1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 If a signal of stream rises above this level it will affect the gain
396 By default it is 0.125. Range is between 0.00097563 and 1.
399 Set a ratio by which the signal is reduced. 1:2 means that if the level
400 rose 4dB above the threshold, it will be only 2dB above after the reduction.
401 Default is 2. Range is between 1 and 20.
404 Amount of milliseconds the signal has to rise above the threshold before gain
405 reduction starts. Default is 20. Range is between 0.01 and 2000.
408 Amount of milliseconds the signal has to fall below the threshold before
409 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
412 Set the amount by how much signal will be amplified after processing.
413 Default is 1. Range is from 1 to 64.
416 Curve the sharp knee around the threshold to enter gain reduction more softly.
417 Default is 2.82843. Range is between 1 and 8.
420 Choose if the @code{average} level between all channels of input stream
421 or the louder(@code{maximum}) channel of input stream affects the
422 reduction. Default is @code{average}.
425 Should the exact signal be taken in case of @code{peak} or an RMS one in case
426 of @code{rms}. Default is @code{rms} which is mostly smoother.
429 How much to use compressed signal in output. Default is 1.
430 Range is between 0 and 1.
434 Simple audio dynamic range commpression/expansion filter.
436 The filter accepts the following options:
440 Set contrast. Default is 33. Allowed range is between 0 and 100.
445 Copy the input audio source unchanged to the output. This is mainly useful for
450 Apply cross fade from one input audio stream to another input audio stream.
451 The cross fade is applied for specified duration near the end of first stream.
453 The filter accepts the following options:
457 Specify the number of samples for which the cross fade effect has to last.
458 At the end of the cross fade effect the first input audio will be completely
459 silent. Default is 44100.
462 Specify the duration of the cross fade effect. See
463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
464 for the accepted syntax.
465 By default the duration is determined by @var{nb_samples}.
466 If set this option is used instead of @var{nb_samples}.
469 Should first stream end overlap with second stream start. Default is enabled.
472 Set curve for cross fade transition for first stream.
475 Set curve for cross fade transition for second stream.
477 For description of available curve types see @ref{afade} filter description.
484 Cross fade from one input to another:
486 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
490 Cross fade from one input to another but without overlapping:
492 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
498 Reduce audio bit resolution.
500 This filter is bit crusher with enhanced functionality. A bit crusher
501 is used to audibly reduce number of bits an audio signal is sampled
502 with. This doesn't change the bit depth at all, it just produces the
503 effect. Material reduced in bit depth sounds more harsh and "digital".
504 This filter is able to even round to continuous values instead of discrete
506 Additionally it has a D/C offset which results in different crushing of
507 the lower and the upper half of the signal.
508 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
510 Another feature of this filter is the logarithmic mode.
511 This setting switches from linear distances between bits to logarithmic ones.
512 The result is a much more "natural" sounding crusher which doesn't gate low
513 signals for example. The human ear has a logarithmic perception,
514 so this kind of crushing is much more pleasant.
515 Logarithmic crushing is also able to get anti-aliased.
517 The filter accepts the following options:
533 Can be linear: @code{lin} or logarithmic: @code{log}.
542 Set sample reduction.
545 Enable LFO. By default disabled.
556 Delay one or more audio channels.
558 Samples in delayed channel are filled with silence.
560 The filter accepts the following option:
564 Set list of delays in milliseconds for each channel separated by '|'.
565 Unused delays will be silently ignored. If number of given delays is
566 smaller than number of channels all remaining channels will not be delayed.
567 If you want to delay exact number of samples, append 'S' to number.
574 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
575 the second channel (and any other channels that may be present) unchanged.
581 Delay second channel by 500 samples, the third channel by 700 samples and leave
582 the first channel (and any other channels that may be present) unchanged.
588 @section aderivative, aintegral
590 Compute derivative/integral of audio stream.
592 Applying both filters one after another produces original audio.
596 Apply echoing to the input audio.
598 Echoes are reflected sound and can occur naturally amongst mountains
599 (and sometimes large buildings) when talking or shouting; digital echo
600 effects emulate this behaviour and are often used to help fill out the
601 sound of a single instrument or vocal. The time difference between the
602 original signal and the reflection is the @code{delay}, and the
603 loudness of the reflected signal is the @code{decay}.
604 Multiple echoes can have different delays and decays.
606 A description of the accepted parameters follows.
610 Set input gain of reflected signal. Default is @code{0.6}.
613 Set output gain of reflected signal. Default is @code{0.3}.
616 Set list of time intervals in milliseconds between original signal and reflections
617 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
618 Default is @code{1000}.
621 Set list of loudness of reflected signals separated by '|'.
622 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
623 Default is @code{0.5}.
630 Make it sound as if there are twice as many instruments as are actually playing:
632 aecho=0.8:0.88:60:0.4
636 If delay is very short, then it sound like a (metallic) robot playing music:
642 A longer delay will sound like an open air concert in the mountains:
644 aecho=0.8:0.9:1000:0.3
648 Same as above but with one more mountain:
650 aecho=0.8:0.9:1000|1800:0.3|0.25
655 Audio emphasis filter creates or restores material directly taken from LPs or
656 emphased CDs with different filter curves. E.g. to store music on vinyl the
657 signal has to be altered by a filter first to even out the disadvantages of
658 this recording medium.
659 Once the material is played back the inverse filter has to be applied to
660 restore the distortion of the frequency response.
662 The filter accepts the following options:
672 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
673 use @code{production} mode. Default is @code{reproduction} mode.
676 Set filter type. Selects medium. Can be one of the following:
688 select Compact Disc (CD).
694 select 50µs (FM-KF).
696 select 75µs (FM-KF).
702 Modify an audio signal according to the specified expressions.
704 This filter accepts one or more expressions (one for each channel),
705 which are evaluated and used to modify a corresponding audio signal.
707 It accepts the following parameters:
711 Set the '|'-separated expressions list for each separate channel. If
712 the number of input channels is greater than the number of
713 expressions, the last specified expression is used for the remaining
716 @item channel_layout, c
717 Set output channel layout. If not specified, the channel layout is
718 specified by the number of expressions. If set to @samp{same}, it will
719 use by default the same input channel layout.
722 Each expression in @var{exprs} can contain the following constants and functions:
726 channel number of the current expression
729 number of the evaluated sample, starting from 0
735 time of the evaluated sample expressed in seconds
738 @item nb_out_channels
739 input and output number of channels
742 the value of input channel with number @var{CH}
745 Note: this filter is slow. For faster processing you should use a
754 aeval=val(ch)/2:c=same
758 Invert phase of the second channel:
767 Apply fade-in/out effect to input audio.
769 A description of the accepted parameters follows.
773 Specify the effect type, can be either @code{in} for fade-in, or
774 @code{out} for a fade-out effect. Default is @code{in}.
776 @item start_sample, ss
777 Specify the number of the start sample for starting to apply the fade
778 effect. Default is 0.
781 Specify the number of samples for which the fade effect has to last. At
782 the end of the fade-in effect the output audio will have the same
783 volume as the input audio, at the end of the fade-out transition
784 the output audio will be silence. Default is 44100.
787 Specify the start time of the fade effect. Default is 0.
788 The value must be specified as a time duration; see
789 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
790 for the accepted syntax.
791 If set this option is used instead of @var{start_sample}.
794 Specify the duration of the fade effect. See
795 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
796 for the accepted syntax.
797 At the end of the fade-in effect the output audio will have the same
798 volume as the input audio, at the end of the fade-out transition
799 the output audio will be silence.
800 By default the duration is determined by @var{nb_samples}.
801 If set this option is used instead of @var{nb_samples}.
804 Set curve for fade transition.
806 It accepts the following values:
809 select triangular, linear slope (default)
811 select quarter of sine wave
813 select half of sine wave
815 select exponential sine wave
819 select inverted parabola
833 select inverted quarter of sine wave
835 select inverted half of sine wave
837 select double-exponential seat
839 select double-exponential sigmoid
847 Fade in first 15 seconds of audio:
853 Fade out last 25 seconds of a 900 seconds audio:
855 afade=t=out:st=875:d=25
860 Apply arbitrary expressions to samples in frequency domain.
864 Set frequency domain real expression for each separate channel separated
865 by '|'. Default is "1".
866 If the number of input channels is greater than the number of
867 expressions, the last specified expression is used for the remaining
871 Set frequency domain imaginary expression for each separate channel
872 separated by '|'. If not set, @var{real} option is used.
874 Each expression in @var{real} and @var{imag} can contain the following
882 current frequency bin number
885 number of available bins
888 channel number of the current expression
900 It accepts the following values:
916 Default is @code{w4096}
919 Set window function. Default is @code{hann}.
922 Set window overlap. If set to 1, the recommended overlap for selected
923 window function will be picked. Default is @code{0.75}.
930 Leave almost only low frequencies in audio:
932 afftfilt="1-clip((b/nb)*b,0,1)"
939 Apply an arbitrary Frequency Impulse Response filter.
941 This filter is designed for applying long FIR filters,
942 up to 30 seconds long.
944 It can be used as component for digital crossover filters,
945 room equalization, cross talk cancellation, wavefield synthesis,
946 auralization, ambiophonics and ambisonics.
948 This filter uses second stream as FIR coefficients.
949 If second stream holds single channel, it will be used
950 for all input channels in first stream, otherwise
951 number of channels in second stream must be same as
952 number of channels in first stream.
954 It accepts the following parameters:
958 Set dry gain. This sets input gain.
961 Set wet gain. This sets final output gain.
964 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
967 Enable applying gain measured from power of IR.
970 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
971 Allowed range is 0.1 to 60 seconds.
974 Show IR frequency reponse, magnitude and phase in additional video stream.
975 By default it is disabled.
978 Set for which IR channel to display frequency response. By default is first channel
979 displayed. This option is used only when @var{response} is enabled.
982 Set video stream size. This option is used only when @var{response} is enabled.
989 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
991 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
998 Set output format constraints for the input audio. The framework will
999 negotiate the most appropriate format to minimize conversions.
1001 It accepts the following parameters:
1005 A '|'-separated list of requested sample formats.
1008 A '|'-separated list of requested sample rates.
1010 @item channel_layouts
1011 A '|'-separated list of requested channel layouts.
1013 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1014 for the required syntax.
1017 If a parameter is omitted, all values are allowed.
1019 Force the output to either unsigned 8-bit or signed 16-bit stereo
1021 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1026 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1027 processing reduces disturbing noise between useful signals.
1029 Gating is done by detecting the volume below a chosen level @var{threshold}
1030 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1031 floor is set via @var{range}. Because an exact manipulation of the signal
1032 would cause distortion of the waveform the reduction can be levelled over
1033 time. This is done by setting @var{attack} and @var{release}.
1035 @var{attack} determines how long the signal has to fall below the threshold
1036 before any reduction will occur and @var{release} sets the time the signal
1037 has to rise above the threshold to reduce the reduction again.
1038 Shorter signals than the chosen attack time will be left untouched.
1042 Set input level before filtering.
1043 Default is 1. Allowed range is from 0.015625 to 64.
1046 Set the level of gain reduction when the signal is below the threshold.
1047 Default is 0.06125. Allowed range is from 0 to 1.
1050 If a signal rises above this level the gain reduction is released.
1051 Default is 0.125. Allowed range is from 0 to 1.
1054 Set a ratio by which the signal is reduced.
1055 Default is 2. Allowed range is from 1 to 9000.
1058 Amount of milliseconds the signal has to rise above the threshold before gain
1060 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1063 Amount of milliseconds the signal has to fall below the threshold before the
1064 reduction is increased again. Default is 250 milliseconds.
1065 Allowed range is from 0.01 to 9000.
1068 Set amount of amplification of signal after processing.
1069 Default is 1. Allowed range is from 1 to 64.
1072 Curve the sharp knee around the threshold to enter gain reduction more softly.
1073 Default is 2.828427125. Allowed range is from 1 to 8.
1076 Choose if exact signal should be taken for detection or an RMS like one.
1077 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1080 Choose if the average level between all channels or the louder channel affects
1082 Default is @code{average}. Can be @code{average} or @code{maximum}.
1087 Apply an arbitrary Infinite Impulse Response filter.
1089 It accepts the following parameters:
1093 Set numerator/zeros coefficients.
1096 Set denominator/poles coefficients.
1108 Set coefficients format.
1114 Z-plane zeros/poles, cartesian (default)
1116 Z-plane zeros/poles, polar radians
1118 Z-plane zeros/poles, polar degrees
1122 Set kind of processing.
1123 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1126 Set filtering precision.
1130 double-precision floating-point (default)
1132 single-precision floating-point
1140 Show IR frequency reponse, magnitude and phase in additional video stream.
1141 By default it is disabled.
1144 Set for which IR channel to display frequency response. By default is first channel
1145 displayed. This option is used only when @var{response} is enabled.
1148 Set video stream size. This option is used only when @var{response} is enabled.
1151 Coefficients in @code{tf} format are separated by spaces and are in ascending
1154 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1155 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1158 Different coefficients and gains can be provided for every channel, in such case
1159 use '|' to separate coefficients or gains. Last provided coefficients will be
1160 used for all remaining channels.
1162 @subsection Examples
1166 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1168 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
1172 Same as above but in @code{zp} format:
1174 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
1180 The limiter prevents an input signal from rising over a desired threshold.
1181 This limiter uses lookahead technology to prevent your signal from distorting.
1182 It means that there is a small delay after the signal is processed. Keep in mind
1183 that the delay it produces is the attack time you set.
1185 The filter accepts the following options:
1189 Set input gain. Default is 1.
1192 Set output gain. Default is 1.
1195 Don't let signals above this level pass the limiter. Default is 1.
1198 The limiter will reach its attenuation level in this amount of time in
1199 milliseconds. Default is 5 milliseconds.
1202 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1203 Default is 50 milliseconds.
1206 When gain reduction is always needed ASC takes care of releasing to an
1207 average reduction level rather than reaching a reduction of 0 in the release
1211 Select how much the release time is affected by ASC, 0 means nearly no changes
1212 in release time while 1 produces higher release times.
1215 Auto level output signal. Default is enabled.
1216 This normalizes audio back to 0dB if enabled.
1219 Depending on picked setting it is recommended to upsample input 2x or 4x times
1220 with @ref{aresample} before applying this filter.
1224 Apply a two-pole all-pass filter with central frequency (in Hz)
1225 @var{frequency}, and filter-width @var{width}.
1226 An all-pass filter changes the audio's frequency to phase relationship
1227 without changing its frequency to amplitude relationship.
1229 The filter accepts the following options:
1233 Set frequency in Hz.
1236 Set method to specify band-width of filter.
1251 Specify the band-width of a filter in width_type units.
1254 Specify which channels to filter, by default all available are filtered.
1257 @subsection Commands
1259 This filter supports the following commands:
1262 Change allpass frequency.
1263 Syntax for the command is : "@var{frequency}"
1266 Change allpass width_type.
1267 Syntax for the command is : "@var{width_type}"
1270 Change allpass width.
1271 Syntax for the command is : "@var{width}"
1278 The filter accepts the following options:
1282 Set the number of loops. Setting this value to -1 will result in infinite loops.
1286 Set maximal number of samples. Default is 0.
1289 Set first sample of loop. Default is 0.
1295 Merge two or more audio streams into a single multi-channel stream.
1297 The filter accepts the following options:
1302 Set the number of inputs. Default is 2.
1306 If the channel layouts of the inputs are disjoint, and therefore compatible,
1307 the channel layout of the output will be set accordingly and the channels
1308 will be reordered as necessary. If the channel layouts of the inputs are not
1309 disjoint, the output will have all the channels of the first input then all
1310 the channels of the second input, in that order, and the channel layout of
1311 the output will be the default value corresponding to the total number of
1314 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1315 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1316 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1317 first input, b1 is the first channel of the second input).
1319 On the other hand, if both input are in stereo, the output channels will be
1320 in the default order: a1, a2, b1, b2, and the channel layout will be
1321 arbitrarily set to 4.0, which may or may not be the expected value.
1323 All inputs must have the same sample rate, and format.
1325 If inputs do not have the same duration, the output will stop with the
1328 @subsection Examples
1332 Merge two mono files into a stereo stream:
1334 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1338 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1340 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
1346 Mixes multiple audio inputs into a single output.
1348 Note that this filter only supports float samples (the @var{amerge}
1349 and @var{pan} audio filters support many formats). If the @var{amix}
1350 input has integer samples then @ref{aresample} will be automatically
1351 inserted to perform the conversion to float samples.
1355 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1357 will mix 3 input audio streams to a single output with the same duration as the
1358 first input and a dropout transition time of 3 seconds.
1360 It accepts the following parameters:
1364 The number of inputs. If unspecified, it defaults to 2.
1367 How to determine the end-of-stream.
1371 The duration of the longest input. (default)
1374 The duration of the shortest input.
1377 The duration of the first input.
1381 @item dropout_transition
1382 The transition time, in seconds, for volume renormalization when an input
1383 stream ends. The default value is 2 seconds.
1386 Specify weight of each input audio stream as sequence.
1387 Each weight is separated by space. By default all inputs have same weight.
1390 @section anequalizer
1392 High-order parametric multiband equalizer for each channel.
1394 It accepts the following parameters:
1398 This option string is in format:
1399 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1400 Each equalizer band is separated by '|'.
1404 Set channel number to which equalization will be applied.
1405 If input doesn't have that channel the entry is ignored.
1408 Set central frequency for band.
1409 If input doesn't have that frequency the entry is ignored.
1412 Set band width in hertz.
1415 Set band gain in dB.
1418 Set filter type for band, optional, can be:
1422 Butterworth, this is default.
1433 With this option activated frequency response of anequalizer is displayed
1437 Set video stream size. Only useful if curves option is activated.
1440 Set max gain that will be displayed. Only useful if curves option is activated.
1441 Setting this to a reasonable value makes it possible to display gain which is derived from
1442 neighbour bands which are too close to each other and thus produce higher gain
1443 when both are activated.
1446 Set frequency scale used to draw frequency response in video output.
1447 Can be linear or logarithmic. Default is logarithmic.
1450 Set color for each channel curve which is going to be displayed in video stream.
1451 This is list of color names separated by space or by '|'.
1452 Unrecognised or missing colors will be replaced by white color.
1455 @subsection Examples
1459 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1460 for first 2 channels using Chebyshev type 1 filter:
1462 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1466 @subsection Commands
1468 This filter supports the following commands:
1471 Alter existing filter parameters.
1472 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1474 @var{fN} is existing filter number, starting from 0, if no such filter is available
1476 @var{freq} set new frequency parameter.
1477 @var{width} set new width parameter in herz.
1478 @var{gain} set new gain parameter in dB.
1480 Full filter invocation with asendcmd may look like this:
1481 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1486 Pass the audio source unchanged to the output.
1490 Pad the end of an audio stream with silence.
1492 This can be used together with @command{ffmpeg} @option{-shortest} to
1493 extend audio streams to the same length as the video stream.
1495 A description of the accepted options follows.
1499 Set silence packet size. Default value is 4096.
1502 Set the number of samples of silence to add to the end. After the
1503 value is reached, the stream is terminated. This option is mutually
1504 exclusive with @option{whole_len}.
1507 Set the minimum total number of samples in the output audio stream. If
1508 the value is longer than the input audio length, silence is added to
1509 the end, until the value is reached. This option is mutually exclusive
1510 with @option{pad_len}.
1513 If neither the @option{pad_len} nor the @option{whole_len} option is
1514 set, the filter will add silence to the end of the input stream
1517 @subsection Examples
1521 Add 1024 samples of silence to the end of the input:
1527 Make sure the audio output will contain at least 10000 samples, pad
1528 the input with silence if required:
1530 apad=whole_len=10000
1534 Use @command{ffmpeg} to pad the audio input with silence, so that the
1535 video stream will always result the shortest and will be converted
1536 until the end in the output file when using the @option{shortest}
1539 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1544 Add a phasing effect to the input audio.
1546 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1547 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1549 A description of the accepted parameters follows.
1553 Set input gain. Default is 0.4.
1556 Set output gain. Default is 0.74
1559 Set delay in milliseconds. Default is 3.0.
1562 Set decay. Default is 0.4.
1565 Set modulation speed in Hz. Default is 0.5.
1568 Set modulation type. Default is triangular.
1570 It accepts the following values:
1579 Audio pulsator is something between an autopanner and a tremolo.
1580 But it can produce funny stereo effects as well. Pulsator changes the volume
1581 of the left and right channel based on a LFO (low frequency oscillator) with
1582 different waveforms and shifted phases.
1583 This filter have the ability to define an offset between left and right
1584 channel. An offset of 0 means that both LFO shapes match each other.
1585 The left and right channel are altered equally - a conventional tremolo.
1586 An offset of 50% means that the shape of the right channel is exactly shifted
1587 in phase (or moved backwards about half of the frequency) - pulsator acts as
1588 an autopanner. At 1 both curves match again. Every setting in between moves the
1589 phase shift gapless between all stages and produces some "bypassing" sounds with
1590 sine and triangle waveforms. The more you set the offset near 1 (starting from
1591 the 0.5) the faster the signal passes from the left to the right speaker.
1593 The filter accepts the following options:
1597 Set input gain. By default it is 1. Range is [0.015625 - 64].
1600 Set output gain. By default it is 1. Range is [0.015625 - 64].
1603 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1604 sawup or sawdown. Default is sine.
1607 Set modulation. Define how much of original signal is affected by the LFO.
1610 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1613 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1616 Set pulse width. Default is 1. Allowed range is [0 - 2].
1619 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1622 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1626 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1630 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1631 if timing is set to hz.
1637 Resample the input audio to the specified parameters, using the
1638 libswresample library. If none are specified then the filter will
1639 automatically convert between its input and output.
1641 This filter is also able to stretch/squeeze the audio data to make it match
1642 the timestamps or to inject silence / cut out audio to make it match the
1643 timestamps, do a combination of both or do neither.
1645 The filter accepts the syntax
1646 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1647 expresses a sample rate and @var{resampler_options} is a list of
1648 @var{key}=@var{value} pairs, separated by ":". See the
1649 @ref{Resampler Options,,"Resampler Options" section in the
1650 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1651 for the complete list of supported options.
1653 @subsection Examples
1657 Resample the input audio to 44100Hz:
1663 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1664 samples per second compensation:
1666 aresample=async=1000
1672 Reverse an audio clip.
1674 Warning: This filter requires memory to buffer the entire clip, so trimming
1677 @subsection Examples
1681 Take the first 5 seconds of a clip, and reverse it.
1683 atrim=end=5,areverse
1687 @section asetnsamples
1689 Set the number of samples per each output audio frame.
1691 The last output packet may contain a different number of samples, as
1692 the filter will flush all the remaining samples when the input audio
1695 The filter accepts the following options:
1699 @item nb_out_samples, n
1700 Set the number of frames per each output audio frame. The number is
1701 intended as the number of samples @emph{per each channel}.
1702 Default value is 1024.
1705 If set to 1, the filter will pad the last audio frame with zeroes, so
1706 that the last frame will contain the same number of samples as the
1707 previous ones. Default value is 1.
1710 For example, to set the number of per-frame samples to 1234 and
1711 disable padding for the last frame, use:
1713 asetnsamples=n=1234:p=0
1718 Set the sample rate without altering the PCM data.
1719 This will result in a change of speed and pitch.
1721 The filter accepts the following options:
1724 @item sample_rate, r
1725 Set the output sample rate. Default is 44100 Hz.
1730 Show a line containing various information for each input audio frame.
1731 The input audio is not modified.
1733 The shown line contains a sequence of key/value pairs of the form
1734 @var{key}:@var{value}.
1736 The following values are shown in the output:
1740 The (sequential) number of the input frame, starting from 0.
1743 The presentation timestamp of the input frame, in time base units; the time base
1744 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1747 The presentation timestamp of the input frame in seconds.
1750 position of the frame in the input stream, -1 if this information in
1751 unavailable and/or meaningless (for example in case of synthetic audio)
1760 The sample rate for the audio frame.
1763 The number of samples (per channel) in the frame.
1766 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1767 audio, the data is treated as if all the planes were concatenated.
1769 @item plane_checksums
1770 A list of Adler-32 checksums for each data plane.
1776 Display time domain statistical information about the audio channels.
1777 Statistics are calculated and displayed for each audio channel and,
1778 where applicable, an overall figure is also given.
1780 It accepts the following option:
1783 Short window length in seconds, used for peak and trough RMS measurement.
1784 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
1788 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1789 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1792 Available keys for each channel are:
1826 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1827 this @code{lavfi.astats.Overall.Peak_count}.
1829 For description what each key means read below.
1832 Set number of frame after which stats are going to be recalculated.
1833 Default is disabled.
1836 A description of each shown parameter follows:
1840 Mean amplitude displacement from zero.
1843 Minimal sample level.
1846 Maximal sample level.
1848 @item Min difference
1849 Minimal difference between two consecutive samples.
1851 @item Max difference
1852 Maximal difference between two consecutive samples.
1854 @item Mean difference
1855 Mean difference between two consecutive samples.
1856 The average of each difference between two consecutive samples.
1858 @item RMS difference
1859 Root Mean Square difference between two consecutive samples.
1863 Standard peak and RMS level measured in dBFS.
1867 Peak and trough values for RMS level measured over a short window.
1870 Standard ratio of peak to RMS level (note: not in dB).
1873 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1874 (i.e. either @var{Min level} or @var{Max level}).
1877 Number of occasions (not the number of samples) that the signal attained either
1878 @var{Min level} or @var{Max level}.
1881 Overall bit depth of audio. Number of bits used for each sample.
1884 Measured dynamic range of audio in dB.
1891 The filter accepts exactly one parameter, the audio tempo. If not
1892 specified then the filter will assume nominal 1.0 tempo. Tempo must
1893 be in the [0.5, 2.0] range.
1895 @subsection Examples
1899 Slow down audio to 80% tempo:
1905 To speed up audio to 125% tempo:
1913 Trim the input so that the output contains one continuous subpart of the input.
1915 It accepts the following parameters:
1918 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1919 sample with the timestamp @var{start} will be the first sample in the output.
1922 Specify time of the first audio sample that will be dropped, i.e. the
1923 audio sample immediately preceding the one with the timestamp @var{end} will be
1924 the last sample in the output.
1927 Same as @var{start}, except this option sets the start timestamp in samples
1931 Same as @var{end}, except this option sets the end timestamp in samples instead
1935 The maximum duration of the output in seconds.
1938 The number of the first sample that should be output.
1941 The number of the first sample that should be dropped.
1944 @option{start}, @option{end}, and @option{duration} are expressed as time
1945 duration specifications; see
1946 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1948 Note that the first two sets of the start/end options and the @option{duration}
1949 option look at the frame timestamp, while the _sample options simply count the
1950 samples that pass through the filter. So start/end_pts and start/end_sample will
1951 give different results when the timestamps are wrong, inexact or do not start at
1952 zero. Also note that this filter does not modify the timestamps. If you wish
1953 to have the output timestamps start at zero, insert the asetpts filter after the
1956 If multiple start or end options are set, this filter tries to be greedy and
1957 keep all samples that match at least one of the specified constraints. To keep
1958 only the part that matches all the constraints at once, chain multiple atrim
1961 The defaults are such that all the input is kept. So it is possible to set e.g.
1962 just the end values to keep everything before the specified time.
1967 Drop everything except the second minute of input:
1969 ffmpeg -i INPUT -af atrim=60:120
1973 Keep only the first 1000 samples:
1975 ffmpeg -i INPUT -af atrim=end_sample=1000
1982 Apply a two-pole Butterworth band-pass filter with central
1983 frequency @var{frequency}, and (3dB-point) band-width width.
1984 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1985 instead of the default: constant 0dB peak gain.
1986 The filter roll off at 6dB per octave (20dB per decade).
1988 The filter accepts the following options:
1992 Set the filter's central frequency. Default is @code{3000}.
1995 Constant skirt gain if set to 1. Defaults to 0.
1998 Set method to specify band-width of filter.
2013 Specify the band-width of a filter in width_type units.
2016 Specify which channels to filter, by default all available are filtered.
2019 @subsection Commands
2021 This filter supports the following commands:
2024 Change bandpass frequency.
2025 Syntax for the command is : "@var{frequency}"
2028 Change bandpass width_type.
2029 Syntax for the command is : "@var{width_type}"
2032 Change bandpass width.
2033 Syntax for the command is : "@var{width}"
2038 Apply a two-pole Butterworth band-reject filter with central
2039 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2040 The filter roll off at 6dB per octave (20dB per decade).
2042 The filter accepts the following options:
2046 Set the filter's central frequency. Default is @code{3000}.
2049 Set method to specify band-width of filter.
2064 Specify the band-width of a filter in width_type units.
2067 Specify which channels to filter, by default all available are filtered.
2070 @subsection Commands
2072 This filter supports the following commands:
2075 Change bandreject frequency.
2076 Syntax for the command is : "@var{frequency}"
2079 Change bandreject width_type.
2080 Syntax for the command is : "@var{width_type}"
2083 Change bandreject width.
2084 Syntax for the command is : "@var{width}"
2087 @section bass, lowshelf
2089 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2090 shelving filter with a response similar to that of a standard
2091 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2093 The filter accepts the following options:
2097 Give the gain at 0 Hz. Its useful range is about -20
2098 (for a large cut) to +20 (for a large boost).
2099 Beware of clipping when using a positive gain.
2102 Set the filter's central frequency and so can be used
2103 to extend or reduce the frequency range to be boosted or cut.
2104 The default value is @code{100} Hz.
2107 Set method to specify band-width of filter.
2122 Determine how steep is the filter's shelf transition.
2125 Specify which channels to filter, by default all available are filtered.
2128 @subsection Commands
2130 This filter supports the following commands:
2133 Change bass frequency.
2134 Syntax for the command is : "@var{frequency}"
2137 Change bass width_type.
2138 Syntax for the command is : "@var{width_type}"
2142 Syntax for the command is : "@var{width}"
2146 Syntax for the command is : "@var{gain}"
2151 Apply a biquad IIR filter with the given coefficients.
2152 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2153 are the numerator and denominator coefficients respectively.
2154 and @var{channels}, @var{c} specify which channels to filter, by default all
2155 available are filtered.
2157 @subsection Commands
2159 This filter supports the following commands:
2167 Change biquad parameter.
2168 Syntax for the command is : "@var{value}"
2172 Bauer stereo to binaural transformation, which improves headphone listening of
2173 stereo audio records.
2175 To enable compilation of this filter you need to configure FFmpeg with
2176 @code{--enable-libbs2b}.
2178 It accepts the following parameters:
2182 Pre-defined crossfeed level.
2186 Default level (fcut=700, feed=50).
2189 Chu Moy circuit (fcut=700, feed=60).
2192 Jan Meier circuit (fcut=650, feed=95).
2197 Cut frequency (in Hz).
2206 Remap input channels to new locations.
2208 It accepts the following parameters:
2211 Map channels from input to output. The argument is a '|'-separated list of
2212 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2213 @var{in_channel} form. @var{in_channel} can be either the name of the input
2214 channel (e.g. FL for front left) or its index in the input channel layout.
2215 @var{out_channel} is the name of the output channel or its index in the output
2216 channel layout. If @var{out_channel} is not given then it is implicitly an
2217 index, starting with zero and increasing by one for each mapping.
2219 @item channel_layout
2220 The channel layout of the output stream.
2223 If no mapping is present, the filter will implicitly map input channels to
2224 output channels, preserving indices.
2226 @subsection Examples
2230 For example, assuming a 5.1+downmix input MOV file,
2232 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2234 will create an output WAV file tagged as stereo from the downmix channels of
2238 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2240 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2244 @section channelsplit
2246 Split each channel from an input audio stream into a separate output stream.
2248 It accepts the following parameters:
2250 @item channel_layout
2251 The channel layout of the input stream. The default is "stereo".
2253 A channel layout describing the channels to be extracted as separate output streams
2254 or "all" to extract each input channel as a separate stream. The default is "all".
2256 Choosing channels not present in channel layout in the input will result in an error.
2259 @subsection Examples
2263 For example, assuming a stereo input MP3 file,
2265 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2267 will create an output Matroska file with two audio streams, one containing only
2268 the left channel and the other the right channel.
2271 Split a 5.1 WAV file into per-channel files:
2273 ffmpeg -i in.wav -filter_complex
2274 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2275 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2276 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2281 Extract only LFE from a 5.1 WAV file:
2283 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2284 -map '[LFE]' lfe.wav
2289 Add a chorus effect to the audio.
2291 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2293 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2294 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2295 The modulation depth defines the range the modulated delay is played before or after
2296 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2297 sound tuned around the original one, like in a chorus where some vocals are slightly
2300 It accepts the following parameters:
2303 Set input gain. Default is 0.4.
2306 Set output gain. Default is 0.4.
2309 Set delays. A typical delay is around 40ms to 60ms.
2321 @subsection Examples
2327 chorus=0.7:0.9:55:0.4:0.25:2
2333 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2337 Fuller sounding chorus with three delays:
2339 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
2344 Compress or expand the audio's dynamic range.
2346 It accepts the following parameters:
2352 A list of times in seconds for each channel over which the instantaneous level
2353 of the input signal is averaged to determine its volume. @var{attacks} refers to
2354 increase of volume and @var{decays} refers to decrease of volume. For most
2355 situations, the attack time (response to the audio getting louder) should be
2356 shorter than the decay time, because the human ear is more sensitive to sudden
2357 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2358 a typical value for decay is 0.8 seconds.
2359 If specified number of attacks & decays is lower than number of channels, the last
2360 set attack/decay will be used for all remaining channels.
2363 A list of points for the transfer function, specified in dB relative to the
2364 maximum possible signal amplitude. Each key points list must be defined using
2365 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2366 @code{x0/y0 x1/y1 x2/y2 ....}
2368 The input values must be in strictly increasing order but the transfer function
2369 does not have to be monotonically rising. The point @code{0/0} is assumed but
2370 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2371 function are @code{-70/-70|-60/-20|1/0}.
2374 Set the curve radius in dB for all joints. It defaults to 0.01.
2377 Set the additional gain in dB to be applied at all points on the transfer
2378 function. This allows for easy adjustment of the overall gain.
2382 Set an initial volume, in dB, to be assumed for each channel when filtering
2383 starts. This permits the user to supply a nominal level initially, so that, for
2384 example, a very large gain is not applied to initial signal levels before the
2385 companding has begun to operate. A typical value for audio which is initially
2386 quiet is -90 dB. It defaults to 0.
2389 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2390 delayed before being fed to the volume adjuster. Specifying a delay
2391 approximately equal to the attack/decay times allows the filter to effectively
2392 operate in predictive rather than reactive mode. It defaults to 0.
2396 @subsection Examples
2400 Make music with both quiet and loud passages suitable for listening to in a
2403 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2406 Another example for audio with whisper and explosion parts:
2408 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2412 A noise gate for when the noise is at a lower level than the signal:
2414 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2418 Here is another noise gate, this time for when the noise is at a higher level
2419 than the signal (making it, in some ways, similar to squelch):
2421 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2425 2:1 compression starting at -6dB:
2427 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2431 2:1 compression starting at -9dB:
2433 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2437 2:1 compression starting at -12dB:
2439 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2443 2:1 compression starting at -18dB:
2445 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2449 3:1 compression starting at -15dB:
2451 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2457 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2463 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
2467 Hard limiter at -6dB:
2469 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2473 Hard limiter at -12dB:
2475 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2479 Hard noise gate at -35 dB:
2481 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2487 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2491 @section compensationdelay
2493 Compensation Delay Line is a metric based delay to compensate differing
2494 positions of microphones or speakers.
2496 For example, you have recorded guitar with two microphones placed in
2497 different location. Because the front of sound wave has fixed speed in
2498 normal conditions, the phasing of microphones can vary and depends on
2499 their location and interposition. The best sound mix can be achieved when
2500 these microphones are in phase (synchronized). Note that distance of
2501 ~30 cm between microphones makes one microphone to capture signal in
2502 antiphase to another microphone. That makes the final mix sounding moody.
2503 This filter helps to solve phasing problems by adding different delays
2504 to each microphone track and make them synchronized.
2506 The best result can be reached when you take one track as base and
2507 synchronize other tracks one by one with it.
2508 Remember that synchronization/delay tolerance depends on sample rate, too.
2509 Higher sample rates will give more tolerance.
2511 It accepts the following parameters:
2515 Set millimeters distance. This is compensation distance for fine tuning.
2519 Set cm distance. This is compensation distance for tightening distance setup.
2523 Set meters distance. This is compensation distance for hard distance setup.
2527 Set dry amount. Amount of unprocessed (dry) signal.
2531 Set wet amount. Amount of processed (wet) signal.
2535 Set temperature degree in Celsius. This is the temperature of the environment.
2540 Apply headphone crossfeed filter.
2542 Crossfeed is the process of blending the left and right channels of stereo
2544 It is mainly used to reduce extreme stereo separation of low frequencies.
2546 The intent is to produce more speaker like sound to the listener.
2548 The filter accepts the following options:
2552 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2553 This sets gain of low shelf filter for side part of stereo image.
2554 Default is -6dB. Max allowed is -30db when strength is set to 1.
2557 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2558 This sets cut off frequency of low shelf filter. Default is cut off near
2559 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2562 Set input gain. Default is 0.9.
2565 Set output gain. Default is 1.
2568 @section crystalizer
2569 Simple algorithm to expand audio dynamic range.
2571 The filter accepts the following options:
2575 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2576 (unchanged sound) to 10.0 (maximum effect).
2579 Enable clipping. By default is enabled.
2583 Apply a DC shift to the audio.
2585 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2586 in the recording chain) from the audio. The effect of a DC offset is reduced
2587 headroom and hence volume. The @ref{astats} filter can be used to determine if
2588 a signal has a DC offset.
2592 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2596 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2597 used to prevent clipping.
2601 Measure audio dynamic range.
2603 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2604 is found in transition material. And anything less that 8 have very poor dynamics
2605 and is very compressed.
2607 The filter accepts the following options:
2611 Set window length in seconds used to split audio into segments of equal length.
2612 Default is 3 seconds.
2616 Dynamic Audio Normalizer.
2618 This filter applies a certain amount of gain to the input audio in order
2619 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2620 contrast to more "simple" normalization algorithms, the Dynamic Audio
2621 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2622 This allows for applying extra gain to the "quiet" sections of the audio
2623 while avoiding distortions or clipping the "loud" sections. In other words:
2624 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2625 sections, in the sense that the volume of each section is brought to the
2626 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2627 this goal *without* applying "dynamic range compressing". It will retain 100%
2628 of the dynamic range *within* each section of the audio file.
2632 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2633 Default is 500 milliseconds.
2634 The Dynamic Audio Normalizer processes the input audio in small chunks,
2635 referred to as frames. This is required, because a peak magnitude has no
2636 meaning for just a single sample value. Instead, we need to determine the
2637 peak magnitude for a contiguous sequence of sample values. While a "standard"
2638 normalizer would simply use the peak magnitude of the complete file, the
2639 Dynamic Audio Normalizer determines the peak magnitude individually for each
2640 frame. The length of a frame is specified in milliseconds. By default, the
2641 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2642 been found to give good results with most files.
2643 Note that the exact frame length, in number of samples, will be determined
2644 automatically, based on the sampling rate of the individual input audio file.
2647 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2648 number. Default is 31.
2649 Probably the most important parameter of the Dynamic Audio Normalizer is the
2650 @code{window size} of the Gaussian smoothing filter. The filter's window size
2651 is specified in frames, centered around the current frame. For the sake of
2652 simplicity, this must be an odd number. Consequently, the default value of 31
2653 takes into account the current frame, as well as the 15 preceding frames and
2654 the 15 subsequent frames. Using a larger window results in a stronger
2655 smoothing effect and thus in less gain variation, i.e. slower gain
2656 adaptation. Conversely, using a smaller window results in a weaker smoothing
2657 effect and thus in more gain variation, i.e. faster gain adaptation.
2658 In other words, the more you increase this value, the more the Dynamic Audio
2659 Normalizer will behave like a "traditional" normalization filter. On the
2660 contrary, the more you decrease this value, the more the Dynamic Audio
2661 Normalizer will behave like a dynamic range compressor.
2664 Set the target peak value. This specifies the highest permissible magnitude
2665 level for the normalized audio input. This filter will try to approach the
2666 target peak magnitude as closely as possible, but at the same time it also
2667 makes sure that the normalized signal will never exceed the peak magnitude.
2668 A frame's maximum local gain factor is imposed directly by the target peak
2669 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2670 It is not recommended to go above this value.
2673 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2674 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2675 factor for each input frame, i.e. the maximum gain factor that does not
2676 result in clipping or distortion. The maximum gain factor is determined by
2677 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2678 additionally bounds the frame's maximum gain factor by a predetermined
2679 (global) maximum gain factor. This is done in order to avoid excessive gain
2680 factors in "silent" or almost silent frames. By default, the maximum gain
2681 factor is 10.0, For most inputs the default value should be sufficient and
2682 it usually is not recommended to increase this value. Though, for input
2683 with an extremely low overall volume level, it may be necessary to allow even
2684 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2685 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2686 Instead, a "sigmoid" threshold function will be applied. This way, the
2687 gain factors will smoothly approach the threshold value, but never exceed that
2691 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2692 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2693 This means that the maximum local gain factor for each frame is defined
2694 (only) by the frame's highest magnitude sample. This way, the samples can
2695 be amplified as much as possible without exceeding the maximum signal
2696 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2697 Normalizer can also take into account the frame's root mean square,
2698 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2699 determine the power of a time-varying signal. It is therefore considered
2700 that the RMS is a better approximation of the "perceived loudness" than
2701 just looking at the signal's peak magnitude. Consequently, by adjusting all
2702 frames to a constant RMS value, a uniform "perceived loudness" can be
2703 established. If a target RMS value has been specified, a frame's local gain
2704 factor is defined as the factor that would result in exactly that RMS value.
2705 Note, however, that the maximum local gain factor is still restricted by the
2706 frame's highest magnitude sample, in order to prevent clipping.
2709 Enable channels coupling. By default is enabled.
2710 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2711 amount. This means the same gain factor will be applied to all channels, i.e.
2712 the maximum possible gain factor is determined by the "loudest" channel.
2713 However, in some recordings, it may happen that the volume of the different
2714 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2715 In this case, this option can be used to disable the channel coupling. This way,
2716 the gain factor will be determined independently for each channel, depending
2717 only on the individual channel's highest magnitude sample. This allows for
2718 harmonizing the volume of the different channels.
2721 Enable DC bias correction. By default is disabled.
2722 An audio signal (in the time domain) is a sequence of sample values.
2723 In the Dynamic Audio Normalizer these sample values are represented in the
2724 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2725 audio signal, or "waveform", should be centered around the zero point.
2726 That means if we calculate the mean value of all samples in a file, or in a
2727 single frame, then the result should be 0.0 or at least very close to that
2728 value. If, however, there is a significant deviation of the mean value from
2729 0.0, in either positive or negative direction, this is referred to as a
2730 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2731 Audio Normalizer provides optional DC bias correction.
2732 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2733 the mean value, or "DC correction" offset, of each input frame and subtract
2734 that value from all of the frame's sample values which ensures those samples
2735 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2736 boundaries, the DC correction offset values will be interpolated smoothly
2737 between neighbouring frames.
2740 Enable alternative boundary mode. By default is disabled.
2741 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2742 around each frame. This includes the preceding frames as well as the
2743 subsequent frames. However, for the "boundary" frames, located at the very
2744 beginning and at the very end of the audio file, not all neighbouring
2745 frames are available. In particular, for the first few frames in the audio
2746 file, the preceding frames are not known. And, similarly, for the last few
2747 frames in the audio file, the subsequent frames are not known. Thus, the
2748 question arises which gain factors should be assumed for the missing frames
2749 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2750 to deal with this situation. The default boundary mode assumes a gain factor
2751 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2752 "fade out" at the beginning and at the end of the input, respectively.
2755 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2756 By default, the Dynamic Audio Normalizer does not apply "traditional"
2757 compression. This means that signal peaks will not be pruned and thus the
2758 full dynamic range will be retained within each local neighbourhood. However,
2759 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2760 normalization algorithm with a more "traditional" compression.
2761 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2762 (thresholding) function. If (and only if) the compression feature is enabled,
2763 all input frames will be processed by a soft knee thresholding function prior
2764 to the actual normalization process. Put simply, the thresholding function is
2765 going to prune all samples whose magnitude exceeds a certain threshold value.
2766 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2767 value. Instead, the threshold value will be adjusted for each individual
2769 In general, smaller parameters result in stronger compression, and vice versa.
2770 Values below 3.0 are not recommended, because audible distortion may appear.
2775 Make audio easier to listen to on headphones.
2777 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2778 so that when listened to on headphones the stereo image is moved from
2779 inside your head (standard for headphones) to outside and in front of
2780 the listener (standard for speakers).
2786 Apply a two-pole peaking equalisation (EQ) filter. With this
2787 filter, the signal-level at and around a selected frequency can
2788 be increased or decreased, whilst (unlike bandpass and bandreject
2789 filters) that at all other frequencies is unchanged.
2791 In order to produce complex equalisation curves, this filter can
2792 be given several times, each with a different central frequency.
2794 The filter accepts the following options:
2798 Set the filter's central frequency in Hz.
2801 Set method to specify band-width of filter.
2816 Specify the band-width of a filter in width_type units.
2819 Set the required gain or attenuation in dB.
2820 Beware of clipping when using a positive gain.
2823 Specify which channels to filter, by default all available are filtered.
2826 @subsection Examples
2829 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2831 equalizer=f=1000:t=h:width=200:g=-10
2835 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2837 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2841 @subsection Commands
2843 This filter supports the following commands:
2846 Change equalizer frequency.
2847 Syntax for the command is : "@var{frequency}"
2850 Change equalizer width_type.
2851 Syntax for the command is : "@var{width_type}"
2854 Change equalizer width.
2855 Syntax for the command is : "@var{width}"
2858 Change equalizer gain.
2859 Syntax for the command is : "@var{gain}"
2862 @section extrastereo
2864 Linearly increases the difference between left and right channels which
2865 adds some sort of "live" effect to playback.
2867 The filter accepts the following options:
2871 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2872 (average of both channels), with 1.0 sound will be unchanged, with
2873 -1.0 left and right channels will be swapped.
2876 Enable clipping. By default is enabled.
2879 @section firequalizer
2880 Apply FIR Equalization using arbitrary frequency response.
2882 The filter accepts the following option:
2886 Set gain curve equation (in dB). The expression can contain variables:
2889 the evaluated frequency
2893 channel number, set to 0 when multichannels evaluation is disabled
2895 channel id, see libavutil/channel_layout.h, set to the first channel id when
2896 multichannels evaluation is disabled
2900 channel_layout, see libavutil/channel_layout.h
2905 @item gain_interpolate(f)
2906 interpolate gain on frequency f based on gain_entry
2907 @item cubic_interpolate(f)
2908 same as gain_interpolate, but smoother
2910 This option is also available as command. Default is @code{gain_interpolate(f)}.
2913 Set gain entry for gain_interpolate function. The expression can
2917 store gain entry at frequency f with value g
2919 This option is also available as command.
2922 Set filter delay in seconds. Higher value means more accurate.
2923 Default is @code{0.01}.
2926 Set filter accuracy in Hz. Lower value means more accurate.
2927 Default is @code{5}.
2930 Set window function. Acceptable values are:
2933 rectangular window, useful when gain curve is already smooth
2935 hann window (default)
2941 3-terms continuous 1st derivative nuttall window
2943 minimum 3-terms discontinuous nuttall window
2945 4-terms continuous 1st derivative nuttall window
2947 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2949 blackman-harris window
2955 If enabled, use fixed number of audio samples. This improves speed when
2956 filtering with large delay. Default is disabled.
2959 Enable multichannels evaluation on gain. Default is disabled.
2962 Enable zero phase mode by subtracting timestamp to compensate delay.
2963 Default is disabled.
2966 Set scale used by gain. Acceptable values are:
2969 linear frequency, linear gain
2971 linear frequency, logarithmic (in dB) gain (default)
2973 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2975 logarithmic frequency, logarithmic gain
2979 Set file for dumping, suitable for gnuplot.
2982 Set scale for dumpfile. Acceptable values are same with scale option.
2986 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2987 Default is disabled.
2990 Enable minimum phase impulse response. Default is disabled.
2993 @subsection Examples
2998 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3001 lowpass at 1000 Hz with gain_entry:
3003 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3006 custom equalization:
3008 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3011 higher delay with zero phase to compensate delay:
3013 firequalizer=delay=0.1:fixed=on:zero_phase=on
3016 lowpass on left channel, highpass on right channel:
3018 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3019 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3024 Apply a flanging effect to the audio.
3026 The filter accepts the following options:
3030 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3033 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3036 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3040 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3041 Default value is 71.
3044 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3047 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3048 Default value is @var{sinusoidal}.
3051 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3052 Default value is 25.
3055 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3056 Default is @var{linear}.
3060 Apply Haas effect to audio.
3062 Note that this makes most sense to apply on mono signals.
3063 With this filter applied to mono signals it give some directionality and
3064 stretches its stereo image.
3066 The filter accepts the following options:
3070 Set input level. By default is @var{1}, or 0dB
3073 Set output level. By default is @var{1}, or 0dB.
3076 Set gain applied to side part of signal. By default is @var{1}.
3079 Set kind of middle source. Can be one of the following:
3089 Pick middle part signal of stereo image.
3092 Pick side part signal of stereo image.
3096 Change middle phase. By default is disabled.
3099 Set left channel delay. By default is @var{2.05} milliseconds.
3102 Set left channel balance. By default is @var{-1}.
3105 Set left channel gain. By default is @var{1}.
3108 Change left phase. By default is disabled.
3111 Set right channel delay. By defaults is @var{2.12} milliseconds.
3114 Set right channel balance. By default is @var{1}.
3117 Set right channel gain. By default is @var{1}.
3120 Change right phase. By default is enabled.
3125 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3126 embedded HDCD codes is expanded into a 20-bit PCM stream.
3128 The filter supports the Peak Extend and Low-level Gain Adjustment features
3129 of HDCD, and detects the Transient Filter flag.
3132 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3135 When using the filter with wav, note the default encoding for wav is 16-bit,
3136 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3137 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3139 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3140 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3143 The filter accepts the following options:
3146 @item disable_autoconvert
3147 Disable any automatic format conversion or resampling in the filter graph.
3149 @item process_stereo
3150 Process the stereo channels together. If target_gain does not match between
3151 channels, consider it invalid and use the last valid target_gain.
3154 Set the code detect timer period in ms.
3157 Always extend peaks above -3dBFS even if PE isn't signaled.
3160 Replace audio with a solid tone and adjust the amplitude to signal some
3161 specific aspect of the decoding process. The output file can be loaded in
3162 an audio editor alongside the original to aid analysis.
3164 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3171 Gain adjustment level at each sample
3173 Samples where peak extend occurs
3175 Samples where the code detect timer is active
3177 Samples where the target gain does not match between channels
3183 Apply head-related transfer functions (HRTFs) to create virtual
3184 loudspeakers around the user for binaural listening via headphones.
3185 The HRIRs are provided via additional streams, for each channel
3186 one stereo input stream is needed.
3188 The filter accepts the following options:
3192 Set mapping of input streams for convolution.
3193 The argument is a '|'-separated list of channel names in order as they
3194 are given as additional stream inputs for filter.
3195 This also specify number of input streams. Number of input streams
3196 must be not less than number of channels in first stream plus one.
3199 Set gain applied to audio. Value is in dB. Default is 0.
3202 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3203 processing audio in time domain which is slow.
3204 @var{freq} is processing audio in frequency domain which is fast.
3205 Default is @var{freq}.
3208 Set custom gain for LFE channels. Value is in dB. Default is 0.
3211 Set size of frame in number of samples which will be processed at once.
3212 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3215 Set format of hrir stream.
3216 Default value is @var{stereo}. Alternative value is @var{multich}.
3217 If value is set to @var{stereo}, number of additional streams should
3218 be greater or equal to number of input channels in first input stream.
3219 Also each additional stream should have stereo number of channels.
3220 If value is set to @var{multich}, number of additional streams should
3221 be exactly one. Also number of input channels of additional stream
3222 should be equal or greater than twice number of channels of first input
3226 @subsection Examples
3230 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3231 each amovie filter use stereo file with IR coefficients as input.
3232 The files give coefficients for each position of virtual loudspeaker:
3234 ffmpeg -i input.wav -lavfi-complex "amovie=azi_270_ele_0_DFC.wav[sr],amovie=azi_90_ele_0_DFC.wav[sl],amovie=azi_225_ele_0_DFC.wav[br],amovie=azi_135_ele_0_DFC.wav[bl],amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe],amovie=azi_35_ele_0_DFC.wav[fl],amovie=azi_325_ele_0_DFC.wav[fr],[a:0][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3239 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3240 but now in @var{multich} @var{hrir} format.
3242 ffmpeg -i input.wav -lavfi-complex "amovie=minp.wav[hrirs],[a:0][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3249 Apply a high-pass filter with 3dB point frequency.
3250 The filter can be either single-pole, or double-pole (the default).
3251 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3253 The filter accepts the following options:
3257 Set frequency in Hz. Default is 3000.
3260 Set number of poles. Default is 2.
3263 Set method to specify band-width of filter.
3278 Specify the band-width of a filter in width_type units.
3279 Applies only to double-pole filter.
3280 The default is 0.707q and gives a Butterworth response.
3283 Specify which channels to filter, by default all available are filtered.
3286 @subsection Commands
3288 This filter supports the following commands:
3291 Change highpass frequency.
3292 Syntax for the command is : "@var{frequency}"
3295 Change highpass width_type.
3296 Syntax for the command is : "@var{width_type}"
3299 Change highpass width.
3300 Syntax for the command is : "@var{width}"
3305 Join multiple input streams into one multi-channel stream.
3307 It accepts the following parameters:
3311 The number of input streams. It defaults to 2.
3313 @item channel_layout
3314 The desired output channel layout. It defaults to stereo.
3317 Map channels from inputs to output. The argument is a '|'-separated list of
3318 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3319 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3320 can be either the name of the input channel (e.g. FL for front left) or its
3321 index in the specified input stream. @var{out_channel} is the name of the output
3325 The filter will attempt to guess the mappings when they are not specified
3326 explicitly. It does so by first trying to find an unused matching input channel
3327 and if that fails it picks the first unused input channel.
3329 Join 3 inputs (with properly set channel layouts):
3331 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3334 Build a 5.1 output from 6 single-channel streams:
3336 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3337 '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'
3343 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3345 To enable compilation of this filter you need to configure FFmpeg with
3346 @code{--enable-ladspa}.
3350 Specifies the name of LADSPA plugin library to load. If the environment
3351 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3352 each one of the directories specified by the colon separated list in
3353 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3354 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3355 @file{/usr/lib/ladspa/}.
3358 Specifies the plugin within the library. Some libraries contain only
3359 one plugin, but others contain many of them. If this is not set filter
3360 will list all available plugins within the specified library.
3363 Set the '|' separated list of controls which are zero or more floating point
3364 values that determine the behavior of the loaded plugin (for example delay,
3366 Controls need to be defined using the following syntax:
3367 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3368 @var{valuei} is the value set on the @var{i}-th control.
3369 Alternatively they can be also defined using the following syntax:
3370 @var{value0}|@var{value1}|@var{value2}|..., where
3371 @var{valuei} is the value set on the @var{i}-th control.
3372 If @option{controls} is set to @code{help}, all available controls and
3373 their valid ranges are printed.
3375 @item sample_rate, s
3376 Specify the sample rate, default to 44100. Only used if plugin have
3380 Set the number of samples per channel per each output frame, default
3381 is 1024. Only used if plugin have zero inputs.
3384 Set the minimum duration of the sourced audio. See
3385 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3386 for the accepted syntax.
3387 Note that the resulting duration may be greater than the specified duration,
3388 as the generated audio is always cut at the end of a complete frame.
3389 If not specified, or the expressed duration is negative, the audio is
3390 supposed to be generated forever.
3391 Only used if plugin have zero inputs.
3395 @subsection Examples
3399 List all available plugins within amp (LADSPA example plugin) library:
3405 List all available controls and their valid ranges for @code{vcf_notch}
3406 plugin from @code{VCF} library:
3408 ladspa=f=vcf:p=vcf_notch:c=help
3412 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3415 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3419 Add reverberation to the audio using TAP-plugins
3420 (Tom's Audio Processing plugins):
3422 ladspa=file=tap_reverb:tap_reverb
3426 Generate white noise, with 0.2 amplitude:
3428 ladspa=file=cmt:noise_source_white:c=c0=.2
3432 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3433 @code{C* Audio Plugin Suite} (CAPS) library:
3435 ladspa=file=caps:Click:c=c1=20'
3439 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3441 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3445 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3446 @code{SWH Plugins} collection:
3448 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3452 Attenuate low frequencies using Multiband EQ from Steve Harris
3453 @code{SWH Plugins} collection:
3455 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3459 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3462 ladspa=caps:Narrower
3466 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3468 ladspa=caps:White:.2
3472 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3474 ladspa=caps:Fractal:c=c1=1
3478 Dynamic volume normalization using @code{VLevel} plugin:
3480 ladspa=vlevel-ladspa:vlevel_mono
3484 @subsection Commands
3486 This filter supports the following commands:
3489 Modify the @var{N}-th control value.
3491 If the specified value is not valid, it is ignored and prior one is kept.
3496 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3497 Support for both single pass (livestreams, files) and double pass (files) modes.
3498 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3499 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3500 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3502 The filter accepts the following options:
3506 Set integrated loudness target.
3507 Range is -70.0 - -5.0. Default value is -24.0.
3510 Set loudness range target.
3511 Range is 1.0 - 20.0. Default value is 7.0.
3514 Set maximum true peak.
3515 Range is -9.0 - +0.0. Default value is -2.0.
3517 @item measured_I, measured_i
3518 Measured IL of input file.
3519 Range is -99.0 - +0.0.
3521 @item measured_LRA, measured_lra
3522 Measured LRA of input file.
3523 Range is 0.0 - 99.0.
3525 @item measured_TP, measured_tp
3526 Measured true peak of input file.
3527 Range is -99.0 - +99.0.
3529 @item measured_thresh
3530 Measured threshold of input file.
3531 Range is -99.0 - +0.0.
3534 Set offset gain. Gain is applied before the true-peak limiter.
3535 Range is -99.0 - +99.0. Default is +0.0.
3538 Normalize linearly if possible.
3539 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3540 to be specified in order to use this mode.
3541 Options are true or false. Default is true.
3544 Treat mono input files as "dual-mono". If a mono file is intended for playback
3545 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3546 If set to @code{true}, this option will compensate for this effect.
3547 Multi-channel input files are not affected by this option.
3548 Options are true or false. Default is false.
3551 Set print format for stats. Options are summary, json, or none.
3552 Default value is none.
3557 Apply a low-pass filter with 3dB point frequency.
3558 The filter can be either single-pole or double-pole (the default).
3559 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3561 The filter accepts the following options:
3565 Set frequency in Hz. Default is 500.
3568 Set number of poles. Default is 2.
3571 Set method to specify band-width of filter.
3586 Specify the band-width of a filter in width_type units.
3587 Applies only to double-pole filter.
3588 The default is 0.707q and gives a Butterworth response.
3591 Specify which channels to filter, by default all available are filtered.
3594 @subsection Examples
3597 Lowpass only LFE channel, it LFE is not present it does nothing:
3603 @subsection Commands
3605 This filter supports the following commands:
3608 Change lowpass frequency.
3609 Syntax for the command is : "@var{frequency}"
3612 Change lowpass width_type.
3613 Syntax for the command is : "@var{width_type}"
3616 Change lowpass width.
3617 Syntax for the command is : "@var{width}"
3622 Load a LV2 (LADSPA Version 2) plugin.
3624 To enable compilation of this filter you need to configure FFmpeg with
3625 @code{--enable-lv2}.
3629 Specifies the plugin URI. You may need to escape ':'.
3632 Set the '|' separated list of controls which are zero or more floating point
3633 values that determine the behavior of the loaded plugin (for example delay,
3635 If @option{controls} is set to @code{help}, all available controls and
3636 their valid ranges are printed.
3638 @item sample_rate, s
3639 Specify the sample rate, default to 44100. Only used if plugin have
3643 Set the number of samples per channel per each output frame, default
3644 is 1024. Only used if plugin have zero inputs.
3647 Set the minimum duration of the sourced audio. See
3648 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3649 for the accepted syntax.
3650 Note that the resulting duration may be greater than the specified duration,
3651 as the generated audio is always cut at the end of a complete frame.
3652 If not specified, or the expressed duration is negative, the audio is
3653 supposed to be generated forever.
3654 Only used if plugin have zero inputs.
3657 @subsection Examples
3661 Apply bass enhancer plugin from Calf:
3663 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3667 Apply vinyl plugin from Calf:
3669 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3673 Apply bit crusher plugin from ArtyFX:
3675 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3680 Multiband Compress or expand the audio's dynamic range.
3682 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3683 This is akin to the crossover of a loudspeaker, and results in flat frequency
3684 response when absent compander action.
3686 It accepts the following parameters:
3690 This option syntax is:
3691 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3692 For explanation of each item refer to compand filter documentation.
3698 Mix channels with specific gain levels. The filter accepts the output
3699 channel layout followed by a set of channels definitions.
3701 This filter is also designed to efficiently remap the channels of an audio
3704 The filter accepts parameters of the form:
3705 "@var{l}|@var{outdef}|@var{outdef}|..."
3709 output channel layout or number of channels
3712 output channel specification, of the form:
3713 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3716 output channel to define, either a channel name (FL, FR, etc.) or a channel
3717 number (c0, c1, etc.)
3720 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3723 input channel to use, see out_name for details; it is not possible to mix
3724 named and numbered input channels
3727 If the `=' in a channel specification is replaced by `<', then the gains for
3728 that specification will be renormalized so that the total is 1, thus
3729 avoiding clipping noise.
3731 @subsection Mixing examples
3733 For example, if you want to down-mix from stereo to mono, but with a bigger
3734 factor for the left channel:
3736 pan=1c|c0=0.9*c0+0.1*c1
3739 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3740 7-channels surround:
3742 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3745 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3746 that should be preferred (see "-ac" option) unless you have very specific
3749 @subsection Remapping examples
3751 The channel remapping will be effective if, and only if:
3754 @item gain coefficients are zeroes or ones,
3755 @item only one input per channel output,
3758 If all these conditions are satisfied, the filter will notify the user ("Pure
3759 channel mapping detected"), and use an optimized and lossless method to do the
3762 For example, if you have a 5.1 source and want a stereo audio stream by
3763 dropping the extra channels:
3765 pan="stereo| c0=FL | c1=FR"
3768 Given the same source, you can also switch front left and front right channels
3769 and keep the input channel layout:
3771 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3774 If the input is a stereo audio stream, you can mute the front left channel (and
3775 still keep the stereo channel layout) with:
3780 Still with a stereo audio stream input, you can copy the right channel in both
3781 front left and right:
3783 pan="stereo| c0=FR | c1=FR"
3788 ReplayGain scanner filter. This filter takes an audio stream as an input and
3789 outputs it unchanged.
3790 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3794 Convert the audio sample format, sample rate and channel layout. It is
3795 not meant to be used directly.
3798 Apply time-stretching and pitch-shifting with librubberband.
3800 The filter accepts the following options:
3804 Set tempo scale factor.
3807 Set pitch scale factor.
3810 Set transients detector.
3811 Possible values are:
3820 Possible values are:
3829 Possible values are:
3836 Set processing window size.
3837 Possible values are:
3846 Possible values are:
3853 Enable formant preservation when shift pitching.
3854 Possible values are:
3862 Possible values are:
3871 Possible values are:
3878 @section sidechaincompress
3880 This filter acts like normal compressor but has the ability to compress
3881 detected signal using second input signal.
3882 It needs two input streams and returns one output stream.
3883 First input stream will be processed depending on second stream signal.
3884 The filtered signal then can be filtered with other filters in later stages of
3885 processing. See @ref{pan} and @ref{amerge} filter.
3887 The filter accepts the following options:
3891 Set input gain. Default is 1. Range is between 0.015625 and 64.
3894 If a signal of second stream raises above this level it will affect the gain
3895 reduction of first stream.
3896 By default is 0.125. Range is between 0.00097563 and 1.
3899 Set a ratio about which the signal is reduced. 1:2 means that if the level
3900 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3901 Default is 2. Range is between 1 and 20.
3904 Amount of milliseconds the signal has to rise above the threshold before gain
3905 reduction starts. Default is 20. Range is between 0.01 and 2000.
3908 Amount of milliseconds the signal has to fall below the threshold before
3909 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3912 Set the amount by how much signal will be amplified after processing.
3913 Default is 1. Range is from 1 to 64.
3916 Curve the sharp knee around the threshold to enter gain reduction more softly.
3917 Default is 2.82843. Range is between 1 and 8.
3920 Choose if the @code{average} level between all channels of side-chain stream
3921 or the louder(@code{maximum}) channel of side-chain stream affects the
3922 reduction. Default is @code{average}.
3925 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3926 of @code{rms}. Default is @code{rms} which is mainly smoother.
3929 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3932 How much to use compressed signal in output. Default is 1.
3933 Range is between 0 and 1.
3936 @subsection Examples
3940 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3941 depending on the signal of 2nd input and later compressed signal to be
3942 merged with 2nd input:
3944 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3948 @section sidechaingate
3950 A sidechain gate acts like a normal (wideband) gate but has the ability to
3951 filter the detected signal before sending it to the gain reduction stage.
3952 Normally a gate uses the full range signal to detect a level above the
3954 For example: If you cut all lower frequencies from your sidechain signal
3955 the gate will decrease the volume of your track only if not enough highs
3956 appear. With this technique you are able to reduce the resonation of a
3957 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3959 It needs two input streams and returns one output stream.
3960 First input stream will be processed depending on second stream signal.
3962 The filter accepts the following options:
3966 Set input level before filtering.
3967 Default is 1. Allowed range is from 0.015625 to 64.
3970 Set the level of gain reduction when the signal is below the threshold.
3971 Default is 0.06125. Allowed range is from 0 to 1.
3974 If a signal rises above this level the gain reduction is released.
3975 Default is 0.125. Allowed range is from 0 to 1.
3978 Set a ratio about which the signal is reduced.
3979 Default is 2. Allowed range is from 1 to 9000.
3982 Amount of milliseconds the signal has to rise above the threshold before gain
3984 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3987 Amount of milliseconds the signal has to fall below the threshold before the
3988 reduction is increased again. Default is 250 milliseconds.
3989 Allowed range is from 0.01 to 9000.
3992 Set amount of amplification of signal after processing.
3993 Default is 1. Allowed range is from 1 to 64.
3996 Curve the sharp knee around the threshold to enter gain reduction more softly.
3997 Default is 2.828427125. Allowed range is from 1 to 8.
4000 Choose if exact signal should be taken for detection or an RMS like one.
4001 Default is rms. Can be peak or rms.
4004 Choose if the average level between all channels or the louder channel affects
4006 Default is average. Can be average or maximum.
4009 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4012 @section silencedetect
4014 Detect silence in an audio stream.
4016 This filter logs a message when it detects that the input audio volume is less
4017 or equal to a noise tolerance value for a duration greater or equal to the
4018 minimum detected noise duration.
4020 The printed times and duration are expressed in seconds.
4022 The filter accepts the following options:
4026 Set silence duration until notification (default is 2 seconds).
4029 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4030 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4033 @subsection Examples
4037 Detect 5 seconds of silence with -50dB noise tolerance:
4039 silencedetect=n=-50dB:d=5
4043 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4044 tolerance in @file{silence.mp3}:
4046 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4050 @section silenceremove
4052 Remove silence from the beginning, middle or end of the audio.
4054 The filter accepts the following options:
4058 This value is used to indicate if audio should be trimmed at beginning of
4059 the audio. A value of zero indicates no silence should be trimmed from the
4060 beginning. When specifying a non-zero value, it trims audio up until it
4061 finds non-silence. Normally, when trimming silence from beginning of audio
4062 the @var{start_periods} will be @code{1} but it can be increased to higher
4063 values to trim all audio up to specific count of non-silence periods.
4064 Default value is @code{0}.
4066 @item start_duration
4067 Specify the amount of time that non-silence must be detected before it stops
4068 trimming audio. By increasing the duration, bursts of noises can be treated
4069 as silence and trimmed off. Default value is @code{0}.
4071 @item start_threshold
4072 This indicates what sample value should be treated as silence. For digital
4073 audio, a value of @code{0} may be fine but for audio recorded from analog,
4074 you may wish to increase the value to account for background noise.
4075 Can be specified in dB (in case "dB" is appended to the specified value)
4076 or amplitude ratio. Default value is @code{0}.
4079 Set the count for trimming silence from the end of audio.
4080 To remove silence from the middle of a file, specify a @var{stop_periods}
4081 that is negative. This value is then treated as a positive value and is
4082 used to indicate the effect should restart processing as specified by
4083 @var{start_periods}, making it suitable for removing periods of silence
4084 in the middle of the audio.
4085 Default value is @code{0}.
4088 Specify a duration of silence that must exist before audio is not copied any
4089 more. By specifying a higher duration, silence that is wanted can be left in
4091 Default value is @code{0}.
4093 @item stop_threshold
4094 This is the same as @option{start_threshold} but for trimming silence from
4096 Can be specified in dB (in case "dB" is appended to the specified value)
4097 or amplitude ratio. Default value is @code{0}.
4100 This indicates that @var{stop_duration} length of audio should be left intact
4101 at the beginning of each period of silence.
4102 For example, if you want to remove long pauses between words but do not want
4103 to remove the pauses completely. Default value is @code{0}.
4106 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4107 and works better with digital silence which is exactly 0.
4108 Default value is @code{rms}.
4111 Set ratio used to calculate size of window for detecting silence.
4112 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4115 @subsection Examples
4119 The following example shows how this filter can be used to start a recording
4120 that does not contain the delay at the start which usually occurs between
4121 pressing the record button and the start of the performance:
4123 silenceremove=1:5:0.02
4127 Trim all silence encountered from beginning to end where there is more than 1
4128 second of silence in audio:
4130 silenceremove=0:0:0:-1:1:-90dB
4136 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4137 loudspeakers around the user for binaural listening via headphones (audio
4138 formats up to 9 channels supported).
4139 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4140 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4141 Austrian Academy of Sciences.
4143 To enable compilation of this filter you need to configure FFmpeg with
4144 @code{--enable-libmysofa}.
4146 The filter accepts the following options:
4150 Set the SOFA file used for rendering.
4153 Set gain applied to audio. Value is in dB. Default is 0.
4156 Set rotation of virtual loudspeakers in deg. Default is 0.
4159 Set elevation of virtual speakers in deg. Default is 0.
4162 Set distance in meters between loudspeakers and the listener with near-field
4163 HRTFs. Default is 1.
4166 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4167 processing audio in time domain which is slow.
4168 @var{freq} is processing audio in frequency domain which is fast.
4169 Default is @var{freq}.
4172 Set custom positions of virtual loudspeakers. Syntax for this option is:
4173 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4174 Each virtual loudspeaker is described with short channel name following with
4175 azimuth and elevation in degrees.
4176 Each virtual loudspeaker description is separated by '|'.
4177 For example to override front left and front right channel positions use:
4178 'speakers=FL 45 15|FR 345 15'.
4179 Descriptions with unrecognised channel names are ignored.
4182 Set custom gain for LFE channels. Value is in dB. Default is 0.
4185 @subsection Examples
4189 Using ClubFritz6 sofa file:
4191 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4195 Using ClubFritz12 sofa file and bigger radius with small rotation:
4197 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4201 Similar as above but with custom speaker positions for front left, front right, back left and back right
4202 and also with custom gain:
4204 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4208 @section stereotools
4210 This filter has some handy utilities to manage stereo signals, for converting
4211 M/S stereo recordings to L/R signal while having control over the parameters
4212 or spreading the stereo image of master track.
4214 The filter accepts the following options:
4218 Set input level before filtering for both channels. Defaults is 1.
4219 Allowed range is from 0.015625 to 64.
4222 Set output level after filtering for both channels. Defaults is 1.
4223 Allowed range is from 0.015625 to 64.
4226 Set input balance between both channels. Default is 0.
4227 Allowed range is from -1 to 1.
4230 Set output balance between both channels. Default is 0.
4231 Allowed range is from -1 to 1.
4234 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4235 clipping. Disabled by default.
4238 Mute the left channel. Disabled by default.
4241 Mute the right channel. Disabled by default.
4244 Change the phase of the left channel. Disabled by default.
4247 Change the phase of the right channel. Disabled by default.
4250 Set stereo mode. Available values are:
4254 Left/Right to Left/Right, this is default.
4257 Left/Right to Mid/Side.
4260 Mid/Side to Left/Right.
4263 Left/Right to Left/Left.
4266 Left/Right to Right/Right.
4269 Left/Right to Left + Right.
4272 Left/Right to Right/Left.
4275 Mid/Side to Left/Left.
4278 Mid/Side to Right/Right.
4282 Set level of side signal. Default is 1.
4283 Allowed range is from 0.015625 to 64.
4286 Set balance of side signal. Default is 0.
4287 Allowed range is from -1 to 1.
4290 Set level of the middle signal. Default is 1.
4291 Allowed range is from 0.015625 to 64.
4294 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4297 Set stereo base between mono and inversed channels. Default is 0.
4298 Allowed range is from -1 to 1.
4301 Set delay in milliseconds how much to delay left from right channel and
4302 vice versa. Default is 0. Allowed range is from -20 to 20.
4305 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4308 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4310 @item bmode_in, bmode_out
4311 Set balance mode for balance_in/balance_out option.
4313 Can be one of the following:
4317 Classic balance mode. Attenuate one channel at time.
4318 Gain is raised up to 1.
4321 Similar as classic mode above but gain is raised up to 2.
4324 Equal power distribution, from -6dB to +6dB range.
4328 @subsection Examples
4332 Apply karaoke like effect:
4334 stereotools=mlev=0.015625
4338 Convert M/S signal to L/R:
4340 "stereotools=mode=ms>lr"
4344 @section stereowiden
4346 This filter enhance the stereo effect by suppressing signal common to both
4347 channels and by delaying the signal of left into right and vice versa,
4348 thereby widening the stereo effect.
4350 The filter accepts the following options:
4354 Time in milliseconds of the delay of left signal into right and vice versa.
4355 Default is 20 milliseconds.
4358 Amount of gain in delayed signal into right and vice versa. Gives a delay
4359 effect of left signal in right output and vice versa which gives widening
4360 effect. Default is 0.3.
4363 Cross feed of left into right with inverted phase. This helps in suppressing
4364 the mono. If the value is 1 it will cancel all the signal common to both
4365 channels. Default is 0.3.
4368 Set level of input signal of original channel. Default is 0.8.
4371 @section superequalizer
4372 Apply 18 band equalizer.
4374 The filter accepts the following options:
4381 Set 131Hz band gain.
4383 Set 185Hz band gain.
4385 Set 262Hz band gain.
4387 Set 370Hz band gain.
4389 Set 523Hz band gain.
4391 Set 740Hz band gain.
4393 Set 1047Hz band gain.
4395 Set 1480Hz band gain.
4397 Set 2093Hz band gain.
4399 Set 2960Hz band gain.
4401 Set 4186Hz band gain.
4403 Set 5920Hz band gain.
4405 Set 8372Hz band gain.
4407 Set 11840Hz band gain.
4409 Set 16744Hz band gain.
4411 Set 20000Hz band gain.
4415 Apply audio surround upmix filter.
4417 This filter allows to produce multichannel output from audio stream.
4419 The filter accepts the following options:
4423 Set output channel layout. By default, this is @var{5.1}.
4425 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4426 for the required syntax.
4429 Set input channel layout. By default, this is @var{stereo}.
4431 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4432 for the required syntax.
4435 Set input volume level. By default, this is @var{1}.
4438 Set output volume level. By default, this is @var{1}.
4441 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4444 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4447 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4450 Set front center input volume. By default, this is @var{1}.
4453 Set front center output volume. By default, this is @var{1}.
4456 Set LFE input volume. By default, this is @var{1}.
4459 Set LFE output volume. By default, this is @var{1}.
4462 @section treble, highshelf
4464 Boost or cut treble (upper) frequencies of the audio using a two-pole
4465 shelving filter with a response similar to that of a standard
4466 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4468 The filter accepts the following options:
4472 Give the gain at whichever is the lower of ~22 kHz and the
4473 Nyquist frequency. Its useful range is about -20 (for a large cut)
4474 to +20 (for a large boost). Beware of clipping when using a positive gain.
4477 Set the filter's central frequency and so can be used
4478 to extend or reduce the frequency range to be boosted or cut.
4479 The default value is @code{3000} Hz.
4482 Set method to specify band-width of filter.
4497 Determine how steep is the filter's shelf transition.
4500 Specify which channels to filter, by default all available are filtered.
4503 @subsection Commands
4505 This filter supports the following commands:
4508 Change treble frequency.
4509 Syntax for the command is : "@var{frequency}"
4512 Change treble width_type.
4513 Syntax for the command is : "@var{width_type}"
4516 Change treble width.
4517 Syntax for the command is : "@var{width}"
4521 Syntax for the command is : "@var{gain}"
4526 Sinusoidal amplitude modulation.
4528 The filter accepts the following options:
4532 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4533 (20 Hz or lower) will result in a tremolo effect.
4534 This filter may also be used as a ring modulator by specifying
4535 a modulation frequency higher than 20 Hz.
4536 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4539 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4540 Default value is 0.5.
4545 Sinusoidal phase modulation.
4547 The filter accepts the following options:
4551 Modulation frequency in Hertz.
4552 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4555 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4556 Default value is 0.5.
4561 Adjust the input audio volume.
4563 It accepts the following parameters:
4567 Set audio volume expression.
4569 Output values are clipped to the maximum value.
4571 The output audio volume is given by the relation:
4573 @var{output_volume} = @var{volume} * @var{input_volume}
4576 The default value for @var{volume} is "1.0".
4579 This parameter represents the mathematical precision.
4581 It determines which input sample formats will be allowed, which affects the
4582 precision of the volume scaling.
4586 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4588 32-bit floating-point; this limits input sample format to FLT. (default)
4590 64-bit floating-point; this limits input sample format to DBL.
4594 Choose the behaviour on encountering ReplayGain side data in input frames.
4598 Remove ReplayGain side data, ignoring its contents (the default).
4601 Ignore ReplayGain side data, but leave it in the frame.
4604 Prefer the track gain, if present.
4607 Prefer the album gain, if present.
4610 @item replaygain_preamp
4611 Pre-amplification gain in dB to apply to the selected replaygain gain.
4613 Default value for @var{replaygain_preamp} is 0.0.
4616 Set when the volume expression is evaluated.
4618 It accepts the following values:
4621 only evaluate expression once during the filter initialization, or
4622 when the @samp{volume} command is sent
4625 evaluate expression for each incoming frame
4628 Default value is @samp{once}.
4631 The volume expression can contain the following parameters.
4635 frame number (starting at zero)
4638 @item nb_consumed_samples
4639 number of samples consumed by the filter
4641 number of samples in the current frame
4643 original frame position in the file
4649 PTS at start of stream
4651 time at start of stream
4657 last set volume value
4660 Note that when @option{eval} is set to @samp{once} only the
4661 @var{sample_rate} and @var{tb} variables are available, all other
4662 variables will evaluate to NAN.
4664 @subsection Commands
4666 This filter supports the following commands:
4669 Modify the volume expression.
4670 The command accepts the same syntax of the corresponding option.
4672 If the specified expression is not valid, it is kept at its current
4674 @item replaygain_noclip
4675 Prevent clipping by limiting the gain applied.
4677 Default value for @var{replaygain_noclip} is 1.
4681 @subsection Examples
4685 Halve the input audio volume:
4689 volume=volume=-6.0206dB
4692 In all the above example the named key for @option{volume} can be
4693 omitted, for example like in:
4699 Increase input audio power by 6 decibels using fixed-point precision:
4701 volume=volume=6dB:precision=fixed
4705 Fade volume after time 10 with an annihilation period of 5 seconds:
4707 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4711 @section volumedetect
4713 Detect the volume of the input video.
4715 The filter has no parameters. The input is not modified. Statistics about
4716 the volume will be printed in the log when the input stream end is reached.
4718 In particular it will show the mean volume (root mean square), maximum
4719 volume (on a per-sample basis), and the beginning of a histogram of the
4720 registered volume values (from the maximum value to a cumulated 1/1000 of
4723 All volumes are in decibels relative to the maximum PCM value.
4725 @subsection Examples
4727 Here is an excerpt of the output:
4729 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4730 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4731 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4732 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4733 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4734 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4735 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4736 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4737 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4743 The mean square energy is approximately -27 dB, or 10^-2.7.
4745 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4747 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4750 In other words, raising the volume by +4 dB does not cause any clipping,
4751 raising it by +5 dB causes clipping for 6 samples, etc.
4753 @c man end AUDIO FILTERS
4755 @chapter Audio Sources
4756 @c man begin AUDIO SOURCES
4758 Below is a description of the currently available audio sources.
4762 Buffer audio frames, and make them available to the filter chain.
4764 This source is mainly intended for a programmatic use, in particular
4765 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4767 It accepts the following parameters:
4771 The timebase which will be used for timestamps of submitted frames. It must be
4772 either a floating-point number or in @var{numerator}/@var{denominator} form.
4775 The sample rate of the incoming audio buffers.
4778 The sample format of the incoming audio buffers.
4779 Either a sample format name or its corresponding integer representation from
4780 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4782 @item channel_layout
4783 The channel layout of the incoming audio buffers.
4784 Either a channel layout name from channel_layout_map in
4785 @file{libavutil/channel_layout.c} or its corresponding integer representation
4786 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4789 The number of channels of the incoming audio buffers.
4790 If both @var{channels} and @var{channel_layout} are specified, then they
4795 @subsection Examples
4798 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4801 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4802 Since the sample format with name "s16p" corresponds to the number
4803 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4806 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4811 Generate an audio signal specified by an expression.
4813 This source accepts in input one or more expressions (one for each
4814 channel), which are evaluated and used to generate a corresponding
4817 This source accepts the following options:
4821 Set the '|'-separated expressions list for each separate channel. In case the
4822 @option{channel_layout} option is not specified, the selected channel layout
4823 depends on the number of provided expressions. Otherwise the last
4824 specified expression is applied to the remaining output channels.
4826 @item channel_layout, c
4827 Set the channel layout. The number of channels in the specified layout
4828 must be equal to the number of specified expressions.
4831 Set the minimum duration of the sourced audio. See
4832 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4833 for the accepted syntax.
4834 Note that the resulting duration may be greater than the specified
4835 duration, as the generated audio is always cut at the end of a
4838 If not specified, or the expressed duration is negative, the audio is
4839 supposed to be generated forever.
4842 Set the number of samples per channel per each output frame,
4845 @item sample_rate, s
4846 Specify the sample rate, default to 44100.
4849 Each expression in @var{exprs} can contain the following constants:
4853 number of the evaluated sample, starting from 0
4856 time of the evaluated sample expressed in seconds, starting from 0
4863 @subsection Examples
4873 Generate a sin signal with frequency of 440 Hz, set sample rate to
4876 aevalsrc="sin(440*2*PI*t):s=8000"
4880 Generate a two channels signal, specify the channel layout (Front
4881 Center + Back Center) explicitly:
4883 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4887 Generate white noise:
4889 aevalsrc="-2+random(0)"
4893 Generate an amplitude modulated signal:
4895 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4899 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4901 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4908 The null audio source, return unprocessed audio frames. It is mainly useful
4909 as a template and to be employed in analysis / debugging tools, or as
4910 the source for filters which ignore the input data (for example the sox
4913 This source accepts the following options:
4917 @item channel_layout, cl
4919 Specifies the channel layout, and can be either an integer or a string
4920 representing a channel layout. The default value of @var{channel_layout}
4923 Check the channel_layout_map definition in
4924 @file{libavutil/channel_layout.c} for the mapping between strings and
4925 channel layout values.
4927 @item sample_rate, r
4928 Specifies the sample rate, and defaults to 44100.
4931 Set the number of samples per requested frames.
4935 @subsection Examples
4939 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4941 anullsrc=r=48000:cl=4
4945 Do the same operation with a more obvious syntax:
4947 anullsrc=r=48000:cl=mono
4951 All the parameters need to be explicitly defined.
4955 Synthesize a voice utterance using the libflite library.
4957 To enable compilation of this filter you need to configure FFmpeg with
4958 @code{--enable-libflite}.
4960 Note that versions of the flite library prior to 2.0 are not thread-safe.
4962 The filter accepts the following options:
4967 If set to 1, list the names of the available voices and exit
4968 immediately. Default value is 0.
4971 Set the maximum number of samples per frame. Default value is 512.
4974 Set the filename containing the text to speak.
4977 Set the text to speak.
4980 Set the voice to use for the speech synthesis. Default value is
4981 @code{kal}. See also the @var{list_voices} option.
4984 @subsection Examples
4988 Read from file @file{speech.txt}, and synthesize the text using the
4989 standard flite voice:
4991 flite=textfile=speech.txt
4995 Read the specified text selecting the @code{slt} voice:
4997 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5001 Input text to ffmpeg:
5003 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5007 Make @file{ffplay} speak the specified text, using @code{flite} and
5008 the @code{lavfi} device:
5010 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5014 For more information about libflite, check:
5015 @url{http://www.festvox.org/flite/}
5019 Generate a noise audio signal.
5021 The filter accepts the following options:
5024 @item sample_rate, r
5025 Specify the sample rate. Default value is 48000 Hz.
5028 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5032 Specify the duration of the generated audio stream. Not specifying this option
5033 results in noise with an infinite length.
5035 @item color, colour, c
5036 Specify the color of noise. Available noise colors are white, pink, brown,
5037 blue and violet. Default color is white.
5040 Specify a value used to seed the PRNG.
5043 Set the number of samples per each output frame, default is 1024.
5046 @subsection Examples
5051 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5053 anoisesrc=d=60:c=pink:r=44100:a=0.5
5059 Generate odd-tap Hilbert transform FIR coefficients.
5061 The resulting stream can be used with @ref{afir} filter for phase-shifting
5062 the signal by 90 degrees.
5064 This is used in many matrix coding schemes and for analytic signal generation.
5065 The process is often written as a multiplication by i (or j), the imaginary unit.
5067 The filter accepts the following options:
5071 @item sample_rate, s
5072 Set sample rate, default is 44100.
5075 Set length of FIR filter, default is 22051.
5078 Set number of samples per each frame.
5081 Set window function to be used when generating FIR coefficients.
5086 Generate an audio signal made of a sine wave with amplitude 1/8.
5088 The audio signal is bit-exact.
5090 The filter accepts the following options:
5095 Set the carrier frequency. Default is 440 Hz.
5097 @item beep_factor, b
5098 Enable a periodic beep every second with frequency @var{beep_factor} times
5099 the carrier frequency. Default is 0, meaning the beep is disabled.
5101 @item sample_rate, r
5102 Specify the sample rate, default is 44100.
5105 Specify the duration of the generated audio stream.
5107 @item samples_per_frame
5108 Set the number of samples per output frame.
5110 The expression can contain the following constants:
5114 The (sequential) number of the output audio frame, starting from 0.
5117 The PTS (Presentation TimeStamp) of the output audio frame,
5118 expressed in @var{TB} units.
5121 The PTS of the output audio frame, expressed in seconds.
5124 The timebase of the output audio frames.
5127 Default is @code{1024}.
5130 @subsection Examples
5135 Generate a simple 440 Hz sine wave:
5141 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5145 sine=frequency=220:beep_factor=4:duration=5
5149 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5152 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5156 @c man end AUDIO SOURCES
5158 @chapter Audio Sinks
5159 @c man begin AUDIO SINKS
5161 Below is a description of the currently available audio sinks.
5163 @section abuffersink
5165 Buffer audio frames, and make them available to the end of filter chain.
5167 This sink is mainly intended for programmatic use, in particular
5168 through the interface defined in @file{libavfilter/buffersink.h}
5169 or the options system.
5171 It accepts a pointer to an AVABufferSinkContext structure, which
5172 defines the incoming buffers' formats, to be passed as the opaque
5173 parameter to @code{avfilter_init_filter} for initialization.
5176 Null audio sink; do absolutely nothing with the input audio. It is
5177 mainly useful as a template and for use in analysis / debugging
5180 @c man end AUDIO SINKS
5182 @chapter Video Filters
5183 @c man begin VIDEO FILTERS
5185 When you configure your FFmpeg build, you can disable any of the
5186 existing filters using @code{--disable-filters}.
5187 The configure output will show the video filters included in your
5190 Below is a description of the currently available video filters.
5192 @section alphaextract
5194 Extract the alpha component from the input as a grayscale video. This
5195 is especially useful with the @var{alphamerge} filter.
5199 Add or replace the alpha component of the primary input with the
5200 grayscale value of a second input. This is intended for use with
5201 @var{alphaextract} to allow the transmission or storage of frame
5202 sequences that have alpha in a format that doesn't support an alpha
5205 For example, to reconstruct full frames from a normal YUV-encoded video
5206 and a separate video created with @var{alphaextract}, you might use:
5208 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5211 Since this filter is designed for reconstruction, it operates on frame
5212 sequences without considering timestamps, and terminates when either
5213 input reaches end of stream. This will cause problems if your encoding
5214 pipeline drops frames. If you're trying to apply an image as an
5215 overlay to a video stream, consider the @var{overlay} filter instead.
5219 Amplify differences between current pixel and pixels of adjacent frames in
5220 same pixel location.
5222 This filter accepts the following options:
5226 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5227 For example radius of 3 will instruct filter to calculate average of 7 frames.
5230 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5233 Set threshold for difference amplification. Any differrence greater or equal to
5234 this value will not alter source pixel. Default is 10.
5235 Allowed range is from 0 to 65535.
5238 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5239 This option controls maximum possible value that will decrease source pixel value.
5242 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5243 This option controls maximum possible value that will increase source pixel value.
5246 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5251 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5252 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5253 Substation Alpha) subtitles files.
5255 This filter accepts the following option in addition to the common options from
5256 the @ref{subtitles} filter:
5260 Set the shaping engine
5262 Available values are:
5265 The default libass shaping engine, which is the best available.
5267 Fast, font-agnostic shaper that can do only substitutions
5269 Slower shaper using OpenType for substitutions and positioning
5272 The default is @code{auto}.
5276 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5278 The filter accepts the following options:
5282 Set threshold A for 1st plane. Default is 0.02.
5283 Valid range is 0 to 0.3.
5286 Set threshold B for 1st plane. Default is 0.04.
5287 Valid range is 0 to 5.
5290 Set threshold A for 2nd plane. Default is 0.02.
5291 Valid range is 0 to 0.3.
5294 Set threshold B for 2nd plane. Default is 0.04.
5295 Valid range is 0 to 5.
5298 Set threshold A for 3rd plane. Default is 0.02.
5299 Valid range is 0 to 0.3.
5302 Set threshold B for 3rd plane. Default is 0.04.
5303 Valid range is 0 to 5.
5305 Threshold A is designed to react on abrupt changes in the input signal and
5306 threshold B is designed to react on continuous changes in the input signal.
5309 Set number of frames filter will use for averaging. Default is 9. Must be odd
5310 number in range [5, 129].
5313 Set what planes of frame filter will use for averaging. Default is all.
5318 Apply average blur filter.
5320 The filter accepts the following options:
5324 Set horizontal kernel size.
5327 Set which planes to filter. By default all planes are filtered.
5330 Set vertical kernel size, if zero it will be same as @code{sizeX}.
5331 Default is @code{0}.
5336 Compute the bounding box for the non-black pixels in the input frame
5339 This filter computes the bounding box containing all the pixels with a
5340 luminance value greater than the minimum allowed value.
5341 The parameters describing the bounding box are printed on the filter
5344 The filter accepts the following option:
5348 Set the minimal luminance value. Default is @code{16}.
5351 @section bitplanenoise
5353 Show and measure bit plane noise.
5355 The filter accepts the following options:
5359 Set which plane to analyze. Default is @code{1}.
5362 Filter out noisy pixels from @code{bitplane} set above.
5363 Default is disabled.
5366 @section blackdetect
5368 Detect video intervals that are (almost) completely black. Can be
5369 useful to detect chapter transitions, commercials, or invalid
5370 recordings. Output lines contains the time for the start, end and
5371 duration of the detected black interval expressed in seconds.
5373 In order to display the output lines, you need to set the loglevel at
5374 least to the AV_LOG_INFO value.
5376 The filter accepts the following options:
5379 @item black_min_duration, d
5380 Set the minimum detected black duration expressed in seconds. It must
5381 be a non-negative floating point number.
5383 Default value is 2.0.
5385 @item picture_black_ratio_th, pic_th
5386 Set the threshold for considering a picture "black".
5387 Express the minimum value for the ratio:
5389 @var{nb_black_pixels} / @var{nb_pixels}
5392 for which a picture is considered black.
5393 Default value is 0.98.
5395 @item pixel_black_th, pix_th
5396 Set the threshold for considering a pixel "black".
5398 The threshold expresses the maximum pixel luminance value for which a
5399 pixel is considered "black". The provided value is scaled according to
5400 the following equation:
5402 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5405 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5406 the input video format, the range is [0-255] for YUV full-range
5407 formats and [16-235] for YUV non full-range formats.
5409 Default value is 0.10.
5412 The following example sets the maximum pixel threshold to the minimum
5413 value, and detects only black intervals of 2 or more seconds:
5415 blackdetect=d=2:pix_th=0.00
5420 Detect frames that are (almost) completely black. Can be useful to
5421 detect chapter transitions or commercials. Output lines consist of
5422 the frame number of the detected frame, the percentage of blackness,
5423 the position in the file if known or -1 and the timestamp in seconds.
5425 In order to display the output lines, you need to set the loglevel at
5426 least to the AV_LOG_INFO value.
5428 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5429 The value represents the percentage of pixels in the picture that
5430 are below the threshold value.
5432 It accepts the following parameters:
5437 The percentage of the pixels that have to be below the threshold; it defaults to
5440 @item threshold, thresh
5441 The threshold below which a pixel value is considered black; it defaults to
5446 @section blend, tblend
5448 Blend two video frames into each other.
5450 The @code{blend} filter takes two input streams and outputs one
5451 stream, the first input is the "top" layer and second input is
5452 "bottom" layer. By default, the output terminates when the longest input terminates.
5454 The @code{tblend} (time blend) filter takes two consecutive frames
5455 from one single stream, and outputs the result obtained by blending
5456 the new frame on top of the old frame.
5458 A description of the accepted options follows.
5466 Set blend mode for specific pixel component or all pixel components in case
5467 of @var{all_mode}. Default value is @code{normal}.
5469 Available values for component modes are:
5511 Set blend opacity for specific pixel component or all pixel components in case
5512 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5519 Set blend expression for specific pixel component or all pixel components in case
5520 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5522 The expressions can use the following variables:
5526 The sequential number of the filtered frame, starting from @code{0}.
5530 the coordinates of the current sample
5534 the width and height of currently filtered plane
5538 Width and height scale depending on the currently filtered plane. It is the
5539 ratio between the corresponding luma plane number of pixels and the current
5540 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5541 @code{0.5,0.5} for chroma planes.
5544 Time of the current frame, expressed in seconds.
5547 Value of pixel component at current location for first video frame (top layer).
5550 Value of pixel component at current location for second video frame (bottom layer).
5554 The @code{blend} filter also supports the @ref{framesync} options.
5556 @subsection Examples
5560 Apply transition from bottom layer to top layer in first 10 seconds:
5562 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5566 Apply linear horizontal transition from top layer to bottom layer:
5568 blend=all_expr='A*(X/W)+B*(1-X/W)'
5572 Apply 1x1 checkerboard effect:
5574 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5578 Apply uncover left effect:
5580 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5584 Apply uncover down effect:
5586 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5590 Apply uncover up-left effect:
5592 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5596 Split diagonally video and shows top and bottom layer on each side:
5598 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5602 Display differences between the current and the previous frame:
5604 tblend=all_mode=grainextract
5610 Apply a boxblur algorithm to the input video.
5612 It accepts the following parameters:
5616 @item luma_radius, lr
5617 @item luma_power, lp
5618 @item chroma_radius, cr
5619 @item chroma_power, cp
5620 @item alpha_radius, ar
5621 @item alpha_power, ap
5625 A description of the accepted options follows.
5628 @item luma_radius, lr
5629 @item chroma_radius, cr
5630 @item alpha_radius, ar
5631 Set an expression for the box radius in pixels used for blurring the
5632 corresponding input plane.
5634 The radius value must be a non-negative number, and must not be
5635 greater than the value of the expression @code{min(w,h)/2} for the
5636 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5639 Default value for @option{luma_radius} is "2". If not specified,
5640 @option{chroma_radius} and @option{alpha_radius} default to the
5641 corresponding value set for @option{luma_radius}.
5643 The expressions can contain the following constants:
5647 The input width and height in pixels.
5651 The input chroma image width and height in pixels.
5655 The horizontal and vertical chroma subsample values. For example, for the
5656 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5659 @item luma_power, lp
5660 @item chroma_power, cp
5661 @item alpha_power, ap
5662 Specify how many times the boxblur filter is applied to the
5663 corresponding plane.
5665 Default value for @option{luma_power} is 2. If not specified,
5666 @option{chroma_power} and @option{alpha_power} default to the
5667 corresponding value set for @option{luma_power}.
5669 A value of 0 will disable the effect.
5672 @subsection Examples
5676 Apply a boxblur filter with the luma, chroma, and alpha radii
5679 boxblur=luma_radius=2:luma_power=1
5684 Set the luma radius to 2, and alpha and chroma radius to 0:
5686 boxblur=2:1:cr=0:ar=0
5690 Set the luma and chroma radii to a fraction of the video dimension:
5692 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5698 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5699 Deinterlacing Filter").
5701 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5702 interpolation algorithms.
5703 It accepts the following parameters:
5707 The interlacing mode to adopt. It accepts one of the following values:
5711 Output one frame for each frame.
5713 Output one frame for each field.
5716 The default value is @code{send_field}.
5719 The picture field parity assumed for the input interlaced video. It accepts one
5720 of the following values:
5724 Assume the top field is first.
5726 Assume the bottom field is first.
5728 Enable automatic detection of field parity.
5731 The default value is @code{auto}.
5732 If the interlacing is unknown or the decoder does not export this information,
5733 top field first will be assumed.
5736 Specify which frames to deinterlace. Accept one of the following
5741 Deinterlace all frames.
5743 Only deinterlace frames marked as interlaced.
5746 The default value is @code{all}.
5750 YUV colorspace color/chroma keying.
5752 The filter accepts the following options:
5756 The color which will be replaced with transparency.
5759 Similarity percentage with the key color.
5761 0.01 matches only the exact key color, while 1.0 matches everything.
5766 0.0 makes pixels either fully transparent, or not transparent at all.
5768 Higher values result in semi-transparent pixels, with a higher transparency
5769 the more similar the pixels color is to the key color.
5772 Signals that the color passed is already in YUV instead of RGB.
5774 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5775 This can be used to pass exact YUV values as hexadecimal numbers.
5778 @subsection Examples
5782 Make every green pixel in the input image transparent:
5784 ffmpeg -i input.png -vf chromakey=green out.png
5788 Overlay a greenscreen-video on top of a static black background.
5790 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
5796 Display CIE color diagram with pixels overlaid onto it.
5798 The filter accepts the following options:
5813 @item uhdtv, rec2020
5826 Set what gamuts to draw.
5828 See @code{system} option for available values.
5831 Set ciescope size, by default set to 512.
5834 Set intensity used to map input pixel values to CIE diagram.
5837 Set contrast used to draw tongue colors that are out of active color system gamut.
5840 Correct gamma displayed on scope, by default enabled.
5843 Show white point on CIE diagram, by default disabled.
5846 Set input gamma. Used only with XYZ input color space.
5851 Visualize information exported by some codecs.
5853 Some codecs can export information through frames using side-data or other
5854 means. For example, some MPEG based codecs export motion vectors through the
5855 @var{export_mvs} flag in the codec @option{flags2} option.
5857 The filter accepts the following option:
5861 Set motion vectors to visualize.
5863 Available flags for @var{mv} are:
5867 forward predicted MVs of P-frames
5869 forward predicted MVs of B-frames
5871 backward predicted MVs of B-frames
5875 Display quantization parameters using the chroma planes.
5878 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5880 Available flags for @var{mv_type} are:
5884 forward predicted MVs
5886 backward predicted MVs
5889 @item frame_type, ft
5890 Set frame type to visualize motion vectors of.
5892 Available flags for @var{frame_type} are:
5896 intra-coded frames (I-frames)
5898 predicted frames (P-frames)
5900 bi-directionally predicted frames (B-frames)
5904 @subsection Examples
5908 Visualize forward predicted MVs of all frames using @command{ffplay}:
5910 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5914 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5916 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5920 @section colorbalance
5921 Modify intensity of primary colors (red, green and blue) of input frames.
5923 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5924 regions for the red-cyan, green-magenta or blue-yellow balance.
5926 A positive adjustment value shifts the balance towards the primary color, a negative
5927 value towards the complementary color.
5929 The filter accepts the following options:
5935 Adjust red, green and blue shadows (darkest pixels).
5940 Adjust red, green and blue midtones (medium pixels).
5945 Adjust red, green and blue highlights (brightest pixels).
5947 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5950 @subsection Examples
5954 Add red color cast to shadows:
5961 RGB colorspace color keying.
5963 The filter accepts the following options:
5967 The color which will be replaced with transparency.
5970 Similarity percentage with the key color.
5972 0.01 matches only the exact key color, while 1.0 matches everything.
5977 0.0 makes pixels either fully transparent, or not transparent at all.
5979 Higher values result in semi-transparent pixels, with a higher transparency
5980 the more similar the pixels color is to the key color.
5983 @subsection Examples
5987 Make every green pixel in the input image transparent:
5989 ffmpeg -i input.png -vf colorkey=green out.png
5993 Overlay a greenscreen-video on top of a static background image.
5995 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
5999 @section colorlevels
6001 Adjust video input frames using levels.
6003 The filter accepts the following options:
6010 Adjust red, green, blue and alpha input black point.
6011 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6017 Adjust red, green, blue and alpha input white point.
6018 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6020 Input levels are used to lighten highlights (bright tones), darken shadows
6021 (dark tones), change the balance of bright and dark tones.
6027 Adjust red, green, blue and alpha output black point.
6028 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6034 Adjust red, green, blue and alpha output white point.
6035 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6037 Output levels allows manual selection of a constrained output level range.
6040 @subsection Examples
6044 Make video output darker:
6046 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6052 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6056 Make video output lighter:
6058 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6062 Increase brightness:
6064 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6068 @section colorchannelmixer
6070 Adjust video input frames by re-mixing color channels.
6072 This filter modifies a color channel by adding the values associated to
6073 the other channels of the same pixels. For example if the value to
6074 modify is red, the output value will be:
6076 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6079 The filter accepts the following options:
6086 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6087 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6093 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6094 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6100 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6101 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6107 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6108 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6110 Allowed ranges for options are @code{[-2.0, 2.0]}.
6113 @subsection Examples
6117 Convert source to grayscale:
6119 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6122 Simulate sepia tones:
6124 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6128 @section colormatrix
6130 Convert color matrix.
6132 The filter accepts the following options:
6137 Specify the source and destination color matrix. Both values must be
6140 The accepted values are:
6168 For example to convert from BT.601 to SMPTE-240M, use the command:
6170 colormatrix=bt601:smpte240m
6175 Convert colorspace, transfer characteristics or color primaries.
6176 Input video needs to have an even size.
6178 The filter accepts the following options:
6183 Specify all color properties at once.
6185 The accepted values are:
6215 Specify output colorspace.
6217 The accepted values are:
6226 BT.470BG or BT.601-6 625
6229 SMPTE-170M or BT.601-6 525
6238 BT.2020 with non-constant luminance
6244 Specify output transfer characteristics.
6246 The accepted values are:
6258 Constant gamma of 2.2
6261 Constant gamma of 2.8
6264 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6282 BT.2020 for 10-bits content
6285 BT.2020 for 12-bits content
6291 Specify output color primaries.
6293 The accepted values are:
6302 BT.470BG or BT.601-6 625
6305 SMPTE-170M or BT.601-6 525
6329 Specify output color range.
6331 The accepted values are:
6334 TV (restricted) range
6337 MPEG (restricted) range
6348 Specify output color format.
6350 The accepted values are:
6353 YUV 4:2:0 planar 8-bits
6356 YUV 4:2:0 planar 10-bits
6359 YUV 4:2:0 planar 12-bits
6362 YUV 4:2:2 planar 8-bits
6365 YUV 4:2:2 planar 10-bits
6368 YUV 4:2:2 planar 12-bits
6371 YUV 4:4:4 planar 8-bits
6374 YUV 4:4:4 planar 10-bits
6377 YUV 4:4:4 planar 12-bits
6382 Do a fast conversion, which skips gamma/primary correction. This will take
6383 significantly less CPU, but will be mathematically incorrect. To get output
6384 compatible with that produced by the colormatrix filter, use fast=1.
6387 Specify dithering mode.
6389 The accepted values are:
6395 Floyd-Steinberg dithering
6399 Whitepoint adaptation mode.
6401 The accepted values are:
6404 Bradford whitepoint adaptation
6407 von Kries whitepoint adaptation
6410 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6414 Override all input properties at once. Same accepted values as @ref{all}.
6417 Override input colorspace. Same accepted values as @ref{space}.
6420 Override input color primaries. Same accepted values as @ref{primaries}.
6423 Override input transfer characteristics. Same accepted values as @ref{trc}.
6426 Override input color range. Same accepted values as @ref{range}.
6430 The filter converts the transfer characteristics, color space and color
6431 primaries to the specified user values. The output value, if not specified,
6432 is set to a default value based on the "all" property. If that property is
6433 also not specified, the filter will log an error. The output color range and
6434 format default to the same value as the input color range and format. The
6435 input transfer characteristics, color space, color primaries and color range
6436 should be set on the input data. If any of these are missing, the filter will
6437 log an error and no conversion will take place.
6439 For example to convert the input to SMPTE-240M, use the command:
6441 colorspace=smpte240m
6444 @section convolution
6446 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
6448 The filter accepts the following options:
6455 Set matrix for each plane.
6456 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
6457 and from 1 to 49 odd number of signed integers in @var{row} mode.
6463 Set multiplier for calculated value for each plane.
6464 If unset or 0, it will be sum of all matrix elements.
6470 Set bias for each plane. This value is added to the result of the multiplication.
6471 Useful for making the overall image brighter or darker. Default is 0.0.
6477 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
6478 Default is @var{square}.
6481 @subsection Examples
6487 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"
6493 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"
6499 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"
6505 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"
6509 Apply laplacian edge detector which includes diagonals:
6511 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"
6517 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"
6523 Apply 2D convolution of video stream in frequency domain using second stream
6526 The filter accepts the following options:
6530 Set which planes to process.
6533 Set which impulse video frames will be processed, can be @var{first}
6534 or @var{all}. Default is @var{all}.
6537 The @code{convolve} filter also supports the @ref{framesync} options.
6541 Copy the input video source unchanged to the output. This is mainly useful for
6546 Video filtering on GPU using Apple's CoreImage API on OSX.
6548 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6549 processed by video hardware. However, software-based OpenGL implementations
6550 exist which means there is no guarantee for hardware processing. It depends on
6553 There are many filters and image generators provided by Apple that come with a
6554 large variety of options. The filter has to be referenced by its name along
6557 The coreimage filter accepts the following options:
6560 List all available filters and generators along with all their respective
6561 options as well as possible minimum and maximum values along with the default
6568 Specify all filters by their respective name and options.
6569 Use @var{list_filters} to determine all valid filter names and options.
6570 Numerical options are specified by a float value and are automatically clamped
6571 to their respective value range. Vector and color options have to be specified
6572 by a list of space separated float values. Character escaping has to be done.
6573 A special option name @code{default} is available to use default options for a
6576 It is required to specify either @code{default} or at least one of the filter options.
6577 All omitted options are used with their default values.
6578 The syntax of the filter string is as follows:
6580 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6584 Specify a rectangle where the output of the filter chain is copied into the
6585 input image. It is given by a list of space separated float values:
6587 output_rect=x\ y\ width\ height
6589 If not given, the output rectangle equals the dimensions of the input image.
6590 The output rectangle is automatically cropped at the borders of the input
6591 image. Negative values are valid for each component.
6593 output_rect=25\ 25\ 100\ 100
6597 Several filters can be chained for successive processing without GPU-HOST
6598 transfers allowing for fast processing of complex filter chains.
6599 Currently, only filters with zero (generators) or exactly one (filters) input
6600 image and one output image are supported. Also, transition filters are not yet
6603 Some filters generate output images with additional padding depending on the
6604 respective filter kernel. The padding is automatically removed to ensure the
6605 filter output has the same size as the input image.
6607 For image generators, the size of the output image is determined by the
6608 previous output image of the filter chain or the input image of the whole
6609 filterchain, respectively. The generators do not use the pixel information of
6610 this image to generate their output. However, the generated output is
6611 blended onto this image, resulting in partial or complete coverage of the
6614 The @ref{coreimagesrc} video source can be used for generating input images
6615 which are directly fed into the filter chain. By using it, providing input
6616 images by another video source or an input video is not required.
6618 @subsection Examples
6623 List all filters available:
6625 coreimage=list_filters=true
6629 Use the CIBoxBlur filter with default options to blur an image:
6631 coreimage=filter=CIBoxBlur@@default
6635 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6636 its center at 100x100 and a radius of 50 pixels:
6638 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6642 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6643 given as complete and escaped command-line for Apple's standard bash shell:
6645 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6651 Crop the input video to given dimensions.
6653 It accepts the following parameters:
6657 The width of the output video. It defaults to @code{iw}.
6658 This expression is evaluated only once during the filter
6659 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6662 The height of the output video. It defaults to @code{ih}.
6663 This expression is evaluated only once during the filter
6664 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6667 The horizontal position, in the input video, of the left edge of the output
6668 video. It defaults to @code{(in_w-out_w)/2}.
6669 This expression is evaluated per-frame.
6672 The vertical position, in the input video, of the top edge of the output video.
6673 It defaults to @code{(in_h-out_h)/2}.
6674 This expression is evaluated per-frame.
6677 If set to 1 will force the output display aspect ratio
6678 to be the same of the input, by changing the output sample aspect
6679 ratio. It defaults to 0.
6682 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6683 width/height/x/y as specified and will not be rounded to nearest smaller value.
6687 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6688 expressions containing the following constants:
6693 The computed values for @var{x} and @var{y}. They are evaluated for
6698 The input width and height.
6702 These are the same as @var{in_w} and @var{in_h}.
6706 The output (cropped) width and height.
6710 These are the same as @var{out_w} and @var{out_h}.
6713 same as @var{iw} / @var{ih}
6716 input sample aspect ratio
6719 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6723 horizontal and vertical chroma subsample values. For example for the
6724 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6727 The number of the input frame, starting from 0.
6730 the position in the file of the input frame, NAN if unknown
6733 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6737 The expression for @var{out_w} may depend on the value of @var{out_h},
6738 and the expression for @var{out_h} may depend on @var{out_w}, but they
6739 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6740 evaluated after @var{out_w} and @var{out_h}.
6742 The @var{x} and @var{y} parameters specify the expressions for the
6743 position of the top-left corner of the output (non-cropped) area. They
6744 are evaluated for each frame. If the evaluated value is not valid, it
6745 is approximated to the nearest valid value.
6747 The expression for @var{x} may depend on @var{y}, and the expression
6748 for @var{y} may depend on @var{x}.
6750 @subsection Examples
6754 Crop area with size 100x100 at position (12,34).
6759 Using named options, the example above becomes:
6761 crop=w=100:h=100:x=12:y=34
6765 Crop the central input area with size 100x100:
6771 Crop the central input area with size 2/3 of the input video:
6773 crop=2/3*in_w:2/3*in_h
6777 Crop the input video central square:
6784 Delimit the rectangle with the top-left corner placed at position
6785 100:100 and the right-bottom corner corresponding to the right-bottom
6786 corner of the input image.
6788 crop=in_w-100:in_h-100:100:100
6792 Crop 10 pixels from the left and right borders, and 20 pixels from
6793 the top and bottom borders
6795 crop=in_w-2*10:in_h-2*20
6799 Keep only the bottom right quarter of the input image:
6801 crop=in_w/2:in_h/2:in_w/2:in_h/2
6805 Crop height for getting Greek harmony:
6807 crop=in_w:1/PHI*in_w
6811 Apply trembling effect:
6813 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)
6817 Apply erratic camera effect depending on timestamp:
6819 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)"
6823 Set x depending on the value of y:
6825 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6829 @subsection Commands
6831 This filter supports the following commands:
6837 Set width/height of the output video and the horizontal/vertical position
6839 The command accepts the same syntax of the corresponding option.
6841 If the specified expression is not valid, it is kept at its current
6847 Auto-detect the crop size.
6849 It calculates the necessary cropping parameters and prints the
6850 recommended parameters via the logging system. The detected dimensions
6851 correspond to the non-black area of the input video.
6853 It accepts the following parameters:
6858 Set higher black value threshold, which can be optionally specified
6859 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6860 value greater to the set value is considered non-black. It defaults to 24.
6861 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6862 on the bitdepth of the pixel format.
6865 The value which the width/height should be divisible by. It defaults to
6866 16. The offset is automatically adjusted to center the video. Use 2 to
6867 get only even dimensions (needed for 4:2:2 video). 16 is best when
6868 encoding to most video codecs.
6870 @item reset_count, reset
6871 Set the counter that determines after how many frames cropdetect will
6872 reset the previously detected largest video area and start over to
6873 detect the current optimal crop area. Default value is 0.
6875 This can be useful when channel logos distort the video area. 0
6876 indicates 'never reset', and returns the largest area encountered during
6883 Apply color adjustments using curves.
6885 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6886 component (red, green and blue) has its values defined by @var{N} key points
6887 tied from each other using a smooth curve. The x-axis represents the pixel
6888 values from the input frame, and the y-axis the new pixel values to be set for
6891 By default, a component curve is defined by the two points @var{(0;0)} and
6892 @var{(1;1)}. This creates a straight line where each original pixel value is
6893 "adjusted" to its own value, which means no change to the image.
6895 The filter allows you to redefine these two points and add some more. A new
6896 curve (using a natural cubic spline interpolation) will be define to pass
6897 smoothly through all these new coordinates. The new defined points needs to be
6898 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6899 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6900 the vector spaces, the values will be clipped accordingly.
6902 The filter accepts the following options:
6906 Select one of the available color presets. This option can be used in addition
6907 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6908 options takes priority on the preset values.
6909 Available presets are:
6912 @item color_negative
6915 @item increase_contrast
6917 @item linear_contrast
6918 @item medium_contrast
6920 @item strong_contrast
6923 Default is @code{none}.
6925 Set the master key points. These points will define a second pass mapping. It
6926 is sometimes called a "luminance" or "value" mapping. It can be used with
6927 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6928 post-processing LUT.
6930 Set the key points for the red component.
6932 Set the key points for the green component.
6934 Set the key points for the blue component.
6936 Set the key points for all components (not including master).
6937 Can be used in addition to the other key points component
6938 options. In this case, the unset component(s) will fallback on this
6939 @option{all} setting.
6941 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6943 Save Gnuplot script of the curves in specified file.
6946 To avoid some filtergraph syntax conflicts, each key points list need to be
6947 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6949 @subsection Examples
6953 Increase slightly the middle level of blue:
6955 curves=blue='0/0 0.5/0.58 1/1'
6961 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'
6963 Here we obtain the following coordinates for each components:
6966 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6968 @code{(0;0) (0.50;0.48) (1;1)}
6970 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6974 The previous example can also be achieved with the associated built-in preset:
6976 curves=preset=vintage
6986 Use a Photoshop preset and redefine the points of the green component:
6988 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6992 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6993 and @command{gnuplot}:
6995 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6996 gnuplot -p /tmp/curves.plt
7002 Video data analysis filter.
7004 This filter shows hexadecimal pixel values of part of video.
7006 The filter accepts the following options:
7010 Set output video size.
7013 Set x offset from where to pick pixels.
7016 Set y offset from where to pick pixels.
7019 Set scope mode, can be one of the following:
7022 Draw hexadecimal pixel values with white color on black background.
7025 Draw hexadecimal pixel values with input video pixel color on black
7029 Draw hexadecimal pixel values on color background picked from input video,
7030 the text color is picked in such way so its always visible.
7034 Draw rows and columns numbers on left and top of video.
7037 Set background opacity.
7042 Denoise frames using 2D DCT (frequency domain filtering).
7044 This filter is not designed for real time.
7046 The filter accepts the following options:
7050 Set the noise sigma constant.
7052 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7053 coefficient (absolute value) below this threshold with be dropped.
7055 If you need a more advanced filtering, see @option{expr}.
7057 Default is @code{0}.
7060 Set number overlapping pixels for each block. Since the filter can be slow, you
7061 may want to reduce this value, at the cost of a less effective filter and the
7062 risk of various artefacts.
7064 If the overlapping value doesn't permit processing the whole input width or
7065 height, a warning will be displayed and according borders won't be denoised.
7067 Default value is @var{blocksize}-1, which is the best possible setting.
7070 Set the coefficient factor expression.
7072 For each coefficient of a DCT block, this expression will be evaluated as a
7073 multiplier value for the coefficient.
7075 If this is option is set, the @option{sigma} option will be ignored.
7077 The absolute value of the coefficient can be accessed through the @var{c}
7081 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7082 @var{blocksize}, which is the width and height of the processed blocks.
7084 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7085 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7086 on the speed processing. Also, a larger block size does not necessarily means a
7090 @subsection Examples
7092 Apply a denoise with a @option{sigma} of @code{4.5}:
7097 The same operation can be achieved using the expression system:
7099 dctdnoiz=e='gte(c, 4.5*3)'
7102 Violent denoise using a block size of @code{16x16}:
7109 Remove banding artifacts from input video.
7110 It works by replacing banded pixels with average value of referenced pixels.
7112 The filter accepts the following options:
7119 Set banding detection threshold for each plane. Default is 0.02.
7120 Valid range is 0.00003 to 0.5.
7121 If difference between current pixel and reference pixel is less than threshold,
7122 it will be considered as banded.
7125 Banding detection range in pixels. Default is 16. If positive, random number
7126 in range 0 to set value will be used. If negative, exact absolute value
7128 The range defines square of four pixels around current pixel.
7131 Set direction in radians from which four pixel will be compared. If positive,
7132 random direction from 0 to set direction will be picked. If negative, exact of
7133 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7134 will pick only pixels on same row and -PI/2 will pick only pixels on same
7138 If enabled, current pixel is compared with average value of all four
7139 surrounding pixels. The default is enabled. If disabled current pixel is
7140 compared with all four surrounding pixels. The pixel is considered banded
7141 if only all four differences with surrounding pixels are less than threshold.
7144 If enabled, current pixel is changed if and only if all pixel components are banded,
7145 e.g. banding detection threshold is triggered for all color components.
7146 The default is disabled.
7151 Remove blocking artifacts from input video.
7153 The filter accepts the following options:
7157 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7158 This controls what kind of deblocking is applied.
7161 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7167 Set blocking detection thresholds. Allowed range is 0 to 1.
7168 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7169 Using higher threshold gives more deblocking strength.
7170 Setting @var{alpha} controls threshold detection at exact edge of block.
7171 Remaining options controls threshold detection near the edge. Each one for
7172 below/above or left/right. Setting any of those to @var{0} disables
7176 Set planes to filter. Default is to filter all available planes.
7179 @subsection Examples
7183 Deblock using weak filter and block size of 4 pixels.
7185 deblock=filter=weak:block=4
7189 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7190 deblocking more edges.
7192 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7196 Similar as above, but filter only first plane.
7198 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7202 Similar as above, but filter only second and third plane.
7204 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7211 Drop duplicated frames at regular intervals.
7213 The filter accepts the following options:
7217 Set the number of frames from which one will be dropped. Setting this to
7218 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7219 Default is @code{5}.
7222 Set the threshold for duplicate detection. If the difference metric for a frame
7223 is less than or equal to this value, then it is declared as duplicate. Default
7227 Set scene change threshold. Default is @code{15}.
7231 Set the size of the x and y-axis blocks used during metric calculations.
7232 Larger blocks give better noise suppression, but also give worse detection of
7233 small movements. Must be a power of two. Default is @code{32}.
7236 Mark main input as a pre-processed input and activate clean source input
7237 stream. This allows the input to be pre-processed with various filters to help
7238 the metrics calculation while keeping the frame selection lossless. When set to
7239 @code{1}, the first stream is for the pre-processed input, and the second
7240 stream is the clean source from where the kept frames are chosen. Default is
7244 Set whether or not chroma is considered in the metric calculations. Default is
7250 Apply 2D deconvolution of video stream in frequency domain using second stream
7253 The filter accepts the following options:
7257 Set which planes to process.
7260 Set which impulse video frames will be processed, can be @var{first}
7261 or @var{all}. Default is @var{all}.
7264 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7265 and height are not same and not power of 2 or if stream prior to convolving
7269 The @code{deconvolve} filter also supports the @ref{framesync} options.
7273 Apply deflate effect to the video.
7275 This filter replaces the pixel by the local(3x3) average by taking into account
7276 only values lower than the pixel.
7278 It accepts the following options:
7285 Limit the maximum change for each plane, default is 65535.
7286 If 0, plane will remain unchanged.
7291 Remove temporal frame luminance variations.
7293 It accepts the following options:
7297 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7300 Set averaging mode to smooth temporal luminance variations.
7302 Available values are:
7327 Do not actually modify frame. Useful when one only wants metadata.
7332 Remove judder produced by partially interlaced telecined content.
7334 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7335 source was partially telecined content then the output of @code{pullup,dejudder}
7336 will have a variable frame rate. May change the recorded frame rate of the
7337 container. Aside from that change, this filter will not affect constant frame
7340 The option available in this filter is:
7344 Specify the length of the window over which the judder repeats.
7346 Accepts any integer greater than 1. Useful values are:
7350 If the original was telecined from 24 to 30 fps (Film to NTSC).
7353 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7356 If a mixture of the two.
7359 The default is @samp{4}.
7364 Suppress a TV station logo by a simple interpolation of the surrounding
7365 pixels. Just set a rectangle covering the logo and watch it disappear
7366 (and sometimes something even uglier appear - your mileage may vary).
7368 It accepts the following parameters:
7373 Specify the top left corner coordinates of the logo. They must be
7378 Specify the width and height of the logo to clear. They must be
7382 Specify the thickness of the fuzzy edge of the rectangle (added to
7383 @var{w} and @var{h}). The default value is 1. This option is
7384 deprecated, setting higher values should no longer be necessary and
7388 When set to 1, a green rectangle is drawn on the screen to simplify
7389 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7390 The default value is 0.
7392 The rectangle is drawn on the outermost pixels which will be (partly)
7393 replaced with interpolated values. The values of the next pixels
7394 immediately outside this rectangle in each direction will be used to
7395 compute the interpolated pixel values inside the rectangle.
7399 @subsection Examples
7403 Set a rectangle covering the area with top left corner coordinates 0,0
7404 and size 100x77, and a band of size 10:
7406 delogo=x=0:y=0:w=100:h=77:band=10
7413 Attempt to fix small changes in horizontal and/or vertical shift. This
7414 filter helps remove camera shake from hand-holding a camera, bumping a
7415 tripod, moving on a vehicle, etc.
7417 The filter accepts the following options:
7425 Specify a rectangular area where to limit the search for motion
7427 If desired the search for motion vectors can be limited to a
7428 rectangular area of the frame defined by its top left corner, width
7429 and height. These parameters have the same meaning as the drawbox
7430 filter which can be used to visualise the position of the bounding
7433 This is useful when simultaneous movement of subjects within the frame
7434 might be confused for camera motion by the motion vector search.
7436 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7437 then the full frame is used. This allows later options to be set
7438 without specifying the bounding box for the motion vector search.
7440 Default - search the whole frame.
7444 Specify the maximum extent of movement in x and y directions in the
7445 range 0-64 pixels. Default 16.
7448 Specify how to generate pixels to fill blanks at the edge of the
7449 frame. Available values are:
7452 Fill zeroes at blank locations
7454 Original image at blank locations
7456 Extruded edge value at blank locations
7458 Mirrored edge at blank locations
7460 Default value is @samp{mirror}.
7463 Specify the blocksize to use for motion search. Range 4-128 pixels,
7467 Specify the contrast threshold for blocks. Only blocks with more than
7468 the specified contrast (difference between darkest and lightest
7469 pixels) will be considered. Range 1-255, default 125.
7472 Specify the search strategy. Available values are:
7475 Set exhaustive search
7477 Set less exhaustive search.
7479 Default value is @samp{exhaustive}.
7482 If set then a detailed log of the motion search is written to the
7489 Remove unwanted contamination of foreground colors, caused by reflected color of
7490 greenscreen or bluescreen.
7492 This filter accepts the following options:
7496 Set what type of despill to use.
7499 Set how spillmap will be generated.
7502 Set how much to get rid of still remaining spill.
7505 Controls amount of red in spill area.
7508 Controls amount of green in spill area.
7509 Should be -1 for greenscreen.
7512 Controls amount of blue in spill area.
7513 Should be -1 for bluescreen.
7516 Controls brightness of spill area, preserving colors.
7519 Modify alpha from generated spillmap.
7524 Apply an exact inverse of the telecine operation. It requires a predefined
7525 pattern specified using the pattern option which must be the same as that passed
7526 to the telecine filter.
7528 This filter accepts the following options:
7537 The default value is @code{top}.
7541 A string of numbers representing the pulldown pattern you wish to apply.
7542 The default value is @code{23}.
7545 A number representing position of the first frame with respect to the telecine
7546 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7551 Apply dilation effect to the video.
7553 This filter replaces the pixel by the local(3x3) maximum.
7555 It accepts the following options:
7562 Limit the maximum change for each plane, default is 65535.
7563 If 0, plane will remain unchanged.
7566 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7569 Flags to local 3x3 coordinates maps like this:
7578 Displace pixels as indicated by second and third input stream.
7580 It takes three input streams and outputs one stream, the first input is the
7581 source, and second and third input are displacement maps.
7583 The second input specifies how much to displace pixels along the
7584 x-axis, while the third input specifies how much to displace pixels
7586 If one of displacement map streams terminates, last frame from that
7587 displacement map will be used.
7589 Note that once generated, displacements maps can be reused over and over again.
7591 A description of the accepted options follows.
7595 Set displace behavior for pixels that are out of range.
7597 Available values are:
7600 Missing pixels are replaced by black pixels.
7603 Adjacent pixels will spread out to replace missing pixels.
7606 Out of range pixels are wrapped so they point to pixels of other side.
7609 Out of range pixels will be replaced with mirrored pixels.
7611 Default is @samp{smear}.
7615 @subsection Examples
7619 Add ripple effect to rgb input of video size hd720:
7621 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
7625 Add wave effect to rgb input of video size hd720:
7627 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
7633 Draw a colored box on the input image.
7635 It accepts the following parameters:
7640 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7644 The expressions which specify the width and height of the box; if 0 they are interpreted as
7645 the input width and height. It defaults to 0.
7648 Specify the color of the box to write. For the general syntax of this option,
7649 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7650 value @code{invert} is used, the box edge color is the same as the
7651 video with inverted luma.
7654 The expression which sets the thickness of the box edge.
7655 A value of @code{fill} will create a filled box. Default value is @code{3}.
7657 See below for the list of accepted constants.
7660 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7661 will overwrite the video's color and alpha pixels.
7662 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7665 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7666 following constants:
7670 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7674 horizontal and vertical chroma subsample values. For example for the
7675 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7679 The input width and height.
7682 The input sample aspect ratio.
7686 The x and y offset coordinates where the box is drawn.
7690 The width and height of the drawn box.
7693 The thickness of the drawn box.
7695 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7696 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7700 @subsection Examples
7704 Draw a black box around the edge of the input image:
7710 Draw a box with color red and an opacity of 50%:
7712 drawbox=10:20:200:60:red@@0.5
7715 The previous example can be specified as:
7717 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7721 Fill the box with pink color:
7723 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7727 Draw a 2-pixel red 2.40:1 mask:
7729 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
7735 Draw a grid on the input image.
7737 It accepts the following parameters:
7742 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7746 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7747 input width and height, respectively, minus @code{thickness}, so image gets
7748 framed. Default to 0.
7751 Specify the color of the grid. For the general syntax of this option,
7752 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7753 value @code{invert} is used, the grid color is the same as the
7754 video with inverted luma.
7757 The expression which sets the thickness of the grid line. Default value is @code{1}.
7759 See below for the list of accepted constants.
7762 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7763 will overwrite the video's color and alpha pixels.
7764 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7767 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7768 following constants:
7772 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7776 horizontal and vertical chroma subsample values. For example for the
7777 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7781 The input grid cell width and height.
7784 The input sample aspect ratio.
7788 The x and y coordinates of some point of grid intersection (meant to configure offset).
7792 The width and height of the drawn cell.
7795 The thickness of the drawn cell.
7797 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7798 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7802 @subsection Examples
7806 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7808 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7812 Draw a white 3x3 grid with an opacity of 50%:
7814 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7821 Draw a text string or text from a specified file on top of a video, using the
7822 libfreetype library.
7824 To enable compilation of this filter, you need to configure FFmpeg with
7825 @code{--enable-libfreetype}.
7826 To enable default font fallback and the @var{font} option you need to
7827 configure FFmpeg with @code{--enable-libfontconfig}.
7828 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7829 @code{--enable-libfribidi}.
7833 It accepts the following parameters:
7838 Used to draw a box around text using the background color.
7839 The value must be either 1 (enable) or 0 (disable).
7840 The default value of @var{box} is 0.
7843 Set the width of the border to be drawn around the box using @var{boxcolor}.
7844 The default value of @var{boxborderw} is 0.
7847 The color to be used for drawing box around text. For the syntax of this
7848 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7850 The default value of @var{boxcolor} is "white".
7853 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7854 The default value of @var{line_spacing} is 0.
7857 Set the width of the border to be drawn around the text using @var{bordercolor}.
7858 The default value of @var{borderw} is 0.
7861 Set the color to be used for drawing border around text. For the syntax of this
7862 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7864 The default value of @var{bordercolor} is "black".
7867 Select how the @var{text} is expanded. Can be either @code{none},
7868 @code{strftime} (deprecated) or
7869 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7873 Set a start time for the count. Value is in microseconds. Only applied
7874 in the deprecated strftime expansion mode. To emulate in normal expansion
7875 mode use the @code{pts} function, supplying the start time (in seconds)
7876 as the second argument.
7879 If true, check and fix text coords to avoid clipping.
7882 The color to be used for drawing fonts. For the syntax of this option, check
7883 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7885 The default value of @var{fontcolor} is "black".
7887 @item fontcolor_expr
7888 String which is expanded the same way as @var{text} to obtain dynamic
7889 @var{fontcolor} value. By default this option has empty value and is not
7890 processed. When this option is set, it overrides @var{fontcolor} option.
7893 The font family to be used for drawing text. By default Sans.
7896 The font file to be used for drawing text. The path must be included.
7897 This parameter is mandatory if the fontconfig support is disabled.
7900 Draw the text applying alpha blending. The value can
7901 be a number between 0.0 and 1.0.
7902 The expression accepts the same variables @var{x, y} as well.
7903 The default value is 1.
7904 Please see @var{fontcolor_expr}.
7907 The font size to be used for drawing text.
7908 The default value of @var{fontsize} is 16.
7911 If set to 1, attempt to shape the text (for example, reverse the order of
7912 right-to-left text and join Arabic characters) before drawing it.
7913 Otherwise, just draw the text exactly as given.
7914 By default 1 (if supported).
7917 The flags to be used for loading the fonts.
7919 The flags map the corresponding flags supported by libfreetype, and are
7920 a combination of the following values:
7927 @item vertical_layout
7928 @item force_autohint
7931 @item ignore_global_advance_width
7933 @item ignore_transform
7939 Default value is "default".
7941 For more information consult the documentation for the FT_LOAD_*
7945 The color to be used for drawing a shadow behind the drawn text. For the
7946 syntax of this option, check the @ref{color syntax,,"Color" section in the
7947 ffmpeg-utils manual,ffmpeg-utils}.
7949 The default value of @var{shadowcolor} is "black".
7953 The x and y offsets for the text shadow position with respect to the
7954 position of the text. They can be either positive or negative
7955 values. The default value for both is "0".
7958 The starting frame number for the n/frame_num variable. The default value
7962 The size in number of spaces to use for rendering the tab.
7966 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7967 format. It can be used with or without text parameter. @var{timecode_rate}
7968 option must be specified.
7970 @item timecode_rate, rate, r
7971 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7972 integer. Minimum value is "1".
7973 Drop-frame timecode is supported for frame rates 30 & 60.
7976 If set to 1, the output of the timecode option will wrap around at 24 hours.
7977 Default is 0 (disabled).
7980 The text string to be drawn. The text must be a sequence of UTF-8
7982 This parameter is mandatory if no file is specified with the parameter
7986 A text file containing text to be drawn. The text must be a sequence
7987 of UTF-8 encoded characters.
7989 This parameter is mandatory if no text string is specified with the
7990 parameter @var{text}.
7992 If both @var{text} and @var{textfile} are specified, an error is thrown.
7995 If set to 1, the @var{textfile} will be reloaded before each frame.
7996 Be sure to update it atomically, or it may be read partially, or even fail.
8000 The expressions which specify the offsets where text will be drawn
8001 within the video frame. They are relative to the top/left border of the
8004 The default value of @var{x} and @var{y} is "0".
8006 See below for the list of accepted constants and functions.
8009 The parameters for @var{x} and @var{y} are expressions containing the
8010 following constants and functions:
8014 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8018 horizontal and vertical chroma subsample values. For example for the
8019 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8022 the height of each text line
8030 @item max_glyph_a, ascent
8031 the maximum distance from the baseline to the highest/upper grid
8032 coordinate used to place a glyph outline point, for all the rendered
8034 It is a positive value, due to the grid's orientation with the Y axis
8037 @item max_glyph_d, descent
8038 the maximum distance from the baseline to the lowest grid coordinate
8039 used to place a glyph outline point, for all the rendered glyphs.
8040 This is a negative value, due to the grid's orientation, with the Y axis
8044 maximum glyph height, that is the maximum height for all the glyphs
8045 contained in the rendered text, it is equivalent to @var{ascent} -
8049 maximum glyph width, that is the maximum width for all the glyphs
8050 contained in the rendered text
8053 the number of input frame, starting from 0
8055 @item rand(min, max)
8056 return a random number included between @var{min} and @var{max}
8059 The input sample aspect ratio.
8062 timestamp expressed in seconds, NAN if the input timestamp is unknown
8065 the height of the rendered text
8068 the width of the rendered text
8072 the x and y offset coordinates where the text is drawn.
8074 These parameters allow the @var{x} and @var{y} expressions to refer
8075 each other, so you can for example specify @code{y=x/dar}.
8078 @anchor{drawtext_expansion}
8079 @subsection Text expansion
8081 If @option{expansion} is set to @code{strftime},
8082 the filter recognizes strftime() sequences in the provided text and
8083 expands them accordingly. Check the documentation of strftime(). This
8084 feature is deprecated.
8086 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8088 If @option{expansion} is set to @code{normal} (which is the default),
8089 the following expansion mechanism is used.
8091 The backslash character @samp{\}, followed by any character, always expands to
8092 the second character.
8094 Sequences of the form @code{%@{...@}} are expanded. The text between the
8095 braces is a function name, possibly followed by arguments separated by ':'.
8096 If the arguments contain special characters or delimiters (':' or '@}'),
8097 they should be escaped.
8099 Note that they probably must also be escaped as the value for the
8100 @option{text} option in the filter argument string and as the filter
8101 argument in the filtergraph description, and possibly also for the shell,
8102 that makes up to four levels of escaping; using a text file avoids these
8105 The following functions are available:
8110 The expression evaluation result.
8112 It must take one argument specifying the expression to be evaluated,
8113 which accepts the same constants and functions as the @var{x} and
8114 @var{y} values. Note that not all constants should be used, for
8115 example the text size is not known when evaluating the expression, so
8116 the constants @var{text_w} and @var{text_h} will have an undefined
8119 @item expr_int_format, eif
8120 Evaluate the expression's value and output as formatted integer.
8122 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8123 The second argument specifies the output format. Allowed values are @samp{x},
8124 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8125 @code{printf} function.
8126 The third parameter is optional and sets the number of positions taken by the output.
8127 It can be used to add padding with zeros from the left.
8130 The time at which the filter is running, expressed in UTC.
8131 It can accept an argument: a strftime() format string.
8134 The time at which the filter is running, expressed in the local time zone.
8135 It can accept an argument: a strftime() format string.
8138 Frame metadata. Takes one or two arguments.
8140 The first argument is mandatory and specifies the metadata key.
8142 The second argument is optional and specifies a default value, used when the
8143 metadata key is not found or empty.
8146 The frame number, starting from 0.
8149 A 1 character description of the current picture type.
8152 The timestamp of the current frame.
8153 It can take up to three arguments.
8155 The first argument is the format of the timestamp; it defaults to @code{flt}
8156 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8157 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8158 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8159 @code{localtime} stands for the timestamp of the frame formatted as
8160 local time zone time.
8162 The second argument is an offset added to the timestamp.
8164 If the format is set to @code{localtime} or @code{gmtime},
8165 a third argument may be supplied: a strftime() format string.
8166 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8169 @subsection Examples
8173 Draw "Test Text" with font FreeSerif, using the default values for the
8174 optional parameters.
8177 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8181 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8182 and y=50 (counting from the top-left corner of the screen), text is
8183 yellow with a red box around it. Both the text and the box have an
8187 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8188 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8191 Note that the double quotes are not necessary if spaces are not used
8192 within the parameter list.
8195 Show the text at the center of the video frame:
8197 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8201 Show the text at a random position, switching to a new position every 30 seconds:
8203 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)"
8207 Show a text line sliding from right to left in the last row of the video
8208 frame. The file @file{LONG_LINE} is assumed to contain a single line
8211 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8215 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8217 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8221 Draw a single green letter "g", at the center of the input video.
8222 The glyph baseline is placed at half screen height.
8224 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8228 Show text for 1 second every 3 seconds:
8230 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8234 Use fontconfig to set the font. Note that the colons need to be escaped.
8236 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8240 Print the date of a real-time encoding (see strftime(3)):
8242 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8246 Show text fading in and out (appearing/disappearing):
8249 DS=1.0 # display start
8250 DE=10.0 # display end
8251 FID=1.5 # fade in duration
8252 FOD=5 # fade out duration
8253 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 @}"
8257 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8258 and the @option{fontsize} value are included in the @option{y} offset.
8260 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8261 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8266 For more information about libfreetype, check:
8267 @url{http://www.freetype.org/}.
8269 For more information about fontconfig, check:
8270 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8272 For more information about libfribidi, check:
8273 @url{http://fribidi.org/}.
8277 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8279 The filter accepts the following options:
8284 Set low and high threshold values used by the Canny thresholding
8287 The high threshold selects the "strong" edge pixels, which are then
8288 connected through 8-connectivity with the "weak" edge pixels selected
8289 by the low threshold.
8291 @var{low} and @var{high} threshold values must be chosen in the range
8292 [0,1], and @var{low} should be lesser or equal to @var{high}.
8294 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8298 Define the drawing mode.
8302 Draw white/gray wires on black background.
8305 Mix the colors to create a paint/cartoon effect.
8308 Apply Canny edge detector on all selected planes.
8310 Default value is @var{wires}.
8313 Select planes for filtering. By default all available planes are filtered.
8316 @subsection Examples
8320 Standard edge detection with custom values for the hysteresis thresholding:
8322 edgedetect=low=0.1:high=0.4
8326 Painting effect without thresholding:
8328 edgedetect=mode=colormix:high=0
8333 Set brightness, contrast, saturation and approximate gamma adjustment.
8335 The filter accepts the following options:
8339 Set the contrast expression. The value must be a float value in range
8340 @code{-2.0} to @code{2.0}. The default value is "1".
8343 Set the brightness expression. The value must be a float value in
8344 range @code{-1.0} to @code{1.0}. The default value is "0".
8347 Set the saturation expression. The value must be a float in
8348 range @code{0.0} to @code{3.0}. The default value is "1".
8351 Set the gamma expression. The value must be a float in range
8352 @code{0.1} to @code{10.0}. The default value is "1".
8355 Set the gamma expression for red. The value must be a float in
8356 range @code{0.1} to @code{10.0}. The default value is "1".
8359 Set the gamma expression for green. The value must be a float in range
8360 @code{0.1} to @code{10.0}. The default value is "1".
8363 Set the gamma expression for blue. The value must be a float in range
8364 @code{0.1} to @code{10.0}. The default value is "1".
8367 Set the gamma weight expression. It can be used to reduce the effect
8368 of a high gamma value on bright image areas, e.g. keep them from
8369 getting overamplified and just plain white. The value must be a float
8370 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8371 gamma correction all the way down while @code{1.0} leaves it at its
8372 full strength. Default is "1".
8375 Set when the expressions for brightness, contrast, saturation and
8376 gamma expressions are evaluated.
8378 It accepts the following values:
8381 only evaluate expressions once during the filter initialization or
8382 when a command is processed
8385 evaluate expressions for each incoming frame
8388 Default value is @samp{init}.
8391 The expressions accept the following parameters:
8394 frame count of the input frame starting from 0
8397 byte position of the corresponding packet in the input file, NAN if
8401 frame rate of the input video, NAN if the input frame rate is unknown
8404 timestamp expressed in seconds, NAN if the input timestamp is unknown
8407 @subsection Commands
8408 The filter supports the following commands:
8412 Set the contrast expression.
8415 Set the brightness expression.
8418 Set the saturation expression.
8421 Set the gamma expression.
8424 Set the gamma_r expression.
8427 Set gamma_g expression.
8430 Set gamma_b expression.
8433 Set gamma_weight expression.
8435 The command accepts the same syntax of the corresponding option.
8437 If the specified expression is not valid, it is kept at its current
8444 Apply erosion effect to the video.
8446 This filter replaces the pixel by the local(3x3) minimum.
8448 It accepts the following options:
8455 Limit the maximum change for each plane, default is 65535.
8456 If 0, plane will remain unchanged.
8459 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8462 Flags to local 3x3 coordinates maps like this:
8469 @section extractplanes
8471 Extract color channel components from input video stream into
8472 separate grayscale video streams.
8474 The filter accepts the following option:
8478 Set plane(s) to extract.
8480 Available values for planes are:
8491 Choosing planes not available in the input will result in an error.
8492 That means you cannot select @code{r}, @code{g}, @code{b} planes
8493 with @code{y}, @code{u}, @code{v} planes at same time.
8496 @subsection Examples
8500 Extract luma, u and v color channel component from input video frame
8501 into 3 grayscale outputs:
8503 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
8509 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8511 For each input image, the filter will compute the optimal mapping from
8512 the input to the output given the codebook length, that is the number
8513 of distinct output colors.
8515 This filter accepts the following options.
8518 @item codebook_length, l
8519 Set codebook length. The value must be a positive integer, and
8520 represents the number of distinct output colors. Default value is 256.
8523 Set the maximum number of iterations to apply for computing the optimal
8524 mapping. The higher the value the better the result and the higher the
8525 computation time. Default value is 1.
8528 Set a random seed, must be an integer included between 0 and
8529 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8530 will try to use a good random seed on a best effort basis.
8533 Set pal8 output pixel format. This option does not work with codebook
8534 length greater than 256.
8539 Measure graylevel entropy in histogram of color channels of video frames.
8541 It accepts the following parameters:
8545 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8547 @var{diff} mode measures entropy of histogram delta values, absolute differences
8548 between neighbour histogram values.
8553 Apply a fade-in/out effect to the input video.
8555 It accepts the following parameters:
8559 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8561 Default is @code{in}.
8563 @item start_frame, s
8564 Specify the number of the frame to start applying the fade
8565 effect at. Default is 0.
8568 The number of frames that the fade effect lasts. At the end of the
8569 fade-in effect, the output video will have the same intensity as the input video.
8570 At the end of the fade-out transition, the output video will be filled with the
8571 selected @option{color}.
8575 If set to 1, fade only alpha channel, if one exists on the input.
8578 @item start_time, st
8579 Specify the timestamp (in seconds) of the frame to start to apply the fade
8580 effect. If both start_frame and start_time are specified, the fade will start at
8581 whichever comes last. Default is 0.
8584 The number of seconds for which the fade effect has to last. At the end of the
8585 fade-in effect the output video will have the same intensity as the input video,
8586 at the end of the fade-out transition the output video will be filled with the
8587 selected @option{color}.
8588 If both duration and nb_frames are specified, duration is used. Default is 0
8589 (nb_frames is used by default).
8592 Specify the color of the fade. Default is "black".
8595 @subsection Examples
8599 Fade in the first 30 frames of video:
8604 The command above is equivalent to:
8610 Fade out the last 45 frames of a 200-frame video:
8613 fade=type=out:start_frame=155:nb_frames=45
8617 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8619 fade=in:0:25, fade=out:975:25
8623 Make the first 5 frames yellow, then fade in from frame 5-24:
8625 fade=in:5:20:color=yellow
8629 Fade in alpha over first 25 frames of video:
8631 fade=in:0:25:alpha=1
8635 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8637 fade=t=in:st=5.5:d=0.5
8643 Apply arbitrary expressions to samples in frequency domain
8647 Adjust the dc value (gain) of the luma plane of the image. The filter
8648 accepts an integer value in range @code{0} to @code{1000}. The default
8649 value is set to @code{0}.
8652 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8653 filter accepts an integer value in range @code{0} to @code{1000}. The
8654 default value is set to @code{0}.
8657 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8658 filter accepts an integer value in range @code{0} to @code{1000}. The
8659 default value is set to @code{0}.
8662 Set the frequency domain weight expression for the luma plane.
8665 Set the frequency domain weight expression for the 1st chroma plane.
8668 Set the frequency domain weight expression for the 2nd chroma plane.
8671 Set when the expressions are evaluated.
8673 It accepts the following values:
8676 Only evaluate expressions once during the filter initialization.
8679 Evaluate expressions for each incoming frame.
8682 Default value is @samp{init}.
8684 The filter accepts the following variables:
8687 The coordinates of the current sample.
8691 The width and height of the image.
8694 The number of input frame, starting from 0.
8697 @subsection Examples
8703 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8709 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8715 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8721 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8727 Denoise frames using 3D FFT (frequency domain filtering).
8729 The filter accepts the following options:
8733 Set the noise sigma constant. This sets denoising strength.
8734 Default value is 1. Allowed range is from 0 to 30.
8735 Using very high sigma with low overlap may give blocking artifacts.
8738 Set amount of denoising. By default all detected noise is reduced.
8739 Default value is 1. Allowed range is from 0 to 1.
8742 Set size of block, Default is 4, can be 3, 4, 5 or 6.
8743 Actual size of block in pixels is 2 to power of @var{block}, so by default
8744 block size in pixels is 2^4 which is 16.
8747 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
8750 Set number of previous frames to use for denoising. By default is set to 0.
8753 Set number of next frames to to use for denoising. By default is set to 0.
8756 Set planes which will be filtered, by default are all available filtered
8762 Extract a single field from an interlaced image using stride
8763 arithmetic to avoid wasting CPU time. The output frames are marked as
8766 The filter accepts the following options:
8770 Specify whether to extract the top (if the value is @code{0} or
8771 @code{top}) or the bottom field (if the value is @code{1} or
8777 Create new frames by copying the top and bottom fields from surrounding frames
8778 supplied as numbers by the hint file.
8782 Set file containing hints: absolute/relative frame numbers.
8784 There must be one line for each frame in a clip. Each line must contain two
8785 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8786 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8787 is current frame number for @code{absolute} mode or out of [-1, 1] range
8788 for @code{relative} mode. First number tells from which frame to pick up top
8789 field and second number tells from which frame to pick up bottom field.
8791 If optionally followed by @code{+} output frame will be marked as interlaced,
8792 else if followed by @code{-} output frame will be marked as progressive, else
8793 it will be marked same as input frame.
8794 If line starts with @code{#} or @code{;} that line is skipped.
8797 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8800 Example of first several lines of @code{hint} file for @code{relative} mode:
8803 1,0 - # second frame, use third's frame top field and second's frame bottom field
8804 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8821 Field matching filter for inverse telecine. It is meant to reconstruct the
8822 progressive frames from a telecined stream. The filter does not drop duplicated
8823 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8824 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8826 The separation of the field matching and the decimation is notably motivated by
8827 the possibility of inserting a de-interlacing filter fallback between the two.
8828 If the source has mixed telecined and real interlaced content,
8829 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8830 But these remaining combed frames will be marked as interlaced, and thus can be
8831 de-interlaced by a later filter such as @ref{yadif} before decimation.
8833 In addition to the various configuration options, @code{fieldmatch} can take an
8834 optional second stream, activated through the @option{ppsrc} option. If
8835 enabled, the frames reconstruction will be based on the fields and frames from
8836 this second stream. This allows the first input to be pre-processed in order to
8837 help the various algorithms of the filter, while keeping the output lossless
8838 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8839 or brightness/contrast adjustments can help.
8841 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8842 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8843 which @code{fieldmatch} is based on. While the semantic and usage are very
8844 close, some behaviour and options names can differ.
8846 The @ref{decimate} filter currently only works for constant frame rate input.
8847 If your input has mixed telecined (30fps) and progressive content with a lower
8848 framerate like 24fps use the following filterchain to produce the necessary cfr
8849 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8851 The filter accepts the following options:
8855 Specify the assumed field order of the input stream. Available values are:
8859 Auto detect parity (use FFmpeg's internal parity value).
8861 Assume bottom field first.
8863 Assume top field first.
8866 Note that it is sometimes recommended not to trust the parity announced by the
8869 Default value is @var{auto}.
8872 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8873 sense that it won't risk creating jerkiness due to duplicate frames when
8874 possible, but if there are bad edits or blended fields it will end up
8875 outputting combed frames when a good match might actually exist. On the other
8876 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8877 but will almost always find a good frame if there is one. The other values are
8878 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8879 jerkiness and creating duplicate frames versus finding good matches in sections
8880 with bad edits, orphaned fields, blended fields, etc.
8882 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8884 Available values are:
8888 2-way matching (p/c)
8890 2-way matching, and trying 3rd match if still combed (p/c + n)
8892 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8894 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8895 still combed (p/c + n + u/b)
8897 3-way matching (p/c/n)
8899 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8900 detected as combed (p/c/n + u/b)
8903 The parenthesis at the end indicate the matches that would be used for that
8904 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8907 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8910 Default value is @var{pc_n}.
8913 Mark the main input stream as a pre-processed input, and enable the secondary
8914 input stream as the clean source to pick the fields from. See the filter
8915 introduction for more details. It is similar to the @option{clip2} feature from
8918 Default value is @code{0} (disabled).
8921 Set the field to match from. It is recommended to set this to the same value as
8922 @option{order} unless you experience matching failures with that setting. In
8923 certain circumstances changing the field that is used to match from can have a
8924 large impact on matching performance. Available values are:
8928 Automatic (same value as @option{order}).
8930 Match from the bottom field.
8932 Match from the top field.
8935 Default value is @var{auto}.
8938 Set whether or not chroma is included during the match comparisons. In most
8939 cases it is recommended to leave this enabled. You should set this to @code{0}
8940 only if your clip has bad chroma problems such as heavy rainbowing or other
8941 artifacts. Setting this to @code{0} could also be used to speed things up at
8942 the cost of some accuracy.
8944 Default value is @code{1}.
8948 These define an exclusion band which excludes the lines between @option{y0} and
8949 @option{y1} from being included in the field matching decision. An exclusion
8950 band can be used to ignore subtitles, a logo, or other things that may
8951 interfere with the matching. @option{y0} sets the starting scan line and
8952 @option{y1} sets the ending line; all lines in between @option{y0} and
8953 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8954 @option{y0} and @option{y1} to the same value will disable the feature.
8955 @option{y0} and @option{y1} defaults to @code{0}.
8958 Set the scene change detection threshold as a percentage of maximum change on
8959 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8960 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8961 @option{scthresh} is @code{[0.0, 100.0]}.
8963 Default value is @code{12.0}.
8966 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8967 account the combed scores of matches when deciding what match to use as the
8968 final match. Available values are:
8972 No final matching based on combed scores.
8974 Combed scores are only used when a scene change is detected.
8976 Use combed scores all the time.
8979 Default is @var{sc}.
8982 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8983 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8984 Available values are:
8988 No forced calculation.
8990 Force p/c/n calculations.
8992 Force p/c/n/u/b calculations.
8995 Default value is @var{none}.
8998 This is the area combing threshold used for combed frame detection. This
8999 essentially controls how "strong" or "visible" combing must be to be detected.
9000 Larger values mean combing must be more visible and smaller values mean combing
9001 can be less visible or strong and still be detected. Valid settings are from
9002 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9003 be detected as combed). This is basically a pixel difference value. A good
9004 range is @code{[8, 12]}.
9006 Default value is @code{9}.
9009 Sets whether or not chroma is considered in the combed frame decision. Only
9010 disable this if your source has chroma problems (rainbowing, etc.) that are
9011 causing problems for the combed frame detection with chroma enabled. Actually,
9012 using @option{chroma}=@var{0} is usually more reliable, except for the case
9013 where there is chroma only combing in the source.
9015 Default value is @code{0}.
9019 Respectively set the x-axis and y-axis size of the window used during combed
9020 frame detection. This has to do with the size of the area in which
9021 @option{combpel} pixels are required to be detected as combed for a frame to be
9022 declared combed. See the @option{combpel} parameter description for more info.
9023 Possible values are any number that is a power of 2 starting at 4 and going up
9026 Default value is @code{16}.
9029 The number of combed pixels inside any of the @option{blocky} by
9030 @option{blockx} size blocks on the frame for the frame to be detected as
9031 combed. While @option{cthresh} controls how "visible" the combing must be, this
9032 setting controls "how much" combing there must be in any localized area (a
9033 window defined by the @option{blockx} and @option{blocky} settings) on the
9034 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9035 which point no frames will ever be detected as combed). This setting is known
9036 as @option{MI} in TFM/VFM vocabulary.
9038 Default value is @code{80}.
9041 @anchor{p/c/n/u/b meaning}
9042 @subsection p/c/n/u/b meaning
9044 @subsubsection p/c/n
9046 We assume the following telecined stream:
9049 Top fields: 1 2 2 3 4
9050 Bottom fields: 1 2 3 4 4
9053 The numbers correspond to the progressive frame the fields relate to. Here, the
9054 first two frames are progressive, the 3rd and 4th are combed, and so on.
9056 When @code{fieldmatch} is configured to run a matching from bottom
9057 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9062 B 1 2 3 4 4 <-- matching reference
9071 As a result of the field matching, we can see that some frames get duplicated.
9072 To perform a complete inverse telecine, you need to rely on a decimation filter
9073 after this operation. See for instance the @ref{decimate} filter.
9075 The same operation now matching from top fields (@option{field}=@var{top})
9080 T 1 2 2 3 4 <-- matching reference
9090 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9091 basically, they refer to the frame and field of the opposite parity:
9094 @item @var{p} matches the field of the opposite parity in the previous frame
9095 @item @var{c} matches the field of the opposite parity in the current frame
9096 @item @var{n} matches the field of the opposite parity in the next frame
9101 The @var{u} and @var{b} matching are a bit special in the sense that they match
9102 from the opposite parity flag. In the following examples, we assume that we are
9103 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9104 'x' is placed above and below each matched fields.
9106 With bottom matching (@option{field}=@var{bottom}):
9111 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9112 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9120 With top matching (@option{field}=@var{top}):
9125 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9126 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9134 @subsection Examples
9136 Simple IVTC of a top field first telecined stream:
9138 fieldmatch=order=tff:combmatch=none, decimate
9141 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9143 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9148 Transform the field order of the input video.
9150 It accepts the following parameters:
9155 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9156 for bottom field first.
9159 The default value is @samp{tff}.
9161 The transformation is done by shifting the picture content up or down
9162 by one line, and filling the remaining line with appropriate picture content.
9163 This method is consistent with most broadcast field order converters.
9165 If the input video is not flagged as being interlaced, or it is already
9166 flagged as being of the required output field order, then this filter does
9167 not alter the incoming video.
9169 It is very useful when converting to or from PAL DV material,
9170 which is bottom field first.
9174 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9177 @section fifo, afifo
9179 Buffer input images and send them when they are requested.
9181 It is mainly useful when auto-inserted by the libavfilter
9184 It does not take parameters.
9186 @section fillborders
9188 Fill borders of the input video, without changing video stream dimensions.
9189 Sometimes video can have garbage at the four edges and you may not want to
9190 crop video input to keep size multiple of some number.
9192 This filter accepts the following options:
9196 Number of pixels to fill from left border.
9199 Number of pixels to fill from right border.
9202 Number of pixels to fill from top border.
9205 Number of pixels to fill from bottom border.
9210 It accepts the following values:
9213 fill pixels using outermost pixels
9216 fill pixels using mirroring
9219 fill pixels with constant value
9222 Default is @var{smear}.
9225 Set color for pixels in fixed mode. Default is @var{black}.
9230 Find a rectangular object
9232 It accepts the following options:
9236 Filepath of the object image, needs to be in gray8.
9239 Detection threshold, default is 0.5.
9242 Number of mipmaps, default is 3.
9244 @item xmin, ymin, xmax, ymax
9245 Specifies the rectangle in which to search.
9248 @subsection Examples
9252 Generate a representative palette of a given video using @command{ffmpeg}:
9254 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9260 Cover a rectangular object
9262 It accepts the following options:
9266 Filepath of the optional cover image, needs to be in yuv420.
9271 It accepts the following values:
9274 cover it by the supplied image
9276 cover it by interpolating the surrounding pixels
9279 Default value is @var{blur}.
9282 @subsection Examples
9286 Generate a representative palette of a given video using @command{ffmpeg}:
9288 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9294 Flood area with values of same pixel components with another values.
9296 It accepts the following options:
9299 Set pixel x coordinate.
9302 Set pixel y coordinate.
9305 Set source #0 component value.
9308 Set source #1 component value.
9311 Set source #2 component value.
9314 Set source #3 component value.
9317 Set destination #0 component value.
9320 Set destination #1 component value.
9323 Set destination #2 component value.
9326 Set destination #3 component value.
9332 Convert the input video to one of the specified pixel formats.
9333 Libavfilter will try to pick one that is suitable as input to
9336 It accepts the following parameters:
9340 A '|'-separated list of pixel format names, such as
9341 "pix_fmts=yuv420p|monow|rgb24".
9345 @subsection Examples
9349 Convert the input video to the @var{yuv420p} format
9351 format=pix_fmts=yuv420p
9354 Convert the input video to any of the formats in the list
9356 format=pix_fmts=yuv420p|yuv444p|yuv410p
9363 Convert the video to specified constant frame rate by duplicating or dropping
9364 frames as necessary.
9366 It accepts the following parameters:
9370 The desired output frame rate. The default is @code{25}.
9373 Assume the first PTS should be the given value, in seconds. This allows for
9374 padding/trimming at the start of stream. By default, no assumption is made
9375 about the first frame's expected PTS, so no padding or trimming is done.
9376 For example, this could be set to 0 to pad the beginning with duplicates of
9377 the first frame if a video stream starts after the audio stream or to trim any
9378 frames with a negative PTS.
9381 Timestamp (PTS) rounding method.
9383 Possible values are:
9390 round towards -infinity
9392 round towards +infinity
9396 The default is @code{near}.
9399 Action performed when reading the last frame.
9401 Possible values are:
9404 Use same timestamp rounding method as used for other frames.
9406 Pass through last frame if input duration has not been reached yet.
9408 The default is @code{round}.
9412 Alternatively, the options can be specified as a flat string:
9413 @var{fps}[:@var{start_time}[:@var{round}]].
9415 See also the @ref{setpts} filter.
9417 @subsection Examples
9421 A typical usage in order to set the fps to 25:
9427 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9429 fps=fps=film:round=near
9435 Pack two different video streams into a stereoscopic video, setting proper
9436 metadata on supported codecs. The two views should have the same size and
9437 framerate and processing will stop when the shorter video ends. Please note
9438 that you may conveniently adjust view properties with the @ref{scale} and
9441 It accepts the following parameters:
9445 The desired packing format. Supported values are:
9450 The views are next to each other (default).
9453 The views are on top of each other.
9456 The views are packed by line.
9459 The views are packed by column.
9462 The views are temporally interleaved.
9471 # Convert left and right views into a frame-sequential video
9472 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9474 # Convert views into a side-by-side video with the same output resolution as the input
9475 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
9480 Change the frame rate by interpolating new video output frames from the source
9483 This filter is not designed to function correctly with interlaced media. If
9484 you wish to change the frame rate of interlaced media then you are required
9485 to deinterlace before this filter and re-interlace after this filter.
9487 A description of the accepted options follows.
9491 Specify the output frames per second. This option can also be specified
9492 as a value alone. The default is @code{50}.
9495 Specify the start of a range where the output frame will be created as a
9496 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9497 the default is @code{15}.
9500 Specify the end of a range where the output frame will be created as a
9501 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9502 the default is @code{240}.
9505 Specify the level at which a scene change is detected as a value between
9506 0 and 100 to indicate a new scene; a low value reflects a low
9507 probability for the current frame to introduce a new scene, while a higher
9508 value means the current frame is more likely to be one.
9509 The default is @code{8.2}.
9512 Specify flags influencing the filter process.
9514 Available value for @var{flags} is:
9517 @item scene_change_detect, scd
9518 Enable scene change detection using the value of the option @var{scene}.
9519 This flag is enabled by default.
9525 Select one frame every N-th frame.
9527 This filter accepts the following option:
9530 Select frame after every @code{step} frames.
9531 Allowed values are positive integers higher than 0. Default value is @code{1}.
9537 Apply a frei0r effect to the input video.
9539 To enable the compilation of this filter, you need to install the frei0r
9540 header and configure FFmpeg with @code{--enable-frei0r}.
9542 It accepts the following parameters:
9547 The name of the frei0r effect to load. If the environment variable
9548 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9549 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9550 Otherwise, the standard frei0r paths are searched, in this order:
9551 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9552 @file{/usr/lib/frei0r-1/}.
9555 A '|'-separated list of parameters to pass to the frei0r effect.
9559 A frei0r effect parameter can be a boolean (its value is either
9560 "y" or "n"), a double, a color (specified as
9561 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9562 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9563 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9564 a position (specified as @var{X}/@var{Y}, where
9565 @var{X} and @var{Y} are floating point numbers) and/or a string.
9567 The number and types of parameters depend on the loaded effect. If an
9568 effect parameter is not specified, the default value is set.
9570 @subsection Examples
9574 Apply the distort0r effect, setting the first two double parameters:
9576 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9580 Apply the colordistance effect, taking a color as the first parameter:
9582 frei0r=colordistance:0.2/0.3/0.4
9583 frei0r=colordistance:violet
9584 frei0r=colordistance:0x112233
9588 Apply the perspective effect, specifying the top left and top right image
9591 frei0r=perspective:0.2/0.2|0.8/0.2
9595 For more information, see
9596 @url{http://frei0r.dyne.org}
9600 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9602 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9603 processing filter, one of them is performed once per block, not per pixel.
9604 This allows for much higher speed.
9606 The filter accepts the following options:
9610 Set quality. This option defines the number of levels for averaging. It accepts
9611 an integer in the range 4-5. Default value is @code{4}.
9614 Force a constant quantization parameter. It accepts an integer in range 0-63.
9615 If not set, the filter will use the QP from the video stream (if available).
9618 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9619 more details but also more artifacts, while higher values make the image smoother
9620 but also blurrier. Default value is @code{0} − PSNR optimal.
9623 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9624 option may cause flicker since the B-Frames have often larger QP. Default is
9625 @code{0} (not enabled).
9631 Apply Gaussian blur filter.
9633 The filter accepts the following options:
9637 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9640 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9643 Set which planes to filter. By default all planes are filtered.
9646 Set vertical sigma, if negative it will be same as @code{sigma}.
9647 Default is @code{-1}.
9652 The filter accepts the following options:
9656 Set the luminance expression.
9658 Set the chrominance blue expression.
9660 Set the chrominance red expression.
9662 Set the alpha expression.
9664 Set the red expression.
9666 Set the green expression.
9668 Set the blue expression.
9671 The colorspace is selected according to the specified options. If one
9672 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9673 options is specified, the filter will automatically select a YCbCr
9674 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9675 @option{blue_expr} options is specified, it will select an RGB
9678 If one of the chrominance expression is not defined, it falls back on the other
9679 one. If no alpha expression is specified it will evaluate to opaque value.
9680 If none of chrominance expressions are specified, they will evaluate
9681 to the luminance expression.
9683 The expressions can use the following variables and functions:
9687 The sequential number of the filtered frame, starting from @code{0}.
9691 The coordinates of the current sample.
9695 The width and height of the image.
9699 Width and height scale depending on the currently filtered plane. It is the
9700 ratio between the corresponding luma plane number of pixels and the current
9701 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9702 @code{0.5,0.5} for chroma planes.
9705 Time of the current frame, expressed in seconds.
9708 Return the value of the pixel at location (@var{x},@var{y}) of the current
9712 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9716 Return the value of the pixel at location (@var{x},@var{y}) of the
9717 blue-difference chroma plane. Return 0 if there is no such plane.
9720 Return the value of the pixel at location (@var{x},@var{y}) of the
9721 red-difference chroma plane. Return 0 if there is no such plane.
9726 Return the value of the pixel at location (@var{x},@var{y}) of the
9727 red/green/blue component. Return 0 if there is no such component.
9730 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9731 plane. Return 0 if there is no such plane.
9734 For functions, if @var{x} and @var{y} are outside the area, the value will be
9735 automatically clipped to the closer edge.
9737 @subsection Examples
9741 Flip the image horizontally:
9747 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9748 wavelength of 100 pixels:
9750 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9754 Generate a fancy enigmatic moving light:
9756 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
9760 Generate a quick emboss effect:
9762 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9766 Modify RGB components depending on pixel position:
9768 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9772 Create a radial gradient that is the same size as the input (also see
9773 the @ref{vignette} filter):
9775 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9781 Fix the banding artifacts that are sometimes introduced into nearly flat
9782 regions by truncation to 8-bit color depth.
9783 Interpolate the gradients that should go where the bands are, and
9786 It is designed for playback only. Do not use it prior to
9787 lossy compression, because compression tends to lose the dither and
9788 bring back the bands.
9790 It accepts the following parameters:
9795 The maximum amount by which the filter will change any one pixel. This is also
9796 the threshold for detecting nearly flat regions. Acceptable values range from
9797 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9801 The neighborhood to fit the gradient to. A larger radius makes for smoother
9802 gradients, but also prevents the filter from modifying the pixels near detailed
9803 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9804 values will be clipped to the valid range.
9808 Alternatively, the options can be specified as a flat string:
9809 @var{strength}[:@var{radius}]
9811 @subsection Examples
9815 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9821 Specify radius, omitting the strength (which will fall-back to the default
9832 Apply a Hald CLUT to a video stream.
9834 First input is the video stream to process, and second one is the Hald CLUT.
9835 The Hald CLUT input can be a simple picture or a complete video stream.
9837 The filter accepts the following options:
9841 Force termination when the shortest input terminates. Default is @code{0}.
9843 Continue applying the last CLUT after the end of the stream. A value of
9844 @code{0} disable the filter after the last frame of the CLUT is reached.
9845 Default is @code{1}.
9848 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9849 filters share the same internals).
9851 More information about the Hald CLUT can be found on Eskil Steenberg's website
9852 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9854 @subsection Workflow examples
9856 @subsubsection Hald CLUT video stream
9858 Generate an identity Hald CLUT stream altered with various effects:
9860 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
9863 Note: make sure you use a lossless codec.
9865 Then use it with @code{haldclut} to apply it on some random stream:
9867 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9870 The Hald CLUT will be applied to the 10 first seconds (duration of
9871 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9872 to the remaining frames of the @code{mandelbrot} stream.
9874 @subsubsection Hald CLUT with preview
9876 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9877 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9878 biggest possible square starting at the top left of the picture. The remaining
9879 padding pixels (bottom or right) will be ignored. This area can be used to add
9880 a preview of the Hald CLUT.
9882 Typically, the following generated Hald CLUT will be supported by the
9883 @code{haldclut} filter:
9886 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9887 pad=iw+320 [padded_clut];
9888 smptebars=s=320x256, split [a][b];
9889 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9890 [main][b] overlay=W-320" -frames:v 1 clut.png
9893 It contains the original and a preview of the effect of the CLUT: SMPTE color
9894 bars are displayed on the right-top, and below the same color bars processed by
9897 Then, the effect of this Hald CLUT can be visualized with:
9899 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9904 Flip the input video horizontally.
9906 For example, to horizontally flip the input video with @command{ffmpeg}:
9908 ffmpeg -i in.avi -vf "hflip" out.avi
9912 This filter applies a global color histogram equalization on a
9915 It can be used to correct video that has a compressed range of pixel
9916 intensities. The filter redistributes the pixel intensities to
9917 equalize their distribution across the intensity range. It may be
9918 viewed as an "automatically adjusting contrast filter". This filter is
9919 useful only for correcting degraded or poorly captured source
9922 The filter accepts the following options:
9926 Determine the amount of equalization to be applied. As the strength
9927 is reduced, the distribution of pixel intensities more-and-more
9928 approaches that of the input frame. The value must be a float number
9929 in the range [0,1] and defaults to 0.200.
9932 Set the maximum intensity that can generated and scale the output
9933 values appropriately. The strength should be set as desired and then
9934 the intensity can be limited if needed to avoid washing-out. The value
9935 must be a float number in the range [0,1] and defaults to 0.210.
9938 Set the antibanding level. If enabled the filter will randomly vary
9939 the luminance of output pixels by a small amount to avoid banding of
9940 the histogram. Possible values are @code{none}, @code{weak} or
9941 @code{strong}. It defaults to @code{none}.
9946 Compute and draw a color distribution histogram for the input video.
9948 The computed histogram is a representation of the color component
9949 distribution in an image.
9951 Standard histogram displays the color components distribution in an image.
9952 Displays color graph for each color component. Shows distribution of
9953 the Y, U, V, A or R, G, B components, depending on input format, in the
9954 current frame. Below each graph a color component scale meter is shown.
9956 The filter accepts the following options:
9960 Set height of level. Default value is @code{200}.
9961 Allowed range is [50, 2048].
9964 Set height of color scale. Default value is @code{12}.
9965 Allowed range is [0, 40].
9969 It accepts the following values:
9972 Per color component graphs are placed below each other.
9975 Per color component graphs are placed side by side.
9978 Presents information identical to that in the @code{parade}, except
9979 that the graphs representing color components are superimposed directly
9982 Default is @code{stack}.
9985 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9986 Default is @code{linear}.
9989 Set what color components to display.
9990 Default is @code{7}.
9993 Set foreground opacity. Default is @code{0.7}.
9996 Set background opacity. Default is @code{0.5}.
9999 @subsection Examples
10004 Calculate and draw histogram:
10006 ffplay -i input -vf histogram
10014 This is a high precision/quality 3d denoise filter. It aims to reduce
10015 image noise, producing smooth images and making still images really
10016 still. It should enhance compressibility.
10018 It accepts the following optional parameters:
10022 A non-negative floating point number which specifies spatial luma strength.
10023 It defaults to 4.0.
10025 @item chroma_spatial
10026 A non-negative floating point number which specifies spatial chroma strength.
10027 It defaults to 3.0*@var{luma_spatial}/4.0.
10030 A floating point number which specifies luma temporal strength. It defaults to
10031 6.0*@var{luma_spatial}/4.0.
10034 A floating point number which specifies chroma temporal strength. It defaults to
10035 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10038 @section hwdownload
10040 Download hardware frames to system memory.
10042 The input must be in hardware frames, and the output a non-hardware format.
10043 Not all formats will be supported on the output - it may be necessary to insert
10044 an additional @option{format} filter immediately following in the graph to get
10045 the output in a supported format.
10049 Map hardware frames to system memory or to another device.
10051 This filter has several different modes of operation; which one is used depends
10052 on the input and output formats:
10055 Hardware frame input, normal frame output
10057 Map the input frames to system memory and pass them to the output. If the
10058 original hardware frame is later required (for example, after overlaying
10059 something else on part of it), the @option{hwmap} filter can be used again
10060 in the next mode to retrieve it.
10062 Normal frame input, hardware frame output
10064 If the input is actually a software-mapped hardware frame, then unmap it -
10065 that is, return the original hardware frame.
10067 Otherwise, a device must be provided. Create new hardware surfaces on that
10068 device for the output, then map them back to the software format at the input
10069 and give those frames to the preceding filter. This will then act like the
10070 @option{hwupload} filter, but may be able to avoid an additional copy when
10071 the input is already in a compatible format.
10073 Hardware frame input and output
10075 A device must be supplied for the output, either directly or with the
10076 @option{derive_device} option. The input and output devices must be of
10077 different types and compatible - the exact meaning of this is
10078 system-dependent, but typically it means that they must refer to the same
10079 underlying hardware context (for example, refer to the same graphics card).
10081 If the input frames were originally created on the output device, then unmap
10082 to retrieve the original frames.
10084 Otherwise, map the frames to the output device - create new hardware frames
10085 on the output corresponding to the frames on the input.
10088 The following additional parameters are accepted:
10092 Set the frame mapping mode. Some combination of:
10095 The mapped frame should be readable.
10097 The mapped frame should be writeable.
10099 The mapping will always overwrite the entire frame.
10101 This may improve performance in some cases, as the original contents of the
10102 frame need not be loaded.
10104 The mapping must not involve any copying.
10106 Indirect mappings to copies of frames are created in some cases where either
10107 direct mapping is not possible or it would have unexpected properties.
10108 Setting this flag ensures that the mapping is direct and will fail if that is
10111 Defaults to @var{read+write} if not specified.
10113 @item derive_device @var{type}
10114 Rather than using the device supplied at initialisation, instead derive a new
10115 device of type @var{type} from the device the input frames exist on.
10118 In a hardware to hardware mapping, map in reverse - create frames in the sink
10119 and map them back to the source. This may be necessary in some cases where
10120 a mapping in one direction is required but only the opposite direction is
10121 supported by the devices being used.
10123 This option is dangerous - it may break the preceding filter in undefined
10124 ways if there are any additional constraints on that filter's output.
10125 Do not use it without fully understanding the implications of its use.
10130 Upload system memory frames to hardware surfaces.
10132 The device to upload to must be supplied when the filter is initialised. If
10133 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10136 @anchor{hwupload_cuda}
10137 @section hwupload_cuda
10139 Upload system memory frames to a CUDA device.
10141 It accepts the following optional parameters:
10145 The number of the CUDA device to use
10150 Apply a high-quality magnification filter designed for pixel art. This filter
10151 was originally created by Maxim Stepin.
10153 It accepts the following option:
10157 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10158 @code{hq3x} and @code{4} for @code{hq4x}.
10159 Default is @code{3}.
10163 Stack input videos horizontally.
10165 All streams must be of same pixel format and of same height.
10167 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10168 to create same output.
10170 The filter accept the following option:
10174 Set number of input streams. Default is 2.
10177 If set to 1, force the output to terminate when the shortest input
10178 terminates. Default value is 0.
10183 Modify the hue and/or the saturation of the input.
10185 It accepts the following parameters:
10189 Specify the hue angle as a number of degrees. It accepts an expression,
10190 and defaults to "0".
10193 Specify the saturation in the [-10,10] range. It accepts an expression and
10197 Specify the hue angle as a number of radians. It accepts an
10198 expression, and defaults to "0".
10201 Specify the brightness in the [-10,10] range. It accepts an expression and
10205 @option{h} and @option{H} are mutually exclusive, and can't be
10206 specified at the same time.
10208 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10209 expressions containing the following constants:
10213 frame count of the input frame starting from 0
10216 presentation timestamp of the input frame expressed in time base units
10219 frame rate of the input video, NAN if the input frame rate is unknown
10222 timestamp expressed in seconds, NAN if the input timestamp is unknown
10225 time base of the input video
10228 @subsection Examples
10232 Set the hue to 90 degrees and the saturation to 1.0:
10238 Same command but expressing the hue in radians:
10244 Rotate hue and make the saturation swing between 0
10245 and 2 over a period of 1 second:
10247 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10251 Apply a 3 seconds saturation fade-in effect starting at 0:
10253 hue="s=min(t/3\,1)"
10256 The general fade-in expression can be written as:
10258 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10262 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10264 hue="s=max(0\, min(1\, (8-t)/3))"
10267 The general fade-out expression can be written as:
10269 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10274 @subsection Commands
10276 This filter supports the following commands:
10282 Modify the hue and/or the saturation and/or brightness of the input video.
10283 The command accepts the same syntax of the corresponding option.
10285 If the specified expression is not valid, it is kept at its current
10289 @section hysteresis
10291 Grow first stream into second stream by connecting components.
10292 This makes it possible to build more robust edge masks.
10294 This filter accepts the following options:
10298 Set which planes will be processed as bitmap, unprocessed planes will be
10299 copied from first stream.
10300 By default value 0xf, all planes will be processed.
10303 Set threshold which is used in filtering. If pixel component value is higher than
10304 this value filter algorithm for connecting components is activated.
10305 By default value is 0.
10310 Detect video interlacing type.
10312 This filter tries to detect if the input frames are interlaced, progressive,
10313 top or bottom field first. It will also try to detect fields that are
10314 repeated between adjacent frames (a sign of telecine).
10316 Single frame detection considers only immediately adjacent frames when classifying each frame.
10317 Multiple frame detection incorporates the classification history of previous frames.
10319 The filter will log these metadata values:
10322 @item single.current_frame
10323 Detected type of current frame using single-frame detection. One of:
10324 ``tff'' (top field first), ``bff'' (bottom field first),
10325 ``progressive'', or ``undetermined''
10328 Cumulative number of frames detected as top field first using single-frame detection.
10331 Cumulative number of frames detected as top field first using multiple-frame detection.
10334 Cumulative number of frames detected as bottom field first using single-frame detection.
10336 @item multiple.current_frame
10337 Detected type of current frame using multiple-frame detection. One of:
10338 ``tff'' (top field first), ``bff'' (bottom field first),
10339 ``progressive'', or ``undetermined''
10342 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10344 @item single.progressive
10345 Cumulative number of frames detected as progressive using single-frame detection.
10347 @item multiple.progressive
10348 Cumulative number of frames detected as progressive using multiple-frame detection.
10350 @item single.undetermined
10351 Cumulative number of frames that could not be classified using single-frame detection.
10353 @item multiple.undetermined
10354 Cumulative number of frames that could not be classified using multiple-frame detection.
10356 @item repeated.current_frame
10357 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10359 @item repeated.neither
10360 Cumulative number of frames with no repeated field.
10363 Cumulative number of frames with the top field repeated from the previous frame's top field.
10365 @item repeated.bottom
10366 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10369 The filter accepts the following options:
10373 Set interlacing threshold.
10375 Set progressive threshold.
10377 Threshold for repeated field detection.
10379 Number of frames after which a given frame's contribution to the
10380 statistics is halved (i.e., it contributes only 0.5 to its
10381 classification). The default of 0 means that all frames seen are given
10382 full weight of 1.0 forever.
10383 @item analyze_interlaced_flag
10384 When this is not 0 then idet will use the specified number of frames to determine
10385 if the interlaced flag is accurate, it will not count undetermined frames.
10386 If the flag is found to be accurate it will be used without any further
10387 computations, if it is found to be inaccurate it will be cleared without any
10388 further computations. This allows inserting the idet filter as a low computational
10389 method to clean up the interlaced flag
10394 Deinterleave or interleave fields.
10396 This filter allows one to process interlaced images fields without
10397 deinterlacing them. Deinterleaving splits the input frame into 2
10398 fields (so called half pictures). Odd lines are moved to the top
10399 half of the output image, even lines to the bottom half.
10400 You can process (filter) them independently and then re-interleave them.
10402 The filter accepts the following options:
10406 @item chroma_mode, c
10407 @item alpha_mode, a
10408 Available values for @var{luma_mode}, @var{chroma_mode} and
10409 @var{alpha_mode} are:
10415 @item deinterleave, d
10416 Deinterleave fields, placing one above the other.
10418 @item interleave, i
10419 Interleave fields. Reverse the effect of deinterleaving.
10421 Default value is @code{none}.
10423 @item luma_swap, ls
10424 @item chroma_swap, cs
10425 @item alpha_swap, as
10426 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10431 Apply inflate effect to the video.
10433 This filter replaces the pixel by the local(3x3) average by taking into account
10434 only values higher than the pixel.
10436 It accepts the following options:
10443 Limit the maximum change for each plane, default is 65535.
10444 If 0, plane will remain unchanged.
10449 Simple interlacing filter from progressive contents. This interleaves upper (or
10450 lower) lines from odd frames with lower (or upper) lines from even frames,
10451 halving the frame rate and preserving image height.
10454 Original Original New Frame
10455 Frame 'j' Frame 'j+1' (tff)
10456 ========== =========== ==================
10457 Line 0 --------------------> Frame 'j' Line 0
10458 Line 1 Line 1 ----> Frame 'j+1' Line 1
10459 Line 2 ---------------------> Frame 'j' Line 2
10460 Line 3 Line 3 ----> Frame 'j+1' Line 3
10462 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10465 It accepts the following optional parameters:
10469 This determines whether the interlaced frame is taken from the even
10470 (tff - default) or odd (bff) lines of the progressive frame.
10473 Vertical lowpass filter to avoid twitter interlacing and
10474 reduce moire patterns.
10478 Disable vertical lowpass filter
10481 Enable linear filter (default)
10484 Enable complex filter. This will slightly less reduce twitter and moire
10485 but better retain detail and subjective sharpness impression.
10492 Deinterlace input video by applying Donald Graft's adaptive kernel
10493 deinterling. Work on interlaced parts of a video to produce
10494 progressive frames.
10496 The description of the accepted parameters follows.
10500 Set the threshold which affects the filter's tolerance when
10501 determining if a pixel line must be processed. It must be an integer
10502 in the range [0,255] and defaults to 10. A value of 0 will result in
10503 applying the process on every pixels.
10506 Paint pixels exceeding the threshold value to white if set to 1.
10510 Set the fields order. Swap fields if set to 1, leave fields alone if
10514 Enable additional sharpening if set to 1. Default is 0.
10517 Enable twoway sharpening if set to 1. Default is 0.
10520 @subsection Examples
10524 Apply default values:
10526 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10530 Enable additional sharpening:
10536 Paint processed pixels in white:
10542 @section lenscorrection
10544 Correct radial lens distortion
10546 This filter can be used to correct for radial distortion as can result from the use
10547 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10548 one can use tools available for example as part of opencv or simply trial-and-error.
10549 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10550 and extract the k1 and k2 coefficients from the resulting matrix.
10552 Note that effectively the same filter is available in the open-source tools Krita and
10553 Digikam from the KDE project.
10555 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10556 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10557 brightness distribution, so you may want to use both filters together in certain
10558 cases, though you will have to take care of ordering, i.e. whether vignetting should
10559 be applied before or after lens correction.
10561 @subsection Options
10563 The filter accepts the following options:
10567 Relative x-coordinate of the focal point of the image, and thereby the center of the
10568 distortion. This value has a range [0,1] and is expressed as fractions of the image
10569 width. Default is 0.5.
10571 Relative y-coordinate of the focal point of the image, and thereby the center of the
10572 distortion. This value has a range [0,1] and is expressed as fractions of the image
10573 height. Default is 0.5.
10575 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
10576 no correction. Default is 0.
10578 Coefficient of the double quadratic correction term. This value has a range [-1,1].
10579 0 means no correction. Default is 0.
10582 The formula that generates the correction is:
10584 @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)
10586 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10587 distances from the focal point in the source and target images, respectively.
10591 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10592 score between two input videos.
10594 The obtained VMAF score is printed through the logging system.
10596 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10597 After installing the library it can be enabled using:
10598 @code{./configure --enable-libvmaf}.
10599 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10601 The filter has following options:
10605 Set the model path which is to be used for SVM.
10606 Default value: @code{"vmaf_v0.6.1.pkl"}
10609 Set the file path to be used to store logs.
10612 Set the format of the log file (xml or json).
10614 @item enable_transform
10615 Enables transform for computing vmaf.
10618 Invokes the phone model which will generate VMAF scores higher than in the
10619 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10622 Enables computing psnr along with vmaf.
10625 Enables computing ssim along with vmaf.
10628 Enables computing ms_ssim along with vmaf.
10631 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10634 This filter also supports the @ref{framesync} options.
10636 On the below examples the input file @file{main.mpg} being processed is
10637 compared with the reference file @file{ref.mpg}.
10640 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10643 Example with options:
10645 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10650 Limits the pixel components values to the specified range [min, max].
10652 The filter accepts the following options:
10656 Lower bound. Defaults to the lowest allowed value for the input.
10659 Upper bound. Defaults to the highest allowed value for the input.
10662 Specify which planes will be processed. Defaults to all available.
10669 The filter accepts the following options:
10673 Set the number of loops. Setting this value to -1 will result in infinite loops.
10677 Set maximal size in number of frames. Default is 0.
10680 Set first frame of loop. Default is 0.
10686 Apply a 3D LUT to an input video.
10688 The filter accepts the following options:
10692 Set the 3D LUT file name.
10694 Currently supported formats:
10706 Select interpolation mode.
10708 Available values are:
10712 Use values from the nearest defined point.
10714 Interpolate values using the 8 points defining a cube.
10716 Interpolate values using a tetrahedron.
10720 This filter also supports the @ref{framesync} options.
10724 Turn certain luma values into transparency.
10726 The filter accepts the following options:
10730 Set the luma which will be used as base for transparency.
10731 Default value is @code{0}.
10734 Set the range of luma values to be keyed out.
10735 Default value is @code{0}.
10738 Set the range of softness. Default value is @code{0}.
10739 Use this to control gradual transition from zero to full transparency.
10742 @section lut, lutrgb, lutyuv
10744 Compute a look-up table for binding each pixel component input value
10745 to an output value, and apply it to the input video.
10747 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10748 to an RGB input video.
10750 These filters accept the following parameters:
10753 set first pixel component expression
10755 set second pixel component expression
10757 set third pixel component expression
10759 set fourth pixel component expression, corresponds to the alpha component
10762 set red component expression
10764 set green component expression
10766 set blue component expression
10768 alpha component expression
10771 set Y/luminance component expression
10773 set U/Cb component expression
10775 set V/Cr component expression
10778 Each of them specifies the expression to use for computing the lookup table for
10779 the corresponding pixel component values.
10781 The exact component associated to each of the @var{c*} options depends on the
10784 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10785 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10787 The expressions can contain the following constants and functions:
10792 The input width and height.
10795 The input value for the pixel component.
10798 The input value, clipped to the @var{minval}-@var{maxval} range.
10801 The maximum value for the pixel component.
10804 The minimum value for the pixel component.
10807 The negated value for the pixel component value, clipped to the
10808 @var{minval}-@var{maxval} range; it corresponds to the expression
10809 "maxval-clipval+minval".
10812 The computed value in @var{val}, clipped to the
10813 @var{minval}-@var{maxval} range.
10815 @item gammaval(gamma)
10816 The computed gamma correction value of the pixel component value,
10817 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10819 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10823 All expressions default to "val".
10825 @subsection Examples
10829 Negate input video:
10831 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10832 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10835 The above is the same as:
10837 lutrgb="r=negval:g=negval:b=negval"
10838 lutyuv="y=negval:u=negval:v=negval"
10848 Remove chroma components, turning the video into a graytone image:
10850 lutyuv="u=128:v=128"
10854 Apply a luma burning effect:
10860 Remove green and blue components:
10866 Set a constant alpha channel value on input:
10868 format=rgba,lutrgb=a="maxval-minval/2"
10872 Correct luminance gamma by a factor of 0.5:
10874 lutyuv=y=gammaval(0.5)
10878 Discard least significant bits of luma:
10880 lutyuv=y='bitand(val, 128+64+32)'
10884 Technicolor like effect:
10886 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10890 @section lut2, tlut2
10892 The @code{lut2} filter takes two input streams and outputs one
10895 The @code{tlut2} (time lut2) filter takes two consecutive frames
10896 from one single stream.
10898 This filter accepts the following parameters:
10901 set first pixel component expression
10903 set second pixel component expression
10905 set third pixel component expression
10907 set fourth pixel component expression, corresponds to the alpha component
10910 Each of them specifies the expression to use for computing the lookup table for
10911 the corresponding pixel component values.
10913 The exact component associated to each of the @var{c*} options depends on the
10916 The expressions can contain the following constants:
10921 The input width and height.
10924 The first input value for the pixel component.
10927 The second input value for the pixel component.
10930 The first input video bit depth.
10933 The second input video bit depth.
10936 All expressions default to "x".
10938 @subsection Examples
10942 Highlight differences between two RGB video streams:
10944 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)'
10948 Highlight differences between two YUV video streams:
10950 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)'
10954 Show max difference between two video streams:
10956 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)))'
10960 @section maskedclamp
10962 Clamp the first input stream with the second input and third input stream.
10964 Returns the value of first stream to be between second input
10965 stream - @code{undershoot} and third input stream + @code{overshoot}.
10967 This filter accepts the following options:
10970 Default value is @code{0}.
10973 Default value is @code{0}.
10976 Set which planes will be processed as bitmap, unprocessed planes will be
10977 copied from first stream.
10978 By default value 0xf, all planes will be processed.
10981 @section maskedmerge
10983 Merge the first input stream with the second input stream using per pixel
10984 weights in the third input stream.
10986 A value of 0 in the third stream pixel component means that pixel component
10987 from first stream is returned unchanged, while maximum value (eg. 255 for
10988 8-bit videos) means that pixel component from second stream is returned
10989 unchanged. Intermediate values define the amount of merging between both
10990 input stream's pixel components.
10992 This filter accepts the following options:
10995 Set which planes will be processed as bitmap, unprocessed planes will be
10996 copied from first stream.
10997 By default value 0xf, all planes will be processed.
11002 Apply motion-compensation deinterlacing.
11004 It needs one field per frame as input and must thus be used together
11005 with yadif=1/3 or equivalent.
11007 This filter accepts the following options:
11010 Set the deinterlacing mode.
11012 It accepts one of the following values:
11017 use iterative motion estimation
11019 like @samp{slow}, but use multiple reference frames.
11021 Default value is @samp{fast}.
11024 Set the picture field parity assumed for the input video. It must be
11025 one of the following values:
11029 assume top field first
11031 assume bottom field first
11034 Default value is @samp{bff}.
11037 Set per-block quantization parameter (QP) used by the internal
11040 Higher values should result in a smoother motion vector field but less
11041 optimal individual vectors. Default value is 1.
11044 @section mergeplanes
11046 Merge color channel components from several video streams.
11048 The filter accepts up to 4 input streams, and merge selected input
11049 planes to the output video.
11051 This filter accepts the following options:
11054 Set input to output plane mapping. Default is @code{0}.
11056 The mappings is specified as a bitmap. It should be specified as a
11057 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
11058 mapping for the first plane of the output stream. 'A' sets the number of
11059 the input stream to use (from 0 to 3), and 'a' the plane number of the
11060 corresponding input to use (from 0 to 3). The rest of the mappings is
11061 similar, 'Bb' describes the mapping for the output stream second
11062 plane, 'Cc' describes the mapping for the output stream third plane and
11063 'Dd' describes the mapping for the output stream fourth plane.
11066 Set output pixel format. Default is @code{yuva444p}.
11069 @subsection Examples
11073 Merge three gray video streams of same width and height into single video stream:
11075 [a0][a1][a2]mergeplanes=0x001020:yuv444p
11079 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
11081 [a0][a1]mergeplanes=0x00010210:yuva444p
11085 Swap Y and A plane in yuva444p stream:
11087 format=yuva444p,mergeplanes=0x03010200:yuva444p
11091 Swap U and V plane in yuv420p stream:
11093 format=yuv420p,mergeplanes=0x000201:yuv420p
11097 Cast a rgb24 clip to yuv444p:
11099 format=rgb24,mergeplanes=0x000102:yuv444p
11105 Estimate and export motion vectors using block matching algorithms.
11106 Motion vectors are stored in frame side data to be used by other filters.
11108 This filter accepts the following options:
11111 Specify the motion estimation method. Accepts one of the following values:
11115 Exhaustive search algorithm.
11117 Three step search algorithm.
11119 Two dimensional logarithmic search algorithm.
11121 New three step search algorithm.
11123 Four step search algorithm.
11125 Diamond search algorithm.
11127 Hexagon-based search algorithm.
11129 Enhanced predictive zonal search algorithm.
11131 Uneven multi-hexagon search algorithm.
11133 Default value is @samp{esa}.
11136 Macroblock size. Default @code{16}.
11139 Search parameter. Default @code{7}.
11142 @section midequalizer
11144 Apply Midway Image Equalization effect using two video streams.
11146 Midway Image Equalization adjusts a pair of images to have the same
11147 histogram, while maintaining their dynamics as much as possible. It's
11148 useful for e.g. matching exposures from a pair of stereo cameras.
11150 This filter has two inputs and one output, which must be of same pixel format, but
11151 may be of different sizes. The output of filter is first input adjusted with
11152 midway histogram of both inputs.
11154 This filter accepts the following option:
11158 Set which planes to process. Default is @code{15}, which is all available planes.
11161 @section minterpolate
11163 Convert the video to specified frame rate using motion interpolation.
11165 This filter accepts the following options:
11168 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}.
11171 Motion interpolation mode. Following values are accepted:
11174 Duplicate previous or next frame for interpolating new ones.
11176 Blend source frames. Interpolated frame is mean of previous and next frames.
11178 Motion compensated interpolation. Following options are effective when this mode is selected:
11182 Motion compensation mode. Following values are accepted:
11185 Overlapped block motion compensation.
11187 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
11189 Default mode is @samp{obmc}.
11192 Motion estimation mode. Following values are accepted:
11195 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
11197 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11199 Default mode is @samp{bilat}.
11202 The algorithm to be used for motion estimation. Following values are accepted:
11205 Exhaustive search algorithm.
11207 Three step search algorithm.
11209 Two dimensional logarithmic search algorithm.
11211 New three step search algorithm.
11213 Four step search algorithm.
11215 Diamond search algorithm.
11217 Hexagon-based search algorithm.
11219 Enhanced predictive zonal search algorithm.
11221 Uneven multi-hexagon search algorithm.
11223 Default algorithm is @samp{epzs}.
11226 Macroblock size. Default @code{16}.
11229 Motion estimation search parameter. Default @code{32}.
11232 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).
11237 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:
11240 Disable scene change detection.
11242 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11244 Default method is @samp{fdiff}.
11246 @item scd_threshold
11247 Scene change detection threshold. Default is @code{5.0}.
11252 Mix several video input streams into one video stream.
11254 A description of the accepted options follows.
11258 The number of inputs. If unspecified, it defaults to 2.
11261 Specify weight of each input video stream as sequence.
11262 Each weight is separated by space. If number of weights
11263 is smaller than number of @var{frames} last specified
11264 weight will be used for all remaining unset weights.
11267 Specify scale, if it is set it will be multiplied with sum
11268 of each weight multiplied with pixel values to give final destination
11269 pixel value. By default @var{scale} is auto scaled to sum of weights.
11272 Specify how end of stream is determined.
11275 The duration of the longest input. (default)
11278 The duration of the shortest input.
11281 The duration of the first input.
11285 @section mpdecimate
11287 Drop frames that do not differ greatly from the previous frame in
11288 order to reduce frame rate.
11290 The main use of this filter is for very-low-bitrate encoding
11291 (e.g. streaming over dialup modem), but it could in theory be used for
11292 fixing movies that were inverse-telecined incorrectly.
11294 A description of the accepted options follows.
11298 Set the maximum number of consecutive frames which can be dropped (if
11299 positive), or the minimum interval between dropped frames (if
11300 negative). If the value is 0, the frame is dropped disregarding the
11301 number of previous sequentially dropped frames.
11303 Default value is 0.
11308 Set the dropping threshold values.
11310 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11311 represent actual pixel value differences, so a threshold of 64
11312 corresponds to 1 unit of difference for each pixel, or the same spread
11313 out differently over the block.
11315 A frame is a candidate for dropping if no 8x8 blocks differ by more
11316 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11317 meaning the whole image) differ by more than a threshold of @option{lo}.
11319 Default value for @option{hi} is 64*12, default value for @option{lo} is
11320 64*5, and default value for @option{frac} is 0.33.
11326 Negate input video.
11328 It accepts an integer in input; if non-zero it negates the
11329 alpha component (if available). The default value in input is 0.
11333 Denoise frames using Non-Local Means algorithm.
11335 Each pixel is adjusted by looking for other pixels with similar contexts. This
11336 context similarity is defined by comparing their surrounding patches of size
11337 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11340 Note that the research area defines centers for patches, which means some
11341 patches will be made of pixels outside that research area.
11343 The filter accepts the following options.
11347 Set denoising strength.
11353 Same as @option{p} but for chroma planes.
11355 The default value is @var{0} and means automatic.
11361 Same as @option{r} but for chroma planes.
11363 The default value is @var{0} and means automatic.
11368 Deinterlace video using neural network edge directed interpolation.
11370 This filter accepts the following options:
11374 Mandatory option, without binary file filter can not work.
11375 Currently file can be found here:
11376 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11379 Set which frames to deinterlace, by default it is @code{all}.
11380 Can be @code{all} or @code{interlaced}.
11383 Set mode of operation.
11385 Can be one of the following:
11389 Use frame flags, both fields.
11391 Use frame flags, single field.
11393 Use top field only.
11395 Use bottom field only.
11397 Use both fields, top first.
11399 Use both fields, bottom first.
11403 Set which planes to process, by default filter process all frames.
11406 Set size of local neighborhood around each pixel, used by the predictor neural
11409 Can be one of the following:
11422 Set the number of neurons in predictor neural network.
11423 Can be one of the following:
11434 Controls the number of different neural network predictions that are blended
11435 together to compute the final output value. Can be @code{fast}, default or
11439 Set which set of weights to use in the predictor.
11440 Can be one of the following:
11444 weights trained to minimize absolute error
11446 weights trained to minimize squared error
11450 Controls whether or not the prescreener neural network is used to decide
11451 which pixels should be processed by the predictor neural network and which
11452 can be handled by simple cubic interpolation.
11453 The prescreener is trained to know whether cubic interpolation will be
11454 sufficient for a pixel or whether it should be predicted by the predictor nn.
11455 The computational complexity of the prescreener nn is much less than that of
11456 the predictor nn. Since most pixels can be handled by cubic interpolation,
11457 using the prescreener generally results in much faster processing.
11458 The prescreener is pretty accurate, so the difference between using it and not
11459 using it is almost always unnoticeable.
11461 Can be one of the following:
11469 Default is @code{new}.
11472 Set various debugging flags.
11477 Force libavfilter not to use any of the specified pixel formats for the
11478 input to the next filter.
11480 It accepts the following parameters:
11484 A '|'-separated list of pixel format names, such as
11485 pix_fmts=yuv420p|monow|rgb24".
11489 @subsection Examples
11493 Force libavfilter to use a format different from @var{yuv420p} for the
11494 input to the vflip filter:
11496 noformat=pix_fmts=yuv420p,vflip
11500 Convert the input video to any of the formats not contained in the list:
11502 noformat=yuv420p|yuv444p|yuv410p
11508 Add noise on video input frame.
11510 The filter accepts the following options:
11518 Set noise seed for specific pixel component or all pixel components in case
11519 of @var{all_seed}. Default value is @code{123457}.
11521 @item all_strength, alls
11522 @item c0_strength, c0s
11523 @item c1_strength, c1s
11524 @item c2_strength, c2s
11525 @item c3_strength, c3s
11526 Set noise strength for specific pixel component or all pixel components in case
11527 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11529 @item all_flags, allf
11530 @item c0_flags, c0f
11531 @item c1_flags, c1f
11532 @item c2_flags, c2f
11533 @item c3_flags, c3f
11534 Set pixel component flags or set flags for all components if @var{all_flags}.
11535 Available values for component flags are:
11538 averaged temporal noise (smoother)
11540 mix random noise with a (semi)regular pattern
11542 temporal noise (noise pattern changes between frames)
11544 uniform noise (gaussian otherwise)
11548 @subsection Examples
11550 Add temporal and uniform noise to input video:
11552 noise=alls=20:allf=t+u
11557 Normalize RGB video (aka histogram stretching, contrast stretching).
11558 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11560 For each channel of each frame, the filter computes the input range and maps
11561 it linearly to the user-specified output range. The output range defaults
11562 to the full dynamic range from pure black to pure white.
11564 Temporal smoothing can be used on the input range to reduce flickering (rapid
11565 changes in brightness) caused when small dark or bright objects enter or leave
11566 the scene. This is similar to the auto-exposure (automatic gain control) on a
11567 video camera, and, like a video camera, it may cause a period of over- or
11568 under-exposure of the video.
11570 The R,G,B channels can be normalized independently, which may cause some
11571 color shifting, or linked together as a single channel, which prevents
11572 color shifting. Linked normalization preserves hue. Independent normalization
11573 does not, so it can be used to remove some color casts. Independent and linked
11574 normalization can be combined in any ratio.
11576 The normalize filter accepts the following options:
11581 Colors which define the output range. The minimum input value is mapped to
11582 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11583 The defaults are black and white respectively. Specifying white for
11584 @var{blackpt} and black for @var{whitept} will give color-inverted,
11585 normalized video. Shades of grey can be used to reduce the dynamic range
11586 (contrast). Specifying saturated colors here can create some interesting
11590 The number of previous frames to use for temporal smoothing. The input range
11591 of each channel is smoothed using a rolling average over the current frame
11592 and the @var{smoothing} previous frames. The default is 0 (no temporal
11596 Controls the ratio of independent (color shifting) channel normalization to
11597 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11598 independent. Defaults to 1.0 (fully independent).
11601 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11602 expensive no-op. Defaults to 1.0 (full strength).
11606 @subsection Examples
11608 Stretch video contrast to use the full dynamic range, with no temporal
11609 smoothing; may flicker depending on the source content:
11611 normalize=blackpt=black:whitept=white:smoothing=0
11614 As above, but with 50 frames of temporal smoothing; flicker should be
11615 reduced, depending on the source content:
11617 normalize=blackpt=black:whitept=white:smoothing=50
11620 As above, but with hue-preserving linked channel normalization:
11622 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11625 As above, but with half strength:
11627 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11630 Map the darkest input color to red, the brightest input color to cyan:
11632 normalize=blackpt=red:whitept=cyan
11637 Pass the video source unchanged to the output.
11640 Optical Character Recognition
11642 This filter uses Tesseract for optical character recognition.
11644 It accepts the following options:
11648 Set datapath to tesseract data. Default is to use whatever was
11649 set at installation.
11652 Set language, default is "eng".
11655 Set character whitelist.
11658 Set character blacklist.
11661 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11665 Apply a video transform using libopencv.
11667 To enable this filter, install the libopencv library and headers and
11668 configure FFmpeg with @code{--enable-libopencv}.
11670 It accepts the following parameters:
11675 The name of the libopencv filter to apply.
11677 @item filter_params
11678 The parameters to pass to the libopencv filter. If not specified, the default
11679 values are assumed.
11683 Refer to the official libopencv documentation for more precise
11685 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11687 Several libopencv filters are supported; see the following subsections.
11692 Dilate an image by using a specific structuring element.
11693 It corresponds to the libopencv function @code{cvDilate}.
11695 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11697 @var{struct_el} represents a structuring element, and has the syntax:
11698 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11700 @var{cols} and @var{rows} represent the number of columns and rows of
11701 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11702 point, and @var{shape} the shape for the structuring element. @var{shape}
11703 must be "rect", "cross", "ellipse", or "custom".
11705 If the value for @var{shape} is "custom", it must be followed by a
11706 string of the form "=@var{filename}". The file with name
11707 @var{filename} is assumed to represent a binary image, with each
11708 printable character corresponding to a bright pixel. When a custom
11709 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11710 or columns and rows of the read file are assumed instead.
11712 The default value for @var{struct_el} is "3x3+0x0/rect".
11714 @var{nb_iterations} specifies the number of times the transform is
11715 applied to the image, and defaults to 1.
11719 # Use the default values
11722 # Dilate using a structuring element with a 5x5 cross, iterating two times
11723 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11725 # Read the shape from the file diamond.shape, iterating two times.
11726 # The file diamond.shape may contain a pattern of characters like this
11732 # The specified columns and rows are ignored
11733 # but the anchor point coordinates are not
11734 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11739 Erode an image by using a specific structuring element.
11740 It corresponds to the libopencv function @code{cvErode}.
11742 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11743 with the same syntax and semantics as the @ref{dilate} filter.
11747 Smooth the input video.
11749 The filter takes the following parameters:
11750 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11752 @var{type} is the type of smooth filter to apply, and must be one of
11753 the following values: "blur", "blur_no_scale", "median", "gaussian",
11754 or "bilateral". The default value is "gaussian".
11756 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11757 depend on the smooth type. @var{param1} and
11758 @var{param2} accept integer positive values or 0. @var{param3} and
11759 @var{param4} accept floating point values.
11761 The default value for @var{param1} is 3. The default value for the
11762 other parameters is 0.
11764 These parameters correspond to the parameters assigned to the
11765 libopencv function @code{cvSmooth}.
11767 @section oscilloscope
11769 2D Video Oscilloscope.
11771 Useful to measure spatial impulse, step responses, chroma delays, etc.
11773 It accepts the following parameters:
11777 Set scope center x position.
11780 Set scope center y position.
11783 Set scope size, relative to frame diagonal.
11786 Set scope tilt/rotation.
11792 Set trace center x position.
11795 Set trace center y position.
11798 Set trace width, relative to width of frame.
11801 Set trace height, relative to height of frame.
11804 Set which components to trace. By default it traces first three components.
11807 Draw trace grid. By default is enabled.
11810 Draw some statistics. By default is enabled.
11813 Draw scope. By default is enabled.
11816 @subsection Examples
11820 Inspect full first row of video frame.
11822 oscilloscope=x=0.5:y=0:s=1
11826 Inspect full last row of video frame.
11828 oscilloscope=x=0.5:y=1:s=1
11832 Inspect full 5th line of video frame of height 1080.
11834 oscilloscope=x=0.5:y=5/1080:s=1
11838 Inspect full last column of video frame.
11840 oscilloscope=x=1:y=0.5:s=1:t=1
11848 Overlay one video on top of another.
11850 It takes two inputs and has one output. The first input is the "main"
11851 video on which the second input is overlaid.
11853 It accepts the following parameters:
11855 A description of the accepted options follows.
11860 Set the expression for the x and y coordinates of the overlaid video
11861 on the main video. Default value is "0" for both expressions. In case
11862 the expression is invalid, it is set to a huge value (meaning that the
11863 overlay will not be displayed within the output visible area).
11866 See @ref{framesync}.
11869 Set when the expressions for @option{x}, and @option{y} are evaluated.
11871 It accepts the following values:
11874 only evaluate expressions once during the filter initialization or
11875 when a command is processed
11878 evaluate expressions for each incoming frame
11881 Default value is @samp{frame}.
11884 See @ref{framesync}.
11887 Set the format for the output video.
11889 It accepts the following values:
11892 force YUV420 output
11895 force YUV422 output
11898 force YUV444 output
11901 force packed RGB output
11904 force planar RGB output
11907 automatically pick format
11910 Default value is @samp{yuv420}.
11913 See @ref{framesync}.
11916 Set format of alpha of the overlaid video, it can be @var{straight} or
11917 @var{premultiplied}. Default is @var{straight}.
11920 The @option{x}, and @option{y} expressions can contain the following
11926 The main input width and height.
11930 The overlay input width and height.
11934 The computed values for @var{x} and @var{y}. They are evaluated for
11939 horizontal and vertical chroma subsample values of the output
11940 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11944 the number of input frame, starting from 0
11947 the position in the file of the input frame, NAN if unknown
11950 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11954 This filter also supports the @ref{framesync} options.
11956 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11957 when evaluation is done @emph{per frame}, and will evaluate to NAN
11958 when @option{eval} is set to @samp{init}.
11960 Be aware that frames are taken from each input video in timestamp
11961 order, hence, if their initial timestamps differ, it is a good idea
11962 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11963 have them begin in the same zero timestamp, as the example for
11964 the @var{movie} filter does.
11966 You can chain together more overlays but you should test the
11967 efficiency of such approach.
11969 @subsection Commands
11971 This filter supports the following commands:
11975 Modify the x and y of the overlay input.
11976 The command accepts the same syntax of the corresponding option.
11978 If the specified expression is not valid, it is kept at its current
11982 @subsection Examples
11986 Draw the overlay at 10 pixels from the bottom right corner of the main
11989 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11992 Using named options the example above becomes:
11994 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11998 Insert a transparent PNG logo in the bottom left corner of the input,
11999 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12001 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12005 Insert 2 different transparent PNG logos (second logo on bottom
12006 right corner) using the @command{ffmpeg} tool:
12008 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
12012 Add a transparent color layer on top of the main video; @code{WxH}
12013 must specify the size of the main input to the overlay filter:
12015 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
12019 Play an original video and a filtered version (here with the deshake
12020 filter) side by side using the @command{ffplay} tool:
12022 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
12025 The above command is the same as:
12027 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
12031 Make a sliding overlay appearing from the left to the right top part of the
12032 screen starting since time 2:
12034 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
12038 Compose output by putting two input videos side to side:
12040 ffmpeg -i left.avi -i right.avi -filter_complex "
12041 nullsrc=size=200x100 [background];
12042 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
12043 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
12044 [background][left] overlay=shortest=1 [background+left];
12045 [background+left][right] overlay=shortest=1:x=100 [left+right]
12050 Mask 10-20 seconds of a video by applying the delogo filter to a section
12052 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
12053 -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]'
12058 Chain several overlays in cascade:
12060 nullsrc=s=200x200 [bg];
12061 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
12062 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
12063 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
12064 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
12065 [in3] null, [mid2] overlay=100:100 [out0]
12072 Apply Overcomplete Wavelet denoiser.
12074 The filter accepts the following options:
12080 Larger depth values will denoise lower frequency components more, but
12081 slow down filtering.
12083 Must be an int in the range 8-16, default is @code{8}.
12085 @item luma_strength, ls
12088 Must be a double value in the range 0-1000, default is @code{1.0}.
12090 @item chroma_strength, cs
12091 Set chroma strength.
12093 Must be a double value in the range 0-1000, default is @code{1.0}.
12099 Add paddings to the input image, and place the original input at the
12100 provided @var{x}, @var{y} coordinates.
12102 It accepts the following parameters:
12107 Specify an expression for the size of the output image with the
12108 paddings added. If the value for @var{width} or @var{height} is 0, the
12109 corresponding input size is used for the output.
12111 The @var{width} expression can reference the value set by the
12112 @var{height} expression, and vice versa.
12114 The default value of @var{width} and @var{height} is 0.
12118 Specify the offsets to place the input image at within the padded area,
12119 with respect to the top/left border of the output image.
12121 The @var{x} expression can reference the value set by the @var{y}
12122 expression, and vice versa.
12124 The default value of @var{x} and @var{y} is 0.
12126 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
12127 so the input image is centered on the padded area.
12130 Specify the color of the padded area. For the syntax of this option,
12131 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
12132 manual,ffmpeg-utils}.
12134 The default value of @var{color} is "black".
12137 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
12139 It accepts the following values:
12143 Only evaluate expressions once during the filter initialization or when
12144 a command is processed.
12147 Evaluate expressions for each incoming frame.
12151 Default value is @samp{init}.
12154 Pad to aspect instead to a resolution.
12158 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
12159 options are expressions containing the following constants:
12164 The input video width and height.
12168 These are the same as @var{in_w} and @var{in_h}.
12172 The output width and height (the size of the padded area), as
12173 specified by the @var{width} and @var{height} expressions.
12177 These are the same as @var{out_w} and @var{out_h}.
12181 The x and y offsets as specified by the @var{x} and @var{y}
12182 expressions, or NAN if not yet specified.
12185 same as @var{iw} / @var{ih}
12188 input sample aspect ratio
12191 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
12195 The horizontal and vertical chroma subsample values. For example for the
12196 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12199 @subsection Examples
12203 Add paddings with the color "violet" to the input video. The output video
12204 size is 640x480, and the top-left corner of the input video is placed at
12207 pad=640:480:0:40:violet
12210 The example above is equivalent to the following command:
12212 pad=width=640:height=480:x=0:y=40:color=violet
12216 Pad the input to get an output with dimensions increased by 3/2,
12217 and put the input video at the center of the padded area:
12219 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12223 Pad the input to get a squared output with size equal to the maximum
12224 value between the input width and height, and put the input video at
12225 the center of the padded area:
12227 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12231 Pad the input to get a final w/h ratio of 16:9:
12233 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12237 In case of anamorphic video, in order to set the output display aspect
12238 correctly, it is necessary to use @var{sar} in the expression,
12239 according to the relation:
12241 (ih * X / ih) * sar = output_dar
12242 X = output_dar / sar
12245 Thus the previous example needs to be modified to:
12247 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12251 Double the output size and put the input video in the bottom-right
12252 corner of the output padded area:
12254 pad="2*iw:2*ih:ow-iw:oh-ih"
12258 @anchor{palettegen}
12259 @section palettegen
12261 Generate one palette for a whole video stream.
12263 It accepts the following options:
12267 Set the maximum number of colors to quantize in the palette.
12268 Note: the palette will still contain 256 colors; the unused palette entries
12271 @item reserve_transparent
12272 Create a palette of 255 colors maximum and reserve the last one for
12273 transparency. Reserving the transparency color is useful for GIF optimization.
12274 If not set, the maximum of colors in the palette will be 256. You probably want
12275 to disable this option for a standalone image.
12278 @item transparency_color
12279 Set the color that will be used as background for transparency.
12282 Set statistics mode.
12284 It accepts the following values:
12287 Compute full frame histograms.
12289 Compute histograms only for the part that differs from previous frame. This
12290 might be relevant to give more importance to the moving part of your input if
12291 the background is static.
12293 Compute new histogram for each frame.
12296 Default value is @var{full}.
12299 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12300 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12301 color quantization of the palette. This information is also visible at
12302 @var{info} logging level.
12304 @subsection Examples
12308 Generate a representative palette of a given video using @command{ffmpeg}:
12310 ffmpeg -i input.mkv -vf palettegen palette.png
12314 @section paletteuse
12316 Use a palette to downsample an input video stream.
12318 The filter takes two inputs: one video stream and a palette. The palette must
12319 be a 256 pixels image.
12321 It accepts the following options:
12325 Select dithering mode. Available algorithms are:
12328 Ordered 8x8 bayer dithering (deterministic)
12330 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12331 Note: this dithering is sometimes considered "wrong" and is included as a
12333 @item floyd_steinberg
12334 Floyd and Steingberg dithering (error diffusion)
12336 Frankie Sierra dithering v2 (error diffusion)
12338 Frankie Sierra dithering v2 "Lite" (error diffusion)
12341 Default is @var{sierra2_4a}.
12344 When @var{bayer} dithering is selected, this option defines the scale of the
12345 pattern (how much the crosshatch pattern is visible). A low value means more
12346 visible pattern for less banding, and higher value means less visible pattern
12347 at the cost of more banding.
12349 The option must be an integer value in the range [0,5]. Default is @var{2}.
12352 If set, define the zone to process
12356 Only the changing rectangle will be reprocessed. This is similar to GIF
12357 cropping/offsetting compression mechanism. This option can be useful for speed
12358 if only a part of the image is changing, and has use cases such as limiting the
12359 scope of the error diffusal @option{dither} to the rectangle that bounds the
12360 moving scene (it leads to more deterministic output if the scene doesn't change
12361 much, and as a result less moving noise and better GIF compression).
12364 Default is @var{none}.
12367 Take new palette for each output frame.
12369 @item alpha_threshold
12370 Sets the alpha threshold for transparency. Alpha values above this threshold
12371 will be treated as completely opaque, and values below this threshold will be
12372 treated as completely transparent.
12374 The option must be an integer value in the range [0,255]. Default is @var{128}.
12377 @subsection Examples
12381 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12382 using @command{ffmpeg}:
12384 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12388 @section perspective
12390 Correct perspective of video not recorded perpendicular to the screen.
12392 A description of the accepted parameters follows.
12403 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12404 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12405 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12406 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12407 then the corners of the source will be sent to the specified coordinates.
12409 The expressions can use the following variables:
12414 the width and height of video frame.
12418 Output frame count.
12421 @item interpolation
12422 Set interpolation for perspective correction.
12424 It accepts the following values:
12430 Default value is @samp{linear}.
12433 Set interpretation of coordinate options.
12435 It accepts the following values:
12439 Send point in the source specified by the given coordinates to
12440 the corners of the destination.
12442 @item 1, destination
12444 Send the corners of the source to the point in the destination specified
12445 by the given coordinates.
12447 Default value is @samp{source}.
12451 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12453 It accepts the following values:
12456 only evaluate expressions once during the filter initialization or
12457 when a command is processed
12460 evaluate expressions for each incoming frame
12463 Default value is @samp{init}.
12468 Delay interlaced video by one field time so that the field order changes.
12470 The intended use is to fix PAL movies that have been captured with the
12471 opposite field order to the film-to-video transfer.
12473 A description of the accepted parameters follows.
12479 It accepts the following values:
12482 Capture field order top-first, transfer bottom-first.
12483 Filter will delay the bottom field.
12486 Capture field order bottom-first, transfer top-first.
12487 Filter will delay the top field.
12490 Capture and transfer with the same field order. This mode only exists
12491 for the documentation of the other options to refer to, but if you
12492 actually select it, the filter will faithfully do nothing.
12495 Capture field order determined automatically by field flags, transfer
12497 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12498 basis using field flags. If no field information is available,
12499 then this works just like @samp{u}.
12502 Capture unknown or varying, transfer opposite.
12503 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12504 analyzing the images and selecting the alternative that produces best
12505 match between the fields.
12508 Capture top-first, transfer unknown or varying.
12509 Filter selects among @samp{t} and @samp{p} using image analysis.
12512 Capture bottom-first, transfer unknown or varying.
12513 Filter selects among @samp{b} and @samp{p} using image analysis.
12516 Capture determined by field flags, transfer unknown or varying.
12517 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12518 image analysis. If no field information is available, then this works just
12519 like @samp{U}. This is the default mode.
12522 Both capture and transfer unknown or varying.
12523 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12527 @section pixdesctest
12529 Pixel format descriptor test filter, mainly useful for internal
12530 testing. The output video should be equal to the input video.
12534 format=monow, pixdesctest
12537 can be used to test the monowhite pixel format descriptor definition.
12541 Display sample values of color channels. Mainly useful for checking color
12542 and levels. Minimum supported resolution is 640x480.
12544 The filters accept the following options:
12548 Set scope X position, relative offset on X axis.
12551 Set scope Y position, relative offset on Y axis.
12560 Set window opacity. This window also holds statistics about pixel area.
12563 Set window X position, relative offset on X axis.
12566 Set window Y position, relative offset on Y axis.
12571 Enable the specified chain of postprocessing subfilters using libpostproc. This
12572 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12573 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12574 Each subfilter and some options have a short and a long name that can be used
12575 interchangeably, i.e. dr/dering are the same.
12577 The filters accept the following options:
12581 Set postprocessing subfilters string.
12584 All subfilters share common options to determine their scope:
12588 Honor the quality commands for this subfilter.
12591 Do chrominance filtering, too (default).
12594 Do luminance filtering only (no chrominance).
12597 Do chrominance filtering only (no luminance).
12600 These options can be appended after the subfilter name, separated by a '|'.
12602 Available subfilters are:
12605 @item hb/hdeblock[|difference[|flatness]]
12606 Horizontal deblocking filter
12609 Difference factor where higher values mean more deblocking (default: @code{32}).
12611 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12614 @item vb/vdeblock[|difference[|flatness]]
12615 Vertical deblocking filter
12618 Difference factor where higher values mean more deblocking (default: @code{32}).
12620 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12623 @item ha/hadeblock[|difference[|flatness]]
12624 Accurate horizontal deblocking filter
12627 Difference factor where higher values mean more deblocking (default: @code{32}).
12629 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12632 @item va/vadeblock[|difference[|flatness]]
12633 Accurate vertical deblocking filter
12636 Difference factor where higher values mean more deblocking (default: @code{32}).
12638 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12642 The horizontal and vertical deblocking filters share the difference and
12643 flatness values so you cannot set different horizontal and vertical
12647 @item h1/x1hdeblock
12648 Experimental horizontal deblocking filter
12650 @item v1/x1vdeblock
12651 Experimental vertical deblocking filter
12656 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12659 larger -> stronger filtering
12661 larger -> stronger filtering
12663 larger -> stronger filtering
12666 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12669 Stretch luminance to @code{0-255}.
12672 @item lb/linblenddeint
12673 Linear blend deinterlacing filter that deinterlaces the given block by
12674 filtering all lines with a @code{(1 2 1)} filter.
12676 @item li/linipoldeint
12677 Linear interpolating deinterlacing filter that deinterlaces the given block by
12678 linearly interpolating every second line.
12680 @item ci/cubicipoldeint
12681 Cubic interpolating deinterlacing filter deinterlaces the given block by
12682 cubically interpolating every second line.
12684 @item md/mediandeint
12685 Median deinterlacing filter that deinterlaces the given block by applying a
12686 median filter to every second line.
12688 @item fd/ffmpegdeint
12689 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12690 second line with a @code{(-1 4 2 4 -1)} filter.
12693 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12694 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12696 @item fq/forceQuant[|quantizer]
12697 Overrides the quantizer table from the input with the constant quantizer you
12705 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12708 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12711 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12714 @subsection Examples
12718 Apply horizontal and vertical deblocking, deringing and automatic
12719 brightness/contrast:
12725 Apply default filters without brightness/contrast correction:
12731 Apply default filters and temporal denoiser:
12733 pp=default/tmpnoise|1|2|3
12737 Apply deblocking on luminance only, and switch vertical deblocking on or off
12738 automatically depending on available CPU time:
12745 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12746 similar to spp = 6 with 7 point DCT, where only the center sample is
12749 The filter accepts the following options:
12753 Force a constant quantization parameter. It accepts an integer in range
12754 0 to 63. If not set, the filter will use the QP from the video stream
12758 Set thresholding mode. Available modes are:
12762 Set hard thresholding.
12764 Set soft thresholding (better de-ringing effect, but likely blurrier).
12766 Set medium thresholding (good results, default).
12770 @section premultiply
12771 Apply alpha premultiply effect to input video stream using first plane
12772 of second stream as alpha.
12774 Both streams must have same dimensions and same pixel format.
12776 The filter accepts the following option:
12780 Set which planes will be processed, unprocessed planes will be copied.
12781 By default value 0xf, all planes will be processed.
12784 Do not require 2nd input for processing, instead use alpha plane from input stream.
12788 Apply prewitt operator to input video stream.
12790 The filter accepts the following option:
12794 Set which planes will be processed, unprocessed planes will be copied.
12795 By default value 0xf, all planes will be processed.
12798 Set value which will be multiplied with filtered result.
12801 Set value which will be added to filtered result.
12804 @anchor{program_opencl}
12805 @section program_opencl
12807 Filter video using an OpenCL program.
12812 OpenCL program source file.
12815 Kernel name in program.
12818 Number of inputs to the filter. Defaults to 1.
12821 Size of output frames. Defaults to the same as the first input.
12825 The program source file must contain a kernel function with the given name,
12826 which will be run once for each plane of the output. Each run on a plane
12827 gets enqueued as a separate 2D global NDRange with one work-item for each
12828 pixel to be generated. The global ID offset for each work-item is therefore
12829 the coordinates of a pixel in the destination image.
12831 The kernel function needs to take the following arguments:
12834 Destination image, @var{__write_only image2d_t}.
12836 This image will become the output; the kernel should write all of it.
12838 Frame index, @var{unsigned int}.
12840 This is a counter starting from zero and increasing by one for each frame.
12842 Source images, @var{__read_only image2d_t}.
12844 These are the most recent images on each input. The kernel may read from
12845 them to generate the output, but they can't be written to.
12852 Copy the input to the output (output must be the same size as the input).
12854 __kernel void copy(__write_only image2d_t destination,
12855 unsigned int index,
12856 __read_only image2d_t source)
12858 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
12860 int2 location = (int2)(get_global_id(0), get_global_id(1));
12862 float4 value = read_imagef(source, sampler, location);
12864 write_imagef(destination, location, value);
12869 Apply a simple transformation, rotating the input by an amount increasing
12870 with the index counter. Pixel values are linearly interpolated by the
12871 sampler, and the output need not have the same dimensions as the input.
12873 __kernel void rotate_image(__write_only image2d_t dst,
12874 unsigned int index,
12875 __read_only image2d_t src)
12877 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12878 CLK_FILTER_LINEAR);
12880 float angle = (float)index / 100.0f;
12882 float2 dst_dim = convert_float2(get_image_dim(dst));
12883 float2 src_dim = convert_float2(get_image_dim(src));
12885 float2 dst_cen = dst_dim / 2.0f;
12886 float2 src_cen = src_dim / 2.0f;
12888 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12890 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
12892 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
12893 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
12895 src_pos = src_pos * src_dim / dst_dim;
12897 float2 src_loc = src_pos + src_cen;
12899 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
12900 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
12901 write_imagef(dst, dst_loc, 0.5f);
12903 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
12908 Blend two inputs together, with the amount of each input used varying
12909 with the index counter.
12911 __kernel void blend_images(__write_only image2d_t dst,
12912 unsigned int index,
12913 __read_only image2d_t src1,
12914 __read_only image2d_t src2)
12916 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12917 CLK_FILTER_LINEAR);
12919 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
12921 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12922 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
12923 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
12925 float4 val1 = read_imagef(src1, sampler, src1_loc);
12926 float4 val2 = read_imagef(src2, sampler, src2_loc);
12928 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
12934 @section pseudocolor
12936 Alter frame colors in video with pseudocolors.
12938 This filter accept the following options:
12942 set pixel first component expression
12945 set pixel second component expression
12948 set pixel third component expression
12951 set pixel fourth component expression, corresponds to the alpha component
12954 set component to use as base for altering colors
12957 Each of them specifies the expression to use for computing the lookup table for
12958 the corresponding pixel component values.
12960 The expressions can contain the following constants and functions:
12965 The input width and height.
12968 The input value for the pixel component.
12970 @item ymin, umin, vmin, amin
12971 The minimum allowed component value.
12973 @item ymax, umax, vmax, amax
12974 The maximum allowed component value.
12977 All expressions default to "val".
12979 @subsection Examples
12983 Change too high luma values to gradient:
12985 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'"
12991 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12992 Ratio) between two input videos.
12994 This filter takes in input two input videos, the first input is
12995 considered the "main" source and is passed unchanged to the
12996 output. The second input is used as a "reference" video for computing
12999 Both video inputs must have the same resolution and pixel format for
13000 this filter to work correctly. Also it assumes that both inputs
13001 have the same number of frames, which are compared one by one.
13003 The obtained average PSNR is printed through the logging system.
13005 The filter stores the accumulated MSE (mean squared error) of each
13006 frame, and at the end of the processing it is averaged across all frames
13007 equally, and the following formula is applied to obtain the PSNR:
13010 PSNR = 10*log10(MAX^2/MSE)
13013 Where MAX is the average of the maximum values of each component of the
13016 The description of the accepted parameters follows.
13019 @item stats_file, f
13020 If specified the filter will use the named file to save the PSNR of
13021 each individual frame. When filename equals "-" the data is sent to
13024 @item stats_version
13025 Specifies which version of the stats file format to use. Details of
13026 each format are written below.
13027 Default value is 1.
13029 @item stats_add_max
13030 Determines whether the max value is output to the stats log.
13031 Default value is 0.
13032 Requires stats_version >= 2. If this is set and stats_version < 2,
13033 the filter will return an error.
13036 This filter also supports the @ref{framesync} options.
13038 The file printed if @var{stats_file} is selected, contains a sequence of
13039 key/value pairs of the form @var{key}:@var{value} for each compared
13042 If a @var{stats_version} greater than 1 is specified, a header line precedes
13043 the list of per-frame-pair stats, with key value pairs following the frame
13044 format with the following parameters:
13047 @item psnr_log_version
13048 The version of the log file format. Will match @var{stats_version}.
13051 A comma separated list of the per-frame-pair parameters included in
13055 A description of each shown per-frame-pair parameter follows:
13059 sequential number of the input frame, starting from 1
13062 Mean Square Error pixel-by-pixel average difference of the compared
13063 frames, averaged over all the image components.
13065 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
13066 Mean Square Error pixel-by-pixel average difference of the compared
13067 frames for the component specified by the suffix.
13069 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
13070 Peak Signal to Noise ratio of the compared frames for the component
13071 specified by the suffix.
13073 @item max_avg, max_y, max_u, max_v
13074 Maximum allowed value for each channel, and average over all
13080 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13081 [main][ref] psnr="stats_file=stats.log" [out]
13084 On this example the input file being processed is compared with the
13085 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
13086 is stored in @file{stats.log}.
13091 Pulldown reversal (inverse telecine) filter, capable of handling mixed
13092 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
13095 The pullup filter is designed to take advantage of future context in making
13096 its decisions. This filter is stateless in the sense that it does not lock
13097 onto a pattern to follow, but it instead looks forward to the following
13098 fields in order to identify matches and rebuild progressive frames.
13100 To produce content with an even framerate, insert the fps filter after
13101 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
13102 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
13104 The filter accepts the following options:
13111 These options set the amount of "junk" to ignore at the left, right, top, and
13112 bottom of the image, respectively. Left and right are in units of 8 pixels,
13113 while top and bottom are in units of 2 lines.
13114 The default is 8 pixels on each side.
13117 Set the strict breaks. Setting this option to 1 will reduce the chances of
13118 filter generating an occasional mismatched frame, but it may also cause an
13119 excessive number of frames to be dropped during high motion sequences.
13120 Conversely, setting it to -1 will make filter match fields more easily.
13121 This may help processing of video where there is slight blurring between
13122 the fields, but may also cause there to be interlaced frames in the output.
13123 Default value is @code{0}.
13126 Set the metric plane to use. It accepts the following values:
13132 Use chroma blue plane.
13135 Use chroma red plane.
13138 This option may be set to use chroma plane instead of the default luma plane
13139 for doing filter's computations. This may improve accuracy on very clean
13140 source material, but more likely will decrease accuracy, especially if there
13141 is chroma noise (rainbow effect) or any grayscale video.
13142 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
13143 load and make pullup usable in realtime on slow machines.
13146 For best results (without duplicated frames in the output file) it is
13147 necessary to change the output frame rate. For example, to inverse
13148 telecine NTSC input:
13150 ffmpeg -i input -vf pullup -r 24000/1001 ...
13155 Change video quantization parameters (QP).
13157 The filter accepts the following option:
13161 Set expression for quantization parameter.
13164 The expression is evaluated through the eval API and can contain, among others,
13165 the following constants:
13169 1 if index is not 129, 0 otherwise.
13172 Sequential index starting from -129 to 128.
13175 @subsection Examples
13179 Some equation like:
13187 Flush video frames from internal cache of frames into a random order.
13188 No frame is discarded.
13189 Inspired by @ref{frei0r} nervous filter.
13193 Set size in number of frames of internal cache, in range from @code{2} to
13194 @code{512}. Default is @code{30}.
13197 Set seed for random number generator, must be an integer included between
13198 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
13199 less than @code{0}, the filter will try to use a good random seed on a
13203 @section readeia608
13205 Read closed captioning (EIA-608) information from the top lines of a video frame.
13207 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13208 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13209 with EIA-608 data (starting from 0). A description of each metadata value follows:
13212 @item lavfi.readeia608.X.cc
13213 The two bytes stored as EIA-608 data (printed in hexadecimal).
13215 @item lavfi.readeia608.X.line
13216 The number of the line on which the EIA-608 data was identified and read.
13219 This filter accepts the following options:
13223 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13226 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13229 Set minimal acceptable amplitude change for sync codes detection.
13230 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13233 Set the ratio of width reserved for sync code detection.
13234 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13237 Set the max peaks height difference for sync code detection.
13238 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13241 Set max peaks period difference for sync code detection.
13242 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13245 Set the first two max start code bits differences.
13246 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13249 Set the minimum ratio of bits height compared to 3rd start code bit.
13250 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13253 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13256 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13259 Enable checking the parity bit. In the event of a parity error, the filter will output
13260 @code{0x00} for that character. Default is false.
13263 @subsection Examples
13267 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13269 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
13275 Read vertical interval timecode (VITC) information from the top lines of a
13278 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13279 timecode value, if a valid timecode has been detected. Further metadata key
13280 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13281 timecode data has been found or not.
13283 This filter accepts the following options:
13287 Set the maximum number of lines to scan for VITC data. If the value is set to
13288 @code{-1} the full video frame is scanned. Default is @code{45}.
13291 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13292 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13295 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13296 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13299 @subsection Examples
13303 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13304 draw @code{--:--:--:--} as a placeholder:
13306 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13312 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13314 Destination pixel at position (X, Y) will be picked from source (x, y) position
13315 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13316 value for pixel will be used for destination pixel.
13318 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13319 will have Xmap/Ymap video stream dimensions.
13320 Xmap and Ymap input video streams are 16bit depth, single channel.
13322 @section removegrain
13324 The removegrain filter is a spatial denoiser for progressive video.
13328 Set mode for the first plane.
13331 Set mode for the second plane.
13334 Set mode for the third plane.
13337 Set mode for the fourth plane.
13340 Range of mode is from 0 to 24. Description of each mode follows:
13344 Leave input plane unchanged. Default.
13347 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13350 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13353 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13356 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13357 This is equivalent to a median filter.
13360 Line-sensitive clipping giving the minimal change.
13363 Line-sensitive clipping, intermediate.
13366 Line-sensitive clipping, intermediate.
13369 Line-sensitive clipping, intermediate.
13372 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13375 Replaces the target pixel with the closest neighbour.
13378 [1 2 1] horizontal and vertical kernel blur.
13384 Bob mode, interpolates top field from the line where the neighbours
13385 pixels are the closest.
13388 Bob mode, interpolates bottom field from the line where the neighbours
13389 pixels are the closest.
13392 Bob mode, interpolates top field. Same as 13 but with a more complicated
13393 interpolation formula.
13396 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13397 interpolation formula.
13400 Clips the pixel with the minimum and maximum of respectively the maximum and
13401 minimum of each pair of opposite neighbour pixels.
13404 Line-sensitive clipping using opposite neighbours whose greatest distance from
13405 the current pixel is minimal.
13408 Replaces the pixel with the average of its 8 neighbours.
13411 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13414 Clips pixels using the averages of opposite neighbour.
13417 Same as mode 21 but simpler and faster.
13420 Small edge and halo removal, but reputed useless.
13426 @section removelogo
13428 Suppress a TV station logo, using an image file to determine which
13429 pixels comprise the logo. It works by filling in the pixels that
13430 comprise the logo with neighboring pixels.
13432 The filter accepts the following options:
13436 Set the filter bitmap file, which can be any image format supported by
13437 libavformat. The width and height of the image file must match those of the
13438 video stream being processed.
13441 Pixels in the provided bitmap image with a value of zero are not
13442 considered part of the logo, non-zero pixels are considered part of
13443 the logo. If you use white (255) for the logo and black (0) for the
13444 rest, you will be safe. For making the filter bitmap, it is
13445 recommended to take a screen capture of a black frame with the logo
13446 visible, and then using a threshold filter followed by the erode
13447 filter once or twice.
13449 If needed, little splotches can be fixed manually. Remember that if
13450 logo pixels are not covered, the filter quality will be much
13451 reduced. Marking too many pixels as part of the logo does not hurt as
13452 much, but it will increase the amount of blurring needed to cover over
13453 the image and will destroy more information than necessary, and extra
13454 pixels will slow things down on a large logo.
13456 @section repeatfields
13458 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13459 fields based on its value.
13463 Reverse a video clip.
13465 Warning: This filter requires memory to buffer the entire clip, so trimming
13468 @subsection Examples
13472 Take the first 5 seconds of a clip, and reverse it.
13479 Apply roberts cross operator to input video stream.
13481 The filter accepts the following option:
13485 Set which planes will be processed, unprocessed planes will be copied.
13486 By default value 0xf, all planes will be processed.
13489 Set value which will be multiplied with filtered result.
13492 Set value which will be added to filtered result.
13497 Rotate video by an arbitrary angle expressed in radians.
13499 The filter accepts the following options:
13501 A description of the optional parameters follows.
13504 Set an expression for the angle by which to rotate the input video
13505 clockwise, expressed as a number of radians. A negative value will
13506 result in a counter-clockwise rotation. By default it is set to "0".
13508 This expression is evaluated for each frame.
13511 Set the output width expression, default value is "iw".
13512 This expression is evaluated just once during configuration.
13515 Set the output height expression, default value is "ih".
13516 This expression is evaluated just once during configuration.
13519 Enable bilinear interpolation if set to 1, a value of 0 disables
13520 it. Default value is 1.
13523 Set the color used to fill the output area not covered by the rotated
13524 image. For the general syntax of this option, check the
13525 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13526 If the special value "none" is selected then no
13527 background is printed (useful for example if the background is never shown).
13529 Default value is "black".
13532 The expressions for the angle and the output size can contain the
13533 following constants and functions:
13537 sequential number of the input frame, starting from 0. It is always NAN
13538 before the first frame is filtered.
13541 time in seconds of the input frame, it is set to 0 when the filter is
13542 configured. It is always NAN before the first frame is filtered.
13546 horizontal and vertical chroma subsample values. For example for the
13547 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13551 the input video width and height
13555 the output width and height, that is the size of the padded area as
13556 specified by the @var{width} and @var{height} expressions
13560 the minimal width/height required for completely containing the input
13561 video rotated by @var{a} radians.
13563 These are only available when computing the @option{out_w} and
13564 @option{out_h} expressions.
13567 @subsection Examples
13571 Rotate the input by PI/6 radians clockwise:
13577 Rotate the input by PI/6 radians counter-clockwise:
13583 Rotate the input by 45 degrees clockwise:
13589 Apply a constant rotation with period T, starting from an angle of PI/3:
13591 rotate=PI/3+2*PI*t/T
13595 Make the input video rotation oscillating with a period of T
13596 seconds and an amplitude of A radians:
13598 rotate=A*sin(2*PI/T*t)
13602 Rotate the video, output size is chosen so that the whole rotating
13603 input video is always completely contained in the output:
13605 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13609 Rotate the video, reduce the output size so that no background is ever
13612 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13616 @subsection Commands
13618 The filter supports the following commands:
13622 Set the angle expression.
13623 The command accepts the same syntax of the corresponding option.
13625 If the specified expression is not valid, it is kept at its current
13631 Apply Shape Adaptive Blur.
13633 The filter accepts the following options:
13636 @item luma_radius, lr
13637 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13638 value is 1.0. A greater value will result in a more blurred image, and
13639 in slower processing.
13641 @item luma_pre_filter_radius, lpfr
13642 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13645 @item luma_strength, ls
13646 Set luma maximum difference between pixels to still be considered, must
13647 be a value in the 0.1-100.0 range, default value is 1.0.
13649 @item chroma_radius, cr
13650 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13651 greater value will result in a more blurred image, and in slower
13654 @item chroma_pre_filter_radius, cpfr
13655 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13657 @item chroma_strength, cs
13658 Set chroma maximum difference between pixels to still be considered,
13659 must be a value in the -0.9-100.0 range.
13662 Each chroma option value, if not explicitly specified, is set to the
13663 corresponding luma option value.
13668 Scale (resize) the input video, using the libswscale library.
13670 The scale filter forces the output display aspect ratio to be the same
13671 of the input, by changing the output sample aspect ratio.
13673 If the input image format is different from the format requested by
13674 the next filter, the scale filter will convert the input to the
13677 @subsection Options
13678 The filter accepts the following options, or any of the options
13679 supported by the libswscale scaler.
13681 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13682 the complete list of scaler options.
13687 Set the output video dimension expression. Default value is the input
13690 If the @var{width} or @var{w} value is 0, the input width is used for
13691 the output. If the @var{height} or @var{h} value is 0, the input height
13692 is used for the output.
13694 If one and only one of the values is -n with n >= 1, the scale filter
13695 will use a value that maintains the aspect ratio of the input image,
13696 calculated from the other specified dimension. After that it will,
13697 however, make sure that the calculated dimension is divisible by n and
13698 adjust the value if necessary.
13700 If both values are -n with n >= 1, the behavior will be identical to
13701 both values being set to 0 as previously detailed.
13703 See below for the list of accepted constants for use in the dimension
13707 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
13711 Only evaluate expressions once during the filter initialization or when a command is processed.
13714 Evaluate expressions for each incoming frame.
13718 Default value is @samp{init}.
13722 Set the interlacing mode. It accepts the following values:
13726 Force interlaced aware scaling.
13729 Do not apply interlaced scaling.
13732 Select interlaced aware scaling depending on whether the source frames
13733 are flagged as interlaced or not.
13736 Default value is @samp{0}.
13739 Set libswscale scaling flags. See
13740 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13741 complete list of values. If not explicitly specified the filter applies
13745 @item param0, param1
13746 Set libswscale input parameters for scaling algorithms that need them. See
13747 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13748 complete documentation. If not explicitly specified the filter applies
13754 Set the video size. For the syntax of this option, check the
13755 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13757 @item in_color_matrix
13758 @item out_color_matrix
13759 Set in/output YCbCr color space type.
13761 This allows the autodetected value to be overridden as well as allows forcing
13762 a specific value used for the output and encoder.
13764 If not specified, the color space type depends on the pixel format.
13770 Choose automatically.
13773 Format conforming to International Telecommunication Union (ITU)
13774 Recommendation BT.709.
13777 Set color space conforming to the United States Federal Communications
13778 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
13781 Set color space conforming to:
13785 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
13788 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
13791 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
13796 Set color space conforming to SMPTE ST 240:1999.
13801 Set in/output YCbCr sample range.
13803 This allows the autodetected value to be overridden as well as allows forcing
13804 a specific value used for the output and encoder. If not specified, the
13805 range depends on the pixel format. Possible values:
13809 Choose automatically.
13812 Set full range (0-255 in case of 8-bit luma).
13814 @item mpeg/limited/tv
13815 Set "MPEG" range (16-235 in case of 8-bit luma).
13818 @item force_original_aspect_ratio
13819 Enable decreasing or increasing output video width or height if necessary to
13820 keep the original aspect ratio. Possible values:
13824 Scale the video as specified and disable this feature.
13827 The output video dimensions will automatically be decreased if needed.
13830 The output video dimensions will automatically be increased if needed.
13834 One useful instance of this option is that when you know a specific device's
13835 maximum allowed resolution, you can use this to limit the output video to
13836 that, while retaining the aspect ratio. For example, device A allows
13837 1280x720 playback, and your video is 1920x800. Using this option (set it to
13838 decrease) and specifying 1280x720 to the command line makes the output
13841 Please note that this is a different thing than specifying -1 for @option{w}
13842 or @option{h}, you still need to specify the output resolution for this option
13847 The values of the @option{w} and @option{h} options are expressions
13848 containing the following constants:
13853 The input width and height
13857 These are the same as @var{in_w} and @var{in_h}.
13861 The output (scaled) width and height
13865 These are the same as @var{out_w} and @var{out_h}
13868 The same as @var{iw} / @var{ih}
13871 input sample aspect ratio
13874 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13878 horizontal and vertical input chroma subsample values. For example for the
13879 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13883 horizontal and vertical output chroma subsample values. For example for the
13884 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13887 @subsection Examples
13891 Scale the input video to a size of 200x100
13896 This is equivalent to:
13907 Specify a size abbreviation for the output size:
13912 which can also be written as:
13918 Scale the input to 2x:
13920 scale=w=2*iw:h=2*ih
13924 The above is the same as:
13926 scale=2*in_w:2*in_h
13930 Scale the input to 2x with forced interlaced scaling:
13932 scale=2*iw:2*ih:interl=1
13936 Scale the input to half size:
13938 scale=w=iw/2:h=ih/2
13942 Increase the width, and set the height to the same size:
13948 Seek Greek harmony:
13955 Increase the height, and set the width to 3/2 of the height:
13957 scale=w=3/2*oh:h=3/5*ih
13961 Increase the size, making the size a multiple of the chroma
13964 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13968 Increase the width to a maximum of 500 pixels,
13969 keeping the same aspect ratio as the input:
13971 scale=w='min(500\, iw*3/2):h=-1'
13975 Make pixels square by combining scale and setsar:
13977 scale='trunc(ih*dar):ih',setsar=1/1
13981 Make pixels square by combining scale and setsar,
13982 making sure the resulting resolution is even (required by some codecs):
13984 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
13988 @subsection Commands
13990 This filter supports the following commands:
13994 Set the output video dimension expression.
13995 The command accepts the same syntax of the corresponding option.
13997 If the specified expression is not valid, it is kept at its current
14003 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
14004 format conversion on CUDA video frames. Setting the output width and height
14005 works in the same way as for the @var{scale} filter.
14007 The following additional options are accepted:
14010 The pixel format of the output CUDA frames. If set to the string "same" (the
14011 default), the input format will be kept. Note that automatic format negotiation
14012 and conversion is not yet supported for hardware frames
14015 The interpolation algorithm used for resizing. One of the following:
14022 @item cubic2p_bspline
14023 2-parameter cubic (B=1, C=0)
14025 @item cubic2p_catmullrom
14026 2-parameter cubic (B=0, C=1/2)
14028 @item cubic2p_b05c03
14029 2-parameter cubic (B=1/2, C=3/10)
14041 Scale (resize) the input video, based on a reference video.
14043 See the scale filter for available options, scale2ref supports the same but
14044 uses the reference video instead of the main input as basis. scale2ref also
14045 supports the following additional constants for the @option{w} and
14046 @option{h} options:
14051 The main input video's width and height
14054 The same as @var{main_w} / @var{main_h}
14057 The main input video's sample aspect ratio
14059 @item main_dar, mdar
14060 The main input video's display aspect ratio. Calculated from
14061 @code{(main_w / main_h) * main_sar}.
14065 The main input video's horizontal and vertical chroma subsample values.
14066 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
14070 @subsection Examples
14074 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
14076 'scale2ref[b][a];[a][b]overlay'
14080 @anchor{selectivecolor}
14081 @section selectivecolor
14083 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
14084 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
14085 by the "purity" of the color (that is, how saturated it already is).
14087 This filter is similar to the Adobe Photoshop Selective Color tool.
14089 The filter accepts the following options:
14092 @item correction_method
14093 Select color correction method.
14095 Available values are:
14098 Specified adjustments are applied "as-is" (added/subtracted to original pixel
14101 Specified adjustments are relative to the original component value.
14103 Default is @code{absolute}.
14105 Adjustments for red pixels (pixels where the red component is the maximum)
14107 Adjustments for yellow pixels (pixels where the blue component is the minimum)
14109 Adjustments for green pixels (pixels where the green component is the maximum)
14111 Adjustments for cyan pixels (pixels where the red component is the minimum)
14113 Adjustments for blue pixels (pixels where the blue component is the maximum)
14115 Adjustments for magenta pixels (pixels where the green component is the minimum)
14117 Adjustments for white pixels (pixels where all components are greater than 128)
14119 Adjustments for all pixels except pure black and pure white
14121 Adjustments for black pixels (pixels where all components are lesser than 128)
14123 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
14126 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
14127 4 space separated floating point adjustment values in the [-1,1] range,
14128 respectively to adjust the amount of cyan, magenta, yellow and black for the
14129 pixels of its range.
14131 @subsection Examples
14135 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
14136 increase magenta by 27% in blue areas:
14138 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
14142 Use a Photoshop selective color preset:
14144 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
14148 @anchor{separatefields}
14149 @section separatefields
14151 The @code{separatefields} takes a frame-based video input and splits
14152 each frame into its components fields, producing a new half height clip
14153 with twice the frame rate and twice the frame count.
14155 This filter use field-dominance information in frame to decide which
14156 of each pair of fields to place first in the output.
14157 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
14159 @section setdar, setsar
14161 The @code{setdar} filter sets the Display Aspect Ratio for the filter
14164 This is done by changing the specified Sample (aka Pixel) Aspect
14165 Ratio, according to the following equation:
14167 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
14170 Keep in mind that the @code{setdar} filter does not modify the pixel
14171 dimensions of the video frame. Also, the display aspect ratio set by
14172 this filter may be changed by later filters in the filterchain,
14173 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
14176 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
14177 the filter output video.
14179 Note that as a consequence of the application of this filter, the
14180 output display aspect ratio will change according to the equation
14183 Keep in mind that the sample aspect ratio set by the @code{setsar}
14184 filter may be changed by later filters in the filterchain, e.g. if
14185 another "setsar" or a "setdar" filter is applied.
14187 It accepts the following parameters:
14190 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
14191 Set the aspect ratio used by the filter.
14193 The parameter can be a floating point number string, an expression, or
14194 a string of the form @var{num}:@var{den}, where @var{num} and
14195 @var{den} are the numerator and denominator of the aspect ratio. If
14196 the parameter is not specified, it is assumed the value "0".
14197 In case the form "@var{num}:@var{den}" is used, the @code{:} character
14201 Set the maximum integer value to use for expressing numerator and
14202 denominator when reducing the expressed aspect ratio to a rational.
14203 Default value is @code{100}.
14207 The parameter @var{sar} is an expression containing
14208 the following constants:
14212 These are approximated values for the mathematical constants e
14213 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14216 The input width and height.
14219 These are the same as @var{w} / @var{h}.
14222 The input sample aspect ratio.
14225 The input display aspect ratio. It is the same as
14226 (@var{w} / @var{h}) * @var{sar}.
14229 Horizontal and vertical chroma subsample values. For example, for the
14230 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14233 @subsection Examples
14238 To change the display aspect ratio to 16:9, specify one of the following:
14245 To change the sample aspect ratio to 10:11, specify:
14251 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14252 1000 in the aspect ratio reduction, use the command:
14254 setdar=ratio=16/9:max=1000
14262 Force field for the output video frame.
14264 The @code{setfield} filter marks the interlace type field for the
14265 output frames. It does not change the input frame, but only sets the
14266 corresponding property, which affects how the frame is treated by
14267 following filters (e.g. @code{fieldorder} or @code{yadif}).
14269 The filter accepts the following options:
14274 Available values are:
14278 Keep the same field property.
14281 Mark the frame as bottom-field-first.
14284 Mark the frame as top-field-first.
14287 Mark the frame as progressive.
14293 Show a line containing various information for each input video frame.
14294 The input video is not modified.
14296 The shown line contains a sequence of key/value pairs of the form
14297 @var{key}:@var{value}.
14299 The following values are shown in the output:
14303 The (sequential) number of the input frame, starting from 0.
14306 The Presentation TimeStamp of the input frame, expressed as a number of
14307 time base units. The time base unit depends on the filter input pad.
14310 The Presentation TimeStamp of the input frame, expressed as a number of
14314 The position of the frame in the input stream, or -1 if this information is
14315 unavailable and/or meaningless (for example in case of synthetic video).
14318 The pixel format name.
14321 The sample aspect ratio of the input frame, expressed in the form
14322 @var{num}/@var{den}.
14325 The size of the input frame. For the syntax of this option, check the
14326 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14329 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14330 for bottom field first).
14333 This is 1 if the frame is a key frame, 0 otherwise.
14336 The picture type of the input frame ("I" for an I-frame, "P" for a
14337 P-frame, "B" for a B-frame, or "?" for an unknown type).
14338 Also refer to the documentation of the @code{AVPictureType} enum and of
14339 the @code{av_get_picture_type_char} function defined in
14340 @file{libavutil/avutil.h}.
14343 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14345 @item plane_checksum
14346 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14347 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14350 @section showpalette
14352 Displays the 256 colors palette of each frame. This filter is only relevant for
14353 @var{pal8} pixel format frames.
14355 It accepts the following option:
14359 Set the size of the box used to represent one palette color entry. Default is
14360 @code{30} (for a @code{30x30} pixel box).
14363 @section shuffleframes
14365 Reorder and/or duplicate and/or drop video frames.
14367 It accepts the following parameters:
14371 Set the destination indexes of input frames.
14372 This is space or '|' separated list of indexes that maps input frames to output
14373 frames. Number of indexes also sets maximal value that each index may have.
14374 '-1' index have special meaning and that is to drop frame.
14377 The first frame has the index 0. The default is to keep the input unchanged.
14379 @subsection Examples
14383 Swap second and third frame of every three frames of the input:
14385 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14389 Swap 10th and 1st frame of every ten frames of the input:
14391 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14395 @section shuffleplanes
14397 Reorder and/or duplicate video planes.
14399 It accepts the following parameters:
14404 The index of the input plane to be used as the first output plane.
14407 The index of the input plane to be used as the second output plane.
14410 The index of the input plane to be used as the third output plane.
14413 The index of the input plane to be used as the fourth output plane.
14417 The first plane has the index 0. The default is to keep the input unchanged.
14419 @subsection Examples
14423 Swap the second and third planes of the input:
14425 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14429 @anchor{signalstats}
14430 @section signalstats
14431 Evaluate various visual metrics that assist in determining issues associated
14432 with the digitization of analog video media.
14434 By default the filter will log these metadata values:
14438 Display the minimal Y value contained within the input frame. Expressed in
14442 Display the Y value at the 10% percentile within the input frame. Expressed in
14446 Display the average Y value within the input frame. Expressed in range of
14450 Display the Y value at the 90% percentile within the input frame. Expressed in
14454 Display the maximum Y value contained within the input frame. Expressed in
14458 Display the minimal U value contained within the input frame. Expressed in
14462 Display the U value at the 10% percentile within the input frame. Expressed in
14466 Display the average U value within the input frame. Expressed in range of
14470 Display the U value at the 90% percentile within the input frame. Expressed in
14474 Display the maximum U value contained within the input frame. Expressed in
14478 Display the minimal V value contained within the input frame. Expressed in
14482 Display the V value at the 10% percentile within the input frame. Expressed in
14486 Display the average V value within the input frame. Expressed in range of
14490 Display the V value at the 90% percentile within the input frame. Expressed in
14494 Display the maximum V value contained within the input frame. Expressed in
14498 Display the minimal saturation value contained within the input frame.
14499 Expressed in range of [0-~181.02].
14502 Display the saturation value at the 10% percentile within the input frame.
14503 Expressed in range of [0-~181.02].
14506 Display the average saturation value within the input frame. Expressed in range
14510 Display the saturation value at the 90% percentile within the input frame.
14511 Expressed in range of [0-~181.02].
14514 Display the maximum saturation value contained within the input frame.
14515 Expressed in range of [0-~181.02].
14518 Display the median value for hue within the input frame. Expressed in range of
14522 Display the average value for hue within the input frame. Expressed in range of
14526 Display the average of sample value difference between all values of the Y
14527 plane in the current frame and corresponding values of the previous input frame.
14528 Expressed in range of [0-255].
14531 Display the average of sample value difference between all values of the U
14532 plane in the current frame and corresponding values of the previous input frame.
14533 Expressed in range of [0-255].
14536 Display the average of sample value difference between all values of the V
14537 plane in the current frame and corresponding values of the previous input frame.
14538 Expressed in range of [0-255].
14541 Display bit depth of Y plane in current frame.
14542 Expressed in range of [0-16].
14545 Display bit depth of U plane in current frame.
14546 Expressed in range of [0-16].
14549 Display bit depth of V plane in current frame.
14550 Expressed in range of [0-16].
14553 The filter accepts the following options:
14559 @option{stat} specify an additional form of image analysis.
14560 @option{out} output video with the specified type of pixel highlighted.
14562 Both options accept the following values:
14566 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14567 unlike the neighboring pixels of the same field. Examples of temporal outliers
14568 include the results of video dropouts, head clogs, or tape tracking issues.
14571 Identify @var{vertical line repetition}. Vertical line repetition includes
14572 similar rows of pixels within a frame. In born-digital video vertical line
14573 repetition is common, but this pattern is uncommon in video digitized from an
14574 analog source. When it occurs in video that results from the digitization of an
14575 analog source it can indicate concealment from a dropout compensator.
14578 Identify pixels that fall outside of legal broadcast range.
14582 Set the highlight color for the @option{out} option. The default color is
14586 @subsection Examples
14590 Output data of various video metrics:
14592 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14596 Output specific data about the minimum and maximum values of the Y plane per frame:
14598 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14602 Playback video while highlighting pixels that are outside of broadcast range in red.
14604 ffplay example.mov -vf signalstats="out=brng:color=red"
14608 Playback video with signalstats metadata drawn over the frame.
14610 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14613 The contents of signalstat_drawtext.txt used in the command are:
14616 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14617 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14618 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14619 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14627 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14628 input. In this case the matching between the inputs can be calculated additionally.
14629 The filter always passes through the first input. The signature of each stream can
14630 be written into a file.
14632 It accepts the following options:
14636 Enable or disable the matching process.
14638 Available values are:
14642 Disable the calculation of a matching (default).
14644 Calculate the matching for the whole video and output whether the whole video
14645 matches or only parts.
14647 Calculate only until a matching is found or the video ends. Should be faster in
14652 Set the number of inputs. The option value must be a non negative integer.
14653 Default value is 1.
14656 Set the path to which the output is written. If there is more than one input,
14657 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14658 integer), that will be replaced with the input number. If no filename is
14659 specified, no output will be written. This is the default.
14662 Choose the output format.
14664 Available values are:
14668 Use the specified binary representation (default).
14670 Use the specified xml representation.
14674 Set threshold to detect one word as similar. The option value must be an integer
14675 greater than zero. The default value is 9000.
14678 Set threshold to detect all words as similar. The option value must be an integer
14679 greater than zero. The default value is 60000.
14682 Set threshold to detect frames as similar. The option value must be an integer
14683 greater than zero. The default value is 116.
14686 Set the minimum length of a sequence in frames to recognize it as matching
14687 sequence. The option value must be a non negative integer value.
14688 The default value is 0.
14691 Set the minimum relation, that matching frames to all frames must have.
14692 The option value must be a double value between 0 and 1. The default value is 0.5.
14695 @subsection Examples
14699 To calculate the signature of an input video and store it in signature.bin:
14701 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14705 To detect whether two videos match and store the signatures in XML format in
14706 signature0.xml and signature1.xml:
14708 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 -
14716 Blur the input video without impacting the outlines.
14718 It accepts the following options:
14721 @item luma_radius, lr
14722 Set the luma radius. The option value must be a float number in
14723 the range [0.1,5.0] that specifies the variance of the gaussian filter
14724 used to blur the image (slower if larger). Default value is 1.0.
14726 @item luma_strength, ls
14727 Set the luma strength. The option value must be a float number
14728 in the range [-1.0,1.0] that configures the blurring. A value included
14729 in [0.0,1.0] will blur the image whereas a value included in
14730 [-1.0,0.0] will sharpen the image. Default value is 1.0.
14732 @item luma_threshold, lt
14733 Set the luma threshold used as a coefficient to determine
14734 whether a pixel should be blurred or not. The option value must be an
14735 integer in the range [-30,30]. A value of 0 will filter all the image,
14736 a value included in [0,30] will filter flat areas and a value included
14737 in [-30,0] will filter edges. Default value is 0.
14739 @item chroma_radius, cr
14740 Set the chroma radius. The option value must be a float number in
14741 the range [0.1,5.0] that specifies the variance of the gaussian filter
14742 used to blur the image (slower if larger). Default value is @option{luma_radius}.
14744 @item chroma_strength, cs
14745 Set the chroma strength. The option value must be a float number
14746 in the range [-1.0,1.0] that configures the blurring. A value included
14747 in [0.0,1.0] will blur the image whereas a value included in
14748 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
14750 @item chroma_threshold, ct
14751 Set the chroma threshold used as a coefficient to determine
14752 whether a pixel should be blurred or not. The option value must be an
14753 integer in the range [-30,30]. A value of 0 will filter all the image,
14754 a value included in [0,30] will filter flat areas and a value included
14755 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
14758 If a chroma option is not explicitly set, the corresponding luma value
14763 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
14765 This filter takes in input two input videos, the first input is
14766 considered the "main" source and is passed unchanged to the
14767 output. The second input is used as a "reference" video for computing
14770 Both video inputs must have the same resolution and pixel format for
14771 this filter to work correctly. Also it assumes that both inputs
14772 have the same number of frames, which are compared one by one.
14774 The filter stores the calculated SSIM of each frame.
14776 The description of the accepted parameters follows.
14779 @item stats_file, f
14780 If specified the filter will use the named file to save the SSIM of
14781 each individual frame. When filename equals "-" the data is sent to
14785 The file printed if @var{stats_file} is selected, contains a sequence of
14786 key/value pairs of the form @var{key}:@var{value} for each compared
14789 A description of each shown parameter follows:
14793 sequential number of the input frame, starting from 1
14795 @item Y, U, V, R, G, B
14796 SSIM of the compared frames for the component specified by the suffix.
14799 SSIM of the compared frames for the whole frame.
14802 Same as above but in dB representation.
14805 This filter also supports the @ref{framesync} options.
14809 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14810 [main][ref] ssim="stats_file=stats.log" [out]
14813 On this example the input file being processed is compared with the
14814 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
14815 is stored in @file{stats.log}.
14817 Another example with both psnr and ssim at same time:
14819 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
14824 Convert between different stereoscopic image formats.
14826 The filters accept the following options:
14830 Set stereoscopic image format of input.
14832 Available values for input image formats are:
14835 side by side parallel (left eye left, right eye right)
14838 side by side crosseye (right eye left, left eye right)
14841 side by side parallel with half width resolution
14842 (left eye left, right eye right)
14845 side by side crosseye with half width resolution
14846 (right eye left, left eye right)
14849 above-below (left eye above, right eye below)
14852 above-below (right eye above, left eye below)
14855 above-below with half height resolution
14856 (left eye above, right eye below)
14859 above-below with half height resolution
14860 (right eye above, left eye below)
14863 alternating frames (left eye first, right eye second)
14866 alternating frames (right eye first, left eye second)
14869 interleaved rows (left eye has top row, right eye starts on next row)
14872 interleaved rows (right eye has top row, left eye starts on next row)
14875 interleaved columns, left eye first
14878 interleaved columns, right eye first
14880 Default value is @samp{sbsl}.
14884 Set stereoscopic image format of output.
14888 side by side parallel (left eye left, right eye right)
14891 side by side crosseye (right eye left, left eye right)
14894 side by side parallel with half width resolution
14895 (left eye left, right eye right)
14898 side by side crosseye with half width resolution
14899 (right eye left, left eye right)
14902 above-below (left eye above, right eye below)
14905 above-below (right eye above, left eye below)
14908 above-below with half height resolution
14909 (left eye above, right eye below)
14912 above-below with half height resolution
14913 (right eye above, left eye below)
14916 alternating frames (left eye first, right eye second)
14919 alternating frames (right eye first, left eye second)
14922 interleaved rows (left eye has top row, right eye starts on next row)
14925 interleaved rows (right eye has top row, left eye starts on next row)
14928 anaglyph red/blue gray
14929 (red filter on left eye, blue filter on right eye)
14932 anaglyph red/green gray
14933 (red filter on left eye, green filter on right eye)
14936 anaglyph red/cyan gray
14937 (red filter on left eye, cyan filter on right eye)
14940 anaglyph red/cyan half colored
14941 (red filter on left eye, cyan filter on right eye)
14944 anaglyph red/cyan color
14945 (red filter on left eye, cyan filter on right eye)
14948 anaglyph red/cyan color optimized with the least squares projection of dubois
14949 (red filter on left eye, cyan filter on right eye)
14952 anaglyph green/magenta gray
14953 (green filter on left eye, magenta filter on right eye)
14956 anaglyph green/magenta half colored
14957 (green filter on left eye, magenta filter on right eye)
14960 anaglyph green/magenta colored
14961 (green filter on left eye, magenta filter on right eye)
14964 anaglyph green/magenta color optimized with the least squares projection of dubois
14965 (green filter on left eye, magenta filter on right eye)
14968 anaglyph yellow/blue gray
14969 (yellow filter on left eye, blue filter on right eye)
14972 anaglyph yellow/blue half colored
14973 (yellow filter on left eye, blue filter on right eye)
14976 anaglyph yellow/blue colored
14977 (yellow filter on left eye, blue filter on right eye)
14980 anaglyph yellow/blue color optimized with the least squares projection of dubois
14981 (yellow filter on left eye, blue filter on right eye)
14984 mono output (left eye only)
14987 mono output (right eye only)
14990 checkerboard, left eye first
14993 checkerboard, right eye first
14996 interleaved columns, left eye first
14999 interleaved columns, right eye first
15005 Default value is @samp{arcd}.
15008 @subsection Examples
15012 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
15018 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
15024 @section streamselect, astreamselect
15025 Select video or audio streams.
15027 The filter accepts the following options:
15031 Set number of inputs. Default is 2.
15034 Set input indexes to remap to outputs.
15037 @subsection Commands
15039 The @code{streamselect} and @code{astreamselect} filter supports the following
15044 Set input indexes to remap to outputs.
15047 @subsection Examples
15051 Select first 5 seconds 1st stream and rest of time 2nd stream:
15053 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
15057 Same as above, but for audio:
15059 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
15064 Apply sobel operator to input video stream.
15066 The filter accepts the following option:
15070 Set which planes will be processed, unprocessed planes will be copied.
15071 By default value 0xf, all planes will be processed.
15074 Set value which will be multiplied with filtered result.
15077 Set value which will be added to filtered result.
15083 Apply a simple postprocessing filter that compresses and decompresses the image
15084 at several (or - in the case of @option{quality} level @code{6} - all) shifts
15085 and average the results.
15087 The filter accepts the following options:
15091 Set quality. This option defines the number of levels for averaging. It accepts
15092 an integer in the range 0-6. If set to @code{0}, the filter will have no
15093 effect. A value of @code{6} means the higher quality. For each increment of
15094 that value the speed drops by a factor of approximately 2. Default value is
15098 Force a constant quantization parameter. If not set, the filter will use the QP
15099 from the video stream (if available).
15102 Set thresholding mode. Available modes are:
15106 Set hard thresholding (default).
15108 Set soft thresholding (better de-ringing effect, but likely blurrier).
15111 @item use_bframe_qp
15112 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
15113 option may cause flicker since the B-Frames have often larger QP. Default is
15114 @code{0} (not enabled).
15120 Draw subtitles on top of input video using the libass library.
15122 To enable compilation of this filter you need to configure FFmpeg with
15123 @code{--enable-libass}. This filter also requires a build with libavcodec and
15124 libavformat to convert the passed subtitles file to ASS (Advanced Substation
15125 Alpha) subtitles format.
15127 The filter accepts the following options:
15131 Set the filename of the subtitle file to read. It must be specified.
15133 @item original_size
15134 Specify the size of the original video, the video for which the ASS file
15135 was composed. For the syntax of this option, check the
15136 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15137 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
15138 correctly scale the fonts if the aspect ratio has been changed.
15141 Set a directory path containing fonts that can be used by the filter.
15142 These fonts will be used in addition to whatever the font provider uses.
15145 Process alpha channel, by default alpha channel is untouched.
15148 Set subtitles input character encoding. @code{subtitles} filter only. Only
15149 useful if not UTF-8.
15151 @item stream_index, si
15152 Set subtitles stream index. @code{subtitles} filter only.
15155 Override default style or script info parameters of the subtitles. It accepts a
15156 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
15159 If the first key is not specified, it is assumed that the first value
15160 specifies the @option{filename}.
15162 For example, to render the file @file{sub.srt} on top of the input
15163 video, use the command:
15168 which is equivalent to:
15170 subtitles=filename=sub.srt
15173 To render the default subtitles stream from file @file{video.mkv}, use:
15175 subtitles=video.mkv
15178 To render the second subtitles stream from that file, use:
15180 subtitles=video.mkv:si=1
15183 To make the subtitles stream from @file{sub.srt} appear in transparent green
15184 @code{DejaVu Serif}, use:
15186 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
15189 @section super2xsai
15191 Scale the input by 2x and smooth using the Super2xSaI (Scale and
15192 Interpolate) pixel art scaling algorithm.
15194 Useful for enlarging pixel art images without reducing sharpness.
15198 Swap two rectangular objects in video.
15200 This filter accepts the following options:
15210 Set 1st rect x coordinate.
15213 Set 1st rect y coordinate.
15216 Set 2nd rect x coordinate.
15219 Set 2nd rect y coordinate.
15221 All expressions are evaluated once for each frame.
15224 The all options are expressions containing the following constants:
15229 The input width and height.
15232 same as @var{w} / @var{h}
15235 input sample aspect ratio
15238 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15241 The number of the input frame, starting from 0.
15244 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15247 the position in the file of the input frame, NAN if unknown
15255 Apply telecine process to the video.
15257 This filter accepts the following options:
15266 The default value is @code{top}.
15270 A string of numbers representing the pulldown pattern you wish to apply.
15271 The default value is @code{23}.
15275 Some typical patterns:
15280 24p: 2332 (preferred)
15287 24p: 222222222223 ("Euro pulldown")
15294 Apply threshold effect to video stream.
15296 This filter needs four video streams to perform thresholding.
15297 First stream is stream we are filtering.
15298 Second stream is holding threshold values, third stream is holding min values,
15299 and last, fourth stream is holding max values.
15301 The filter accepts the following option:
15305 Set which planes will be processed, unprocessed planes will be copied.
15306 By default value 0xf, all planes will be processed.
15309 For example if first stream pixel's component value is less then threshold value
15310 of pixel component from 2nd threshold stream, third stream value will picked,
15311 otherwise fourth stream pixel component value will be picked.
15313 Using color source filter one can perform various types of thresholding:
15315 @subsection Examples
15319 Binary threshold, using gray color as threshold:
15321 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15325 Inverted binary threshold, using gray color as threshold:
15327 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15331 Truncate binary threshold, using gray color as threshold:
15333 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15337 Threshold to zero, using gray color as threshold:
15339 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15343 Inverted threshold to zero, using gray color as threshold:
15345 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15350 Select the most representative frame in a given sequence of consecutive frames.
15352 The filter accepts the following options:
15356 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15357 will pick one of them, and then handle the next batch of @var{n} frames until
15358 the end. Default is @code{100}.
15361 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15362 value will result in a higher memory usage, so a high value is not recommended.
15364 @subsection Examples
15368 Extract one picture each 50 frames:
15374 Complete example of a thumbnail creation with @command{ffmpeg}:
15376 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15382 Tile several successive frames together.
15384 The filter accepts the following options:
15389 Set the grid size (i.e. the number of lines and columns). For the syntax of
15390 this option, check the
15391 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15394 Set the maximum number of frames to render in the given area. It must be less
15395 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15396 the area will be used.
15399 Set the outer border margin in pixels.
15402 Set the inner border thickness (i.e. the number of pixels between frames). For
15403 more advanced padding options (such as having different values for the edges),
15404 refer to the pad video filter.
15407 Specify the color of the unused area. For the syntax of this option, check the
15408 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15409 The default value of @var{color} is "black".
15412 Set the number of frames to overlap when tiling several successive frames together.
15413 The value must be between @code{0} and @var{nb_frames - 1}.
15416 Set the number of frames to initially be empty before displaying first output frame.
15417 This controls how soon will one get first output frame.
15418 The value must be between @code{0} and @var{nb_frames - 1}.
15421 @subsection Examples
15425 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15427 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15429 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15430 duplicating each output frame to accommodate the originally detected frame
15434 Display @code{5} pictures in an area of @code{3x2} frames,
15435 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15436 mixed flat and named options:
15438 tile=3x2:nb_frames=5:padding=7:margin=2
15442 @section tinterlace
15444 Perform various types of temporal field interlacing.
15446 Frames are counted starting from 1, so the first input frame is
15449 The filter accepts the following options:
15454 Specify the mode of the interlacing. This option can also be specified
15455 as a value alone. See below for a list of values for this option.
15457 Available values are:
15461 Move odd frames into the upper field, even into the lower field,
15462 generating a double height frame at half frame rate.
15466 Frame 1 Frame 2 Frame 3 Frame 4
15468 11111 22222 33333 44444
15469 11111 22222 33333 44444
15470 11111 22222 33333 44444
15471 11111 22222 33333 44444
15485 Only output odd frames, even frames are dropped, generating a frame with
15486 unchanged height at half frame rate.
15491 Frame 1 Frame 2 Frame 3 Frame 4
15493 11111 22222 33333 44444
15494 11111 22222 33333 44444
15495 11111 22222 33333 44444
15496 11111 22222 33333 44444
15506 Only output even frames, odd frames are dropped, generating a frame with
15507 unchanged height at half frame rate.
15512 Frame 1 Frame 2 Frame 3 Frame 4
15514 11111 22222 33333 44444
15515 11111 22222 33333 44444
15516 11111 22222 33333 44444
15517 11111 22222 33333 44444
15527 Expand each frame to full height, but pad alternate lines with black,
15528 generating a frame with double height at the same input frame rate.
15533 Frame 1 Frame 2 Frame 3 Frame 4
15535 11111 22222 33333 44444
15536 11111 22222 33333 44444
15537 11111 22222 33333 44444
15538 11111 22222 33333 44444
15541 11111 ..... 33333 .....
15542 ..... 22222 ..... 44444
15543 11111 ..... 33333 .....
15544 ..... 22222 ..... 44444
15545 11111 ..... 33333 .....
15546 ..... 22222 ..... 44444
15547 11111 ..... 33333 .....
15548 ..... 22222 ..... 44444
15552 @item interleave_top, 4
15553 Interleave the upper field from odd frames with the lower field from
15554 even frames, generating a frame with unchanged height at half frame rate.
15559 Frame 1 Frame 2 Frame 3 Frame 4
15561 11111<- 22222 33333<- 44444
15562 11111 22222<- 33333 44444<-
15563 11111<- 22222 33333<- 44444
15564 11111 22222<- 33333 44444<-
15574 @item interleave_bottom, 5
15575 Interleave the lower field from odd frames with the upper field from
15576 even frames, generating a frame with unchanged height at half frame rate.
15581 Frame 1 Frame 2 Frame 3 Frame 4
15583 11111 22222<- 33333 44444<-
15584 11111<- 22222 33333<- 44444
15585 11111 22222<- 33333 44444<-
15586 11111<- 22222 33333<- 44444
15596 @item interlacex2, 6
15597 Double frame rate with unchanged height. Frames are inserted each
15598 containing the second temporal field from the previous input frame and
15599 the first temporal field from the next input frame. This mode relies on
15600 the top_field_first flag. Useful for interlaced video displays with no
15601 field synchronisation.
15606 Frame 1 Frame 2 Frame 3 Frame 4
15608 11111 22222 33333 44444
15609 11111 22222 33333 44444
15610 11111 22222 33333 44444
15611 11111 22222 33333 44444
15614 11111 22222 22222 33333 33333 44444 44444
15615 11111 11111 22222 22222 33333 33333 44444
15616 11111 22222 22222 33333 33333 44444 44444
15617 11111 11111 22222 22222 33333 33333 44444
15622 Move odd frames into the upper field, even into the lower field,
15623 generating a double height frame at same frame rate.
15628 Frame 1 Frame 2 Frame 3 Frame 4
15630 11111 22222 33333 44444
15631 11111 22222 33333 44444
15632 11111 22222 33333 44444
15633 11111 22222 33333 44444
15636 11111 33333 33333 55555
15637 22222 22222 44444 44444
15638 11111 33333 33333 55555
15639 22222 22222 44444 44444
15640 11111 33333 33333 55555
15641 22222 22222 44444 44444
15642 11111 33333 33333 55555
15643 22222 22222 44444 44444
15648 Numeric values are deprecated but are accepted for backward
15649 compatibility reasons.
15651 Default mode is @code{merge}.
15654 Specify flags influencing the filter process.
15656 Available value for @var{flags} is:
15659 @item low_pass_filter, vlfp
15660 Enable linear vertical low-pass filtering in the filter.
15661 Vertical low-pass filtering is required when creating an interlaced
15662 destination from a progressive source which contains high-frequency
15663 vertical detail. Filtering will reduce interlace 'twitter' and Moire
15666 @item complex_filter, cvlfp
15667 Enable complex vertical low-pass filtering.
15668 This will slightly less reduce interlace 'twitter' and Moire
15669 patterning but better retain detail and subjective sharpness impression.
15673 Vertical low-pass filtering can only be enabled for @option{mode}
15674 @var{interleave_top} and @var{interleave_bottom}.
15680 Mix successive video frames.
15682 A description of the accepted options follows.
15686 The number of successive frames to mix. If unspecified, it defaults to 3.
15689 Specify weight of each input video frame.
15690 Each weight is separated by space. If number of weights is smaller than
15691 number of @var{frames} last specified weight will be used for all remaining
15695 Specify scale, if it is set it will be multiplied with sum
15696 of each weight multiplied with pixel values to give final destination
15697 pixel value. By default @var{scale} is auto scaled to sum of weights.
15700 @subsection Examples
15704 Average 7 successive frames:
15706 tmix=frames=7:weights="1 1 1 1 1 1 1"
15710 Apply simple temporal convolution:
15712 tmix=frames=3:weights="-1 3 -1"
15716 Similar as above but only showing temporal differences:
15718 tmix=frames=3:weights="-1 2 -1":scale=1
15723 Tone map colors from different dynamic ranges.
15725 This filter expects data in single precision floating point, as it needs to
15726 operate on (and can output) out-of-range values. Another filter, such as
15727 @ref{zscale}, is needed to convert the resulting frame to a usable format.
15729 The tonemapping algorithms implemented only work on linear light, so input
15730 data should be linearized beforehand (and possibly correctly tagged).
15733 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
15736 @subsection Options
15737 The filter accepts the following options.
15741 Set the tone map algorithm to use.
15743 Possible values are:
15746 Do not apply any tone map, only desaturate overbright pixels.
15749 Hard-clip any out-of-range values. Use it for perfect color accuracy for
15750 in-range values, while distorting out-of-range values.
15753 Stretch the entire reference gamut to a linear multiple of the display.
15756 Fit a logarithmic transfer between the tone curves.
15759 Preserve overall image brightness with a simple curve, using nonlinear
15760 contrast, which results in flattening details and degrading color accuracy.
15763 Preserve both dark and bright details better than @var{reinhard}, at the cost
15764 of slightly darkening everything. Use it when detail preservation is more
15765 important than color and brightness accuracy.
15768 Smoothly map out-of-range values, while retaining contrast and colors for
15769 in-range material as much as possible. Use it when color accuracy is more
15770 important than detail preservation.
15776 Tune the tone mapping algorithm.
15778 This affects the following algorithms:
15784 Specifies the scale factor to use while stretching.
15788 Specifies the exponent of the function.
15792 Specify an extra linear coefficient to multiply into the signal before clipping.
15796 Specify the local contrast coefficient at the display peak.
15797 Default to 0.5, which means that in-gamut values will be about half as bright
15804 Specify the transition point from linear to mobius transform. Every value
15805 below this point is guaranteed to be mapped 1:1. The higher the value, the
15806 more accurate the result will be, at the cost of losing bright details.
15807 Default to 0.3, which due to the steep initial slope still preserves in-range
15808 colors fairly accurately.
15812 Apply desaturation for highlights that exceed this level of brightness. The
15813 higher the parameter, the more color information will be preserved. This
15814 setting helps prevent unnaturally blown-out colors for super-highlights, by
15815 (smoothly) turning into white instead. This makes images feel more natural,
15816 at the cost of reducing information about out-of-range colors.
15818 The default of 2.0 is somewhat conservative and will mostly just apply to
15819 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
15821 This option works only if the input frame has a supported color tag.
15824 Override signal/nominal/reference peak with this value. Useful when the
15825 embedded peak information in display metadata is not reliable or when tone
15826 mapping from a lower range to a higher range.
15831 Transpose rows with columns in the input video and optionally flip it.
15833 It accepts the following parameters:
15838 Specify the transposition direction.
15840 Can assume the following values:
15842 @item 0, 4, cclock_flip
15843 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
15851 Rotate by 90 degrees clockwise, that is:
15859 Rotate by 90 degrees counterclockwise, that is:
15866 @item 3, 7, clock_flip
15867 Rotate by 90 degrees clockwise and vertically flip, that is:
15875 For values between 4-7, the transposition is only done if the input
15876 video geometry is portrait and not landscape. These values are
15877 deprecated, the @code{passthrough} option should be used instead.
15879 Numerical values are deprecated, and should be dropped in favor of
15880 symbolic constants.
15883 Do not apply the transposition if the input geometry matches the one
15884 specified by the specified value. It accepts the following values:
15887 Always apply transposition.
15889 Preserve portrait geometry (when @var{height} >= @var{width}).
15891 Preserve landscape geometry (when @var{width} >= @var{height}).
15894 Default value is @code{none}.
15897 For example to rotate by 90 degrees clockwise and preserve portrait
15900 transpose=dir=1:passthrough=portrait
15903 The command above can also be specified as:
15905 transpose=1:portrait
15909 Trim the input so that the output contains one continuous subpart of the input.
15911 It accepts the following parameters:
15914 Specify the time of the start of the kept section, i.e. the frame with the
15915 timestamp @var{start} will be the first frame in the output.
15918 Specify the time of the first frame that will be dropped, i.e. the frame
15919 immediately preceding the one with the timestamp @var{end} will be the last
15920 frame in the output.
15923 This is the same as @var{start}, except this option sets the start timestamp
15924 in timebase units instead of seconds.
15927 This is the same as @var{end}, except this option sets the end timestamp
15928 in timebase units instead of seconds.
15931 The maximum duration of the output in seconds.
15934 The number of the first frame that should be passed to the output.
15937 The number of the first frame that should be dropped.
15940 @option{start}, @option{end}, and @option{duration} are expressed as time
15941 duration specifications; see
15942 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15943 for the accepted syntax.
15945 Note that the first two sets of the start/end options and the @option{duration}
15946 option look at the frame timestamp, while the _frame variants simply count the
15947 frames that pass through the filter. Also note that this filter does not modify
15948 the timestamps. If you wish for the output timestamps to start at zero, insert a
15949 setpts filter after the trim filter.
15951 If multiple start or end options are set, this filter tries to be greedy and
15952 keep all the frames that match at least one of the specified constraints. To keep
15953 only the part that matches all the constraints at once, chain multiple trim
15956 The defaults are such that all the input is kept. So it is possible to set e.g.
15957 just the end values to keep everything before the specified time.
15962 Drop everything except the second minute of input:
15964 ffmpeg -i INPUT -vf trim=60:120
15968 Keep only the first second:
15970 ffmpeg -i INPUT -vf trim=duration=1
15975 @section unpremultiply
15976 Apply alpha unpremultiply effect to input video stream using first plane
15977 of second stream as alpha.
15979 Both streams must have same dimensions and same pixel format.
15981 The filter accepts the following option:
15985 Set which planes will be processed, unprocessed planes will be copied.
15986 By default value 0xf, all planes will be processed.
15988 If the format has 1 or 2 components, then luma is bit 0.
15989 If the format has 3 or 4 components:
15990 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15991 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15992 If present, the alpha channel is always the last bit.
15995 Do not require 2nd input for processing, instead use alpha plane from input stream.
16001 Sharpen or blur the input video.
16003 It accepts the following parameters:
16006 @item luma_msize_x, lx
16007 Set the luma matrix horizontal size. It must be an odd integer between
16008 3 and 23. The default value is 5.
16010 @item luma_msize_y, ly
16011 Set the luma matrix vertical size. It must be an odd integer between 3
16012 and 23. The default value is 5.
16014 @item luma_amount, la
16015 Set the luma effect strength. It must be a floating point number, reasonable
16016 values lay between -1.5 and 1.5.
16018 Negative values will blur the input video, while positive values will
16019 sharpen it, a value of zero will disable the effect.
16021 Default value is 1.0.
16023 @item chroma_msize_x, cx
16024 Set the chroma matrix horizontal size. It must be an odd integer
16025 between 3 and 23. The default value is 5.
16027 @item chroma_msize_y, cy
16028 Set the chroma matrix vertical size. It must be an odd integer
16029 between 3 and 23. The default value is 5.
16031 @item chroma_amount, ca
16032 Set the chroma effect strength. It must be a floating point number, reasonable
16033 values lay between -1.5 and 1.5.
16035 Negative values will blur the input video, while positive values will
16036 sharpen it, a value of zero will disable the effect.
16038 Default value is 0.0.
16042 All parameters are optional and default to the equivalent of the
16043 string '5:5:1.0:5:5:0.0'.
16045 @subsection Examples
16049 Apply strong luma sharpen effect:
16051 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
16055 Apply a strong blur of both luma and chroma parameters:
16057 unsharp=7:7:-2:7:7:-2
16063 Apply ultra slow/simple postprocessing filter that compresses and decompresses
16064 the image at several (or - in the case of @option{quality} level @code{8} - all)
16065 shifts and average the results.
16067 The way this differs from the behavior of spp is that uspp actually encodes &
16068 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
16069 DCT similar to MJPEG.
16071 The filter accepts the following options:
16075 Set quality. This option defines the number of levels for averaging. It accepts
16076 an integer in the range 0-8. If set to @code{0}, the filter will have no
16077 effect. A value of @code{8} means the higher quality. For each increment of
16078 that value the speed drops by a factor of approximately 2. Default value is
16082 Force a constant quantization parameter. If not set, the filter will use the QP
16083 from the video stream (if available).
16086 @section vaguedenoiser
16088 Apply a wavelet based denoiser.
16090 It transforms each frame from the video input into the wavelet domain,
16091 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
16092 the obtained coefficients. It does an inverse wavelet transform after.
16093 Due to wavelet properties, it should give a nice smoothed result, and
16094 reduced noise, without blurring picture features.
16096 This filter accepts the following options:
16100 The filtering strength. The higher, the more filtered the video will be.
16101 Hard thresholding can use a higher threshold than soft thresholding
16102 before the video looks overfiltered. Default value is 2.
16105 The filtering method the filter will use.
16107 It accepts the following values:
16110 All values under the threshold will be zeroed.
16113 All values under the threshold will be zeroed. All values above will be
16114 reduced by the threshold.
16117 Scales or nullifies coefficients - intermediary between (more) soft and
16118 (less) hard thresholding.
16121 Default is garrote.
16124 Number of times, the wavelet will decompose the picture. Picture can't
16125 be decomposed beyond a particular point (typically, 8 for a 640x480
16126 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
16129 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
16132 A list of the planes to process. By default all planes are processed.
16135 @section vectorscope
16137 Display 2 color component values in the two dimensional graph (which is called
16140 This filter accepts the following options:
16144 Set vectorscope mode.
16146 It accepts the following values:
16149 Gray values are displayed on graph, higher brightness means more pixels have
16150 same component color value on location in graph. This is the default mode.
16153 Gray values are displayed on graph. Surrounding pixels values which are not
16154 present in video frame are drawn in gradient of 2 color components which are
16155 set by option @code{x} and @code{y}. The 3rd color component is static.
16158 Actual color components values present in video frame are displayed on graph.
16161 Similar as color2 but higher frequency of same values @code{x} and @code{y}
16162 on graph increases value of another color component, which is luminance by
16163 default values of @code{x} and @code{y}.
16166 Actual colors present in video frame are displayed on graph. If two different
16167 colors map to same position on graph then color with higher value of component
16168 not present in graph is picked.
16171 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
16172 component picked from radial gradient.
16176 Set which color component will be represented on X-axis. Default is @code{1}.
16179 Set which color component will be represented on Y-axis. Default is @code{2}.
16182 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
16183 of color component which represents frequency of (X, Y) location in graph.
16188 No envelope, this is default.
16191 Instant envelope, even darkest single pixel will be clearly highlighted.
16194 Hold maximum and minimum values presented in graph over time. This way you
16195 can still spot out of range values without constantly looking at vectorscope.
16198 Peak and instant envelope combined together.
16202 Set what kind of graticule to draw.
16210 Set graticule opacity.
16213 Set graticule flags.
16217 Draw graticule for white point.
16220 Draw graticule for black point.
16223 Draw color points short names.
16227 Set background opacity.
16229 @item lthreshold, l
16230 Set low threshold for color component not represented on X or Y axis.
16231 Values lower than this value will be ignored. Default is 0.
16232 Note this value is multiplied with actual max possible value one pixel component
16233 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
16236 @item hthreshold, h
16237 Set high threshold for color component not represented on X or Y axis.
16238 Values higher than this value will be ignored. Default is 1.
16239 Note this value is multiplied with actual max possible value one pixel component
16240 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
16241 is 0.9 * 255 = 230.
16243 @item colorspace, c
16244 Set what kind of colorspace to use when drawing graticule.
16253 @anchor{vidstabdetect}
16254 @section vidstabdetect
16256 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16257 @ref{vidstabtransform} for pass 2.
16259 This filter generates a file with relative translation and rotation
16260 transform information about subsequent frames, which is then used by
16261 the @ref{vidstabtransform} filter.
16263 To enable compilation of this filter you need to configure FFmpeg with
16264 @code{--enable-libvidstab}.
16266 This filter accepts the following options:
16270 Set the path to the file used to write the transforms information.
16271 Default value is @file{transforms.trf}.
16274 Set how shaky the video is and how quick the camera is. It accepts an
16275 integer in the range 1-10, a value of 1 means little shakiness, a
16276 value of 10 means strong shakiness. Default value is 5.
16279 Set the accuracy of the detection process. It must be a value in the
16280 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16281 accuracy. Default value is 15.
16284 Set stepsize of the search process. The region around minimum is
16285 scanned with 1 pixel resolution. Default value is 6.
16288 Set minimum contrast. Below this value a local measurement field is
16289 discarded. Must be a floating point value in the range 0-1. Default
16293 Set reference frame number for tripod mode.
16295 If enabled, the motion of the frames is compared to a reference frame
16296 in the filtered stream, identified by the specified number. The idea
16297 is to compensate all movements in a more-or-less static scene and keep
16298 the camera view absolutely still.
16300 If set to 0, it is disabled. The frames are counted starting from 1.
16303 Show fields and transforms in the resulting frames. It accepts an
16304 integer in the range 0-2. Default value is 0, which disables any
16308 @subsection Examples
16312 Use default values:
16318 Analyze strongly shaky movie and put the results in file
16319 @file{mytransforms.trf}:
16321 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16325 Visualize the result of internal transformations in the resulting
16328 vidstabdetect=show=1
16332 Analyze a video with medium shakiness using @command{ffmpeg}:
16334 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16338 @anchor{vidstabtransform}
16339 @section vidstabtransform
16341 Video stabilization/deshaking: pass 2 of 2,
16342 see @ref{vidstabdetect} for pass 1.
16344 Read a file with transform information for each frame and
16345 apply/compensate them. Together with the @ref{vidstabdetect}
16346 filter this can be used to deshake videos. See also
16347 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16348 the @ref{unsharp} filter, see below.
16350 To enable compilation of this filter you need to configure FFmpeg with
16351 @code{--enable-libvidstab}.
16353 @subsection Options
16357 Set path to the file used to read the transforms. Default value is
16358 @file{transforms.trf}.
16361 Set the number of frames (value*2 + 1) used for lowpass filtering the
16362 camera movements. Default value is 10.
16364 For example a number of 10 means that 21 frames are used (10 in the
16365 past and 10 in the future) to smoothen the motion in the video. A
16366 larger value leads to a smoother video, but limits the acceleration of
16367 the camera (pan/tilt movements). 0 is a special case where a static
16368 camera is simulated.
16371 Set the camera path optimization algorithm.
16373 Accepted values are:
16376 gaussian kernel low-pass filter on camera motion (default)
16378 averaging on transformations
16382 Set maximal number of pixels to translate frames. Default value is -1,
16386 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16387 value is -1, meaning no limit.
16390 Specify how to deal with borders that may be visible due to movement
16393 Available values are:
16396 keep image information from previous frame (default)
16398 fill the border black
16402 Invert transforms if set to 1. Default value is 0.
16405 Consider transforms as relative to previous frame if set to 1,
16406 absolute if set to 0. Default value is 0.
16409 Set percentage to zoom. A positive value will result in a zoom-in
16410 effect, a negative value in a zoom-out effect. Default value is 0 (no
16414 Set optimal zooming to avoid borders.
16416 Accepted values are:
16421 optimal static zoom value is determined (only very strong movements
16422 will lead to visible borders) (default)
16424 optimal adaptive zoom value is determined (no borders will be
16425 visible), see @option{zoomspeed}
16428 Note that the value given at zoom is added to the one calculated here.
16431 Set percent to zoom maximally each frame (enabled when
16432 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16436 Specify type of interpolation.
16438 Available values are:
16443 linear only horizontal
16445 linear in both directions (default)
16447 cubic in both directions (slow)
16451 Enable virtual tripod mode if set to 1, which is equivalent to
16452 @code{relative=0:smoothing=0}. Default value is 0.
16454 Use also @code{tripod} option of @ref{vidstabdetect}.
16457 Increase log verbosity if set to 1. Also the detected global motions
16458 are written to the temporary file @file{global_motions.trf}. Default
16462 @subsection Examples
16466 Use @command{ffmpeg} for a typical stabilization with default values:
16468 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16471 Note the use of the @ref{unsharp} filter which is always recommended.
16474 Zoom in a bit more and load transform data from a given file:
16476 vidstabtransform=zoom=5:input="mytransforms.trf"
16480 Smoothen the video even more:
16482 vidstabtransform=smoothing=30
16488 Flip the input video vertically.
16490 For example, to vertically flip a video with @command{ffmpeg}:
16492 ffmpeg -i in.avi -vf "vflip" out.avi
16497 Detect variable frame rate video.
16499 This filter tries to detect if the input is variable or constant frame rate.
16501 At end it will output number of frames detected as having variable delta pts,
16502 and ones with constant delta pts.
16503 If there was frames with variable delta, than it will also show min and max delta
16509 Make or reverse a natural vignetting effect.
16511 The filter accepts the following options:
16515 Set lens angle expression as a number of radians.
16517 The value is clipped in the @code{[0,PI/2]} range.
16519 Default value: @code{"PI/5"}
16523 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16527 Set forward/backward mode.
16529 Available modes are:
16532 The larger the distance from the central point, the darker the image becomes.
16535 The larger the distance from the central point, the brighter the image becomes.
16536 This can be used to reverse a vignette effect, though there is no automatic
16537 detection to extract the lens @option{angle} and other settings (yet). It can
16538 also be used to create a burning effect.
16541 Default value is @samp{forward}.
16544 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16546 It accepts the following values:
16549 Evaluate expressions only once during the filter initialization.
16552 Evaluate expressions for each incoming frame. This is way slower than the
16553 @samp{init} mode since it requires all the scalers to be re-computed, but it
16554 allows advanced dynamic expressions.
16557 Default value is @samp{init}.
16560 Set dithering to reduce the circular banding effects. Default is @code{1}
16564 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16565 Setting this value to the SAR of the input will make a rectangular vignetting
16566 following the dimensions of the video.
16568 Default is @code{1/1}.
16571 @subsection Expressions
16573 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16574 following parameters.
16579 input width and height
16582 the number of input frame, starting from 0
16585 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16586 @var{TB} units, NAN if undefined
16589 frame rate of the input video, NAN if the input frame rate is unknown
16592 the PTS (Presentation TimeStamp) of the filtered video frame,
16593 expressed in seconds, NAN if undefined
16596 time base of the input video
16600 @subsection Examples
16604 Apply simple strong vignetting effect:
16610 Make a flickering vignetting:
16612 vignette='PI/4+random(1)*PI/50':eval=frame
16617 @section vmafmotion
16619 Obtain the average vmaf motion score of a video.
16620 It is one of the component filters of VMAF.
16622 The obtained average motion score is printed through the logging system.
16624 In the below example the input file @file{ref.mpg} is being processed and score
16628 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16632 Stack input videos vertically.
16634 All streams must be of same pixel format and of same width.
16636 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16637 to create same output.
16639 The filter accept the following option:
16643 Set number of input streams. Default is 2.
16646 If set to 1, force the output to terminate when the shortest input
16647 terminates. Default value is 0.
16652 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
16653 Deinterlacing Filter").
16655 Based on the process described by Martin Weston for BBC R&D, and
16656 implemented based on the de-interlace algorithm written by Jim
16657 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
16658 uses filter coefficients calculated by BBC R&D.
16660 There are two sets of filter coefficients, so called "simple":
16661 and "complex". Which set of filter coefficients is used can
16662 be set by passing an optional parameter:
16666 Set the interlacing filter coefficients. Accepts one of the following values:
16670 Simple filter coefficient set.
16672 More-complex filter coefficient set.
16674 Default value is @samp{complex}.
16677 Specify which frames to deinterlace. Accept one of the following values:
16681 Deinterlace all frames,
16683 Only deinterlace frames marked as interlaced.
16686 Default value is @samp{all}.
16690 Video waveform monitor.
16692 The waveform monitor plots color component intensity. By default luminance
16693 only. Each column of the waveform corresponds to a column of pixels in the
16696 It accepts the following options:
16700 Can be either @code{row}, or @code{column}. Default is @code{column}.
16701 In row mode, the graph on the left side represents color component value 0 and
16702 the right side represents value = 255. In column mode, the top side represents
16703 color component value = 0 and bottom side represents value = 255.
16706 Set intensity. Smaller values are useful to find out how many values of the same
16707 luminance are distributed across input rows/columns.
16708 Default value is @code{0.04}. Allowed range is [0, 1].
16711 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
16712 In mirrored mode, higher values will be represented on the left
16713 side for @code{row} mode and at the top for @code{column} mode. Default is
16714 @code{1} (mirrored).
16718 It accepts the following values:
16721 Presents information identical to that in the @code{parade}, except
16722 that the graphs representing color components are superimposed directly
16725 This display mode makes it easier to spot relative differences or similarities
16726 in overlapping areas of the color components that are supposed to be identical,
16727 such as neutral whites, grays, or blacks.
16730 Display separate graph for the color components side by side in
16731 @code{row} mode or one below the other in @code{column} mode.
16734 Display separate graph for the color components side by side in
16735 @code{column} mode or one below the other in @code{row} mode.
16737 Using this display mode makes it easy to spot color casts in the highlights
16738 and shadows of an image, by comparing the contours of the top and the bottom
16739 graphs of each waveform. Since whites, grays, and blacks are characterized
16740 by exactly equal amounts of red, green, and blue, neutral areas of the picture
16741 should display three waveforms of roughly equal width/height. If not, the
16742 correction is easy to perform by making level adjustments the three waveforms.
16744 Default is @code{stack}.
16746 @item components, c
16747 Set which color components to display. Default is 1, which means only luminance
16748 or red color component if input is in RGB colorspace. If is set for example to
16749 7 it will display all 3 (if) available color components.
16754 No envelope, this is default.
16757 Instant envelope, minimum and maximum values presented in graph will be easily
16758 visible even with small @code{step} value.
16761 Hold minimum and maximum values presented in graph across time. This way you
16762 can still spot out of range values without constantly looking at waveforms.
16765 Peak and instant envelope combined together.
16771 No filtering, this is default.
16774 Luma and chroma combined together.
16777 Similar as above, but shows difference between blue and red chroma.
16780 Similar as above, but use different colors.
16783 Displays only chroma.
16786 Displays actual color value on waveform.
16789 Similar as above, but with luma showing frequency of chroma values.
16793 Set which graticule to display.
16797 Do not display graticule.
16800 Display green graticule showing legal broadcast ranges.
16803 Display orange graticule showing legal broadcast ranges.
16807 Set graticule opacity.
16810 Set graticule flags.
16814 Draw numbers above lines. By default enabled.
16817 Draw dots instead of lines.
16821 Set scale used for displaying graticule.
16828 Default is digital.
16831 Set background opacity.
16834 @section weave, doubleweave
16836 The @code{weave} takes a field-based video input and join
16837 each two sequential fields into single frame, producing a new double
16838 height clip with half the frame rate and half the frame count.
16840 The @code{doubleweave} works same as @code{weave} but without
16841 halving frame rate and frame count.
16843 It accepts the following option:
16847 Set first field. Available values are:
16851 Set the frame as top-field-first.
16854 Set the frame as bottom-field-first.
16858 @subsection Examples
16862 Interlace video using @ref{select} and @ref{separatefields} filter:
16864 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
16869 Apply the xBR high-quality magnification filter which is designed for pixel
16870 art. It follows a set of edge-detection rules, see
16871 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
16873 It accepts the following option:
16877 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
16878 @code{3xBR} and @code{4} for @code{4xBR}.
16879 Default is @code{3}.
16885 Deinterlace the input video ("yadif" means "yet another deinterlacing
16888 It accepts the following parameters:
16894 The interlacing mode to adopt. It accepts one of the following values:
16897 @item 0, send_frame
16898 Output one frame for each frame.
16899 @item 1, send_field
16900 Output one frame for each field.
16901 @item 2, send_frame_nospatial
16902 Like @code{send_frame}, but it skips the spatial interlacing check.
16903 @item 3, send_field_nospatial
16904 Like @code{send_field}, but it skips the spatial interlacing check.
16907 The default value is @code{send_frame}.
16910 The picture field parity assumed for the input interlaced video. It accepts one
16911 of the following values:
16915 Assume the top field is first.
16917 Assume the bottom field is first.
16919 Enable automatic detection of field parity.
16922 The default value is @code{auto}.
16923 If the interlacing is unknown or the decoder does not export this information,
16924 top field first will be assumed.
16927 Specify which frames to deinterlace. Accept one of the following
16932 Deinterlace all frames.
16933 @item 1, interlaced
16934 Only deinterlace frames marked as interlaced.
16937 The default value is @code{all}.
16942 Apply Zoom & Pan effect.
16944 This filter accepts the following options:
16948 Set the zoom expression. Default is 1.
16952 Set the x and y expression. Default is 0.
16955 Set the duration expression in number of frames.
16956 This sets for how many number of frames effect will last for
16957 single input image.
16960 Set the output image size, default is 'hd720'.
16963 Set the output frame rate, default is '25'.
16966 Each expression can contain the following constants:
16985 Output frame count.
16989 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16990 for current input frame.
16994 'x' and 'y' of last output frame of previous input frame or 0 when there was
16995 not yet such frame (first input frame).
16998 Last calculated zoom from 'z' expression for current input frame.
17001 Last calculated zoom of last output frame of previous input frame.
17004 Number of output frames for current input frame. Calculated from 'd' expression
17005 for each input frame.
17008 number of output frames created for previous input frame
17011 Rational number: input width / input height
17014 sample aspect ratio
17017 display aspect ratio
17021 @subsection Examples
17025 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
17027 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
17031 Zoom-in up to 1.5 and pan always at center of picture:
17033 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17037 Same as above but without pausing:
17039 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17045 Scale (resize) the input video, using the z.lib library:
17046 https://github.com/sekrit-twc/zimg.
17048 The zscale filter forces the output display aspect ratio to be the same
17049 as the input, by changing the output sample aspect ratio.
17051 If the input image format is different from the format requested by
17052 the next filter, the zscale filter will convert the input to the
17055 @subsection Options
17056 The filter accepts the following options.
17061 Set the output video dimension expression. Default value is the input
17064 If the @var{width} or @var{w} value is 0, the input width is used for
17065 the output. If the @var{height} or @var{h} value is 0, the input height
17066 is used for the output.
17068 If one and only one of the values is -n with n >= 1, the zscale filter
17069 will use a value that maintains the aspect ratio of the input image,
17070 calculated from the other specified dimension. After that it will,
17071 however, make sure that the calculated dimension is divisible by n and
17072 adjust the value if necessary.
17074 If both values are -n with n >= 1, the behavior will be identical to
17075 both values being set to 0 as previously detailed.
17077 See below for the list of accepted constants for use in the dimension
17081 Set the video size. For the syntax of this option, check the
17082 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17085 Set the dither type.
17087 Possible values are:
17092 @item error_diffusion
17098 Set the resize filter type.
17100 Possible values are:
17110 Default is bilinear.
17113 Set the color range.
17115 Possible values are:
17122 Default is same as input.
17125 Set the color primaries.
17127 Possible values are:
17137 Default is same as input.
17140 Set the transfer characteristics.
17142 Possible values are:
17156 Default is same as input.
17159 Set the colorspace matrix.
17161 Possible value are:
17172 Default is same as input.
17175 Set the input color range.
17177 Possible values are:
17184 Default is same as input.
17186 @item primariesin, pin
17187 Set the input color primaries.
17189 Possible values are:
17199 Default is same as input.
17201 @item transferin, tin
17202 Set the input transfer characteristics.
17204 Possible values are:
17215 Default is same as input.
17217 @item matrixin, min
17218 Set the input colorspace matrix.
17220 Possible value are:
17232 Set the output chroma location.
17234 Possible values are:
17245 @item chromalin, cin
17246 Set the input chroma location.
17248 Possible values are:
17260 Set the nominal peak luminance.
17263 The values of the @option{w} and @option{h} options are expressions
17264 containing the following constants:
17269 The input width and height
17273 These are the same as @var{in_w} and @var{in_h}.
17277 The output (scaled) width and height
17281 These are the same as @var{out_w} and @var{out_h}
17284 The same as @var{iw} / @var{ih}
17287 input sample aspect ratio
17290 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17294 horizontal and vertical input chroma subsample values. For example for the
17295 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17299 horizontal and vertical output chroma subsample values. For example for the
17300 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17306 @c man end VIDEO FILTERS
17308 @chapter Video Sources
17309 @c man begin VIDEO SOURCES
17311 Below is a description of the currently available video sources.
17315 Buffer video frames, and make them available to the filter chain.
17317 This source is mainly intended for a programmatic use, in particular
17318 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17320 It accepts the following parameters:
17325 Specify the size (width and height) of the buffered video frames. For the
17326 syntax of this option, check the
17327 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17330 The input video width.
17333 The input video height.
17336 A string representing the pixel format of the buffered video frames.
17337 It may be a number corresponding to a pixel format, or a pixel format
17341 Specify the timebase assumed by the timestamps of the buffered frames.
17344 Specify the frame rate expected for the video stream.
17346 @item pixel_aspect, sar
17347 The sample (pixel) aspect ratio of the input video.
17350 Specify the optional parameters to be used for the scale filter which
17351 is automatically inserted when an input change is detected in the
17352 input size or format.
17354 @item hw_frames_ctx
17355 When using a hardware pixel format, this should be a reference to an
17356 AVHWFramesContext describing input frames.
17361 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17364 will instruct the source to accept video frames with size 320x240 and
17365 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17366 square pixels (1:1 sample aspect ratio).
17367 Since the pixel format with name "yuv410p" corresponds to the number 6
17368 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17369 this example corresponds to:
17371 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17374 Alternatively, the options can be specified as a flat string, but this
17375 syntax is deprecated:
17377 @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}]
17381 Create a pattern generated by an elementary cellular automaton.
17383 The initial state of the cellular automaton can be defined through the
17384 @option{filename} and @option{pattern} options. If such options are
17385 not specified an initial state is created randomly.
17387 At each new frame a new row in the video is filled with the result of
17388 the cellular automaton next generation. The behavior when the whole
17389 frame is filled is defined by the @option{scroll} option.
17391 This source accepts the following options:
17395 Read the initial cellular automaton state, i.e. the starting row, from
17396 the specified file.
17397 In the file, each non-whitespace character is considered an alive
17398 cell, a newline will terminate the row, and further characters in the
17399 file will be ignored.
17402 Read the initial cellular automaton state, i.e. the starting row, from
17403 the specified string.
17405 Each non-whitespace character in the string is considered an alive
17406 cell, a newline will terminate the row, and further characters in the
17407 string will be ignored.
17410 Set the video rate, that is the number of frames generated per second.
17413 @item random_fill_ratio, ratio
17414 Set the random fill ratio for the initial cellular automaton row. It
17415 is a floating point number value ranging from 0 to 1, defaults to
17418 This option is ignored when a file or a pattern is specified.
17420 @item random_seed, seed
17421 Set the seed for filling randomly the initial row, must be an integer
17422 included between 0 and UINT32_MAX. If not specified, or if explicitly
17423 set to -1, the filter will try to use a good random seed on a best
17427 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17428 Default value is 110.
17431 Set the size of the output video. For the syntax of this option, check the
17432 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17434 If @option{filename} or @option{pattern} is specified, the size is set
17435 by default to the width of the specified initial state row, and the
17436 height is set to @var{width} * PHI.
17438 If @option{size} is set, it must contain the width of the specified
17439 pattern string, and the specified pattern will be centered in the
17442 If a filename or a pattern string is not specified, the size value
17443 defaults to "320x518" (used for a randomly generated initial state).
17446 If set to 1, scroll the output upward when all the rows in the output
17447 have been already filled. If set to 0, the new generated row will be
17448 written over the top row just after the bottom row is filled.
17451 @item start_full, full
17452 If set to 1, completely fill the output with generated rows before
17453 outputting the first frame.
17454 This is the default behavior, for disabling set the value to 0.
17457 If set to 1, stitch the left and right row edges together.
17458 This is the default behavior, for disabling set the value to 0.
17461 @subsection Examples
17465 Read the initial state from @file{pattern}, and specify an output of
17468 cellauto=f=pattern:s=200x400
17472 Generate a random initial row with a width of 200 cells, with a fill
17475 cellauto=ratio=2/3:s=200x200
17479 Create a pattern generated by rule 18 starting by a single alive cell
17480 centered on an initial row with width 100:
17482 cellauto=p=@@:s=100x400:full=0:rule=18
17486 Specify a more elaborated initial pattern:
17488 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17493 @anchor{coreimagesrc}
17494 @section coreimagesrc
17495 Video source generated on GPU using Apple's CoreImage API on OSX.
17497 This video source is a specialized version of the @ref{coreimage} video filter.
17498 Use a core image generator at the beginning of the applied filterchain to
17499 generate the content.
17501 The coreimagesrc video source accepts the following options:
17503 @item list_generators
17504 List all available generators along with all their respective options as well as
17505 possible minimum and maximum values along with the default values.
17507 list_generators=true
17511 Specify the size of the sourced video. For the syntax of this option, check the
17512 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17513 The default value is @code{320x240}.
17516 Specify the frame rate of the sourced video, as the number of frames
17517 generated per second. It has to be a string in the format
17518 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17519 number or a valid video frame rate abbreviation. The default value is
17523 Set the sample aspect ratio of the sourced video.
17526 Set the duration of the sourced video. See
17527 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17528 for the accepted syntax.
17530 If not specified, or the expressed duration is negative, the video is
17531 supposed to be generated forever.
17534 Additionally, all options of the @ref{coreimage} video filter are accepted.
17535 A complete filterchain can be used for further processing of the
17536 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17537 and examples for details.
17539 @subsection Examples
17544 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17545 given as complete and escaped command-line for Apple's standard bash shell:
17547 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17549 This example is equivalent to the QRCode example of @ref{coreimage} without the
17550 need for a nullsrc video source.
17554 @section mandelbrot
17556 Generate a Mandelbrot set fractal, and progressively zoom towards the
17557 point specified with @var{start_x} and @var{start_y}.
17559 This source accepts the following options:
17564 Set the terminal pts value. Default value is 400.
17567 Set the terminal scale value.
17568 Must be a floating point value. Default value is 0.3.
17571 Set the inner coloring mode, that is the algorithm used to draw the
17572 Mandelbrot fractal internal region.
17574 It shall assume one of the following values:
17579 Show time until convergence.
17581 Set color based on point closest to the origin of the iterations.
17586 Default value is @var{mincol}.
17589 Set the bailout value. Default value is 10.0.
17592 Set the maximum of iterations performed by the rendering
17593 algorithm. Default value is 7189.
17596 Set outer coloring mode.
17597 It shall assume one of following values:
17599 @item iteration_count
17600 Set iteration cound mode.
17601 @item normalized_iteration_count
17602 set normalized iteration count mode.
17604 Default value is @var{normalized_iteration_count}.
17607 Set frame rate, expressed as number of frames per second. Default
17611 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17612 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17615 Set the initial scale value. Default value is 3.0.
17618 Set the initial x position. Must be a floating point value between
17619 -100 and 100. Default value is -0.743643887037158704752191506114774.
17622 Set the initial y position. Must be a floating point value between
17623 -100 and 100. Default value is -0.131825904205311970493132056385139.
17628 Generate various test patterns, as generated by the MPlayer test filter.
17630 The size of the generated video is fixed, and is 256x256.
17631 This source is useful in particular for testing encoding features.
17633 This source accepts the following options:
17638 Specify the frame rate of the sourced video, as the number of frames
17639 generated per second. It has to be a string in the format
17640 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17641 number or a valid video frame rate abbreviation. The default value is
17645 Set the duration of the sourced video. See
17646 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17647 for the accepted syntax.
17649 If not specified, or the expressed duration is negative, the video is
17650 supposed to be generated forever.
17654 Set the number or the name of the test to perform. Supported tests are:
17670 Default value is "all", which will cycle through the list of all tests.
17675 mptestsrc=t=dc_luma
17678 will generate a "dc_luma" test pattern.
17680 @section frei0r_src
17682 Provide a frei0r source.
17684 To enable compilation of this filter you need to install the frei0r
17685 header and configure FFmpeg with @code{--enable-frei0r}.
17687 This source accepts the following parameters:
17692 The size of the video to generate. For the syntax of this option, check the
17693 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17696 The framerate of the generated video. It may be a string of the form
17697 @var{num}/@var{den} or a frame rate abbreviation.
17700 The name to the frei0r source to load. For more information regarding frei0r and
17701 how to set the parameters, read the @ref{frei0r} section in the video filters
17704 @item filter_params
17705 A '|'-separated list of parameters to pass to the frei0r source.
17709 For example, to generate a frei0r partik0l source with size 200x200
17710 and frame rate 10 which is overlaid on the overlay filter main input:
17712 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
17717 Generate a life pattern.
17719 This source is based on a generalization of John Conway's life game.
17721 The sourced input represents a life grid, each pixel represents a cell
17722 which can be in one of two possible states, alive or dead. Every cell
17723 interacts with its eight neighbours, which are the cells that are
17724 horizontally, vertically, or diagonally adjacent.
17726 At each interaction the grid evolves according to the adopted rule,
17727 which specifies the number of neighbor alive cells which will make a
17728 cell stay alive or born. The @option{rule} option allows one to specify
17731 This source accepts the following options:
17735 Set the file from which to read the initial grid state. In the file,
17736 each non-whitespace character is considered an alive cell, and newline
17737 is used to delimit the end of each row.
17739 If this option is not specified, the initial grid is generated
17743 Set the video rate, that is the number of frames generated per second.
17746 @item random_fill_ratio, ratio
17747 Set the random fill ratio for the initial random grid. It is a
17748 floating point number value ranging from 0 to 1, defaults to 1/PHI.
17749 It is ignored when a file is specified.
17751 @item random_seed, seed
17752 Set the seed for filling the initial random grid, must be an integer
17753 included between 0 and UINT32_MAX. If not specified, or if explicitly
17754 set to -1, the filter will try to use a good random seed on a best
17760 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
17761 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
17762 @var{NS} specifies the number of alive neighbor cells which make a
17763 live cell stay alive, and @var{NB} the number of alive neighbor cells
17764 which make a dead cell to become alive (i.e. to "born").
17765 "s" and "b" can be used in place of "S" and "B", respectively.
17767 Alternatively a rule can be specified by an 18-bits integer. The 9
17768 high order bits are used to encode the next cell state if it is alive
17769 for each number of neighbor alive cells, the low order bits specify
17770 the rule for "borning" new cells. Higher order bits encode for an
17771 higher number of neighbor cells.
17772 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
17773 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
17775 Default value is "S23/B3", which is the original Conway's game of life
17776 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
17777 cells, and will born a new cell if there are three alive cells around
17781 Set the size of the output video. For the syntax of this option, check the
17782 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17784 If @option{filename} is specified, the size is set by default to the
17785 same size of the input file. If @option{size} is set, it must contain
17786 the size specified in the input file, and the initial grid defined in
17787 that file is centered in the larger resulting area.
17789 If a filename is not specified, the size value defaults to "320x240"
17790 (used for a randomly generated initial grid).
17793 If set to 1, stitch the left and right grid edges together, and the
17794 top and bottom edges also. Defaults to 1.
17797 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
17798 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
17799 value from 0 to 255.
17802 Set the color of living (or new born) cells.
17805 Set the color of dead cells. If @option{mold} is set, this is the first color
17806 used to represent a dead cell.
17809 Set mold color, for definitely dead and moldy cells.
17811 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
17812 ffmpeg-utils manual,ffmpeg-utils}.
17815 @subsection Examples
17819 Read a grid from @file{pattern}, and center it on a grid of size
17822 life=f=pattern:s=300x300
17826 Generate a random grid of size 200x200, with a fill ratio of 2/3:
17828 life=ratio=2/3:s=200x200
17832 Specify a custom rule for evolving a randomly generated grid:
17838 Full example with slow death effect (mold) using @command{ffplay}:
17840 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
17847 @anchor{haldclutsrc}
17850 @anchor{pal100bars}
17851 @anchor{rgbtestsrc}
17853 @anchor{smptehdbars}
17856 @anchor{yuvtestsrc}
17857 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
17859 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
17861 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
17863 The @code{color} source provides an uniformly colored input.
17865 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
17866 @ref{haldclut} filter.
17868 The @code{nullsrc} source returns unprocessed video frames. It is
17869 mainly useful to be employed in analysis / debugging tools, or as the
17870 source for filters which ignore the input data.
17872 The @code{pal75bars} source generates a color bars pattern, based on
17873 EBU PAL recommendations with 75% color levels.
17875 The @code{pal100bars} source generates a color bars pattern, based on
17876 EBU PAL recommendations with 100% color levels.
17878 The @code{rgbtestsrc} source generates an RGB test pattern useful for
17879 detecting RGB vs BGR issues. You should see a red, green and blue
17880 stripe from top to bottom.
17882 The @code{smptebars} source generates a color bars pattern, based on
17883 the SMPTE Engineering Guideline EG 1-1990.
17885 The @code{smptehdbars} source generates a color bars pattern, based on
17886 the SMPTE RP 219-2002.
17888 The @code{testsrc} source generates a test video pattern, showing a
17889 color pattern, a scrolling gradient and a timestamp. This is mainly
17890 intended for testing purposes.
17892 The @code{testsrc2} source is similar to testsrc, but supports more
17893 pixel formats instead of just @code{rgb24}. This allows using it as an
17894 input for other tests without requiring a format conversion.
17896 The @code{yuvtestsrc} source generates an YUV test pattern. You should
17897 see a y, cb and cr stripe from top to bottom.
17899 The sources accept the following parameters:
17904 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17905 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17906 pixels to be used as identity matrix for 3D lookup tables. Each component is
17907 coded on a @code{1/(N*N)} scale.
17910 Specify the color of the source, only available in the @code{color}
17911 source. For the syntax of this option, check the
17912 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17915 Specify the size of the sourced video. For the syntax of this option, check the
17916 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17917 The default value is @code{320x240}.
17919 This option is not available with the @code{allrgb}, @code{allyuv}, and
17920 @code{haldclutsrc} filters.
17923 Specify the frame rate of the sourced video, as the number of frames
17924 generated per second. It has to be a string in the format
17925 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17926 number or a valid video frame rate abbreviation. The default value is
17930 Set the duration of the sourced video. See
17931 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17932 for the accepted syntax.
17934 If not specified, or the expressed duration is negative, the video is
17935 supposed to be generated forever.
17938 Set the sample aspect ratio of the sourced video.
17941 Specify the alpha (opacity) of the background, only available in the
17942 @code{testsrc2} source. The value must be between 0 (fully transparent) and
17943 255 (fully opaque, the default).
17946 Set the number of decimals to show in the timestamp, only available in the
17947 @code{testsrc} source.
17949 The displayed timestamp value will correspond to the original
17950 timestamp value multiplied by the power of 10 of the specified
17951 value. Default value is 0.
17954 @subsection Examples
17958 Generate a video with a duration of 5.3 seconds, with size
17959 176x144 and a frame rate of 10 frames per second:
17961 testsrc=duration=5.3:size=qcif:rate=10
17965 The following graph description will generate a red source
17966 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17969 color=c=red@@0.2:s=qcif:r=10
17973 If the input content is to be ignored, @code{nullsrc} can be used. The
17974 following command generates noise in the luminance plane by employing
17975 the @code{geq} filter:
17977 nullsrc=s=256x256, geq=random(1)*255:128:128
17981 @subsection Commands
17983 The @code{color} source supports the following commands:
17987 Set the color of the created image. Accepts the same syntax of the
17988 corresponding @option{color} option.
17993 Generate video using an OpenCL program.
17998 OpenCL program source file.
18001 Kernel name in program.
18004 Size of frames to generate. This must be set.
18007 Pixel format to use for the generated frames. This must be set.
18010 Number of frames generated every second. Default value is '25'.
18014 For details of how the program loading works, see the @ref{program_opencl}
18021 Generate a colour ramp by setting pixel values from the position of the pixel
18022 in the output image. (Note that this will work with all pixel formats, but
18023 the generated output will not be the same.)
18025 __kernel void ramp(__write_only image2d_t dst,
18026 unsigned int index)
18028 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18031 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
18033 write_imagef(dst, loc, val);
18038 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
18040 __kernel void sierpinski_carpet(__write_only image2d_t dst,
18041 unsigned int index)
18043 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18045 float4 value = 0.0f;
18046 int x = loc.x + index;
18047 int y = loc.y + index;
18048 while (x > 0 || y > 0) {
18049 if (x % 3 == 1 && y % 3 == 1) {
18057 write_imagef(dst, loc, value);
18063 @c man end VIDEO SOURCES
18065 @chapter Video Sinks
18066 @c man begin VIDEO SINKS
18068 Below is a description of the currently available video sinks.
18070 @section buffersink
18072 Buffer video frames, and make them available to the end of the filter
18075 This sink is mainly intended for programmatic use, in particular
18076 through the interface defined in @file{libavfilter/buffersink.h}
18077 or the options system.
18079 It accepts a pointer to an AVBufferSinkContext structure, which
18080 defines the incoming buffers' formats, to be passed as the opaque
18081 parameter to @code{avfilter_init_filter} for initialization.
18085 Null video sink: do absolutely nothing with the input video. It is
18086 mainly useful as a template and for use in analysis / debugging
18089 @c man end VIDEO SINKS
18091 @chapter Multimedia Filters
18092 @c man begin MULTIMEDIA FILTERS
18094 Below is a description of the currently available multimedia filters.
18098 Convert input audio to a video output, displaying the audio bit scope.
18100 The filter accepts the following options:
18104 Set frame rate, expressed as number of frames per second. Default
18108 Specify the video size for the output. For the syntax of this option, check the
18109 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18110 Default value is @code{1024x256}.
18113 Specify list of colors separated by space or by '|' which will be used to
18114 draw channels. Unrecognized or missing colors will be replaced
18118 @section ahistogram
18120 Convert input audio to a video output, displaying the volume histogram.
18122 The filter accepts the following options:
18126 Specify how histogram is calculated.
18128 It accepts the following values:
18131 Use single histogram for all channels.
18133 Use separate histogram for each channel.
18135 Default is @code{single}.
18138 Set frame rate, expressed as number of frames per second. Default
18142 Specify the video size for the output. For the syntax of this option, check the
18143 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18144 Default value is @code{hd720}.
18149 It accepts the following values:
18160 reverse logarithmic
18162 Default is @code{log}.
18165 Set amplitude scale.
18167 It accepts the following values:
18174 Default is @code{log}.
18177 Set how much frames to accumulate in histogram.
18178 Defauls is 1. Setting this to -1 accumulates all frames.
18181 Set histogram ratio of window height.
18184 Set sonogram sliding.
18186 It accepts the following values:
18189 replace old rows with new ones.
18191 scroll from top to bottom.
18193 Default is @code{replace}.
18196 @section aphasemeter
18198 Convert input audio to a video output, displaying the audio phase.
18200 The filter accepts the following options:
18204 Set the output frame rate. Default value is @code{25}.
18207 Set the video size for the output. For the syntax of this option, check the
18208 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18209 Default value is @code{800x400}.
18214 Specify the red, green, blue contrast. Default values are @code{2},
18215 @code{7} and @code{1}.
18216 Allowed range is @code{[0, 255]}.
18219 Set color which will be used for drawing median phase. If color is
18220 @code{none} which is default, no median phase value will be drawn.
18223 Enable video output. Default is enabled.
18226 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
18227 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
18228 The @code{-1} means left and right channels are completely out of phase and
18229 @code{1} means channels are in phase.
18231 @section avectorscope
18233 Convert input audio to a video output, representing the audio vector
18236 The filter is used to measure the difference between channels of stereo
18237 audio stream. A monoaural signal, consisting of identical left and right
18238 signal, results in straight vertical line. Any stereo separation is visible
18239 as a deviation from this line, creating a Lissajous figure.
18240 If the straight (or deviation from it) but horizontal line appears this
18241 indicates that the left and right channels are out of phase.
18243 The filter accepts the following options:
18247 Set the vectorscope mode.
18249 Available values are:
18252 Lissajous rotated by 45 degrees.
18255 Same as above but not rotated.
18258 Shape resembling half of circle.
18261 Default value is @samp{lissajous}.
18264 Set the video size for the output. For the syntax of this option, check the
18265 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18266 Default value is @code{400x400}.
18269 Set the output frame rate. Default value is @code{25}.
18275 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18276 @code{160}, @code{80} and @code{255}.
18277 Allowed range is @code{[0, 255]}.
18283 Specify the red, green, blue and alpha fade. Default values are @code{15},
18284 @code{10}, @code{5} and @code{5}.
18285 Allowed range is @code{[0, 255]}.
18288 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18289 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18292 Set the vectorscope drawing mode.
18294 Available values are:
18297 Draw dot for each sample.
18300 Draw line between previous and current sample.
18303 Default value is @samp{dot}.
18306 Specify amplitude scale of audio samples.
18308 Available values are:
18324 Swap left channel axis with right channel axis.
18334 Mirror only x axis.
18337 Mirror only y axis.
18345 @subsection Examples
18349 Complete example using @command{ffplay}:
18351 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18352 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18356 @section bench, abench
18358 Benchmark part of a filtergraph.
18360 The filter accepts the following options:
18364 Start or stop a timer.
18366 Available values are:
18369 Get the current time, set it as frame metadata (using the key
18370 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18373 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18374 the input frame metadata to get the time difference. Time difference, average,
18375 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18376 @code{min}) are then printed. The timestamps are expressed in seconds.
18380 @subsection Examples
18384 Benchmark @ref{selectivecolor} filter:
18386 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18392 Concatenate audio and video streams, joining them together one after the
18395 The filter works on segments of synchronized video and audio streams. All
18396 segments must have the same number of streams of each type, and that will
18397 also be the number of streams at output.
18399 The filter accepts the following options:
18404 Set the number of segments. Default is 2.
18407 Set the number of output video streams, that is also the number of video
18408 streams in each segment. Default is 1.
18411 Set the number of output audio streams, that is also the number of audio
18412 streams in each segment. Default is 0.
18415 Activate unsafe mode: do not fail if segments have a different format.
18419 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18420 @var{a} audio outputs.
18422 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18423 segment, in the same order as the outputs, then the inputs for the second
18426 Related streams do not always have exactly the same duration, for various
18427 reasons including codec frame size or sloppy authoring. For that reason,
18428 related synchronized streams (e.g. a video and its audio track) should be
18429 concatenated at once. The concat filter will use the duration of the longest
18430 stream in each segment (except the last one), and if necessary pad shorter
18431 audio streams with silence.
18433 For this filter to work correctly, all segments must start at timestamp 0.
18435 All corresponding streams must have the same parameters in all segments; the
18436 filtering system will automatically select a common pixel format for video
18437 streams, and a common sample format, sample rate and channel layout for
18438 audio streams, but other settings, such as resolution, must be converted
18439 explicitly by the user.
18441 Different frame rates are acceptable but will result in variable frame rate
18442 at output; be sure to configure the output file to handle it.
18444 @subsection Examples
18448 Concatenate an opening, an episode and an ending, all in bilingual version
18449 (video in stream 0, audio in streams 1 and 2):
18451 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18452 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18453 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18454 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18458 Concatenate two parts, handling audio and video separately, using the
18459 (a)movie sources, and adjusting the resolution:
18461 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18462 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18463 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18465 Note that a desync will happen at the stitch if the audio and video streams
18466 do not have exactly the same duration in the first file.
18470 @subsection Commands
18472 This filter supports the following commands:
18475 Close the current segment and step to the next one
18478 @section drawgraph, adrawgraph
18480 Draw a graph using input video or audio metadata.
18482 It accepts the following parameters:
18486 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18489 Set 1st foreground color expression.
18492 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18495 Set 2nd foreground color expression.
18498 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18501 Set 3rd foreground color expression.
18504 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18507 Set 4th foreground color expression.
18510 Set minimal value of metadata value.
18513 Set maximal value of metadata value.
18516 Set graph background color. Default is white.
18521 Available values for mode is:
18528 Default is @code{line}.
18533 Available values for slide is:
18536 Draw new frame when right border is reached.
18539 Replace old columns with new ones.
18542 Scroll from right to left.
18545 Scroll from left to right.
18548 Draw single picture.
18551 Default is @code{frame}.
18554 Set size of graph video. For the syntax of this option, check the
18555 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18556 The default value is @code{900x256}.
18558 The foreground color expressions can use the following variables:
18561 Minimal value of metadata value.
18564 Maximal value of metadata value.
18567 Current metadata key value.
18570 The color is defined as 0xAABBGGRR.
18573 Example using metadata from @ref{signalstats} filter:
18575 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18578 Example using metadata from @ref{ebur128} filter:
18580 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18586 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18587 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18588 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18589 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18591 The filter also has a video output (see the @var{video} option) with a real
18592 time graph to observe the loudness evolution. The graphic contains the logged
18593 message mentioned above, so it is not printed anymore when this option is set,
18594 unless the verbose logging is set. The main graphing area contains the
18595 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18596 the momentary loudness (400 milliseconds).
18598 More information about the Loudness Recommendation EBU R128 on
18599 @url{http://tech.ebu.ch/loudness}.
18601 The filter accepts the following options:
18606 Activate the video output. The audio stream is passed unchanged whether this
18607 option is set or no. The video stream will be the first output stream if
18608 activated. Default is @code{0}.
18611 Set the video size. This option is for video only. For the syntax of this
18613 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18614 Default and minimum resolution is @code{640x480}.
18617 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18618 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18619 other integer value between this range is allowed.
18622 Set metadata injection. If set to @code{1}, the audio input will be segmented
18623 into 100ms output frames, each of them containing various loudness information
18624 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18626 Default is @code{0}.
18629 Force the frame logging level.
18631 Available values are:
18634 information logging level
18636 verbose logging level
18639 By default, the logging level is set to @var{info}. If the @option{video} or
18640 the @option{metadata} options are set, it switches to @var{verbose}.
18645 Available modes can be cumulated (the option is a @code{flag} type). Possible
18649 Disable any peak mode (default).
18651 Enable sample-peak mode.
18653 Simple peak mode looking for the higher sample value. It logs a message
18654 for sample-peak (identified by @code{SPK}).
18656 Enable true-peak mode.
18658 If enabled, the peak lookup is done on an over-sampled version of the input
18659 stream for better peak accuracy. It logs a message for true-peak.
18660 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
18661 This mode requires a build with @code{libswresample}.
18665 Treat mono input files as "dual mono". If a mono file is intended for playback
18666 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
18667 If set to @code{true}, this option will compensate for this effect.
18668 Multi-channel input files are not affected by this option.
18671 Set a specific pan law to be used for the measurement of dual mono files.
18672 This parameter is optional, and has a default value of -3.01dB.
18675 @subsection Examples
18679 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
18681 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
18685 Run an analysis with @command{ffmpeg}:
18687 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
18691 @section interleave, ainterleave
18693 Temporally interleave frames from several inputs.
18695 @code{interleave} works with video inputs, @code{ainterleave} with audio.
18697 These filters read frames from several inputs and send the oldest
18698 queued frame to the output.
18700 Input streams must have well defined, monotonically increasing frame
18703 In order to submit one frame to output, these filters need to enqueue
18704 at least one frame for each input, so they cannot work in case one
18705 input is not yet terminated and will not receive incoming frames.
18707 For example consider the case when one input is a @code{select} filter
18708 which always drops input frames. The @code{interleave} filter will keep
18709 reading from that input, but it will never be able to send new frames
18710 to output until the input sends an end-of-stream signal.
18712 Also, depending on inputs synchronization, the filters will drop
18713 frames in case one input receives more frames than the other ones, and
18714 the queue is already filled.
18716 These filters accept the following options:
18720 Set the number of different inputs, it is 2 by default.
18723 @subsection Examples
18727 Interleave frames belonging to different streams using @command{ffmpeg}:
18729 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
18733 Add flickering blur effect:
18735 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
18739 @section metadata, ametadata
18741 Manipulate frame metadata.
18743 This filter accepts the following options:
18747 Set mode of operation of the filter.
18749 Can be one of the following:
18753 If both @code{value} and @code{key} is set, select frames
18754 which have such metadata. If only @code{key} is set, select
18755 every frame that has such key in metadata.
18758 Add new metadata @code{key} and @code{value}. If key is already available
18762 Modify value of already present key.
18765 If @code{value} is set, delete only keys that have such value.
18766 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
18770 Print key and its value if metadata was found. If @code{key} is not set print all
18771 metadata values available in frame.
18775 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
18778 Set metadata value which will be used. This option is mandatory for
18779 @code{modify} and @code{add} mode.
18782 Which function to use when comparing metadata value and @code{value}.
18784 Can be one of following:
18788 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
18791 Values are interpreted as strings, returns true if metadata value starts with
18792 the @code{value} option string.
18795 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
18798 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
18801 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
18804 Values are interpreted as floats, returns true if expression from option @code{expr}
18809 Set expression which is used when @code{function} is set to @code{expr}.
18810 The expression is evaluated through the eval API and can contain the following
18815 Float representation of @code{value} from metadata key.
18818 Float representation of @code{value} as supplied by user in @code{value} option.
18822 If specified in @code{print} mode, output is written to the named file. Instead of
18823 plain filename any writable url can be specified. Filename ``-'' is a shorthand
18824 for standard output. If @code{file} option is not set, output is written to the log
18825 with AV_LOG_INFO loglevel.
18829 @subsection Examples
18833 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
18836 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
18839 Print silencedetect output to file @file{metadata.txt}.
18841 silencedetect,ametadata=mode=print:file=metadata.txt
18844 Direct all metadata to a pipe with file descriptor 4.
18846 metadata=mode=print:file='pipe\:4'
18850 @section perms, aperms
18852 Set read/write permissions for the output frames.
18854 These filters are mainly aimed at developers to test direct path in the
18855 following filter in the filtergraph.
18857 The filters accept the following options:
18861 Select the permissions mode.
18863 It accepts the following values:
18866 Do nothing. This is the default.
18868 Set all the output frames read-only.
18870 Set all the output frames directly writable.
18872 Make the frame read-only if writable, and writable if read-only.
18874 Set each output frame read-only or writable randomly.
18878 Set the seed for the @var{random} mode, must be an integer included between
18879 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
18880 @code{-1}, the filter will try to use a good random seed on a best effort
18884 Note: in case of auto-inserted filter between the permission filter and the
18885 following one, the permission might not be received as expected in that
18886 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
18887 perms/aperms filter can avoid this problem.
18889 @section realtime, arealtime
18891 Slow down filtering to match real time approximately.
18893 These filters will pause the filtering for a variable amount of time to
18894 match the output rate with the input timestamps.
18895 They are similar to the @option{re} option to @code{ffmpeg}.
18897 They accept the following options:
18901 Time limit for the pauses. Any pause longer than that will be considered
18902 a timestamp discontinuity and reset the timer. Default is 2 seconds.
18906 @section select, aselect
18908 Select frames to pass in output.
18910 This filter accepts the following options:
18915 Set expression, which is evaluated for each input frame.
18917 If the expression is evaluated to zero, the frame is discarded.
18919 If the evaluation result is negative or NaN, the frame is sent to the
18920 first output; otherwise it is sent to the output with index
18921 @code{ceil(val)-1}, assuming that the input index starts from 0.
18923 For example a value of @code{1.2} corresponds to the output with index
18924 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
18927 Set the number of outputs. The output to which to send the selected
18928 frame is based on the result of the evaluation. Default value is 1.
18931 The expression can contain the following constants:
18935 The (sequential) number of the filtered frame, starting from 0.
18938 The (sequential) number of the selected frame, starting from 0.
18940 @item prev_selected_n
18941 The sequential number of the last selected frame. It's NAN if undefined.
18944 The timebase of the input timestamps.
18947 The PTS (Presentation TimeStamp) of the filtered video frame,
18948 expressed in @var{TB} units. It's NAN if undefined.
18951 The PTS of the filtered video frame,
18952 expressed in seconds. It's NAN if undefined.
18955 The PTS of the previously filtered video frame. It's NAN if undefined.
18957 @item prev_selected_pts
18958 The PTS of the last previously filtered video frame. It's NAN if undefined.
18960 @item prev_selected_t
18961 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
18964 The PTS of the first video frame in the video. It's NAN if undefined.
18967 The time of the first video frame in the video. It's NAN if undefined.
18969 @item pict_type @emph{(video only)}
18970 The type of the filtered frame. It can assume one of the following
18982 @item interlace_type @emph{(video only)}
18983 The frame interlace type. It can assume one of the following values:
18986 The frame is progressive (not interlaced).
18988 The frame is top-field-first.
18990 The frame is bottom-field-first.
18993 @item consumed_sample_n @emph{(audio only)}
18994 the number of selected samples before the current frame
18996 @item samples_n @emph{(audio only)}
18997 the number of samples in the current frame
18999 @item sample_rate @emph{(audio only)}
19000 the input sample rate
19003 This is 1 if the filtered frame is a key-frame, 0 otherwise.
19006 the position in the file of the filtered frame, -1 if the information
19007 is not available (e.g. for synthetic video)
19009 @item scene @emph{(video only)}
19010 value between 0 and 1 to indicate a new scene; a low value reflects a low
19011 probability for the current frame to introduce a new scene, while a higher
19012 value means the current frame is more likely to be one (see the example below)
19014 @item concatdec_select
19015 The concat demuxer can select only part of a concat input file by setting an
19016 inpoint and an outpoint, but the output packets may not be entirely contained
19017 in the selected interval. By using this variable, it is possible to skip frames
19018 generated by the concat demuxer which are not exactly contained in the selected
19021 This works by comparing the frame pts against the @var{lavf.concat.start_time}
19022 and the @var{lavf.concat.duration} packet metadata values which are also
19023 present in the decoded frames.
19025 The @var{concatdec_select} variable is -1 if the frame pts is at least
19026 start_time and either the duration metadata is missing or the frame pts is less
19027 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
19030 That basically means that an input frame is selected if its pts is within the
19031 interval set by the concat demuxer.
19035 The default value of the select expression is "1".
19037 @subsection Examples
19041 Select all frames in input:
19046 The example above is the same as:
19058 Select only I-frames:
19060 select='eq(pict_type\,I)'
19064 Select one frame every 100:
19066 select='not(mod(n\,100))'
19070 Select only frames contained in the 10-20 time interval:
19072 select=between(t\,10\,20)
19076 Select only I-frames contained in the 10-20 time interval:
19078 select=between(t\,10\,20)*eq(pict_type\,I)
19082 Select frames with a minimum distance of 10 seconds:
19084 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
19088 Use aselect to select only audio frames with samples number > 100:
19090 aselect='gt(samples_n\,100)'
19094 Create a mosaic of the first scenes:
19096 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
19099 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
19103 Send even and odd frames to separate outputs, and compose them:
19105 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
19109 Select useful frames from an ffconcat file which is using inpoints and
19110 outpoints but where the source files are not intra frame only.
19112 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
19116 @section sendcmd, asendcmd
19118 Send commands to filters in the filtergraph.
19120 These filters read commands to be sent to other filters in the
19123 @code{sendcmd} must be inserted between two video filters,
19124 @code{asendcmd} must be inserted between two audio filters, but apart
19125 from that they act the same way.
19127 The specification of commands can be provided in the filter arguments
19128 with the @var{commands} option, or in a file specified by the
19129 @var{filename} option.
19131 These filters accept the following options:
19134 Set the commands to be read and sent to the other filters.
19136 Set the filename of the commands to be read and sent to the other
19140 @subsection Commands syntax
19142 A commands description consists of a sequence of interval
19143 specifications, comprising a list of commands to be executed when a
19144 particular event related to that interval occurs. The occurring event
19145 is typically the current frame time entering or leaving a given time
19148 An interval is specified by the following syntax:
19150 @var{START}[-@var{END}] @var{COMMANDS};
19153 The time interval is specified by the @var{START} and @var{END} times.
19154 @var{END} is optional and defaults to the maximum time.
19156 The current frame time is considered within the specified interval if
19157 it is included in the interval [@var{START}, @var{END}), that is when
19158 the time is greater or equal to @var{START} and is lesser than
19161 @var{COMMANDS} consists of a sequence of one or more command
19162 specifications, separated by ",", relating to that interval. The
19163 syntax of a command specification is given by:
19165 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
19168 @var{FLAGS} is optional and specifies the type of events relating to
19169 the time interval which enable sending the specified command, and must
19170 be a non-null sequence of identifier flags separated by "+" or "|" and
19171 enclosed between "[" and "]".
19173 The following flags are recognized:
19176 The command is sent when the current frame timestamp enters the
19177 specified interval. In other words, the command is sent when the
19178 previous frame timestamp was not in the given interval, and the
19182 The command is sent when the current frame timestamp leaves the
19183 specified interval. In other words, the command is sent when the
19184 previous frame timestamp was in the given interval, and the
19188 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
19191 @var{TARGET} specifies the target of the command, usually the name of
19192 the filter class or a specific filter instance name.
19194 @var{COMMAND} specifies the name of the command for the target filter.
19196 @var{ARG} is optional and specifies the optional list of argument for
19197 the given @var{COMMAND}.
19199 Between one interval specification and another, whitespaces, or
19200 sequences of characters starting with @code{#} until the end of line,
19201 are ignored and can be used to annotate comments.
19203 A simplified BNF description of the commands specification syntax
19206 @var{COMMAND_FLAG} ::= "enter" | "leave"
19207 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
19208 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
19209 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
19210 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
19211 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
19214 @subsection Examples
19218 Specify audio tempo change at second 4:
19220 asendcmd=c='4.0 atempo tempo 1.5',atempo
19224 Target a specific filter instance:
19226 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
19230 Specify a list of drawtext and hue commands in a file.
19232 # show text in the interval 5-10
19233 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
19234 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
19236 # desaturate the image in the interval 15-20
19237 15.0-20.0 [enter] hue s 0,
19238 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
19240 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
19242 # apply an exponential saturation fade-out effect, starting from time 25
19243 25 [enter] hue s exp(25-t)
19246 A filtergraph allowing to read and process the above command list
19247 stored in a file @file{test.cmd}, can be specified with:
19249 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
19254 @section setpts, asetpts
19256 Change the PTS (presentation timestamp) of the input frames.
19258 @code{setpts} works on video frames, @code{asetpts} on audio frames.
19260 This filter accepts the following options:
19265 The expression which is evaluated for each frame to construct its timestamp.
19269 The expression is evaluated through the eval API and can contain the following
19274 frame rate, only defined for constant frame-rate video
19277 The presentation timestamp in input
19280 The count of the input frame for video or the number of consumed samples,
19281 not including the current frame for audio, starting from 0.
19283 @item NB_CONSUMED_SAMPLES
19284 The number of consumed samples, not including the current frame (only
19287 @item NB_SAMPLES, S
19288 The number of samples in the current frame (only audio)
19290 @item SAMPLE_RATE, SR
19291 The audio sample rate.
19294 The PTS of the first frame.
19297 the time in seconds of the first frame
19300 State whether the current frame is interlaced.
19303 the time in seconds of the current frame
19306 original position in the file of the frame, or undefined if undefined
19307 for the current frame
19310 The previous input PTS.
19313 previous input time in seconds
19316 The previous output PTS.
19319 previous output time in seconds
19322 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19326 The wallclock (RTC) time at the start of the movie in microseconds.
19329 The timebase of the input timestamps.
19333 @subsection Examples
19337 Start counting PTS from zero
19339 setpts=PTS-STARTPTS
19343 Apply fast motion effect:
19349 Apply slow motion effect:
19355 Set fixed rate of 25 frames per second:
19361 Set fixed rate 25 fps with some jitter:
19363 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19367 Apply an offset of 10 seconds to the input PTS:
19373 Generate timestamps from a "live source" and rebase onto the current timebase:
19375 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19379 Generate timestamps by counting samples:
19388 Force color range for the output video frame.
19390 The @code{setrange} filter marks the color range property for the
19391 output frames. It does not change the input frame, but only sets the
19392 corresponding property, which affects how the frame is treated by
19395 The filter accepts the following options:
19400 Available values are:
19404 Keep the same color range property.
19406 @item unspecified, unknown
19407 Set the color range as unspecified.
19409 @item limited, tv, mpeg
19410 Set the color range as limited.
19412 @item full, pc, jpeg
19413 Set the color range as full.
19417 @section settb, asettb
19419 Set the timebase to use for the output frames timestamps.
19420 It is mainly useful for testing timebase configuration.
19422 It accepts the following parameters:
19427 The expression which is evaluated into the output timebase.
19431 The value for @option{tb} is an arithmetic expression representing a
19432 rational. The expression can contain the constants "AVTB" (the default
19433 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19434 audio only). Default value is "intb".
19436 @subsection Examples
19440 Set the timebase to 1/25:
19446 Set the timebase to 1/10:
19452 Set the timebase to 1001/1000:
19458 Set the timebase to 2*intb:
19464 Set the default timebase value:
19471 Convert input audio to a video output representing frequency spectrum
19472 logarithmically using Brown-Puckette constant Q transform algorithm with
19473 direct frequency domain coefficient calculation (but the transform itself
19474 is not really constant Q, instead the Q factor is actually variable/clamped),
19475 with musical tone scale, from E0 to D#10.
19477 The filter accepts the following options:
19481 Specify the video size for the output. It must be even. For the syntax of this option,
19482 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19483 Default value is @code{1920x1080}.
19486 Set the output frame rate. Default value is @code{25}.
19489 Set the bargraph height. It must be even. Default value is @code{-1} which
19490 computes the bargraph height automatically.
19493 Set the axis height. It must be even. Default value is @code{-1} which computes
19494 the axis height automatically.
19497 Set the sonogram height. It must be even. Default value is @code{-1} which
19498 computes the sonogram height automatically.
19501 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19502 instead. Default value is @code{1}.
19504 @item sono_v, volume
19505 Specify the sonogram volume expression. It can contain variables:
19508 the @var{bar_v} evaluated expression
19509 @item frequency, freq, f
19510 the frequency where it is evaluated
19511 @item timeclamp, tc
19512 the value of @var{timeclamp} option
19516 @item a_weighting(f)
19517 A-weighting of equal loudness
19518 @item b_weighting(f)
19519 B-weighting of equal loudness
19520 @item c_weighting(f)
19521 C-weighting of equal loudness.
19523 Default value is @code{16}.
19525 @item bar_v, volume2
19526 Specify the bargraph volume expression. It can contain variables:
19529 the @var{sono_v} evaluated expression
19530 @item frequency, freq, f
19531 the frequency where it is evaluated
19532 @item timeclamp, tc
19533 the value of @var{timeclamp} option
19537 @item a_weighting(f)
19538 A-weighting of equal loudness
19539 @item b_weighting(f)
19540 B-weighting of equal loudness
19541 @item c_weighting(f)
19542 C-weighting of equal loudness.
19544 Default value is @code{sono_v}.
19546 @item sono_g, gamma
19547 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19548 higher gamma makes the spectrum having more range. Default value is @code{3}.
19549 Acceptable range is @code{[1, 7]}.
19551 @item bar_g, gamma2
19552 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19556 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19557 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19559 @item timeclamp, tc
19560 Specify the transform timeclamp. At low frequency, there is trade-off between
19561 accuracy in time domain and frequency domain. If timeclamp is lower,
19562 event in time domain is represented more accurately (such as fast bass drum),
19563 otherwise event in frequency domain is represented more accurately
19564 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19567 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19568 limits future samples by applying asymmetric windowing in time domain, useful
19569 when low latency is required. Accepted range is @code{[0, 1]}.
19572 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19573 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19576 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19577 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19580 This option is deprecated and ignored.
19583 Specify the transform length in time domain. Use this option to control accuracy
19584 trade-off between time domain and frequency domain at every frequency sample.
19585 It can contain variables:
19587 @item frequency, freq, f
19588 the frequency where it is evaluated
19589 @item timeclamp, tc
19590 the value of @var{timeclamp} option.
19592 Default value is @code{384*tc/(384+tc*f)}.
19595 Specify the transform count for every video frame. Default value is @code{6}.
19596 Acceptable range is @code{[1, 30]}.
19599 Specify the transform count for every single pixel. Default value is @code{0},
19600 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19603 Specify font file for use with freetype to draw the axis. If not specified,
19604 use embedded font. Note that drawing with font file or embedded font is not
19605 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19609 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19610 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19613 Specify font color expression. This is arithmetic expression that should return
19614 integer value 0xRRGGBB. It can contain variables:
19616 @item frequency, freq, f
19617 the frequency where it is evaluated
19618 @item timeclamp, tc
19619 the value of @var{timeclamp} option
19624 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19625 @item r(x), g(x), b(x)
19626 red, green, and blue value of intensity x.
19628 Default value is @code{st(0, (midi(f)-59.5)/12);
19629 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19630 r(1-ld(1)) + b(ld(1))}.
19633 Specify image file to draw the axis. This option override @var{fontfile} and
19634 @var{fontcolor} option.
19637 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19638 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19639 Default value is @code{1}.
19642 Set colorspace. The accepted values are:
19645 Unspecified (default)
19654 BT.470BG or BT.601-6 625
19657 SMPTE-170M or BT.601-6 525
19663 BT.2020 with non-constant luminance
19668 Set spectrogram color scheme. This is list of floating point values with format
19669 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
19670 The default is @code{1|0.5|0|0|0.5|1}.
19674 @subsection Examples
19678 Playing audio while showing the spectrum:
19680 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
19684 Same as above, but with frame rate 30 fps:
19686 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
19690 Playing at 1280x720:
19692 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
19696 Disable sonogram display:
19702 A1 and its harmonics: A1, A2, (near)E3, A3:
19704 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),
19705 asplit[a][out1]; [a] showcqt [out0]'
19709 Same as above, but with more accuracy in frequency domain:
19711 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),
19712 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
19718 bar_v=10:sono_v=bar_v*a_weighting(f)
19722 Custom gamma, now spectrum is linear to the amplitude.
19728 Custom tlength equation:
19730 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)))'
19734 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
19736 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
19740 Custom font using fontconfig:
19742 font='Courier New,Monospace,mono|bold'
19746 Custom frequency range with custom axis using image file:
19748 axisfile=myaxis.png:basefreq=40:endfreq=10000
19754 Convert input audio to video output representing the audio power spectrum.
19755 Audio amplitude is on Y-axis while frequency is on X-axis.
19757 The filter accepts the following options:
19761 Specify size of video. For the syntax of this option, check the
19762 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19763 Default is @code{1024x512}.
19767 This set how each frequency bin will be represented.
19769 It accepts the following values:
19775 Default is @code{bar}.
19778 Set amplitude scale.
19780 It accepts the following values:
19794 Default is @code{log}.
19797 Set frequency scale.
19799 It accepts the following values:
19808 Reverse logarithmic scale.
19810 Default is @code{lin}.
19815 It accepts the following values:
19831 Default is @code{w2048}
19834 Set windowing function.
19836 It accepts the following values:
19858 Default is @code{hanning}.
19861 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19862 which means optimal overlap for selected window function will be picked.
19865 Set time averaging. Setting this to 0 will display current maximal peaks.
19866 Default is @code{1}, which means time averaging is disabled.
19869 Specify list of colors separated by space or by '|' which will be used to
19870 draw channel frequencies. Unrecognized or missing colors will be replaced
19874 Set channel display mode.
19876 It accepts the following values:
19881 Default is @code{combined}.
19884 Set minimum amplitude used in @code{log} amplitude scaler.
19888 @anchor{showspectrum}
19889 @section showspectrum
19891 Convert input audio to a video output, representing the audio frequency
19894 The filter accepts the following options:
19898 Specify the video size for the output. For the syntax of this option, check the
19899 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19900 Default value is @code{640x512}.
19903 Specify how the spectrum should slide along the window.
19905 It accepts the following values:
19908 the samples start again on the left when they reach the right
19910 the samples scroll from right to left
19912 frames are only produced when the samples reach the right
19914 the samples scroll from left to right
19917 Default value is @code{replace}.
19920 Specify display mode.
19922 It accepts the following values:
19925 all channels are displayed in the same row
19927 all channels are displayed in separate rows
19930 Default value is @samp{combined}.
19933 Specify display color mode.
19935 It accepts the following values:
19938 each channel is displayed in a separate color
19940 each channel is displayed using the same color scheme
19942 each channel is displayed using the rainbow color scheme
19944 each channel is displayed using the moreland color scheme
19946 each channel is displayed using the nebulae color scheme
19948 each channel is displayed using the fire color scheme
19950 each channel is displayed using the fiery color scheme
19952 each channel is displayed using the fruit color scheme
19954 each channel is displayed using the cool color scheme
19957 Default value is @samp{channel}.
19960 Specify scale used for calculating intensity color values.
19962 It accepts the following values:
19967 square root, default
19978 Default value is @samp{sqrt}.
19981 Set saturation modifier for displayed colors. Negative values provide
19982 alternative color scheme. @code{0} is no saturation at all.
19983 Saturation must be in [-10.0, 10.0] range.
19984 Default value is @code{1}.
19987 Set window function.
19989 It accepts the following values:
20013 Default value is @code{hann}.
20016 Set orientation of time vs frequency axis. Can be @code{vertical} or
20017 @code{horizontal}. Default is @code{vertical}.
20020 Set ratio of overlap window. Default value is @code{0}.
20021 When value is @code{1} overlap is set to recommended size for specific
20022 window function currently used.
20025 Set scale gain for calculating intensity color values.
20026 Default value is @code{1}.
20029 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
20032 Set color rotation, must be in [-1.0, 1.0] range.
20033 Default value is @code{0}.
20036 The usage is very similar to the showwaves filter; see the examples in that
20039 @subsection Examples
20043 Large window with logarithmic color scaling:
20045 showspectrum=s=1280x480:scale=log
20049 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
20051 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20052 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
20056 @section showspectrumpic
20058 Convert input audio to a single video frame, representing the audio frequency
20061 The filter accepts the following options:
20065 Specify the video size for the output. For the syntax of this option, check the
20066 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20067 Default value is @code{4096x2048}.
20070 Specify display mode.
20072 It accepts the following values:
20075 all channels are displayed in the same row
20077 all channels are displayed in separate rows
20079 Default value is @samp{combined}.
20082 Specify display color mode.
20084 It accepts the following values:
20087 each channel is displayed in a separate color
20089 each channel is displayed using the same color scheme
20091 each channel is displayed using the rainbow color scheme
20093 each channel is displayed using the moreland color scheme
20095 each channel is displayed using the nebulae color scheme
20097 each channel is displayed using the fire color scheme
20099 each channel is displayed using the fiery color scheme
20101 each channel is displayed using the fruit color scheme
20103 each channel is displayed using the cool color scheme
20105 Default value is @samp{intensity}.
20108 Specify scale used for calculating intensity color values.
20110 It accepts the following values:
20115 square root, default
20125 Default value is @samp{log}.
20128 Set saturation modifier for displayed colors. Negative values provide
20129 alternative color scheme. @code{0} is no saturation at all.
20130 Saturation must be in [-10.0, 10.0] range.
20131 Default value is @code{1}.
20134 Set window function.
20136 It accepts the following values:
20159 Default value is @code{hann}.
20162 Set orientation of time vs frequency axis. Can be @code{vertical} or
20163 @code{horizontal}. Default is @code{vertical}.
20166 Set scale gain for calculating intensity color values.
20167 Default value is @code{1}.
20170 Draw time and frequency axes and legends. Default is enabled.
20173 Set color rotation, must be in [-1.0, 1.0] range.
20174 Default value is @code{0}.
20177 @subsection Examples
20181 Extract an audio spectrogram of a whole audio track
20182 in a 1024x1024 picture using @command{ffmpeg}:
20184 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
20188 @section showvolume
20190 Convert input audio volume to a video output.
20192 The filter accepts the following options:
20199 Set border width, allowed range is [0, 5]. Default is 1.
20202 Set channel width, allowed range is [80, 8192]. Default is 400.
20205 Set channel height, allowed range is [1, 900]. Default is 20.
20208 Set fade, allowed range is [0, 1]. Default is 0.95.
20211 Set volume color expression.
20213 The expression can use the following variables:
20217 Current max volume of channel in dB.
20223 Current channel number, starting from 0.
20227 If set, displays channel names. Default is enabled.
20230 If set, displays volume values. Default is enabled.
20233 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
20234 default is @code{h}.
20237 Set step size, allowed range is [0, 5]. Default is 0, which means
20241 Set background opacity, allowed range is [0, 1]. Default is 0.
20244 Set metering mode, can be peak: @code{p} or rms: @code{r},
20245 default is @code{p}.
20248 Set display scale, can be linear: @code{lin} or log: @code{log},
20249 default is @code{lin}.
20253 If set to > 0., display a line for the max level
20254 in the previous seconds.
20255 default is disabled: @code{0.}
20258 The color of the max line. Use when @code{dm} option is set to > 0.
20259 default is: @code{orange}
20264 Convert input audio to a video output, representing the samples waves.
20266 The filter accepts the following options:
20270 Specify the video size for the output. For the syntax of this option, check the
20271 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20272 Default value is @code{600x240}.
20277 Available values are:
20280 Draw a point for each sample.
20283 Draw a vertical line for each sample.
20286 Draw a point for each sample and a line between them.
20289 Draw a centered vertical line for each sample.
20292 Default value is @code{point}.
20295 Set the number of samples which are printed on the same column. A
20296 larger value will decrease the frame rate. Must be a positive
20297 integer. This option can be set only if the value for @var{rate}
20298 is not explicitly specified.
20301 Set the (approximate) output frame rate. This is done by setting the
20302 option @var{n}. Default value is "25".
20304 @item split_channels
20305 Set if channels should be drawn separately or overlap. Default value is 0.
20308 Set colors separated by '|' which are going to be used for drawing of each channel.
20311 Set amplitude scale.
20313 Available values are:
20331 Set the draw mode. This is mostly useful to set for high @var{n}.
20333 Available values are:
20336 Scale pixel values for each drawn sample.
20339 Draw every sample directly.
20342 Default value is @code{scale}.
20345 @subsection Examples
20349 Output the input file audio and the corresponding video representation
20352 amovie=a.mp3,asplit[out0],showwaves[out1]
20356 Create a synthetic signal and show it with showwaves, forcing a
20357 frame rate of 30 frames per second:
20359 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20363 @section showwavespic
20365 Convert input audio to a single video frame, representing the samples waves.
20367 The filter accepts the following options:
20371 Specify the video size for the output. For the syntax of this option, check the
20372 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20373 Default value is @code{600x240}.
20375 @item split_channels
20376 Set if channels should be drawn separately or overlap. Default value is 0.
20379 Set colors separated by '|' which are going to be used for drawing of each channel.
20382 Set amplitude scale.
20384 Available values are:
20402 @subsection Examples
20406 Extract a channel split representation of the wave form of a whole audio track
20407 in a 1024x800 picture using @command{ffmpeg}:
20409 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20413 @section sidedata, asidedata
20415 Delete frame side data, or select frames based on it.
20417 This filter accepts the following options:
20421 Set mode of operation of the filter.
20423 Can be one of the following:
20427 Select every frame with side data of @code{type}.
20430 Delete side data of @code{type}. If @code{type} is not set, delete all side
20436 Set side data type used with all modes. Must be set for @code{select} mode. For
20437 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20438 in @file{libavutil/frame.h}. For example, to choose
20439 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20443 @section spectrumsynth
20445 Sythesize audio from 2 input video spectrums, first input stream represents
20446 magnitude across time and second represents phase across time.
20447 The filter will transform from frequency domain as displayed in videos back
20448 to time domain as presented in audio output.
20450 This filter is primarily created for reversing processed @ref{showspectrum}
20451 filter outputs, but can synthesize sound from other spectrograms too.
20452 But in such case results are going to be poor if the phase data is not
20453 available, because in such cases phase data need to be recreated, usually
20454 its just recreated from random noise.
20455 For best results use gray only output (@code{channel} color mode in
20456 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20457 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20458 @code{data} option. Inputs videos should generally use @code{fullframe}
20459 slide mode as that saves resources needed for decoding video.
20461 The filter accepts the following options:
20465 Specify sample rate of output audio, the sample rate of audio from which
20466 spectrum was generated may differ.
20469 Set number of channels represented in input video spectrums.
20472 Set scale which was used when generating magnitude input spectrum.
20473 Can be @code{lin} or @code{log}. Default is @code{log}.
20476 Set slide which was used when generating inputs spectrums.
20477 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20478 Default is @code{fullframe}.
20481 Set window function used for resynthesis.
20484 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20485 which means optimal overlap for selected window function will be picked.
20488 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20489 Default is @code{vertical}.
20492 @subsection Examples
20496 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20497 then resynthesize videos back to audio with spectrumsynth:
20499 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
20500 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
20501 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20505 @section split, asplit
20507 Split input into several identical outputs.
20509 @code{asplit} works with audio input, @code{split} with video.
20511 The filter accepts a single parameter which specifies the number of outputs. If
20512 unspecified, it defaults to 2.
20514 @subsection Examples
20518 Create two separate outputs from the same input:
20520 [in] split [out0][out1]
20524 To create 3 or more outputs, you need to specify the number of
20527 [in] asplit=3 [out0][out1][out2]
20531 Create two separate outputs from the same input, one cropped and
20534 [in] split [splitout1][splitout2];
20535 [splitout1] crop=100:100:0:0 [cropout];
20536 [splitout2] pad=200:200:100:100 [padout];
20540 Create 5 copies of the input audio with @command{ffmpeg}:
20542 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20548 Receive commands sent through a libzmq client, and forward them to
20549 filters in the filtergraph.
20551 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20552 must be inserted between two video filters, @code{azmq} between two
20553 audio filters. Both are capable to send messages to any filter type.
20555 To enable these filters you need to install the libzmq library and
20556 headers and configure FFmpeg with @code{--enable-libzmq}.
20558 For more information about libzmq see:
20559 @url{http://www.zeromq.org/}
20561 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20562 receives messages sent through a network interface defined by the
20563 @option{bind_address} (or the abbreviation "@option{b}") option.
20564 Default value of this option is @file{tcp://localhost:5555}. You may
20565 want to alter this value to your needs, but do not forget to escape any
20566 ':' signs (see @ref{filtergraph escaping}).
20568 The received message must be in the form:
20570 @var{TARGET} @var{COMMAND} [@var{ARG}]
20573 @var{TARGET} specifies the target of the command, usually the name of
20574 the filter class or a specific filter instance name. The default
20575 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
20576 but you can override this by using the @samp{filter_name@@id} syntax
20577 (see @ref{Filtergraph syntax}).
20579 @var{COMMAND} specifies the name of the command for the target filter.
20581 @var{ARG} is optional and specifies the optional argument list for the
20582 given @var{COMMAND}.
20584 Upon reception, the message is processed and the corresponding command
20585 is injected into the filtergraph. Depending on the result, the filter
20586 will send a reply to the client, adopting the format:
20588 @var{ERROR_CODE} @var{ERROR_REASON}
20592 @var{MESSAGE} is optional.
20594 @subsection Examples
20596 Look at @file{tools/zmqsend} for an example of a zmq client which can
20597 be used to send commands processed by these filters.
20599 Consider the following filtergraph generated by @command{ffplay}.
20600 In this example the last overlay filter has an instance name. All other
20601 filters will have default instance names.
20604 ffplay -dumpgraph 1 -f lavfi "
20605 color=s=100x100:c=red [l];
20606 color=s=100x100:c=blue [r];
20607 nullsrc=s=200x100, zmq [bg];
20608 [bg][l] overlay [bg+l];
20609 [bg+l][r] overlay@@my=x=100 "
20612 To change the color of the left side of the video, the following
20613 command can be used:
20615 echo Parsed_color_0 c yellow | tools/zmqsend
20618 To change the right side:
20620 echo Parsed_color_1 c pink | tools/zmqsend
20623 To change the position of the right side:
20625 echo overlay@@my x 150 | tools/zmqsend
20629 @c man end MULTIMEDIA FILTERS
20631 @chapter Multimedia Sources
20632 @c man begin MULTIMEDIA SOURCES
20634 Below is a description of the currently available multimedia sources.
20638 This is the same as @ref{movie} source, except it selects an audio
20644 Read audio and/or video stream(s) from a movie container.
20646 It accepts the following parameters:
20650 The name of the resource to read (not necessarily a file; it can also be a
20651 device or a stream accessed through some protocol).
20653 @item format_name, f
20654 Specifies the format assumed for the movie to read, and can be either
20655 the name of a container or an input device. If not specified, the
20656 format is guessed from @var{movie_name} or by probing.
20658 @item seek_point, sp
20659 Specifies the seek point in seconds. The frames will be output
20660 starting from this seek point. The parameter is evaluated with
20661 @code{av_strtod}, so the numerical value may be suffixed by an IS
20662 postfix. The default value is "0".
20665 Specifies the streams to read. Several streams can be specified,
20666 separated by "+". The source will then have as many outputs, in the
20667 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
20668 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
20669 respectively the default (best suited) video and audio stream. Default
20670 is "dv", or "da" if the filter is called as "amovie".
20672 @item stream_index, si
20673 Specifies the index of the video stream to read. If the value is -1,
20674 the most suitable video stream will be automatically selected. The default
20675 value is "-1". Deprecated. If the filter is called "amovie", it will select
20676 audio instead of video.
20679 Specifies how many times to read the stream in sequence.
20680 If the value is 0, the stream will be looped infinitely.
20681 Default value is "1".
20683 Note that when the movie is looped the source timestamps are not
20684 changed, so it will generate non monotonically increasing timestamps.
20686 @item discontinuity
20687 Specifies the time difference between frames above which the point is
20688 considered a timestamp discontinuity which is removed by adjusting the later
20692 It allows overlaying a second video on top of the main input of
20693 a filtergraph, as shown in this graph:
20695 input -----------> deltapts0 --> overlay --> output
20698 movie --> scale--> deltapts1 -------+
20700 @subsection Examples
20704 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
20705 on top of the input labelled "in":
20707 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
20708 [in] setpts=PTS-STARTPTS [main];
20709 [main][over] overlay=16:16 [out]
20713 Read from a video4linux2 device, and overlay it on top of the input
20716 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
20717 [in] setpts=PTS-STARTPTS [main];
20718 [main][over] overlay=16:16 [out]
20722 Read the first video stream and the audio stream with id 0x81 from
20723 dvd.vob; the video is connected to the pad named "video" and the audio is
20724 connected to the pad named "audio":
20726 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
20730 @subsection Commands
20732 Both movie and amovie support the following commands:
20735 Perform seek using "av_seek_frame".
20736 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
20739 @var{stream_index}: If stream_index is -1, a default
20740 stream is selected, and @var{timestamp} is automatically converted
20741 from AV_TIME_BASE units to the stream specific time_base.
20743 @var{timestamp}: Timestamp in AVStream.time_base units
20744 or, if no stream is specified, in AV_TIME_BASE units.
20746 @var{flags}: Flags which select direction and seeking mode.
20750 Get movie duration in AV_TIME_BASE units.
20754 @c man end MULTIMEDIA SOURCES