as component for digital crossover filters,
1170 room equalization, cross talk cancellation, wavefield synthesis,
1171 auralization, ambiophonics, ambisonics and spatialization.
1173 This filter uses second stream as FIR coefficients.
1174 If second stream holds single channel, it will be used
1175 for all input channels in first stream, otherwise
1176 number of channels in second stream must be same as
1177 number of channels in first stream.
1179 It accepts the following parameters:
1183 Set dry gain. This sets input gain.
1186 Set wet gain. This sets final output gain.
1189 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1192 Enable applying gain measured from power of IR.
1194 Set which approach to use for auto gain measurement.
1198 Do not apply any gain.
1201 select peak gain, very conservative approach. This is default value.
1204 select DC gain, limited application.
1207 select gain to noise approach, this is most popular one.
1211 Set gain to be applied to IR coefficients before filtering.
1212 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1215 Set format of IR stream. Can be @code{mono} or @code{input}.
1216 Default is @code{input}.
1219 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1220 Allowed range is 0.1 to 60 seconds.
1223 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1224 By default it is disabled.
1227 Set for which IR channel to display frequency response. By default is first channel
1228 displayed. This option is used only when @var{response} is enabled.
1231 Set video stream size. This option is used only when @var{response} is enabled.
1234 Set video stream frame rate. This option is used only when @var{response} is enabled.
1237 Set minimal partition size used for convolution. Default is @var{8192}.
1238 Allowed range is from @var{8} to @var{32768}.
1239 Lower values decreases latency at cost of higher CPU usage.
1242 Set maximal partition size used for convolution. Default is @var{8192}.
1243 Allowed range is from @var{8} to @var{32768}.
1244 Lower values may increase CPU usage.
1247 @subsection Examples
1251 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1253 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1260 Set output format constraints for the input audio. The framework will
1261 negotiate the most appropriate format to minimize conversions.
1263 It accepts the following parameters:
1267 A '|'-separated list of requested sample formats.
1270 A '|'-separated list of requested sample rates.
1272 @item channel_layouts
1273 A '|'-separated list of requested channel layouts.
1275 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1276 for the required syntax.
1279 If a parameter is omitted, all values are allowed.
1281 Force the output to either unsigned 8-bit or signed 16-bit stereo
1283 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1288 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1289 processing reduces disturbing noise between useful signals.
1291 Gating is done by detecting the volume below a chosen level @var{threshold}
1292 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1293 floor is set via @var{range}. Because an exact manipulation of the signal
1294 would cause distortion of the waveform the reduction can be levelled over
1295 time. This is done by setting @var{attack} and @var{release}.
1297 @var{attack} determines how long the signal has to fall below the threshold
1298 before any reduction will occur and @var{release} sets the time the signal
1299 has to rise above the threshold to reduce the reduction again.
1300 Shorter signals than the chosen attack time will be left untouched.
1304 Set input level before filtering.
1305 Default is 1. Allowed range is from 0.015625 to 64.
1308 Set the level of gain reduction when the signal is below the threshold.
1309 Default is 0.06125. Allowed range is from 0 to 1.
1312 If a signal rises above this level the gain reduction is released.
1313 Default is 0.125. Allowed range is from 0 to 1.
1316 Set a ratio by which the signal is reduced.
1317 Default is 2. Allowed range is from 1 to 9000.
1320 Amount of milliseconds the signal has to rise above the threshold before gain
1322 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1325 Amount of milliseconds the signal has to fall below the threshold before the
1326 reduction is increased again. Default is 250 milliseconds.
1327 Allowed range is from 0.01 to 9000.
1330 Set amount of amplification of signal after processing.
1331 Default is 1. Allowed range is from 1 to 64.
1334 Curve the sharp knee around the threshold to enter gain reduction more softly.
1335 Default is 2.828427125. Allowed range is from 1 to 8.
1338 Choose if exact signal should be taken for detection or an RMS like one.
1339 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1342 Choose if the average level between all channels or the louder channel affects
1344 Default is @code{average}. Can be @code{average} or @code{maximum}.
1349 Apply an arbitrary Infinite Impulse Response filter.
1351 It accepts the following parameters:
1355 Set numerator/zeros coefficients.
1358 Set denominator/poles coefficients.
1370 Set coefficients format.
1376 Z-plane zeros/poles, cartesian (default)
1378 Z-plane zeros/poles, polar radians
1380 Z-plane zeros/poles, polar degrees
1384 Set kind of processing.
1385 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1388 Set filtering precision.
1392 double-precision floating-point (default)
1394 single-precision floating-point
1402 Show IR frequency response, magnitude and phase in additional video stream.
1403 By default it is disabled.
1406 Set for which IR channel to display frequency response. By default is first channel
1407 displayed. This option is used only when @var{response} is enabled.
1410 Set video stream size. This option is used only when @var{response} is enabled.
1413 Coefficients in @code{tf} format are separated by spaces and are in ascending
1416 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1417 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1420 Different coefficients and gains can be provided for every channel, in such case
1421 use '|' to separate coefficients or gains. Last provided coefficients will be
1422 used for all remaining channels.
1424 @subsection Examples
1428 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1430 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1434 Same as above but in @code{zp} format:
1436 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1442 The limiter prevents an input signal from rising over a desired threshold.
1443 This limiter uses lookahead technology to prevent your signal from distorting.
1444 It means that there is a small delay after the signal is processed. Keep in mind
1445 that the delay it produces is the attack time you set.
1447 The filter accepts the following options:
1451 Set input gain. Default is 1.
1454 Set output gain. Default is 1.
1457 Don't let signals above this level pass the limiter. Default is 1.
1460 The limiter will reach its attenuation level in this amount of time in
1461 milliseconds. Default is 5 milliseconds.
1464 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1465 Default is 50 milliseconds.
1468 When gain reduction is always needed ASC takes care of releasing to an
1469 average reduction level rather than reaching a reduction of 0 in the release
1473 Select how much the release time is affected by ASC, 0 means nearly no changes
1474 in release time while 1 produces higher release times.
1477 Auto level output signal. Default is enabled.
1478 This normalizes audio back to 0dB if enabled.
1481 Depending on picked setting it is recommended to upsample input 2x or 4x times
1482 with @ref{aresample} before applying this filter.
1486 Apply a two-pole all-pass filter with central frequency (in Hz)
1487 @var{frequency}, and filter-width @var{width}.
1488 An all-pass filter changes the audio's frequency to phase relationship
1489 without changing its frequency to amplitude relationship.
1491 The filter accepts the following options:
1495 Set frequency in Hz.
1498 Set method to specify band-width of filter.
1513 Specify the band-width of a filter in width_type units.
1516 Specify which channels to filter, by default all available are filtered.
1519 @subsection Commands
1521 This filter supports the following commands:
1524 Change allpass frequency.
1525 Syntax for the command is : "@var{frequency}"
1528 Change allpass width_type.
1529 Syntax for the command is : "@var{width_type}"
1532 Change allpass width.
1533 Syntax for the command is : "@var{width}"
1540 The filter accepts the following options:
1544 Set the number of loops. Setting this value to -1 will result in infinite loops.
1548 Set maximal number of samples. Default is 0.
1551 Set first sample of loop. Default is 0.
1557 Merge two or more audio streams into a single multi-channel stream.
1559 The filter accepts the following options:
1564 Set the number of inputs. Default is 2.
1568 If the channel layouts of the inputs are disjoint, and therefore compatible,
1569 the channel layout of the output will be set accordingly and the channels
1570 will be reordered as necessary. If the channel layouts of the inputs are not
1571 disjoint, the output will have all the channels of the first input then all
1572 the channels of the second input, in that order, and the channel layout of
1573 the output will be the default value corresponding to the total number of
1576 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1577 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1578 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1579 first input, b1 is the first channel of the second input).
1581 On the other hand, if both input are in stereo, the output channels will be
1582 in the default order: a1, a2, b1, b2, and the channel layout will be
1583 arbitrarily set to 4.0, which may or may not be the expected value.
1585 All inputs must have the same sample rate, and format.
1587 If inputs do not have the same duration, the output will stop with the
1590 @subsection Examples
1594 Merge two mono files into a stereo stream:
1596 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1600 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1602 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1608 Mixes multiple audio inputs into a single output.
1610 Note that this filter only supports float samples (the @var{amerge}
1611 and @var{pan} audio filters support many formats). If the @var{amix}
1612 input has integer samples then @ref{aresample} will be automatically
1613 inserted to perform the conversion to float samples.
1617 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1619 will mix 3 input audio streams to a single output with the same duration as the
1620 first input and a dropout transition time of 3 seconds.
1622 It accepts the following parameters:
1626 The number of inputs. If unspecified, it defaults to 2.
1629 How to determine the end-of-stream.
1633 The duration of the longest input. (default)
1636 The duration of the shortest input.
1639 The duration of the first input.
1643 @item dropout_transition
1644 The transition time, in seconds, for volume renormalization when an input
1645 stream ends. The default value is 2 seconds.
1648 Specify weight of each input audio stream as sequence.
1649 Each weight is separated by space. By default all inputs have same weight.
1654 Multiply first audio stream with second audio stream and store result
1655 in output audio stream. Multiplication is done by multiplying each
1656 sample from first stream with sample at same position from second stream.
1658 With this element-wise multiplication one can create amplitude fades and
1659 amplitude modulations.
1661 @section anequalizer
1663 High-order parametric multiband equalizer for each channel.
1665 It accepts the following parameters:
1669 This option string is in format:
1670 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1671 Each equalizer band is separated by '|'.
1675 Set channel number to which equalization will be applied.
1676 If input doesn't have that channel the entry is ignored.
1679 Set central frequency for band.
1680 If input doesn't have that frequency the entry is ignored.
1683 Set band width in hertz.
1686 Set band gain in dB.
1689 Set filter type for band, optional, can be:
1693 Butterworth, this is default.
1704 With this option activated frequency response of anequalizer is displayed
1708 Set video stream size. Only useful if curves option is activated.
1711 Set max gain that will be displayed. Only useful if curves option is activated.
1712 Setting this to a reasonable value makes it possible to display gain which is derived from
1713 neighbour bands which are too close to each other and thus produce higher gain
1714 when both are activated.
1717 Set frequency scale used to draw frequency response in video output.
1718 Can be linear or logarithmic. Default is logarithmic.
1721 Set color for each channel curve which is going to be displayed in video stream.
1722 This is list of color names separated by space or by '|'.
1723 Unrecognised or missing colors will be replaced by white color.
1726 @subsection Examples
1730 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1731 for first 2 channels using Chebyshev type 1 filter:
1733 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1737 @subsection Commands
1739 This filter supports the following commands:
1742 Alter existing filter parameters.
1743 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1745 @var{fN} is existing filter number, starting from 0, if no such filter is available
1747 @var{freq} set new frequency parameter.
1748 @var{width} set new width parameter in herz.
1749 @var{gain} set new gain parameter in dB.
1751 Full filter invocation with asendcmd may look like this:
1752 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1757 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1759 Each sample is adjusted by looking for other samples with similar contexts. This
1760 context similarity is defined by comparing their surrounding patches of size
1761 @option{p}. Patches are searched in an area of @option{r} around the sample.
1763 The filter accepts the following options.
1767 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1770 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1771 Default value is 2 milliseconds.
1774 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1775 Default value is 6 milliseconds.
1778 Set the output mode.
1780 It accepts the following values:
1783 Pass input unchanged.
1786 Pass noise filtered out.
1791 Default value is @var{o}.
1797 Pass the audio source unchanged to the output.
1801 Pad the end of an audio stream with silence.
1803 This can be used together with @command{ffmpeg} @option{-shortest} to
1804 extend audio streams to the same length as the video stream.
1806 A description of the accepted options follows.
1810 Set silence packet size. Default value is 4096.
1813 Set the number of samples of silence to add to the end. After the
1814 value is reached, the stream is terminated. This option is mutually
1815 exclusive with @option{whole_len}.
1818 Set the minimum total number of samples in the output audio stream. If
1819 the value is longer than the input audio length, silence is added to
1820 the end, until the value is reached. This option is mutually exclusive
1821 with @option{pad_len}.
1824 Specify the duration of samples of silence to add. See
1825 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1826 for the accepted syntax. Used only if set to non-zero value.
1829 Specify the minimum total duration in the output audio stream. See
1830 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1831 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1832 the input audio length, silence is added to the end, until the value is reached.
1833 This option is mutually exclusive with @option{pad_dur}
1836 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1837 nor @option{whole_dur} option is set, the filter will add silence to the end of
1838 the input stream indefinitely.
1840 @subsection Examples
1844 Add 1024 samples of silence to the end of the input:
1850 Make sure the audio output will contain at least 10000 samples, pad
1851 the input with silence if required:
1853 apad=whole_len=10000
1857 Use @command{ffmpeg} to pad the audio input with silence, so that the
1858 video stream will always result the shortest and will be converted
1859 until the end in the output file when using the @option{shortest}
1862 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1867 Add a phasing effect to the input audio.
1869 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1870 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1872 A description of the accepted parameters follows.
1876 Set input gain. Default is 0.4.
1879 Set output gain. Default is 0.74
1882 Set delay in milliseconds. Default is 3.0.
1885 Set decay. Default is 0.4.
1888 Set modulation speed in Hz. Default is 0.5.
1891 Set modulation type. Default is triangular.
1893 It accepts the following values:
1902 Audio pulsator is something between an autopanner and a tremolo.
1903 But it can produce funny stereo effects as well. Pulsator changes the volume
1904 of the left and right channel based on a LFO (low frequency oscillator) with
1905 different waveforms and shifted phases.
1906 This filter have the ability to define an offset between left and right
1907 channel. An offset of 0 means that both LFO shapes match each other.
1908 The left and right channel are altered equally - a conventional tremolo.
1909 An offset of 50% means that the shape of the right channel is exactly shifted
1910 in phase (or moved backwards about half of the frequency) - pulsator acts as
1911 an autopanner. At 1 both curves match again. Every setting in between moves the
1912 phase shift gapless between all stages and produces some "bypassing" sounds with
1913 sine and triangle waveforms. The more you set the offset near 1 (starting from
1914 the 0.5) the faster the signal passes from the left to the right speaker.
1916 The filter accepts the following options:
1920 Set input gain. By default it is 1. Range is [0.015625 - 64].
1923 Set output gain. By default it is 1. Range is [0.015625 - 64].
1926 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1927 sawup or sawdown. Default is sine.
1930 Set modulation. Define how much of original signal is affected by the LFO.
1933 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1936 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1939 Set pulse width. Default is 1. Allowed range is [0 - 2].
1942 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1945 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1949 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1953 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1954 if timing is set to hz.
1960 Resample the input audio to the specified parameters, using the
1961 libswresample library. If none are specified then the filter will
1962 automatically convert between its input and output.
1964 This filter is also able to stretch/squeeze the audio data to make it match
1965 the timestamps or to inject silence / cut out audio to make it match the
1966 timestamps, do a combination of both or do neither.
1968 The filter accepts the syntax
1969 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1970 expresses a sample rate and @var{resampler_options} is a list of
1971 @var{key}=@var{value} pairs, separated by ":". See the
1972 @ref{Resampler Options,,"Resampler Options" section in the
1973 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1974 for the complete list of supported options.
1976 @subsection Examples
1980 Resample the input audio to 44100Hz:
1986 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1987 samples per second compensation:
1989 aresample=async=1000
1995 Reverse an audio clip.
1997 Warning: This filter requires memory to buffer the entire clip, so trimming
2000 @subsection Examples
2004 Take the first 5 seconds of a clip, and reverse it.
2006 atrim=end=5,areverse
2010 @section asetnsamples
2012 Set the number of samples per each output audio frame.
2014 The last output packet may contain a different number of samples, as
2015 the filter will flush all the remaining samples when the input audio
2018 The filter accepts the following options:
2022 @item nb_out_samples, n
2023 Set the number of frames per each output audio frame. The number is
2024 intended as the number of samples @emph{per each channel}.
2025 Default value is 1024.
2028 If set to 1, the filter will pad the last audio frame with zeroes, so
2029 that the last frame will contain the same number of samples as the
2030 previous ones. Default value is 1.
2033 For example, to set the number of per-frame samples to 1234 and
2034 disable padding for the last frame, use:
2036 asetnsamples=n=1234:p=0
2041 Set the sample rate without altering the PCM data.
2042 This will result in a change of speed and pitch.
2044 The filter accepts the following options:
2047 @item sample_rate, r
2048 Set the output sample rate. Default is 44100 Hz.
2053 Show a line containing various information for each input audio frame.
2054 The input audio is not modified.
2056 The shown line contains a sequence of key/value pairs of the form
2057 @var{key}:@var{value}.
2059 The following values are shown in the output:
2063 The (sequential) number of the input frame, starting from 0.
2066 The presentation timestamp of the input frame, in time base units; the time base
2067 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2070 The presentation timestamp of the input frame in seconds.
2073 position of the frame in the input stream, -1 if this information in
2074 unavailable and/or meaningless (for example in case of synthetic audio)
2083 The sample rate for the audio frame.
2086 The number of samples (per channel) in the frame.
2089 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2090 audio, the data is treated as if all the planes were concatenated.
2092 @item plane_checksums
2093 A list of Adler-32 checksums for each data plane.
2099 Display time domain statistical information about the audio channels.
2100 Statistics are calculated and displayed for each audio channel and,
2101 where applicable, an overall figure is also given.
2103 It accepts the following option:
2106 Short window length in seconds, used for peak and trough RMS measurement.
2107 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2111 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2112 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2115 Available keys for each channel are:
2151 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2152 this @code{lavfi.astats.Overall.Peak_count}.
2154 For description what each key means read below.
2157 Set number of frame after which stats are going to be recalculated.
2158 Default is disabled.
2160 @item measure_perchannel
2161 Select the entries which need to be measured per channel. The metadata keys can
2162 be used as flags, default is @option{all} which measures everything.
2163 @option{none} disables all per channel measurement.
2165 @item measure_overall
2166 Select the entries which need to be measured overall. The metadata keys can
2167 be used as flags, default is @option{all} which measures everything.
2168 @option{none} disables all overall measurement.
2172 A description of each shown parameter follows:
2176 Mean amplitude displacement from zero.
2179 Minimal sample level.
2182 Maximal sample level.
2184 @item Min difference
2185 Minimal difference between two consecutive samples.
2187 @item Max difference
2188 Maximal difference between two consecutive samples.
2190 @item Mean difference
2191 Mean difference between two consecutive samples.
2192 The average of each difference between two consecutive samples.
2194 @item RMS difference
2195 Root Mean Square difference between two consecutive samples.
2199 Standard peak and RMS level measured in dBFS.
2203 Peak and trough values for RMS level measured over a short window.
2206 Standard ratio of peak to RMS level (note: not in dB).
2209 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2210 (i.e. either @var{Min level} or @var{Max level}).
2213 Number of occasions (not the number of samples) that the signal attained either
2214 @var{Min level} or @var{Max level}.
2217 Overall bit depth of audio. Number of bits used for each sample.
2220 Measured dynamic range of audio in dB.
2222 @item Zero crossings
2223 Number of points where the waveform crosses the zero level axis.
2225 @item Zero crossings rate
2226 Rate of Zero crossings and number of audio samples.
2233 The filter accepts exactly one parameter, the audio tempo. If not
2234 specified then the filter will assume nominal 1.0 tempo. Tempo must
2235 be in the [0.5, 100.0] range.
2237 Note that tempo greater than 2 will skip some samples rather than
2238 blend them in. If for any reason this is a concern it is always
2239 possible to daisy-chain several instances of atempo to achieve the
2240 desired product tempo.
2242 @subsection Examples
2246 Slow down audio to 80% tempo:
2252 To speed up audio to 300% tempo:
2258 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2260 atempo=sqrt(3),atempo=sqrt(3)
2266 Trim the input so that the output contains one continuous subpart of the input.
2268 It accepts the following parameters:
2271 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2272 sample with the timestamp @var{start} will be the first sample in the output.
2275 Specify time of the first audio sample that will be dropped, i.e. the
2276 audio sample immediately preceding the one with the timestamp @var{end} will be
2277 the last sample in the output.
2280 Same as @var{start}, except this option sets the start timestamp in samples
2284 Same as @var{end}, except this option sets the end timestamp in samples instead
2288 The maximum duration of the output in seconds.
2291 The number of the first sample that should be output.
2294 The number of the first sample that should be dropped.
2297 @option{start}, @option{end}, and @option{duration} are expressed as time
2298 duration specifications; see
2299 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2301 Note that the first two sets of the start/end options and the @option{duration}
2302 option look at the frame timestamp, while the _sample options simply count the
2303 samples that pass through the filter. So start/end_pts and start/end_sample will
2304 give different results when the timestamps are wrong, inexact or do not start at
2305 zero. Also note that this filter does not modify the timestamps. If you wish
2306 to have the output timestamps start at zero, insert the asetpts filter after the
2309 If multiple start or end options are set, this filter tries to be greedy and
2310 keep all samples that match at least one of the specified constraints. To keep
2311 only the part that matches all the constraints at once, chain multiple atrim
2314 The defaults are such that all the input is kept. So it is possible to set e.g.
2315 just the end values to keep everything before the specified time.
2320 Drop everything except the second minute of input:
2322 ffmpeg -i INPUT -af atrim=60:120
2326 Keep only the first 1000 samples:
2328 ffmpeg -i INPUT -af atrim=end_sample=1000
2335 Apply a two-pole Butterworth band-pass filter with central
2336 frequency @var{frequency}, and (3dB-point) band-width width.
2337 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2338 instead of the default: constant 0dB peak gain.
2339 The filter roll off at 6dB per octave (20dB per decade).
2341 The filter accepts the following options:
2345 Set the filter's central frequency. Default is @code{3000}.
2348 Constant skirt gain if set to 1. Defaults to 0.
2351 Set method to specify band-width of filter.
2366 Specify the band-width of a filter in width_type units.
2369 Specify which channels to filter, by default all available are filtered.
2372 @subsection Commands
2374 This filter supports the following commands:
2377 Change bandpass frequency.
2378 Syntax for the command is : "@var{frequency}"
2381 Change bandpass width_type.
2382 Syntax for the command is : "@var{width_type}"
2385 Change bandpass width.
2386 Syntax for the command is : "@var{width}"
2391 Apply a two-pole Butterworth band-reject filter with central
2392 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2393 The filter roll off at 6dB per octave (20dB per decade).
2395 The filter accepts the following options:
2399 Set the filter's central frequency. Default is @code{3000}.
2402 Set method to specify band-width of filter.
2417 Specify the band-width of a filter in width_type units.
2420 Specify which channels to filter, by default all available are filtered.
2423 @subsection Commands
2425 This filter supports the following commands:
2428 Change bandreject frequency.
2429 Syntax for the command is : "@var{frequency}"
2432 Change bandreject width_type.
2433 Syntax for the command is : "@var{width_type}"
2436 Change bandreject width.
2437 Syntax for the command is : "@var{width}"
2440 @section bass, lowshelf
2442 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2443 shelving filter with a response similar to that of a standard
2444 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2446 The filter accepts the following options:
2450 Give the gain at 0 Hz. Its useful range is about -20
2451 (for a large cut) to +20 (for a large boost).
2452 Beware of clipping when using a positive gain.
2455 Set the filter's central frequency and so can be used
2456 to extend or reduce the frequency range to be boosted or cut.
2457 The default value is @code{100} Hz.
2460 Set method to specify band-width of filter.
2475 Determine how steep is the filter's shelf transition.
2478 Specify which channels to filter, by default all available are filtered.
2481 @subsection Commands
2483 This filter supports the following commands:
2486 Change bass frequency.
2487 Syntax for the command is : "@var{frequency}"
2490 Change bass width_type.
2491 Syntax for the command is : "@var{width_type}"
2495 Syntax for the command is : "@var{width}"
2499 Syntax for the command is : "@var{gain}"
2504 Apply a biquad IIR filter with the given coefficients.
2505 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2506 are the numerator and denominator coefficients respectively.
2507 and @var{channels}, @var{c} specify which channels to filter, by default all
2508 available are filtered.
2510 @subsection Commands
2512 This filter supports the following commands:
2520 Change biquad parameter.
2521 Syntax for the command is : "@var{value}"
2525 Bauer stereo to binaural transformation, which improves headphone listening of
2526 stereo audio records.
2528 To enable compilation of this filter you need to configure FFmpeg with
2529 @code{--enable-libbs2b}.
2531 It accepts the following parameters:
2535 Pre-defined crossfeed level.
2539 Default level (fcut=700, feed=50).
2542 Chu Moy circuit (fcut=700, feed=60).
2545 Jan Meier circuit (fcut=650, feed=95).
2550 Cut frequency (in Hz).
2559 Remap input channels to new locations.
2561 It accepts the following parameters:
2564 Map channels from input to output. The argument is a '|'-separated list of
2565 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2566 @var{in_channel} form. @var{in_channel} can be either the name of the input
2567 channel (e.g. FL for front left) or its index in the input channel layout.
2568 @var{out_channel} is the name of the output channel or its index in the output
2569 channel layout. If @var{out_channel} is not given then it is implicitly an
2570 index, starting with zero and increasing by one for each mapping.
2572 @item channel_layout
2573 The channel layout of the output stream.
2576 If no mapping is present, the filter will implicitly map input channels to
2577 output channels, preserving indices.
2579 @subsection Examples
2583 For example, assuming a 5.1+downmix input MOV file,
2585 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2587 will create an output WAV file tagged as stereo from the downmix channels of
2591 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2593 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2597 @section channelsplit
2599 Split each channel from an input audio stream into a separate output stream.
2601 It accepts the following parameters:
2603 @item channel_layout
2604 The channel layout of the input stream. The default is "stereo".
2606 A channel layout describing the channels to be extracted as separate output streams
2607 or "all" to extract each input channel as a separate stream. The default is "all".
2609 Choosing channels not present in channel layout in the input will result in an error.
2612 @subsection Examples
2616 For example, assuming a stereo input MP3 file,
2618 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2620 will create an output Matroska file with two audio streams, one containing only
2621 the left channel and the other the right channel.
2624 Split a 5.1 WAV file into per-channel files:
2626 ffmpeg -i in.wav -filter_complex
2627 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2628 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2629 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2634 Extract only LFE from a 5.1 WAV file:
2636 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2637 -map '[LFE]' lfe.wav
2642 Add a chorus effect to the audio.
2644 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2646 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2647 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2648 The modulation depth defines the range the modulated delay is played before or after
2649 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2650 sound tuned around the original one, like in a chorus where some vocals are slightly
2653 It accepts the following parameters:
2656 Set input gain. Default is 0.4.
2659 Set output gain. Default is 0.4.
2662 Set delays. A typical delay is around 40ms to 60ms.
2674 @subsection Examples
2680 chorus=0.7:0.9:55:0.4:0.25:2
2686 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2690 Fuller sounding chorus with three delays:
2692 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
2697 Compress or expand the audio's dynamic range.
2699 It accepts the following parameters:
2705 A list of times in seconds for each channel over which the instantaneous level
2706 of the input signal is averaged to determine its volume. @var{attacks} refers to
2707 increase of volume and @var{decays} refers to decrease of volume. For most
2708 situations, the attack time (response to the audio getting louder) should be
2709 shorter than the decay time, because the human ear is more sensitive to sudden
2710 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2711 a typical value for decay is 0.8 seconds.
2712 If specified number of attacks & decays is lower than number of channels, the last
2713 set attack/decay will be used for all remaining channels.
2716 A list of points for the transfer function, specified in dB relative to the
2717 maximum possible signal amplitude. Each key points list must be defined using
2718 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2719 @code{x0/y0 x1/y1 x2/y2 ....}
2721 The input values must be in strictly increasing order but the transfer function
2722 does not have to be monotonically rising. The point @code{0/0} is assumed but
2723 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2724 function are @code{-70/-70|-60/-20|1/0}.
2727 Set the curve radius in dB for all joints. It defaults to 0.01.
2730 Set the additional gain in dB to be applied at all points on the transfer
2731 function. This allows for easy adjustment of the overall gain.
2735 Set an initial volume, in dB, to be assumed for each channel when filtering
2736 starts. This permits the user to supply a nominal level initially, so that, for
2737 example, a very large gain is not applied to initial signal levels before the
2738 companding has begun to operate. A typical value for audio which is initially
2739 quiet is -90 dB. It defaults to 0.
2742 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2743 delayed before being fed to the volume adjuster. Specifying a delay
2744 approximately equal to the attack/decay times allows the filter to effectively
2745 operate in predictive rather than reactive mode. It defaults to 0.
2749 @subsection Examples
2753 Make music with both quiet and loud passages suitable for listening to in a
2756 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2759 Another example for audio with whisper and explosion parts:
2761 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2765 A noise gate for when the noise is at a lower level than the signal:
2767 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2771 Here is another noise gate, this time for when the noise is at a higher level
2772 than the signal (making it, in some ways, similar to squelch):
2774 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2778 2:1 compression starting at -6dB:
2780 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2784 2:1 compression starting at -9dB:
2786 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2790 2:1 compression starting at -12dB:
2792 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2796 2:1 compression starting at -18dB:
2798 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2802 3:1 compression starting at -15dB:
2804 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2810 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2816 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
2820 Hard limiter at -6dB:
2822 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2826 Hard limiter at -12dB:
2828 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2832 Hard noise gate at -35 dB:
2834 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2840 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2844 @section compensationdelay
2846 Compensation Delay Line is a metric based delay to compensate differing
2847 positions of microphones or speakers.
2849 For example, you have recorded guitar with two microphones placed in
2850 different location. Because the front of sound wave has fixed speed in
2851 normal conditions, the phasing of microphones can vary and depends on
2852 their location and interposition. The best sound mix can be achieved when
2853 these microphones are in phase (synchronized). Note that distance of
2854 ~30 cm between microphones makes one microphone to capture signal in
2855 antiphase to another microphone. That makes the final mix sounding moody.
2856 This filter helps to solve phasing problems by adding different delays
2857 to each microphone track and make them synchronized.
2859 The best result can be reached when you take one track as base and
2860 synchronize other tracks one by one with it.
2861 Remember that synchronization/delay tolerance depends on sample rate, too.
2862 Higher sample rates will give more tolerance.
2864 It accepts the following parameters:
2868 Set millimeters distance. This is compensation distance for fine tuning.
2872 Set cm distance. This is compensation distance for tightening distance setup.
2876 Set meters distance. This is compensation distance for hard distance setup.
2880 Set dry amount. Amount of unprocessed (dry) signal.
2884 Set wet amount. Amount of processed (wet) signal.
2888 Set temperature degree in Celsius. This is the temperature of the environment.
2893 Apply headphone crossfeed filter.
2895 Crossfeed is the process of blending the left and right channels of stereo
2897 It is mainly used to reduce extreme stereo separation of low frequencies.
2899 The intent is to produce more speaker like sound to the listener.
2901 The filter accepts the following options:
2905 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2906 This sets gain of low shelf filter for side part of stereo image.
2907 Default is -6dB. Max allowed is -30db when strength is set to 1.
2910 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2911 This sets cut off frequency of low shelf filter. Default is cut off near
2912 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2915 Set input gain. Default is 0.9.
2918 Set output gain. Default is 1.
2921 @section crystalizer
2922 Simple algorithm to expand audio dynamic range.
2924 The filter accepts the following options:
2928 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2929 (unchanged sound) to 10.0 (maximum effect).
2932 Enable clipping. By default is enabled.
2936 Apply a DC shift to the audio.
2938 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2939 in the recording chain) from the audio. The effect of a DC offset is reduced
2940 headroom and hence volume. The @ref{astats} filter can be used to determine if
2941 a signal has a DC offset.
2945 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2949 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2950 used to prevent clipping.
2954 Measure audio dynamic range.
2956 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2957 is found in transition material. And anything less that 8 have very poor dynamics
2958 and is very compressed.
2960 The filter accepts the following options:
2964 Set window length in seconds used to split audio into segments of equal length.
2965 Default is 3 seconds.
2969 Dynamic Audio Normalizer.
2971 This filter applies a certain amount of gain to the input audio in order
2972 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2973 contrast to more "simple" normalization algorithms, the Dynamic Audio
2974 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2975 This allows for applying extra gain to the "quiet" sections of the audio
2976 while avoiding distortions or clipping the "loud" sections. In other words:
2977 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2978 sections, in the sense that the volume of each section is brought to the
2979 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2980 this goal *without* applying "dynamic range compressing". It will retain 100%
2981 of the dynamic range *within* each section of the audio file.
2985 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2986 Default is 500 milliseconds.
2987 The Dynamic Audio Normalizer processes the input audio in small chunks,
2988 referred to as frames. This is required, because a peak magnitude has no
2989 meaning for just a single sample value. Instead, we need to determine the
2990 peak magnitude for a contiguous sequence of sample values. While a "standard"
2991 normalizer would simply use the peak magnitude of the complete file, the
2992 Dynamic Audio Normalizer determines the peak magnitude individually for each
2993 frame. The length of a frame is specified in milliseconds. By default, the
2994 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2995 been found to give good results with most files.
2996 Note that the exact frame length, in number of samples, will be determined
2997 automatically, based on the sampling rate of the individual input audio file.
3000 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3001 number. Default is 31.
3002 Probably the most important parameter of the Dynamic Audio Normalizer is the
3003 @code{window size} of the Gaussian smoothing filter. The filter's window size
3004 is specified in frames, centered around the current frame. For the sake of
3005 simplicity, this must be an odd number. Consequently, the default value of 31
3006 takes into account the current frame, as well as the 15 preceding frames and
3007 the 15 subsequent frames. Using a larger window results in a stronger
3008 smoothing effect and thus in less gain variation, i.e. slower gain
3009 adaptation. Conversely, using a smaller window results in a weaker smoothing
3010 effect and thus in more gain variation, i.e. faster gain adaptation.
3011 In other words, the more you increase this value, the more the Dynamic Audio
3012 Normalizer will behave like a "traditional" normalization filter. On the
3013 contrary, the more you decrease this value, the more the Dynamic Audio
3014 Normalizer will behave like a dynamic range compressor.
3017 Set the target peak value. This specifies the highest permissible magnitude
3018 level for the normalized audio input. This filter will try to approach the
3019 target peak magnitude as closely as possible, but at the same time it also
3020 makes sure that the normalized signal will never exceed the peak magnitude.
3021 A frame's maximum local gain factor is imposed directly by the target peak
3022 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3023 It is not recommended to go above this value.
3026 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3027 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3028 factor for each input frame, i.e. the maximum gain factor that does not
3029 result in clipping or distortion. The maximum gain factor is determined by
3030 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3031 additionally bounds the frame's maximum gain factor by a predetermined
3032 (global) maximum gain factor. This is done in order to avoid excessive gain
3033 factors in "silent" or almost silent frames. By default, the maximum gain
3034 factor is 10.0, For most inputs the default value should be sufficient and
3035 it usually is not recommended to increase this value. Though, for input
3036 with an extremely low overall volume level, it may be necessary to allow even
3037 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3038 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3039 Instead, a "sigmoid" threshold function will be applied. This way, the
3040 gain factors will smoothly approach the threshold value, but never exceed that
3044 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3045 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3046 This means that the maximum local gain factor for each frame is defined
3047 (only) by the frame's highest magnitude sample. This way, the samples can
3048 be amplified as much as possible without exceeding the maximum signal
3049 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3050 Normalizer can also take into account the frame's root mean square,
3051 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3052 determine the power of a time-varying signal. It is therefore considered
3053 that the RMS is a better approximation of the "perceived loudness" than
3054 just looking at the signal's peak magnitude. Consequently, by adjusting all
3055 frames to a constant RMS value, a uniform "perceived loudness" can be
3056 established. If a target RMS value has been specified, a frame's local gain
3057 factor is defined as the factor that would result in exactly that RMS value.
3058 Note, however, that the maximum local gain factor is still restricted by the
3059 frame's highest magnitude sample, in order to prevent clipping.
3062 Enable channels coupling. By default is enabled.
3063 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3064 amount. This means the same gain factor will be applied to all channels, i.e.
3065 the maximum possible gain factor is determined by the "loudest" channel.
3066 However, in some recordings, it may happen that the volume of the different
3067 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3068 In this case, this option can be used to disable the channel coupling. This way,
3069 the gain factor will be determined independently for each channel, depending
3070 only on the individual channel's highest magnitude sample. This allows for
3071 harmonizing the volume of the different channels.
3074 Enable DC bias correction. By default is disabled.
3075 An audio signal (in the time domain) is a sequence of sample values.
3076 In the Dynamic Audio Normalizer these sample values are represented in the
3077 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3078 audio signal, or "waveform", should be centered around the zero point.
3079 That means if we calculate the mean value of all samples in a file, or in a
3080 single frame, then the result should be 0.0 or at least very close to that
3081 value. If, however, there is a significant deviation of the mean value from
3082 0.0, in either positive or negative direction, this is referred to as a
3083 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3084 Audio Normalizer provides optional DC bias correction.
3085 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3086 the mean value, or "DC correction" offset, of each input frame and subtract
3087 that value from all of the frame's sample values which ensures those samples
3088 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3089 boundaries, the DC correction offset values will be interpolated smoothly
3090 between neighbouring frames.
3093 Enable alternative boundary mode. By default is disabled.
3094 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3095 around each frame. This includes the preceding frames as well as the
3096 subsequent frames. However, for the "boundary" frames, located at the very
3097 beginning and at the very end of the audio file, not all neighbouring
3098 frames are available. In particular, for the first few frames in the audio
3099 file, the preceding frames are not known. And, similarly, for the last few
3100 frames in the audio file, the subsequent frames are not known. Thus, the
3101 question arises which gain factors should be assumed for the missing frames
3102 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3103 to deal with this situation. The default boundary mode assumes a gain factor
3104 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3105 "fade out" at the beginning and at the end of the input, respectively.
3108 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3109 By default, the Dynamic Audio Normalizer does not apply "traditional"
3110 compression. This means that signal peaks will not be pruned and thus the
3111 full dynamic range will be retained within each local neighbourhood. However,
3112 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3113 normalization algorithm with a more "traditional" compression.
3114 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3115 (thresholding) function. If (and only if) the compression feature is enabled,
3116 all input frames will be processed by a soft knee thresholding function prior
3117 to the actual normalization process. Put simply, the thresholding function is
3118 going to prune all samples whose magnitude exceeds a certain threshold value.
3119 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3120 value. Instead, the threshold value will be adjusted for each individual
3122 In general, smaller parameters result in stronger compression, and vice versa.
3123 Values below 3.0 are not recommended, because audible distortion may appear.
3128 Make audio easier to listen to on headphones.
3130 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3131 so that when listened to on headphones the stereo image is moved from
3132 inside your head (standard for headphones) to outside and in front of
3133 the listener (standard for speakers).
3139 Apply a two-pole peaking equalisation (EQ) filter. With this
3140 filter, the signal-level at and around a selected frequency can
3141 be increased or decreased, whilst (unlike bandpass and bandreject
3142 filters) that at all other frequencies is unchanged.
3144 In order to produce complex equalisation curves, this filter can
3145 be given several times, each with a different central frequency.
3147 The filter accepts the following options:
3151 Set the filter's central frequency in Hz.
3154 Set method to specify band-width of filter.
3169 Specify the band-width of a filter in width_type units.
3172 Set the required gain or attenuation in dB.
3173 Beware of clipping when using a positive gain.
3176 Specify which channels to filter, by default all available are filtered.
3179 @subsection Examples
3182 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3184 equalizer=f=1000:t=h:width=200:g=-10
3188 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3190 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3194 @subsection Commands
3196 This filter supports the following commands:
3199 Change equalizer frequency.
3200 Syntax for the command is : "@var{frequency}"
3203 Change equalizer width_type.
3204 Syntax for the command is : "@var{width_type}"
3207 Change equalizer width.
3208 Syntax for the command is : "@var{width}"
3211 Change equalizer gain.
3212 Syntax for the command is : "@var{gain}"
3215 @section extrastereo
3217 Linearly increases the difference between left and right channels which
3218 adds some sort of "live" effect to playback.
3220 The filter accepts the following options:
3224 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3225 (average of both channels), with 1.0 sound will be unchanged, with
3226 -1.0 left and right channels will be swapped.
3229 Enable clipping. By default is enabled.
3232 @section firequalizer
3233 Apply FIR Equalization using arbitrary frequency response.
3235 The filter accepts the following option:
3239 Set gain curve equation (in dB). The expression can contain variables:
3242 the evaluated frequency
3246 channel number, set to 0 when multichannels evaluation is disabled
3248 channel id, see libavutil/channel_layout.h, set to the first channel id when
3249 multichannels evaluation is disabled
3253 channel_layout, see libavutil/channel_layout.h
3258 @item gain_interpolate(f)
3259 interpolate gain on frequency f based on gain_entry
3260 @item cubic_interpolate(f)
3261 same as gain_interpolate, but smoother
3263 This option is also available as command. Default is @code{gain_interpolate(f)}.
3266 Set gain entry for gain_interpolate function. The expression can
3270 store gain entry at frequency f with value g
3272 This option is also available as command.
3275 Set filter delay in seconds. Higher value means more accurate.
3276 Default is @code{0.01}.
3279 Set filter accuracy in Hz. Lower value means more accurate.
3280 Default is @code{5}.
3283 Set window function. Acceptable values are:
3286 rectangular window, useful when gain curve is already smooth
3288 hann window (default)
3294 3-terms continuous 1st derivative nuttall window
3296 minimum 3-terms discontinuous nuttall window
3298 4-terms continuous 1st derivative nuttall window
3300 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3302 blackman-harris window
3308 If enabled, use fixed number of audio samples. This improves speed when
3309 filtering with large delay. Default is disabled.
3312 Enable multichannels evaluation on gain. Default is disabled.
3315 Enable zero phase mode by subtracting timestamp to compensate delay.
3316 Default is disabled.
3319 Set scale used by gain. Acceptable values are:
3322 linear frequency, linear gain
3324 linear frequency, logarithmic (in dB) gain (default)
3326 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3328 logarithmic frequency, logarithmic gain
3332 Set file for dumping, suitable for gnuplot.
3335 Set scale for dumpfile. Acceptable values are same with scale option.
3339 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3340 Default is disabled.
3343 Enable minimum phase impulse response. Default is disabled.
3346 @subsection Examples
3351 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3354 lowpass at 1000 Hz with gain_entry:
3356 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3359 custom equalization:
3361 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3364 higher delay with zero phase to compensate delay:
3366 firequalizer=delay=0.1:fixed=on:zero_phase=on
3369 lowpass on left channel, highpass on right channel:
3371 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3372 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3377 Apply a flanging effect to the audio.
3379 The filter accepts the following options:
3383 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3386 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3389 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3393 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3394 Default value is 71.
3397 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3400 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3401 Default value is @var{sinusoidal}.
3404 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3405 Default value is 25.
3408 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3409 Default is @var{linear}.
3413 Apply Haas effect to audio.
3415 Note that this makes most sense to apply on mono signals.
3416 With this filter applied to mono signals it give some directionality and
3417 stretches its stereo image.
3419 The filter accepts the following options:
3423 Set input level. By default is @var{1}, or 0dB
3426 Set output level. By default is @var{1}, or 0dB.
3429 Set gain applied to side part of signal. By default is @var{1}.
3432 Set kind of middle source. Can be one of the following:
3442 Pick middle part signal of stereo image.
3445 Pick side part signal of stereo image.
3449 Change middle phase. By default is disabled.
3452 Set left channel delay. By default is @var{2.05} milliseconds.
3455 Set left channel balance. By default is @var{-1}.
3458 Set left channel gain. By default is @var{1}.
3461 Change left phase. By default is disabled.
3464 Set right channel delay. By defaults is @var{2.12} milliseconds.
3467 Set right channel balance. By default is @var{1}.
3470 Set right channel gain. By default is @var{1}.
3473 Change right phase. By default is enabled.
3478 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3479 embedded HDCD codes is expanded into a 20-bit PCM stream.
3481 The filter supports the Peak Extend and Low-level Gain Adjustment features
3482 of HDCD, and detects the Transient Filter flag.
3485 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3488 When using the filter with wav, note the default encoding for wav is 16-bit,
3489 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3490 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3492 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3493 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3496 The filter accepts the following options:
3499 @item disable_autoconvert
3500 Disable any automatic format conversion or resampling in the filter graph.
3502 @item process_stereo
3503 Process the stereo channels together. If target_gain does not match between
3504 channels, consider it invalid and use the last valid target_gain.
3507 Set the code detect timer period in ms.
3510 Always extend peaks above -3dBFS even if PE isn't signaled.
3513 Replace audio with a solid tone and adjust the amplitude to signal some
3514 specific aspect of the decoding process. The output file can be loaded in
3515 an audio editor alongside the original to aid analysis.
3517 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3524 Gain adjustment level at each sample
3526 Samples where peak extend occurs
3528 Samples where the code detect timer is active
3530 Samples where the target gain does not match between channels
3536 Apply head-related transfer functions (HRTFs) to create virtual
3537 loudspeakers around the user for binaural listening via headphones.
3538 The HRIRs are provided via additional streams, for each channel
3539 one stereo input stream is needed.
3541 The filter accepts the following options:
3545 Set mapping of input streams for convolution.
3546 The argument is a '|'-separated list of channel names in order as they
3547 are given as additional stream inputs for filter.
3548 This also specify number of input streams. Number of input streams
3549 must be not less than number of channels in first stream plus one.
3552 Set gain applied to audio. Value is in dB. Default is 0.
3555 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3556 processing audio in time domain which is slow.
3557 @var{freq} is processing audio in frequency domain which is fast.
3558 Default is @var{freq}.
3561 Set custom gain for LFE channels. Value is in dB. Default is 0.
3564 Set size of frame in number of samples which will be processed at once.
3565 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3568 Set format of hrir stream.
3569 Default value is @var{stereo}. Alternative value is @var{multich}.
3570 If value is set to @var{stereo}, number of additional streams should
3571 be greater or equal to number of input channels in first input stream.
3572 Also each additional stream should have stereo number of channels.
3573 If value is set to @var{multich}, number of additional streams should
3574 be exactly one. Also number of input channels of additional stream
3575 should be equal or greater than twice number of channels of first input
3579 @subsection Examples
3583 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3584 each amovie filter use stereo file with IR coefficients as input.
3585 The files give coefficients for each position of virtual loudspeaker:
3588 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3593 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3594 but now in @var{multich} @var{hrir} format.
3596 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3603 Apply a high-pass filter with 3dB point frequency.
3604 The filter can be either single-pole, or double-pole (the default).
3605 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3607 The filter accepts the following options:
3611 Set frequency in Hz. Default is 3000.
3614 Set number of poles. Default is 2.
3617 Set method to specify band-width of filter.
3632 Specify the band-width of a filter in width_type units.
3633 Applies only to double-pole filter.
3634 The default is 0.707q and gives a Butterworth response.
3637 Specify which channels to filter, by default all available are filtered.
3640 @subsection Commands
3642 This filter supports the following commands:
3645 Change highpass frequency.
3646 Syntax for the command is : "@var{frequency}"
3649 Change highpass width_type.
3650 Syntax for the command is : "@var{width_type}"
3653 Change highpass width.
3654 Syntax for the command is : "@var{width}"
3659 Join multiple input streams into one multi-channel stream.
3661 It accepts the following parameters:
3665 The number of input streams. It defaults to 2.
3667 @item channel_layout
3668 The desired output channel layout. It defaults to stereo.
3671 Map channels from inputs to output. The argument is a '|'-separated list of
3672 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3673 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3674 can be either the name of the input channel (e.g. FL for front left) or its
3675 index in the specified input stream. @var{out_channel} is the name of the output
3679 The filter will attempt to guess the mappings when they are not specified
3680 explicitly. It does so by first trying to find an unused matching input channel
3681 and if that fails it picks the first unused input channel.
3683 Join 3 inputs (with properly set channel layouts):
3685 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3688 Build a 5.1 output from 6 single-channel streams:
3690 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3691 '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'
3697 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3699 To enable compilation of this filter you need to configure FFmpeg with
3700 @code{--enable-ladspa}.
3704 Specifies the name of LADSPA plugin library to load. If the environment
3705 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3706 each one of the directories specified by the colon separated list in
3707 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3708 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3709 @file{/usr/lib/ladspa/}.
3712 Specifies the plugin within the library. Some libraries contain only
3713 one plugin, but others contain many of them. If this is not set filter
3714 will list all available plugins within the specified library.
3717 Set the '|' separated list of controls which are zero or more floating point
3718 values that determine the behavior of the loaded plugin (for example delay,
3720 Controls need to be defined using the following syntax:
3721 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3722 @var{valuei} is the value set on the @var{i}-th control.
3723 Alternatively they can be also defined using the following syntax:
3724 @var{value0}|@var{value1}|@var{value2}|..., where
3725 @var{valuei} is the value set on the @var{i}-th control.
3726 If @option{controls} is set to @code{help}, all available controls and
3727 their valid ranges are printed.
3729 @item sample_rate, s
3730 Specify the sample rate, default to 44100. Only used if plugin have
3734 Set the number of samples per channel per each output frame, default
3735 is 1024. Only used if plugin have zero inputs.
3738 Set the minimum duration of the sourced audio. See
3739 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3740 for the accepted syntax.
3741 Note that the resulting duration may be greater than the specified duration,
3742 as the generated audio is always cut at the end of a complete frame.
3743 If not specified, or the expressed duration is negative, the audio is
3744 supposed to be generated forever.
3745 Only used if plugin have zero inputs.
3749 @subsection Examples
3753 List all available plugins within amp (LADSPA example plugin) library:
3759 List all available controls and their valid ranges for @code{vcf_notch}
3760 plugin from @code{VCF} library:
3762 ladspa=f=vcf:p=vcf_notch:c=help
3766 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3769 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3773 Add reverberation to the audio using TAP-plugins
3774 (Tom's Audio Processing plugins):
3776 ladspa=file=tap_reverb:tap_reverb
3780 Generate white noise, with 0.2 amplitude:
3782 ladspa=file=cmt:noise_source_white:c=c0=.2
3786 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3787 @code{C* Audio Plugin Suite} (CAPS) library:
3789 ladspa=file=caps:Click:c=c1=20'
3793 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3795 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3799 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3800 @code{SWH Plugins} collection:
3802 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3806 Attenuate low frequencies using Multiband EQ from Steve Harris
3807 @code{SWH Plugins} collection:
3809 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3813 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3816 ladspa=caps:Narrower
3820 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3822 ladspa=caps:White:.2
3826 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3828 ladspa=caps:Fractal:c=c1=1
3832 Dynamic volume normalization using @code{VLevel} plugin:
3834 ladspa=vlevel-ladspa:vlevel_mono
3838 @subsection Commands
3840 This filter supports the following commands:
3843 Modify the @var{N}-th control value.
3845 If the specified value is not valid, it is ignored and prior one is kept.
3850 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3851 Support for both single pass (livestreams, files) and double pass (files) modes.
3852 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3853 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3854 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3856 The filter accepts the following options:
3860 Set integrated loudness target.
3861 Range is -70.0 - -5.0. Default value is -24.0.
3864 Set loudness range target.
3865 Range is 1.0 - 20.0. Default value is 7.0.
3868 Set maximum true peak.
3869 Range is -9.0 - +0.0. Default value is -2.0.
3871 @item measured_I, measured_i
3872 Measured IL of input file.
3873 Range is -99.0 - +0.0.
3875 @item measured_LRA, measured_lra
3876 Measured LRA of input file.
3877 Range is 0.0 - 99.0.
3879 @item measured_TP, measured_tp
3880 Measured true peak of input file.
3881 Range is -99.0 - +99.0.
3883 @item measured_thresh
3884 Measured threshold of input file.
3885 Range is -99.0 - +0.0.
3888 Set offset gain. Gain is applied before the true-peak limiter.
3889 Range is -99.0 - +99.0. Default is +0.0.
3892 Normalize linearly if possible.
3893 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3894 to be specified in order to use this mode.
3895 Options are true or false. Default is true.
3898 Treat mono input files as "dual-mono". If a mono file is intended for playback
3899 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3900 If set to @code{true}, this option will compensate for this effect.
3901 Multi-channel input files are not affected by this option.
3902 Options are true or false. Default is false.
3905 Set print format for stats. Options are summary, json, or none.
3906 Default value is none.
3911 Apply a low-pass filter with 3dB point frequency.
3912 The filter can be either single-pole or double-pole (the default).
3913 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3915 The filter accepts the following options:
3919 Set frequency in Hz. Default is 500.
3922 Set number of poles. Default is 2.
3925 Set method to specify band-width of filter.
3940 Specify the band-width of a filter in width_type units.
3941 Applies only to double-pole filter.
3942 The default is 0.707q and gives a Butterworth response.
3945 Specify which channels to filter, by default all available are filtered.
3948 @subsection Examples
3951 Lowpass only LFE channel, it LFE is not present it does nothing:
3957 @subsection Commands
3959 This filter supports the following commands:
3962 Change lowpass frequency.
3963 Syntax for the command is : "@var{frequency}"
3966 Change lowpass width_type.
3967 Syntax for the command is : "@var{width_type}"
3970 Change lowpass width.
3971 Syntax for the command is : "@var{width}"
3976 Load a LV2 (LADSPA Version 2) plugin.
3978 To enable compilation of this filter you need to configure FFmpeg with
3979 @code{--enable-lv2}.
3983 Specifies the plugin URI. You may need to escape ':'.
3986 Set the '|' separated list of controls which are zero or more floating point
3987 values that determine the behavior of the loaded plugin (for example delay,
3989 If @option{controls} is set to @code{help}, all available controls and
3990 their valid ranges are printed.
3992 @item sample_rate, s
3993 Specify the sample rate, default to 44100. Only used if plugin have
3997 Set the number of samples per channel per each output frame, default
3998 is 1024. Only used if plugin have zero inputs.
4001 Set the minimum duration of the sourced audio. See
4002 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4003 for the accepted syntax.
4004 Note that the resulting duration may be greater than the specified duration,
4005 as the generated audio is always cut at the end of a complete frame.
4006 If not specified, or the expressed duration is negative, the audio is
4007 supposed to be generated forever.
4008 Only used if plugin have zero inputs.
4011 @subsection Examples
4015 Apply bass enhancer plugin from Calf:
4017 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4021 Apply vinyl plugin from Calf:
4023 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4027 Apply bit crusher plugin from ArtyFX:
4029 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4034 Multiband Compress or expand the audio's dynamic range.
4036 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4037 This is akin to the crossover of a loudspeaker, and results in flat frequency
4038 response when absent compander action.
4040 It accepts the following parameters:
4044 This option syntax is:
4045 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4046 For explanation of each item refer to compand filter documentation.
4052 Mix channels with specific gain levels. The filter accepts the output
4053 channel layout followed by a set of channels definitions.
4055 This filter is also designed to efficiently remap the channels of an audio
4058 The filter accepts parameters of the form:
4059 "@var{l}|@var{outdef}|@var{outdef}|..."
4063 output channel layout or number of channels
4066 output channel specification, of the form:
4067 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4070 output channel to define, either a channel name (FL, FR, etc.) or a channel
4071 number (c0, c1, etc.)
4074 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4077 input channel to use, see out_name for details; it is not possible to mix
4078 named and numbered input channels
4081 If the `=' in a channel specification is replaced by `<', then the gains for
4082 that specification will be renormalized so that the total is 1, thus
4083 avoiding clipping noise.
4085 @subsection Mixing examples
4087 For example, if you want to down-mix from stereo to mono, but with a bigger
4088 factor for the left channel:
4090 pan=1c|c0=0.9*c0+0.1*c1
4093 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4094 7-channels surround:
4096 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4099 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4100 that should be preferred (see "-ac" option) unless you have very specific
4103 @subsection Remapping examples
4105 The channel remapping will be effective if, and only if:
4108 @item gain coefficients are zeroes or ones,
4109 @item only one input per channel output,
4112 If all these conditions are satisfied, the filter will notify the user ("Pure
4113 channel mapping detected"), and use an optimized and lossless method to do the
4116 For example, if you have a 5.1 source and want a stereo audio stream by
4117 dropping the extra channels:
4119 pan="stereo| c0=FL | c1=FR"
4122 Given the same source, you can also switch front left and front right channels
4123 and keep the input channel layout:
4125 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4128 If the input is a stereo audio stream, you can mute the front left channel (and
4129 still keep the stereo channel layout) with:
4134 Still with a stereo audio stream input, you can copy the right channel in both
4135 front left and right:
4137 pan="stereo| c0=FR | c1=FR"
4142 ReplayGain scanner filter. This filter takes an audio stream as an input and
4143 outputs it unchanged.
4144 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4148 Convert the audio sample format, sample rate and channel layout. It is
4149 not meant to be used directly.
4152 Apply time-stretching and pitch-shifting with librubberband.
4154 To enable compilation of this filter, you need to configure FFmpeg with
4155 @code{--enable-librubberband}.
4157 The filter accepts the following options:
4161 Set tempo scale factor.
4164 Set pitch scale factor.
4167 Set transients detector.
4168 Possible values are:
4177 Possible values are:
4186 Possible values are:
4193 Set processing window size.
4194 Possible values are:
4203 Possible values are:
4210 Enable formant preservation when shift pitching.
4211 Possible values are:
4219 Possible values are:
4228 Possible values are:
4235 @section sidechaincompress
4237 This filter acts like normal compressor but has the ability to compress
4238 detected signal using second input signal.
4239 It needs two input streams and returns one output stream.
4240 First input stream will be processed depending on second stream signal.
4241 The filtered signal then can be filtered with other filters in later stages of
4242 processing. See @ref{pan} and @ref{amerge} filter.
4244 The filter accepts the following options:
4248 Set input gain. Default is 1. Range is between 0.015625 and 64.
4251 If a signal of second stream raises above this level it will affect the gain
4252 reduction of first stream.
4253 By default is 0.125. Range is between 0.00097563 and 1.
4256 Set a ratio about which the signal is reduced. 1:2 means that if the level
4257 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4258 Default is 2. Range is between 1 and 20.
4261 Amount of milliseconds the signal has to rise above the threshold before gain
4262 reduction starts. Default is 20. Range is between 0.01 and 2000.
4265 Amount of milliseconds the signal has to fall below the threshold before
4266 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4269 Set the amount by how much signal will be amplified after processing.
4270 Default is 1. Range is from 1 to 64.
4273 Curve the sharp knee around the threshold to enter gain reduction more softly.
4274 Default is 2.82843. Range is between 1 and 8.
4277 Choose if the @code{average} level between all channels of side-chain stream
4278 or the louder(@code{maximum}) channel of side-chain stream affects the
4279 reduction. Default is @code{average}.
4282 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4283 of @code{rms}. Default is @code{rms} which is mainly smoother.
4286 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4289 How much to use compressed signal in output. Default is 1.
4290 Range is between 0 and 1.
4293 @subsection Examples
4297 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4298 depending on the signal of 2nd input and later compressed signal to be
4299 merged with 2nd input:
4301 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4305 @section sidechaingate
4307 A sidechain gate acts like a normal (wideband) gate but has the ability to
4308 filter the detected signal before sending it to the gain reduction stage.
4309 Normally a gate uses the full range signal to detect a level above the
4311 For example: If you cut all lower frequencies from your sidechain signal
4312 the gate will decrease the volume of your track only if not enough highs
4313 appear. With this technique you are able to reduce the resonation of a
4314 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4316 It needs two input streams and returns one output stream.
4317 First input stream will be processed depending on second stream signal.
4319 The filter accepts the following options:
4323 Set input level before filtering.
4324 Default is 1. Allowed range is from 0.015625 to 64.
4327 Set the level of gain reduction when the signal is below the threshold.
4328 Default is 0.06125. Allowed range is from 0 to 1.
4331 If a signal rises above this level the gain reduction is released.
4332 Default is 0.125. Allowed range is from 0 to 1.
4335 Set a ratio about which the signal is reduced.
4336 Default is 2. Allowed range is from 1 to 9000.
4339 Amount of milliseconds the signal has to rise above the threshold before gain
4341 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4344 Amount of milliseconds the signal has to fall below the threshold before the
4345 reduction is increased again. Default is 250 milliseconds.
4346 Allowed range is from 0.01 to 9000.
4349 Set amount of amplification of signal after processing.
4350 Default is 1. Allowed range is from 1 to 64.
4353 Curve the sharp knee around the threshold to enter gain reduction more softly.
4354 Default is 2.828427125. Allowed range is from 1 to 8.
4357 Choose if exact signal should be taken for detection or an RMS like one.
4358 Default is rms. Can be peak or rms.
4361 Choose if the average level between all channels or the louder channel affects
4363 Default is average. Can be average or maximum.
4366 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4369 @section silencedetect
4371 Detect silence in an audio stream.
4373 This filter logs a message when it detects that the input audio volume is less
4374 or equal to a noise tolerance value for a duration greater or equal to the
4375 minimum detected noise duration.
4377 The printed times and duration are expressed in seconds.
4379 The filter accepts the following options:
4383 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4384 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4387 Set silence duration until notification (default is 2 seconds).
4390 Process each channel separately, instead of combined. By default is disabled.
4393 @subsection Examples
4397 Detect 5 seconds of silence with -50dB noise tolerance:
4399 silencedetect=n=-50dB:d=5
4403 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4404 tolerance in @file{silence.mp3}:
4406 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4410 @section silenceremove
4412 Remove silence from the beginning, middle or end of the audio.
4414 The filter accepts the following options:
4418 This value is used to indicate if audio should be trimmed at beginning of
4419 the audio. A value of zero indicates no silence should be trimmed from the
4420 beginning. When specifying a non-zero value, it trims audio up until it
4421 finds non-silence. Normally, when trimming silence from beginning of audio
4422 the @var{start_periods} will be @code{1} but it can be increased to higher
4423 values to trim all audio up to specific count of non-silence periods.
4424 Default value is @code{0}.
4426 @item start_duration
4427 Specify the amount of time that non-silence must be detected before it stops
4428 trimming audio. By increasing the duration, bursts of noises can be treated
4429 as silence and trimmed off. Default value is @code{0}.
4431 @item start_threshold
4432 This indicates what sample value should be treated as silence. For digital
4433 audio, a value of @code{0} may be fine but for audio recorded from analog,
4434 you may wish to increase the value to account for background noise.
4435 Can be specified in dB (in case "dB" is appended to the specified value)
4436 or amplitude ratio. Default value is @code{0}.
4439 Specify max duration of silence at beginning that will be kept after
4440 trimming. Default is 0, which is equal to trimming all samples detected
4444 Specify mode of detection of silence end in start of multi-channel audio.
4445 Can be @var{any} or @var{all}. Default is @var{any}.
4446 With @var{any}, any sample that is detected as non-silence will cause
4447 stopped trimming of silence.
4448 With @var{all}, only if all channels are detected as non-silence will cause
4449 stopped trimming of silence.
4452 Set the count for trimming silence from the end of audio.
4453 To remove silence from the middle of a file, specify a @var{stop_periods}
4454 that is negative. This value is then treated as a positive value and is
4455 used to indicate the effect should restart processing as specified by
4456 @var{start_periods}, making it suitable for removing periods of silence
4457 in the middle of the audio.
4458 Default value is @code{0}.
4461 Specify a duration of silence that must exist before audio is not copied any
4462 more. By specifying a higher duration, silence that is wanted can be left in
4464 Default value is @code{0}.
4466 @item stop_threshold
4467 This is the same as @option{start_threshold} but for trimming silence from
4469 Can be specified in dB (in case "dB" is appended to the specified value)
4470 or amplitude ratio. Default value is @code{0}.
4473 Specify max duration of silence at end that will be kept after
4474 trimming. Default is 0, which is equal to trimming all samples detected
4478 Specify mode of detection of silence start in end of multi-channel audio.
4479 Can be @var{any} or @var{all}. Default is @var{any}.
4480 With @var{any}, any sample that is detected as non-silence will cause
4481 stopped trimming of silence.
4482 With @var{all}, only if all channels are detected as non-silence will cause
4483 stopped trimming of silence.
4486 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4487 and works better with digital silence which is exactly 0.
4488 Default value is @code{rms}.
4491 Set duration in number of seconds used to calculate size of window in number
4492 of samples for detecting silence.
4493 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4496 @subsection Examples
4500 The following example shows how this filter can be used to start a recording
4501 that does not contain the delay at the start which usually occurs between
4502 pressing the record button and the start of the performance:
4504 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4508 Trim all silence encountered from beginning to end where there is more than 1
4509 second of silence in audio:
4511 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4517 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4518 loudspeakers around the user for binaural listening via headphones (audio
4519 formats up to 9 channels supported).
4520 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4521 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4522 Austrian Academy of Sciences.
4524 To enable compilation of this filter you need to configure FFmpeg with
4525 @code{--enable-libmysofa}.
4527 The filter accepts the following options:
4531 Set the SOFA file used for rendering.
4534 Set gain applied to audio. Value is in dB. Default is 0.
4537 Set rotation of virtual loudspeakers in deg. Default is 0.
4540 Set elevation of virtual speakers in deg. Default is 0.
4543 Set distance in meters between loudspeakers and the listener with near-field
4544 HRTFs. Default is 1.
4547 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4548 processing audio in time domain which is slow.
4549 @var{freq} is processing audio in frequency domain which is fast.
4550 Default is @var{freq}.
4553 Set custom positions of virtual loudspeakers. Syntax for this option is:
4554 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4555 Each virtual loudspeaker is described with short channel name following with
4556 azimuth and elevation in degrees.
4557 Each virtual loudspeaker description is separated by '|'.
4558 For example to override front left and front right channel positions use:
4559 'speakers=FL 45 15|FR 345 15'.
4560 Descriptions with unrecognised channel names are ignored.
4563 Set custom gain for LFE channels. Value is in dB. Default is 0.
4566 Set custom frame size in number of samples. Default is 1024.
4567 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4568 is set to @var{freq}.
4571 Should all IRs be normalized upon importing SOFA file.
4572 By default is enabled.
4575 Should nearest IRs be interpolated with neighbor IRs if exact position
4576 does not match. By default is disabled.
4579 Minphase all IRs upon loading of SOFA file. By default is disabled.
4582 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4585 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4588 @subsection Examples
4592 Using ClubFritz6 sofa file:
4594 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4598 Using ClubFritz12 sofa file and bigger radius with small rotation:
4600 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4604 Similar as above but with custom speaker positions for front left, front right, back left and back right
4605 and also with custom gain:
4607 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4611 @section stereotools
4613 This filter has some handy utilities to manage stereo signals, for converting
4614 M/S stereo recordings to L/R signal while having control over the parameters
4615 or spreading the stereo image of master track.
4617 The filter accepts the following options:
4621 Set input level before filtering for both channels. Defaults is 1.
4622 Allowed range is from 0.015625 to 64.
4625 Set output level after filtering for both channels. Defaults is 1.
4626 Allowed range is from 0.015625 to 64.
4629 Set input balance between both channels. Default is 0.
4630 Allowed range is from -1 to 1.
4633 Set output balance between both channels. Default is 0.
4634 Allowed range is from -1 to 1.
4637 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4638 clipping. Disabled by default.
4641 Mute the left channel. Disabled by default.
4644 Mute the right channel. Disabled by default.
4647 Change the phase of the left channel. Disabled by default.
4650 Change the phase of the right channel. Disabled by default.
4653 Set stereo mode. Available values are:
4657 Left/Right to Left/Right, this is default.
4660 Left/Right to Mid/Side.
4663 Mid/Side to Left/Right.
4666 Left/Right to Left/Left.
4669 Left/Right to Right/Right.
4672 Left/Right to Left + Right.
4675 Left/Right to Right/Left.
4678 Mid/Side to Left/Left.
4681 Mid/Side to Right/Right.
4685 Set level of side signal. Default is 1.
4686 Allowed range is from 0.015625 to 64.
4689 Set balance of side signal. Default is 0.
4690 Allowed range is from -1 to 1.
4693 Set level of the middle signal. Default is 1.
4694 Allowed range is from 0.015625 to 64.
4697 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4700 Set stereo base between mono and inversed channels. Default is 0.
4701 Allowed range is from -1 to 1.
4704 Set delay in milliseconds how much to delay left from right channel and
4705 vice versa. Default is 0. Allowed range is from -20 to 20.
4708 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4711 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4713 @item bmode_in, bmode_out
4714 Set balance mode for balance_in/balance_out option.
4716 Can be one of the following:
4720 Classic balance mode. Attenuate one channel at time.
4721 Gain is raised up to 1.
4724 Similar as classic mode above but gain is raised up to 2.
4727 Equal power distribution, from -6dB to +6dB range.
4731 @subsection Examples
4735 Apply karaoke like effect:
4737 stereotools=mlev=0.015625
4741 Convert M/S signal to L/R:
4743 "stereotools=mode=ms>lr"
4747 @section stereowiden
4749 This filter enhance the stereo effect by suppressing signal common to both
4750 channels and by delaying the signal of left into right and vice versa,
4751 thereby widening the stereo effect.
4753 The filter accepts the following options:
4757 Time in milliseconds of the delay of left signal into right and vice versa.
4758 Default is 20 milliseconds.
4761 Amount of gain in delayed signal into right and vice versa. Gives a delay
4762 effect of left signal in right output and vice versa which gives widening
4763 effect. Default is 0.3.
4766 Cross feed of left into right with inverted phase. This helps in suppressing
4767 the mono. If the value is 1 it will cancel all the signal common to both
4768 channels. Default is 0.3.
4771 Set level of input signal of original channel. Default is 0.8.
4774 @section superequalizer
4775 Apply 18 band equalizer.
4777 The filter accepts the following options:
4784 Set 131Hz band gain.
4786 Set 185Hz band gain.
4788 Set 262Hz band gain.
4790 Set 370Hz band gain.
4792 Set 523Hz band gain.
4794 Set 740Hz band gain.
4796 Set 1047Hz band gain.
4798 Set 1480Hz band gain.
4800 Set 2093Hz band gain.
4802 Set 2960Hz band gain.
4804 Set 4186Hz band gain.
4806 Set 5920Hz band gain.
4808 Set 8372Hz band gain.
4810 Set 11840Hz band gain.
4812 Set 16744Hz band gain.
4814 Set 20000Hz band gain.
4818 Apply audio surround upmix filter.
4820 This filter allows to produce multichannel output from audio stream.
4822 The filter accepts the following options:
4826 Set output channel layout. By default, this is @var{5.1}.
4828 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4829 for the required syntax.
4832 Set input channel layout. By default, this is @var{stereo}.
4834 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4835 for the required syntax.
4838 Set input volume level. By default, this is @var{1}.
4841 Set output volume level. By default, this is @var{1}.
4844 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4847 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4850 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4853 Set front center input volume. By default, this is @var{1}.
4856 Set front center output volume. By default, this is @var{1}.
4859 Set LFE input volume. By default, this is @var{1}.
4862 Set LFE output volume. By default, this is @var{1}.
4865 @section treble, highshelf
4867 Boost or cut treble (upper) frequencies of the audio using a two-pole
4868 shelving filter with a response similar to that of a standard
4869 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4871 The filter accepts the following options:
4875 Give the gain at whichever is the lower of ~22 kHz and the
4876 Nyquist frequency. Its useful range is about -20 (for a large cut)
4877 to +20 (for a large boost). Beware of clipping when using a positive gain.
4880 Set the filter's central frequency and so can be used
4881 to extend or reduce the frequency range to be boosted or cut.
4882 The default value is @code{3000} Hz.
4885 Set method to specify band-width of filter.
4900 Determine how steep is the filter's shelf transition.
4903 Specify which channels to filter, by default all available are filtered.
4906 @subsection Commands
4908 This filter supports the following commands:
4911 Change treble frequency.
4912 Syntax for the command is : "@var{frequency}"
4915 Change treble width_type.
4916 Syntax for the command is : "@var{width_type}"
4919 Change treble width.
4920 Syntax for the command is : "@var{width}"
4924 Syntax for the command is : "@var{gain}"
4929 Sinusoidal amplitude modulation.
4931 The filter accepts the following options:
4935 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4936 (20 Hz or lower) will result in a tremolo effect.
4937 This filter may also be used as a ring modulator by specifying
4938 a modulation frequency higher than 20 Hz.
4939 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4942 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4943 Default value is 0.5.
4948 Sinusoidal phase modulation.
4950 The filter accepts the following options:
4954 Modulation frequency in Hertz.
4955 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4958 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4959 Default value is 0.5.
4964 Adjust the input audio volume.
4966 It accepts the following parameters:
4970 Set audio volume expression.
4972 Output values are clipped to the maximum value.
4974 The output audio volume is given by the relation:
4976 @var{output_volume} = @var{volume} * @var{input_volume}
4979 The default value for @var{volume} is "1.0".
4982 This parameter represents the mathematical precision.
4984 It determines which input sample formats will be allowed, which affects the
4985 precision of the volume scaling.
4989 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4991 32-bit floating-point; this limits input sample format to FLT. (default)
4993 64-bit floating-point; this limits input sample format to DBL.
4997 Choose the behaviour on encountering ReplayGain side data in input frames.
5001 Remove ReplayGain side data, ignoring its contents (the default).
5004 Ignore ReplayGain side data, but leave it in the frame.
5007 Prefer the track gain, if present.
5010 Prefer the album gain, if present.
5013 @item replaygain_preamp
5014 Pre-amplification gain in dB to apply to the selected replaygain gain.
5016 Default value for @var{replaygain_preamp} is 0.0.
5019 Set when the volume expression is evaluated.
5021 It accepts the following values:
5024 only evaluate expression once during the filter initialization, or
5025 when the @samp{volume} command is sent
5028 evaluate expression for each incoming frame
5031 Default value is @samp{once}.
5034 The volume expression can contain the following parameters.
5038 frame number (starting at zero)
5041 @item nb_consumed_samples
5042 number of samples consumed by the filter
5044 number of samples in the current frame
5046 original frame position in the file
5052 PTS at start of stream
5054 time at start of stream
5060 last set volume value
5063 Note that when @option{eval} is set to @samp{once} only the
5064 @var{sample_rate} and @var{tb} variables are available, all other
5065 variables will evaluate to NAN.
5067 @subsection Commands
5069 This filter supports the following commands:
5072 Modify the volume expression.
5073 The command accepts the same syntax of the corresponding option.
5075 If the specified expression is not valid, it is kept at its current
5077 @item replaygain_noclip
5078 Prevent clipping by limiting the gain applied.
5080 Default value for @var{replaygain_noclip} is 1.
5084 @subsection Examples
5088 Halve the input audio volume:
5092 volume=volume=-6.0206dB
5095 In all the above example the named key for @option{volume} can be
5096 omitted, for example like in:
5102 Increase input audio power by 6 decibels using fixed-point precision:
5104 volume=volume=6dB:precision=fixed
5108 Fade volume after time 10 with an annihilation period of 5 seconds:
5110 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5114 @section volumedetect
5116 Detect the volume of the input video.
5118 The filter has no parameters. The input is not modified. Statistics about
5119 the volume will be printed in the log when the input stream end is reached.
5121 In particular it will show the mean volume (root mean square), maximum
5122 volume (on a per-sample basis), and the beginning of a histogram of the
5123 registered volume values (from the maximum value to a cumulated 1/1000 of
5126 All volumes are in decibels relative to the maximum PCM value.
5128 @subsection Examples
5130 Here is an excerpt of the output:
5132 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5133 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5134 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5135 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5136 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5137 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5138 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5139 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5140 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5146 The mean square energy is approximately -27 dB, or 10^-2.7.
5148 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5150 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5153 In other words, raising the volume by +4 dB does not cause any clipping,
5154 raising it by +5 dB causes clipping for 6 samples, etc.
5156 @c man end AUDIO FILTERS
5158 @chapter Audio Sources
5159 @c man begin AUDIO SOURCES
5161 Below is a description of the currently available audio sources.
5165 Buffer audio frames, and make them available to the filter chain.
5167 This source is mainly intended for a programmatic use, in particular
5168 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5170 It accepts the following parameters:
5174 The timebase which will be used for timestamps of submitted frames. It must be
5175 either a floating-point number or in @var{numerator}/@var{denominator} form.
5178 The sample rate of the incoming audio buffers.
5181 The sample format of the incoming audio buffers.
5182 Either a sample format name or its corresponding integer representation from
5183 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5185 @item channel_layout
5186 The channel layout of the incoming audio buffers.
5187 Either a channel layout name from channel_layout_map in
5188 @file{libavutil/channel_layout.c} or its corresponding integer representation
5189 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5192 The number of channels of the incoming audio buffers.
5193 If both @var{channels} and @var{channel_layout} are specified, then they
5198 @subsection Examples
5201 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5204 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5205 Since the sample format with name "s16p" corresponds to the number
5206 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5209 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5214 Generate an audio signal specified by an expression.
5216 This source accepts in input one or more expressions (one for each
5217 channel), which are evaluated and used to generate a corresponding
5220 This source accepts the following options:
5224 Set the '|'-separated expressions list for each separate channel. In case the
5225 @option{channel_layout} option is not specified, the selected channel layout
5226 depends on the number of provided expressions. Otherwise the last
5227 specified expression is applied to the remaining output channels.
5229 @item channel_layout, c
5230 Set the channel layout. The number of channels in the specified layout
5231 must be equal to the number of specified expressions.
5234 Set the minimum duration of the sourced audio. See
5235 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5236 for the accepted syntax.
5237 Note that the resulting duration may be greater than the specified
5238 duration, as the generated audio is always cut at the end of a
5241 If not specified, or the expressed duration is negative, the audio is
5242 supposed to be generated forever.
5245 Set the number of samples per channel per each output frame,
5248 @item sample_rate, s
5249 Specify the sample rate, default to 44100.
5252 Each expression in @var{exprs} can contain the following constants:
5256 number of the evaluated sample, starting from 0
5259 time of the evaluated sample expressed in seconds, starting from 0
5266 @subsection Examples
5276 Generate a sin signal with frequency of 440 Hz, set sample rate to
5279 aevalsrc="sin(440*2*PI*t):s=8000"
5283 Generate a two channels signal, specify the channel layout (Front
5284 Center + Back Center) explicitly:
5286 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5290 Generate white noise:
5292 aevalsrc="-2+random(0)"
5296 Generate an amplitude modulated signal:
5298 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5302 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5304 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5311 The null audio source, return unprocessed audio frames. It is mainly useful
5312 as a template and to be employed in analysis / debugging tools, or as
5313 the source for filters which ignore the input data (for example the sox
5316 This source accepts the following options:
5320 @item channel_layout, cl
5322 Specifies the channel layout, and can be either an integer or a string
5323 representing a channel layout. The default value of @var{channel_layout}
5326 Check the channel_layout_map definition in
5327 @file{libavutil/channel_layout.c} for the mapping between strings and
5328 channel layout values.
5330 @item sample_rate, r
5331 Specifies the sample rate, and defaults to 44100.
5334 Set the number of samples per requested frames.
5338 @subsection Examples
5342 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5344 anullsrc=r=48000:cl=4
5348 Do the same operation with a more obvious syntax:
5350 anullsrc=r=48000:cl=mono
5354 All the parameters need to be explicitly defined.
5358 Synthesize a voice utterance using the libflite library.
5360 To enable compilation of this filter you need to configure FFmpeg with
5361 @code{--enable-libflite}.
5363 Note that versions of the flite library prior to 2.0 are not thread-safe.
5365 The filter accepts the following options:
5370 If set to 1, list the names of the available voices and exit
5371 immediately. Default value is 0.
5374 Set the maximum number of samples per frame. Default value is 512.
5377 Set the filename containing the text to speak.
5380 Set the text to speak.
5383 Set the voice to use for the speech synthesis. Default value is
5384 @code{kal}. See also the @var{list_voices} option.
5387 @subsection Examples
5391 Read from file @file{speech.txt}, and synthesize the text using the
5392 standard flite voice:
5394 flite=textfile=speech.txt
5398 Read the specified text selecting the @code{slt} voice:
5400 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5404 Input text to ffmpeg:
5406 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5410 Make @file{ffplay} speak the specified text, using @code{flite} and
5411 the @code{lavfi} device:
5413 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5417 For more information about libflite, check:
5418 @url{http://www.festvox.org/flite/}
5422 Generate a noise audio signal.
5424 The filter accepts the following options:
5427 @item sample_rate, r
5428 Specify the sample rate. Default value is 48000 Hz.
5431 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5435 Specify the duration of the generated audio stream. Not specifying this option
5436 results in noise with an infinite length.
5438 @item color, colour, c
5439 Specify the color of noise. Available noise colors are white, pink, brown,
5440 blue and violet. Default color is white.
5443 Specify a value used to seed the PRNG.
5446 Set the number of samples per each output frame, default is 1024.
5449 @subsection Examples
5454 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5456 anoisesrc=d=60:c=pink:r=44100:a=0.5
5462 Generate odd-tap Hilbert transform FIR coefficients.
5464 The resulting stream can be used with @ref{afir} filter for phase-shifting
5465 the signal by 90 degrees.
5467 This is used in many matrix coding schemes and for analytic signal generation.
5468 The process is often written as a multiplication by i (or j), the imaginary unit.
5470 The filter accepts the following options:
5474 @item sample_rate, s
5475 Set sample rate, default is 44100.
5478 Set length of FIR filter, default is 22051.
5481 Set number of samples per each frame.
5484 Set window function to be used when generating FIR coefficients.
5489 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5491 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5493 The filter accepts the following options:
5496 @item sample_rate, r
5497 Set sample rate, default is 44100.
5500 Set number of samples per each frame. Default is 1024.
5503 Set high-pass frequency. Default is 0.
5506 Set low-pass frequency. Default is 0.
5507 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5508 is higher than 0 then filter will create band-pass filter coefficients,
5509 otherwise band-reject filter coefficients.
5512 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5515 Set Kaiser window beta.
5518 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5521 Enable rounding, by default is disabled.
5524 Set number of taps for high-pass filter.
5527 Set number of taps for low-pass filter.
5532 Generate an audio signal made of a sine wave with amplitude 1/8.
5534 The audio signal is bit-exact.
5536 The filter accepts the following options:
5541 Set the carrier frequency. Default is 440 Hz.
5543 @item beep_factor, b
5544 Enable a periodic beep every second with frequency @var{beep_factor} times
5545 the carrier frequency. Default is 0, meaning the beep is disabled.
5547 @item sample_rate, r
5548 Specify the sample rate, default is 44100.
5551 Specify the duration of the generated audio stream.
5553 @item samples_per_frame
5554 Set the number of samples per output frame.
5556 The expression can contain the following constants:
5560 The (sequential) number of the output audio frame, starting from 0.
5563 The PTS (Presentation TimeStamp) of the output audio frame,
5564 expressed in @var{TB} units.
5567 The PTS of the output audio frame, expressed in seconds.
5570 The timebase of the output audio frames.
5573 Default is @code{1024}.
5576 @subsection Examples
5581 Generate a simple 440 Hz sine wave:
5587 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5591 sine=frequency=220:beep_factor=4:duration=5
5595 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5598 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5602 @c man end AUDIO SOURCES
5604 @chapter Audio Sinks
5605 @c man begin AUDIO SINKS
5607 Below is a description of the currently available audio sinks.
5609 @section abuffersink
5611 Buffer audio frames, and make them available to the end of filter chain.
5613 This sink is mainly intended for programmatic use, in particular
5614 through the interface defined in @file{libavfilter/buffersink.h}
5615 or the options system.
5617 It accepts a pointer to an AVABufferSinkContext structure, which
5618 defines the incoming buffers' formats, to be passed as the opaque
5619 parameter to @code{avfilter_init_filter} for initialization.
5622 Null audio sink; do absolutely nothing with the input audio. It is
5623 mainly useful as a template and for use in analysis / debugging
5626 @c man end AUDIO SINKS
5628 @chapter Video Filters
5629 @c man begin VIDEO FILTERS
5631 When you configure your FFmpeg build, you can disable any of the
5632 existing filters using @code{--disable-filters}.
5633 The configure output will show the video filters included in your
5636 Below is a description of the currently available video filters.
5638 @section alphaextract
5640 Extract the alpha component from the input as a grayscale video. This
5641 is especially useful with the @var{alphamerge} filter.
5645 Add or replace the alpha component of the primary input with the
5646 grayscale value of a second input. This is intended for use with
5647 @var{alphaextract} to allow the transmission or storage of frame
5648 sequences that have alpha in a format that doesn't support an alpha
5651 For example, to reconstruct full frames from a normal YUV-encoded video
5652 and a separate video created with @var{alphaextract}, you might use:
5654 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5657 Since this filter is designed for reconstruction, it operates on frame
5658 sequences without considering timestamps, and terminates when either
5659 input reaches end of stream. This will cause problems if your encoding
5660 pipeline drops frames. If you're trying to apply an image as an
5661 overlay to a video stream, consider the @var{overlay} filter instead.
5665 Amplify differences between current pixel and pixels of adjacent frames in
5666 same pixel location.
5668 This filter accepts the following options:
5672 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5673 For example radius of 3 will instruct filter to calculate average of 7 frames.
5676 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5679 Set threshold for difference amplification. Any difference greater or equal to
5680 this value will not alter source pixel. Default is 10.
5681 Allowed range is from 0 to 65535.
5684 Set tolerance for difference amplification. Any difference lower to
5685 this value will not alter source pixel. Default is 0.
5686 Allowed range is from 0 to 65535.
5689 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5690 This option controls maximum possible value that will decrease source pixel value.
5693 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5694 This option controls maximum possible value that will increase source pixel value.
5697 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5702 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5703 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5704 Substation Alpha) subtitles files.
5706 This filter accepts the following option in addition to the common options from
5707 the @ref{subtitles} filter:
5711 Set the shaping engine
5713 Available values are:
5716 The default libass shaping engine, which is the best available.
5718 Fast, font-agnostic shaper that can do only substitutions
5720 Slower shaper using OpenType for substitutions and positioning
5723 The default is @code{auto}.
5727 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5729 The filter accepts the following options:
5733 Set threshold A for 1st plane. Default is 0.02.
5734 Valid range is 0 to 0.3.
5737 Set threshold B for 1st plane. Default is 0.04.
5738 Valid range is 0 to 5.
5741 Set threshold A for 2nd plane. Default is 0.02.
5742 Valid range is 0 to 0.3.
5745 Set threshold B for 2nd plane. Default is 0.04.
5746 Valid range is 0 to 5.
5749 Set threshold A for 3rd plane. Default is 0.02.
5750 Valid range is 0 to 0.3.
5753 Set threshold B for 3rd plane. Default is 0.04.
5754 Valid range is 0 to 5.
5756 Threshold A is designed to react on abrupt changes in the input signal and
5757 threshold B is designed to react on continuous changes in the input signal.
5760 Set number of frames filter will use for averaging. Default is 9. Must be odd
5761 number in range [5, 129].
5764 Set what planes of frame filter will use for averaging. Default is all.
5769 Apply average blur filter.
5771 The filter accepts the following options:
5775 Set horizontal radius size.
5778 Set which planes to filter. By default all planes are filtered.
5781 Set vertical radius size, if zero it will be same as @code{sizeX}.
5782 Default is @code{0}.
5787 Compute the bounding box for the non-black pixels in the input frame
5790 This filter computes the bounding box containing all the pixels with a
5791 luminance value greater than the minimum allowed value.
5792 The parameters describing the bounding box are printed on the filter
5795 The filter accepts the following option:
5799 Set the minimal luminance value. Default is @code{16}.
5802 @section bitplanenoise
5804 Show and measure bit plane noise.
5806 The filter accepts the following options:
5810 Set which plane to analyze. Default is @code{1}.
5813 Filter out noisy pixels from @code{bitplane} set above.
5814 Default is disabled.
5817 @section blackdetect
5819 Detect video intervals that are (almost) completely black. Can be
5820 useful to detect chapter transitions, commercials, or invalid
5821 recordings. Output lines contains the time for the start, end and
5822 duration of the detected black interval expressed in seconds.
5824 In order to display the output lines, you need to set the loglevel at
5825 least to the AV_LOG_INFO value.
5827 The filter accepts the following options:
5830 @item black_min_duration, d
5831 Set the minimum detected black duration expressed in seconds. It must
5832 be a non-negative floating point number.
5834 Default value is 2.0.
5836 @item picture_black_ratio_th, pic_th
5837 Set the threshold for considering a picture "black".
5838 Express the minimum value for the ratio:
5840 @var{nb_black_pixels} / @var{nb_pixels}
5843 for which a picture is considered black.
5844 Default value is 0.98.
5846 @item pixel_black_th, pix_th
5847 Set the threshold for considering a pixel "black".
5849 The threshold expresses the maximum pixel luminance value for which a
5850 pixel is considered "black". The provided value is scaled according to
5851 the following equation:
5853 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5856 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5857 the input video format, the range is [0-255] for YUV full-range
5858 formats and [16-235] for YUV non full-range formats.
5860 Default value is 0.10.
5863 The following example sets the maximum pixel threshold to the minimum
5864 value, and detects only black intervals of 2 or more seconds:
5866 blackdetect=d=2:pix_th=0.00
5871 Detect frames that are (almost) completely black. Can be useful to
5872 detect chapter transitions or commercials. Output lines consist of
5873 the frame number of the detected frame, the percentage of blackness,
5874 the position in the file if known or -1 and the timestamp in seconds.
5876 In order to display the output lines, you need to set the loglevel at
5877 least to the AV_LOG_INFO value.
5879 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5880 The value represents the percentage of pixels in the picture that
5881 are below the threshold value.
5883 It accepts the following parameters:
5888 The percentage of the pixels that have to be below the threshold; it defaults to
5891 @item threshold, thresh
5892 The threshold below which a pixel value is considered black; it defaults to
5897 @section blend, tblend
5899 Blend two video frames into each other.
5901 The @code{blend} filter takes two input streams and outputs one
5902 stream, the first input is the "top" layer and second input is
5903 "bottom" layer. By default, the output terminates when the longest input terminates.
5905 The @code{tblend} (time blend) filter takes two consecutive frames
5906 from one single stream, and outputs the result obtained by blending
5907 the new frame on top of the old frame.
5909 A description of the accepted options follows.
5917 Set blend mode for specific pixel component or all pixel components in case
5918 of @var{all_mode}. Default value is @code{normal}.
5920 Available values for component modes are:
5962 Set blend opacity for specific pixel component or all pixel components in case
5963 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5970 Set blend expression for specific pixel component or all pixel components in case
5971 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5973 The expressions can use the following variables:
5977 The sequential number of the filtered frame, starting from @code{0}.
5981 the coordinates of the current sample
5985 the width and height of currently filtered plane
5989 Width and height scale for the plane being filtered. It is the
5990 ratio between the dimensions of the current plane to the luma plane,
5991 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
5992 the luma plane and @code{0.5,0.5} for the chroma planes.
5995 Time of the current frame, expressed in seconds.
5998 Value of pixel component at current location for first video frame (top layer).
6001 Value of pixel component at current location for second video frame (bottom layer).
6005 The @code{blend} filter also supports the @ref{framesync} options.
6007 @subsection Examples
6011 Apply transition from bottom layer to top layer in first 10 seconds:
6013 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6017 Apply linear horizontal transition from top layer to bottom layer:
6019 blend=all_expr='A*(X/W)+B*(1-X/W)'
6023 Apply 1x1 checkerboard effect:
6025 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6029 Apply uncover left effect:
6031 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6035 Apply uncover down effect:
6037 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6041 Apply uncover up-left effect:
6043 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6047 Split diagonally video and shows top and bottom layer on each side:
6049 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6053 Display differences between the current and the previous frame:
6055 tblend=all_mode=grainextract
6061 Denoise frames using Block-Matching 3D algorithm.
6063 The filter accepts the following options.
6067 Set denoising strength. Default value is 1.
6068 Allowed range is from 0 to 999.9.
6069 The denoising algorithm is very sensitive to sigma, so adjust it
6070 according to the source.
6073 Set local patch size. This sets dimensions in 2D.
6076 Set sliding step for processing blocks. Default value is 4.
6077 Allowed range is from 1 to 64.
6078 Smaller values allows processing more reference blocks and is slower.
6081 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6082 When set to 1, no block matching is done. Larger values allows more blocks
6084 Allowed range is from 1 to 256.
6087 Set radius for search block matching. Default is 9.
6088 Allowed range is from 1 to INT32_MAX.
6091 Set step between two search locations for block matching. Default is 1.
6092 Allowed range is from 1 to 64. Smaller is slower.
6095 Set threshold of mean square error for block matching. Valid range is 0 to
6099 Set thresholding parameter for hard thresholding in 3D transformed domain.
6100 Larger values results in stronger hard-thresholding filtering in frequency
6104 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6105 Default is @code{basic}.
6108 If enabled, filter will use 2nd stream for block matching.
6109 Default is disabled for @code{basic} value of @var{estim} option,
6110 and always enabled if value of @var{estim} is @code{final}.
6113 Set planes to filter. Default is all available except alpha.
6116 @subsection Examples
6120 Basic filtering with bm3d:
6122 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6126 Same as above, but filtering only luma:
6128 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6132 Same as above, but with both estimation modes:
6134 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
6138 Same as above, but prefilter with @ref{nlmeans} filter instead:
6140 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
6146 Apply a boxblur algorithm to the input video.
6148 It accepts the following parameters:
6152 @item luma_radius, lr
6153 @item luma_power, lp
6154 @item chroma_radius, cr
6155 @item chroma_power, cp
6156 @item alpha_radius, ar
6157 @item alpha_power, ap
6161 A description of the accepted options follows.
6164 @item luma_radius, lr
6165 @item chroma_radius, cr
6166 @item alpha_radius, ar
6167 Set an expression for the box radius in pixels used for blurring the
6168 corresponding input plane.
6170 The radius value must be a non-negative number, and must not be
6171 greater than the value of the expression @code{min(w,h)/2} for the
6172 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6175 Default value for @option{luma_radius} is "2". If not specified,
6176 @option{chroma_radius} and @option{alpha_radius} default to the
6177 corresponding value set for @option{luma_radius}.
6179 The expressions can contain the following constants:
6183 The input width and height in pixels.
6187 The input chroma image width and height in pixels.
6191 The horizontal and vertical chroma subsample values. For example, for the
6192 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6195 @item luma_power, lp
6196 @item chroma_power, cp
6197 @item alpha_power, ap
6198 Specify how many times the boxblur filter is applied to the
6199 corresponding plane.
6201 Default value for @option{luma_power} is 2. If not specified,
6202 @option{chroma_power} and @option{alpha_power} default to the
6203 corresponding value set for @option{luma_power}.
6205 A value of 0 will disable the effect.
6208 @subsection Examples
6212 Apply a boxblur filter with the luma, chroma, and alpha radii
6215 boxblur=luma_radius=2:luma_power=1
6220 Set the luma radius to 2, and alpha and chroma radius to 0:
6222 boxblur=2:1:cr=0:ar=0
6226 Set the luma and chroma radii to a fraction of the video dimension:
6228 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6234 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6235 Deinterlacing Filter").
6237 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6238 interpolation algorithms.
6239 It accepts the following parameters:
6243 The interlacing mode to adopt. It accepts one of the following values:
6247 Output one frame for each frame.
6249 Output one frame for each field.
6252 The default value is @code{send_field}.
6255 The picture field parity assumed for the input interlaced video. It accepts one
6256 of the following values:
6260 Assume the top field is first.
6262 Assume the bottom field is first.
6264 Enable automatic detection of field parity.
6267 The default value is @code{auto}.
6268 If the interlacing is unknown or the decoder does not export this information,
6269 top field first will be assumed.
6272 Specify which frames to deinterlace. Accept one of the following
6277 Deinterlace all frames.
6279 Only deinterlace frames marked as interlaced.
6282 The default value is @code{all}.
6286 Remove all color information for all colors except for certain one.
6288 The filter accepts the following options:
6292 The color which will not be replaced with neutral chroma.
6295 Similarity percentage with the above color.
6296 0.01 matches only the exact key color, while 1.0 matches everything.
6299 Signals that the color passed is already in YUV instead of RGB.
6301 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6302 This can be used to pass exact YUV values as hexadecimal numbers.
6306 YUV colorspace color/chroma keying.
6308 The filter accepts the following options:
6312 The color which will be replaced with transparency.
6315 Similarity percentage with the key color.
6317 0.01 matches only the exact key color, while 1.0 matches everything.
6322 0.0 makes pixels either fully transparent, or not transparent at all.
6324 Higher values result in semi-transparent pixels, with a higher transparency
6325 the more similar the pixels color is to the key color.
6328 Signals that the color passed is already in YUV instead of RGB.
6330 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6331 This can be used to pass exact YUV values as hexadecimal numbers.
6334 @subsection Examples
6338 Make every green pixel in the input image transparent:
6340 ffmpeg -i input.png -vf chromakey=green out.png
6344 Overlay a greenscreen-video on top of a static black background.
6346 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
6350 @section chromashift
6351 Shift chroma pixels horizontally and/or vertically.
6353 The filter accepts the following options:
6356 Set amount to shift chroma-blue horizontally.
6358 Set amount to shift chroma-blue vertically.
6360 Set amount to shift chroma-red horizontally.
6362 Set amount to shift chroma-red vertically.
6364 Set edge mode, can be @var{smear}, default, or @var{warp}.
6369 Display CIE color diagram with pixels overlaid onto it.
6371 The filter accepts the following options:
6386 @item uhdtv, rec2020
6399 Set what gamuts to draw.
6401 See @code{system} option for available values.
6404 Set ciescope size, by default set to 512.
6407 Set intensity used to map input pixel values to CIE diagram.
6410 Set contrast used to draw tongue colors that are out of active color system gamut.
6413 Correct gamma displayed on scope, by default enabled.
6416 Show white point on CIE diagram, by default disabled.
6419 Set input gamma. Used only with XYZ input color space.
6424 Visualize information exported by some codecs.
6426 Some codecs can export information through frames using side-data or other
6427 means. For example, some MPEG based codecs export motion vectors through the
6428 @var{export_mvs} flag in the codec @option{flags2} option.
6430 The filter accepts the following option:
6434 Set motion vectors to visualize.
6436 Available flags for @var{mv} are:
6440 forward predicted MVs of P-frames
6442 forward predicted MVs of B-frames
6444 backward predicted MVs of B-frames
6448 Display quantization parameters using the chroma planes.
6451 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6453 Available flags for @var{mv_type} are:
6457 forward predicted MVs
6459 backward predicted MVs
6462 @item frame_type, ft
6463 Set frame type to visualize motion vectors of.
6465 Available flags for @var{frame_type} are:
6469 intra-coded frames (I-frames)
6471 predicted frames (P-frames)
6473 bi-directionally predicted frames (B-frames)
6477 @subsection Examples
6481 Visualize forward predicted MVs of all frames using @command{ffplay}:
6483 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6487 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6489 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6493 @section colorbalance
6494 Modify intensity of primary colors (red, green and blue) of input frames.
6496 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6497 regions for the red-cyan, green-magenta or blue-yellow balance.
6499 A positive adjustment value shifts the balance towards the primary color, a negative
6500 value towards the complementary color.
6502 The filter accepts the following options:
6508 Adjust red, green and blue shadows (darkest pixels).
6513 Adjust red, green and blue midtones (medium pixels).
6518 Adjust red, green and blue highlights (brightest pixels).
6520 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6523 @subsection Examples
6527 Add red color cast to shadows:
6534 RGB colorspace color keying.
6536 The filter accepts the following options:
6540 The color which will be replaced with transparency.
6543 Similarity percentage with the key color.
6545 0.01 matches only the exact key color, while 1.0 matches everything.
6550 0.0 makes pixels either fully transparent, or not transparent at all.
6552 Higher values result in semi-transparent pixels, with a higher transparency
6553 the more similar the pixels color is to the key color.
6556 @subsection Examples
6560 Make every green pixel in the input image transparent:
6562 ffmpeg -i input.png -vf colorkey=green out.png
6566 Overlay a greenscreen-video on top of a static background image.
6568 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
6572 @section colorlevels
6574 Adjust video input frames using levels.
6576 The filter accepts the following options:
6583 Adjust red, green, blue and alpha input black point.
6584 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6590 Adjust red, green, blue and alpha input white point.
6591 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6593 Input levels are used to lighten highlights (bright tones), darken shadows
6594 (dark tones), change the balance of bright and dark tones.
6600 Adjust red, green, blue and alpha output black point.
6601 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6607 Adjust red, green, blue and alpha output white point.
6608 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6610 Output levels allows manual selection of a constrained output level range.
6613 @subsection Examples
6617 Make video output darker:
6619 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6625 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6629 Make video output lighter:
6631 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6635 Increase brightness:
6637 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6641 @section colorchannelmixer
6643 Adjust video input frames by re-mixing color channels.
6645 This filter modifies a color channel by adding the values associated to
6646 the other channels of the same pixels. For example if the value to
6647 modify is red, the output value will be:
6649 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6652 The filter accepts the following options:
6659 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6660 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6666 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6667 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6673 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6674 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6680 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6681 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6683 Allowed ranges for options are @code{[-2.0, 2.0]}.
6686 @subsection Examples
6690 Convert source to grayscale:
6692 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6695 Simulate sepia tones:
6697 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6701 @section colormatrix
6703 Convert color matrix.
6705 The filter accepts the following options:
6710 Specify the source and destination color matrix. Both values must be
6713 The accepted values are:
6741 For example to convert from BT.601 to SMPTE-240M, use the command:
6743 colormatrix=bt601:smpte240m
6748 Convert colorspace, transfer characteristics or color primaries.
6749 Input video needs to have an even size.
6751 The filter accepts the following options:
6756 Specify all color properties at once.
6758 The accepted values are:
6788 Specify output colorspace.
6790 The accepted values are:
6799 BT.470BG or BT.601-6 625
6802 SMPTE-170M or BT.601-6 525
6811 BT.2020 with non-constant luminance
6817 Specify output transfer characteristics.
6819 The accepted values are:
6831 Constant gamma of 2.2
6834 Constant gamma of 2.8
6837 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6855 BT.2020 for 10-bits content
6858 BT.2020 for 12-bits content
6864 Specify output color primaries.
6866 The accepted values are:
6875 BT.470BG or BT.601-6 625
6878 SMPTE-170M or BT.601-6 525
6902 Specify output color range.
6904 The accepted values are:
6907 TV (restricted) range
6910 MPEG (restricted) range
6921 Specify output color format.
6923 The accepted values are:
6926 YUV 4:2:0 planar 8-bits
6929 YUV 4:2:0 planar 10-bits
6932 YUV 4:2:0 planar 12-bits
6935 YUV 4:2:2 planar 8-bits
6938 YUV 4:2:2 planar 10-bits
6941 YUV 4:2:2 planar 12-bits
6944 YUV 4:4:4 planar 8-bits
6947 YUV 4:4:4 planar 10-bits
6950 YUV 4:4:4 planar 12-bits
6955 Do a fast conversion, which skips gamma/primary correction. This will take
6956 significantly less CPU, but will be mathematically incorrect. To get output
6957 compatible with that produced by the colormatrix filter, use fast=1.
6960 Specify dithering mode.
6962 The accepted values are:
6968 Floyd-Steinberg dithering
6972 Whitepoint adaptation mode.
6974 The accepted values are:
6977 Bradford whitepoint adaptation
6980 von Kries whitepoint adaptation
6983 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6987 Override all input properties at once. Same accepted values as @ref{all}.
6990 Override input colorspace. Same accepted values as @ref{space}.
6993 Override input color primaries. Same accepted values as @ref{primaries}.
6996 Override input transfer characteristics. Same accepted values as @ref{trc}.
6999 Override input color range. Same accepted values as @ref{range}.
7003 The filter converts the transfer characteristics, color space and color
7004 primaries to the specified user values. The output value, if not specified,
7005 is set to a default value based on the "all" property. If that property is
7006 also not specified, the filter will log an error. The output color range and
7007 format default to the same value as the input color range and format. The
7008 input transfer characteristics, color space, color primaries and color range
7009 should be set on the input data. If any of these are missing, the filter will
7010 log an error and no conversion will take place.
7012 For example to convert the input to SMPTE-240M, use the command:
7014 colorspace=smpte240m
7017 @section convolution
7019 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7021 The filter accepts the following options:
7028 Set matrix for each plane.
7029 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7030 and from 1 to 49 odd number of signed integers in @var{row} mode.
7036 Set multiplier for calculated value for each plane.
7037 If unset or 0, it will be sum of all matrix elements.
7043 Set bias for each plane. This value is added to the result of the multiplication.
7044 Useful for making the overall image brighter or darker. Default is 0.0.
7050 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7051 Default is @var{square}.
7054 @subsection Examples
7060 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"
7066 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"
7072 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"
7078 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"
7082 Apply laplacian edge detector which includes diagonals:
7084 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"
7090 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"
7096 Apply 2D convolution of video stream in frequency domain using second stream
7099 The filter accepts the following options:
7103 Set which planes to process.
7106 Set which impulse video frames will be processed, can be @var{first}
7107 or @var{all}. Default is @var{all}.
7110 The @code{convolve} filter also supports the @ref{framesync} options.
7114 Copy the input video source unchanged to the output. This is mainly useful for
7119 Video filtering on GPU using Apple's CoreImage API on OSX.
7121 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7122 processed by video hardware. However, software-based OpenGL implementations
7123 exist which means there is no guarantee for hardware processing. It depends on
7126 There are many filters and image generators provided by Apple that come with a
7127 large variety of options. The filter has to be referenced by its name along
7130 The coreimage filter accepts the following options:
7133 List all available filters and generators along with all their respective
7134 options as well as possible minimum and maximum values along with the default
7141 Specify all filters by their respective name and options.
7142 Use @var{list_filters} to determine all valid filter names and options.
7143 Numerical options are specified by a float value and are automatically clamped
7144 to their respective value range. Vector and color options have to be specified
7145 by a list of space separated float values. Character escaping has to be done.
7146 A special option name @code{default} is available to use default options for a
7149 It is required to specify either @code{default} or at least one of the filter options.
7150 All omitted options are used with their default values.
7151 The syntax of the filter string is as follows:
7153 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7157 Specify a rectangle where the output of the filter chain is copied into the
7158 input image. It is given by a list of space separated float values:
7160 output_rect=x\ y\ width\ height
7162 If not given, the output rectangle equals the dimensions of the input image.
7163 The output rectangle is automatically cropped at the borders of the input
7164 image. Negative values are valid for each component.
7166 output_rect=25\ 25\ 100\ 100
7170 Several filters can be chained for successive processing without GPU-HOST
7171 transfers allowing for fast processing of complex filter chains.
7172 Currently, only filters with zero (generators) or exactly one (filters) input
7173 image and one output image are supported. Also, transition filters are not yet
7176 Some filters generate output images with additional padding depending on the
7177 respective filter kernel. The padding is automatically removed to ensure the
7178 filter output has the same size as the input image.
7180 For image generators, the size of the output image is determined by the
7181 previous output image of the filter chain or the input image of the whole
7182 filterchain, respectively. The generators do not use the pixel information of
7183 this image to generate their output. However, the generated output is
7184 blended onto this image, resulting in partial or complete coverage of the
7187 The @ref{coreimagesrc} video source can be used for generating input images
7188 which are directly fed into the filter chain. By using it, providing input
7189 images by another video source or an input video is not required.
7191 @subsection Examples
7196 List all filters available:
7198 coreimage=list_filters=true
7202 Use the CIBoxBlur filter with default options to blur an image:
7204 coreimage=filter=CIBoxBlur@@default
7208 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7209 its center at 100x100 and a radius of 50 pixels:
7211 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7215 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7216 given as complete and escaped command-line for Apple's standard bash shell:
7218 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7224 Crop the input video to given dimensions.
7226 It accepts the following parameters:
7230 The width of the output video. It defaults to @code{iw}.
7231 This expression is evaluated only once during the filter
7232 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7235 The height of the output video. It defaults to @code{ih}.
7236 This expression is evaluated only once during the filter
7237 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7240 The horizontal position, in the input video, of the left edge of the output
7241 video. It defaults to @code{(in_w-out_w)/2}.
7242 This expression is evaluated per-frame.
7245 The vertical position, in the input video, of the top edge of the output video.
7246 It defaults to @code{(in_h-out_h)/2}.
7247 This expression is evaluated per-frame.
7250 If set to 1 will force the output display aspect ratio
7251 to be the same of the input, by changing the output sample aspect
7252 ratio. It defaults to 0.
7255 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7256 width/height/x/y as specified and will not be rounded to nearest smaller value.
7260 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7261 expressions containing the following constants:
7266 The computed values for @var{x} and @var{y}. They are evaluated for
7271 The input width and height.
7275 These are the same as @var{in_w} and @var{in_h}.
7279 The output (cropped) width and height.
7283 These are the same as @var{out_w} and @var{out_h}.
7286 same as @var{iw} / @var{ih}
7289 input sample aspect ratio
7292 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7296 horizontal and vertical chroma subsample values. For example for the
7297 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7300 The number of the input frame, starting from 0.
7303 the position in the file of the input frame, NAN if unknown
7306 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7310 The expression for @var{out_w} may depend on the value of @var{out_h},
7311 and the expression for @var{out_h} may depend on @var{out_w}, but they
7312 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7313 evaluated after @var{out_w} and @var{out_h}.
7315 The @var{x} and @var{y} parameters specify the expressions for the
7316 position of the top-left corner of the output (non-cropped) area. They
7317 are evaluated for each frame. If the evaluated value is not valid, it
7318 is approximated to the nearest valid value.
7320 The expression for @var{x} may depend on @var{y}, and the expression
7321 for @var{y} may depend on @var{x}.
7323 @subsection Examples
7327 Crop area with size 100x100 at position (12,34).
7332 Using named options, the example above becomes:
7334 crop=w=100:h=100:x=12:y=34
7338 Crop the central input area with size 100x100:
7344 Crop the central input area with size 2/3 of the input video:
7346 crop=2/3*in_w:2/3*in_h
7350 Crop the input video central square:
7357 Delimit the rectangle with the top-left corner placed at position
7358 100:100 and the right-bottom corner corresponding to the right-bottom
7359 corner of the input image.
7361 crop=in_w-100:in_h-100:100:100
7365 Crop 10 pixels from the left and right borders, and 20 pixels from
7366 the top and bottom borders
7368 crop=in_w-2*10:in_h-2*20
7372 Keep only the bottom right quarter of the input image:
7374 crop=in_w/2:in_h/2:in_w/2:in_h/2
7378 Crop height for getting Greek harmony:
7380 crop=in_w:1/PHI*in_w
7384 Apply trembling effect:
7386 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)
7390 Apply erratic camera effect depending on timestamp:
7392 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)"
7396 Set x depending on the value of y:
7398 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7402 @subsection Commands
7404 This filter supports the following commands:
7410 Set width/height of the output video and the horizontal/vertical position
7412 The command accepts the same syntax of the corresponding option.
7414 If the specified expression is not valid, it is kept at its current
7420 Auto-detect the crop size.
7422 It calculates the necessary cropping parameters and prints the
7423 recommended parameters via the logging system. The detected dimensions
7424 correspond to the non-black area of the input video.
7426 It accepts the following parameters:
7431 Set higher black value threshold, which can be optionally specified
7432 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7433 value greater to the set value is considered non-black. It defaults to 24.
7434 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7435 on the bitdepth of the pixel format.
7438 The value which the width/height should be divisible by. It defaults to
7439 16. The offset is automatically adjusted to center the video. Use 2 to
7440 get only even dimensions (needed for 4:2:2 video). 16 is best when
7441 encoding to most video codecs.
7443 @item reset_count, reset
7444 Set the counter that determines after how many frames cropdetect will
7445 reset the previously detected largest video area and start over to
7446 detect the current optimal crop area. Default value is 0.
7448 This can be useful when channel logos distort the video area. 0
7449 indicates 'never reset', and returns the largest area encountered during
7456 Delay video filtering until a given wallclock timestamp. The filter first
7457 passes on @option{preroll} amount of frames, then it buffers at most
7458 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7459 it forwards the buffered frames and also any subsequent frames coming in its
7462 The filter can be used synchronize the output of multiple ffmpeg processes for
7463 realtime output devices like decklink. By putting the delay in the filtering
7464 chain and pre-buffering frames the process can pass on data to output almost
7465 immediately after the target wallclock timestamp is reached.
7467 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7473 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7476 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7479 The maximum duration of content to buffer before waiting for the cue expressed
7480 in seconds. Default is 0.
7487 Apply color adjustments using curves.
7489 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7490 component (red, green and blue) has its values defined by @var{N} key points
7491 tied from each other using a smooth curve. The x-axis represents the pixel
7492 values from the input frame, and the y-axis the new pixel values to be set for
7495 By default, a component curve is defined by the two points @var{(0;0)} and
7496 @var{(1;1)}. This creates a straight line where each original pixel value is
7497 "adjusted" to its own value, which means no change to the image.
7499 The filter allows you to redefine these two points and add some more. A new
7500 curve (using a natural cubic spline interpolation) will be define to pass
7501 smoothly through all these new coordinates. The new defined points needs to be
7502 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7503 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7504 the vector spaces, the values will be clipped accordingly.
7506 The filter accepts the following options:
7510 Select one of the available color presets. This option can be used in addition
7511 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7512 options takes priority on the preset values.
7513 Available presets are:
7516 @item color_negative
7519 @item increase_contrast
7521 @item linear_contrast
7522 @item medium_contrast
7524 @item strong_contrast
7527 Default is @code{none}.
7529 Set the master key points. These points will define a second pass mapping. It
7530 is sometimes called a "luminance" or "value" mapping. It can be used with
7531 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7532 post-processing LUT.
7534 Set the key points for the red component.
7536 Set the key points for the green component.
7538 Set the key points for the blue component.
7540 Set the key points for all components (not including master).
7541 Can be used in addition to the other key points component
7542 options. In this case, the unset component(s) will fallback on this
7543 @option{all} setting.
7545 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7547 Save Gnuplot script of the curves in specified file.
7550 To avoid some filtergraph syntax conflicts, each key points list need to be
7551 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7553 @subsection Examples
7557 Increase slightly the middle level of blue:
7559 curves=blue='0/0 0.5/0.58 1/1'
7565 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'
7567 Here we obtain the following coordinates for each components:
7570 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7572 @code{(0;0) (0.50;0.48) (1;1)}
7574 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7578 The previous example can also be achieved with the associated built-in preset:
7580 curves=preset=vintage
7590 Use a Photoshop preset and redefine the points of the green component:
7592 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7596 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7597 and @command{gnuplot}:
7599 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7600 gnuplot -p /tmp/curves.plt
7606 Video data analysis filter.
7608 This filter shows hexadecimal pixel values of part of video.
7610 The filter accepts the following options:
7614 Set output video size.
7617 Set x offset from where to pick pixels.
7620 Set y offset from where to pick pixels.
7623 Set scope mode, can be one of the following:
7626 Draw hexadecimal pixel values with white color on black background.
7629 Draw hexadecimal pixel values with input video pixel color on black
7633 Draw hexadecimal pixel values on color background picked from input video,
7634 the text color is picked in such way so its always visible.
7638 Draw rows and columns numbers on left and top of video.
7641 Set background opacity.
7646 Denoise frames using 2D DCT (frequency domain filtering).
7648 This filter is not designed for real time.
7650 The filter accepts the following options:
7654 Set the noise sigma constant.
7656 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7657 coefficient (absolute value) below this threshold with be dropped.
7659 If you need a more advanced filtering, see @option{expr}.
7661 Default is @code{0}.
7664 Set number overlapping pixels for each block. Since the filter can be slow, you
7665 may want to reduce this value, at the cost of a less effective filter and the
7666 risk of various artefacts.
7668 If the overlapping value doesn't permit processing the whole input width or
7669 height, a warning will be displayed and according borders won't be denoised.
7671 Default value is @var{blocksize}-1, which is the best possible setting.
7674 Set the coefficient factor expression.
7676 For each coefficient of a DCT block, this expression will be evaluated as a
7677 multiplier value for the coefficient.
7679 If this is option is set, the @option{sigma} option will be ignored.
7681 The absolute value of the coefficient can be accessed through the @var{c}
7685 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7686 @var{blocksize}, which is the width and height of the processed blocks.
7688 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7689 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7690 on the speed processing. Also, a larger block size does not necessarily means a
7694 @subsection Examples
7696 Apply a denoise with a @option{sigma} of @code{4.5}:
7701 The same operation can be achieved using the expression system:
7703 dctdnoiz=e='gte(c, 4.5*3)'
7706 Violent denoise using a block size of @code{16x16}:
7713 Remove banding artifacts from input video.
7714 It works by replacing banded pixels with average value of referenced pixels.
7716 The filter accepts the following options:
7723 Set banding detection threshold for each plane. Default is 0.02.
7724 Valid range is 0.00003 to 0.5.
7725 If difference between current pixel and reference pixel is less than threshold,
7726 it will be considered as banded.
7729 Banding detection range in pixels. Default is 16. If positive, random number
7730 in range 0 to set value will be used. If negative, exact absolute value
7732 The range defines square of four pixels around current pixel.
7735 Set direction in radians from which four pixel will be compared. If positive,
7736 random direction from 0 to set direction will be picked. If negative, exact of
7737 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7738 will pick only pixels on same row and -PI/2 will pick only pixels on same
7742 If enabled, current pixel is compared with average value of all four
7743 surrounding pixels. The default is enabled. If disabled current pixel is
7744 compared with all four surrounding pixels. The pixel is considered banded
7745 if only all four differences with surrounding pixels are less than threshold.
7748 If enabled, current pixel is changed if and only if all pixel components are banded,
7749 e.g. banding detection threshold is triggered for all color components.
7750 The default is disabled.
7755 Remove blocking artifacts from input video.
7757 The filter accepts the following options:
7761 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7762 This controls what kind of deblocking is applied.
7765 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7771 Set blocking detection thresholds. Allowed range is 0 to 1.
7772 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7773 Using higher threshold gives more deblocking strength.
7774 Setting @var{alpha} controls threshold detection at exact edge of block.
7775 Remaining options controls threshold detection near the edge. Each one for
7776 below/above or left/right. Setting any of those to @var{0} disables
7780 Set planes to filter. Default is to filter all available planes.
7783 @subsection Examples
7787 Deblock using weak filter and block size of 4 pixels.
7789 deblock=filter=weak:block=4
7793 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7794 deblocking more edges.
7796 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7800 Similar as above, but filter only first plane.
7802 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7806 Similar as above, but filter only second and third plane.
7808 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7815 Drop duplicated frames at regular intervals.
7817 The filter accepts the following options:
7821 Set the number of frames from which one will be dropped. Setting this to
7822 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7823 Default is @code{5}.
7826 Set the threshold for duplicate detection. If the difference metric for a frame
7827 is less than or equal to this value, then it is declared as duplicate. Default
7831 Set scene change threshold. Default is @code{15}.
7835 Set the size of the x and y-axis blocks used during metric calculations.
7836 Larger blocks give better noise suppression, but also give worse detection of
7837 small movements. Must be a power of two. Default is @code{32}.
7840 Mark main input as a pre-processed input and activate clean source input
7841 stream. This allows the input to be pre-processed with various filters to help
7842 the metrics calculation while keeping the frame selection lossless. When set to
7843 @code{1}, the first stream is for the pre-processed input, and the second
7844 stream is the clean source from where the kept frames are chosen. Default is
7848 Set whether or not chroma is considered in the metric calculations. Default is
7854 Apply 2D deconvolution of video stream in frequency domain using second stream
7857 The filter accepts the following options:
7861 Set which planes to process.
7864 Set which impulse video frames will be processed, can be @var{first}
7865 or @var{all}. Default is @var{all}.
7868 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7869 and height are not same and not power of 2 or if stream prior to convolving
7873 The @code{deconvolve} filter also supports the @ref{framesync} options.
7877 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7879 It accepts the following options:
7883 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7884 @var{rainbows} for cross-color reduction.
7887 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7890 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7893 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7896 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7901 Apply deflate effect to the video.
7903 This filter replaces the pixel by the local(3x3) average by taking into account
7904 only values lower than the pixel.
7906 It accepts the following options:
7913 Limit the maximum change for each plane, default is 65535.
7914 If 0, plane will remain unchanged.
7919 Remove temporal frame luminance variations.
7921 It accepts the following options:
7925 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7928 Set averaging mode to smooth temporal luminance variations.
7930 Available values are:
7955 Do not actually modify frame. Useful when one only wants metadata.
7960 Remove judder produced by partially interlaced telecined content.
7962 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7963 source was partially telecined content then the output of @code{pullup,dejudder}
7964 will have a variable frame rate. May change the recorded frame rate of the
7965 container. Aside from that change, this filter will not affect constant frame
7968 The option available in this filter is:
7972 Specify the length of the window over which the judder repeats.
7974 Accepts any integer greater than 1. Useful values are:
7978 If the original was telecined from 24 to 30 fps (Film to NTSC).
7981 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7984 If a mixture of the two.
7987 The default is @samp{4}.
7992 Suppress a TV station logo by a simple interpolation of the surrounding
7993 pixels. Just set a rectangle covering the logo and watch it disappear
7994 (and sometimes something even uglier appear - your mileage may vary).
7996 It accepts the following parameters:
8001 Specify the top left corner coordinates of the logo. They must be
8006 Specify the width and height of the logo to clear. They must be
8010 Specify the thickness of the fuzzy edge of the rectangle (added to
8011 @var{w} and @var{h}). The default value is 1. This option is
8012 deprecated, setting higher values should no longer be necessary and
8016 When set to 1, a green rectangle is drawn on the screen to simplify
8017 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8018 The default value is 0.
8020 The rectangle is drawn on the outermost pixels which will be (partly)
8021 replaced with interpolated values. The values of the next pixels
8022 immediately outside this rectangle in each direction will be used to
8023 compute the interpolated pixel values inside the rectangle.
8027 @subsection Examples
8031 Set a rectangle covering the area with top left corner coordinates 0,0
8032 and size 100x77, and a band of size 10:
8034 delogo=x=0:y=0:w=100:h=77:band=10
8041 Attempt to fix small changes in horizontal and/or vertical shift. This
8042 filter helps remove camera shake from hand-holding a camera, bumping a
8043 tripod, moving on a vehicle, etc.
8045 The filter accepts the following options:
8053 Specify a rectangular area where to limit the search for motion
8055 If desired the search for motion vectors can be limited to a
8056 rectangular area of the frame defined by its top left corner, width
8057 and height. These parameters have the same meaning as the drawbox
8058 filter which can be used to visualise the position of the bounding
8061 This is useful when simultaneous movement of subjects within the frame
8062 might be confused for camera motion by the motion vector search.
8064 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8065 then the full frame is used. This allows later options to be set
8066 without specifying the bounding box for the motion vector search.
8068 Default - search the whole frame.
8072 Specify the maximum extent of movement in x and y directions in the
8073 range 0-64 pixels. Default 16.
8076 Specify how to generate pixels to fill blanks at the edge of the
8077 frame. Available values are:
8080 Fill zeroes at blank locations
8082 Original image at blank locations
8084 Extruded edge value at blank locations
8086 Mirrored edge at blank locations
8088 Default value is @samp{mirror}.
8091 Specify the blocksize to use for motion search. Range 4-128 pixels,
8095 Specify the contrast threshold for blocks. Only blocks with more than
8096 the specified contrast (difference between darkest and lightest
8097 pixels) will be considered. Range 1-255, default 125.
8100 Specify the search strategy. Available values are:
8103 Set exhaustive search
8105 Set less exhaustive search.
8107 Default value is @samp{exhaustive}.
8110 If set then a detailed log of the motion search is written to the
8117 Remove unwanted contamination of foreground colors, caused by reflected color of
8118 greenscreen or bluescreen.
8120 This filter accepts the following options:
8124 Set what type of despill to use.
8127 Set how spillmap will be generated.
8130 Set how much to get rid of still remaining spill.
8133 Controls amount of red in spill area.
8136 Controls amount of green in spill area.
8137 Should be -1 for greenscreen.
8140 Controls amount of blue in spill area.
8141 Should be -1 for bluescreen.
8144 Controls brightness of spill area, preserving colors.
8147 Modify alpha from generated spillmap.
8152 Apply an exact inverse of the telecine operation. It requires a predefined
8153 pattern specified using the pattern option which must be the same as that passed
8154 to the telecine filter.
8156 This filter accepts the following options:
8165 The default value is @code{top}.
8169 A string of numbers representing the pulldown pattern you wish to apply.
8170 The default value is @code{23}.
8173 A number representing position of the first frame with respect to the telecine
8174 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8179 Apply dilation effect to the video.
8181 This filter replaces the pixel by the local(3x3) maximum.
8183 It accepts the following options:
8190 Limit the maximum change for each plane, default is 65535.
8191 If 0, plane will remain unchanged.
8194 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8197 Flags to local 3x3 coordinates maps like this:
8206 Displace pixels as indicated by second and third input stream.
8208 It takes three input streams and outputs one stream, the first input is the
8209 source, and second and third input are displacement maps.
8211 The second input specifies how much to displace pixels along the
8212 x-axis, while the third input specifies how much to displace pixels
8214 If one of displacement map streams terminates, last frame from that
8215 displacement map will be used.
8217 Note that once generated, displacements maps can be reused over and over again.
8219 A description of the accepted options follows.
8223 Set displace behavior for pixels that are out of range.
8225 Available values are:
8228 Missing pixels are replaced by black pixels.
8231 Adjacent pixels will spread out to replace missing pixels.
8234 Out of range pixels are wrapped so they point to pixels of other side.
8237 Out of range pixels will be replaced with mirrored pixels.
8239 Default is @samp{smear}.
8243 @subsection Examples
8247 Add ripple effect to rgb input of video size hd720:
8249 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
8253 Add wave effect to rgb input of video size hd720:
8255 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
8261 Draw a colored box on the input image.
8263 It accepts the following parameters:
8268 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8272 The expressions which specify the width and height of the box; if 0 they are interpreted as
8273 the input width and height. It defaults to 0.
8276 Specify the color of the box to write. For the general syntax of this option,
8277 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8278 value @code{invert} is used, the box edge color is the same as the
8279 video with inverted luma.
8282 The expression which sets the thickness of the box edge.
8283 A value of @code{fill} will create a filled box. Default value is @code{3}.
8285 See below for the list of accepted constants.
8288 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8289 will overwrite the video's color and alpha pixels.
8290 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8293 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8294 following constants:
8298 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8302 horizontal and vertical chroma subsample values. For example for the
8303 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8307 The input width and height.
8310 The input sample aspect ratio.
8314 The x and y offset coordinates where the box is drawn.
8318 The width and height of the drawn box.
8321 The thickness of the drawn box.
8323 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8324 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8328 @subsection Examples
8332 Draw a black box around the edge of the input image:
8338 Draw a box with color red and an opacity of 50%:
8340 drawbox=10:20:200:60:red@@0.5
8343 The previous example can be specified as:
8345 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8349 Fill the box with pink color:
8351 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8355 Draw a 2-pixel red 2.40:1 mask:
8357 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
8363 Draw a grid on the input image.
8365 It accepts the following parameters:
8370 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8374 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8375 input width and height, respectively, minus @code{thickness}, so image gets
8376 framed. Default to 0.
8379 Specify the color of the grid. For the general syntax of this option,
8380 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8381 value @code{invert} is used, the grid color is the same as the
8382 video with inverted luma.
8385 The expression which sets the thickness of the grid line. Default value is @code{1}.
8387 See below for the list of accepted constants.
8390 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8391 will overwrite the video's color and alpha pixels.
8392 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8395 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8396 following constants:
8400 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8404 horizontal and vertical chroma subsample values. For example for the
8405 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8409 The input grid cell width and height.
8412 The input sample aspect ratio.
8416 The x and y coordinates of some point of grid intersection (meant to configure offset).
8420 The width and height of the drawn cell.
8423 The thickness of the drawn cell.
8425 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8426 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8430 @subsection Examples
8434 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8436 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8440 Draw a white 3x3 grid with an opacity of 50%:
8442 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8449 Draw a text string or text from a specified file on top of a video, using the
8450 libfreetype library.
8452 To enable compilation of this filter, you need to configure FFmpeg with
8453 @code{--enable-libfreetype}.
8454 To enable default font fallback and the @var{font} option you need to
8455 configure FFmpeg with @code{--enable-libfontconfig}.
8456 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8457 @code{--enable-libfribidi}.
8461 It accepts the following parameters:
8466 Used to draw a box around text using the background color.
8467 The value must be either 1 (enable) or 0 (disable).
8468 The default value of @var{box} is 0.
8471 Set the width of the border to be drawn around the box using @var{boxcolor}.
8472 The default value of @var{boxborderw} is 0.
8475 The color to be used for drawing box around text. For the syntax of this
8476 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8478 The default value of @var{boxcolor} is "white".
8481 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8482 The default value of @var{line_spacing} is 0.
8485 Set the width of the border to be drawn around the text using @var{bordercolor}.
8486 The default value of @var{borderw} is 0.
8489 Set the color to be used for drawing border around text. For the syntax of this
8490 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8492 The default value of @var{bordercolor} is "black".
8495 Select how the @var{text} is expanded. Can be either @code{none},
8496 @code{strftime} (deprecated) or
8497 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8501 Set a start time for the count. Value is in microseconds. Only applied
8502 in the deprecated strftime expansion mode. To emulate in normal expansion
8503 mode use the @code{pts} function, supplying the start time (in seconds)
8504 as the second argument.
8507 If true, check and fix text coords to avoid clipping.
8510 The color to be used for drawing fonts. For the syntax of this option, check
8511 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8513 The default value of @var{fontcolor} is "black".
8515 @item fontcolor_expr
8516 String which is expanded the same way as @var{text} to obtain dynamic
8517 @var{fontcolor} value. By default this option has empty value and is not
8518 processed. When this option is set, it overrides @var{fontcolor} option.
8521 The font family to be used for drawing text. By default Sans.
8524 The font file to be used for drawing text. The path must be included.
8525 This parameter is mandatory if the fontconfig support is disabled.
8528 Draw the text applying alpha blending. The value can
8529 be a number between 0.0 and 1.0.
8530 The expression accepts the same variables @var{x, y} as well.
8531 The default value is 1.
8532 Please see @var{fontcolor_expr}.
8535 The font size to be used for drawing text.
8536 The default value of @var{fontsize} is 16.
8539 If set to 1, attempt to shape the text (for example, reverse the order of
8540 right-to-left text and join Arabic characters) before drawing it.
8541 Otherwise, just draw the text exactly as given.
8542 By default 1 (if supported).
8545 The flags to be used for loading the fonts.
8547 The flags map the corresponding flags supported by libfreetype, and are
8548 a combination of the following values:
8555 @item vertical_layout
8556 @item force_autohint
8559 @item ignore_global_advance_width
8561 @item ignore_transform
8567 Default value is "default".
8569 For more information consult the documentation for the FT_LOAD_*
8573 The color to be used for drawing a shadow behind the drawn text. For the
8574 syntax of this option, check the @ref{color syntax,,"Color" section in the
8575 ffmpeg-utils manual,ffmpeg-utils}.
8577 The default value of @var{shadowcolor} is "black".
8581 The x and y offsets for the text shadow position with respect to the
8582 position of the text. They can be either positive or negative
8583 values. The default value for both is "0".
8586 The starting frame number for the n/frame_num variable. The default value
8590 The size in number of spaces to use for rendering the tab.
8594 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8595 format. It can be used with or without text parameter. @var{timecode_rate}
8596 option must be specified.
8598 @item timecode_rate, rate, r
8599 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8600 integer. Minimum value is "1".
8601 Drop-frame timecode is supported for frame rates 30 & 60.
8604 If set to 1, the output of the timecode option will wrap around at 24 hours.
8605 Default is 0 (disabled).
8608 The text string to be drawn. The text must be a sequence of UTF-8
8610 This parameter is mandatory if no file is specified with the parameter
8614 A text file containing text to be drawn. The text must be a sequence
8615 of UTF-8 encoded characters.
8617 This parameter is mandatory if no text string is specified with the
8618 parameter @var{text}.
8620 If both @var{text} and @var{textfile} are specified, an error is thrown.
8623 If set to 1, the @var{textfile} will be reloaded before each frame.
8624 Be sure to update it atomically, or it may be read partially, or even fail.
8628 The expressions which specify the offsets where text will be drawn
8629 within the video frame. They are relative to the top/left border of the
8632 The default value of @var{x} and @var{y} is "0".
8634 See below for the list of accepted constants and functions.
8637 The parameters for @var{x} and @var{y} are expressions containing the
8638 following constants and functions:
8642 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8646 horizontal and vertical chroma subsample values. For example for the
8647 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8650 the height of each text line
8658 @item max_glyph_a, ascent
8659 the maximum distance from the baseline to the highest/upper grid
8660 coordinate used to place a glyph outline point, for all the rendered
8662 It is a positive value, due to the grid's orientation with the Y axis
8665 @item max_glyph_d, descent
8666 the maximum distance from the baseline to the lowest grid coordinate
8667 used to place a glyph outline point, for all the rendered glyphs.
8668 This is a negative value, due to the grid's orientation, with the Y axis
8672 maximum glyph height, that is the maximum height for all the glyphs
8673 contained in the rendered text, it is equivalent to @var{ascent} -
8677 maximum glyph width, that is the maximum width for all the glyphs
8678 contained in the rendered text
8681 the number of input frame, starting from 0
8683 @item rand(min, max)
8684 return a random number included between @var{min} and @var{max}
8687 The input sample aspect ratio.
8690 timestamp expressed in seconds, NAN if the input timestamp is unknown
8693 the height of the rendered text
8696 the width of the rendered text
8700 the x and y offset coordinates where the text is drawn.
8702 These parameters allow the @var{x} and @var{y} expressions to refer
8703 each other, so you can for example specify @code{y=x/dar}.
8706 @anchor{drawtext_expansion}
8707 @subsection Text expansion
8709 If @option{expansion} is set to @code{strftime},
8710 the filter recognizes strftime() sequences in the provided text and
8711 expands them accordingly. Check the documentation of strftime(). This
8712 feature is deprecated.
8714 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8716 If @option{expansion} is set to @code{normal} (which is the default),
8717 the following expansion mechanism is used.
8719 The backslash character @samp{\}, followed by any character, always expands to
8720 the second character.
8722 Sequences of the form @code{%@{...@}} are expanded. The text between the
8723 braces is a function name, possibly followed by arguments separated by ':'.
8724 If the arguments contain special characters or delimiters (':' or '@}'),
8725 they should be escaped.
8727 Note that they probably must also be escaped as the value for the
8728 @option{text} option in the filter argument string and as the filter
8729 argument in the filtergraph description, and possibly also for the shell,
8730 that makes up to four levels of escaping; using a text file avoids these
8733 The following functions are available:
8738 The expression evaluation result.
8740 It must take one argument specifying the expression to be evaluated,
8741 which accepts the same constants and functions as the @var{x} and
8742 @var{y} values. Note that not all constants should be used, for
8743 example the text size is not known when evaluating the expression, so
8744 the constants @var{text_w} and @var{text_h} will have an undefined
8747 @item expr_int_format, eif
8748 Evaluate the expression's value and output as formatted integer.
8750 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8751 The second argument specifies the output format. Allowed values are @samp{x},
8752 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8753 @code{printf} function.
8754 The third parameter is optional and sets the number of positions taken by the output.
8755 It can be used to add padding with zeros from the left.
8758 The time at which the filter is running, expressed in UTC.
8759 It can accept an argument: a strftime() format string.
8762 The time at which the filter is running, expressed in the local time zone.
8763 It can accept an argument: a strftime() format string.
8766 Frame metadata. Takes one or two arguments.
8768 The first argument is mandatory and specifies the metadata key.
8770 The second argument is optional and specifies a default value, used when the
8771 metadata key is not found or empty.
8774 The frame number, starting from 0.
8777 A 1 character description of the current picture type.
8780 The timestamp of the current frame.
8781 It can take up to three arguments.
8783 The first argument is the format of the timestamp; it defaults to @code{flt}
8784 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8785 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8786 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8787 @code{localtime} stands for the timestamp of the frame formatted as
8788 local time zone time.
8790 The second argument is an offset added to the timestamp.
8792 If the format is set to @code{hms}, a third argument @code{24HH} may be
8793 supplied to present the hour part of the formatted timestamp in 24h format
8796 If the format is set to @code{localtime} or @code{gmtime},
8797 a third argument may be supplied: a strftime() format string.
8798 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8801 @subsection Examples
8805 Draw "Test Text" with font FreeSerif, using the default values for the
8806 optional parameters.
8809 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8813 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8814 and y=50 (counting from the top-left corner of the screen), text is
8815 yellow with a red box around it. Both the text and the box have an
8819 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8820 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8823 Note that the double quotes are not necessary if spaces are not used
8824 within the parameter list.
8827 Show the text at the center of the video frame:
8829 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8833 Show the text at a random position, switching to a new position every 30 seconds:
8835 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)"
8839 Show a text line sliding from right to left in the last row of the video
8840 frame. The file @file{LONG_LINE} is assumed to contain a single line
8843 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8847 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8849 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8853 Draw a single green letter "g", at the center of the input video.
8854 The glyph baseline is placed at half screen height.
8856 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8860 Show text for 1 second every 3 seconds:
8862 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8866 Use fontconfig to set the font. Note that the colons need to be escaped.
8868 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8872 Print the date of a real-time encoding (see strftime(3)):
8874 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8878 Show text fading in and out (appearing/disappearing):
8881 DS=1.0 # display start
8882 DE=10.0 # display end
8883 FID=1.5 # fade in duration
8884 FOD=5 # fade out duration
8885 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 @}"
8889 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8890 and the @option{fontsize} value are included in the @option{y} offset.
8892 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8893 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8898 For more information about libfreetype, check:
8899 @url{http://www.freetype.org/}.
8901 For more information about fontconfig, check:
8902 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8904 For more information about libfribidi, check:
8905 @url{http://fribidi.org/}.
8909 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8911 The filter accepts the following options:
8916 Set low and high threshold values used by the Canny thresholding
8919 The high threshold selects the "strong" edge pixels, which are then
8920 connected through 8-connectivity with the "weak" edge pixels selected
8921 by the low threshold.
8923 @var{low} and @var{high} threshold values must be chosen in the range
8924 [0,1], and @var{low} should be lesser or equal to @var{high}.
8926 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8930 Define the drawing mode.
8934 Draw white/gray wires on black background.
8937 Mix the colors to create a paint/cartoon effect.
8940 Apply Canny edge detector on all selected planes.
8942 Default value is @var{wires}.
8945 Select planes for filtering. By default all available planes are filtered.
8948 @subsection Examples
8952 Standard edge detection with custom values for the hysteresis thresholding:
8954 edgedetect=low=0.1:high=0.4
8958 Painting effect without thresholding:
8960 edgedetect=mode=colormix:high=0
8965 Set brightness, contrast, saturation and approximate gamma adjustment.
8967 The filter accepts the following options:
8971 Set the contrast expression. The value must be a float value in range
8972 @code{-2.0} to @code{2.0}. The default value is "1".
8975 Set the brightness expression. The value must be a float value in
8976 range @code{-1.0} to @code{1.0}. The default value is "0".
8979 Set the saturation expression. The value must be a float in
8980 range @code{0.0} to @code{3.0}. The default value is "1".
8983 Set the gamma expression. The value must be a float in range
8984 @code{0.1} to @code{10.0}. The default value is "1".
8987 Set the gamma expression for red. The value must be a float in
8988 range @code{0.1} to @code{10.0}. The default value is "1".
8991 Set the gamma expression for green. The value must be a float in range
8992 @code{0.1} to @code{10.0}. The default value is "1".
8995 Set the gamma expression for blue. The value must be a float in range
8996 @code{0.1} to @code{10.0}. The default value is "1".
8999 Set the gamma weight expression. It can be used to reduce the effect
9000 of a high gamma value on bright image areas, e.g. keep them from
9001 getting overamplified and just plain white. The value must be a float
9002 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9003 gamma correction all the way down while @code{1.0} leaves it at its
9004 full strength. Default is "1".
9007 Set when the expressions for brightness, contrast, saturation and
9008 gamma expressions are evaluated.
9010 It accepts the following values:
9013 only evaluate expressions once during the filter initialization or
9014 when a command is processed
9017 evaluate expressions for each incoming frame
9020 Default value is @samp{init}.
9023 The expressions accept the following parameters:
9026 frame count of the input frame starting from 0
9029 byte position of the corresponding packet in the input file, NAN if
9033 frame rate of the input video, NAN if the input frame rate is unknown
9036 timestamp expressed in seconds, NAN if the input timestamp is unknown
9039 @subsection Commands
9040 The filter supports the following commands:
9044 Set the contrast expression.
9047 Set the brightness expression.
9050 Set the saturation expression.
9053 Set the gamma expression.
9056 Set the gamma_r expression.
9059 Set gamma_g expression.
9062 Set gamma_b expression.
9065 Set gamma_weight expression.
9067 The command accepts the same syntax of the corresponding option.
9069 If the specified expression is not valid, it is kept at its current
9076 Apply erosion effect to the video.
9078 This filter replaces the pixel by the local(3x3) minimum.
9080 It accepts the following options:
9087 Limit the maximum change for each plane, default is 65535.
9088 If 0, plane will remain unchanged.
9091 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9094 Flags to local 3x3 coordinates maps like this:
9101 @section extractplanes
9103 Extract color channel components from input video stream into
9104 separate grayscale video streams.
9106 The filter accepts the following option:
9110 Set plane(s) to extract.
9112 Available values for planes are:
9123 Choosing planes not available in the input will result in an error.
9124 That means you cannot select @code{r}, @code{g}, @code{b} planes
9125 with @code{y}, @code{u}, @code{v} planes at same time.
9128 @subsection Examples
9132 Extract luma, u and v color channel component from input video frame
9133 into 3 grayscale outputs:
9135 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
9141 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9143 For each input image, the filter will compute the optimal mapping from
9144 the input to the output given the codebook length, that is the number
9145 of distinct output colors.
9147 This filter accepts the following options.
9150 @item codebook_length, l
9151 Set codebook length. The value must be a positive integer, and
9152 represents the number of distinct output colors. Default value is 256.
9155 Set the maximum number of iterations to apply for computing the optimal
9156 mapping. The higher the value the better the result and the higher the
9157 computation time. Default value is 1.
9160 Set a random seed, must be an integer included between 0 and
9161 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9162 will try to use a good random seed on a best effort basis.
9165 Set pal8 output pixel format. This option does not work with codebook
9166 length greater than 256.
9171 Measure graylevel entropy in histogram of color channels of video frames.
9173 It accepts the following parameters:
9177 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9179 @var{diff} mode measures entropy of histogram delta values, absolute differences
9180 between neighbour histogram values.
9185 Apply a fade-in/out effect to the input video.
9187 It accepts the following parameters:
9191 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9193 Default is @code{in}.
9195 @item start_frame, s
9196 Specify the number of the frame to start applying the fade
9197 effect at. Default is 0.
9200 The number of frames that the fade effect lasts. At the end of the
9201 fade-in effect, the output video will have the same intensity as the input video.
9202 At the end of the fade-out transition, the output video will be filled with the
9203 selected @option{color}.
9207 If set to 1, fade only alpha channel, if one exists on the input.
9210 @item start_time, st
9211 Specify the timestamp (in seconds) of the frame to start to apply the fade
9212 effect. If both start_frame and start_time are specified, the fade will start at
9213 whichever comes last. Default is 0.
9216 The number of seconds for which the fade effect has to last. At the end of the
9217 fade-in effect the output video will have the same intensity as the input video,
9218 at the end of the fade-out transition the output video will be filled with the
9219 selected @option{color}.
9220 If both duration and nb_frames are specified, duration is used. Default is 0
9221 (nb_frames is used by default).
9224 Specify the color of the fade. Default is "black".
9227 @subsection Examples
9231 Fade in the first 30 frames of video:
9236 The command above is equivalent to:
9242 Fade out the last 45 frames of a 200-frame video:
9245 fade=type=out:start_frame=155:nb_frames=45
9249 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9251 fade=in:0:25, fade=out:975:25
9255 Make the first 5 frames yellow, then fade in from frame 5-24:
9257 fade=in:5:20:color=yellow
9261 Fade in alpha over first 25 frames of video:
9263 fade=in:0:25:alpha=1
9267 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9269 fade=t=in:st=5.5:d=0.5
9275 Apply arbitrary expressions to samples in frequency domain
9279 Adjust the dc value (gain) of the luma plane of the image. The filter
9280 accepts an integer value in range @code{0} to @code{1000}. The default
9281 value is set to @code{0}.
9284 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9285 filter accepts an integer value in range @code{0} to @code{1000}. The
9286 default value is set to @code{0}.
9289 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9290 filter accepts an integer value in range @code{0} to @code{1000}. The
9291 default value is set to @code{0}.
9294 Set the frequency domain weight expression for the luma plane.
9297 Set the frequency domain weight expression for the 1st chroma plane.
9300 Set the frequency domain weight expression for the 2nd chroma plane.
9303 Set when the expressions are evaluated.
9305 It accepts the following values:
9308 Only evaluate expressions once during the filter initialization.
9311 Evaluate expressions for each incoming frame.
9314 Default value is @samp{init}.
9316 The filter accepts the following variables:
9319 The coordinates of the current sample.
9323 The width and height of the image.
9326 The number of input frame, starting from 0.
9329 @subsection Examples
9335 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9341 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9347 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9353 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9359 Denoise frames using 3D FFT (frequency domain filtering).
9361 The filter accepts the following options:
9365 Set the noise sigma constant. This sets denoising strength.
9366 Default value is 1. Allowed range is from 0 to 30.
9367 Using very high sigma with low overlap may give blocking artifacts.
9370 Set amount of denoising. By default all detected noise is reduced.
9371 Default value is 1. Allowed range is from 0 to 1.
9374 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9375 Actual size of block in pixels is 2 to power of @var{block}, so by default
9376 block size in pixels is 2^4 which is 16.
9379 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9382 Set number of previous frames to use for denoising. By default is set to 0.
9385 Set number of next frames to to use for denoising. By default is set to 0.
9388 Set planes which will be filtered, by default are all available filtered
9394 Extract a single field from an interlaced image using stride
9395 arithmetic to avoid wasting CPU time. The output frames are marked as
9398 The filter accepts the following options:
9402 Specify whether to extract the top (if the value is @code{0} or
9403 @code{top}) or the bottom field (if the value is @code{1} or
9409 Create new frames by copying the top and bottom fields from surrounding frames
9410 supplied as numbers by the hint file.
9414 Set file containing hints: absolute/relative frame numbers.
9416 There must be one line for each frame in a clip. Each line must contain two
9417 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9418 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9419 is current frame number for @code{absolute} mode or out of [-1, 1] range
9420 for @code{relative} mode. First number tells from which frame to pick up top
9421 field and second number tells from which frame to pick up bottom field.
9423 If optionally followed by @code{+} output frame will be marked as interlaced,
9424 else if followed by @code{-} output frame will be marked as progressive, else
9425 it will be marked same as input frame.
9426 If line starts with @code{#} or @code{;} that line is skipped.
9429 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9432 Example of first several lines of @code{hint} file for @code{relative} mode:
9435 1,0 - # second frame, use third's frame top field and second's frame bottom field
9436 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9453 Field matching filter for inverse telecine. It is meant to reconstruct the
9454 progressive frames from a telecined stream. The filter does not drop duplicated
9455 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9456 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9458 The separation of the field matching and the decimation is notably motivated by
9459 the possibility of inserting a de-interlacing filter fallback between the two.
9460 If the source has mixed telecined and real interlaced content,
9461 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9462 But these remaining combed frames will be marked as interlaced, and thus can be
9463 de-interlaced by a later filter such as @ref{yadif} before decimation.
9465 In addition to the various configuration options, @code{fieldmatch} can take an
9466 optional second stream, activated through the @option{ppsrc} option. If
9467 enabled, the frames reconstruction will be based on the fields and frames from
9468 this second stream. This allows the first input to be pre-processed in order to
9469 help the various algorithms of the filter, while keeping the output lossless
9470 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9471 or brightness/contrast adjustments can help.
9473 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9474 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9475 which @code{fieldmatch} is based on. While the semantic and usage are very
9476 close, some behaviour and options names can differ.
9478 The @ref{decimate} filter currently only works for constant frame rate input.
9479 If your input has mixed telecined (30fps) and progressive content with a lower
9480 framerate like 24fps use the following filterchain to produce the necessary cfr
9481 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9483 The filter accepts the following options:
9487 Specify the assumed field order of the input stream. Available values are:
9491 Auto detect parity (use FFmpeg's internal parity value).
9493 Assume bottom field first.
9495 Assume top field first.
9498 Note that it is sometimes recommended not to trust the parity announced by the
9501 Default value is @var{auto}.
9504 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9505 sense that it won't risk creating jerkiness due to duplicate frames when
9506 possible, but if there are bad edits or blended fields it will end up
9507 outputting combed frames when a good match might actually exist. On the other
9508 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9509 but will almost always find a good frame if there is one. The other values are
9510 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9511 jerkiness and creating duplicate frames versus finding good matches in sections
9512 with bad edits, orphaned fields, blended fields, etc.
9514 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9516 Available values are:
9520 2-way matching (p/c)
9522 2-way matching, and trying 3rd match if still combed (p/c + n)
9524 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9526 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9527 still combed (p/c + n + u/b)
9529 3-way matching (p/c/n)
9531 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9532 detected as combed (p/c/n + u/b)
9535 The parenthesis at the end indicate the matches that would be used for that
9536 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9539 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9542 Default value is @var{pc_n}.
9545 Mark the main input stream as a pre-processed input, and enable the secondary
9546 input stream as the clean source to pick the fields from. See the filter
9547 introduction for more details. It is similar to the @option{clip2} feature from
9550 Default value is @code{0} (disabled).
9553 Set the field to match from. It is recommended to set this to the same value as
9554 @option{order} unless you experience matching failures with that setting. In
9555 certain circumstances changing the field that is used to match from can have a
9556 large impact on matching performance. Available values are:
9560 Automatic (same value as @option{order}).
9562 Match from the bottom field.
9564 Match from the top field.
9567 Default value is @var{auto}.
9570 Set whether or not chroma is included during the match comparisons. In most
9571 cases it is recommended to leave this enabled. You should set this to @code{0}
9572 only if your clip has bad chroma problems such as heavy rainbowing or other
9573 artifacts. Setting this to @code{0} could also be used to speed things up at
9574 the cost of some accuracy.
9576 Default value is @code{1}.
9580 These define an exclusion band which excludes the lines between @option{y0} and
9581 @option{y1} from being included in the field matching decision. An exclusion
9582 band can be used to ignore subtitles, a logo, or other things that may
9583 interfere with the matching. @option{y0} sets the starting scan line and
9584 @option{y1} sets the ending line; all lines in between @option{y0} and
9585 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9586 @option{y0} and @option{y1} to the same value will disable the feature.
9587 @option{y0} and @option{y1} defaults to @code{0}.
9590 Set the scene change detection threshold as a percentage of maximum change on
9591 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9592 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9593 @option{scthresh} is @code{[0.0, 100.0]}.
9595 Default value is @code{12.0}.
9598 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9599 account the combed scores of matches when deciding what match to use as the
9600 final match. Available values are:
9604 No final matching based on combed scores.
9606 Combed scores are only used when a scene change is detected.
9608 Use combed scores all the time.
9611 Default is @var{sc}.
9614 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9615 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9616 Available values are:
9620 No forced calculation.
9622 Force p/c/n calculations.
9624 Force p/c/n/u/b calculations.
9627 Default value is @var{none}.
9630 This is the area combing threshold used for combed frame detection. This
9631 essentially controls how "strong" or "visible" combing must be to be detected.
9632 Larger values mean combing must be more visible and smaller values mean combing
9633 can be less visible or strong and still be detected. Valid settings are from
9634 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9635 be detected as combed). This is basically a pixel difference value. A good
9636 range is @code{[8, 12]}.
9638 Default value is @code{9}.
9641 Sets whether or not chroma is considered in the combed frame decision. Only
9642 disable this if your source has chroma problems (rainbowing, etc.) that are
9643 causing problems for the combed frame detection with chroma enabled. Actually,
9644 using @option{chroma}=@var{0} is usually more reliable, except for the case
9645 where there is chroma only combing in the source.
9647 Default value is @code{0}.
9651 Respectively set the x-axis and y-axis size of the window used during combed
9652 frame detection. This has to do with the size of the area in which
9653 @option{combpel} pixels are required to be detected as combed for a frame to be
9654 declared combed. See the @option{combpel} parameter description for more info.
9655 Possible values are any number that is a power of 2 starting at 4 and going up
9658 Default value is @code{16}.
9661 The number of combed pixels inside any of the @option{blocky} by
9662 @option{blockx} size blocks on the frame for the frame to be detected as
9663 combed. While @option{cthresh} controls how "visible" the combing must be, this
9664 setting controls "how much" combing there must be in any localized area (a
9665 window defined by the @option{blockx} and @option{blocky} settings) on the
9666 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9667 which point no frames will ever be detected as combed). This setting is known
9668 as @option{MI} in TFM/VFM vocabulary.
9670 Default value is @code{80}.
9673 @anchor{p/c/n/u/b meaning}
9674 @subsection p/c/n/u/b meaning
9676 @subsubsection p/c/n
9678 We assume the following telecined stream:
9681 Top fields: 1 2 2 3 4
9682 Bottom fields: 1 2 3 4 4
9685 The numbers correspond to the progressive frame the fields relate to. Here, the
9686 first two frames are progressive, the 3rd and 4th are combed, and so on.
9688 When @code{fieldmatch} is configured to run a matching from bottom
9689 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9694 B 1 2 3 4 4 <-- matching reference
9703 As a result of the field matching, we can see that some frames get duplicated.
9704 To perform a complete inverse telecine, you need to rely on a decimation filter
9705 after this operation. See for instance the @ref{decimate} filter.
9707 The same operation now matching from top fields (@option{field}=@var{top})
9712 T 1 2 2 3 4 <-- matching reference
9722 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9723 basically, they refer to the frame and field of the opposite parity:
9726 @item @var{p} matches the field of the opposite parity in the previous frame
9727 @item @var{c} matches the field of the opposite parity in the current frame
9728 @item @var{n} matches the field of the opposite parity in the next frame
9733 The @var{u} and @var{b} matching are a bit special in the sense that they match
9734 from the opposite parity flag. In the following examples, we assume that we are
9735 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9736 'x' is placed above and below each matched fields.
9738 With bottom matching (@option{field}=@var{bottom}):
9743 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9744 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9752 With top matching (@option{field}=@var{top}):
9757 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9758 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9766 @subsection Examples
9768 Simple IVTC of a top field first telecined stream:
9770 fieldmatch=order=tff:combmatch=none, decimate
9773 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9775 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9780 Transform the field order of the input video.
9782 It accepts the following parameters:
9787 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9788 for bottom field first.
9791 The default value is @samp{tff}.
9793 The transformation is done by shifting the picture content up or down
9794 by one line, and filling the remaining line with appropriate picture content.
9795 This method is consistent with most broadcast field order converters.
9797 If the input video is not flagged as being interlaced, or it is already
9798 flagged as being of the required output field order, then this filter does
9799 not alter the incoming video.
9801 It is very useful when converting to or from PAL DV material,
9802 which is bottom field first.
9806 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9809 @section fifo, afifo
9811 Buffer input images and send them when they are requested.
9813 It is mainly useful when auto-inserted by the libavfilter
9816 It does not take parameters.
9818 @section fillborders
9820 Fill borders of the input video, without changing video stream dimensions.
9821 Sometimes video can have garbage at the four edges and you may not want to
9822 crop video input to keep size multiple of some number.
9824 This filter accepts the following options:
9828 Number of pixels to fill from left border.
9831 Number of pixels to fill from right border.
9834 Number of pixels to fill from top border.
9837 Number of pixels to fill from bottom border.
9842 It accepts the following values:
9845 fill pixels using outermost pixels
9848 fill pixels using mirroring
9851 fill pixels with constant value
9854 Default is @var{smear}.
9857 Set color for pixels in fixed mode. Default is @var{black}.
9862 Find a rectangular object
9864 It accepts the following options:
9868 Filepath of the object image, needs to be in gray8.
9871 Detection threshold, default is 0.5.
9874 Number of mipmaps, default is 3.
9876 @item xmin, ymin, xmax, ymax
9877 Specifies the rectangle in which to search.
9880 @subsection Examples
9884 Generate a representative palette of a given video using @command{ffmpeg}:
9886 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9892 Cover a rectangular object
9894 It accepts the following options:
9898 Filepath of the optional cover image, needs to be in yuv420.
9903 It accepts the following values:
9906 cover it by the supplied image
9908 cover it by interpolating the surrounding pixels
9911 Default value is @var{blur}.
9914 @subsection Examples
9918 Generate a representative palette of a given video using @command{ffmpeg}:
9920 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9926 Flood area with values of same pixel components with another values.
9928 It accepts the following options:
9931 Set pixel x coordinate.
9934 Set pixel y coordinate.
9937 Set source #0 component value.
9940 Set source #1 component value.
9943 Set source #2 component value.
9946 Set source #3 component value.
9949 Set destination #0 component value.
9952 Set destination #1 component value.
9955 Set destination #2 component value.
9958 Set destination #3 component value.
9964 Convert the input video to one of the specified pixel formats.
9965 Libavfilter will try to pick one that is suitable as input to
9968 It accepts the following parameters:
9972 A '|'-separated list of pixel format names, such as
9973 "pix_fmts=yuv420p|monow|rgb24".
9977 @subsection Examples
9981 Convert the input video to the @var{yuv420p} format
9983 format=pix_fmts=yuv420p
9986 Convert the input video to any of the formats in the list
9988 format=pix_fmts=yuv420p|yuv444p|yuv410p
9995 Convert the video to specified constant frame rate by duplicating or dropping
9996 frames as necessary.
9998 It accepts the following parameters:
10002 The desired output frame rate. The default is @code{25}.
10005 Assume the first PTS should be the given value, in seconds. This allows for
10006 padding/trimming at the start of stream. By default, no assumption is made
10007 about the first frame's expected PTS, so no padding or trimming is done.
10008 For example, this could be set to 0 to pad the beginning with duplicates of
10009 the first frame if a video stream starts after the audio stream or to trim any
10010 frames with a negative PTS.
10013 Timestamp (PTS) rounding method.
10015 Possible values are:
10022 round towards -infinity
10024 round towards +infinity
10028 The default is @code{near}.
10031 Action performed when reading the last frame.
10033 Possible values are:
10036 Use same timestamp rounding method as used for other frames.
10038 Pass through last frame if input duration has not been reached yet.
10040 The default is @code{round}.
10044 Alternatively, the options can be specified as a flat string:
10045 @var{fps}[:@var{start_time}[:@var{round}]].
10047 See also the @ref{setpts} filter.
10049 @subsection Examples
10053 A typical usage in order to set the fps to 25:
10059 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10061 fps=fps=film:round=near
10067 Pack two different video streams into a stereoscopic video, setting proper
10068 metadata on supported codecs. The two views should have the same size and
10069 framerate and processing will stop when the shorter video ends. Please note
10070 that you may conveniently adjust view properties with the @ref{scale} and
10073 It accepts the following parameters:
10077 The desired packing format. Supported values are:
10082 The views are next to each other (default).
10085 The views are on top of each other.
10088 The views are packed by line.
10091 The views are packed by column.
10094 The views are temporally interleaved.
10103 # Convert left and right views into a frame-sequential video
10104 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10106 # Convert views into a side-by-side video with the same output resolution as the input
10107 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
10112 Change the frame rate by interpolating new video output frames from the source
10115 This filter is not designed to function correctly with interlaced media. If
10116 you wish to change the frame rate of interlaced media then you are required
10117 to deinterlace before this filter and re-interlace after this filter.
10119 A description of the accepted options follows.
10123 Specify the output frames per second. This option can also be specified
10124 as a value alone. The default is @code{50}.
10127 Specify the start of a range where the output frame will be created as a
10128 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10129 the default is @code{15}.
10132 Specify the end of a range where the output frame will be created as a
10133 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10134 the default is @code{240}.
10137 Specify the level at which a scene change is detected as a value between
10138 0 and 100 to indicate a new scene; a low value reflects a low
10139 probability for the current frame to introduce a new scene, while a higher
10140 value means the current frame is more likely to be one.
10141 The default is @code{8.2}.
10144 Specify flags influencing the filter process.
10146 Available value for @var{flags} is:
10149 @item scene_change_detect, scd
10150 Enable scene change detection using the value of the option @var{scene}.
10151 This flag is enabled by default.
10157 Select one frame every N-th frame.
10159 This filter accepts the following option:
10162 Select frame after every @code{step} frames.
10163 Allowed values are positive integers higher than 0. Default value is @code{1}.
10166 @section freezedetect
10168 Detect frozen video.
10170 This filter logs a message and sets frame metadata when it detects that the
10171 input video has no significant change in content during a specified duration.
10172 Video freeze detection calculates the mean average absolute difference of all
10173 the components of video frames and compares it to a noise floor.
10175 The printed times and duration are expressed in seconds. The
10176 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10177 whose timestamp equals or exceeds the detection duration and it contains the
10178 timestamp of the first frame of the freeze. The
10179 @code{lavfi.freezedetect.freeze_duration} and
10180 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10183 The filter accepts the following options:
10187 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10188 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10192 Set freeze duration until notification (default is 2 seconds).
10198 Apply a frei0r effect to the input video.
10200 To enable the compilation of this filter, you need to install the frei0r
10201 header and configure FFmpeg with @code{--enable-frei0r}.
10203 It accepts the following parameters:
10208 The name of the frei0r effect to load. If the environment variable
10209 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10210 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10211 Otherwise, the standard frei0r paths are searched, in this order:
10212 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10213 @file{/usr/lib/frei0r-1/}.
10215 @item filter_params
10216 A '|'-separated list of parameters to pass to the frei0r effect.
10220 A frei0r effect parameter can be a boolean (its value is either
10221 "y" or "n"), a double, a color (specified as
10222 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10223 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10224 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10225 a position (specified as @var{X}/@var{Y}, where
10226 @var{X} and @var{Y} are floating point numbers) and/or a string.
10228 The number and types of parameters depend on the loaded effect. If an
10229 effect parameter is not specified, the default value is set.
10231 @subsection Examples
10235 Apply the distort0r effect, setting the first two double parameters:
10237 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10241 Apply the colordistance effect, taking a color as the first parameter:
10243 frei0r=colordistance:0.2/0.3/0.4
10244 frei0r=colordistance:violet
10245 frei0r=colordistance:0x112233
10249 Apply the perspective effect, specifying the top left and top right image
10252 frei0r=perspective:0.2/0.2|0.8/0.2
10256 For more information, see
10257 @url{http://frei0r.dyne.org}
10261 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10263 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10264 processing filter, one of them is performed once per block, not per pixel.
10265 This allows for much higher speed.
10267 The filter accepts the following options:
10271 Set quality. This option defines the number of levels for averaging. It accepts
10272 an integer in the range 4-5. Default value is @code{4}.
10275 Force a constant quantization parameter. It accepts an integer in range 0-63.
10276 If not set, the filter will use the QP from the video stream (if available).
10279 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10280 more details but also more artifacts, while higher values make the image smoother
10281 but also blurrier. Default value is @code{0} − PSNR optimal.
10283 @item use_bframe_qp
10284 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10285 option may cause flicker since the B-Frames have often larger QP. Default is
10286 @code{0} (not enabled).
10292 Apply Gaussian blur filter.
10294 The filter accepts the following options:
10298 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10301 Set number of steps for Gaussian approximation. Default is @code{1}.
10304 Set which planes to filter. By default all planes are filtered.
10307 Set vertical sigma, if negative it will be same as @code{sigma}.
10308 Default is @code{-1}.
10313 Apply generic equation to each pixel.
10315 The filter accepts the following options:
10318 @item lum_expr, lum
10319 Set the luminance expression.
10321 Set the chrominance blue expression.
10323 Set the chrominance red expression.
10324 @item alpha_expr, a
10325 Set the alpha expression.
10327 Set the red expression.
10328 @item green_expr, g
10329 Set the green expression.
10331 Set the blue expression.
10334 The colorspace is selected according to the specified options. If one
10335 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10336 options is specified, the filter will automatically select a YCbCr
10337 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10338 @option{blue_expr} options is specified, it will select an RGB
10341 If one of the chrominance expression is not defined, it falls back on the other
10342 one. If no alpha expression is specified it will evaluate to opaque value.
10343 If none of chrominance expressions are specified, they will evaluate
10344 to the luminance expression.
10346 The expressions can use the following variables and functions:
10350 The sequential number of the filtered frame, starting from @code{0}.
10354 The coordinates of the current sample.
10358 The width and height of the image.
10362 Width and height scale depending on the currently filtered plane. It is the
10363 ratio between the corresponding luma plane number of pixels and the current
10364 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10365 @code{0.5,0.5} for chroma planes.
10368 Time of the current frame, expressed in seconds.
10371 Return the value of the pixel at location (@var{x},@var{y}) of the current
10375 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10379 Return the value of the pixel at location (@var{x},@var{y}) of the
10380 blue-difference chroma plane. Return 0 if there is no such plane.
10383 Return the value of the pixel at location (@var{x},@var{y}) of the
10384 red-difference chroma plane. Return 0 if there is no such plane.
10389 Return the value of the pixel at location (@var{x},@var{y}) of the
10390 red/green/blue component. Return 0 if there is no such component.
10393 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10394 plane. Return 0 if there is no such plane.
10397 For functions, if @var{x} and @var{y} are outside the area, the value will be
10398 automatically clipped to the closer edge.
10400 @subsection Examples
10404 Flip the image horizontally:
10410 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10411 wavelength of 100 pixels:
10413 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10417 Generate a fancy enigmatic moving light:
10419 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
10423 Generate a quick emboss effect:
10425 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10429 Modify RGB components depending on pixel position:
10431 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10435 Create a radial gradient that is the same size as the input (also see
10436 the @ref{vignette} filter):
10438 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10444 Fix the banding artifacts that are sometimes introduced into nearly flat
10445 regions by truncation to 8-bit color depth.
10446 Interpolate the gradients that should go where the bands are, and
10449 It is designed for playback only. Do not use it prior to
10450 lossy compression, because compression tends to lose the dither and
10451 bring back the bands.
10453 It accepts the following parameters:
10458 The maximum amount by which the filter will change any one pixel. This is also
10459 the threshold for detecting nearly flat regions. Acceptable values range from
10460 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10464 The neighborhood to fit the gradient to. A larger radius makes for smoother
10465 gradients, but also prevents the filter from modifying the pixels near detailed
10466 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10467 values will be clipped to the valid range.
10471 Alternatively, the options can be specified as a flat string:
10472 @var{strength}[:@var{radius}]
10474 @subsection Examples
10478 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10484 Specify radius, omitting the strength (which will fall-back to the default
10492 @section graphmonitor, agraphmonitor
10493 Show various filtergraph stats.
10495 With this filter one can debug complete filtergraph.
10496 Especially issues with links filling with queued frames.
10498 The filter accepts the following options:
10502 Set video output size. Default is @var{hd720}.
10505 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10508 Set output mode, can be @var{fulll} or @var{compact}.
10509 In @var{compact} mode only filters with some queued frames have displayed stats.
10512 Set flags which enable which stats are shown in video.
10514 Available values for flags are:
10517 Display number of queued frames in each link.
10519 @item frame_count_in
10520 Display number of frames taken from filter.
10522 @item frame_count_out
10523 Display number of frames given out from filter.
10526 Display current filtered frame pts.
10529 Display current filtered frame time.
10532 Display time base for filter link.
10535 Display used format for filter link.
10538 Display video size or number of audio channels in case of audio used by filter link.
10541 Display video frame rate or sample rate in case of audio used by filter link.
10545 Set upper limit for video rate of output stream, Default value is @var{25}.
10546 This guarantee that output video frame rate will not be higher than this value.
10550 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10551 and corrects the scene colors accordingly.
10553 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10555 The filter accepts the following options:
10559 The order of differentiation to be applied on the scene. Must be chosen in the range
10560 [0,2] and default value is 1.
10563 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10564 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10565 max value instead of calculating Minkowski distance.
10568 The standard deviation of Gaussian blur to be applied on the scene. Must be
10569 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10570 can't be equal to 0 if @var{difford} is greater than 0.
10573 @subsection Examples
10579 greyedge=difford=1:minknorm=5:sigma=2
10585 greyedge=difford=1:minknorm=0:sigma=2
10593 Apply a Hald CLUT to a video stream.
10595 First input is the video stream to process, and second one is the Hald CLUT.
10596 The Hald CLUT input can be a simple picture or a complete video stream.
10598 The filter accepts the following options:
10602 Force termination when the shortest input terminates. Default is @code{0}.
10604 Continue applying the last CLUT after the end of the stream. A value of
10605 @code{0} disable the filter after the last frame of the CLUT is reached.
10606 Default is @code{1}.
10609 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10610 filters share the same internals).
10612 More information about the Hald CLUT can be found on Eskil Steenberg's website
10613 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10615 @subsection Workflow examples
10617 @subsubsection Hald CLUT video stream
10619 Generate an identity Hald CLUT stream altered with various effects:
10621 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
10624 Note: make sure you use a lossless codec.
10626 Then use it with @code{haldclut} to apply it on some random stream:
10628 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10631 The Hald CLUT will be applied to the 10 first seconds (duration of
10632 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10633 to the remaining frames of the @code{mandelbrot} stream.
10635 @subsubsection Hald CLUT with preview
10637 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10638 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10639 biggest possible square starting at the top left of the picture. The remaining
10640 padding pixels (bottom or right) will be ignored. This area can be used to add
10641 a preview of the Hald CLUT.
10643 Typically, the following generated Hald CLUT will be supported by the
10644 @code{haldclut} filter:
10647 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10648 pad=iw+320 [padded_clut];
10649 smptebars=s=320x256, split [a][b];
10650 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10651 [main][b] overlay=W-320" -frames:v 1 clut.png
10654 It contains the original and a preview of the effect of the CLUT: SMPTE color
10655 bars are displayed on the right-top, and below the same color bars processed by
10658 Then, the effect of this Hald CLUT can be visualized with:
10660 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10665 Flip the input video horizontally.
10667 For example, to horizontally flip the input video with @command{ffmpeg}:
10669 ffmpeg -i in.avi -vf "hflip" out.avi
10673 This filter applies a global color histogram equalization on a
10676 It can be used to correct video that has a compressed range of pixel
10677 intensities. The filter redistributes the pixel intensities to
10678 equalize their distribution across the intensity range. It may be
10679 viewed as an "automatically adjusting contrast filter". This filter is
10680 useful only for correcting degraded or poorly captured source
10683 The filter accepts the following options:
10687 Determine the amount of equalization to be applied. As the strength
10688 is reduced, the distribution of pixel intensities more-and-more
10689 approaches that of the input frame. The value must be a float number
10690 in the range [0,1] and defaults to 0.200.
10693 Set the maximum intensity that can generated and scale the output
10694 values appropriately. The strength should be set as desired and then
10695 the intensity can be limited if needed to avoid washing-out. The value
10696 must be a float number in the range [0,1] and defaults to 0.210.
10699 Set the antibanding level. If enabled the filter will randomly vary
10700 the luminance of output pixels by a small amount to avoid banding of
10701 the histogram. Possible values are @code{none}, @code{weak} or
10702 @code{strong}. It defaults to @code{none}.
10707 Compute and draw a color distribution histogram for the input video.
10709 The computed histogram is a representation of the color component
10710 distribution in an image.
10712 Standard histogram displays the color components distribution in an image.
10713 Displays color graph for each color component. Shows distribution of
10714 the Y, U, V, A or R, G, B components, depending on input format, in the
10715 current frame. Below each graph a color component scale meter is shown.
10717 The filter accepts the following options:
10721 Set height of level. Default value is @code{200}.
10722 Allowed range is [50, 2048].
10725 Set height of color scale. Default value is @code{12}.
10726 Allowed range is [0, 40].
10730 It accepts the following values:
10733 Per color component graphs are placed below each other.
10736 Per color component graphs are placed side by side.
10739 Presents information identical to that in the @code{parade}, except
10740 that the graphs representing color components are superimposed directly
10743 Default is @code{stack}.
10746 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10747 Default is @code{linear}.
10750 Set what color components to display.
10751 Default is @code{7}.
10754 Set foreground opacity. Default is @code{0.7}.
10757 Set background opacity. Default is @code{0.5}.
10760 @subsection Examples
10765 Calculate and draw histogram:
10767 ffplay -i input -vf histogram
10775 This is a high precision/quality 3d denoise filter. It aims to reduce
10776 image noise, producing smooth images and making still images really
10777 still. It should enhance compressibility.
10779 It accepts the following optional parameters:
10783 A non-negative floating point number which specifies spatial luma strength.
10784 It defaults to 4.0.
10786 @item chroma_spatial
10787 A non-negative floating point number which specifies spatial chroma strength.
10788 It defaults to 3.0*@var{luma_spatial}/4.0.
10791 A floating point number which specifies luma temporal strength. It defaults to
10792 6.0*@var{luma_spatial}/4.0.
10795 A floating point number which specifies chroma temporal strength. It defaults to
10796 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10799 @anchor{hwdownload}
10800 @section hwdownload
10802 Download hardware frames to system memory.
10804 The input must be in hardware frames, and the output a non-hardware format.
10805 Not all formats will be supported on the output - it may be necessary to insert
10806 an additional @option{format} filter immediately following in the graph to get
10807 the output in a supported format.
10811 Map hardware frames to system memory or to another device.
10813 This filter has several different modes of operation; which one is used depends
10814 on the input and output formats:
10817 Hardware frame input, normal frame output
10819 Map the input frames to system memory and pass them to the output. If the
10820 original hardware frame is later required (for example, after overlaying
10821 something else on part of it), the @option{hwmap} filter can be used again
10822 in the next mode to retrieve it.
10824 Normal frame input, hardware frame output
10826 If the input is actually a software-mapped hardware frame, then unmap it -
10827 that is, return the original hardware frame.
10829 Otherwise, a device must be provided. Create new hardware surfaces on that
10830 device for the output, then map them back to the software format at the input
10831 and give those frames to the preceding filter. This will then act like the
10832 @option{hwupload} filter, but may be able to avoid an additional copy when
10833 the input is already in a compatible format.
10835 Hardware frame input and output
10837 A device must be supplied for the output, either directly or with the
10838 @option{derive_device} option. The input and output devices must be of
10839 different types and compatible - the exact meaning of this is
10840 system-dependent, but typically it means that they must refer to the same
10841 underlying hardware context (for example, refer to the same graphics card).
10843 If the input frames were originally created on the output device, then unmap
10844 to retrieve the original frames.
10846 Otherwise, map the frames to the output device - create new hardware frames
10847 on the output corresponding to the frames on the input.
10850 The following additional parameters are accepted:
10854 Set the frame mapping mode. Some combination of:
10857 The mapped frame should be readable.
10859 The mapped frame should be writeable.
10861 The mapping will always overwrite the entire frame.
10863 This may improve performance in some cases, as the original contents of the
10864 frame need not be loaded.
10866 The mapping must not involve any copying.
10868 Indirect mappings to copies of frames are created in some cases where either
10869 direct mapping is not possible or it would have unexpected properties.
10870 Setting this flag ensures that the mapping is direct and will fail if that is
10873 Defaults to @var{read+write} if not specified.
10875 @item derive_device @var{type}
10876 Rather than using the device supplied at initialisation, instead derive a new
10877 device of type @var{type} from the device the input frames exist on.
10880 In a hardware to hardware mapping, map in reverse - create frames in the sink
10881 and map them back to the source. This may be necessary in some cases where
10882 a mapping in one direction is required but only the opposite direction is
10883 supported by the devices being used.
10885 This option is dangerous - it may break the preceding filter in undefined
10886 ways if there are any additional constraints on that filter's output.
10887 Do not use it without fully understanding the implications of its use.
10893 Upload system memory frames to hardware surfaces.
10895 The device to upload to must be supplied when the filter is initialised. If
10896 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10899 @anchor{hwupload_cuda}
10900 @section hwupload_cuda
10902 Upload system memory frames to a CUDA device.
10904 It accepts the following optional parameters:
10908 The number of the CUDA device to use
10913 Apply a high-quality magnification filter designed for pixel art. This filter
10914 was originally created by Maxim Stepin.
10916 It accepts the following option:
10920 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10921 @code{hq3x} and @code{4} for @code{hq4x}.
10922 Default is @code{3}.
10926 Stack input videos horizontally.
10928 All streams must be of same pixel format and of same height.
10930 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10931 to create same output.
10933 The filter accept the following option:
10937 Set number of input streams. Default is 2.
10940 If set to 1, force the output to terminate when the shortest input
10941 terminates. Default value is 0.
10946 Modify the hue and/or the saturation of the input.
10948 It accepts the following parameters:
10952 Specify the hue angle as a number of degrees. It accepts an expression,
10953 and defaults to "0".
10956 Specify the saturation in the [-10,10] range. It accepts an expression and
10960 Specify the hue angle as a number of radians. It accepts an
10961 expression, and defaults to "0".
10964 Specify the brightness in the [-10,10] range. It accepts an expression and
10968 @option{h} and @option{H} are mutually exclusive, and can't be
10969 specified at the same time.
10971 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10972 expressions containing the following constants:
10976 frame count of the input frame starting from 0
10979 presentation timestamp of the input frame expressed in time base units
10982 frame rate of the input video, NAN if the input frame rate is unknown
10985 timestamp expressed in seconds, NAN if the input timestamp is unknown
10988 time base of the input video
10991 @subsection Examples
10995 Set the hue to 90 degrees and the saturation to 1.0:
11001 Same command but expressing the hue in radians:
11007 Rotate hue and make the saturation swing between 0
11008 and 2 over a period of 1 second:
11010 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11014 Apply a 3 seconds saturation fade-in effect starting at 0:
11016 hue="s=min(t/3\,1)"
11019 The general fade-in expression can be written as:
11021 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11025 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11027 hue="s=max(0\, min(1\, (8-t)/3))"
11030 The general fade-out expression can be written as:
11032 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11037 @subsection Commands
11039 This filter supports the following commands:
11045 Modify the hue and/or the saturation and/or brightness of the input video.
11046 The command accepts the same syntax of the corresponding option.
11048 If the specified expression is not valid, it is kept at its current
11052 @section hysteresis
11054 Grow first stream into second stream by connecting components.
11055 This makes it possible to build more robust edge masks.
11057 This filter accepts the following options:
11061 Set which planes will be processed as bitmap, unprocessed planes will be
11062 copied from first stream.
11063 By default value 0xf, all planes will be processed.
11066 Set threshold which is used in filtering. If pixel component value is higher than
11067 this value filter algorithm for connecting components is activated.
11068 By default value is 0.
11073 Detect video interlacing type.
11075 This filter tries to detect if the input frames are interlaced, progressive,
11076 top or bottom field first. It will also try to detect fields that are
11077 repeated between adjacent frames (a sign of telecine).
11079 Single frame detection considers only immediately adjacent frames when classifying each frame.
11080 Multiple frame detection incorporates the classification history of previous frames.
11082 The filter will log these metadata values:
11085 @item single.current_frame
11086 Detected type of current frame using single-frame detection. One of:
11087 ``tff'' (top field first), ``bff'' (bottom field first),
11088 ``progressive'', or ``undetermined''
11091 Cumulative number of frames detected as top field first using single-frame detection.
11094 Cumulative number of frames detected as top field first using multiple-frame detection.
11097 Cumulative number of frames detected as bottom field first using single-frame detection.
11099 @item multiple.current_frame
11100 Detected type of current frame using multiple-frame detection. One of:
11101 ``tff'' (top field first), ``bff'' (bottom field first),
11102 ``progressive'', or ``undetermined''
11105 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11107 @item single.progressive
11108 Cumulative number of frames detected as progressive using single-frame detection.
11110 @item multiple.progressive
11111 Cumulative number of frames detected as progressive using multiple-frame detection.
11113 @item single.undetermined
11114 Cumulative number of frames that could not be classified using single-frame detection.
11116 @item multiple.undetermined
11117 Cumulative number of frames that could not be classified using multiple-frame detection.
11119 @item repeated.current_frame
11120 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11122 @item repeated.neither
11123 Cumulative number of frames with no repeated field.
11126 Cumulative number of frames with the top field repeated from the previous frame's top field.
11128 @item repeated.bottom
11129 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11132 The filter accepts the following options:
11136 Set interlacing threshold.
11138 Set progressive threshold.
11140 Threshold for repeated field detection.
11142 Number of frames after which a given frame's contribution to the
11143 statistics is halved (i.e., it contributes only 0.5 to its
11144 classification). The default of 0 means that all frames seen are given
11145 full weight of 1.0 forever.
11146 @item analyze_interlaced_flag
11147 When this is not 0 then idet will use the specified number of frames to determine
11148 if the interlaced flag is accurate, it will not count undetermined frames.
11149 If the flag is found to be accurate it will be used without any further
11150 computations, if it is found to be inaccurate it will be cleared without any
11151 further computations. This allows inserting the idet filter as a low computational
11152 method to clean up the interlaced flag
11157 Deinterleave or interleave fields.
11159 This filter allows one to process interlaced images fields without
11160 deinterlacing them. Deinterleaving splits the input frame into 2
11161 fields (so called half pictures). Odd lines are moved to the top
11162 half of the output image, even lines to the bottom half.
11163 You can process (filter) them independently and then re-interleave them.
11165 The filter accepts the following options:
11169 @item chroma_mode, c
11170 @item alpha_mode, a
11171 Available values for @var{luma_mode}, @var{chroma_mode} and
11172 @var{alpha_mode} are:
11178 @item deinterleave, d
11179 Deinterleave fields, placing one above the other.
11181 @item interleave, i
11182 Interleave fields. Reverse the effect of deinterleaving.
11184 Default value is @code{none}.
11186 @item luma_swap, ls
11187 @item chroma_swap, cs
11188 @item alpha_swap, as
11189 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11194 Apply inflate effect to the video.
11196 This filter replaces the pixel by the local(3x3) average by taking into account
11197 only values higher than the pixel.
11199 It accepts the following options:
11206 Limit the maximum change for each plane, default is 65535.
11207 If 0, plane will remain unchanged.
11212 Simple interlacing filter from progressive contents. This interleaves upper (or
11213 lower) lines from odd frames with lower (or upper) lines from even frames,
11214 halving the frame rate and preserving image height.
11217 Original Original New Frame
11218 Frame 'j' Frame 'j+1' (tff)
11219 ========== =========== ==================
11220 Line 0 --------------------> Frame 'j' Line 0
11221 Line 1 Line 1 ----> Frame 'j+1' Line 1
11222 Line 2 ---------------------> Frame 'j' Line 2
11223 Line 3 Line 3 ----> Frame 'j+1' Line 3
11225 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11228 It accepts the following optional parameters:
11232 This determines whether the interlaced frame is taken from the even
11233 (tff - default) or odd (bff) lines of the progressive frame.
11236 Vertical lowpass filter to avoid twitter interlacing and
11237 reduce moire patterns.
11241 Disable vertical lowpass filter
11244 Enable linear filter (default)
11247 Enable complex filter. This will slightly less reduce twitter and moire
11248 but better retain detail and subjective sharpness impression.
11255 Deinterlace input video by applying Donald Graft's adaptive kernel
11256 deinterling. Work on interlaced parts of a video to produce
11257 progressive frames.
11259 The description of the accepted parameters follows.
11263 Set the threshold which affects the filter's tolerance when
11264 determining if a pixel line must be processed. It must be an integer
11265 in the range [0,255] and defaults to 10. A value of 0 will result in
11266 applying the process on every pixels.
11269 Paint pixels exceeding the threshold value to white if set to 1.
11273 Set the fields order. Swap fields if set to 1, leave fields alone if
11277 Enable additional sharpening if set to 1. Default is 0.
11280 Enable twoway sharpening if set to 1. Default is 0.
11283 @subsection Examples
11287 Apply default values:
11289 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11293 Enable additional sharpening:
11299 Paint processed pixels in white:
11305 @section lenscorrection
11307 Correct radial lens distortion
11309 This filter can be used to correct for radial distortion as can result from the use
11310 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11311 one can use tools available for example as part of opencv or simply trial-and-error.
11312 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11313 and extract the k1 and k2 coefficients from the resulting matrix.
11315 Note that effectively the same filter is available in the open-source tools Krita and
11316 Digikam from the KDE project.
11318 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11319 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11320 brightness distribution, so you may want to use both filters together in certain
11321 cases, though you will have to take care of ordering, i.e. whether vignetting should
11322 be applied before or after lens correction.
11324 @subsection Options
11326 The filter accepts the following options:
11330 Relative x-coordinate of the focal point of the image, and thereby the center of the
11331 distortion. This value has a range [0,1] and is expressed as fractions of the image
11332 width. Default is 0.5.
11334 Relative y-coordinate of the focal point of the image, and thereby the center of the
11335 distortion. This value has a range [0,1] and is expressed as fractions of the image
11336 height. Default is 0.5.
11338 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11339 no correction. Default is 0.
11341 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11342 0 means no correction. Default is 0.
11345 The formula that generates the correction is:
11347 @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)
11349 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11350 distances from the focal point in the source and target images, respectively.
11354 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11356 The @code{lensfun} filter requires the camera make, camera model, and lens model
11357 to apply the lens correction. The filter will load the lensfun database and
11358 query it to find the corresponding camera and lens entries in the database. As
11359 long as these entries can be found with the given options, the filter can
11360 perform corrections on frames. Note that incomplete strings will result in the
11361 filter choosing the best match with the given options, and the filter will
11362 output the chosen camera and lens models (logged with level "info"). You must
11363 provide the make, camera model, and lens model as they are required.
11365 The filter accepts the following options:
11369 The make of the camera (for example, "Canon"). This option is required.
11372 The model of the camera (for example, "Canon EOS 100D"). This option is
11376 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11377 option is required.
11380 The type of correction to apply. The following values are valid options:
11384 Enables fixing lens vignetting.
11387 Enables fixing lens geometry. This is the default.
11390 Enables fixing chromatic aberrations.
11393 Enables fixing lens vignetting and lens geometry.
11396 Enables fixing lens vignetting and chromatic aberrations.
11399 Enables fixing both lens geometry and chromatic aberrations.
11402 Enables all possible corrections.
11406 The focal length of the image/video (zoom; expected constant for video). For
11407 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11408 range should be chosen when using that lens. Default 18.
11411 The aperture of the image/video (expected constant for video). Note that
11412 aperture is only used for vignetting correction. Default 3.5.
11414 @item focus_distance
11415 The focus distance of the image/video (expected constant for video). Note that
11416 focus distance is only used for vignetting and only slightly affects the
11417 vignetting correction process. If unknown, leave it at the default value (which
11420 @item target_geometry
11421 The target geometry of the output image/video. The following values are valid
11425 @item rectilinear (default)
11428 @item equirectangular
11429 @item fisheye_orthographic
11430 @item fisheye_stereographic
11431 @item fisheye_equisolid
11432 @item fisheye_thoby
11435 Apply the reverse of image correction (instead of correcting distortion, apply
11438 @item interpolation
11439 The type of interpolation used when correcting distortion. The following values
11444 @item linear (default)
11449 @subsection Examples
11453 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11454 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11458 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
11462 Apply the same as before, but only for the first 5 seconds of video.
11465 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
11472 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11473 score between two input videos.
11475 The obtained VMAF score is printed through the logging system.
11477 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11478 After installing the library it can be enabled using:
11479 @code{./configure --enable-libvmaf --enable-version3}.
11480 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11482 The filter has following options:
11486 Set the model path which is to be used for SVM.
11487 Default value: @code{"vmaf_v0.6.1.pkl"}
11490 Set the file path to be used to store logs.
11493 Set the format of the log file (xml or json).
11495 @item enable_transform
11496 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11497 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11498 Default value: @code{false}
11501 Invokes the phone model which will generate VMAF scores higher than in the
11502 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11505 Enables computing psnr along with vmaf.
11508 Enables computing ssim along with vmaf.
11511 Enables computing ms_ssim along with vmaf.
11514 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11517 Set number of threads to be used when computing vmaf.
11520 Set interval for frame subsampling used when computing vmaf.
11522 @item enable_conf_interval
11523 Enables confidence interval.
11526 This filter also supports the @ref{framesync} options.
11528 On the below examples the input file @file{main.mpg} being processed is
11529 compared with the reference file @file{ref.mpg}.
11532 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11535 Example with options:
11537 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11542 Limits the pixel components values to the specified range [min, max].
11544 The filter accepts the following options:
11548 Lower bound. Defaults to the lowest allowed value for the input.
11551 Upper bound. Defaults to the highest allowed value for the input.
11554 Specify which planes will be processed. Defaults to all available.
11561 The filter accepts the following options:
11565 Set the number of loops. Setting this value to -1 will result in infinite loops.
11569 Set maximal size in number of frames. Default is 0.
11572 Set first frame of loop. Default is 0.
11575 @subsection Examples
11579 Loop single first frame infinitely:
11581 loop=loop=-1:size=1:start=0
11585 Loop single first frame 10 times:
11587 loop=loop=10:size=1:start=0
11591 Loop 10 first frames 5 times:
11593 loop=loop=5:size=10:start=0
11599 Apply a 1D LUT to an input video.
11601 The filter accepts the following options:
11605 Set the 1D LUT file name.
11607 Currently supported formats:
11614 Select interpolation mode.
11616 Available values are:
11620 Use values from the nearest defined point.
11622 Interpolate values using the linear interpolation.
11624 Interpolate values using the cosine interpolation.
11626 Interpolate values using the cubic interpolation.
11628 Interpolate values using the spline interpolation.
11635 Apply a 3D LUT to an input video.
11637 The filter accepts the following options:
11641 Set the 3D LUT file name.
11643 Currently supported formats:
11655 Select interpolation mode.
11657 Available values are:
11661 Use values from the nearest defined point.
11663 Interpolate values using the 8 points defining a cube.
11665 Interpolate values using a tetrahedron.
11669 This filter also supports the @ref{framesync} options.
11673 Turn certain luma values into transparency.
11675 The filter accepts the following options:
11679 Set the luma which will be used as base for transparency.
11680 Default value is @code{0}.
11683 Set the range of luma values to be keyed out.
11684 Default value is @code{0}.
11687 Set the range of softness. Default value is @code{0}.
11688 Use this to control gradual transition from zero to full transparency.
11691 @section lut, lutrgb, lutyuv
11693 Compute a look-up table for binding each pixel component input value
11694 to an output value, and apply it to the input video.
11696 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11697 to an RGB input video.
11699 These filters accept the following parameters:
11702 set first pixel component expression
11704 set second pixel component expression
11706 set third pixel component expression
11708 set fourth pixel component expression, corresponds to the alpha component
11711 set red component expression
11713 set green component expression
11715 set blue component expression
11717 alpha component expression
11720 set Y/luminance component expression
11722 set U/Cb component expression
11724 set V/Cr component expression
11727 Each of them specifies the expression to use for computing the lookup table for
11728 the corresponding pixel component values.
11730 The exact component associated to each of the @var{c*} options depends on the
11733 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11734 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11736 The expressions can contain the following constants and functions:
11741 The input width and height.
11744 The input value for the pixel component.
11747 The input value, clipped to the @var{minval}-@var{maxval} range.
11750 The maximum value for the pixel component.
11753 The minimum value for the pixel component.
11756 The negated value for the pixel component value, clipped to the
11757 @var{minval}-@var{maxval} range; it corresponds to the expression
11758 "maxval-clipval+minval".
11761 The computed value in @var{val}, clipped to the
11762 @var{minval}-@var{maxval} range.
11764 @item gammaval(gamma)
11765 The computed gamma correction value of the pixel component value,
11766 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11768 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11772 All expressions default to "val".
11774 @subsection Examples
11778 Negate input video:
11780 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11781 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11784 The above is the same as:
11786 lutrgb="r=negval:g=negval:b=negval"
11787 lutyuv="y=negval:u=negval:v=negval"
11797 Remove chroma components, turning the video into a graytone image:
11799 lutyuv="u=128:v=128"
11803 Apply a luma burning effect:
11809 Remove green and blue components:
11815 Set a constant alpha channel value on input:
11817 format=rgba,lutrgb=a="maxval-minval/2"
11821 Correct luminance gamma by a factor of 0.5:
11823 lutyuv=y=gammaval(0.5)
11827 Discard least significant bits of luma:
11829 lutyuv=y='bitand(val, 128+64+32)'
11833 Technicolor like effect:
11835 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11839 @section lut2, tlut2
11841 The @code{lut2} filter takes two input streams and outputs one
11844 The @code{tlut2} (time lut2) filter takes two consecutive frames
11845 from one single stream.
11847 This filter accepts the following parameters:
11850 set first pixel component expression
11852 set second pixel component expression
11854 set third pixel component expression
11856 set fourth pixel component expression, corresponds to the alpha component
11859 set output bit depth, only available for @code{lut2} filter. By default is 0,
11860 which means bit depth is automatically picked from first input format.
11863 Each of them specifies the expression to use for computing the lookup table for
11864 the corresponding pixel component values.
11866 The exact component associated to each of the @var{c*} options depends on the
11869 The expressions can contain the following constants:
11874 The input width and height.
11877 The first input value for the pixel component.
11880 The second input value for the pixel component.
11883 The first input video bit depth.
11886 The second input video bit depth.
11889 All expressions default to "x".
11891 @subsection Examples
11895 Highlight differences between two RGB video streams:
11897 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)'
11901 Highlight differences between two YUV video streams:
11903 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)'
11907 Show max difference between two video streams:
11909 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)))'
11913 @section maskedclamp
11915 Clamp the first input stream with the second input and third input stream.
11917 Returns the value of first stream to be between second input
11918 stream - @code{undershoot} and third input stream + @code{overshoot}.
11920 This filter accepts the following options:
11923 Default value is @code{0}.
11926 Default value is @code{0}.
11929 Set which planes will be processed as bitmap, unprocessed planes will be
11930 copied from first stream.
11931 By default value 0xf, all planes will be processed.
11934 @section maskedmerge
11936 Merge the first input stream with the second input stream using per pixel
11937 weights in the third input stream.
11939 A value of 0 in the third stream pixel component means that pixel component
11940 from first stream is returned unchanged, while maximum value (eg. 255 for
11941 8-bit videos) means that pixel component from second stream is returned
11942 unchanged. Intermediate values define the amount of merging between both
11943 input stream's pixel components.
11945 This filter accepts the following options:
11948 Set which planes will be processed as bitmap, unprocessed planes will be
11949 copied from first stream.
11950 By default value 0xf, all planes will be processed.
11954 Create mask from input video.
11956 For example it is useful to create motion masks after @code{tblend} filter.
11958 This filter accepts the following options:
11962 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
11965 Set high threshold. Any pixel component higher than this value will be set to max value
11966 allowed for current pixel format.
11969 Set planes to filter, by default all available planes are filtered.
11972 Fill all frame pixels with this value.
11975 Set max average pixel value for frame. If sum of all pixel components is higher that this
11976 average, output frame will be completely filled with value set by @var{fill} option.
11977 Typically useful for scene changes when used in combination with @code{tblend} filter.
11982 Apply motion-compensation deinterlacing.
11984 It needs one field per frame as input and must thus be used together
11985 with yadif=1/3 or equivalent.
11987 This filter accepts the following options:
11990 Set the deinterlacing mode.
11992 It accepts one of the following values:
11997 use iterative motion estimation
11999 like @samp{slow}, but use multiple reference frames.
12001 Default value is @samp{fast}.
12004 Set the picture field parity assumed for the input video. It must be
12005 one of the following values:
12009 assume top field first
12011 assume bottom field first
12014 Default value is @samp{bff}.
12017 Set per-block quantization parameter (QP) used by the internal
12020 Higher values should result in a smoother motion vector field but less
12021 optimal individual vectors. Default value is 1.
12024 @section mergeplanes
12026 Merge color channel components from several video streams.
12028 The filter accepts up to 4 input streams, and merge selected input
12029 planes to the output video.
12031 This filter accepts the following options:
12034 Set input to output plane mapping. Default is @code{0}.
12036 The mappings is specified as a bitmap. It should be specified as a
12037 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12038 mapping for the first plane of the output stream. 'A' sets the number of
12039 the input stream to use (from 0 to 3), and 'a' the plane number of the
12040 corresponding input to use (from 0 to 3). The rest of the mappings is
12041 similar, 'Bb' describes the mapping for the output stream second
12042 plane, 'Cc' describes the mapping for the output stream third plane and
12043 'Dd' describes the mapping for the output stream fourth plane.
12046 Set output pixel format. Default is @code{yuva444p}.
12049 @subsection Examples
12053 Merge three gray video streams of same width and height into single video stream:
12055 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12059 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12061 [a0][a1]mergeplanes=0x00010210:yuva444p
12065 Swap Y and A plane in yuva444p stream:
12067 format=yuva444p,mergeplanes=0x03010200:yuva444p
12071 Swap U and V plane in yuv420p stream:
12073 format=yuv420p,mergeplanes=0x000201:yuv420p
12077 Cast a rgb24 clip to yuv444p:
12079 format=rgb24,mergeplanes=0x000102:yuv444p
12085 Estimate and export motion vectors using block matching algorithms.
12086 Motion vectors are stored in frame side data to be used by other filters.
12088 This filter accepts the following options:
12091 Specify the motion estimation method. Accepts one of the following values:
12095 Exhaustive search algorithm.
12097 Three step search algorithm.
12099 Two dimensional logarithmic search algorithm.
12101 New three step search algorithm.
12103 Four step search algorithm.
12105 Diamond search algorithm.
12107 Hexagon-based search algorithm.
12109 Enhanced predictive zonal search algorithm.
12111 Uneven multi-hexagon search algorithm.
12113 Default value is @samp{esa}.
12116 Macroblock size. Default @code{16}.
12119 Search parameter. Default @code{7}.
12122 @section midequalizer
12124 Apply Midway Image Equalization effect using two video streams.
12126 Midway Image Equalization adjusts a pair of images to have the same
12127 histogram, while maintaining their dynamics as much as possible. It's
12128 useful for e.g. matching exposures from a pair of stereo cameras.
12130 This filter has two inputs and one output, which must be of same pixel format, but
12131 may be of different sizes. The output of filter is first input adjusted with
12132 midway histogram of both inputs.
12134 This filter accepts the following option:
12138 Set which planes to process. Default is @code{15}, which is all available planes.
12141 @section minterpolate
12143 Convert the video to specified frame rate using motion interpolation.
12145 This filter accepts the following options:
12148 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}.
12151 Motion interpolation mode. Following values are accepted:
12154 Duplicate previous or next frame for interpolating new ones.
12156 Blend source frames. Interpolated frame is mean of previous and next frames.
12158 Motion compensated interpolation. Following options are effective when this mode is selected:
12162 Motion compensation mode. Following values are accepted:
12165 Overlapped block motion compensation.
12167 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12169 Default mode is @samp{obmc}.
12172 Motion estimation mode. Following values are accepted:
12175 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12177 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12179 Default mode is @samp{bilat}.
12182 The algorithm to be used for motion estimation. Following values are accepted:
12185 Exhaustive search algorithm.
12187 Three step search algorithm.
12189 Two dimensional logarithmic search algorithm.
12191 New three step search algorithm.
12193 Four step search algorithm.
12195 Diamond search algorithm.
12197 Hexagon-based search algorithm.
12199 Enhanced predictive zonal search algorithm.
12201 Uneven multi-hexagon search algorithm.
12203 Default algorithm is @samp{epzs}.
12206 Macroblock size. Default @code{16}.
12209 Motion estimation search parameter. Default @code{32}.
12212 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).
12217 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:
12220 Disable scene change detection.
12222 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12224 Default method is @samp{fdiff}.
12226 @item scd_threshold
12227 Scene change detection threshold. Default is @code{5.0}.
12232 Mix several video input streams into one video stream.
12234 A description of the accepted options follows.
12238 The number of inputs. If unspecified, it defaults to 2.
12241 Specify weight of each input video stream as sequence.
12242 Each weight is separated by space. If number of weights
12243 is smaller than number of @var{frames} last specified
12244 weight will be used for all remaining unset weights.
12247 Specify scale, if it is set it will be multiplied with sum
12248 of each weight multiplied with pixel values to give final destination
12249 pixel value. By default @var{scale} is auto scaled to sum of weights.
12252 Specify how end of stream is determined.
12255 The duration of the longest input. (default)
12258 The duration of the shortest input.
12261 The duration of the first input.
12265 @section mpdecimate
12267 Drop frames that do not differ greatly from the previous frame in
12268 order to reduce frame rate.
12270 The main use of this filter is for very-low-bitrate encoding
12271 (e.g. streaming over dialup modem), but it could in theory be used for
12272 fixing movies that were inverse-telecined incorrectly.
12274 A description of the accepted options follows.
12278 Set the maximum number of consecutive frames which can be dropped (if
12279 positive), or the minimum interval between dropped frames (if
12280 negative). If the value is 0, the frame is dropped disregarding the
12281 number of previous sequentially dropped frames.
12283 Default value is 0.
12288 Set the dropping threshold values.
12290 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12291 represent actual pixel value differences, so a threshold of 64
12292 corresponds to 1 unit of difference for each pixel, or the same spread
12293 out differently over the block.
12295 A frame is a candidate for dropping if no 8x8 blocks differ by more
12296 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12297 meaning the whole image) differ by more than a threshold of @option{lo}.
12299 Default value for @option{hi} is 64*12, default value for @option{lo} is
12300 64*5, and default value for @option{frac} is 0.33.
12306 Negate (invert) the input video.
12308 It accepts the following option:
12313 With value 1, it negates the alpha component, if present. Default value is 0.
12319 Denoise frames using Non-Local Means algorithm.
12321 Each pixel is adjusted by looking for other pixels with similar contexts. This
12322 context similarity is defined by comparing their surrounding patches of size
12323 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12326 Note that the research area defines centers for patches, which means some
12327 patches will be made of pixels outside that research area.
12329 The filter accepts the following options.
12333 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12336 Set patch size. Default is 7. Must be odd number in range [0, 99].
12339 Same as @option{p} but for chroma planes.
12341 The default value is @var{0} and means automatic.
12344 Set research size. Default is 15. Must be odd number in range [0, 99].
12347 Same as @option{r} but for chroma planes.
12349 The default value is @var{0} and means automatic.
12354 Deinterlace video using neural network edge directed interpolation.
12356 This filter accepts the following options:
12360 Mandatory option, without binary file filter can not work.
12361 Currently file can be found here:
12362 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12365 Set which frames to deinterlace, by default it is @code{all}.
12366 Can be @code{all} or @code{interlaced}.
12369 Set mode of operation.
12371 Can be one of the following:
12375 Use frame flags, both fields.
12377 Use frame flags, single field.
12379 Use top field only.
12381 Use bottom field only.
12383 Use both fields, top first.
12385 Use both fields, bottom first.
12389 Set which planes to process, by default filter process all frames.
12392 Set size of local neighborhood around each pixel, used by the predictor neural
12395 Can be one of the following:
12408 Set the number of neurons in predictor neural network.
12409 Can be one of the following:
12420 Controls the number of different neural network predictions that are blended
12421 together to compute the final output value. Can be @code{fast}, default or
12425 Set which set of weights to use in the predictor.
12426 Can be one of the following:
12430 weights trained to minimize absolute error
12432 weights trained to minimize squared error
12436 Controls whether or not the prescreener neural network is used to decide
12437 which pixels should be processed by the predictor neural network and which
12438 can be handled by simple cubic interpolation.
12439 The prescreener is trained to know whether cubic interpolation will be
12440 sufficient for a pixel or whether it should be predicted by the predictor nn.
12441 The computational complexity of the prescreener nn is much less than that of
12442 the predictor nn. Since most pixels can be handled by cubic interpolation,
12443 using the prescreener generally results in much faster processing.
12444 The prescreener is pretty accurate, so the difference between using it and not
12445 using it is almost always unnoticeable.
12447 Can be one of the following:
12455 Default is @code{new}.
12458 Set various debugging flags.
12463 Force libavfilter not to use any of the specified pixel formats for the
12464 input to the next filter.
12466 It accepts the following parameters:
12470 A '|'-separated list of pixel format names, such as
12471 pix_fmts=yuv420p|monow|rgb24".
12475 @subsection Examples
12479 Force libavfilter to use a format different from @var{yuv420p} for the
12480 input to the vflip filter:
12482 noformat=pix_fmts=yuv420p,vflip
12486 Convert the input video to any of the formats not contained in the list:
12488 noformat=yuv420p|yuv444p|yuv410p
12494 Add noise on video input frame.
12496 The filter accepts the following options:
12504 Set noise seed for specific pixel component or all pixel components in case
12505 of @var{all_seed}. Default value is @code{123457}.
12507 @item all_strength, alls
12508 @item c0_strength, c0s
12509 @item c1_strength, c1s
12510 @item c2_strength, c2s
12511 @item c3_strength, c3s
12512 Set noise strength for specific pixel component or all pixel components in case
12513 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12515 @item all_flags, allf
12516 @item c0_flags, c0f
12517 @item c1_flags, c1f
12518 @item c2_flags, c2f
12519 @item c3_flags, c3f
12520 Set pixel component flags or set flags for all components if @var{all_flags}.
12521 Available values for component flags are:
12524 averaged temporal noise (smoother)
12526 mix random noise with a (semi)regular pattern
12528 temporal noise (noise pattern changes between frames)
12530 uniform noise (gaussian otherwise)
12534 @subsection Examples
12536 Add temporal and uniform noise to input video:
12538 noise=alls=20:allf=t+u
12543 Normalize RGB video (aka histogram stretching, contrast stretching).
12544 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12546 For each channel of each frame, the filter computes the input range and maps
12547 it linearly to the user-specified output range. The output range defaults
12548 to the full dynamic range from pure black to pure white.
12550 Temporal smoothing can be used on the input range to reduce flickering (rapid
12551 changes in brightness) caused when small dark or bright objects enter or leave
12552 the scene. This is similar to the auto-exposure (automatic gain control) on a
12553 video camera, and, like a video camera, it may cause a period of over- or
12554 under-exposure of the video.
12556 The R,G,B channels can be normalized independently, which may cause some
12557 color shifting, or linked together as a single channel, which prevents
12558 color shifting. Linked normalization preserves hue. Independent normalization
12559 does not, so it can be used to remove some color casts. Independent and linked
12560 normalization can be combined in any ratio.
12562 The normalize filter accepts the following options:
12567 Colors which define the output range. The minimum input value is mapped to
12568 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12569 The defaults are black and white respectively. Specifying white for
12570 @var{blackpt} and black for @var{whitept} will give color-inverted,
12571 normalized video. Shades of grey can be used to reduce the dynamic range
12572 (contrast). Specifying saturated colors here can create some interesting
12576 The number of previous frames to use for temporal smoothing. The input range
12577 of each channel is smoothed using a rolling average over the current frame
12578 and the @var{smoothing} previous frames. The default is 0 (no temporal
12582 Controls the ratio of independent (color shifting) channel normalization to
12583 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12584 independent. Defaults to 1.0 (fully independent).
12587 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12588 expensive no-op. Defaults to 1.0 (full strength).
12592 @subsection Examples
12594 Stretch video contrast to use the full dynamic range, with no temporal
12595 smoothing; may flicker depending on the source content:
12597 normalize=blackpt=black:whitept=white:smoothing=0
12600 As above, but with 50 frames of temporal smoothing; flicker should be
12601 reduced, depending on the source content:
12603 normalize=blackpt=black:whitept=white:smoothing=50
12606 As above, but with hue-preserving linked channel normalization:
12608 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12611 As above, but with half strength:
12613 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12616 Map the darkest input color to red, the brightest input color to cyan:
12618 normalize=blackpt=red:whitept=cyan
12623 Pass the video source unchanged to the output.
12626 Optical Character Recognition
12628 This filter uses Tesseract for optical character recognition. To enable
12629 compilation of this filter, you need to configure FFmpeg with
12630 @code{--enable-libtesseract}.
12632 It accepts the following options:
12636 Set datapath to tesseract data. Default is to use whatever was
12637 set at installation.
12640 Set language, default is "eng".
12643 Set character whitelist.
12646 Set character blacklist.
12649 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12653 Apply a video transform using libopencv.
12655 To enable this filter, install the libopencv library and headers and
12656 configure FFmpeg with @code{--enable-libopencv}.
12658 It accepts the following parameters:
12663 The name of the libopencv filter to apply.
12665 @item filter_params
12666 The parameters to pass to the libopencv filter. If not specified, the default
12667 values are assumed.
12671 Refer to the official libopencv documentation for more precise
12673 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12675 Several libopencv filters are supported; see the following subsections.
12680 Dilate an image by using a specific structuring element.
12681 It corresponds to the libopencv function @code{cvDilate}.
12683 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12685 @var{struct_el} represents a structuring element, and has the syntax:
12686 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12688 @var{cols} and @var{rows} represent the number of columns and rows of
12689 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12690 point, and @var{shape} the shape for the structuring element. @var{shape}
12691 must be "rect", "cross", "ellipse", or "custom".
12693 If the value for @var{shape} is "custom", it must be followed by a
12694 string of the form "=@var{filename}". The file with name
12695 @var{filename} is assumed to represent a binary image, with each
12696 printable character corresponding to a bright pixel. When a custom
12697 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12698 or columns and rows of the read file are assumed instead.
12700 The default value for @var{struct_el} is "3x3+0x0/rect".
12702 @var{nb_iterations} specifies the number of times the transform is
12703 applied to the image, and defaults to 1.
12707 # Use the default values
12710 # Dilate using a structuring element with a 5x5 cross, iterating two times
12711 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12713 # Read the shape from the file diamond.shape, iterating two times.
12714 # The file diamond.shape may contain a pattern of characters like this
12720 # The specified columns and rows are ignored
12721 # but the anchor point coordinates are not
12722 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12727 Erode an image by using a specific structuring element.
12728 It corresponds to the libopencv function @code{cvErode}.
12730 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12731 with the same syntax and semantics as the @ref{dilate} filter.
12735 Smooth the input video.
12737 The filter takes the following parameters:
12738 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12740 @var{type} is the type of smooth filter to apply, and must be one of
12741 the following values: "blur", "blur_no_scale", "median", "gaussian",
12742 or "bilateral". The default value is "gaussian".
12744 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12745 depend on the smooth type. @var{param1} and
12746 @var{param2} accept integer positive values or 0. @var{param3} and
12747 @var{param4} accept floating point values.
12749 The default value for @var{param1} is 3. The default value for the
12750 other parameters is 0.
12752 These parameters correspond to the parameters assigned to the
12753 libopencv function @code{cvSmooth}.
12755 @section oscilloscope
12757 2D Video Oscilloscope.
12759 Useful to measure spatial impulse, step responses, chroma delays, etc.
12761 It accepts the following parameters:
12765 Set scope center x position.
12768 Set scope center y position.
12771 Set scope size, relative to frame diagonal.
12774 Set scope tilt/rotation.
12780 Set trace center x position.
12783 Set trace center y position.
12786 Set trace width, relative to width of frame.
12789 Set trace height, relative to height of frame.
12792 Set which components to trace. By default it traces first three components.
12795 Draw trace grid. By default is enabled.
12798 Draw some statistics. By default is enabled.
12801 Draw scope. By default is enabled.
12804 @subsection Examples
12808 Inspect full first row of video frame.
12810 oscilloscope=x=0.5:y=0:s=1
12814 Inspect full last row of video frame.
12816 oscilloscope=x=0.5:y=1:s=1
12820 Inspect full 5th line of video frame of height 1080.
12822 oscilloscope=x=0.5:y=5/1080:s=1
12826 Inspect full last column of video frame.
12828 oscilloscope=x=1:y=0.5:s=1:t=1
12836 Overlay one video on top of another.
12838 It takes two inputs and has one output. The first input is the "main"
12839 video on which the second input is overlaid.
12841 It accepts the following parameters:
12843 A description of the accepted options follows.
12848 Set the expression for the x and y coordinates of the overlaid video
12849 on the main video. Default value is "0" for both expressions. In case
12850 the expression is invalid, it is set to a huge value (meaning that the
12851 overlay will not be displayed within the output visible area).
12854 See @ref{framesync}.
12857 Set when the expressions for @option{x}, and @option{y} are evaluated.
12859 It accepts the following values:
12862 only evaluate expressions once during the filter initialization or
12863 when a command is processed
12866 evaluate expressions for each incoming frame
12869 Default value is @samp{frame}.
12872 See @ref{framesync}.
12875 Set the format for the output video.
12877 It accepts the following values:
12880 force YUV420 output
12883 force YUV422 output
12886 force YUV444 output
12889 force packed RGB output
12892 force planar RGB output
12895 automatically pick format
12898 Default value is @samp{yuv420}.
12901 See @ref{framesync}.
12904 Set format of alpha of the overlaid video, it can be @var{straight} or
12905 @var{premultiplied}. Default is @var{straight}.
12908 The @option{x}, and @option{y} expressions can contain the following
12914 The main input width and height.
12918 The overlay input width and height.
12922 The computed values for @var{x} and @var{y}. They are evaluated for
12927 horizontal and vertical chroma subsample values of the output
12928 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12932 the number of input frame, starting from 0
12935 the position in the file of the input frame, NAN if unknown
12938 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12942 This filter also supports the @ref{framesync} options.
12944 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12945 when evaluation is done @emph{per frame}, and will evaluate to NAN
12946 when @option{eval} is set to @samp{init}.
12948 Be aware that frames are taken from each input video in timestamp
12949 order, hence, if their initial timestamps differ, it is a good idea
12950 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12951 have them begin in the same zero timestamp, as the example for
12952 the @var{movie} filter does.
12954 You can chain together more overlays but you should test the
12955 efficiency of such approach.
12957 @subsection Commands
12959 This filter supports the following commands:
12963 Modify the x and y of the overlay input.
12964 The command accepts the same syntax of the corresponding option.
12966 If the specified expression is not valid, it is kept at its current
12970 @subsection Examples
12974 Draw the overlay at 10 pixels from the bottom right corner of the main
12977 overlay=main_w-overlay_w-10:main_h-overlay_h-10
12980 Using named options the example above becomes:
12982 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
12986 Insert a transparent PNG logo in the bottom left corner of the input,
12987 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12989 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12993 Insert 2 different transparent PNG logos (second logo on bottom
12994 right corner) using the @command{ffmpeg} tool:
12996 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
13000 Add a transparent color layer on top of the main video; @code{WxH}
13001 must specify the size of the main input to the overlay filter:
13003 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13007 Play an original video and a filtered version (here with the deshake
13008 filter) side by side using the @command{ffplay} tool:
13010 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13013 The above command is the same as:
13015 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13019 Make a sliding overlay appearing from the left to the right top part of the
13020 screen starting since time 2:
13022 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13026 Compose output by putting two input videos side to side:
13028 ffmpeg -i left.avi -i right.avi -filter_complex "
13029 nullsrc=size=200x100 [background];
13030 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13031 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13032 [background][left] overlay=shortest=1 [background+left];
13033 [background+left][right] overlay=shortest=1:x=100 [left+right]
13038 Mask 10-20 seconds of a video by applying the delogo filter to a section
13040 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13041 -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]'
13046 Chain several overlays in cascade:
13048 nullsrc=s=200x200 [bg];
13049 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13050 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13051 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13052 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13053 [in3] null, [mid2] overlay=100:100 [out0]
13060 Apply Overcomplete Wavelet denoiser.
13062 The filter accepts the following options:
13068 Larger depth values will denoise lower frequency components more, but
13069 slow down filtering.
13071 Must be an int in the range 8-16, default is @code{8}.
13073 @item luma_strength, ls
13076 Must be a double value in the range 0-1000, default is @code{1.0}.
13078 @item chroma_strength, cs
13079 Set chroma strength.
13081 Must be a double value in the range 0-1000, default is @code{1.0}.
13087 Add paddings to the input image, and place the original input at the
13088 provided @var{x}, @var{y} coordinates.
13090 It accepts the following parameters:
13095 Specify an expression for the size of the output image with the
13096 paddings added. If the value for @var{width} or @var{height} is 0, the
13097 corresponding input size is used for the output.
13099 The @var{width} expression can reference the value set by the
13100 @var{height} expression, and vice versa.
13102 The default value of @var{width} and @var{height} is 0.
13106 Specify the offsets to place the input image at within the padded area,
13107 with respect to the top/left border of the output image.
13109 The @var{x} expression can reference the value set by the @var{y}
13110 expression, and vice versa.
13112 The default value of @var{x} and @var{y} is 0.
13114 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13115 so the input image is centered on the padded area.
13118 Specify the color of the padded area. For the syntax of this option,
13119 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13120 manual,ffmpeg-utils}.
13122 The default value of @var{color} is "black".
13125 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13127 It accepts the following values:
13131 Only evaluate expressions once during the filter initialization or when
13132 a command is processed.
13135 Evaluate expressions for each incoming frame.
13139 Default value is @samp{init}.
13142 Pad to aspect instead to a resolution.
13146 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13147 options are expressions containing the following constants:
13152 The input video width and height.
13156 These are the same as @var{in_w} and @var{in_h}.
13160 The output width and height (the size of the padded area), as
13161 specified by the @var{width} and @var{height} expressions.
13165 These are the same as @var{out_w} and @var{out_h}.
13169 The x and y offsets as specified by the @var{x} and @var{y}
13170 expressions, or NAN if not yet specified.
13173 same as @var{iw} / @var{ih}
13176 input sample aspect ratio
13179 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13183 The horizontal and vertical chroma subsample values. For example for the
13184 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13187 @subsection Examples
13191 Add paddings with the color "violet" to the input video. The output video
13192 size is 640x480, and the top-left corner of the input video is placed at
13195 pad=640:480:0:40:violet
13198 The example above is equivalent to the following command:
13200 pad=width=640:height=480:x=0:y=40:color=violet
13204 Pad the input to get an output with dimensions increased by 3/2,
13205 and put the input video at the center of the padded area:
13207 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13211 Pad the input to get a squared output with size equal to the maximum
13212 value between the input width and height, and put the input video at
13213 the center of the padded area:
13215 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13219 Pad the input to get a final w/h ratio of 16:9:
13221 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13225 In case of anamorphic video, in order to set the output display aspect
13226 correctly, it is necessary to use @var{sar} in the expression,
13227 according to the relation:
13229 (ih * X / ih) * sar = output_dar
13230 X = output_dar / sar
13233 Thus the previous example needs to be modified to:
13235 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13239 Double the output size and put the input video in the bottom-right
13240 corner of the output padded area:
13242 pad="2*iw:2*ih:ow-iw:oh-ih"
13246 @anchor{palettegen}
13247 @section palettegen
13249 Generate one palette for a whole video stream.
13251 It accepts the following options:
13255 Set the maximum number of colors to quantize in the palette.
13256 Note: the palette will still contain 256 colors; the unused palette entries
13259 @item reserve_transparent
13260 Create a palette of 255 colors maximum and reserve the last one for
13261 transparency. Reserving the transparency color is useful for GIF optimization.
13262 If not set, the maximum of colors in the palette will be 256. You probably want
13263 to disable this option for a standalone image.
13266 @item transparency_color
13267 Set the color that will be used as background for transparency.
13270 Set statistics mode.
13272 It accepts the following values:
13275 Compute full frame histograms.
13277 Compute histograms only for the part that differs from previous frame. This
13278 might be relevant to give more importance to the moving part of your input if
13279 the background is static.
13281 Compute new histogram for each frame.
13284 Default value is @var{full}.
13287 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13288 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13289 color quantization of the palette. This information is also visible at
13290 @var{info} logging level.
13292 @subsection Examples
13296 Generate a representative palette of a given video using @command{ffmpeg}:
13298 ffmpeg -i input.mkv -vf palettegen palette.png
13302 @section paletteuse
13304 Use a palette to downsample an input video stream.
13306 The filter takes two inputs: one video stream and a palette. The palette must
13307 be a 256 pixels image.
13309 It accepts the following options:
13313 Select dithering mode. Available algorithms are:
13316 Ordered 8x8 bayer dithering (deterministic)
13318 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13319 Note: this dithering is sometimes considered "wrong" and is included as a
13321 @item floyd_steinberg
13322 Floyd and Steingberg dithering (error diffusion)
13324 Frankie Sierra dithering v2 (error diffusion)
13326 Frankie Sierra dithering v2 "Lite" (error diffusion)
13329 Default is @var{sierra2_4a}.
13332 When @var{bayer} dithering is selected, this option defines the scale of the
13333 pattern (how much the crosshatch pattern is visible). A low value means more
13334 visible pattern for less banding, and higher value means less visible pattern
13335 at the cost of more banding.
13337 The option must be an integer value in the range [0,5]. Default is @var{2}.
13340 If set, define the zone to process
13344 Only the changing rectangle will be reprocessed. This is similar to GIF
13345 cropping/offsetting compression mechanism. This option can be useful for speed
13346 if only a part of the image is changing, and has use cases such as limiting the
13347 scope of the error diffusal @option{dither} to the rectangle that bounds the
13348 moving scene (it leads to more deterministic output if the scene doesn't change
13349 much, and as a result less moving noise and better GIF compression).
13352 Default is @var{none}.
13355 Take new palette for each output frame.
13357 @item alpha_threshold
13358 Sets the alpha threshold for transparency. Alpha values above this threshold
13359 will be treated as completely opaque, and values below this threshold will be
13360 treated as completely transparent.
13362 The option must be an integer value in the range [0,255]. Default is @var{128}.
13365 @subsection Examples
13369 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13370 using @command{ffmpeg}:
13372 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13376 @section perspective
13378 Correct perspective of video not recorded perpendicular to the screen.
13380 A description of the accepted parameters follows.
13391 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13392 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13393 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13394 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13395 then the corners of the source will be sent to the specified coordinates.
13397 The expressions can use the following variables:
13402 the width and height of video frame.
13406 Output frame count.
13409 @item interpolation
13410 Set interpolation for perspective correction.
13412 It accepts the following values:
13418 Default value is @samp{linear}.
13421 Set interpretation of coordinate options.
13423 It accepts the following values:
13427 Send point in the source specified by the given coordinates to
13428 the corners of the destination.
13430 @item 1, destination
13432 Send the corners of the source to the point in the destination specified
13433 by the given coordinates.
13435 Default value is @samp{source}.
13439 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13441 It accepts the following values:
13444 only evaluate expressions once during the filter initialization or
13445 when a command is processed
13448 evaluate expressions for each incoming frame
13451 Default value is @samp{init}.
13456 Delay interlaced video by one field time so that the field order changes.
13458 The intended use is to fix PAL movies that have been captured with the
13459 opposite field order to the film-to-video transfer.
13461 A description of the accepted parameters follows.
13467 It accepts the following values:
13470 Capture field order top-first, transfer bottom-first.
13471 Filter will delay the bottom field.
13474 Capture field order bottom-first, transfer top-first.
13475 Filter will delay the top field.
13478 Capture and transfer with the same field order. This mode only exists
13479 for the documentation of the other options to refer to, but if you
13480 actually select it, the filter will faithfully do nothing.
13483 Capture field order determined automatically by field flags, transfer
13485 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13486 basis using field flags. If no field information is available,
13487 then this works just like @samp{u}.
13490 Capture unknown or varying, transfer opposite.
13491 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13492 analyzing the images and selecting the alternative that produces best
13493 match between the fields.
13496 Capture top-first, transfer unknown or varying.
13497 Filter selects among @samp{t} and @samp{p} using image analysis.
13500 Capture bottom-first, transfer unknown or varying.
13501 Filter selects among @samp{b} and @samp{p} using image analysis.
13504 Capture determined by field flags, transfer unknown or varying.
13505 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13506 image analysis. If no field information is available, then this works just
13507 like @samp{U}. This is the default mode.
13510 Both capture and transfer unknown or varying.
13511 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13515 @section pixdesctest
13517 Pixel format descriptor test filter, mainly useful for internal
13518 testing. The output video should be equal to the input video.
13522 format=monow, pixdesctest
13525 can be used to test the monowhite pixel format descriptor definition.
13529 Display sample values of color channels. Mainly useful for checking color
13530 and levels. Minimum supported resolution is 640x480.
13532 The filters accept the following options:
13536 Set scope X position, relative offset on X axis.
13539 Set scope Y position, relative offset on Y axis.
13548 Set window opacity. This window also holds statistics about pixel area.
13551 Set window X position, relative offset on X axis.
13554 Set window Y position, relative offset on Y axis.
13559 Enable the specified chain of postprocessing subfilters using libpostproc. This
13560 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13561 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13562 Each subfilter and some options have a short and a long name that can be used
13563 interchangeably, i.e. dr/dering are the same.
13565 The filters accept the following options:
13569 Set postprocessing subfilters string.
13572 All subfilters share common options to determine their scope:
13576 Honor the quality commands for this subfilter.
13579 Do chrominance filtering, too (default).
13582 Do luminance filtering only (no chrominance).
13585 Do chrominance filtering only (no luminance).
13588 These options can be appended after the subfilter name, separated by a '|'.
13590 Available subfilters are:
13593 @item hb/hdeblock[|difference[|flatness]]
13594 Horizontal deblocking filter
13597 Difference factor where higher values mean more deblocking (default: @code{32}).
13599 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13602 @item vb/vdeblock[|difference[|flatness]]
13603 Vertical deblocking filter
13606 Difference factor where higher values mean more deblocking (default: @code{32}).
13608 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13611 @item ha/hadeblock[|difference[|flatness]]
13612 Accurate horizontal deblocking filter
13615 Difference factor where higher values mean more deblocking (default: @code{32}).
13617 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13620 @item va/vadeblock[|difference[|flatness]]
13621 Accurate vertical deblocking filter
13624 Difference factor where higher values mean more deblocking (default: @code{32}).
13626 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13630 The horizontal and vertical deblocking filters share the difference and
13631 flatness values so you cannot set different horizontal and vertical
13635 @item h1/x1hdeblock
13636 Experimental horizontal deblocking filter
13638 @item v1/x1vdeblock
13639 Experimental vertical deblocking filter
13644 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13647 larger -> stronger filtering
13649 larger -> stronger filtering
13651 larger -> stronger filtering
13654 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13657 Stretch luminance to @code{0-255}.
13660 @item lb/linblenddeint
13661 Linear blend deinterlacing filter that deinterlaces the given block by
13662 filtering all lines with a @code{(1 2 1)} filter.
13664 @item li/linipoldeint
13665 Linear interpolating deinterlacing filter that deinterlaces the given block by
13666 linearly interpolating every second line.
13668 @item ci/cubicipoldeint
13669 Cubic interpolating deinterlacing filter deinterlaces the given block by
13670 cubically interpolating every second line.
13672 @item md/mediandeint
13673 Median deinterlacing filter that deinterlaces the given block by applying a
13674 median filter to every second line.
13676 @item fd/ffmpegdeint
13677 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13678 second line with a @code{(-1 4 2 4 -1)} filter.
13681 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13682 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13684 @item fq/forceQuant[|quantizer]
13685 Overrides the quantizer table from the input with the constant quantizer you
13693 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13696 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13699 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13702 @subsection Examples
13706 Apply horizontal and vertical deblocking, deringing and automatic
13707 brightness/contrast:
13713 Apply default filters without brightness/contrast correction:
13719 Apply default filters and temporal denoiser:
13721 pp=default/tmpnoise|1|2|3
13725 Apply deblocking on luminance only, and switch vertical deblocking on or off
13726 automatically depending on available CPU time:
13733 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13734 similar to spp = 6 with 7 point DCT, where only the center sample is
13737 The filter accepts the following options:
13741 Force a constant quantization parameter. It accepts an integer in range
13742 0 to 63. If not set, the filter will use the QP from the video stream
13746 Set thresholding mode. Available modes are:
13750 Set hard thresholding.
13752 Set soft thresholding (better de-ringing effect, but likely blurrier).
13754 Set medium thresholding (good results, default).
13758 @section premultiply
13759 Apply alpha premultiply effect to input video stream using first plane
13760 of second stream as alpha.
13762 Both streams must have same dimensions and same pixel format.
13764 The filter accepts the following option:
13768 Set which planes will be processed, unprocessed planes will be copied.
13769 By default value 0xf, all planes will be processed.
13772 Do not require 2nd input for processing, instead use alpha plane from input stream.
13776 Apply prewitt operator to input video stream.
13778 The filter accepts the following option:
13782 Set which planes will be processed, unprocessed planes will be copied.
13783 By default value 0xf, all planes will be processed.
13786 Set value which will be multiplied with filtered result.
13789 Set value which will be added to filtered result.
13792 @anchor{program_opencl}
13793 @section program_opencl
13795 Filter video using an OpenCL program.
13800 OpenCL program source file.
13803 Kernel name in program.
13806 Number of inputs to the filter. Defaults to 1.
13809 Size of output frames. Defaults to the same as the first input.
13813 The program source file must contain a kernel function with the given name,
13814 which will be run once for each plane of the output. Each run on a plane
13815 gets enqueued as a separate 2D global NDRange with one work-item for each
13816 pixel to be generated. The global ID offset for each work-item is therefore
13817 the coordinates of a pixel in the destination image.
13819 The kernel function needs to take the following arguments:
13822 Destination image, @var{__write_only image2d_t}.
13824 This image will become the output; the kernel should write all of it.
13826 Frame index, @var{unsigned int}.
13828 This is a counter starting from zero and increasing by one for each frame.
13830 Source images, @var{__read_only image2d_t}.
13832 These are the most recent images on each input. The kernel may read from
13833 them to generate the output, but they can't be written to.
13840 Copy the input to the output (output must be the same size as the input).
13842 __kernel void copy(__write_only image2d_t destination,
13843 unsigned int index,
13844 __read_only image2d_t source)
13846 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13848 int2 location = (int2)(get_global_id(0), get_global_id(1));
13850 float4 value = read_imagef(source, sampler, location);
13852 write_imagef(destination, location, value);
13857 Apply a simple transformation, rotating the input by an amount increasing
13858 with the index counter. Pixel values are linearly interpolated by the
13859 sampler, and the output need not have the same dimensions as the input.
13861 __kernel void rotate_image(__write_only image2d_t dst,
13862 unsigned int index,
13863 __read_only image2d_t src)
13865 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13866 CLK_FILTER_LINEAR);
13868 float angle = (float)index / 100.0f;
13870 float2 dst_dim = convert_float2(get_image_dim(dst));
13871 float2 src_dim = convert_float2(get_image_dim(src));
13873 float2 dst_cen = dst_dim / 2.0f;
13874 float2 src_cen = src_dim / 2.0f;
13876 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13878 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13880 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13881 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13883 src_pos = src_pos * src_dim / dst_dim;
13885 float2 src_loc = src_pos + src_cen;
13887 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13888 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13889 write_imagef(dst, dst_loc, 0.5f);
13891 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13896 Blend two inputs together, with the amount of each input used varying
13897 with the index counter.
13899 __kernel void blend_images(__write_only image2d_t dst,
13900 unsigned int index,
13901 __read_only image2d_t src1,
13902 __read_only image2d_t src2)
13904 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13905 CLK_FILTER_LINEAR);
13907 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13909 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13910 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13911 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13913 float4 val1 = read_imagef(src1, sampler, src1_loc);
13914 float4 val2 = read_imagef(src2, sampler, src2_loc);
13916 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13922 @section pseudocolor
13924 Alter frame colors in video with pseudocolors.
13926 This filter accept the following options:
13930 set pixel first component expression
13933 set pixel second component expression
13936 set pixel third component expression
13939 set pixel fourth component expression, corresponds to the alpha component
13942 set component to use as base for altering colors
13945 Each of them specifies the expression to use for computing the lookup table for
13946 the corresponding pixel component values.
13948 The expressions can contain the following constants and functions:
13953 The input width and height.
13956 The input value for the pixel component.
13958 @item ymin, umin, vmin, amin
13959 The minimum allowed component value.
13961 @item ymax, umax, vmax, amax
13962 The maximum allowed component value.
13965 All expressions default to "val".
13967 @subsection Examples
13971 Change too high luma values to gradient:
13973 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'"
13979 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
13980 Ratio) between two input videos.
13982 This filter takes in input two input videos, the first input is
13983 considered the "main" source and is passed unchanged to the
13984 output. The second input is used as a "reference" video for computing
13987 Both video inputs must have the same resolution and pixel format for
13988 this filter to work correctly. Also it assumes that both inputs
13989 have the same number of frames, which are compared one by one.
13991 The obtained average PSNR is printed through the logging system.
13993 The filter stores the accumulated MSE (mean squared error) of each
13994 frame, and at the end of the processing it is averaged across all frames
13995 equally, and the following formula is applied to obtain the PSNR:
13998 PSNR = 10*log10(MAX^2/MSE)
14001 Where MAX is the average of the maximum values of each component of the
14004 The description of the accepted parameters follows.
14007 @item stats_file, f
14008 If specified the filter will use the named file to save the PSNR of
14009 each individual frame. When filename equals "-" the data is sent to
14012 @item stats_version
14013 Specifies which version of the stats file format to use. Details of
14014 each format are written below.
14015 Default value is 1.
14017 @item stats_add_max
14018 Determines whether the max value is output to the stats log.
14019 Default value is 0.
14020 Requires stats_version >= 2. If this is set and stats_version < 2,
14021 the filter will return an error.
14024 This filter also supports the @ref{framesync} options.
14026 The file printed if @var{stats_file} is selected, contains a sequence of
14027 key/value pairs of the form @var{key}:@var{value} for each compared
14030 If a @var{stats_version} greater than 1 is specified, a header line precedes
14031 the list of per-frame-pair stats, with key value pairs following the frame
14032 format with the following parameters:
14035 @item psnr_log_version
14036 The version of the log file format. Will match @var{stats_version}.
14039 A comma separated list of the per-frame-pair parameters included in
14043 A description of each shown per-frame-pair parameter follows:
14047 sequential number of the input frame, starting from 1
14050 Mean Square Error pixel-by-pixel average difference of the compared
14051 frames, averaged over all the image components.
14053 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14054 Mean Square Error pixel-by-pixel average difference of the compared
14055 frames for the component specified by the suffix.
14057 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14058 Peak Signal to Noise ratio of the compared frames for the component
14059 specified by the suffix.
14061 @item max_avg, max_y, max_u, max_v
14062 Maximum allowed value for each channel, and average over all
14068 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14069 [main][ref] psnr="stats_file=stats.log" [out]
14072 On this example the input file being processed is compared with the
14073 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14074 is stored in @file{stats.log}.
14079 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14080 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14083 The pullup filter is designed to take advantage of future context in making
14084 its decisions. This filter is stateless in the sense that it does not lock
14085 onto a pattern to follow, but it instead looks forward to the following
14086 fields in order to identify matches and rebuild progressive frames.
14088 To produce content with an even framerate, insert the fps filter after
14089 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14090 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14092 The filter accepts the following options:
14099 These options set the amount of "junk" to ignore at the left, right, top, and
14100 bottom of the image, respectively. Left and right are in units of 8 pixels,
14101 while top and bottom are in units of 2 lines.
14102 The default is 8 pixels on each side.
14105 Set the strict breaks. Setting this option to 1 will reduce the chances of
14106 filter generating an occasional mismatched frame, but it may also cause an
14107 excessive number of frames to be dropped during high motion sequences.
14108 Conversely, setting it to -1 will make filter match fields more easily.
14109 This may help processing of video where there is slight blurring between
14110 the fields, but may also cause there to be interlaced frames in the output.
14111 Default value is @code{0}.
14114 Set the metric plane to use. It accepts the following values:
14120 Use chroma blue plane.
14123 Use chroma red plane.
14126 This option may be set to use chroma plane instead of the default luma plane
14127 for doing filter's computations. This may improve accuracy on very clean
14128 source material, but more likely will decrease accuracy, especially if there
14129 is chroma noise (rainbow effect) or any grayscale video.
14130 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14131 load and make pullup usable in realtime on slow machines.
14134 For best results (without duplicated frames in the output file) it is
14135 necessary to change the output frame rate. For example, to inverse
14136 telecine NTSC input:
14138 ffmpeg -i input -vf pullup -r 24000/1001 ...
14143 Change video quantization parameters (QP).
14145 The filter accepts the following option:
14149 Set expression for quantization parameter.
14152 The expression is evaluated through the eval API and can contain, among others,
14153 the following constants:
14157 1 if index is not 129, 0 otherwise.
14160 Sequential index starting from -129 to 128.
14163 @subsection Examples
14167 Some equation like:
14175 Flush video frames from internal cache of frames into a random order.
14176 No frame is discarded.
14177 Inspired by @ref{frei0r} nervous filter.
14181 Set size in number of frames of internal cache, in range from @code{2} to
14182 @code{512}. Default is @code{30}.
14185 Set seed for random number generator, must be an integer included between
14186 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14187 less than @code{0}, the filter will try to use a good random seed on a
14191 @section readeia608
14193 Read closed captioning (EIA-608) information from the top lines of a video frame.
14195 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14196 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14197 with EIA-608 data (starting from 0). A description of each metadata value follows:
14200 @item lavfi.readeia608.X.cc
14201 The two bytes stored as EIA-608 data (printed in hexadecimal).
14203 @item lavfi.readeia608.X.line
14204 The number of the line on which the EIA-608 data was identified and read.
14207 This filter accepts the following options:
14211 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14214 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14217 Set minimal acceptable amplitude change for sync codes detection.
14218 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14221 Set the ratio of width reserved for sync code detection.
14222 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14225 Set the max peaks height difference for sync code detection.
14226 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14229 Set max peaks period difference for sync code detection.
14230 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14233 Set the first two max start code bits differences.
14234 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14237 Set the minimum ratio of bits height compared to 3rd start code bit.
14238 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14241 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14244 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14247 Enable checking the parity bit. In the event of a parity error, the filter will output
14248 @code{0x00} for that character. Default is false.
14251 @subsection Examples
14255 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14257 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
14263 Read vertical interval timecode (VITC) information from the top lines of a
14266 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14267 timecode value, if a valid timecode has been detected. Further metadata key
14268 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14269 timecode data has been found or not.
14271 This filter accepts the following options:
14275 Set the maximum number of lines to scan for VITC data. If the value is set to
14276 @code{-1} the full video frame is scanned. Default is @code{45}.
14279 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14280 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14283 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14284 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14287 @subsection Examples
14291 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14292 draw @code{--:--:--:--} as a placeholder:
14294 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14300 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14302 Destination pixel at position (X, Y) will be picked from source (x, y) position
14303 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14304 value for pixel will be used for destination pixel.
14306 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14307 will have Xmap/Ymap video stream dimensions.
14308 Xmap and Ymap input video streams are 16bit depth, single channel.
14310 @section removegrain
14312 The removegrain filter is a spatial denoiser for progressive video.
14316 Set mode for the first plane.
14319 Set mode for the second plane.
14322 Set mode for the third plane.
14325 Set mode for the fourth plane.
14328 Range of mode is from 0 to 24. Description of each mode follows:
14332 Leave input plane unchanged. Default.
14335 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14338 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14341 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14344 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14345 This is equivalent to a median filter.
14348 Line-sensitive clipping giving the minimal change.
14351 Line-sensitive clipping, intermediate.
14354 Line-sensitive clipping, intermediate.
14357 Line-sensitive clipping, intermediate.
14360 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14363 Replaces the target pixel with the closest neighbour.
14366 [1 2 1] horizontal and vertical kernel blur.
14372 Bob mode, interpolates top field from the line where the neighbours
14373 pixels are the closest.
14376 Bob mode, interpolates bottom field from the line where the neighbours
14377 pixels are the closest.
14380 Bob mode, interpolates top field. Same as 13 but with a more complicated
14381 interpolation formula.
14384 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14385 interpolation formula.
14388 Clips the pixel with the minimum and maximum of respectively the maximum and
14389 minimum of each pair of opposite neighbour pixels.
14392 Line-sensitive clipping using opposite neighbours whose greatest distance from
14393 the current pixel is minimal.
14396 Replaces the pixel with the average of its 8 neighbours.
14399 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14402 Clips pixels using the averages of opposite neighbour.
14405 Same as mode 21 but simpler and faster.
14408 Small edge and halo removal, but reputed useless.
14414 @section removelogo
14416 Suppress a TV station logo, using an image file to determine which
14417 pixels comprise the logo. It works by filling in the pixels that
14418 comprise the logo with neighboring pixels.
14420 The filter accepts the following options:
14424 Set the filter bitmap file, which can be any image format supported by
14425 libavformat. The width and height of the image file must match those of the
14426 video stream being processed.
14429 Pixels in the provided bitmap image with a value of zero are not
14430 considered part of the logo, non-zero pixels are considered part of
14431 the logo. If you use white (255) for the logo and black (0) for the
14432 rest, you will be safe. For making the filter bitmap, it is
14433 recommended to take a screen capture of a black frame with the logo
14434 visible, and then using a threshold filter followed by the erode
14435 filter once or twice.
14437 If needed, little splotches can be fixed manually. Remember that if
14438 logo pixels are not covered, the filter quality will be much
14439 reduced. Marking too many pixels as part of the logo does not hurt as
14440 much, but it will increase the amount of blurring needed to cover over
14441 the image and will destroy more information than necessary, and extra
14442 pixels will slow things down on a large logo.
14444 @section repeatfields
14446 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14447 fields based on its value.
14451 Reverse a video clip.
14453 Warning: This filter requires memory to buffer the entire clip, so trimming
14456 @subsection Examples
14460 Take the first 5 seconds of a clip, and reverse it.
14467 Shift R/G/B/A pixels horizontally and/or vertically.
14469 The filter accepts the following options:
14472 Set amount to shift red horizontally.
14474 Set amount to shift red vertically.
14476 Set amount to shift green horizontally.
14478 Set amount to shift green vertically.
14480 Set amount to shift blue horizontally.
14482 Set amount to shift blue vertically.
14484 Set amount to shift alpha horizontally.
14486 Set amount to shift alpha vertically.
14488 Set edge mode, can be @var{smear}, default, or @var{warp}.
14492 Apply roberts cross operator to input video stream.
14494 The filter accepts the following option:
14498 Set which planes will be processed, unprocessed planes will be copied.
14499 By default value 0xf, all planes will be processed.
14502 Set value which will be multiplied with filtered result.
14505 Set value which will be added to filtered result.
14510 Rotate video by an arbitrary angle expressed in radians.
14512 The filter accepts the following options:
14514 A description of the optional parameters follows.
14517 Set an expression for the angle by which to rotate the input video
14518 clockwise, expressed as a number of radians. A negative value will
14519 result in a counter-clockwise rotation. By default it is set to "0".
14521 This expression is evaluated for each frame.
14524 Set the output width expression, default value is "iw".
14525 This expression is evaluated just once during configuration.
14528 Set the output height expression, default value is "ih".
14529 This expression is evaluated just once during configuration.
14532 Enable bilinear interpolation if set to 1, a value of 0 disables
14533 it. Default value is 1.
14536 Set the color used to fill the output area not covered by the rotated
14537 image. For the general syntax of this option, check the
14538 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14539 If the special value "none" is selected then no
14540 background is printed (useful for example if the background is never shown).
14542 Default value is "black".
14545 The expressions for the angle and the output size can contain the
14546 following constants and functions:
14550 sequential number of the input frame, starting from 0. It is always NAN
14551 before the first frame is filtered.
14554 time in seconds of the input frame, it is set to 0 when the filter is
14555 configured. It is always NAN before the first frame is filtered.
14559 horizontal and vertical chroma subsample values. For example for the
14560 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14564 the input video width and height
14568 the output width and height, that is the size of the padded area as
14569 specified by the @var{width} and @var{height} expressions
14573 the minimal width/height required for completely containing the input
14574 video rotated by @var{a} radians.
14576 These are only available when computing the @option{out_w} and
14577 @option{out_h} expressions.
14580 @subsection Examples
14584 Rotate the input by PI/6 radians clockwise:
14590 Rotate the input by PI/6 radians counter-clockwise:
14596 Rotate the input by 45 degrees clockwise:
14602 Apply a constant rotation with period T, starting from an angle of PI/3:
14604 rotate=PI/3+2*PI*t/T
14608 Make the input video rotation oscillating with a period of T
14609 seconds and an amplitude of A radians:
14611 rotate=A*sin(2*PI/T*t)
14615 Rotate the video, output size is chosen so that the whole rotating
14616 input video is always completely contained in the output:
14618 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14622 Rotate the video, reduce the output size so that no background is ever
14625 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14629 @subsection Commands
14631 The filter supports the following commands:
14635 Set the angle expression.
14636 The command accepts the same syntax of the corresponding option.
14638 If the specified expression is not valid, it is kept at its current
14644 Apply Shape Adaptive Blur.
14646 The filter accepts the following options:
14649 @item luma_radius, lr
14650 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14651 value is 1.0. A greater value will result in a more blurred image, and
14652 in slower processing.
14654 @item luma_pre_filter_radius, lpfr
14655 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14658 @item luma_strength, ls
14659 Set luma maximum difference between pixels to still be considered, must
14660 be a value in the 0.1-100.0 range, default value is 1.0.
14662 @item chroma_radius, cr
14663 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14664 greater value will result in a more blurred image, and in slower
14667 @item chroma_pre_filter_radius, cpfr
14668 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14670 @item chroma_strength, cs
14671 Set chroma maximum difference between pixels to still be considered,
14672 must be a value in the -0.9-100.0 range.
14675 Each chroma option value, if not explicitly specified, is set to the
14676 corresponding luma option value.
14681 Scale (resize) the input video, using the libswscale library.
14683 The scale filter forces the output display aspect ratio to be the same
14684 of the input, by changing the output sample aspect ratio.
14686 If the input image format is different from the format requested by
14687 the next filter, the scale filter will convert the input to the
14690 @subsection Options
14691 The filter accepts the following options, or any of the options
14692 supported by the libswscale scaler.
14694 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14695 the complete list of scaler options.
14700 Set the output video dimension expression. Default value is the input
14703 If the @var{width} or @var{w} value is 0, the input width is used for
14704 the output. If the @var{height} or @var{h} value is 0, the input height
14705 is used for the output.
14707 If one and only one of the values is -n with n >= 1, the scale filter
14708 will use a value that maintains the aspect ratio of the input image,
14709 calculated from the other specified dimension. After that it will,
14710 however, make sure that the calculated dimension is divisible by n and
14711 adjust the value if necessary.
14713 If both values are -n with n >= 1, the behavior will be identical to
14714 both values being set to 0 as previously detailed.
14716 See below for the list of accepted constants for use in the dimension
14720 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14724 Only evaluate expressions once during the filter initialization or when a command is processed.
14727 Evaluate expressions for each incoming frame.
14731 Default value is @samp{init}.
14735 Set the interlacing mode. It accepts the following values:
14739 Force interlaced aware scaling.
14742 Do not apply interlaced scaling.
14745 Select interlaced aware scaling depending on whether the source frames
14746 are flagged as interlaced or not.
14749 Default value is @samp{0}.
14752 Set libswscale scaling flags. See
14753 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14754 complete list of values. If not explicitly specified the filter applies
14758 @item param0, param1
14759 Set libswscale input parameters for scaling algorithms that need them. See
14760 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14761 complete documentation. If not explicitly specified the filter applies
14767 Set the video size. For the syntax of this option, check the
14768 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14770 @item in_color_matrix
14771 @item out_color_matrix
14772 Set in/output YCbCr color space type.
14774 This allows the autodetected value to be overridden as well as allows forcing
14775 a specific value used for the output and encoder.
14777 If not specified, the color space type depends on the pixel format.
14783 Choose automatically.
14786 Format conforming to International Telecommunication Union (ITU)
14787 Recommendation BT.709.
14790 Set color space conforming to the United States Federal Communications
14791 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14794 Set color space conforming to:
14798 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14801 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14804 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14809 Set color space conforming to SMPTE ST 240:1999.
14814 Set in/output YCbCr sample range.
14816 This allows the autodetected value to be overridden as well as allows forcing
14817 a specific value used for the output and encoder. If not specified, the
14818 range depends on the pixel format. Possible values:
14822 Choose automatically.
14825 Set full range (0-255 in case of 8-bit luma).
14827 @item mpeg/limited/tv
14828 Set "MPEG" range (16-235 in case of 8-bit luma).
14831 @item force_original_aspect_ratio
14832 Enable decreasing or increasing output video width or height if necessary to
14833 keep the original aspect ratio. Possible values:
14837 Scale the video as specified and disable this feature.
14840 The output video dimensions will automatically be decreased if needed.
14843 The output video dimensions will automatically be increased if needed.
14847 One useful instance of this option is that when you know a specific device's
14848 maximum allowed resolution, you can use this to limit the output video to
14849 that, while retaining the aspect ratio. For example, device A allows
14850 1280x720 playback, and your video is 1920x800. Using this option (set it to
14851 decrease) and specifying 1280x720 to the command line makes the output
14854 Please note that this is a different thing than specifying -1 for @option{w}
14855 or @option{h}, you still need to specify the output resolution for this option
14860 The values of the @option{w} and @option{h} options are expressions
14861 containing the following constants:
14866 The input width and height
14870 These are the same as @var{in_w} and @var{in_h}.
14874 The output (scaled) width and height
14878 These are the same as @var{out_w} and @var{out_h}
14881 The same as @var{iw} / @var{ih}
14884 input sample aspect ratio
14887 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14891 horizontal and vertical input chroma subsample values. For example for the
14892 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14896 horizontal and vertical output chroma subsample values. For example for the
14897 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14900 @subsection Examples
14904 Scale the input video to a size of 200x100
14909 This is equivalent to:
14920 Specify a size abbreviation for the output size:
14925 which can also be written as:
14931 Scale the input to 2x:
14933 scale=w=2*iw:h=2*ih
14937 The above is the same as:
14939 scale=2*in_w:2*in_h
14943 Scale the input to 2x with forced interlaced scaling:
14945 scale=2*iw:2*ih:interl=1
14949 Scale the input to half size:
14951 scale=w=iw/2:h=ih/2
14955 Increase the width, and set the height to the same size:
14961 Seek Greek harmony:
14968 Increase the height, and set the width to 3/2 of the height:
14970 scale=w=3/2*oh:h=3/5*ih
14974 Increase the size, making the size a multiple of the chroma
14977 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
14981 Increase the width to a maximum of 500 pixels,
14982 keeping the same aspect ratio as the input:
14984 scale=w='min(500\, iw*3/2):h=-1'
14988 Make pixels square by combining scale and setsar:
14990 scale='trunc(ih*dar):ih',setsar=1/1
14994 Make pixels square by combining scale and setsar,
14995 making sure the resulting resolution is even (required by some codecs):
14997 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15001 @subsection Commands
15003 This filter supports the following commands:
15007 Set the output video dimension expression.
15008 The command accepts the same syntax of the corresponding option.
15010 If the specified expression is not valid, it is kept at its current
15016 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15017 format conversion on CUDA video frames. Setting the output width and height
15018 works in the same way as for the @var{scale} filter.
15020 The following additional options are accepted:
15023 The pixel format of the output CUDA frames. If set to the string "same" (the
15024 default), the input format will be kept. Note that automatic format negotiation
15025 and conversion is not yet supported for hardware frames
15028 The interpolation algorithm used for resizing. One of the following:
15035 @item cubic2p_bspline
15036 2-parameter cubic (B=1, C=0)
15038 @item cubic2p_catmullrom
15039 2-parameter cubic (B=0, C=1/2)
15041 @item cubic2p_b05c03
15042 2-parameter cubic (B=1/2, C=3/10)
15054 Scale (resize) the input video, based on a reference video.
15056 See the scale filter for available options, scale2ref supports the same but
15057 uses the reference video instead of the main input as basis. scale2ref also
15058 supports the following additional constants for the @option{w} and
15059 @option{h} options:
15064 The main input video's width and height
15067 The same as @var{main_w} / @var{main_h}
15070 The main input video's sample aspect ratio
15072 @item main_dar, mdar
15073 The main input video's display aspect ratio. Calculated from
15074 @code{(main_w / main_h) * main_sar}.
15078 The main input video's horizontal and vertical chroma subsample values.
15079 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15083 @subsection Examples
15087 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15089 'scale2ref[b][a];[a][b]overlay'
15093 @anchor{selectivecolor}
15094 @section selectivecolor
15096 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15097 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15098 by the "purity" of the color (that is, how saturated it already is).
15100 This filter is similar to the Adobe Photoshop Selective Color tool.
15102 The filter accepts the following options:
15105 @item correction_method
15106 Select color correction method.
15108 Available values are:
15111 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15114 Specified adjustments are relative to the original component value.
15116 Default is @code{absolute}.
15118 Adjustments for red pixels (pixels where the red component is the maximum)
15120 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15122 Adjustments for green pixels (pixels where the green component is the maximum)
15124 Adjustments for cyan pixels (pixels where the red component is the minimum)
15126 Adjustments for blue pixels (pixels where the blue component is the maximum)
15128 Adjustments for magenta pixels (pixels where the green component is the minimum)
15130 Adjustments for white pixels (pixels where all components are greater than 128)
15132 Adjustments for all pixels except pure black and pure white
15134 Adjustments for black pixels (pixels where all components are lesser than 128)
15136 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15139 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15140 4 space separated floating point adjustment values in the [-1,1] range,
15141 respectively to adjust the amount of cyan, magenta, yellow and black for the
15142 pixels of its range.
15144 @subsection Examples
15148 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15149 increase magenta by 27% in blue areas:
15151 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15155 Use a Photoshop selective color preset:
15157 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15161 @anchor{separatefields}
15162 @section separatefields
15164 The @code{separatefields} takes a frame-based video input and splits
15165 each frame into its components fields, producing a new half height clip
15166 with twice the frame rate and twice the frame count.
15168 This filter use field-dominance information in frame to decide which
15169 of each pair of fields to place first in the output.
15170 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15172 @section setdar, setsar
15174 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15177 This is done by changing the specified Sample (aka Pixel) Aspect
15178 Ratio, according to the following equation:
15180 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15183 Keep in mind that the @code{setdar} filter does not modify the pixel
15184 dimensions of the video frame. Also, the display aspect ratio set by
15185 this filter may be changed by later filters in the filterchain,
15186 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15189 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15190 the filter output video.
15192 Note that as a consequence of the application of this filter, the
15193 output display aspect ratio will change according to the equation
15196 Keep in mind that the sample aspect ratio set by the @code{setsar}
15197 filter may be changed by later filters in the filterchain, e.g. if
15198 another "setsar" or a "setdar" filter is applied.
15200 It accepts the following parameters:
15203 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15204 Set the aspect ratio used by the filter.
15206 The parameter can be a floating point number string, an expression, or
15207 a string of the form @var{num}:@var{den}, where @var{num} and
15208 @var{den} are the numerator and denominator of the aspect ratio. If
15209 the parameter is not specified, it is assumed the value "0".
15210 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15214 Set the maximum integer value to use for expressing numerator and
15215 denominator when reducing the expressed aspect ratio to a rational.
15216 Default value is @code{100}.
15220 The parameter @var{sar} is an expression containing
15221 the following constants:
15225 These are approximated values for the mathematical constants e
15226 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15229 The input width and height.
15232 These are the same as @var{w} / @var{h}.
15235 The input sample aspect ratio.
15238 The input display aspect ratio. It is the same as
15239 (@var{w} / @var{h}) * @var{sar}.
15242 Horizontal and vertical chroma subsample values. For example, for the
15243 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15246 @subsection Examples
15251 To change the display aspect ratio to 16:9, specify one of the following:
15258 To change the sample aspect ratio to 10:11, specify:
15264 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15265 1000 in the aspect ratio reduction, use the command:
15267 setdar=ratio=16/9:max=1000
15275 Force field for the output video frame.
15277 The @code{setfield} filter marks the interlace type field for the
15278 output frames. It does not change the input frame, but only sets the
15279 corresponding property, which affects how the frame is treated by
15280 following filters (e.g. @code{fieldorder} or @code{yadif}).
15282 The filter accepts the following options:
15287 Available values are:
15291 Keep the same field property.
15294 Mark the frame as bottom-field-first.
15297 Mark the frame as top-field-first.
15300 Mark the frame as progressive.
15307 Force frame parameter for the output video frame.
15309 The @code{setparams} filter marks interlace and color range for the
15310 output frames. It does not change the input frame, but only sets the
15311 corresponding property, which affects how the frame is treated by
15316 Available values are:
15320 Keep the same field property (default).
15323 Mark the frame as bottom-field-first.
15326 Mark the frame as top-field-first.
15329 Mark the frame as progressive.
15333 Available values are:
15337 Keep the same color range property (default).
15339 @item unspecified, unknown
15340 Mark the frame as unspecified color range.
15342 @item limited, tv, mpeg
15343 Mark the frame as limited range.
15345 @item full, pc, jpeg
15346 Mark the frame as full range.
15349 @item color_primaries
15350 Set the color primaries.
15351 Available values are:
15355 Keep the same color primaries property (default).
15372 Set the color transfer.
15373 Available values are:
15377 Keep the same color trc property (default).
15399 Set the colorspace.
15400 Available values are:
15404 Keep the same colorspace property (default).
15417 @item chroma-derived-nc
15418 @item chroma-derived-c
15425 Show a line containing various information for each input video frame.
15426 The input video is not modified.
15428 This filter supports the following options:
15432 Calculate checksums of each plane. By default enabled.
15435 The shown line contains a sequence of key/value pairs of the form
15436 @var{key}:@var{value}.
15438 The following values are shown in the output:
15442 The (sequential) number of the input frame, starting from 0.
15445 The Presentation TimeStamp of the input frame, expressed as a number of
15446 time base units. The time base unit depends on the filter input pad.
15449 The Presentation TimeStamp of the input frame, expressed as a number of
15453 The position of the frame in the input stream, or -1 if this information is
15454 unavailable and/or meaningless (for example in case of synthetic video).
15457 The pixel format name.
15460 The sample aspect ratio of the input frame, expressed in the form
15461 @var{num}/@var{den}.
15464 The size of the input frame. For the syntax of this option, check the
15465 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15468 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15469 for bottom field first).
15472 This is 1 if the frame is a key frame, 0 otherwise.
15475 The picture type of the input frame ("I" for an I-frame, "P" for a
15476 P-frame, "B" for a B-frame, or "?" for an unknown type).
15477 Also refer to the documentation of the @code{AVPictureType} enum and of
15478 the @code{av_get_picture_type_char} function defined in
15479 @file{libavutil/avutil.h}.
15482 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15484 @item plane_checksum
15485 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15486 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15489 @section showpalette
15491 Displays the 256 colors palette of each frame. This filter is only relevant for
15492 @var{pal8} pixel format frames.
15494 It accepts the following option:
15498 Set the size of the box used to represent one palette color entry. Default is
15499 @code{30} (for a @code{30x30} pixel box).
15502 @section shuffleframes
15504 Reorder and/or duplicate and/or drop video frames.
15506 It accepts the following parameters:
15510 Set the destination indexes of input frames.
15511 This is space or '|' separated list of indexes that maps input frames to output
15512 frames. Number of indexes also sets maximal value that each index may have.
15513 '-1' index have special meaning and that is to drop frame.
15516 The first frame has the index 0. The default is to keep the input unchanged.
15518 @subsection Examples
15522 Swap second and third frame of every three frames of the input:
15524 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15528 Swap 10th and 1st frame of every ten frames of the input:
15530 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15534 @section shuffleplanes
15536 Reorder and/or duplicate video planes.
15538 It accepts the following parameters:
15543 The index of the input plane to be used as the first output plane.
15546 The index of the input plane to be used as the second output plane.
15549 The index of the input plane to be used as the third output plane.
15552 The index of the input plane to be used as the fourth output plane.
15556 The first plane has the index 0. The default is to keep the input unchanged.
15558 @subsection Examples
15562 Swap the second and third planes of the input:
15564 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15568 @anchor{signalstats}
15569 @section signalstats
15570 Evaluate various visual metrics that assist in determining issues associated
15571 with the digitization of analog video media.
15573 By default the filter will log these metadata values:
15577 Display the minimal Y value contained within the input frame. Expressed in
15581 Display the Y value at the 10% percentile within the input frame. Expressed in
15585 Display the average Y value within the input frame. Expressed in range of
15589 Display the Y value at the 90% percentile within the input frame. Expressed in
15593 Display the maximum Y value contained within the input frame. Expressed in
15597 Display the minimal U value contained within the input frame. Expressed in
15601 Display the U value at the 10% percentile within the input frame. Expressed in
15605 Display the average U value within the input frame. Expressed in range of
15609 Display the U value at the 90% percentile within the input frame. Expressed in
15613 Display the maximum U value contained within the input frame. Expressed in
15617 Display the minimal V value contained within the input frame. Expressed in
15621 Display the V value at the 10% percentile within the input frame. Expressed in
15625 Display the average V value within the input frame. Expressed in range of
15629 Display the V value at the 90% percentile within the input frame. Expressed in
15633 Display the maximum V value contained within the input frame. Expressed in
15637 Display the minimal saturation value contained within the input frame.
15638 Expressed in range of [0-~181.02].
15641 Display the saturation value at the 10% percentile within the input frame.
15642 Expressed in range of [0-~181.02].
15645 Display the average saturation value within the input frame. Expressed in range
15649 Display the saturation value at the 90% percentile within the input frame.
15650 Expressed in range of [0-~181.02].
15653 Display the maximum saturation value contained within the input frame.
15654 Expressed in range of [0-~181.02].
15657 Display the median value for hue within the input frame. Expressed in range of
15661 Display the average value for hue within the input frame. Expressed in range of
15665 Display the average of sample value difference between all values of the Y
15666 plane in the current frame and corresponding values of the previous input frame.
15667 Expressed in range of [0-255].
15670 Display the average of sample value difference between all values of the U
15671 plane in the current frame and corresponding values of the previous input frame.
15672 Expressed in range of [0-255].
15675 Display the average of sample value difference between all values of the V
15676 plane in the current frame and corresponding values of the previous input frame.
15677 Expressed in range of [0-255].
15680 Display bit depth of Y plane in current frame.
15681 Expressed in range of [0-16].
15684 Display bit depth of U plane in current frame.
15685 Expressed in range of [0-16].
15688 Display bit depth of V plane in current frame.
15689 Expressed in range of [0-16].
15692 The filter accepts the following options:
15698 @option{stat} specify an additional form of image analysis.
15699 @option{out} output video with the specified type of pixel highlighted.
15701 Both options accept the following values:
15705 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15706 unlike the neighboring pixels of the same field. Examples of temporal outliers
15707 include the results of video dropouts, head clogs, or tape tracking issues.
15710 Identify @var{vertical line repetition}. Vertical line repetition includes
15711 similar rows of pixels within a frame. In born-digital video vertical line
15712 repetition is common, but this pattern is uncommon in video digitized from an
15713 analog source. When it occurs in video that results from the digitization of an
15714 analog source it can indicate concealment from a dropout compensator.
15717 Identify pixels that fall outside of legal broadcast range.
15721 Set the highlight color for the @option{out} option. The default color is
15725 @subsection Examples
15729 Output data of various video metrics:
15731 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15735 Output specific data about the minimum and maximum values of the Y plane per frame:
15737 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15741 Playback video while highlighting pixels that are outside of broadcast range in red.
15743 ffplay example.mov -vf signalstats="out=brng:color=red"
15747 Playback video with signalstats metadata drawn over the frame.
15749 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15752 The contents of signalstat_drawtext.txt used in the command are:
15755 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15756 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15757 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15758 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15766 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15767 input. In this case the matching between the inputs can be calculated additionally.
15768 The filter always passes through the first input. The signature of each stream can
15769 be written into a file.
15771 It accepts the following options:
15775 Enable or disable the matching process.
15777 Available values are:
15781 Disable the calculation of a matching (default).
15783 Calculate the matching for the whole video and output whether the whole video
15784 matches or only parts.
15786 Calculate only until a matching is found or the video ends. Should be faster in
15791 Set the number of inputs. The option value must be a non negative integer.
15792 Default value is 1.
15795 Set the path to which the output is written. If there is more than one input,
15796 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15797 integer), that will be replaced with the input number. If no filename is
15798 specified, no output will be written. This is the default.
15801 Choose the output format.
15803 Available values are:
15807 Use the specified binary representation (default).
15809 Use the specified xml representation.
15813 Set threshold to detect one word as similar. The option value must be an integer
15814 greater than zero. The default value is 9000.
15817 Set threshold to detect all words as similar. The option value must be an integer
15818 greater than zero. The default value is 60000.
15821 Set threshold to detect frames as similar. The option value must be an integer
15822 greater than zero. The default value is 116.
15825 Set the minimum length of a sequence in frames to recognize it as matching
15826 sequence. The option value must be a non negative integer value.
15827 The default value is 0.
15830 Set the minimum relation, that matching frames to all frames must have.
15831 The option value must be a double value between 0 and 1. The default value is 0.5.
15834 @subsection Examples
15838 To calculate the signature of an input video and store it in signature.bin:
15840 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15844 To detect whether two videos match and store the signatures in XML format in
15845 signature0.xml and signature1.xml:
15847 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 -
15855 Blur the input video without impacting the outlines.
15857 It accepts the following options:
15860 @item luma_radius, lr
15861 Set the luma radius. The option value must be a float number in
15862 the range [0.1,5.0] that specifies the variance of the gaussian filter
15863 used to blur the image (slower if larger). Default value is 1.0.
15865 @item luma_strength, ls
15866 Set the luma strength. The option value must be a float number
15867 in the range [-1.0,1.0] that configures the blurring. A value included
15868 in [0.0,1.0] will blur the image whereas a value included in
15869 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15871 @item luma_threshold, lt
15872 Set the luma threshold used as a coefficient to determine
15873 whether a pixel should be blurred or not. The option value must be an
15874 integer in the range [-30,30]. A value of 0 will filter all the image,
15875 a value included in [0,30] will filter flat areas and a value included
15876 in [-30,0] will filter edges. Default value is 0.
15878 @item chroma_radius, cr
15879 Set the chroma radius. The option value must be a float number in
15880 the range [0.1,5.0] that specifies the variance of the gaussian filter
15881 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15883 @item chroma_strength, cs
15884 Set the chroma strength. The option value must be a float number
15885 in the range [-1.0,1.0] that configures the blurring. A value included
15886 in [0.0,1.0] will blur the image whereas a value included in
15887 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15889 @item chroma_threshold, ct
15890 Set the chroma threshold used as a coefficient to determine
15891 whether a pixel should be blurred or not. The option value must be an
15892 integer in the range [-30,30]. A value of 0 will filter all the image,
15893 a value included in [0,30] will filter flat areas and a value included
15894 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15897 If a chroma option is not explicitly set, the corresponding luma value
15902 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15904 This filter takes in input two input videos, the first input is
15905 considered the "main" source and is passed unchanged to the
15906 output. The second input is used as a "reference" video for computing
15909 Both video inputs must have the same resolution and pixel format for
15910 this filter to work correctly. Also it assumes that both inputs
15911 have the same number of frames, which are compared one by one.
15913 The filter stores the calculated SSIM of each frame.
15915 The description of the accepted parameters follows.
15918 @item stats_file, f
15919 If specified the filter will use the named file to save the SSIM of
15920 each individual frame. When filename equals "-" the data is sent to
15924 The file printed if @var{stats_file} is selected, contains a sequence of
15925 key/value pairs of the form @var{key}:@var{value} for each compared
15928 A description of each shown parameter follows:
15932 sequential number of the input frame, starting from 1
15934 @item Y, U, V, R, G, B
15935 SSIM of the compared frames for the component specified by the suffix.
15938 SSIM of the compared frames for the whole frame.
15941 Same as above but in dB representation.
15944 This filter also supports the @ref{framesync} options.
15948 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15949 [main][ref] ssim="stats_file=stats.log" [out]
15952 On this example the input file being processed is compared with the
15953 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15954 is stored in @file{stats.log}.
15956 Another example with both psnr and ssim at same time:
15958 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
15963 Convert between different stereoscopic image formats.
15965 The filters accept the following options:
15969 Set stereoscopic image format of input.
15971 Available values for input image formats are:
15974 side by side parallel (left eye left, right eye right)
15977 side by side crosseye (right eye left, left eye right)
15980 side by side parallel with half width resolution
15981 (left eye left, right eye right)
15984 side by side crosseye with half width resolution
15985 (right eye left, left eye right)
15988 above-below (left eye above, right eye below)
15991 above-below (right eye above, left eye below)
15994 above-below with half height resolution
15995 (left eye above, right eye below)
15998 above-below with half height resolution
15999 (right eye above, left eye below)
16002 alternating frames (left eye first, right eye second)
16005 alternating frames (right eye first, left eye second)
16008 interleaved rows (left eye has top row, right eye starts on next row)
16011 interleaved rows (right eye has top row, left eye starts on next row)
16014 interleaved columns, left eye first
16017 interleaved columns, right eye first
16019 Default value is @samp{sbsl}.
16023 Set stereoscopic image format of output.
16027 side by side parallel (left eye left, right eye right)
16030 side by side crosseye (right eye left, left eye right)
16033 side by side parallel with half width resolution
16034 (left eye left, right eye right)
16037 side by side crosseye with half width resolution
16038 (right eye left, left eye right)
16041 above-below (left eye above, right eye below)
16044 above-below (right eye above, left eye below)
16047 above-below with half height resolution
16048 (left eye above, right eye below)
16051 above-below with half height resolution
16052 (right eye above, left eye below)
16055 alternating frames (left eye first, right eye second)
16058 alternating frames (right eye first, left eye second)
16061 interleaved rows (left eye has top row, right eye starts on next row)
16064 interleaved rows (right eye has top row, left eye starts on next row)
16067 anaglyph red/blue gray
16068 (red filter on left eye, blue filter on right eye)
16071 anaglyph red/green gray
16072 (red filter on left eye, green filter on right eye)
16075 anaglyph red/cyan gray
16076 (red filter on left eye, cyan filter on right eye)
16079 anaglyph red/cyan half colored
16080 (red filter on left eye, cyan filter on right eye)
16083 anaglyph red/cyan color
16084 (red filter on left eye, cyan filter on right eye)
16087 anaglyph red/cyan color optimized with the least squares projection of dubois
16088 (red filter on left eye, cyan filter on right eye)
16091 anaglyph green/magenta gray
16092 (green filter on left eye, magenta filter on right eye)
16095 anaglyph green/magenta half colored
16096 (green filter on left eye, magenta filter on right eye)
16099 anaglyph green/magenta colored
16100 (green filter on left eye, magenta filter on right eye)
16103 anaglyph green/magenta color optimized with the least squares projection of dubois
16104 (green filter on left eye, magenta filter on right eye)
16107 anaglyph yellow/blue gray
16108 (yellow filter on left eye, blue filter on right eye)
16111 anaglyph yellow/blue half colored
16112 (yellow filter on left eye, blue filter on right eye)
16115 anaglyph yellow/blue colored
16116 (yellow filter on left eye, blue filter on right eye)
16119 anaglyph yellow/blue color optimized with the least squares projection of dubois
16120 (yellow filter on left eye, blue filter on right eye)
16123 mono output (left eye only)
16126 mono output (right eye only)
16129 checkerboard, left eye first
16132 checkerboard, right eye first
16135 interleaved columns, left eye first
16138 interleaved columns, right eye first
16144 Default value is @samp{arcd}.
16147 @subsection Examples
16151 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16157 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16163 @section streamselect, astreamselect
16164 Select video or audio streams.
16166 The filter accepts the following options:
16170 Set number of inputs. Default is 2.
16173 Set input indexes to remap to outputs.
16176 @subsection Commands
16178 The @code{streamselect} and @code{astreamselect} filter supports the following
16183 Set input indexes to remap to outputs.
16186 @subsection Examples
16190 Select first 5 seconds 1st stream and rest of time 2nd stream:
16192 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16196 Same as above, but for audio:
16198 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16203 Apply sobel operator to input video stream.
16205 The filter accepts the following option:
16209 Set which planes will be processed, unprocessed planes will be copied.
16210 By default value 0xf, all planes will be processed.
16213 Set value which will be multiplied with filtered result.
16216 Set value which will be added to filtered result.
16222 Apply a simple postprocessing filter that compresses and decompresses the image
16223 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16224 and average the results.
16226 The filter accepts the following options:
16230 Set quality. This option defines the number of levels for averaging. It accepts
16231 an integer in the range 0-6. If set to @code{0}, the filter will have no
16232 effect. A value of @code{6} means the higher quality. For each increment of
16233 that value the speed drops by a factor of approximately 2. Default value is
16237 Force a constant quantization parameter. If not set, the filter will use the QP
16238 from the video stream (if available).
16241 Set thresholding mode. Available modes are:
16245 Set hard thresholding (default).
16247 Set soft thresholding (better de-ringing effect, but likely blurrier).
16250 @item use_bframe_qp
16251 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16252 option may cause flicker since the B-Frames have often larger QP. Default is
16253 @code{0} (not enabled).
16258 Scale the input by applying one of the super-resolution methods based on
16259 convolutional neural networks. Supported models:
16263 Super-Resolution Convolutional Neural Network model (SRCNN).
16264 See @url{https://arxiv.org/abs/1501.00092}.
16267 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16268 See @url{https://arxiv.org/abs/1609.05158}.
16271 Training scripts as well as scripts for model generation are provided in
16272 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16274 The filter accepts the following options:
16278 Specify which DNN backend to use for model loading and execution. This option accepts
16279 the following values:
16283 Native implementation of DNN loading and execution.
16286 TensorFlow backend. To enable this backend you
16287 need to install the TensorFlow for C library (see
16288 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16289 @code{--enable-libtensorflow}
16292 Default value is @samp{native}.
16295 Set path to model file specifying network architecture and its parameters.
16296 Note that different backends use different file formats. TensorFlow backend
16297 can load files for both formats, while native backend can load files for only
16301 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16302 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16303 input upscaled using bicubic upscaling with proper scale factor.
16309 Draw subtitles on top of input video using the libass library.
16311 To enable compilation of this filter you need to configure FFmpeg with
16312 @code{--enable-libass}. This filter also requires a build with libavcodec and
16313 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16314 Alpha) subtitles format.
16316 The filter accepts the following options:
16320 Set the filename of the subtitle file to read. It must be specified.
16322 @item original_size
16323 Specify the size of the original video, the video for which the ASS file
16324 was composed. For the syntax of this option, check the
16325 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16326 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16327 correctly scale the fonts if the aspect ratio has been changed.
16330 Set a directory path containing fonts that can be used by the filter.
16331 These fonts will be used in addition to whatever the font provider uses.
16334 Process alpha channel, by default alpha channel is untouched.
16337 Set subtitles input character encoding. @code{subtitles} filter only. Only
16338 useful if not UTF-8.
16340 @item stream_index, si
16341 Set subtitles stream index. @code{subtitles} filter only.
16344 Override default style or script info parameters of the subtitles. It accepts a
16345 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16348 If the first key is not specified, it is assumed that the first value
16349 specifies the @option{filename}.
16351 For example, to render the file @file{sub.srt} on top of the input
16352 video, use the command:
16357 which is equivalent to:
16359 subtitles=filename=sub.srt
16362 To render the default subtitles stream from file @file{video.mkv}, use:
16364 subtitles=video.mkv
16367 To render the second subtitles stream from that file, use:
16369 subtitles=video.mkv:si=1
16372 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16373 @code{DejaVu Serif}, use:
16375 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16378 @section super2xsai
16380 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16381 Interpolate) pixel art scaling algorithm.
16383 Useful for enlarging pixel art images without reducing sharpness.
16387 Swap two rectangular objects in video.
16389 This filter accepts the following options:
16399 Set 1st rect x coordinate.
16402 Set 1st rect y coordinate.
16405 Set 2nd rect x coordinate.
16408 Set 2nd rect y coordinate.
16410 All expressions are evaluated once for each frame.
16413 The all options are expressions containing the following constants:
16418 The input width and height.
16421 same as @var{w} / @var{h}
16424 input sample aspect ratio
16427 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16430 The number of the input frame, starting from 0.
16433 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16436 the position in the file of the input frame, NAN if unknown
16444 Apply telecine process to the video.
16446 This filter accepts the following options:
16455 The default value is @code{top}.
16459 A string of numbers representing the pulldown pattern you wish to apply.
16460 The default value is @code{23}.
16464 Some typical patterns:
16469 24p: 2332 (preferred)
16476 24p: 222222222223 ("Euro pulldown")
16483 Apply threshold effect to video stream.
16485 This filter needs four video streams to perform thresholding.
16486 First stream is stream we are filtering.
16487 Second stream is holding threshold values, third stream is holding min values,
16488 and last, fourth stream is holding max values.
16490 The filter accepts the following option:
16494 Set which planes will be processed, unprocessed planes will be copied.
16495 By default value 0xf, all planes will be processed.
16498 For example if first stream pixel's component value is less then threshold value
16499 of pixel component from 2nd threshold stream, third stream value will picked,
16500 otherwise fourth stream pixel component value will be picked.
16502 Using color source filter one can perform various types of thresholding:
16504 @subsection Examples
16508 Binary threshold, using gray color as threshold:
16510 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16514 Inverted binary threshold, using gray color as threshold:
16516 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16520 Truncate binary threshold, using gray color as threshold:
16522 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16526 Threshold to zero, using gray color as threshold:
16528 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16532 Inverted threshold to zero, using gray color as threshold:
16534 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16539 Select the most representative frame in a given sequence of consecutive frames.
16541 The filter accepts the following options:
16545 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16546 will pick one of them, and then handle the next batch of @var{n} frames until
16547 the end. Default is @code{100}.
16550 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16551 value will result in a higher memory usage, so a high value is not recommended.
16553 @subsection Examples
16557 Extract one picture each 50 frames:
16563 Complete example of a thumbnail creation with @command{ffmpeg}:
16565 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16571 Tile several successive frames together.
16573 The filter accepts the following options:
16578 Set the grid size (i.e. the number of lines and columns). For the syntax of
16579 this option, check the
16580 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16583 Set the maximum number of frames to render in the given area. It must be less
16584 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16585 the area will be used.
16588 Set the outer border margin in pixels.
16591 Set the inner border thickness (i.e. the number of pixels between frames). For
16592 more advanced padding options (such as having different values for the edges),
16593 refer to the pad video filter.
16596 Specify the color of the unused area. For the syntax of this option, check the
16597 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16598 The default value of @var{color} is "black".
16601 Set the number of frames to overlap when tiling several successive frames together.
16602 The value must be between @code{0} and @var{nb_frames - 1}.
16605 Set the number of frames to initially be empty before displaying first output frame.
16606 This controls how soon will one get first output frame.
16607 The value must be between @code{0} and @var{nb_frames - 1}.
16610 @subsection Examples
16614 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16616 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16618 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16619 duplicating each output frame to accommodate the originally detected frame
16623 Display @code{5} pictures in an area of @code{3x2} frames,
16624 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16625 mixed flat and named options:
16627 tile=3x2:nb_frames=5:padding=7:margin=2
16631 @section tinterlace
16633 Perform various types of temporal field interlacing.
16635 Frames are counted starting from 1, so the first input frame is
16638 The filter accepts the following options:
16643 Specify the mode of the interlacing. This option can also be specified
16644 as a value alone. See below for a list of values for this option.
16646 Available values are:
16650 Move odd frames into the upper field, even into the lower field,
16651 generating a double height frame at half frame rate.
16655 Frame 1 Frame 2 Frame 3 Frame 4
16657 11111 22222 33333 44444
16658 11111 22222 33333 44444
16659 11111 22222 33333 44444
16660 11111 22222 33333 44444
16674 Only output odd frames, even frames are dropped, generating a frame with
16675 unchanged height at half frame rate.
16680 Frame 1 Frame 2 Frame 3 Frame 4
16682 11111 22222 33333 44444
16683 11111 22222 33333 44444
16684 11111 22222 33333 44444
16685 11111 22222 33333 44444
16695 Only output even frames, odd frames are dropped, generating a frame with
16696 unchanged height at half frame rate.
16701 Frame 1 Frame 2 Frame 3 Frame 4
16703 11111 22222 33333 44444
16704 11111 22222 33333 44444
16705 11111 22222 33333 44444
16706 11111 22222 33333 44444
16716 Expand each frame to full height, but pad alternate lines with black,
16717 generating a frame with double height at the same input frame rate.
16722 Frame 1 Frame 2 Frame 3 Frame 4
16724 11111 22222 33333 44444
16725 11111 22222 33333 44444
16726 11111 22222 33333 44444
16727 11111 22222 33333 44444
16730 11111 ..... 33333 .....
16731 ..... 22222 ..... 44444
16732 11111 ..... 33333 .....
16733 ..... 22222 ..... 44444
16734 11111 ..... 33333 .....
16735 ..... 22222 ..... 44444
16736 11111 ..... 33333 .....
16737 ..... 22222 ..... 44444
16741 @item interleave_top, 4
16742 Interleave the upper field from odd frames with the lower field from
16743 even frames, generating a frame with unchanged height at half frame rate.
16748 Frame 1 Frame 2 Frame 3 Frame 4
16750 11111<- 22222 33333<- 44444
16751 11111 22222<- 33333 44444<-
16752 11111<- 22222 33333<- 44444
16753 11111 22222<- 33333 44444<-
16763 @item interleave_bottom, 5
16764 Interleave the lower field from odd frames with the upper field from
16765 even frames, generating a frame with unchanged height at half frame rate.
16770 Frame 1 Frame 2 Frame 3 Frame 4
16772 11111 22222<- 33333 44444<-
16773 11111<- 22222 33333<- 44444
16774 11111 22222<- 33333 44444<-
16775 11111<- 22222 33333<- 44444
16785 @item interlacex2, 6
16786 Double frame rate with unchanged height. Frames are inserted each
16787 containing the second temporal field from the previous input frame and
16788 the first temporal field from the next input frame. This mode relies on
16789 the top_field_first flag. Useful for interlaced video displays with no
16790 field synchronisation.
16795 Frame 1 Frame 2 Frame 3 Frame 4
16797 11111 22222 33333 44444
16798 11111 22222 33333 44444
16799 11111 22222 33333 44444
16800 11111 22222 33333 44444
16803 11111 22222 22222 33333 33333 44444 44444
16804 11111 11111 22222 22222 33333 33333 44444
16805 11111 22222 22222 33333 33333 44444 44444
16806 11111 11111 22222 22222 33333 33333 44444
16811 Move odd frames into the upper field, even into the lower field,
16812 generating a double height frame at same frame rate.
16817 Frame 1 Frame 2 Frame 3 Frame 4
16819 11111 22222 33333 44444
16820 11111 22222 33333 44444
16821 11111 22222 33333 44444
16822 11111 22222 33333 44444
16825 11111 33333 33333 55555
16826 22222 22222 44444 44444
16827 11111 33333 33333 55555
16828 22222 22222 44444 44444
16829 11111 33333 33333 55555
16830 22222 22222 44444 44444
16831 11111 33333 33333 55555
16832 22222 22222 44444 44444
16837 Numeric values are deprecated but are accepted for backward
16838 compatibility reasons.
16840 Default mode is @code{merge}.
16843 Specify flags influencing the filter process.
16845 Available value for @var{flags} is:
16848 @item low_pass_filter, vlfp
16849 Enable linear vertical low-pass filtering in the filter.
16850 Vertical low-pass filtering is required when creating an interlaced
16851 destination from a progressive source which contains high-frequency
16852 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16855 @item complex_filter, cvlfp
16856 Enable complex vertical low-pass filtering.
16857 This will slightly less reduce interlace 'twitter' and Moire
16858 patterning but better retain detail and subjective sharpness impression.
16862 Vertical low-pass filtering can only be enabled for @option{mode}
16863 @var{interleave_top} and @var{interleave_bottom}.
16869 Mix successive video frames.
16871 A description of the accepted options follows.
16875 The number of successive frames to mix. If unspecified, it defaults to 3.
16878 Specify weight of each input video frame.
16879 Each weight is separated by space. If number of weights is smaller than
16880 number of @var{frames} last specified weight will be used for all remaining
16884 Specify scale, if it is set it will be multiplied with sum
16885 of each weight multiplied with pixel values to give final destination
16886 pixel value. By default @var{scale} is auto scaled to sum of weights.
16889 @subsection Examples
16893 Average 7 successive frames:
16895 tmix=frames=7:weights="1 1 1 1 1 1 1"
16899 Apply simple temporal convolution:
16901 tmix=frames=3:weights="-1 3 -1"
16905 Similar as above but only showing temporal differences:
16907 tmix=frames=3:weights="-1 2 -1":scale=1
16913 Tone map colors from different dynamic ranges.
16915 This filter expects data in single precision floating point, as it needs to
16916 operate on (and can output) out-of-range values. Another filter, such as
16917 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16919 The tonemapping algorithms implemented only work on linear light, so input
16920 data should be linearized beforehand (and possibly correctly tagged).
16923 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16926 @subsection Options
16927 The filter accepts the following options.
16931 Set the tone map algorithm to use.
16933 Possible values are:
16936 Do not apply any tone map, only desaturate overbright pixels.
16939 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16940 in-range values, while distorting out-of-range values.
16943 Stretch the entire reference gamut to a linear multiple of the display.
16946 Fit a logarithmic transfer between the tone curves.
16949 Preserve overall image brightness with a simple curve, using nonlinear
16950 contrast, which results in flattening details and degrading color accuracy.
16953 Preserve both dark and bright details better than @var{reinhard}, at the cost
16954 of slightly darkening everything. Use it when detail preservation is more
16955 important than color and brightness accuracy.
16958 Smoothly map out-of-range values, while retaining contrast and colors for
16959 in-range material as much as possible. Use it when color accuracy is more
16960 important than detail preservation.
16966 Tune the tone mapping algorithm.
16968 This affects the following algorithms:
16974 Specifies the scale factor to use while stretching.
16978 Specifies the exponent of the function.
16982 Specify an extra linear coefficient to multiply into the signal before clipping.
16986 Specify the local contrast coefficient at the display peak.
16987 Default to 0.5, which means that in-gamut values will be about half as bright
16994 Specify the transition point from linear to mobius transform. Every value
16995 below this point is guaranteed to be mapped 1:1. The higher the value, the
16996 more accurate the result will be, at the cost of losing bright details.
16997 Default to 0.3, which due to the steep initial slope still preserves in-range
16998 colors fairly accurately.
17002 Apply desaturation for highlights that exceed this level of brightness. The
17003 higher the parameter, the more color information will be preserved. This
17004 setting helps prevent unnaturally blown-out colors for super-highlights, by
17005 (smoothly) turning into white instead. This makes images feel more natural,
17006 at the cost of reducing information about out-of-range colors.
17008 The default of 2.0 is somewhat conservative and will mostly just apply to
17009 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17011 This option works only if the input frame has a supported color tag.
17014 Override signal/nominal/reference peak with this value. Useful when the
17015 embedded peak information in display metadata is not reliable or when tone
17016 mapping from a lower range to a higher range.
17021 Temporarily pad video frames.
17023 The filter accepts the following options:
17027 Specify number of delay frames before input video stream.
17030 Specify number of padding frames after input video stream.
17031 Set to -1 to pad indefinitely.
17034 Set kind of frames added to beginning of stream.
17035 Can be either @var{add} or @var{clone}.
17036 With @var{add} frames of solid-color are added.
17037 With @var{clone} frames are clones of first frame.
17040 Set kind of frames added to end of stream.
17041 Can be either @var{add} or @var{clone}.
17042 With @var{add} frames of solid-color are added.
17043 With @var{clone} frames are clones of last frame.
17045 @item start_duration, stop_duration
17046 Specify the duration of the start/stop delay. See
17047 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17048 for the accepted syntax.
17049 These options override @var{start} and @var{stop}.
17052 Specify the color of the padded area. For the syntax of this option,
17053 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17054 manual,ffmpeg-utils}.
17056 The default value of @var{color} is "black".
17062 Transpose rows with columns in the input video and optionally flip it.
17064 It accepts the following parameters:
17069 Specify the transposition direction.
17071 Can assume the following values:
17073 @item 0, 4, cclock_flip
17074 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17082 Rotate by 90 degrees clockwise, that is:
17090 Rotate by 90 degrees counterclockwise, that is:
17097 @item 3, 7, clock_flip
17098 Rotate by 90 degrees clockwise and vertically flip, that is:
17106 For values between 4-7, the transposition is only done if the input
17107 video geometry is portrait and not landscape. These values are
17108 deprecated, the @code{passthrough} option should be used instead.
17110 Numerical values are deprecated, and should be dropped in favor of
17111 symbolic constants.
17114 Do not apply the transposition if the input geometry matches the one
17115 specified by the specified value. It accepts the following values:
17118 Always apply transposition.
17120 Preserve portrait geometry (when @var{height} >= @var{width}).
17122 Preserve landscape geometry (when @var{width} >= @var{height}).
17125 Default value is @code{none}.
17128 For example to rotate by 90 degrees clockwise and preserve portrait
17131 transpose=dir=1:passthrough=portrait
17134 The command above can also be specified as:
17136 transpose=1:portrait
17139 @section transpose_npp
17141 Transpose rows with columns in the input video and optionally flip it.
17142 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17144 It accepts the following parameters:
17149 Specify the transposition direction.
17151 Can assume the following values:
17154 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17157 Rotate by 90 degrees clockwise.
17160 Rotate by 90 degrees counterclockwise.
17163 Rotate by 90 degrees clockwise and vertically flip.
17167 Do not apply the transposition if the input geometry matches the one
17168 specified by the specified value. It accepts the following values:
17171 Always apply transposition. (default)
17173 Preserve portrait geometry (when @var{height} >= @var{width}).
17175 Preserve landscape geometry (when @var{width} >= @var{height}).
17181 Trim the input so that the output contains one continuous subpart of the input.
17183 It accepts the following parameters:
17186 Specify the time of the start of the kept section, i.e. the frame with the
17187 timestamp @var{start} will be the first frame in the output.
17190 Specify the time of the first frame that will be dropped, i.e. the frame
17191 immediately preceding the one with the timestamp @var{end} will be the last
17192 frame in the output.
17195 This is the same as @var{start}, except this option sets the start timestamp
17196 in timebase units instead of seconds.
17199 This is the same as @var{end}, except this option sets the end timestamp
17200 in timebase units instead of seconds.
17203 The maximum duration of the output in seconds.
17206 The number of the first frame that should be passed to the output.
17209 The number of the first frame that should be dropped.
17212 @option{start}, @option{end}, and @option{duration} are expressed as time
17213 duration specifications; see
17214 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17215 for the accepted syntax.
17217 Note that the first two sets of the start/end options and the @option{duration}
17218 option look at the frame timestamp, while the _frame variants simply count the
17219 frames that pass through the filter. Also note that this filter does not modify
17220 the timestamps. If you wish for the output timestamps to start at zero, insert a
17221 setpts filter after the trim filter.
17223 If multiple start or end options are set, this filter tries to be greedy and
17224 keep all the frames that match at least one of the specified constraints. To keep
17225 only the part that matches all the constraints at once, chain multiple trim
17228 The defaults are such that all the input is kept. So it is possible to set e.g.
17229 just the end values to keep everything before the specified time.
17234 Drop everything except the second minute of input:
17236 ffmpeg -i INPUT -vf trim=60:120
17240 Keep only the first second:
17242 ffmpeg -i INPUT -vf trim=duration=1
17247 @section unpremultiply
17248 Apply alpha unpremultiply effect to input video stream using first plane
17249 of second stream as alpha.
17251 Both streams must have same dimensions and same pixel format.
17253 The filter accepts the following option:
17257 Set which planes will be processed, unprocessed planes will be copied.
17258 By default value 0xf, all planes will be processed.
17260 If the format has 1 or 2 components, then luma is bit 0.
17261 If the format has 3 or 4 components:
17262 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17263 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17264 If present, the alpha channel is always the last bit.
17267 Do not require 2nd input for processing, instead use alpha plane from input stream.
17273 Sharpen or blur the input video.
17275 It accepts the following parameters:
17278 @item luma_msize_x, lx
17279 Set the luma matrix horizontal size. It must be an odd integer between
17280 3 and 23. The default value is 5.
17282 @item luma_msize_y, ly
17283 Set the luma matrix vertical size. It must be an odd integer between 3
17284 and 23. The default value is 5.
17286 @item luma_amount, la
17287 Set the luma effect strength. It must be a floating point number, reasonable
17288 values lay between -1.5 and 1.5.
17290 Negative values will blur the input video, while positive values will
17291 sharpen it, a value of zero will disable the effect.
17293 Default value is 1.0.
17295 @item chroma_msize_x, cx
17296 Set the chroma matrix horizontal size. It must be an odd integer
17297 between 3 and 23. The default value is 5.
17299 @item chroma_msize_y, cy
17300 Set the chroma matrix vertical size. It must be an odd integer
17301 between 3 and 23. The default value is 5.
17303 @item chroma_amount, ca
17304 Set the chroma effect strength. It must be a floating point number, reasonable
17305 values lay between -1.5 and 1.5.
17307 Negative values will blur the input video, while positive values will
17308 sharpen it, a value of zero will disable the effect.
17310 Default value is 0.0.
17314 All parameters are optional and default to the equivalent of the
17315 string '5:5:1.0:5:5:0.0'.
17317 @subsection Examples
17321 Apply strong luma sharpen effect:
17323 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17327 Apply a strong blur of both luma and chroma parameters:
17329 unsharp=7:7:-2:7:7:-2
17335 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17336 the image at several (or - in the case of @option{quality} level @code{8} - all)
17337 shifts and average the results.
17339 The way this differs from the behavior of spp is that uspp actually encodes &
17340 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17341 DCT similar to MJPEG.
17343 The filter accepts the following options:
17347 Set quality. This option defines the number of levels for averaging. It accepts
17348 an integer in the range 0-8. If set to @code{0}, the filter will have no
17349 effect. A value of @code{8} means the higher quality. For each increment of
17350 that value the speed drops by a factor of approximately 2. Default value is
17354 Force a constant quantization parameter. If not set, the filter will use the QP
17355 from the video stream (if available).
17358 @section vaguedenoiser
17360 Apply a wavelet based denoiser.
17362 It transforms each frame from the video input into the wavelet domain,
17363 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17364 the obtained coefficients. It does an inverse wavelet transform after.
17365 Due to wavelet properties, it should give a nice smoothed result, and
17366 reduced noise, without blurring picture features.
17368 This filter accepts the following options:
17372 The filtering strength. The higher, the more filtered the video will be.
17373 Hard thresholding can use a higher threshold than soft thresholding
17374 before the video looks overfiltered. Default value is 2.
17377 The filtering method the filter will use.
17379 It accepts the following values:
17382 All values under the threshold will be zeroed.
17385 All values under the threshold will be zeroed. All values above will be
17386 reduced by the threshold.
17389 Scales or nullifies coefficients - intermediary between (more) soft and
17390 (less) hard thresholding.
17393 Default is garrote.
17396 Number of times, the wavelet will decompose the picture. Picture can't
17397 be decomposed beyond a particular point (typically, 8 for a 640x480
17398 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17401 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17404 A list of the planes to process. By default all planes are processed.
17407 @section vectorscope
17409 Display 2 color component values in the two dimensional graph (which is called
17412 This filter accepts the following options:
17416 Set vectorscope mode.
17418 It accepts the following values:
17421 Gray values are displayed on graph, higher brightness means more pixels have
17422 same component color value on location in graph. This is the default mode.
17425 Gray values are displayed on graph. Surrounding pixels values which are not
17426 present in video frame are drawn in gradient of 2 color components which are
17427 set by option @code{x} and @code{y}. The 3rd color component is static.
17430 Actual color components values present in video frame are displayed on graph.
17433 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17434 on graph increases value of another color component, which is luminance by
17435 default values of @code{x} and @code{y}.
17438 Actual colors present in video frame are displayed on graph. If two different
17439 colors map to same position on graph then color with higher value of component
17440 not present in graph is picked.
17443 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17444 component picked from radial gradient.
17448 Set which color component will be represented on X-axis. Default is @code{1}.
17451 Set which color component will be represented on Y-axis. Default is @code{2}.
17454 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17455 of color component which represents frequency of (X, Y) location in graph.
17460 No envelope, this is default.
17463 Instant envelope, even darkest single pixel will be clearly highlighted.
17466 Hold maximum and minimum values presented in graph over time. This way you
17467 can still spot out of range values without constantly looking at vectorscope.
17470 Peak and instant envelope combined together.
17474 Set what kind of graticule to draw.
17482 Set graticule opacity.
17485 Set graticule flags.
17489 Draw graticule for white point.
17492 Draw graticule for black point.
17495 Draw color points short names.
17499 Set background opacity.
17501 @item lthreshold, l
17502 Set low threshold for color component not represented on X or Y axis.
17503 Values lower than this value will be ignored. Default is 0.
17504 Note this value is multiplied with actual max possible value one pixel component
17505 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17508 @item hthreshold, h
17509 Set high threshold for color component not represented on X or Y axis.
17510 Values higher than this value will be ignored. Default is 1.
17511 Note this value is multiplied with actual max possible value one pixel component
17512 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17513 is 0.9 * 255 = 230.
17515 @item colorspace, c
17516 Set what kind of colorspace to use when drawing graticule.
17525 @anchor{vidstabdetect}
17526 @section vidstabdetect
17528 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17529 @ref{vidstabtransform} for pass 2.
17531 This filter generates a file with relative translation and rotation
17532 transform information about subsequent frames, which is then used by
17533 the @ref{vidstabtransform} filter.
17535 To enable compilation of this filter you need to configure FFmpeg with
17536 @code{--enable-libvidstab}.
17538 This filter accepts the following options:
17542 Set the path to the file used to write the transforms information.
17543 Default value is @file{transforms.trf}.
17546 Set how shaky the video is and how quick the camera is. It accepts an
17547 integer in the range 1-10, a value of 1 means little shakiness, a
17548 value of 10 means strong shakiness. Default value is 5.
17551 Set the accuracy of the detection process. It must be a value in the
17552 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17553 accuracy. Default value is 15.
17556 Set stepsize of the search process. The region around minimum is
17557 scanned with 1 pixel resolution. Default value is 6.
17560 Set minimum contrast. Below this value a local measurement field is
17561 discarded. Must be a floating point value in the range 0-1. Default
17565 Set reference frame number for tripod mode.
17567 If enabled, the motion of the frames is compared to a reference frame
17568 in the filtered stream, identified by the specified number. The idea
17569 is to compensate all movements in a more-or-less static scene and keep
17570 the camera view absolutely still.
17572 If set to 0, it is disabled. The frames are counted starting from 1.
17575 Show fields and transforms in the resulting frames. It accepts an
17576 integer in the range 0-2. Default value is 0, which disables any
17580 @subsection Examples
17584 Use default values:
17590 Analyze strongly shaky movie and put the results in file
17591 @file{mytransforms.trf}:
17593 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17597 Visualize the result of internal transformations in the resulting
17600 vidstabdetect=show=1
17604 Analyze a video with medium shakiness using @command{ffmpeg}:
17606 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17610 @anchor{vidstabtransform}
17611 @section vidstabtransform
17613 Video stabilization/deshaking: pass 2 of 2,
17614 see @ref{vidstabdetect} for pass 1.
17616 Read a file with transform information for each frame and
17617 apply/compensate them. Together with the @ref{vidstabdetect}
17618 filter this can be used to deshake videos. See also
17619 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17620 the @ref{unsharp} filter, see below.
17622 To enable compilation of this filter you need to configure FFmpeg with
17623 @code{--enable-libvidstab}.
17625 @subsection Options
17629 Set path to the file used to read the transforms. Default value is
17630 @file{transforms.trf}.
17633 Set the number of frames (value*2 + 1) used for lowpass filtering the
17634 camera movements. Default value is 10.
17636 For example a number of 10 means that 21 frames are used (10 in the
17637 past and 10 in the future) to smoothen the motion in the video. A
17638 larger value leads to a smoother video, but limits the acceleration of
17639 the camera (pan/tilt movements). 0 is a special case where a static
17640 camera is simulated.
17643 Set the camera path optimization algorithm.
17645 Accepted values are:
17648 gaussian kernel low-pass filter on camera motion (default)
17650 averaging on transformations
17654 Set maximal number of pixels to translate frames. Default value is -1,
17658 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17659 value is -1, meaning no limit.
17662 Specify how to deal with borders that may be visible due to movement
17665 Available values are:
17668 keep image information from previous frame (default)
17670 fill the border black
17674 Invert transforms if set to 1. Default value is 0.
17677 Consider transforms as relative to previous frame if set to 1,
17678 absolute if set to 0. Default value is 0.
17681 Set percentage to zoom. A positive value will result in a zoom-in
17682 effect, a negative value in a zoom-out effect. Default value is 0 (no
17686 Set optimal zooming to avoid borders.
17688 Accepted values are:
17693 optimal static zoom value is determined (only very strong movements
17694 will lead to visible borders) (default)
17696 optimal adaptive zoom value is determined (no borders will be
17697 visible), see @option{zoomspeed}
17700 Note that the value given at zoom is added to the one calculated here.
17703 Set percent to zoom maximally each frame (enabled when
17704 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17708 Specify type of interpolation.
17710 Available values are:
17715 linear only horizontal
17717 linear in both directions (default)
17719 cubic in both directions (slow)
17723 Enable virtual tripod mode if set to 1, which is equivalent to
17724 @code{relative=0:smoothing=0}. Default value is 0.
17726 Use also @code{tripod} option of @ref{vidstabdetect}.
17729 Increase log verbosity if set to 1. Also the detected global motions
17730 are written to the temporary file @file{global_motions.trf}. Default
17734 @subsection Examples
17738 Use @command{ffmpeg} for a typical stabilization with default values:
17740 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17743 Note the use of the @ref{unsharp} filter which is always recommended.
17746 Zoom in a bit more and load transform data from a given file:
17748 vidstabtransform=zoom=5:input="mytransforms.trf"
17752 Smoothen the video even more:
17754 vidstabtransform=smoothing=30
17760 Flip the input video vertically.
17762 For example, to vertically flip a video with @command{ffmpeg}:
17764 ffmpeg -i in.avi -vf "vflip" out.avi
17769 Detect variable frame rate video.
17771 This filter tries to detect if the input is variable or constant frame rate.
17773 At end it will output number of frames detected as having variable delta pts,
17774 and ones with constant delta pts.
17775 If there was frames with variable delta, than it will also show min and max delta
17780 Boost or alter saturation.
17782 The filter accepts the following options:
17785 Set strength of boost if positive value or strength of alter if negative value.
17786 Default is 0. Allowed range is from -2 to 2.
17789 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17792 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17795 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17798 Set the red luma coefficient.
17801 Set the green luma coefficient.
17804 Set the blue luma coefficient.
17810 Make or reverse a natural vignetting effect.
17812 The filter accepts the following options:
17816 Set lens angle expression as a number of radians.
17818 The value is clipped in the @code{[0,PI/2]} range.
17820 Default value: @code{"PI/5"}
17824 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17828 Set forward/backward mode.
17830 Available modes are:
17833 The larger the distance from the central point, the darker the image becomes.
17836 The larger the distance from the central point, the brighter the image becomes.
17837 This can be used to reverse a vignette effect, though there is no automatic
17838 detection to extract the lens @option{angle} and other settings (yet). It can
17839 also be used to create a burning effect.
17842 Default value is @samp{forward}.
17845 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17847 It accepts the following values:
17850 Evaluate expressions only once during the filter initialization.
17853 Evaluate expressions for each incoming frame. This is way slower than the
17854 @samp{init} mode since it requires all the scalers to be re-computed, but it
17855 allows advanced dynamic expressions.
17858 Default value is @samp{init}.
17861 Set dithering to reduce the circular banding effects. Default is @code{1}
17865 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17866 Setting this value to the SAR of the input will make a rectangular vignetting
17867 following the dimensions of the video.
17869 Default is @code{1/1}.
17872 @subsection Expressions
17874 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17875 following parameters.
17880 input width and height
17883 the number of input frame, starting from 0
17886 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17887 @var{TB} units, NAN if undefined
17890 frame rate of the input video, NAN if the input frame rate is unknown
17893 the PTS (Presentation TimeStamp) of the filtered video frame,
17894 expressed in seconds, NAN if undefined
17897 time base of the input video
17901 @subsection Examples
17905 Apply simple strong vignetting effect:
17911 Make a flickering vignetting:
17913 vignette='PI/4+random(1)*PI/50':eval=frame
17918 @section vmafmotion
17920 Obtain the average vmaf motion score of a video.
17921 It is one of the component filters of VMAF.
17923 The obtained average motion score is printed through the logging system.
17925 In the below example the input file @file{ref.mpg} is being processed and score
17929 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
17933 Stack input videos vertically.
17935 All streams must be of same pixel format and of same width.
17937 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
17938 to create same output.
17940 The filter accept the following option:
17944 Set number of input streams. Default is 2.
17947 If set to 1, force the output to terminate when the shortest input
17948 terminates. Default value is 0.
17953 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17954 Deinterlacing Filter").
17956 Based on the process described by Martin Weston for BBC R&D, and
17957 implemented based on the de-interlace algorithm written by Jim
17958 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17959 uses filter coefficients calculated by BBC R&D.
17961 There are two sets of filter coefficients, so called "simple":
17962 and "complex". Which set of filter coefficients is used can
17963 be set by passing an optional parameter:
17967 Set the interlacing filter coefficients. Accepts one of the following values:
17971 Simple filter coefficient set.
17973 More-complex filter coefficient set.
17975 Default value is @samp{complex}.
17978 Specify which frames to deinterlace. Accept one of the following values:
17982 Deinterlace all frames,
17984 Only deinterlace frames marked as interlaced.
17987 Default value is @samp{all}.
17991 Video waveform monitor.
17993 The waveform monitor plots color component intensity. By default luminance
17994 only. Each column of the waveform corresponds to a column of pixels in the
17997 It accepts the following options:
18001 Can be either @code{row}, or @code{column}. Default is @code{column}.
18002 In row mode, the graph on the left side represents color component value 0 and
18003 the right side represents value = 255. In column mode, the top side represents
18004 color component value = 0 and bottom side represents value = 255.
18007 Set intensity. Smaller values are useful to find out how many values of the same
18008 luminance are distributed across input rows/columns.
18009 Default value is @code{0.04}. Allowed range is [0, 1].
18012 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18013 In mirrored mode, higher values will be represented on the left
18014 side for @code{row} mode and at the top for @code{column} mode. Default is
18015 @code{1} (mirrored).
18019 It accepts the following values:
18022 Presents information identical to that in the @code{parade}, except
18023 that the graphs representing color components are superimposed directly
18026 This display mode makes it easier to spot relative differences or similarities
18027 in overlapping areas of the color components that are supposed to be identical,
18028 such as neutral whites, grays, or blacks.
18031 Display separate graph for the color components side by side in
18032 @code{row} mode or one below the other in @code{column} mode.
18035 Display separate graph for the color components side by side in
18036 @code{column} mode or one below the other in @code{row} mode.
18038 Using this display mode makes it easy to spot color casts in the highlights
18039 and shadows of an image, by comparing the contours of the top and the bottom
18040 graphs of each waveform. Since whites, grays, and blacks are characterized
18041 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18042 should display three waveforms of roughly equal width/height. If not, the
18043 correction is easy to perform by making level adjustments the three waveforms.
18045 Default is @code{stack}.
18047 @item components, c
18048 Set which color components to display. Default is 1, which means only luminance
18049 or red color component if input is in RGB colorspace. If is set for example to
18050 7 it will display all 3 (if) available color components.
18055 No envelope, this is default.
18058 Instant envelope, minimum and maximum values presented in graph will be easily
18059 visible even with small @code{step} value.
18062 Hold minimum and maximum values presented in graph across time. This way you
18063 can still spot out of range values without constantly looking at waveforms.
18066 Peak and instant envelope combined together.
18072 No filtering, this is default.
18075 Luma and chroma combined together.
18078 Similar as above, but shows difference between blue and red chroma.
18081 Similar as above, but use different colors.
18084 Displays only chroma.
18087 Displays actual color value on waveform.
18090 Similar as above, but with luma showing frequency of chroma values.
18094 Set which graticule to display.
18098 Do not display graticule.
18101 Display green graticule showing legal broadcast ranges.
18104 Display orange graticule showing legal broadcast ranges.
18108 Set graticule opacity.
18111 Set graticule flags.
18115 Draw numbers above lines. By default enabled.
18118 Draw dots instead of lines.
18122 Set scale used for displaying graticule.
18129 Default is digital.
18132 Set background opacity.
18135 @section weave, doubleweave
18137 The @code{weave} takes a field-based video input and join
18138 each two sequential fields into single frame, producing a new double
18139 height clip with half the frame rate and half the frame count.
18141 The @code{doubleweave} works same as @code{weave} but without
18142 halving frame rate and frame count.
18144 It accepts the following option:
18148 Set first field. Available values are:
18152 Set the frame as top-field-first.
18155 Set the frame as bottom-field-first.
18159 @subsection Examples
18163 Interlace video using @ref{select} and @ref{separatefields} filter:
18165 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18170 Apply the xBR high-quality magnification filter which is designed for pixel
18171 art. It follows a set of edge-detection rules, see
18172 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18174 It accepts the following option:
18178 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18179 @code{3xBR} and @code{4} for @code{4xBR}.
18180 Default is @code{3}.
18184 Stack video inputs into custom layout.
18186 All streams must be of same pixel format.
18188 The filter accept the following option:
18192 Set number of input streams. Default is 2.
18195 Specify layout of inputs.
18196 This option requires the desired layout configuration to be explicitly set by the user.
18197 This sets position of each video input in output. Each input
18198 is separated by '|'.
18199 The first number represents the column, and the second number represents the row.
18200 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18201 where X is video input from which to take width or height.
18202 Multiple values can be used when separated by '+'. In such
18203 case values are summed together.
18206 If set to 1, force the output to terminate when the shortest input
18207 terminates. Default value is 0.
18210 @subsection Examples
18214 Display 4 inputs into 2x2 grid,
18215 note that if inputs are of different sizes unused gaps might appear,
18216 as not all of output video is used.
18218 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18222 Display 4 inputs into 1x4 grid,
18223 note that if inputs are of different sizes unused gaps might appear,
18224 as not all of output video is used.
18226 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18230 Display 9 inputs into 3x3 grid,
18231 note that if inputs are of different sizes unused gaps might appear,
18232 as not all of output video is used.
18234 xstack=inputs=9:layout=w3_0|w3_h0+h2|w3_h0|0_h4|0_0|w3+w1_0|0_h1+h2|w3+w1_h0|w3+w1_h1+h2
18241 Deinterlace the input video ("yadif" means "yet another deinterlacing
18244 It accepts the following parameters:
18250 The interlacing mode to adopt. It accepts one of the following values:
18253 @item 0, send_frame
18254 Output one frame for each frame.
18255 @item 1, send_field
18256 Output one frame for each field.
18257 @item 2, send_frame_nospatial
18258 Like @code{send_frame}, but it skips the spatial interlacing check.
18259 @item 3, send_field_nospatial
18260 Like @code{send_field}, but it skips the spatial interlacing check.
18263 The default value is @code{send_frame}.
18266 The picture field parity assumed for the input interlaced video. It accepts one
18267 of the following values:
18271 Assume the top field is first.
18273 Assume the bottom field is first.
18275 Enable automatic detection of field parity.
18278 The default value is @code{auto}.
18279 If the interlacing is unknown or the decoder does not export this information,
18280 top field first will be assumed.
18283 Specify which frames to deinterlace. Accept one of the following
18288 Deinterlace all frames.
18289 @item 1, interlaced
18290 Only deinterlace frames marked as interlaced.
18293 The default value is @code{all}.
18296 @section yadif_cuda
18298 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18299 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18302 It accepts the following parameters:
18308 The interlacing mode to adopt. It accepts one of the following values:
18311 @item 0, send_frame
18312 Output one frame for each frame.
18313 @item 1, send_field
18314 Output one frame for each field.
18315 @item 2, send_frame_nospatial
18316 Like @code{send_frame}, but it skips the spatial interlacing check.
18317 @item 3, send_field_nospatial
18318 Like @code{send_field}, but it skips the spatial interlacing check.
18321 The default value is @code{send_frame}.
18324 The picture field parity assumed for the input interlaced video. It accepts one
18325 of the following values:
18329 Assume the top field is first.
18331 Assume the bottom field is first.
18333 Enable automatic detection of field parity.
18336 The default value is @code{auto}.
18337 If the interlacing is unknown or the decoder does not export this information,
18338 top field first will be assumed.
18341 Specify which frames to deinterlace. Accept one of the following
18346 Deinterlace all frames.
18347 @item 1, interlaced
18348 Only deinterlace frames marked as interlaced.
18351 The default value is @code{all}.
18356 Apply Zoom & Pan effect.
18358 This filter accepts the following options:
18362 Set the zoom expression. Default is 1.
18366 Set the x and y expression. Default is 0.
18369 Set the duration expression in number of frames.
18370 This sets for how many number of frames effect will last for
18371 single input image.
18374 Set the output image size, default is 'hd720'.
18377 Set the output frame rate, default is '25'.
18380 Each expression can contain the following constants:
18399 Output frame count.
18403 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18404 for current input frame.
18408 'x' and 'y' of last output frame of previous input frame or 0 when there was
18409 not yet such frame (first input frame).
18412 Last calculated zoom from 'z' expression for current input frame.
18415 Last calculated zoom of last output frame of previous input frame.
18418 Number of output frames for current input frame. Calculated from 'd' expression
18419 for each input frame.
18422 number of output frames created for previous input frame
18425 Rational number: input width / input height
18428 sample aspect ratio
18431 display aspect ratio
18435 @subsection Examples
18439 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18441 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
18445 Zoom-in up to 1.5 and pan always at center of picture:
18447 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18451 Same as above but without pausing:
18453 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18459 Scale (resize) the input video, using the z.lib library:
18460 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18461 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18463 The zscale filter forces the output display aspect ratio to be the same
18464 as the input, by changing the output sample aspect ratio.
18466 If the input image format is different from the format requested by
18467 the next filter, the zscale filter will convert the input to the
18470 @subsection Options
18471 The filter accepts the following options.
18476 Set the output video dimension expression. Default value is the input
18479 If the @var{width} or @var{w} value is 0, the input width is used for
18480 the output. If the @var{height} or @var{h} value is 0, the input height
18481 is used for the output.
18483 If one and only one of the values is -n with n >= 1, the zscale filter
18484 will use a value that maintains the aspect ratio of the input image,
18485 calculated from the other specified dimension. After that it will,
18486 however, make sure that the calculated dimension is divisible by n and
18487 adjust the value if necessary.
18489 If both values are -n with n >= 1, the behavior will be identical to
18490 both values being set to 0 as previously detailed.
18492 See below for the list of accepted constants for use in the dimension
18496 Set the video size. For the syntax of this option, check the
18497 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18500 Set the dither type.
18502 Possible values are:
18507 @item error_diffusion
18513 Set the resize filter type.
18515 Possible values are:
18525 Default is bilinear.
18528 Set the color range.
18530 Possible values are:
18537 Default is same as input.
18540 Set the color primaries.
18542 Possible values are:
18552 Default is same as input.
18555 Set the transfer characteristics.
18557 Possible values are:
18571 Default is same as input.
18574 Set the colorspace matrix.
18576 Possible value are:
18587 Default is same as input.
18590 Set the input color range.
18592 Possible values are:
18599 Default is same as input.
18601 @item primariesin, pin
18602 Set the input color primaries.
18604 Possible values are:
18614 Default is same as input.
18616 @item transferin, tin
18617 Set the input transfer characteristics.
18619 Possible values are:
18630 Default is same as input.
18632 @item matrixin, min
18633 Set the input colorspace matrix.
18635 Possible value are:
18647 Set the output chroma location.
18649 Possible values are:
18660 @item chromalin, cin
18661 Set the input chroma location.
18663 Possible values are:
18675 Set the nominal peak luminance.
18678 The values of the @option{w} and @option{h} options are expressions
18679 containing the following constants:
18684 The input width and height
18688 These are the same as @var{in_w} and @var{in_h}.
18692 The output (scaled) width and height
18696 These are the same as @var{out_w} and @var{out_h}
18699 The same as @var{iw} / @var{ih}
18702 input sample aspect ratio
18705 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18709 horizontal and vertical input chroma subsample values. For example for the
18710 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18714 horizontal and vertical output chroma subsample values. For example for the
18715 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18721 @c man end VIDEO FILTERS
18723 @chapter OpenCL Video Filters
18724 @c man begin OPENCL VIDEO FILTERS
18726 Below is a description of the currently available OpenCL video filters.
18728 To enable compilation of these filters you need to configure FFmpeg with
18729 @code{--enable-opencl}.
18731 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18734 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18735 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18736 given device parameters.
18738 @item -filter_hw_device @var{name}
18739 Pass the hardware device called @var{name} to all filters in any filter graph.
18743 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18747 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18749 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18753 Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(@ref{hwupload}) to hardware surfaces connected to the appropriate device before being used and then downloaded(@ref{hwdownload}) back to normal memory. Note that @ref{hwupload} will upload to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before to get the input into the right format and @ref{hwdownload} does not support all formats on the output - it may be necessary to insert an additional @ref{format} filter immediately following in the graph to get the output in a supported format.
18755 @section avgblur_opencl
18757 Apply average blur filter.
18759 The filter accepts the following options:
18763 Set horizontal radius size.
18764 Range is @code{[1, 1024]} and default value is @code{1}.
18767 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18770 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18773 @subsection Example
18777 Apply average blur filter with horizontal and vertical size of 3, setting each pixel of the output to the average value of the 7x7 region centered on it in the input. For pixels on the edges of the image, the region does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
18779 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18783 @section boxblur_opencl
18785 Apply a boxblur algorithm to the input video.
18787 It accepts the following parameters:
18791 @item luma_radius, lr
18792 @item luma_power, lp
18793 @item chroma_radius, cr
18794 @item chroma_power, cp
18795 @item alpha_radius, ar
18796 @item alpha_power, ap
18800 A description of the accepted options follows.
18803 @item luma_radius, lr
18804 @item chroma_radius, cr
18805 @item alpha_radius, ar
18806 Set an expression for the box radius in pixels used for blurring the
18807 corresponding input plane.
18809 The radius value must be a non-negative number, and must not be
18810 greater than the value of the expression @code{min(w,h)/2} for the
18811 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18814 Default value for @option{luma_radius} is "2". If not specified,
18815 @option{chroma_radius} and @option{alpha_radius} default to the
18816 corresponding value set for @option{luma_radius}.
18818 The expressions can contain the following constants:
18822 The input width and height in pixels.
18826 The input chroma image width and height in pixels.
18830 The horizontal and vertical chroma subsample values. For example, for the
18831 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18834 @item luma_power, lp
18835 @item chroma_power, cp
18836 @item alpha_power, ap
18837 Specify how many times the boxblur filter is applied to the
18838 corresponding plane.
18840 Default value for @option{luma_power} is 2. If not specified,
18841 @option{chroma_power} and @option{alpha_power} default to the
18842 corresponding value set for @option{luma_power}.
18844 A value of 0 will disable the effect.
18847 @subsection Examples
18849 Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses @var{luma_radius}, @var{chroma_radius}, @var{alpha_radius} for each plane respectively. The filter will apply @var{luma_power}, @var{chroma_power}, @var{alpha_power} times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
18853 Apply a boxblur filter with the luma, chroma, and alpha radius
18854 set to 2 and luma, chroma, and alpha power set to 3. The filter will run 3 times with box-radius set to 2 for every plane of the image.
18856 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18857 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18861 Apply a boxblur filter with luma radius set to 2, luma_power to 1, chroma_radius to 4, chroma_power to 5, alpha_radius to 3 and alpha_power to 7.
18863 For the luma plane, a 2x2 box radius will be run once.
18865 For the chroma plane, a 4x4 box radius will be run 5 times.
18867 For the alpha plane, a 3x3 box radius will be run 7 times.
18869 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18873 @section convolution_opencl
18875 Apply convolution of 3x3, 5x5, 7x7 matrix.
18877 The filter accepts the following options:
18884 Set matrix for each plane.
18885 Matrix is sequence of 9, 25 or 49 signed numbers.
18886 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18892 Set multiplier for calculated value for each plane.
18893 If unset or 0, it will be sum of all matrix elements.
18894 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
18900 Set bias for each plane. This value is added to the result of the multiplication.
18901 Useful for making the overall image brighter or darker.
18902 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
18906 @subsection Examples
18912 -i INPUT -vf "hwupload, convolution_opencl=0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0, hwdownload" OUTPUT
18918 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9, hwdownload" OUTPUT
18922 Apply edge enhance:
18924 -i INPUT -vf "hwupload, convolution_opencl=0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128, hwdownload" OUTPUT
18930 -i INPUT -vf "hwupload, convolution_opencl=0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128, hwdownload" OUTPUT
18934 Apply laplacian edge detector which includes diagonals:
18936 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
18942 -i INPUT -vf "hwupload, convolution_opencl=-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2, hwdownload" OUTPUT
18946 @section dilation_opencl
18948 Apply dilation effect to the video.
18950 This filter replaces the pixel by the local(3x3) maximum.
18952 It accepts the following options:
18959 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18960 If @code{0}, plane will remain unchanged.
18963 Flag which specifies the pixel to refer to.
18964 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
18966 Flags to local 3x3 coordinates region centered on @code{x}:
18975 @subsection Example
18979 Apply dilation filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local maximum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local maximum is more then threshold of the corresponding plane, output pixel will be set to input pixel + threshold of corresponding plane.
18981 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
18985 @section erosion_opencl
18987 Apply erosion effect to the video.
18989 This filter replaces the pixel by the local(3x3) minimum.
18991 It accepts the following options:
18998 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18999 If @code{0}, plane will remain unchanged.
19002 Flag which specifies the pixel to refer to.
19003 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19005 Flags to local 3x3 coordinates region centered on @code{x}:
19014 @subsection Example
19018 Apply erosion filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local minimum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local minimum is more then threshold of the corresponding plane, output pixel will be set to input pixel - threshold of corresponding plane.
19020 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19024 @section overlay_opencl
19026 Overlay one video on top of another.
19028 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19029 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19031 The filter accepts the following options:
19036 Set the x coordinate of the overlaid video on the main video.
19037 Default value is @code{0}.
19040 Set the x coordinate of the overlaid video on the main video.
19041 Default value is @code{0}.
19045 @subsection Examples
19049 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19051 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19054 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19056 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19061 @section prewitt_opencl
19063 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19065 The filter accepts the following option:
19069 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19072 Set value which will be multiplied with filtered result.
19073 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19076 Set value which will be added to filtered result.
19077 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19080 @subsection Example
19084 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19086 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19090 @section roberts_opencl
19091 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19093 The filter accepts the following option:
19097 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19100 Set value which will be multiplied with filtered result.
19101 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19104 Set value which will be added to filtered result.
19105 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19108 @subsection Example
19112 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19114 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19118 @section sobel_opencl
19120 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19122 The filter accepts the following option:
19126 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19129 Set value which will be multiplied with filtered result.
19130 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19133 Set value which will be added to filtered result.
19134 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19137 @subsection Example
19141 Apply sobel operator with scale set to 2 and delta set to 10
19143 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19147 @section tonemap_opencl
19149 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19151 It accepts the following parameters:
19155 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19158 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19161 Apply desaturation for highlights that exceed this level of brightness. The
19162 higher the parameter, the more color information will be preserved. This
19163 setting helps prevent unnaturally blown-out colors for super-highlights, by
19164 (smoothly) turning into white instead. This makes images feel more natural,
19165 at the cost of reducing information about out-of-range colors.
19167 The default value is 0.5, and the algorithm here is a little different from
19168 the cpu version tonemap currently. A setting of 0.0 disables this option.
19171 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19172 is used to detect whether the scene has changed or not. If the distance between
19173 the current frame average brightness and the current running average exceeds
19174 a threshold value, we would re-calculate scene average and peak brightness.
19175 The default value is 0.2.
19178 Specify the output pixel format.
19180 Currently supported formats are:
19187 Set the output color range.
19189 Possible values are:
19195 Default is same as input.
19198 Set the output color primaries.
19200 Possible values are:
19206 Default is same as input.
19209 Set the output transfer characteristics.
19211 Possible values are:
19220 Set the output colorspace matrix.
19222 Possible value are:
19228 Default is same as input.
19232 @subsection Example
19236 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19238 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19242 @section unsharp_opencl
19244 Sharpen or blur the input video.
19246 It accepts the following parameters:
19249 @item luma_msize_x, lx
19250 Set the luma matrix horizontal size.
19251 Range is @code{[1, 23]} and default value is @code{5}.
19253 @item luma_msize_y, ly
19254 Set the luma matrix vertical size.
19255 Range is @code{[1, 23]} and default value is @code{5}.
19257 @item luma_amount, la
19258 Set the luma effect strength.
19259 Range is @code{[-10, 10]} and default value is @code{1.0}.
19261 Negative values will blur the input video, while positive values will
19262 sharpen it, a value of zero will disable the effect.
19264 @item chroma_msize_x, cx
19265 Set the chroma matrix horizontal size.
19266 Range is @code{[1, 23]} and default value is @code{5}.
19268 @item chroma_msize_y, cy
19269 Set the chroma matrix vertical size.
19270 Range is @code{[1, 23]} and default value is @code{5}.
19272 @item chroma_amount, ca
19273 Set the chroma effect strength.
19274 Range is @code{[-10, 10]} and default value is @code{0.0}.
19276 Negative values will blur the input video, while positive values will
19277 sharpen it, a value of zero will disable the effect.
19281 All parameters are optional and default to the equivalent of the
19282 string '5:5:1.0:5:5:0.0'.
19284 @subsection Examples
19288 Apply strong luma sharpen effect:
19290 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19294 Apply a strong blur of both luma and chroma parameters:
19296 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19300 @c man end OPENCL VIDEO FILTERS
19302 @chapter Video Sources
19303 @c man begin VIDEO SOURCES
19305 Below is a description of the currently available video sources.
19309 Buffer video frames, and make them available to the filter chain.
19311 This source is mainly intended for a programmatic use, in particular
19312 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19314 It accepts the following parameters:
19319 Specify the size (width and height) of the buffered video frames. For the
19320 syntax of this option, check the
19321 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19324 The input video width.
19327 The input video height.
19330 A string representing the pixel format of the buffered video frames.
19331 It may be a number corresponding to a pixel format, or a pixel format
19335 Specify the timebase assumed by the timestamps of the buffered frames.
19338 Specify the frame rate expected for the video stream.
19340 @item pixel_aspect, sar
19341 The sample (pixel) aspect ratio of the input video.
19344 Specify the optional parameters to be used for the scale filter which
19345 is automatically inserted when an input change is detected in the
19346 input size or format.
19348 @item hw_frames_ctx
19349 When using a hardware pixel format, this should be a reference to an
19350 AVHWFramesContext describing input frames.
19355 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19358 will instruct the source to accept video frames with size 320x240 and
19359 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19360 square pixels (1:1 sample aspect ratio).
19361 Since the pixel format with name "yuv410p" corresponds to the number 6
19362 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19363 this example corresponds to:
19365 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19368 Alternatively, the options can be specified as a flat string, but this
19369 syntax is deprecated:
19371 @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}]
19375 Create a pattern generated by an elementary cellular automaton.
19377 The initial state of the cellular automaton can be defined through the
19378 @option{filename} and @option{pattern} options. If such options are
19379 not specified an initial state is created randomly.
19381 At each new frame a new row in the video is filled with the result of
19382 the cellular automaton next generation. The behavior when the whole
19383 frame is filled is defined by the @option{scroll} option.
19385 This source accepts the following options:
19389 Read the initial cellular automaton state, i.e. the starting row, from
19390 the specified file.
19391 In the file, each non-whitespace character is considered an alive
19392 cell, a newline will terminate the row, and further characters in the
19393 file will be ignored.
19396 Read the initial cellular automaton state, i.e. the starting row, from
19397 the specified string.
19399 Each non-whitespace character in the string is considered an alive
19400 cell, a newline will terminate the row, and further characters in the
19401 string will be ignored.
19404 Set the video rate, that is the number of frames generated per second.
19407 @item random_fill_ratio, ratio
19408 Set the random fill ratio for the initial cellular automaton row. It
19409 is a floating point number value ranging from 0 to 1, defaults to
19412 This option is ignored when a file or a pattern is specified.
19414 @item random_seed, seed
19415 Set the seed for filling randomly the initial row, must be an integer
19416 included between 0 and UINT32_MAX. If not specified, or if explicitly
19417 set to -1, the filter will try to use a good random seed on a best
19421 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19422 Default value is 110.
19425 Set the size of the output video. For the syntax of this option, check the
19426 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19428 If @option{filename} or @option{pattern} is specified, the size is set
19429 by default to the width of the specified initial state row, and the
19430 height is set to @var{width} * PHI.
19432 If @option{size} is set, it must contain the width of the specified
19433 pattern string, and the specified pattern will be centered in the
19436 If a filename or a pattern string is not specified, the size value
19437 defaults to "320x518" (used for a randomly generated initial state).
19440 If set to 1, scroll the output upward when all the rows in the output
19441 have been already filled. If set to 0, the new generated row will be
19442 written over the top row just after the bottom row is filled.
19445 @item start_full, full
19446 If set to 1, completely fill the output with generated rows before
19447 outputting the first frame.
19448 This is the default behavior, for disabling set the value to 0.
19451 If set to 1, stitch the left and right row edges together.
19452 This is the default behavior, for disabling set the value to 0.
19455 @subsection Examples
19459 Read the initial state from @file{pattern}, and specify an output of
19462 cellauto=f=pattern:s=200x400
19466 Generate a random initial row with a width of 200 cells, with a fill
19469 cellauto=ratio=2/3:s=200x200
19473 Create a pattern generated by rule 18 starting by a single alive cell
19474 centered on an initial row with width 100:
19476 cellauto=p=@@:s=100x400:full=0:rule=18
19480 Specify a more elaborated initial pattern:
19482 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19487 @anchor{coreimagesrc}
19488 @section coreimagesrc
19489 Video source generated on GPU using Apple's CoreImage API on OSX.
19491 This video source is a specialized version of the @ref{coreimage} video filter.
19492 Use a core image generator at the beginning of the applied filterchain to
19493 generate the content.
19495 The coreimagesrc video source accepts the following options:
19497 @item list_generators
19498 List all available generators along with all their respective options as well as
19499 possible minimum and maximum values along with the default values.
19501 list_generators=true
19505 Specify the size of the sourced video. For the syntax of this option, check the
19506 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19507 The default value is @code{320x240}.
19510 Specify the frame rate of the sourced video, as the number of frames
19511 generated per second. It has to be a string in the format
19512 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19513 number or a valid video frame rate abbreviation. The default value is
19517 Set the sample aspect ratio of the sourced video.
19520 Set the duration of the sourced video. See
19521 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19522 for the accepted syntax.
19524 If not specified, or the expressed duration is negative, the video is
19525 supposed to be generated forever.
19528 Additionally, all options of the @ref{coreimage} video filter are accepted.
19529 A complete filterchain can be used for further processing of the
19530 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19531 and examples for details.
19533 @subsection Examples
19538 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19539 given as complete and escaped command-line for Apple's standard bash shell:
19541 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19543 This example is equivalent to the QRCode example of @ref{coreimage} without the
19544 need for a nullsrc video source.
19548 @section mandelbrot
19550 Generate a Mandelbrot set fractal, and progressively zoom towards the
19551 point specified with @var{start_x} and @var{start_y}.
19553 This source accepts the following options:
19558 Set the terminal pts value. Default value is 400.
19561 Set the terminal scale value.
19562 Must be a floating point value. Default value is 0.3.
19565 Set the inner coloring mode, that is the algorithm used to draw the
19566 Mandelbrot fractal internal region.
19568 It shall assume one of the following values:
19573 Show time until convergence.
19575 Set color based on point closest to the origin of the iterations.
19580 Default value is @var{mincol}.
19583 Set the bailout value. Default value is 10.0.
19586 Set the maximum of iterations performed by the rendering
19587 algorithm. Default value is 7189.
19590 Set outer coloring mode.
19591 It shall assume one of following values:
19593 @item iteration_count
19594 Set iteration count mode.
19595 @item normalized_iteration_count
19596 set normalized iteration count mode.
19598 Default value is @var{normalized_iteration_count}.
19601 Set frame rate, expressed as number of frames per second. Default
19605 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19606 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19609 Set the initial scale value. Default value is 3.0.
19612 Set the initial x position. Must be a floating point value between
19613 -100 and 100. Default value is -0.743643887037158704752191506114774.
19616 Set the initial y position. Must be a floating point value between
19617 -100 and 100. Default value is -0.131825904205311970493132056385139.
19622 Generate various test patterns, as generated by the MPlayer test filter.
19624 The size of the generated video is fixed, and is 256x256.
19625 This source is useful in particular for testing encoding features.
19627 This source accepts the following options:
19632 Specify the frame rate of the sourced video, as the number of frames
19633 generated per second. It has to be a string in the format
19634 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19635 number or a valid video frame rate abbreviation. The default value is
19639 Set the duration of the sourced video. See
19640 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19641 for the accepted syntax.
19643 If not specified, or the expressed duration is negative, the video is
19644 supposed to be generated forever.
19648 Set the number or the name of the test to perform. Supported tests are:
19664 Default value is "all", which will cycle through the list of all tests.
19669 mptestsrc=t=dc_luma
19672 will generate a "dc_luma" test pattern.
19674 @section frei0r_src
19676 Provide a frei0r source.
19678 To enable compilation of this filter you need to install the frei0r
19679 header and configure FFmpeg with @code{--enable-frei0r}.
19681 This source accepts the following parameters:
19686 The size of the video to generate. For the syntax of this option, check the
19687 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19690 The framerate of the generated video. It may be a string of the form
19691 @var{num}/@var{den} or a frame rate abbreviation.
19694 The name to the frei0r source to load. For more information regarding frei0r and
19695 how to set the parameters, read the @ref{frei0r} section in the video filters
19698 @item filter_params
19699 A '|'-separated list of parameters to pass to the frei0r source.
19703 For example, to generate a frei0r partik0l source with size 200x200
19704 and frame rate 10 which is overlaid on the overlay filter main input:
19706 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19711 Generate a life pattern.
19713 This source is based on a generalization of John Conway's life game.
19715 The sourced input represents a life grid, each pixel represents a cell
19716 which can be in one of two possible states, alive or dead. Every cell
19717 interacts with its eight neighbours, which are the cells that are
19718 horizontally, vertically, or diagonally adjacent.
19720 At each interaction the grid evolves according to the adopted rule,
19721 which specifies the number of neighbor alive cells which will make a
19722 cell stay alive or born. The @option{rule} option allows one to specify
19725 This source accepts the following options:
19729 Set the file from which to read the initial grid state. In the file,
19730 each non-whitespace character is considered an alive cell, and newline
19731 is used to delimit the end of each row.
19733 If this option is not specified, the initial grid is generated
19737 Set the video rate, that is the number of frames generated per second.
19740 @item random_fill_ratio, ratio
19741 Set the random fill ratio for the initial random grid. It is a
19742 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19743 It is ignored when a file is specified.
19745 @item random_seed, seed
19746 Set the seed for filling the initial random grid, must be an integer
19747 included between 0 and UINT32_MAX. If not specified, or if explicitly
19748 set to -1, the filter will try to use a good random seed on a best
19754 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19755 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19756 @var{NS} specifies the number of alive neighbor cells which make a
19757 live cell stay alive, and @var{NB} the number of alive neighbor cells
19758 which make a dead cell to become alive (i.e. to "born").
19759 "s" and "b" can be used in place of "S" and "B", respectively.
19761 Alternatively a rule can be specified by an 18-bits integer. The 9
19762 high order bits are used to encode the next cell state if it is alive
19763 for each number of neighbor alive cells, the low order bits specify
19764 the rule for "borning" new cells. Higher order bits encode for an
19765 higher number of neighbor cells.
19766 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19767 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19769 Default value is "S23/B3", which is the original Conway's game of life
19770 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19771 cells, and will born a new cell if there are three alive cells around
19775 Set the size of the output video. For the syntax of this option, check the
19776 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19778 If @option{filename} is specified, the size is set by default to the
19779 same size of the input file. If @option{size} is set, it must contain
19780 the size specified in the input file, and the initial grid defined in
19781 that file is centered in the larger resulting area.
19783 If a filename is not specified, the size value defaults to "320x240"
19784 (used for a randomly generated initial grid).
19787 If set to 1, stitch the left and right grid edges together, and the
19788 top and bottom edges also. Defaults to 1.
19791 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19792 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19793 value from 0 to 255.
19796 Set the color of living (or new born) cells.
19799 Set the color of dead cells. If @option{mold} is set, this is the first color
19800 used to represent a dead cell.
19803 Set mold color, for definitely dead and moldy cells.
19805 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19806 ffmpeg-utils manual,ffmpeg-utils}.
19809 @subsection Examples
19813 Read a grid from @file{pattern}, and center it on a grid of size
19816 life=f=pattern:s=300x300
19820 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19822 life=ratio=2/3:s=200x200
19826 Specify a custom rule for evolving a randomly generated grid:
19832 Full example with slow death effect (mold) using @command{ffplay}:
19834 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19841 @anchor{haldclutsrc}
19844 @anchor{pal100bars}
19845 @anchor{rgbtestsrc}
19847 @anchor{smptehdbars}
19850 @anchor{yuvtestsrc}
19851 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19853 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19855 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19857 The @code{color} source provides an uniformly colored input.
19859 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19860 @ref{haldclut} filter.
19862 The @code{nullsrc} source returns unprocessed video frames. It is
19863 mainly useful to be employed in analysis / debugging tools, or as the
19864 source for filters which ignore the input data.
19866 The @code{pal75bars} source generates a color bars pattern, based on
19867 EBU PAL recommendations with 75% color levels.
19869 The @code{pal100bars} source generates a color bars pattern, based on
19870 EBU PAL recommendations with 100% color levels.
19872 The @code{rgbtestsrc} source generates an RGB test pattern useful for
19873 detecting RGB vs BGR issues. You should see a red, green and blue
19874 stripe from top to bottom.
19876 The @code{smptebars} source generates a color bars pattern, based on
19877 the SMPTE Engineering Guideline EG 1-1990.
19879 The @code{smptehdbars} source generates a color bars pattern, based on
19880 the SMPTE RP 219-2002.
19882 The @code{testsrc} source generates a test video pattern, showing a
19883 color pattern, a scrolling gradient and a timestamp. This is mainly
19884 intended for testing purposes.
19886 The @code{testsrc2} source is similar to testsrc, but supports more
19887 pixel formats instead of just @code{rgb24}. This allows using it as an
19888 input for other tests without requiring a format conversion.
19890 The @code{yuvtestsrc} source generates an YUV test pattern. You should
19891 see a y, cb and cr stripe from top to bottom.
19893 The sources accept the following parameters:
19898 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
19899 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
19900 pixels to be used as identity matrix for 3D lookup tables. Each component is
19901 coded on a @code{1/(N*N)} scale.
19904 Specify the color of the source, only available in the @code{color}
19905 source. For the syntax of this option, check the
19906 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19909 Specify the size of the sourced video. For the syntax of this option, check the
19910 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19911 The default value is @code{320x240}.
19913 This option is not available with the @code{allrgb}, @code{allyuv}, and
19914 @code{haldclutsrc} filters.
19917 Specify the frame rate of the sourced video, as the number of frames
19918 generated per second. It has to be a string in the format
19919 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19920 number or a valid video frame rate abbreviation. The default value is
19924 Set the duration of the sourced video. See
19925 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19926 for the accepted syntax.
19928 If not specified, or the expressed duration is negative, the video is
19929 supposed to be generated forever.
19932 Set the sample aspect ratio of the sourced video.
19935 Specify the alpha (opacity) of the background, only available in the
19936 @code{testsrc2} source. The value must be between 0 (fully transparent) and
19937 255 (fully opaque, the default).
19940 Set the number of decimals to show in the timestamp, only available in the
19941 @code{testsrc} source.
19943 The displayed timestamp value will correspond to the original
19944 timestamp value multiplied by the power of 10 of the specified
19945 value. Default value is 0.
19948 @subsection Examples
19952 Generate a video with a duration of 5.3 seconds, with size
19953 176x144 and a frame rate of 10 frames per second:
19955 testsrc=duration=5.3:size=qcif:rate=10
19959 The following graph description will generate a red source
19960 with an opacity of 0.2, with size "qcif" and a frame rate of 10
19963 color=c=red@@0.2:s=qcif:r=10
19967 If the input content is to be ignored, @code{nullsrc} can be used. The
19968 following command generates noise in the luminance plane by employing
19969 the @code{geq} filter:
19971 nullsrc=s=256x256, geq=random(1)*255:128:128
19975 @subsection Commands
19977 The @code{color} source supports the following commands:
19981 Set the color of the created image. Accepts the same syntax of the
19982 corresponding @option{color} option.
19987 Generate video using an OpenCL program.
19992 OpenCL program source file.
19995 Kernel name in program.
19998 Size of frames to generate. This must be set.
20001 Pixel format to use for the generated frames. This must be set.
20004 Number of frames generated every second. Default value is '25'.
20008 For details of how the program loading works, see the @ref{program_opencl}
20015 Generate a colour ramp by setting pixel values from the position of the pixel
20016 in the output image. (Note that this will work with all pixel formats, but
20017 the generated output will not be the same.)
20019 __kernel void ramp(__write_only image2d_t dst,
20020 unsigned int index)
20022 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20025 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20027 write_imagef(dst, loc, val);
20032 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20034 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20035 unsigned int index)
20037 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20039 float4 value = 0.0f;
20040 int x = loc.x + index;
20041 int y = loc.y + index;
20042 while (x > 0 || y > 0) {
20043 if (x % 3 == 1 && y % 3 == 1) {
20051 write_imagef(dst, loc, value);
20057 @c man end VIDEO SOURCES
20059 @chapter Video Sinks
20060 @c man begin VIDEO SINKS
20062 Below is a description of the currently available video sinks.
20064 @section buffersink
20066 Buffer video frames, and make them available to the end of the filter
20069 This sink is mainly intended for programmatic use, in particular
20070 through the interface defined in @file{libavfilter/buffersink.h}
20071 or the options system.
20073 It accepts a pointer to an AVBufferSinkContext structure, which
20074 defines the incoming buffers' formats, to be passed as the opaque
20075 parameter to @code{avfilter_init_filter} for initialization.
20079 Null video sink: do absolutely nothing with the input video. It is
20080 mainly useful as a template and for use in analysis / debugging
20083 @c man end VIDEO SINKS
20085 @chapter Multimedia Filters
20086 @c man begin MULTIMEDIA FILTERS
20088 Below is a description of the currently available multimedia filters.
20092 Convert input audio to a video output, displaying the audio bit scope.
20094 The filter accepts the following options:
20098 Set frame rate, expressed as number of frames per second. Default
20102 Specify the video size for the output. For the syntax of this option, check the
20103 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20104 Default value is @code{1024x256}.
20107 Specify list of colors separated by space or by '|' which will be used to
20108 draw channels. Unrecognized or missing colors will be replaced
20112 @section ahistogram
20114 Convert input audio to a video output, displaying the volume histogram.
20116 The filter accepts the following options:
20120 Specify how histogram is calculated.
20122 It accepts the following values:
20125 Use single histogram for all channels.
20127 Use separate histogram for each channel.
20129 Default is @code{single}.
20132 Set frame rate, expressed as number of frames per second. Default
20136 Specify the video size for the output. For the syntax of this option, check the
20137 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20138 Default value is @code{hd720}.
20143 It accepts the following values:
20154 reverse logarithmic
20156 Default is @code{log}.
20159 Set amplitude scale.
20161 It accepts the following values:
20168 Default is @code{log}.
20171 Set how much frames to accumulate in histogram.
20172 Default is 1. Setting this to -1 accumulates all frames.
20175 Set histogram ratio of window height.
20178 Set sonogram sliding.
20180 It accepts the following values:
20183 replace old rows with new ones.
20185 scroll from top to bottom.
20187 Default is @code{replace}.
20190 @section aphasemeter
20192 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20193 representing mean phase of current audio frame. A video output can also be produced and is
20194 enabled by default. The audio is passed through as first output.
20196 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20197 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20198 and @code{1} means channels are in phase.
20200 The filter accepts the following options, all related to its video output:
20204 Set the output frame rate. Default value is @code{25}.
20207 Set the video size for the output. For the syntax of this option, check the
20208 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20209 Default value is @code{800x400}.
20214 Specify the red, green, blue contrast. Default values are @code{2},
20215 @code{7} and @code{1}.
20216 Allowed range is @code{[0, 255]}.
20219 Set color which will be used for drawing median phase. If color is
20220 @code{none} which is default, no median phase value will be drawn.
20223 Enable video output. Default is enabled.
20226 @section avectorscope
20228 Convert input audio to a video output, representing the audio vector
20231 The filter is used to measure the difference between channels of stereo
20232 audio stream. A monoaural signal, consisting of identical left and right
20233 signal, results in straight vertical line. Any stereo separation is visible
20234 as a deviation from this line, creating a Lissajous figure.
20235 If the straight (or deviation from it) but horizontal line appears this
20236 indicates that the left and right channels are out of phase.
20238 The filter accepts the following options:
20242 Set the vectorscope mode.
20244 Available values are:
20247 Lissajous rotated by 45 degrees.
20250 Same as above but not rotated.
20253 Shape resembling half of circle.
20256 Default value is @samp{lissajous}.
20259 Set the video size for the output. For the syntax of this option, check the
20260 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20261 Default value is @code{400x400}.
20264 Set the output frame rate. Default value is @code{25}.
20270 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20271 @code{160}, @code{80} and @code{255}.
20272 Allowed range is @code{[0, 255]}.
20278 Specify the red, green, blue and alpha fade. Default values are @code{15},
20279 @code{10}, @code{5} and @code{5}.
20280 Allowed range is @code{[0, 255]}.
20283 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20284 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20287 Set the vectorscope drawing mode.
20289 Available values are:
20292 Draw dot for each sample.
20295 Draw line between previous and current sample.
20298 Default value is @samp{dot}.
20301 Specify amplitude scale of audio samples.
20303 Available values are:
20319 Swap left channel axis with right channel axis.
20329 Mirror only x axis.
20332 Mirror only y axis.
20340 @subsection Examples
20344 Complete example using @command{ffplay}:
20346 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20347 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20351 @section bench, abench
20353 Benchmark part of a filtergraph.
20355 The filter accepts the following options:
20359 Start or stop a timer.
20361 Available values are:
20364 Get the current time, set it as frame metadata (using the key
20365 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20368 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20369 the input frame metadata to get the time difference. Time difference, average,
20370 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20371 @code{min}) are then printed. The timestamps are expressed in seconds.
20375 @subsection Examples
20379 Benchmark @ref{selectivecolor} filter:
20381 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20387 Concatenate audio and video streams, joining them together one after the
20390 The filter works on segments of synchronized video and audio streams. All
20391 segments must have the same number of streams of each type, and that will
20392 also be the number of streams at output.
20394 The filter accepts the following options:
20399 Set the number of segments. Default is 2.
20402 Set the number of output video streams, that is also the number of video
20403 streams in each segment. Default is 1.
20406 Set the number of output audio streams, that is also the number of audio
20407 streams in each segment. Default is 0.
20410 Activate unsafe mode: do not fail if segments have a different format.
20414 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20415 @var{a} audio outputs.
20417 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20418 segment, in the same order as the outputs, then the inputs for the second
20421 Related streams do not always have exactly the same duration, for various
20422 reasons including codec frame size or sloppy authoring. For that reason,
20423 related synchronized streams (e.g. a video and its audio track) should be
20424 concatenated at once. The concat filter will use the duration of the longest
20425 stream in each segment (except the last one), and if necessary pad shorter
20426 audio streams with silence.
20428 For this filter to work correctly, all segments must start at timestamp 0.
20430 All corresponding streams must have the same parameters in all segments; the
20431 filtering system will automatically select a common pixel format for video
20432 streams, and a common sample format, sample rate and channel layout for
20433 audio streams, but other settings, such as resolution, must be converted
20434 explicitly by the user.
20436 Different frame rates are acceptable but will result in variable frame rate
20437 at output; be sure to configure the output file to handle it.
20439 @subsection Examples
20443 Concatenate an opening, an episode and an ending, all in bilingual version
20444 (video in stream 0, audio in streams 1 and 2):
20446 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20447 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20448 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20449 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20453 Concatenate two parts, handling audio and video separately, using the
20454 (a)movie sources, and adjusting the resolution:
20456 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20457 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20458 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20460 Note that a desync will happen at the stitch if the audio and video streams
20461 do not have exactly the same duration in the first file.
20465 @subsection Commands
20467 This filter supports the following commands:
20470 Close the current segment and step to the next one
20473 @section drawgraph, adrawgraph
20475 Draw a graph using input video or audio metadata.
20477 It accepts the following parameters:
20481 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20484 Set 1st foreground color expression.
20487 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20490 Set 2nd foreground color expression.
20493 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20496 Set 3rd foreground color expression.
20499 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20502 Set 4th foreground color expression.
20505 Set minimal value of metadata value.
20508 Set maximal value of metadata value.
20511 Set graph background color. Default is white.
20516 Available values for mode is:
20523 Default is @code{line}.
20528 Available values for slide is:
20531 Draw new frame when right border is reached.
20534 Replace old columns with new ones.
20537 Scroll from right to left.
20540 Scroll from left to right.
20543 Draw single picture.
20546 Default is @code{frame}.
20549 Set size of graph video. For the syntax of this option, check the
20550 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20551 The default value is @code{900x256}.
20553 The foreground color expressions can use the following variables:
20556 Minimal value of metadata value.
20559 Maximal value of metadata value.
20562 Current metadata key value.
20565 The color is defined as 0xAABBGGRR.
20568 Example using metadata from @ref{signalstats} filter:
20570 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20573 Example using metadata from @ref{ebur128} filter:
20575 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20581 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
20582 it unchanged. By default, it logs a message at a frequency of 10Hz with the
20583 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20584 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20586 The filter also has a video output (see the @var{video} option) with a real
20587 time graph to observe the loudness evolution. The graphic contains the logged
20588 message mentioned above, so it is not printed anymore when this option is set,
20589 unless the verbose logging is set. The main graphing area contains the
20590 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20591 the momentary loudness (400 milliseconds), but can optionally be configured
20592 to instead display short-term loudness (see @var{gauge}).
20594 The green area marks a +/- 1LU target range around the target loudness
20595 (-23LUFS by default, unless modified through @var{target}).
20597 More information about the Loudness Recommendation EBU R128 on
20598 @url{http://tech.ebu.ch/loudness}.
20600 The filter accepts the following options:
20605 Activate the video output. The audio stream is passed unchanged whether this
20606 option is set or no. The video stream will be the first output stream if
20607 activated. Default is @code{0}.
20610 Set the video size. This option is for video only. For the syntax of this
20612 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20613 Default and minimum resolution is @code{640x480}.
20616 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20617 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20618 other integer value between this range is allowed.
20621 Set metadata injection. If set to @code{1}, the audio input will be segmented
20622 into 100ms output frames, each of them containing various loudness information
20623 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20625 Default is @code{0}.
20628 Force the frame logging level.
20630 Available values are:
20633 information logging level
20635 verbose logging level
20638 By default, the logging level is set to @var{info}. If the @option{video} or
20639 the @option{metadata} options are set, it switches to @var{verbose}.
20644 Available modes can be cumulated (the option is a @code{flag} type). Possible
20648 Disable any peak mode (default).
20650 Enable sample-peak mode.
20652 Simple peak mode looking for the higher sample value. It logs a message
20653 for sample-peak (identified by @code{SPK}).
20655 Enable true-peak mode.
20657 If enabled, the peak lookup is done on an over-sampled version of the input
20658 stream for better peak accuracy. It logs a message for true-peak.
20659 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20660 This mode requires a build with @code{libswresample}.
20664 Treat mono input files as "dual mono". If a mono file is intended for playback
20665 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20666 If set to @code{true}, this option will compensate for this effect.
20667 Multi-channel input files are not affected by this option.
20670 Set a specific pan law to be used for the measurement of dual mono files.
20671 This parameter is optional, and has a default value of -3.01dB.
20674 Set a specific target level (in LUFS) used as relative zero in the visualization.
20675 This parameter is optional and has a default value of -23LUFS as specified
20676 by EBU R128. However, material published online may prefer a level of -16LUFS
20677 (e.g. for use with podcasts or video platforms).
20680 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20681 @code{shortterm}. By default the momentary value will be used, but in certain
20682 scenarios it may be more useful to observe the short term value instead (e.g.
20686 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20687 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20688 video output, not the summary or continuous log output.
20691 @subsection Examples
20695 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20697 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20701 Run an analysis with @command{ffmpeg}:
20703 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20707 @section interleave, ainterleave
20709 Temporally interleave frames from several inputs.
20711 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20713 These filters read frames from several inputs and send the oldest
20714 queued frame to the output.
20716 Input streams must have well defined, monotonically increasing frame
20719 In order to submit one frame to output, these filters need to enqueue
20720 at least one frame for each input, so they cannot work in case one
20721 input is not yet terminated and will not receive incoming frames.
20723 For example consider the case when one input is a @code{select} filter
20724 which always drops input frames. The @code{interleave} filter will keep
20725 reading from that input, but it will never be able to send new frames
20726 to output until the input sends an end-of-stream signal.
20728 Also, depending on inputs synchronization, the filters will drop
20729 frames in case one input receives more frames than the other ones, and
20730 the queue is already filled.
20732 These filters accept the following options:
20736 Set the number of different inputs, it is 2 by default.
20739 @subsection Examples
20743 Interleave frames belonging to different streams using @command{ffmpeg}:
20745 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20749 Add flickering blur effect:
20751 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20755 @section metadata, ametadata
20757 Manipulate frame metadata.
20759 This filter accepts the following options:
20763 Set mode of operation of the filter.
20765 Can be one of the following:
20769 If both @code{value} and @code{key} is set, select frames
20770 which have such metadata. If only @code{key} is set, select
20771 every frame that has such key in metadata.
20774 Add new metadata @code{key} and @code{value}. If key is already available
20778 Modify value of already present key.
20781 If @code{value} is set, delete only keys that have such value.
20782 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20786 Print key and its value if metadata was found. If @code{key} is not set print all
20787 metadata values available in frame.
20791 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20794 Set metadata value which will be used. This option is mandatory for
20795 @code{modify} and @code{add} mode.
20798 Which function to use when comparing metadata value and @code{value}.
20800 Can be one of following:
20804 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20807 Values are interpreted as strings, returns true if metadata value starts with
20808 the @code{value} option string.
20811 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20814 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20817 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20820 Values are interpreted as floats, returns true if expression from option @code{expr}
20825 Set expression which is used when @code{function} is set to @code{expr}.
20826 The expression is evaluated through the eval API and can contain the following
20831 Float representation of @code{value} from metadata key.
20834 Float representation of @code{value} as supplied by user in @code{value} option.
20838 If specified in @code{print} mode, output is written to the named file. Instead of
20839 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20840 for standard output. If @code{file} option is not set, output is written to the log
20841 with AV_LOG_INFO loglevel.
20845 @subsection Examples
20849 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20852 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20855 Print silencedetect output to file @file{metadata.txt}.
20857 silencedetect,ametadata=mode=print:file=metadata.txt
20860 Direct all metadata to a pipe with file descriptor 4.
20862 metadata=mode=print:file='pipe\:4'
20866 @section perms, aperms
20868 Set read/write permissions for the output frames.
20870 These filters are mainly aimed at developers to test direct path in the
20871 following filter in the filtergraph.
20873 The filters accept the following options:
20877 Select the permissions mode.
20879 It accepts the following values:
20882 Do nothing. This is the default.
20884 Set all the output frames read-only.
20886 Set all the output frames directly writable.
20888 Make the frame read-only if writable, and writable if read-only.
20890 Set each output frame read-only or writable randomly.
20894 Set the seed for the @var{random} mode, must be an integer included between
20895 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
20896 @code{-1}, the filter will try to use a good random seed on a best effort
20900 Note: in case of auto-inserted filter between the permission filter and the
20901 following one, the permission might not be received as expected in that
20902 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
20903 perms/aperms filter can avoid this problem.
20905 @section realtime, arealtime
20907 Slow down filtering to match real time approximately.
20909 These filters will pause the filtering for a variable amount of time to
20910 match the output rate with the input timestamps.
20911 They are similar to the @option{re} option to @code{ffmpeg}.
20913 They accept the following options:
20917 Time limit for the pauses. Any pause longer than that will be considered
20918 a timestamp discontinuity and reset the timer. Default is 2 seconds.
20922 @section select, aselect
20924 Select frames to pass in output.
20926 This filter accepts the following options:
20931 Set expression, which is evaluated for each input frame.
20933 If the expression is evaluated to zero, the frame is discarded.
20935 If the evaluation result is negative or NaN, the frame is sent to the
20936 first output; otherwise it is sent to the output with index
20937 @code{ceil(val)-1}, assuming that the input index starts from 0.
20939 For example a value of @code{1.2} corresponds to the output with index
20940 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
20943 Set the number of outputs. The output to which to send the selected
20944 frame is based on the result of the evaluation. Default value is 1.
20947 The expression can contain the following constants:
20951 The (sequential) number of the filtered frame, starting from 0.
20954 The (sequential) number of the selected frame, starting from 0.
20956 @item prev_selected_n
20957 The sequential number of the last selected frame. It's NAN if undefined.
20960 The timebase of the input timestamps.
20963 The PTS (Presentation TimeStamp) of the filtered video frame,
20964 expressed in @var{TB} units. It's NAN if undefined.
20967 The PTS of the filtered video frame,
20968 expressed in seconds. It's NAN if undefined.
20971 The PTS of the previously filtered video frame. It's NAN if undefined.
20973 @item prev_selected_pts
20974 The PTS of the last previously filtered video frame. It's NAN if undefined.
20976 @item prev_selected_t
20977 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
20980 The PTS of the first video frame in the video. It's NAN if undefined.
20983 The time of the first video frame in the video. It's NAN if undefined.
20985 @item pict_type @emph{(video only)}
20986 The type of the filtered frame. It can assume one of the following
20998 @item interlace_type @emph{(video only)}
20999 The frame interlace type. It can assume one of the following values:
21002 The frame is progressive (not interlaced).
21004 The frame is top-field-first.
21006 The frame is bottom-field-first.
21009 @item consumed_sample_n @emph{(audio only)}
21010 the number of selected samples before the current frame
21012 @item samples_n @emph{(audio only)}
21013 the number of samples in the current frame
21015 @item sample_rate @emph{(audio only)}
21016 the input sample rate
21019 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21022 the position in the file of the filtered frame, -1 if the information
21023 is not available (e.g. for synthetic video)
21025 @item scene @emph{(video only)}
21026 value between 0 and 1 to indicate a new scene; a low value reflects a low
21027 probability for the current frame to introduce a new scene, while a higher
21028 value means the current frame is more likely to be one (see the example below)
21030 @item concatdec_select
21031 The concat demuxer can select only part of a concat input file by setting an
21032 inpoint and an outpoint, but the output packets may not be entirely contained
21033 in the selected interval. By using this variable, it is possible to skip frames
21034 generated by the concat demuxer which are not exactly contained in the selected
21037 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21038 and the @var{lavf.concat.duration} packet metadata values which are also
21039 present in the decoded frames.
21041 The @var{concatdec_select} variable is -1 if the frame pts is at least
21042 start_time and either the duration metadata is missing or the frame pts is less
21043 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21046 That basically means that an input frame is selected if its pts is within the
21047 interval set by the concat demuxer.
21051 The default value of the select expression is "1".
21053 @subsection Examples
21057 Select all frames in input:
21062 The example above is the same as:
21074 Select only I-frames:
21076 select='eq(pict_type\,I)'
21080 Select one frame every 100:
21082 select='not(mod(n\,100))'
21086 Select only frames contained in the 10-20 time interval:
21088 select=between(t\,10\,20)
21092 Select only I-frames contained in the 10-20 time interval:
21094 select=between(t\,10\,20)*eq(pict_type\,I)
21098 Select frames with a minimum distance of 10 seconds:
21100 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21104 Use aselect to select only audio frames with samples number > 100:
21106 aselect='gt(samples_n\,100)'
21110 Create a mosaic of the first scenes:
21112 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21115 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21119 Send even and odd frames to separate outputs, and compose them:
21121 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21125 Select useful frames from an ffconcat file which is using inpoints and
21126 outpoints but where the source files are not intra frame only.
21128 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21132 @section sendcmd, asendcmd
21134 Send commands to filters in the filtergraph.
21136 These filters read commands to be sent to other filters in the
21139 @code{sendcmd} must be inserted between two video filters,
21140 @code{asendcmd} must be inserted between two audio filters, but apart
21141 from that they act the same way.
21143 The specification of commands can be provided in the filter arguments
21144 with the @var{commands} option, or in a file specified by the
21145 @var{filename} option.
21147 These filters accept the following options:
21150 Set the commands to be read and sent to the other filters.
21152 Set the filename of the commands to be read and sent to the other
21156 @subsection Commands syntax
21158 A commands description consists of a sequence of interval
21159 specifications, comprising a list of commands to be executed when a
21160 particular event related to that interval occurs. The occurring event
21161 is typically the current frame time entering or leaving a given time
21164 An interval is specified by the following syntax:
21166 @var{START}[-@var{END}] @var{COMMANDS};
21169 The time interval is specified by the @var{START} and @var{END} times.
21170 @var{END} is optional and defaults to the maximum time.
21172 The current frame time is considered within the specified interval if
21173 it is included in the interval [@var{START}, @var{END}), that is when
21174 the time is greater or equal to @var{START} and is lesser than
21177 @var{COMMANDS} consists of a sequence of one or more command
21178 specifications, separated by ",", relating to that interval. The
21179 syntax of a command specification is given by:
21181 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21184 @var{FLAGS} is optional and specifies the type of events relating to
21185 the time interval which enable sending the specified command, and must
21186 be a non-null sequence of identifier flags separated by "+" or "|" and
21187 enclosed between "[" and "]".
21189 The following flags are recognized:
21192 The command is sent when the current frame timestamp enters the
21193 specified interval. In other words, the command is sent when the
21194 previous frame timestamp was not in the given interval, and the
21198 The command is sent when the current frame timestamp leaves the
21199 specified interval. In other words, the command is sent when the
21200 previous frame timestamp was in the given interval, and the
21204 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21207 @var{TARGET} specifies the target of the command, usually the name of
21208 the filter class or a specific filter instance name.
21210 @var{COMMAND} specifies the name of the command for the target filter.
21212 @var{ARG} is optional and specifies the optional list of argument for
21213 the given @var{COMMAND}.
21215 Between one interval specification and another, whitespaces, or
21216 sequences of characters starting with @code{#} until the end of line,
21217 are ignored and can be used to annotate comments.
21219 A simplified BNF description of the commands specification syntax
21222 @var{COMMAND_FLAG} ::= "enter" | "leave"
21223 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21224 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21225 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21226 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21227 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21230 @subsection Examples
21234 Specify audio tempo change at second 4:
21236 asendcmd=c='4.0 atempo tempo 1.5',atempo
21240 Target a specific filter instance:
21242 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21246 Specify a list of drawtext and hue commands in a file.
21248 # show text in the interval 5-10
21249 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21250 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21252 # desaturate the image in the interval 15-20
21253 15.0-20.0 [enter] hue s 0,
21254 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21256 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21258 # apply an exponential saturation fade-out effect, starting from time 25
21259 25 [enter] hue s exp(25-t)
21262 A filtergraph allowing to read and process the above command list
21263 stored in a file @file{test.cmd}, can be specified with:
21265 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21270 @section setpts, asetpts
21272 Change the PTS (presentation timestamp) of the input frames.
21274 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21276 This filter accepts the following options:
21281 The expression which is evaluated for each frame to construct its timestamp.
21285 The expression is evaluated through the eval API and can contain the following
21289 @item FRAME_RATE, FR
21290 frame rate, only defined for constant frame-rate video
21293 The presentation timestamp in input
21296 The count of the input frame for video or the number of consumed samples,
21297 not including the current frame for audio, starting from 0.
21299 @item NB_CONSUMED_SAMPLES
21300 The number of consumed samples, not including the current frame (only
21303 @item NB_SAMPLES, S
21304 The number of samples in the current frame (only audio)
21306 @item SAMPLE_RATE, SR
21307 The audio sample rate.
21310 The PTS of the first frame.
21313 the time in seconds of the first frame
21316 State whether the current frame is interlaced.
21319 the time in seconds of the current frame
21322 original position in the file of the frame, or undefined if undefined
21323 for the current frame
21326 The previous input PTS.
21329 previous input time in seconds
21332 The previous output PTS.
21335 previous output time in seconds
21338 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21342 The wallclock (RTC) time at the start of the movie in microseconds.
21345 The timebase of the input timestamps.
21349 @subsection Examples
21353 Start counting PTS from zero
21355 setpts=PTS-STARTPTS
21359 Apply fast motion effect:
21365 Apply slow motion effect:
21371 Set fixed rate of 25 frames per second:
21377 Set fixed rate 25 fps with some jitter:
21379 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21383 Apply an offset of 10 seconds to the input PTS:
21389 Generate timestamps from a "live source" and rebase onto the current timebase:
21391 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21395 Generate timestamps by counting samples:
21404 Force color range for the output video frame.
21406 The @code{setrange} filter marks the color range property for the
21407 output frames. It does not change the input frame, but only sets the
21408 corresponding property, which affects how the frame is treated by
21411 The filter accepts the following options:
21416 Available values are:
21420 Keep the same color range property.
21422 @item unspecified, unknown
21423 Set the color range as unspecified.
21425 @item limited, tv, mpeg
21426 Set the color range as limited.
21428 @item full, pc, jpeg
21429 Set the color range as full.
21433 @section settb, asettb
21435 Set the timebase to use for the output frames timestamps.
21436 It is mainly useful for testing timebase configuration.
21438 It accepts the following parameters:
21443 The expression which is evaluated into the output timebase.
21447 The value for @option{tb} is an arithmetic expression representing a
21448 rational. The expression can contain the constants "AVTB" (the default
21449 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21450 audio only). Default value is "intb".
21452 @subsection Examples
21456 Set the timebase to 1/25:
21462 Set the timebase to 1/10:
21468 Set the timebase to 1001/1000:
21474 Set the timebase to 2*intb:
21480 Set the default timebase value:
21487 Convert input audio to a video output representing frequency spectrum
21488 logarithmically using Brown-Puckette constant Q transform algorithm with
21489 direct frequency domain coefficient calculation (but the transform itself
21490 is not really constant Q, instead the Q factor is actually variable/clamped),
21491 with musical tone scale, from E0 to D#10.
21493 The filter accepts the following options:
21497 Specify the video size for the output. It must be even. For the syntax of this option,
21498 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21499 Default value is @code{1920x1080}.
21502 Set the output frame rate. Default value is @code{25}.
21505 Set the bargraph height. It must be even. Default value is @code{-1} which
21506 computes the bargraph height automatically.
21509 Set the axis height. It must be even. Default value is @code{-1} which computes
21510 the axis height automatically.
21513 Set the sonogram height. It must be even. Default value is @code{-1} which
21514 computes the sonogram height automatically.
21517 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21518 instead. Default value is @code{1}.
21520 @item sono_v, volume
21521 Specify the sonogram volume expression. It can contain variables:
21524 the @var{bar_v} evaluated expression
21525 @item frequency, freq, f
21526 the frequency where it is evaluated
21527 @item timeclamp, tc
21528 the value of @var{timeclamp} option
21532 @item a_weighting(f)
21533 A-weighting of equal loudness
21534 @item b_weighting(f)
21535 B-weighting of equal loudness
21536 @item c_weighting(f)
21537 C-weighting of equal loudness.
21539 Default value is @code{16}.
21541 @item bar_v, volume2
21542 Specify the bargraph volume expression. It can contain variables:
21545 the @var{sono_v} evaluated expression
21546 @item frequency, freq, f
21547 the frequency where it is evaluated
21548 @item timeclamp, tc
21549 the value of @var{timeclamp} option
21553 @item a_weighting(f)
21554 A-weighting of equal loudness
21555 @item b_weighting(f)
21556 B-weighting of equal loudness
21557 @item c_weighting(f)
21558 C-weighting of equal loudness.
21560 Default value is @code{sono_v}.
21562 @item sono_g, gamma
21563 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21564 higher gamma makes the spectrum having more range. Default value is @code{3}.
21565 Acceptable range is @code{[1, 7]}.
21567 @item bar_g, gamma2
21568 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21572 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21573 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21575 @item timeclamp, tc
21576 Specify the transform timeclamp. At low frequency, there is trade-off between
21577 accuracy in time domain and frequency domain. If timeclamp is lower,
21578 event in time domain is represented more accurately (such as fast bass drum),
21579 otherwise event in frequency domain is represented more accurately
21580 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21583 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21584 limits future samples by applying asymmetric windowing in time domain, useful
21585 when low latency is required. Accepted range is @code{[0, 1]}.
21588 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21589 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21592 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21593 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21596 This option is deprecated and ignored.
21599 Specify the transform length in time domain. Use this option to control accuracy
21600 trade-off between time domain and frequency domain at every frequency sample.
21601 It can contain variables:
21603 @item frequency, freq, f
21604 the frequency where it is evaluated
21605 @item timeclamp, tc
21606 the value of @var{timeclamp} option.
21608 Default value is @code{384*tc/(384+tc*f)}.
21611 Specify the transform count for every video frame. Default value is @code{6}.
21612 Acceptable range is @code{[1, 30]}.
21615 Specify the transform count for every single pixel. Default value is @code{0},
21616 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21619 Specify font file for use with freetype to draw the axis. If not specified,
21620 use embedded font. Note that drawing with font file or embedded font is not
21621 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21625 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21626 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21629 Specify font color expression. This is arithmetic expression that should return
21630 integer value 0xRRGGBB. It can contain variables:
21632 @item frequency, freq, f
21633 the frequency where it is evaluated
21634 @item timeclamp, tc
21635 the value of @var{timeclamp} option
21640 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21641 @item r(x), g(x), b(x)
21642 red, green, and blue value of intensity x.
21644 Default value is @code{st(0, (midi(f)-59.5)/12);
21645 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21646 r(1-ld(1)) + b(ld(1))}.
21649 Specify image file to draw the axis. This option override @var{fontfile} and
21650 @var{fontcolor} option.
21653 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21654 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21655 Default value is @code{1}.
21658 Set colorspace. The accepted values are:
21661 Unspecified (default)
21670 BT.470BG or BT.601-6 625
21673 SMPTE-170M or BT.601-6 525
21679 BT.2020 with non-constant luminance
21684 Set spectrogram color scheme. This is list of floating point values with format
21685 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21686 The default is @code{1|0.5|0|0|0.5|1}.
21690 @subsection Examples
21694 Playing audio while showing the spectrum:
21696 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21700 Same as above, but with frame rate 30 fps:
21702 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21706 Playing at 1280x720:
21708 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21712 Disable sonogram display:
21718 A1 and its harmonics: A1, A2, (near)E3, A3:
21720 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),
21721 asplit[a][out1]; [a] showcqt [out0]'
21725 Same as above, but with more accuracy in frequency domain:
21727 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),
21728 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21734 bar_v=10:sono_v=bar_v*a_weighting(f)
21738 Custom gamma, now spectrum is linear to the amplitude.
21744 Custom tlength equation:
21746 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)))'
21750 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21752 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21756 Custom font using fontconfig:
21758 font='Courier New,Monospace,mono|bold'
21762 Custom frequency range with custom axis using image file:
21764 axisfile=myaxis.png:basefreq=40:endfreq=10000
21770 Convert input audio to video output representing the audio power spectrum.
21771 Audio amplitude is on Y-axis while frequency is on X-axis.
21773 The filter accepts the following options:
21777 Specify size of video. For the syntax of this option, check the
21778 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21779 Default is @code{1024x512}.
21783 This set how each frequency bin will be represented.
21785 It accepts the following values:
21791 Default is @code{bar}.
21794 Set amplitude scale.
21796 It accepts the following values:
21810 Default is @code{log}.
21813 Set frequency scale.
21815 It accepts the following values:
21824 Reverse logarithmic scale.
21826 Default is @code{lin}.
21831 It accepts the following values:
21847 Default is @code{w2048}
21850 Set windowing function.
21852 It accepts the following values:
21875 Default is @code{hanning}.
21878 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
21879 which means optimal overlap for selected window function will be picked.
21882 Set time averaging. Setting this to 0 will display current maximal peaks.
21883 Default is @code{1}, which means time averaging is disabled.
21886 Specify list of colors separated by space or by '|' which will be used to
21887 draw channel frequencies. Unrecognized or missing colors will be replaced
21891 Set channel display mode.
21893 It accepts the following values:
21898 Default is @code{combined}.
21901 Set minimum amplitude used in @code{log} amplitude scaler.
21905 @anchor{showspectrum}
21906 @section showspectrum
21908 Convert input audio to a video output, representing the audio frequency
21911 The filter accepts the following options:
21915 Specify the video size for the output. For the syntax of this option, check the
21916 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21917 Default value is @code{640x512}.
21920 Specify how the spectrum should slide along the window.
21922 It accepts the following values:
21925 the samples start again on the left when they reach the right
21927 the samples scroll from right to left
21929 frames are only produced when the samples reach the right
21931 the samples scroll from left to right
21934 Default value is @code{replace}.
21937 Specify display mode.
21939 It accepts the following values:
21942 all channels are displayed in the same row
21944 all channels are displayed in separate rows
21947 Default value is @samp{combined}.
21950 Specify display color mode.
21952 It accepts the following values:
21955 each channel is displayed in a separate color
21957 each channel is displayed using the same color scheme
21959 each channel is displayed using the rainbow color scheme
21961 each channel is displayed using the moreland color scheme
21963 each channel is displayed using the nebulae color scheme
21965 each channel is displayed using the fire color scheme
21967 each channel is displayed using the fiery color scheme
21969 each channel is displayed using the fruit color scheme
21971 each channel is displayed using the cool color scheme
21973 each channel is displayed using the magma color scheme
21975 each channel is displayed using the green color scheme
21977 each channel is displayed using the viridis color scheme
21979 each channel is displayed using the plasma color scheme
21981 each channel is displayed using the cividis color scheme
21983 each channel is displayed using the terrain color scheme
21986 Default value is @samp{channel}.
21989 Specify scale used for calculating intensity color values.
21991 It accepts the following values:
21996 square root, default
22007 Default value is @samp{sqrt}.
22010 Set saturation modifier for displayed colors. Negative values provide
22011 alternative color scheme. @code{0} is no saturation at all.
22012 Saturation must be in [-10.0, 10.0] range.
22013 Default value is @code{1}.
22016 Set window function.
22018 It accepts the following values:
22043 Default value is @code{hann}.
22046 Set orientation of time vs frequency axis. Can be @code{vertical} or
22047 @code{horizontal}. Default is @code{vertical}.
22050 Set ratio of overlap window. Default value is @code{0}.
22051 When value is @code{1} overlap is set to recommended size for specific
22052 window function currently used.
22055 Set scale gain for calculating intensity color values.
22056 Default value is @code{1}.
22059 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22062 Set color rotation, must be in [-1.0, 1.0] range.
22063 Default value is @code{0}.
22066 Set start frequency from which to display spectrogram. Default is @code{0}.
22069 Set stop frequency to which to display spectrogram. Default is @code{0}.
22072 Set upper frame rate limit. Default is @code{auto}, unlimited.
22075 Draw time and frequency axes and legends. Default is disabled.
22078 The usage is very similar to the showwaves filter; see the examples in that
22081 @subsection Examples
22085 Large window with logarithmic color scaling:
22087 showspectrum=s=1280x480:scale=log
22091 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22093 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22094 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22098 @section showspectrumpic
22100 Convert input audio to a single video frame, representing the audio frequency
22103 The filter accepts the following options:
22107 Specify the video size for the output. For the syntax of this option, check the
22108 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22109 Default value is @code{4096x2048}.
22112 Specify display mode.
22114 It accepts the following values:
22117 all channels are displayed in the same row
22119 all channels are displayed in separate rows
22121 Default value is @samp{combined}.
22124 Specify display color mode.
22126 It accepts the following values:
22129 each channel is displayed in a separate color
22131 each channel is displayed using the same color scheme
22133 each channel is displayed using the rainbow color scheme
22135 each channel is displayed using the moreland color scheme
22137 each channel is displayed using the nebulae color scheme
22139 each channel is displayed using the fire color scheme
22141 each channel is displayed using the fiery color scheme
22143 each channel is displayed using the fruit color scheme
22145 each channel is displayed using the cool color scheme
22147 each channel is displayed using the magma color scheme
22149 each channel is displayed using the green color scheme
22151 each channel is displayed using the viridis color scheme
22153 each channel is displayed using the plasma color scheme
22155 each channel is displayed using the cividis color scheme
22157 each channel is displayed using the terrain color scheme
22159 Default value is @samp{intensity}.
22162 Specify scale used for calculating intensity color values.
22164 It accepts the following values:
22169 square root, default
22179 Default value is @samp{log}.
22182 Set saturation modifier for displayed colors. Negative values provide
22183 alternative color scheme. @code{0} is no saturation at all.
22184 Saturation must be in [-10.0, 10.0] range.
22185 Default value is @code{1}.
22188 Set window function.
22190 It accepts the following values:
22214 Default value is @code{hann}.
22217 Set orientation of time vs frequency axis. Can be @code{vertical} or
22218 @code{horizontal}. Default is @code{vertical}.
22221 Set scale gain for calculating intensity color values.
22222 Default value is @code{1}.
22225 Draw time and frequency axes and legends. Default is enabled.
22228 Set color rotation, must be in [-1.0, 1.0] range.
22229 Default value is @code{0}.
22232 Set start frequency from which to display spectrogram. Default is @code{0}.
22235 Set stop frequency to which to display spectrogram. Default is @code{0}.
22238 @subsection Examples
22242 Extract an audio spectrogram of a whole audio track
22243 in a 1024x1024 picture using @command{ffmpeg}:
22245 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22249 @section showvolume
22251 Convert input audio volume to a video output.
22253 The filter accepts the following options:
22260 Set border width, allowed range is [0, 5]. Default is 1.
22263 Set channel width, allowed range is [80, 8192]. Default is 400.
22266 Set channel height, allowed range is [1, 900]. Default is 20.
22269 Set fade, allowed range is [0, 1]. Default is 0.95.
22272 Set volume color expression.
22274 The expression can use the following variables:
22278 Current max volume of channel in dB.
22284 Current channel number, starting from 0.
22288 If set, displays channel names. Default is enabled.
22291 If set, displays volume values. Default is enabled.
22294 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22295 default is @code{h}.
22298 Set step size, allowed range is [0, 5]. Default is 0, which means
22302 Set background opacity, allowed range is [0, 1]. Default is 0.
22305 Set metering mode, can be peak: @code{p} or rms: @code{r},
22306 default is @code{p}.
22309 Set display scale, can be linear: @code{lin} or log: @code{log},
22310 default is @code{lin}.
22314 If set to > 0., display a line for the max level
22315 in the previous seconds.
22316 default is disabled: @code{0.}
22319 The color of the max line. Use when @code{dm} option is set to > 0.
22320 default is: @code{orange}
22325 Convert input audio to a video output, representing the samples waves.
22327 The filter accepts the following options:
22331 Specify the video size for the output. For the syntax of this option, check the
22332 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22333 Default value is @code{600x240}.
22338 Available values are:
22341 Draw a point for each sample.
22344 Draw a vertical line for each sample.
22347 Draw a point for each sample and a line between them.
22350 Draw a centered vertical line for each sample.
22353 Default value is @code{point}.
22356 Set the number of samples which are printed on the same column. A
22357 larger value will decrease the frame rate. Must be a positive
22358 integer. This option can be set only if the value for @var{rate}
22359 is not explicitly specified.
22362 Set the (approximate) output frame rate. This is done by setting the
22363 option @var{n}. Default value is "25".
22365 @item split_channels
22366 Set if channels should be drawn separately or overlap. Default value is 0.
22369 Set colors separated by '|' which are going to be used for drawing of each channel.
22372 Set amplitude scale.
22374 Available values are:
22392 Set the draw mode. This is mostly useful to set for high @var{n}.
22394 Available values are:
22397 Scale pixel values for each drawn sample.
22400 Draw every sample directly.
22403 Default value is @code{scale}.
22406 @subsection Examples
22410 Output the input file audio and the corresponding video representation
22413 amovie=a.mp3,asplit[out0],showwaves[out1]
22417 Create a synthetic signal and show it with showwaves, forcing a
22418 frame rate of 30 frames per second:
22420 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22424 @section showwavespic
22426 Convert input audio to a single video frame, representing the samples waves.
22428 The filter accepts the following options:
22432 Specify the video size for the output. For the syntax of this option, check the
22433 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22434 Default value is @code{600x240}.
22436 @item split_channels
22437 Set if channels should be drawn separately or overlap. Default value is 0.
22440 Set colors separated by '|' which are going to be used for drawing of each channel.
22443 Set amplitude scale.
22445 Available values are:
22463 @subsection Examples
22467 Extract a channel split representation of the wave form of a whole audio track
22468 in a 1024x800 picture using @command{ffmpeg}:
22470 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22474 @section sidedata, asidedata
22476 Delete frame side data, or select frames based on it.
22478 This filter accepts the following options:
22482 Set mode of operation of the filter.
22484 Can be one of the following:
22488 Select every frame with side data of @code{type}.
22491 Delete side data of @code{type}. If @code{type} is not set, delete all side
22497 Set side data type used with all modes. Must be set for @code{select} mode. For
22498 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22499 in @file{libavutil/frame.h}. For example, to choose
22500 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22504 @section spectrumsynth
22506 Sythesize audio from 2 input video spectrums, first input stream represents
22507 magnitude across time and second represents phase across time.
22508 The filter will transform from frequency domain as displayed in videos back
22509 to time domain as presented in audio output.
22511 This filter is primarily created for reversing processed @ref{showspectrum}
22512 filter outputs, but can synthesize sound from other spectrograms too.
22513 But in such case results are going to be poor if the phase data is not
22514 available, because in such cases phase data need to be recreated, usually
22515 it's just recreated from random noise.
22516 For best results use gray only output (@code{channel} color mode in
22517 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22518 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22519 @code{data} option. Inputs videos should generally use @code{fullframe}
22520 slide mode as that saves resources needed for decoding video.
22522 The filter accepts the following options:
22526 Specify sample rate of output audio, the sample rate of audio from which
22527 spectrum was generated may differ.
22530 Set number of channels represented in input video spectrums.
22533 Set scale which was used when generating magnitude input spectrum.
22534 Can be @code{lin} or @code{log}. Default is @code{log}.
22537 Set slide which was used when generating inputs spectrums.
22538 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22539 Default is @code{fullframe}.
22542 Set window function used for resynthesis.
22545 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22546 which means optimal overlap for selected window function will be picked.
22549 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22550 Default is @code{vertical}.
22553 @subsection Examples
22557 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22558 then resynthesize videos back to audio with spectrumsynth:
22560 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
22561 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
22562 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22566 @section split, asplit
22568 Split input into several identical outputs.
22570 @code{asplit} works with audio input, @code{split} with video.
22572 The filter accepts a single parameter which specifies the number of outputs. If
22573 unspecified, it defaults to 2.
22575 @subsection Examples
22579 Create two separate outputs from the same input:
22581 [in] split [out0][out1]
22585 To create 3 or more outputs, you need to specify the number of
22588 [in] asplit=3 [out0][out1][out2]
22592 Create two separate outputs from the same input, one cropped and
22595 [in] split [splitout1][splitout2];
22596 [splitout1] crop=100:100:0:0 [cropout];
22597 [splitout2] pad=200:200:100:100 [padout];
22601 Create 5 copies of the input audio with @command{ffmpeg}:
22603 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22609 Receive commands sent through a libzmq client, and forward them to
22610 filters in the filtergraph.
22612 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22613 must be inserted between two video filters, @code{azmq} between two
22614 audio filters. Both are capable to send messages to any filter type.
22616 To enable these filters you need to install the libzmq library and
22617 headers and configure FFmpeg with @code{--enable-libzmq}.
22619 For more information about libzmq see:
22620 @url{http://www.zeromq.org/}
22622 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22623 receives messages sent through a network interface defined by the
22624 @option{bind_address} (or the abbreviation "@option{b}") option.
22625 Default value of this option is @file{tcp://localhost:5555}. You may
22626 want to alter this value to your needs, but do not forget to escape any
22627 ':' signs (see @ref{filtergraph escaping}).
22629 The received message must be in the form:
22631 @var{TARGET} @var{COMMAND} [@var{ARG}]
22634 @var{TARGET} specifies the target of the command, usually the name of
22635 the filter class or a specific filter instance name. The default
22636 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22637 but you can override this by using the @samp{filter_name@@id} syntax
22638 (see @ref{Filtergraph syntax}).
22640 @var{COMMAND} specifies the name of the command for the target filter.
22642 @var{ARG} is optional and specifies the optional argument list for the
22643 given @var{COMMAND}.
22645 Upon reception, the message is processed and the corresponding command
22646 is injected into the filtergraph. Depending on the result, the filter
22647 will send a reply to the client, adopting the format:
22649 @var{ERROR_CODE} @var{ERROR_REASON}
22653 @var{MESSAGE} is optional.
22655 @subsection Examples
22657 Look at @file{tools/zmqsend} for an example of a zmq client which can
22658 be used to send commands processed by these filters.
22660 Consider the following filtergraph generated by @command{ffplay}.
22661 In this example the last overlay filter has an instance name. All other
22662 filters will have default instance names.
22665 ffplay -dumpgraph 1 -f lavfi "
22666 color=s=100x100:c=red [l];
22667 color=s=100x100:c=blue [r];
22668 nullsrc=s=200x100, zmq [bg];
22669 [bg][l] overlay [bg+l];
22670 [bg+l][r] overlay@@my=x=100 "
22673 To change the color of the left side of the video, the following
22674 command can be used:
22676 echo Parsed_color_0 c yellow | tools/zmqsend
22679 To change the right side:
22681 echo Parsed_color_1 c pink | tools/zmqsend
22684 To change the position of the right side:
22686 echo overlay@@my x 150 | tools/zmqsend
22690 @c man end MULTIMEDIA FILTERS
22692 @chapter Multimedia Sources
22693 @c man begin MULTIMEDIA SOURCES
22695 Below is a description of the currently available multimedia sources.
22699 This is the same as @ref{movie} source, except it selects an audio
22705 Read audio and/or video stream(s) from a movie container.
22707 It accepts the following parameters:
22711 The name of the resource to read (not necessarily a file; it can also be a
22712 device or a stream accessed through some protocol).
22714 @item format_name, f
22715 Specifies the format assumed for the movie to read, and can be either
22716 the name of a container or an input device. If not specified, the
22717 format is guessed from @var{movie_name} or by probing.
22719 @item seek_point, sp
22720 Specifies the seek point in seconds. The frames will be output
22721 starting from this seek point. The parameter is evaluated with
22722 @code{av_strtod}, so the numerical value may be suffixed by an IS
22723 postfix. The default value is "0".
22726 Specifies the streams to read. Several streams can be specified,
22727 separated by "+". The source will then have as many outputs, in the
22728 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22729 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22730 respectively the default (best suited) video and audio stream. Default
22731 is "dv", or "da" if the filter is called as "amovie".
22733 @item stream_index, si
22734 Specifies the index of the video stream to read. If the value is -1,
22735 the most suitable video stream will be automatically selected. The default
22736 value is "-1". Deprecated. If the filter is called "amovie", it will select
22737 audio instead of video.
22740 Specifies how many times to read the stream in sequence.
22741 If the value is 0, the stream will be looped infinitely.
22742 Default value is "1".
22744 Note that when the movie is looped the source timestamps are not
22745 changed, so it will generate non monotonically increasing timestamps.
22747 @item discontinuity
22748 Specifies the time difference between frames above which the point is
22749 considered a timestamp discontinuity which is removed by adjusting the later
22753 It allows overlaying a second video on top of the main input of
22754 a filtergraph, as shown in this graph:
22756 input -----------> deltapts0 --> overlay --> output
22759 movie --> scale--> deltapts1 -------+
22761 @subsection Examples
22765 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22766 on top of the input labelled "in":
22768 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22769 [in] setpts=PTS-STARTPTS [main];
22770 [main][over] overlay=16:16 [out]
22774 Read from a video4linux2 device, and overlay it on top of the input
22777 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22778 [in] setpts=PTS-STARTPTS [main];
22779 [main][over] overlay=16:16 [out]
22783 Read the first video stream and the audio stream with id 0x81 from
22784 dvd.vob; the video is connected to the pad named "video" and the audio is
22785 connected to the pad named "audio":
22787 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22791 @subsection Commands
22793 Both movie and amovie support the following commands:
22796 Perform seek using "av_seek_frame".
22797 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22800 @var{stream_index}: If stream_index is -1, a default
22801 stream is selected, and @var{timestamp} is automatically converted
22802 from AV_TIME_BASE units to the stream specific time_base.
22804 @var{timestamp}: Timestamp in AVStream.time_base units
22805 or, if no stream is specified, in AV_TIME_BASE units.
22807 @var{flags}: Flags which select direction and seeking mode.
22811 Get movie duration in AV_TIME_BASE units.
22815 @c man end MULTIMEDIA SOURCES