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.
555 Remove impulsive noise from input audio.
557 Samples detected as impulsive noise are replaced by interpolated samples using
558 autoregressive modelling.
562 Set window size, in milliseconds. Allowed range is from @code{10} to
563 @code{100}. Default value is @code{55} milliseconds.
564 This sets size of window which will be processed at once.
567 Set window overlap, in percentage of window size. Allowed range is from
568 @code{50} to @code{95}. Default value is @code{75} percent.
569 Setting this to a very high value increases impulsive noise removal but makes
570 whole process much slower.
573 Set autoregression order, in percentage of window size. Allowed range is from
574 @code{0} to @code{25}. Default value is @code{2} percent. This option also
575 controls quality of interpolated samples using neighbour good samples.
578 Set threshold value. Allowed range is from @code{1} to @code{100}.
579 Default value is @code{2}.
580 This controls the strength of impulsive noise which is going to be removed.
581 The lower value, the more samples will be detected as impulsive noise.
584 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
585 @code{10}. Default value is @code{2}.
586 If any two samples deteced as noise are spaced less than this value then any
587 sample inbetween those two samples will be also detected as noise.
592 It accepts the following values:
595 Select overlap-add method. Even not interpolated samples are slightly
596 changed with this method.
599 Select overlap-save method. Not interpolated samples remain unchanged.
602 Default value is @code{a}.
606 Remove clipped samples from input audio.
608 Samples detected as clipped are replaced by interpolated samples using
609 autoregressive modelling.
613 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
614 Default value is @code{55} milliseconds.
615 This sets size of window which will be processed at once.
618 Set window overlap, in percentage of window size. Allowed range is from @code{50}
619 to @code{95}. Default value is @code{75} percent.
622 Set autoregression order, in percentage of window size. Allowed range is from
623 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
624 quality of interpolated samples using neighbour good samples.
627 Set threshold value. Allowed range is from @code{1} to @code{100}.
628 Default value is @code{10}. Higher values make clip detection less aggressive.
631 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
632 Default value is @code{1000}. Higher values make clip detection less aggressive.
637 It accepts the following values:
640 Select overlap-add method. Even not interpolated samples are slightly changed
644 Select overlap-save method. Not interpolated samples remain unchanged.
647 Default value is @code{a}.
652 Delay one or more audio channels.
654 Samples in delayed channel are filled with silence.
656 The filter accepts the following option:
660 Set list of delays in milliseconds for each channel separated by '|'.
661 Unused delays will be silently ignored. If number of given delays is
662 smaller than number of channels all remaining channels will not be delayed.
663 If you want to delay exact number of samples, append 'S' to number.
670 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
671 the second channel (and any other channels that may be present) unchanged.
677 Delay second channel by 500 samples, the third channel by 700 samples and leave
678 the first channel (and any other channels that may be present) unchanged.
684 @section aderivative, aintegral
686 Compute derivative/integral of audio stream.
688 Applying both filters one after another produces original audio.
692 Apply echoing to the input audio.
694 Echoes are reflected sound and can occur naturally amongst mountains
695 (and sometimes large buildings) when talking or shouting; digital echo
696 effects emulate this behaviour and are often used to help fill out the
697 sound of a single instrument or vocal. The time difference between the
698 original signal and the reflection is the @code{delay}, and the
699 loudness of the reflected signal is the @code{decay}.
700 Multiple echoes can have different delays and decays.
702 A description of the accepted parameters follows.
706 Set input gain of reflected signal. Default is @code{0.6}.
709 Set output gain of reflected signal. Default is @code{0.3}.
712 Set list of time intervals in milliseconds between original signal and reflections
713 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
714 Default is @code{1000}.
717 Set list of loudness of reflected signals separated by '|'.
718 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
719 Default is @code{0.5}.
726 Make it sound as if there are twice as many instruments as are actually playing:
728 aecho=0.8:0.88:60:0.4
732 If delay is very short, then it sound like a (metallic) robot playing music:
738 A longer delay will sound like an open air concert in the mountains:
740 aecho=0.8:0.9:1000:0.3
744 Same as above but with one more mountain:
746 aecho=0.8:0.9:1000|1800:0.3|0.25
751 Audio emphasis filter creates or restores material directly taken from LPs or
752 emphased CDs with different filter curves. E.g. to store music on vinyl the
753 signal has to be altered by a filter first to even out the disadvantages of
754 this recording medium.
755 Once the material is played back the inverse filter has to be applied to
756 restore the distortion of the frequency response.
758 The filter accepts the following options:
768 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
769 use @code{production} mode. Default is @code{reproduction} mode.
772 Set filter type. Selects medium. Can be one of the following:
784 select Compact Disc (CD).
790 select 50µs (FM-KF).
792 select 75µs (FM-KF).
798 Modify an audio signal according to the specified expressions.
800 This filter accepts one or more expressions (one for each channel),
801 which are evaluated and used to modify a corresponding audio signal.
803 It accepts the following parameters:
807 Set the '|'-separated expressions list for each separate channel. If
808 the number of input channels is greater than the number of
809 expressions, the last specified expression is used for the remaining
812 @item channel_layout, c
813 Set output channel layout. If not specified, the channel layout is
814 specified by the number of expressions. If set to @samp{same}, it will
815 use by default the same input channel layout.
818 Each expression in @var{exprs} can contain the following constants and functions:
822 channel number of the current expression
825 number of the evaluated sample, starting from 0
831 time of the evaluated sample expressed in seconds
834 @item nb_out_channels
835 input and output number of channels
838 the value of input channel with number @var{CH}
841 Note: this filter is slow. For faster processing you should use a
850 aeval=val(ch)/2:c=same
854 Invert phase of the second channel:
863 Apply fade-in/out effect to input audio.
865 A description of the accepted parameters follows.
869 Specify the effect type, can be either @code{in} for fade-in, or
870 @code{out} for a fade-out effect. Default is @code{in}.
872 @item start_sample, ss
873 Specify the number of the start sample for starting to apply the fade
874 effect. Default is 0.
877 Specify the number of samples for which the fade effect has to last. At
878 the end of the fade-in effect the output audio will have the same
879 volume as the input audio, at the end of the fade-out transition
880 the output audio will be silence. Default is 44100.
883 Specify the start time of the fade effect. Default is 0.
884 The value must be specified as a time duration; see
885 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
886 for the accepted syntax.
887 If set this option is used instead of @var{start_sample}.
890 Specify the duration of the fade effect. See
891 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
892 for the accepted syntax.
893 At the end of the fade-in effect the output audio will have the same
894 volume as the input audio, at the end of the fade-out transition
895 the output audio will be silence.
896 By default the duration is determined by @var{nb_samples}.
897 If set this option is used instead of @var{nb_samples}.
900 Set curve for fade transition.
902 It accepts the following values:
905 select triangular, linear slope (default)
907 select quarter of sine wave
909 select half of sine wave
911 select exponential sine wave
915 select inverted parabola
929 select inverted quarter of sine wave
931 select inverted half of sine wave
933 select double-exponential seat
935 select double-exponential sigmoid
943 Fade in first 15 seconds of audio:
949 Fade out last 25 seconds of a 900 seconds audio:
951 afade=t=out:st=875:d=25
956 Apply arbitrary expressions to samples in frequency domain.
960 Set frequency domain real expression for each separate channel separated
961 by '|'. Default is "1".
962 If the number of input channels is greater than the number of
963 expressions, the last specified expression is used for the remaining
967 Set frequency domain imaginary expression for each separate channel
968 separated by '|'. If not set, @var{real} option is used.
970 Each expression in @var{real} and @var{imag} can contain the following
978 current frequency bin number
981 number of available bins
984 channel number of the current expression
996 It accepts the following values:
1012 Default is @code{w4096}
1015 Set window function. Default is @code{hann}.
1018 Set window overlap. If set to 1, the recommended overlap for selected
1019 window function will be picked. Default is @code{0.75}.
1022 @subsection Examples
1026 Leave almost only low frequencies in audio:
1028 afftfilt="1-clip((b/nb)*b,0,1)"
1035 Apply an arbitrary Frequency Impulse Response filter.
1037 This filter is designed for applying long FIR filters,
1038 up to 30 seconds long.
1040 It can be used as component for digital crossover filters,
1041 room equalization, cross talk cancellation, wavefield synthesis,
1042 auralization, ambiophonics and ambisonics.
1044 This filter uses second stream as FIR coefficients.
1045 If second stream holds single channel, it will be used
1046 for all input channels in first stream, otherwise
1047 number of channels in second stream must be same as
1048 number of channels in first stream.
1050 It accepts the following parameters:
1054 Set dry gain. This sets input gain.
1057 Set wet gain. This sets final output gain.
1060 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1063 Enable applying gain measured from power of IR.
1066 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1067 Allowed range is 0.1 to 60 seconds.
1070 Show IR frequency reponse, magnitude and phase in additional video stream.
1071 By default it is disabled.
1074 Set for which IR channel to display frequency response. By default is first channel
1075 displayed. This option is used only when @var{response} is enabled.
1078 Set video stream size. This option is used only when @var{response} is enabled.
1081 @subsection Examples
1085 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1087 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1094 Set output format constraints for the input audio. The framework will
1095 negotiate the most appropriate format to minimize conversions.
1097 It accepts the following parameters:
1101 A '|'-separated list of requested sample formats.
1104 A '|'-separated list of requested sample rates.
1106 @item channel_layouts
1107 A '|'-separated list of requested channel layouts.
1109 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1110 for the required syntax.
1113 If a parameter is omitted, all values are allowed.
1115 Force the output to either unsigned 8-bit or signed 16-bit stereo
1117 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1122 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1123 processing reduces disturbing noise between useful signals.
1125 Gating is done by detecting the volume below a chosen level @var{threshold}
1126 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1127 floor is set via @var{range}. Because an exact manipulation of the signal
1128 would cause distortion of the waveform the reduction can be levelled over
1129 time. This is done by setting @var{attack} and @var{release}.
1131 @var{attack} determines how long the signal has to fall below the threshold
1132 before any reduction will occur and @var{release} sets the time the signal
1133 has to rise above the threshold to reduce the reduction again.
1134 Shorter signals than the chosen attack time will be left untouched.
1138 Set input level before filtering.
1139 Default is 1. Allowed range is from 0.015625 to 64.
1142 Set the level of gain reduction when the signal is below the threshold.
1143 Default is 0.06125. Allowed range is from 0 to 1.
1146 If a signal rises above this level the gain reduction is released.
1147 Default is 0.125. Allowed range is from 0 to 1.
1150 Set a ratio by which the signal is reduced.
1151 Default is 2. Allowed range is from 1 to 9000.
1154 Amount of milliseconds the signal has to rise above the threshold before gain
1156 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1159 Amount of milliseconds the signal has to fall below the threshold before the
1160 reduction is increased again. Default is 250 milliseconds.
1161 Allowed range is from 0.01 to 9000.
1164 Set amount of amplification of signal after processing.
1165 Default is 1. Allowed range is from 1 to 64.
1168 Curve the sharp knee around the threshold to enter gain reduction more softly.
1169 Default is 2.828427125. Allowed range is from 1 to 8.
1172 Choose if exact signal should be taken for detection or an RMS like one.
1173 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1176 Choose if the average level between all channels or the louder channel affects
1178 Default is @code{average}. Can be @code{average} or @code{maximum}.
1183 Apply an arbitrary Infinite Impulse Response filter.
1185 It accepts the following parameters:
1189 Set numerator/zeros coefficients.
1192 Set denominator/poles coefficients.
1204 Set coefficients format.
1210 Z-plane zeros/poles, cartesian (default)
1212 Z-plane zeros/poles, polar radians
1214 Z-plane zeros/poles, polar degrees
1218 Set kind of processing.
1219 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1222 Set filtering precision.
1226 double-precision floating-point (default)
1228 single-precision floating-point
1236 Show IR frequency reponse, magnitude and phase in additional video stream.
1237 By default it is disabled.
1240 Set for which IR channel to display frequency response. By default is first channel
1241 displayed. This option is used only when @var{response} is enabled.
1244 Set video stream size. This option is used only when @var{response} is enabled.
1247 Coefficients in @code{tf} format are separated by spaces and are in ascending
1250 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1251 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1254 Different coefficients and gains can be provided for every channel, in such case
1255 use '|' to separate coefficients or gains. Last provided coefficients will be
1256 used for all remaining channels.
1258 @subsection Examples
1262 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1264 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
1268 Same as above but in @code{zp} format:
1270 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
1276 The limiter prevents an input signal from rising over a desired threshold.
1277 This limiter uses lookahead technology to prevent your signal from distorting.
1278 It means that there is a small delay after the signal is processed. Keep in mind
1279 that the delay it produces is the attack time you set.
1281 The filter accepts the following options:
1285 Set input gain. Default is 1.
1288 Set output gain. Default is 1.
1291 Don't let signals above this level pass the limiter. Default is 1.
1294 The limiter will reach its attenuation level in this amount of time in
1295 milliseconds. Default is 5 milliseconds.
1298 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1299 Default is 50 milliseconds.
1302 When gain reduction is always needed ASC takes care of releasing to an
1303 average reduction level rather than reaching a reduction of 0 in the release
1307 Select how much the release time is affected by ASC, 0 means nearly no changes
1308 in release time while 1 produces higher release times.
1311 Auto level output signal. Default is enabled.
1312 This normalizes audio back to 0dB if enabled.
1315 Depending on picked setting it is recommended to upsample input 2x or 4x times
1316 with @ref{aresample} before applying this filter.
1320 Apply a two-pole all-pass filter with central frequency (in Hz)
1321 @var{frequency}, and filter-width @var{width}.
1322 An all-pass filter changes the audio's frequency to phase relationship
1323 without changing its frequency to amplitude relationship.
1325 The filter accepts the following options:
1329 Set frequency in Hz.
1332 Set method to specify band-width of filter.
1347 Specify the band-width of a filter in width_type units.
1350 Specify which channels to filter, by default all available are filtered.
1353 @subsection Commands
1355 This filter supports the following commands:
1358 Change allpass frequency.
1359 Syntax for the command is : "@var{frequency}"
1362 Change allpass width_type.
1363 Syntax for the command is : "@var{width_type}"
1366 Change allpass width.
1367 Syntax for the command is : "@var{width}"
1374 The filter accepts the following options:
1378 Set the number of loops. Setting this value to -1 will result in infinite loops.
1382 Set maximal number of samples. Default is 0.
1385 Set first sample of loop. Default is 0.
1391 Merge two or more audio streams into a single multi-channel stream.
1393 The filter accepts the following options:
1398 Set the number of inputs. Default is 2.
1402 If the channel layouts of the inputs are disjoint, and therefore compatible,
1403 the channel layout of the output will be set accordingly and the channels
1404 will be reordered as necessary. If the channel layouts of the inputs are not
1405 disjoint, the output will have all the channels of the first input then all
1406 the channels of the second input, in that order, and the channel layout of
1407 the output will be the default value corresponding to the total number of
1410 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1411 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1412 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1413 first input, b1 is the first channel of the second input).
1415 On the other hand, if both input are in stereo, the output channels will be
1416 in the default order: a1, a2, b1, b2, and the channel layout will be
1417 arbitrarily set to 4.0, which may or may not be the expected value.
1419 All inputs must have the same sample rate, and format.
1421 If inputs do not have the same duration, the output will stop with the
1424 @subsection Examples
1428 Merge two mono files into a stereo stream:
1430 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1434 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1436 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
1442 Mixes multiple audio inputs into a single output.
1444 Note that this filter only supports float samples (the @var{amerge}
1445 and @var{pan} audio filters support many formats). If the @var{amix}
1446 input has integer samples then @ref{aresample} will be automatically
1447 inserted to perform the conversion to float samples.
1451 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1453 will mix 3 input audio streams to a single output with the same duration as the
1454 first input and a dropout transition time of 3 seconds.
1456 It accepts the following parameters:
1460 The number of inputs. If unspecified, it defaults to 2.
1463 How to determine the end-of-stream.
1467 The duration of the longest input. (default)
1470 The duration of the shortest input.
1473 The duration of the first input.
1477 @item dropout_transition
1478 The transition time, in seconds, for volume renormalization when an input
1479 stream ends. The default value is 2 seconds.
1482 Specify weight of each input audio stream as sequence.
1483 Each weight is separated by space. By default all inputs have same weight.
1486 @section anequalizer
1488 High-order parametric multiband equalizer for each channel.
1490 It accepts the following parameters:
1494 This option string is in format:
1495 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1496 Each equalizer band is separated by '|'.
1500 Set channel number to which equalization will be applied.
1501 If input doesn't have that channel the entry is ignored.
1504 Set central frequency for band.
1505 If input doesn't have that frequency the entry is ignored.
1508 Set band width in hertz.
1511 Set band gain in dB.
1514 Set filter type for band, optional, can be:
1518 Butterworth, this is default.
1529 With this option activated frequency response of anequalizer is displayed
1533 Set video stream size. Only useful if curves option is activated.
1536 Set max gain that will be displayed. Only useful if curves option is activated.
1537 Setting this to a reasonable value makes it possible to display gain which is derived from
1538 neighbour bands which are too close to each other and thus produce higher gain
1539 when both are activated.
1542 Set frequency scale used to draw frequency response in video output.
1543 Can be linear or logarithmic. Default is logarithmic.
1546 Set color for each channel curve which is going to be displayed in video stream.
1547 This is list of color names separated by space or by '|'.
1548 Unrecognised or missing colors will be replaced by white color.
1551 @subsection Examples
1555 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1556 for first 2 channels using Chebyshev type 1 filter:
1558 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1562 @subsection Commands
1564 This filter supports the following commands:
1567 Alter existing filter parameters.
1568 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1570 @var{fN} is existing filter number, starting from 0, if no such filter is available
1572 @var{freq} set new frequency parameter.
1573 @var{width} set new width parameter in herz.
1574 @var{gain} set new gain parameter in dB.
1576 Full filter invocation with asendcmd may look like this:
1577 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1582 Pass the audio source unchanged to the output.
1586 Pad the end of an audio stream with silence.
1588 This can be used together with @command{ffmpeg} @option{-shortest} to
1589 extend audio streams to the same length as the video stream.
1591 A description of the accepted options follows.
1595 Set silence packet size. Default value is 4096.
1598 Set the number of samples of silence to add to the end. After the
1599 value is reached, the stream is terminated. This option is mutually
1600 exclusive with @option{whole_len}.
1603 Set the minimum total number of samples in the output audio stream. If
1604 the value is longer than the input audio length, silence is added to
1605 the end, until the value is reached. This option is mutually exclusive
1606 with @option{pad_len}.
1609 If neither the @option{pad_len} nor the @option{whole_len} option is
1610 set, the filter will add silence to the end of the input stream
1613 @subsection Examples
1617 Add 1024 samples of silence to the end of the input:
1623 Make sure the audio output will contain at least 10000 samples, pad
1624 the input with silence if required:
1626 apad=whole_len=10000
1630 Use @command{ffmpeg} to pad the audio input with silence, so that the
1631 video stream will always result the shortest and will be converted
1632 until the end in the output file when using the @option{shortest}
1635 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1640 Add a phasing effect to the input audio.
1642 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1643 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1645 A description of the accepted parameters follows.
1649 Set input gain. Default is 0.4.
1652 Set output gain. Default is 0.74
1655 Set delay in milliseconds. Default is 3.0.
1658 Set decay. Default is 0.4.
1661 Set modulation speed in Hz. Default is 0.5.
1664 Set modulation type. Default is triangular.
1666 It accepts the following values:
1675 Audio pulsator is something between an autopanner and a tremolo.
1676 But it can produce funny stereo effects as well. Pulsator changes the volume
1677 of the left and right channel based on a LFO (low frequency oscillator) with
1678 different waveforms and shifted phases.
1679 This filter have the ability to define an offset between left and right
1680 channel. An offset of 0 means that both LFO shapes match each other.
1681 The left and right channel are altered equally - a conventional tremolo.
1682 An offset of 50% means that the shape of the right channel is exactly shifted
1683 in phase (or moved backwards about half of the frequency) - pulsator acts as
1684 an autopanner. At 1 both curves match again. Every setting in between moves the
1685 phase shift gapless between all stages and produces some "bypassing" sounds with
1686 sine and triangle waveforms. The more you set the offset near 1 (starting from
1687 the 0.5) the faster the signal passes from the left to the right speaker.
1689 The filter accepts the following options:
1693 Set input gain. By default it is 1. Range is [0.015625 - 64].
1696 Set output gain. By default it is 1. Range is [0.015625 - 64].
1699 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1700 sawup or sawdown. Default is sine.
1703 Set modulation. Define how much of original signal is affected by the LFO.
1706 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1709 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1712 Set pulse width. Default is 1. Allowed range is [0 - 2].
1715 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1718 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1722 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1726 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1727 if timing is set to hz.
1733 Resample the input audio to the specified parameters, using the
1734 libswresample library. If none are specified then the filter will
1735 automatically convert between its input and output.
1737 This filter is also able to stretch/squeeze the audio data to make it match
1738 the timestamps or to inject silence / cut out audio to make it match the
1739 timestamps, do a combination of both or do neither.
1741 The filter accepts the syntax
1742 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1743 expresses a sample rate and @var{resampler_options} is a list of
1744 @var{key}=@var{value} pairs, separated by ":". See the
1745 @ref{Resampler Options,,"Resampler Options" section in the
1746 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1747 for the complete list of supported options.
1749 @subsection Examples
1753 Resample the input audio to 44100Hz:
1759 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1760 samples per second compensation:
1762 aresample=async=1000
1768 Reverse an audio clip.
1770 Warning: This filter requires memory to buffer the entire clip, so trimming
1773 @subsection Examples
1777 Take the first 5 seconds of a clip, and reverse it.
1779 atrim=end=5,areverse
1783 @section asetnsamples
1785 Set the number of samples per each output audio frame.
1787 The last output packet may contain a different number of samples, as
1788 the filter will flush all the remaining samples when the input audio
1791 The filter accepts the following options:
1795 @item nb_out_samples, n
1796 Set the number of frames per each output audio frame. The number is
1797 intended as the number of samples @emph{per each channel}.
1798 Default value is 1024.
1801 If set to 1, the filter will pad the last audio frame with zeroes, so
1802 that the last frame will contain the same number of samples as the
1803 previous ones. Default value is 1.
1806 For example, to set the number of per-frame samples to 1234 and
1807 disable padding for the last frame, use:
1809 asetnsamples=n=1234:p=0
1814 Set the sample rate without altering the PCM data.
1815 This will result in a change of speed and pitch.
1817 The filter accepts the following options:
1820 @item sample_rate, r
1821 Set the output sample rate. Default is 44100 Hz.
1826 Show a line containing various information for each input audio frame.
1827 The input audio is not modified.
1829 The shown line contains a sequence of key/value pairs of the form
1830 @var{key}:@var{value}.
1832 The following values are shown in the output:
1836 The (sequential) number of the input frame, starting from 0.
1839 The presentation timestamp of the input frame, in time base units; the time base
1840 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1843 The presentation timestamp of the input frame in seconds.
1846 position of the frame in the input stream, -1 if this information in
1847 unavailable and/or meaningless (for example in case of synthetic audio)
1856 The sample rate for the audio frame.
1859 The number of samples (per channel) in the frame.
1862 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1863 audio, the data is treated as if all the planes were concatenated.
1865 @item plane_checksums
1866 A list of Adler-32 checksums for each data plane.
1872 Display time domain statistical information about the audio channels.
1873 Statistics are calculated and displayed for each audio channel and,
1874 where applicable, an overall figure is also given.
1876 It accepts the following option:
1879 Short window length in seconds, used for peak and trough RMS measurement.
1880 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
1884 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1885 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1888 Available keys for each channel are:
1922 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1923 this @code{lavfi.astats.Overall.Peak_count}.
1925 For description what each key means read below.
1928 Set number of frame after which stats are going to be recalculated.
1929 Default is disabled.
1932 A description of each shown parameter follows:
1936 Mean amplitude displacement from zero.
1939 Minimal sample level.
1942 Maximal sample level.
1944 @item Min difference
1945 Minimal difference between two consecutive samples.
1947 @item Max difference
1948 Maximal difference between two consecutive samples.
1950 @item Mean difference
1951 Mean difference between two consecutive samples.
1952 The average of each difference between two consecutive samples.
1954 @item RMS difference
1955 Root Mean Square difference between two consecutive samples.
1959 Standard peak and RMS level measured in dBFS.
1963 Peak and trough values for RMS level measured over a short window.
1966 Standard ratio of peak to RMS level (note: not in dB).
1969 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1970 (i.e. either @var{Min level} or @var{Max level}).
1973 Number of occasions (not the number of samples) that the signal attained either
1974 @var{Min level} or @var{Max level}.
1977 Overall bit depth of audio. Number of bits used for each sample.
1980 Measured dynamic range of audio in dB.
1987 The filter accepts exactly one parameter, the audio tempo. If not
1988 specified then the filter will assume nominal 1.0 tempo. Tempo must
1989 be in the [0.5, 100.0] range.
1991 Note that tempo greater than 2 will skip some samples rather than
1992 blend them in. If for any reason this is a concern it is always
1993 possible to daisy-chain several instances of atempo to achieve the
1994 desired product tempo.
1996 @subsection Examples
2000 Slow down audio to 80% tempo:
2006 To speed up audio to 300% tempo:
2012 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2014 atempo=sqrt(3),atempo=sqrt(3)
2020 Trim the input so that the output contains one continuous subpart of the input.
2022 It accepts the following parameters:
2025 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2026 sample with the timestamp @var{start} will be the first sample in the output.
2029 Specify time of the first audio sample that will be dropped, i.e. the
2030 audio sample immediately preceding the one with the timestamp @var{end} will be
2031 the last sample in the output.
2034 Same as @var{start}, except this option sets the start timestamp in samples
2038 Same as @var{end}, except this option sets the end timestamp in samples instead
2042 The maximum duration of the output in seconds.
2045 The number of the first sample that should be output.
2048 The number of the first sample that should be dropped.
2051 @option{start}, @option{end}, and @option{duration} are expressed as time
2052 duration specifications; see
2053 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2055 Note that the first two sets of the start/end options and the @option{duration}
2056 option look at the frame timestamp, while the _sample options simply count the
2057 samples that pass through the filter. So start/end_pts and start/end_sample will
2058 give different results when the timestamps are wrong, inexact or do not start at
2059 zero. Also note that this filter does not modify the timestamps. If you wish
2060 to have the output timestamps start at zero, insert the asetpts filter after the
2063 If multiple start or end options are set, this filter tries to be greedy and
2064 keep all samples that match at least one of the specified constraints. To keep
2065 only the part that matches all the constraints at once, chain multiple atrim
2068 The defaults are such that all the input is kept. So it is possible to set e.g.
2069 just the end values to keep everything before the specified time.
2074 Drop everything except the second minute of input:
2076 ffmpeg -i INPUT -af atrim=60:120
2080 Keep only the first 1000 samples:
2082 ffmpeg -i INPUT -af atrim=end_sample=1000
2089 Apply a two-pole Butterworth band-pass filter with central
2090 frequency @var{frequency}, and (3dB-point) band-width width.
2091 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2092 instead of the default: constant 0dB peak gain.
2093 The filter roll off at 6dB per octave (20dB per decade).
2095 The filter accepts the following options:
2099 Set the filter's central frequency. Default is @code{3000}.
2102 Constant skirt gain if set to 1. Defaults to 0.
2105 Set method to specify band-width of filter.
2120 Specify the band-width of a filter in width_type units.
2123 Specify which channels to filter, by default all available are filtered.
2126 @subsection Commands
2128 This filter supports the following commands:
2131 Change bandpass frequency.
2132 Syntax for the command is : "@var{frequency}"
2135 Change bandpass width_type.
2136 Syntax for the command is : "@var{width_type}"
2139 Change bandpass width.
2140 Syntax for the command is : "@var{width}"
2145 Apply a two-pole Butterworth band-reject filter with central
2146 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2147 The filter roll off at 6dB per octave (20dB per decade).
2149 The filter accepts the following options:
2153 Set the filter's central frequency. Default is @code{3000}.
2156 Set method to specify band-width of filter.
2171 Specify the band-width of a filter in width_type units.
2174 Specify which channels to filter, by default all available are filtered.
2177 @subsection Commands
2179 This filter supports the following commands:
2182 Change bandreject frequency.
2183 Syntax for the command is : "@var{frequency}"
2186 Change bandreject width_type.
2187 Syntax for the command is : "@var{width_type}"
2190 Change bandreject width.
2191 Syntax for the command is : "@var{width}"
2194 @section bass, lowshelf
2196 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2197 shelving filter with a response similar to that of a standard
2198 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2200 The filter accepts the following options:
2204 Give the gain at 0 Hz. Its useful range is about -20
2205 (for a large cut) to +20 (for a large boost).
2206 Beware of clipping when using a positive gain.
2209 Set the filter's central frequency and so can be used
2210 to extend or reduce the frequency range to be boosted or cut.
2211 The default value is @code{100} Hz.
2214 Set method to specify band-width of filter.
2229 Determine how steep is the filter's shelf transition.
2232 Specify which channels to filter, by default all available are filtered.
2235 @subsection Commands
2237 This filter supports the following commands:
2240 Change bass frequency.
2241 Syntax for the command is : "@var{frequency}"
2244 Change bass width_type.
2245 Syntax for the command is : "@var{width_type}"
2249 Syntax for the command is : "@var{width}"
2253 Syntax for the command is : "@var{gain}"
2258 Apply a biquad IIR filter with the given coefficients.
2259 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2260 are the numerator and denominator coefficients respectively.
2261 and @var{channels}, @var{c} specify which channels to filter, by default all
2262 available are filtered.
2264 @subsection Commands
2266 This filter supports the following commands:
2274 Change biquad parameter.
2275 Syntax for the command is : "@var{value}"
2279 Bauer stereo to binaural transformation, which improves headphone listening of
2280 stereo audio records.
2282 To enable compilation of this filter you need to configure FFmpeg with
2283 @code{--enable-libbs2b}.
2285 It accepts the following parameters:
2289 Pre-defined crossfeed level.
2293 Default level (fcut=700, feed=50).
2296 Chu Moy circuit (fcut=700, feed=60).
2299 Jan Meier circuit (fcut=650, feed=95).
2304 Cut frequency (in Hz).
2313 Remap input channels to new locations.
2315 It accepts the following parameters:
2318 Map channels from input to output. The argument is a '|'-separated list of
2319 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2320 @var{in_channel} form. @var{in_channel} can be either the name of the input
2321 channel (e.g. FL for front left) or its index in the input channel layout.
2322 @var{out_channel} is the name of the output channel or its index in the output
2323 channel layout. If @var{out_channel} is not given then it is implicitly an
2324 index, starting with zero and increasing by one for each mapping.
2326 @item channel_layout
2327 The channel layout of the output stream.
2330 If no mapping is present, the filter will implicitly map input channels to
2331 output channels, preserving indices.
2333 @subsection Examples
2337 For example, assuming a 5.1+downmix input MOV file,
2339 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2341 will create an output WAV file tagged as stereo from the downmix channels of
2345 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2347 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2351 @section channelsplit
2353 Split each channel from an input audio stream into a separate output stream.
2355 It accepts the following parameters:
2357 @item channel_layout
2358 The channel layout of the input stream. The default is "stereo".
2360 A channel layout describing the channels to be extracted as separate output streams
2361 or "all" to extract each input channel as a separate stream. The default is "all".
2363 Choosing channels not present in channel layout in the input will result in an error.
2366 @subsection Examples
2370 For example, assuming a stereo input MP3 file,
2372 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2374 will create an output Matroska file with two audio streams, one containing only
2375 the left channel and the other the right channel.
2378 Split a 5.1 WAV file into per-channel files:
2380 ffmpeg -i in.wav -filter_complex
2381 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2382 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2383 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2388 Extract only LFE from a 5.1 WAV file:
2390 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2391 -map '[LFE]' lfe.wav
2396 Add a chorus effect to the audio.
2398 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2400 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2401 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2402 The modulation depth defines the range the modulated delay is played before or after
2403 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2404 sound tuned around the original one, like in a chorus where some vocals are slightly
2407 It accepts the following parameters:
2410 Set input gain. Default is 0.4.
2413 Set output gain. Default is 0.4.
2416 Set delays. A typical delay is around 40ms to 60ms.
2428 @subsection Examples
2434 chorus=0.7:0.9:55:0.4:0.25:2
2440 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2444 Fuller sounding chorus with three delays:
2446 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
2451 Compress or expand the audio's dynamic range.
2453 It accepts the following parameters:
2459 A list of times in seconds for each channel over which the instantaneous level
2460 of the input signal is averaged to determine its volume. @var{attacks} refers to
2461 increase of volume and @var{decays} refers to decrease of volume. For most
2462 situations, the attack time (response to the audio getting louder) should be
2463 shorter than the decay time, because the human ear is more sensitive to sudden
2464 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2465 a typical value for decay is 0.8 seconds.
2466 If specified number of attacks & decays is lower than number of channels, the last
2467 set attack/decay will be used for all remaining channels.
2470 A list of points for the transfer function, specified in dB relative to the
2471 maximum possible signal amplitude. Each key points list must be defined using
2472 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2473 @code{x0/y0 x1/y1 x2/y2 ....}
2475 The input values must be in strictly increasing order but the transfer function
2476 does not have to be monotonically rising. The point @code{0/0} is assumed but
2477 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2478 function are @code{-70/-70|-60/-20|1/0}.
2481 Set the curve radius in dB for all joints. It defaults to 0.01.
2484 Set the additional gain in dB to be applied at all points on the transfer
2485 function. This allows for easy adjustment of the overall gain.
2489 Set an initial volume, in dB, to be assumed for each channel when filtering
2490 starts. This permits the user to supply a nominal level initially, so that, for
2491 example, a very large gain is not applied to initial signal levels before the
2492 companding has begun to operate. A typical value for audio which is initially
2493 quiet is -90 dB. It defaults to 0.
2496 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2497 delayed before being fed to the volume adjuster. Specifying a delay
2498 approximately equal to the attack/decay times allows the filter to effectively
2499 operate in predictive rather than reactive mode. It defaults to 0.
2503 @subsection Examples
2507 Make music with both quiet and loud passages suitable for listening to in a
2510 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2513 Another example for audio with whisper and explosion parts:
2515 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2519 A noise gate for when the noise is at a lower level than the signal:
2521 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2525 Here is another noise gate, this time for when the noise is at a higher level
2526 than the signal (making it, in some ways, similar to squelch):
2528 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2532 2:1 compression starting at -6dB:
2534 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2538 2:1 compression starting at -9dB:
2540 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2544 2:1 compression starting at -12dB:
2546 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2550 2:1 compression starting at -18dB:
2552 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2556 3:1 compression starting at -15dB:
2558 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2564 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2570 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
2574 Hard limiter at -6dB:
2576 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2580 Hard limiter at -12dB:
2582 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2586 Hard noise gate at -35 dB:
2588 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2594 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2598 @section compensationdelay
2600 Compensation Delay Line is a metric based delay to compensate differing
2601 positions of microphones or speakers.
2603 For example, you have recorded guitar with two microphones placed in
2604 different location. Because the front of sound wave has fixed speed in
2605 normal conditions, the phasing of microphones can vary and depends on
2606 their location and interposition. The best sound mix can be achieved when
2607 these microphones are in phase (synchronized). Note that distance of
2608 ~30 cm between microphones makes one microphone to capture signal in
2609 antiphase to another microphone. That makes the final mix sounding moody.
2610 This filter helps to solve phasing problems by adding different delays
2611 to each microphone track and make them synchronized.
2613 The best result can be reached when you take one track as base and
2614 synchronize other tracks one by one with it.
2615 Remember that synchronization/delay tolerance depends on sample rate, too.
2616 Higher sample rates will give more tolerance.
2618 It accepts the following parameters:
2622 Set millimeters distance. This is compensation distance for fine tuning.
2626 Set cm distance. This is compensation distance for tightening distance setup.
2630 Set meters distance. This is compensation distance for hard distance setup.
2634 Set dry amount. Amount of unprocessed (dry) signal.
2638 Set wet amount. Amount of processed (wet) signal.
2642 Set temperature degree in Celsius. This is the temperature of the environment.
2647 Apply headphone crossfeed filter.
2649 Crossfeed is the process of blending the left and right channels of stereo
2651 It is mainly used to reduce extreme stereo separation of low frequencies.
2653 The intent is to produce more speaker like sound to the listener.
2655 The filter accepts the following options:
2659 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2660 This sets gain of low shelf filter for side part of stereo image.
2661 Default is -6dB. Max allowed is -30db when strength is set to 1.
2664 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2665 This sets cut off frequency of low shelf filter. Default is cut off near
2666 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2669 Set input gain. Default is 0.9.
2672 Set output gain. Default is 1.
2675 @section crystalizer
2676 Simple algorithm to expand audio dynamic range.
2678 The filter accepts the following options:
2682 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2683 (unchanged sound) to 10.0 (maximum effect).
2686 Enable clipping. By default is enabled.
2690 Apply a DC shift to the audio.
2692 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2693 in the recording chain) from the audio. The effect of a DC offset is reduced
2694 headroom and hence volume. The @ref{astats} filter can be used to determine if
2695 a signal has a DC offset.
2699 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2703 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2704 used to prevent clipping.
2708 Measure audio dynamic range.
2710 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2711 is found in transition material. And anything less that 8 have very poor dynamics
2712 and is very compressed.
2714 The filter accepts the following options:
2718 Set window length in seconds used to split audio into segments of equal length.
2719 Default is 3 seconds.
2723 Dynamic Audio Normalizer.
2725 This filter applies a certain amount of gain to the input audio in order
2726 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2727 contrast to more "simple" normalization algorithms, the Dynamic Audio
2728 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2729 This allows for applying extra gain to the "quiet" sections of the audio
2730 while avoiding distortions or clipping the "loud" sections. In other words:
2731 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2732 sections, in the sense that the volume of each section is brought to the
2733 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2734 this goal *without* applying "dynamic range compressing". It will retain 100%
2735 of the dynamic range *within* each section of the audio file.
2739 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2740 Default is 500 milliseconds.
2741 The Dynamic Audio Normalizer processes the input audio in small chunks,
2742 referred to as frames. This is required, because a peak magnitude has no
2743 meaning for just a single sample value. Instead, we need to determine the
2744 peak magnitude for a contiguous sequence of sample values. While a "standard"
2745 normalizer would simply use the peak magnitude of the complete file, the
2746 Dynamic Audio Normalizer determines the peak magnitude individually for each
2747 frame. The length of a frame is specified in milliseconds. By default, the
2748 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2749 been found to give good results with most files.
2750 Note that the exact frame length, in number of samples, will be determined
2751 automatically, based on the sampling rate of the individual input audio file.
2754 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2755 number. Default is 31.
2756 Probably the most important parameter of the Dynamic Audio Normalizer is the
2757 @code{window size} of the Gaussian smoothing filter. The filter's window size
2758 is specified in frames, centered around the current frame. For the sake of
2759 simplicity, this must be an odd number. Consequently, the default value of 31
2760 takes into account the current frame, as well as the 15 preceding frames and
2761 the 15 subsequent frames. Using a larger window results in a stronger
2762 smoothing effect and thus in less gain variation, i.e. slower gain
2763 adaptation. Conversely, using a smaller window results in a weaker smoothing
2764 effect and thus in more gain variation, i.e. faster gain adaptation.
2765 In other words, the more you increase this value, the more the Dynamic Audio
2766 Normalizer will behave like a "traditional" normalization filter. On the
2767 contrary, the more you decrease this value, the more the Dynamic Audio
2768 Normalizer will behave like a dynamic range compressor.
2771 Set the target peak value. This specifies the highest permissible magnitude
2772 level for the normalized audio input. This filter will try to approach the
2773 target peak magnitude as closely as possible, but at the same time it also
2774 makes sure that the normalized signal will never exceed the peak magnitude.
2775 A frame's maximum local gain factor is imposed directly by the target peak
2776 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2777 It is not recommended to go above this value.
2780 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2781 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2782 factor for each input frame, i.e. the maximum gain factor that does not
2783 result in clipping or distortion. The maximum gain factor is determined by
2784 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2785 additionally bounds the frame's maximum gain factor by a predetermined
2786 (global) maximum gain factor. This is done in order to avoid excessive gain
2787 factors in "silent" or almost silent frames. By default, the maximum gain
2788 factor is 10.0, For most inputs the default value should be sufficient and
2789 it usually is not recommended to increase this value. Though, for input
2790 with an extremely low overall volume level, it may be necessary to allow even
2791 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2792 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2793 Instead, a "sigmoid" threshold function will be applied. This way, the
2794 gain factors will smoothly approach the threshold value, but never exceed that
2798 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2799 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2800 This means that the maximum local gain factor for each frame is defined
2801 (only) by the frame's highest magnitude sample. This way, the samples can
2802 be amplified as much as possible without exceeding the maximum signal
2803 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2804 Normalizer can also take into account the frame's root mean square,
2805 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2806 determine the power of a time-varying signal. It is therefore considered
2807 that the RMS is a better approximation of the "perceived loudness" than
2808 just looking at the signal's peak magnitude. Consequently, by adjusting all
2809 frames to a constant RMS value, a uniform "perceived loudness" can be
2810 established. If a target RMS value has been specified, a frame's local gain
2811 factor is defined as the factor that would result in exactly that RMS value.
2812 Note, however, that the maximum local gain factor is still restricted by the
2813 frame's highest magnitude sample, in order to prevent clipping.
2816 Enable channels coupling. By default is enabled.
2817 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2818 amount. This means the same gain factor will be applied to all channels, i.e.
2819 the maximum possible gain factor is determined by the "loudest" channel.
2820 However, in some recordings, it may happen that the volume of the different
2821 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2822 In this case, this option can be used to disable the channel coupling. This way,
2823 the gain factor will be determined independently for each channel, depending
2824 only on the individual channel's highest magnitude sample. This allows for
2825 harmonizing the volume of the different channels.
2828 Enable DC bias correction. By default is disabled.
2829 An audio signal (in the time domain) is a sequence of sample values.
2830 In the Dynamic Audio Normalizer these sample values are represented in the
2831 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2832 audio signal, or "waveform", should be centered around the zero point.
2833 That means if we calculate the mean value of all samples in a file, or in a
2834 single frame, then the result should be 0.0 or at least very close to that
2835 value. If, however, there is a significant deviation of the mean value from
2836 0.0, in either positive or negative direction, this is referred to as a
2837 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2838 Audio Normalizer provides optional DC bias correction.
2839 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2840 the mean value, or "DC correction" offset, of each input frame and subtract
2841 that value from all of the frame's sample values which ensures those samples
2842 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2843 boundaries, the DC correction offset values will be interpolated smoothly
2844 between neighbouring frames.
2847 Enable alternative boundary mode. By default is disabled.
2848 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2849 around each frame. This includes the preceding frames as well as the
2850 subsequent frames. However, for the "boundary" frames, located at the very
2851 beginning and at the very end of the audio file, not all neighbouring
2852 frames are available. In particular, for the first few frames in the audio
2853 file, the preceding frames are not known. And, similarly, for the last few
2854 frames in the audio file, the subsequent frames are not known. Thus, the
2855 question arises which gain factors should be assumed for the missing frames
2856 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2857 to deal with this situation. The default boundary mode assumes a gain factor
2858 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2859 "fade out" at the beginning and at the end of the input, respectively.
2862 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2863 By default, the Dynamic Audio Normalizer does not apply "traditional"
2864 compression. This means that signal peaks will not be pruned and thus the
2865 full dynamic range will be retained within each local neighbourhood. However,
2866 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2867 normalization algorithm with a more "traditional" compression.
2868 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2869 (thresholding) function. If (and only if) the compression feature is enabled,
2870 all input frames will be processed by a soft knee thresholding function prior
2871 to the actual normalization process. Put simply, the thresholding function is
2872 going to prune all samples whose magnitude exceeds a certain threshold value.
2873 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2874 value. Instead, the threshold value will be adjusted for each individual
2876 In general, smaller parameters result in stronger compression, and vice versa.
2877 Values below 3.0 are not recommended, because audible distortion may appear.
2882 Make audio easier to listen to on headphones.
2884 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2885 so that when listened to on headphones the stereo image is moved from
2886 inside your head (standard for headphones) to outside and in front of
2887 the listener (standard for speakers).
2893 Apply a two-pole peaking equalisation (EQ) filter. With this
2894 filter, the signal-level at and around a selected frequency can
2895 be increased or decreased, whilst (unlike bandpass and bandreject
2896 filters) that at all other frequencies is unchanged.
2898 In order to produce complex equalisation curves, this filter can
2899 be given several times, each with a different central frequency.
2901 The filter accepts the following options:
2905 Set the filter's central frequency in Hz.
2908 Set method to specify band-width of filter.
2923 Specify the band-width of a filter in width_type units.
2926 Set the required gain or attenuation in dB.
2927 Beware of clipping when using a positive gain.
2930 Specify which channels to filter, by default all available are filtered.
2933 @subsection Examples
2936 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2938 equalizer=f=1000:t=h:width=200:g=-10
2942 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2944 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2948 @subsection Commands
2950 This filter supports the following commands:
2953 Change equalizer frequency.
2954 Syntax for the command is : "@var{frequency}"
2957 Change equalizer width_type.
2958 Syntax for the command is : "@var{width_type}"
2961 Change equalizer width.
2962 Syntax for the command is : "@var{width}"
2965 Change equalizer gain.
2966 Syntax for the command is : "@var{gain}"
2969 @section extrastereo
2971 Linearly increases the difference between left and right channels which
2972 adds some sort of "live" effect to playback.
2974 The filter accepts the following options:
2978 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2979 (average of both channels), with 1.0 sound will be unchanged, with
2980 -1.0 left and right channels will be swapped.
2983 Enable clipping. By default is enabled.
2986 @section firequalizer
2987 Apply FIR Equalization using arbitrary frequency response.
2989 The filter accepts the following option:
2993 Set gain curve equation (in dB). The expression can contain variables:
2996 the evaluated frequency
3000 channel number, set to 0 when multichannels evaluation is disabled
3002 channel id, see libavutil/channel_layout.h, set to the first channel id when
3003 multichannels evaluation is disabled
3007 channel_layout, see libavutil/channel_layout.h
3012 @item gain_interpolate(f)
3013 interpolate gain on frequency f based on gain_entry
3014 @item cubic_interpolate(f)
3015 same as gain_interpolate, but smoother
3017 This option is also available as command. Default is @code{gain_interpolate(f)}.
3020 Set gain entry for gain_interpolate function. The expression can
3024 store gain entry at frequency f with value g
3026 This option is also available as command.
3029 Set filter delay in seconds. Higher value means more accurate.
3030 Default is @code{0.01}.
3033 Set filter accuracy in Hz. Lower value means more accurate.
3034 Default is @code{5}.
3037 Set window function. Acceptable values are:
3040 rectangular window, useful when gain curve is already smooth
3042 hann window (default)
3048 3-terms continuous 1st derivative nuttall window
3050 minimum 3-terms discontinuous nuttall window
3052 4-terms continuous 1st derivative nuttall window
3054 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3056 blackman-harris window
3062 If enabled, use fixed number of audio samples. This improves speed when
3063 filtering with large delay. Default is disabled.
3066 Enable multichannels evaluation on gain. Default is disabled.
3069 Enable zero phase mode by subtracting timestamp to compensate delay.
3070 Default is disabled.
3073 Set scale used by gain. Acceptable values are:
3076 linear frequency, linear gain
3078 linear frequency, logarithmic (in dB) gain (default)
3080 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3082 logarithmic frequency, logarithmic gain
3086 Set file for dumping, suitable for gnuplot.
3089 Set scale for dumpfile. Acceptable values are same with scale option.
3093 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3094 Default is disabled.
3097 Enable minimum phase impulse response. Default is disabled.
3100 @subsection Examples
3105 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3108 lowpass at 1000 Hz with gain_entry:
3110 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3113 custom equalization:
3115 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3118 higher delay with zero phase to compensate delay:
3120 firequalizer=delay=0.1:fixed=on:zero_phase=on
3123 lowpass on left channel, highpass on right channel:
3125 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3126 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3131 Apply a flanging effect to the audio.
3133 The filter accepts the following options:
3137 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3140 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3143 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3147 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3148 Default value is 71.
3151 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3154 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3155 Default value is @var{sinusoidal}.
3158 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3159 Default value is 25.
3162 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3163 Default is @var{linear}.
3167 Apply Haas effect to audio.
3169 Note that this makes most sense to apply on mono signals.
3170 With this filter applied to mono signals it give some directionality and
3171 stretches its stereo image.
3173 The filter accepts the following options:
3177 Set input level. By default is @var{1}, or 0dB
3180 Set output level. By default is @var{1}, or 0dB.
3183 Set gain applied to side part of signal. By default is @var{1}.
3186 Set kind of middle source. Can be one of the following:
3196 Pick middle part signal of stereo image.
3199 Pick side part signal of stereo image.
3203 Change middle phase. By default is disabled.
3206 Set left channel delay. By default is @var{2.05} milliseconds.
3209 Set left channel balance. By default is @var{-1}.
3212 Set left channel gain. By default is @var{1}.
3215 Change left phase. By default is disabled.
3218 Set right channel delay. By defaults is @var{2.12} milliseconds.
3221 Set right channel balance. By default is @var{1}.
3224 Set right channel gain. By default is @var{1}.
3227 Change right phase. By default is enabled.
3232 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3233 embedded HDCD codes is expanded into a 20-bit PCM stream.
3235 The filter supports the Peak Extend and Low-level Gain Adjustment features
3236 of HDCD, and detects the Transient Filter flag.
3239 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3242 When using the filter with wav, note the default encoding for wav is 16-bit,
3243 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3244 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3246 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3247 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3250 The filter accepts the following options:
3253 @item disable_autoconvert
3254 Disable any automatic format conversion or resampling in the filter graph.
3256 @item process_stereo
3257 Process the stereo channels together. If target_gain does not match between
3258 channels, consider it invalid and use the last valid target_gain.
3261 Set the code detect timer period in ms.
3264 Always extend peaks above -3dBFS even if PE isn't signaled.
3267 Replace audio with a solid tone and adjust the amplitude to signal some
3268 specific aspect of the decoding process. The output file can be loaded in
3269 an audio editor alongside the original to aid analysis.
3271 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3278 Gain adjustment level at each sample
3280 Samples where peak extend occurs
3282 Samples where the code detect timer is active
3284 Samples where the target gain does not match between channels
3290 Apply head-related transfer functions (HRTFs) to create virtual
3291 loudspeakers around the user for binaural listening via headphones.
3292 The HRIRs are provided via additional streams, for each channel
3293 one stereo input stream is needed.
3295 The filter accepts the following options:
3299 Set mapping of input streams for convolution.
3300 The argument is a '|'-separated list of channel names in order as they
3301 are given as additional stream inputs for filter.
3302 This also specify number of input streams. Number of input streams
3303 must be not less than number of channels in first stream plus one.
3306 Set gain applied to audio. Value is in dB. Default is 0.
3309 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3310 processing audio in time domain which is slow.
3311 @var{freq} is processing audio in frequency domain which is fast.
3312 Default is @var{freq}.
3315 Set custom gain for LFE channels. Value is in dB. Default is 0.
3318 Set size of frame in number of samples which will be processed at once.
3319 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3322 Set format of hrir stream.
3323 Default value is @var{stereo}. Alternative value is @var{multich}.
3324 If value is set to @var{stereo}, number of additional streams should
3325 be greater or equal to number of input channels in first input stream.
3326 Also each additional stream should have stereo number of channels.
3327 If value is set to @var{multich}, number of additional streams should
3328 be exactly one. Also number of input channels of additional stream
3329 should be equal or greater than twice number of channels of first input
3333 @subsection Examples
3337 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3338 each amovie filter use stereo file with IR coefficients as input.
3339 The files give coefficients for each position of virtual loudspeaker:
3341 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"
3346 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3347 but now in @var{multich} @var{hrir} format.
3349 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"
3356 Apply a high-pass filter with 3dB point frequency.
3357 The filter can be either single-pole, or double-pole (the default).
3358 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3360 The filter accepts the following options:
3364 Set frequency in Hz. Default is 3000.
3367 Set number of poles. Default is 2.
3370 Set method to specify band-width of filter.
3385 Specify the band-width of a filter in width_type units.
3386 Applies only to double-pole filter.
3387 The default is 0.707q and gives a Butterworth response.
3390 Specify which channels to filter, by default all available are filtered.
3393 @subsection Commands
3395 This filter supports the following commands:
3398 Change highpass frequency.
3399 Syntax for the command is : "@var{frequency}"
3402 Change highpass width_type.
3403 Syntax for the command is : "@var{width_type}"
3406 Change highpass width.
3407 Syntax for the command is : "@var{width}"
3412 Join multiple input streams into one multi-channel stream.
3414 It accepts the following parameters:
3418 The number of input streams. It defaults to 2.
3420 @item channel_layout
3421 The desired output channel layout. It defaults to stereo.
3424 Map channels from inputs to output. The argument is a '|'-separated list of
3425 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3426 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3427 can be either the name of the input channel (e.g. FL for front left) or its
3428 index in the specified input stream. @var{out_channel} is the name of the output
3432 The filter will attempt to guess the mappings when they are not specified
3433 explicitly. It does so by first trying to find an unused matching input channel
3434 and if that fails it picks the first unused input channel.
3436 Join 3 inputs (with properly set channel layouts):
3438 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3441 Build a 5.1 output from 6 single-channel streams:
3443 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3444 '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'
3450 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3452 To enable compilation of this filter you need to configure FFmpeg with
3453 @code{--enable-ladspa}.
3457 Specifies the name of LADSPA plugin library to load. If the environment
3458 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3459 each one of the directories specified by the colon separated list in
3460 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3461 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3462 @file{/usr/lib/ladspa/}.
3465 Specifies the plugin within the library. Some libraries contain only
3466 one plugin, but others contain many of them. If this is not set filter
3467 will list all available plugins within the specified library.
3470 Set the '|' separated list of controls which are zero or more floating point
3471 values that determine the behavior of the loaded plugin (for example delay,
3473 Controls need to be defined using the following syntax:
3474 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3475 @var{valuei} is the value set on the @var{i}-th control.
3476 Alternatively they can be also defined using the following syntax:
3477 @var{value0}|@var{value1}|@var{value2}|..., where
3478 @var{valuei} is the value set on the @var{i}-th control.
3479 If @option{controls} is set to @code{help}, all available controls and
3480 their valid ranges are printed.
3482 @item sample_rate, s
3483 Specify the sample rate, default to 44100. Only used if plugin have
3487 Set the number of samples per channel per each output frame, default
3488 is 1024. Only used if plugin have zero inputs.
3491 Set the minimum duration of the sourced audio. See
3492 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3493 for the accepted syntax.
3494 Note that the resulting duration may be greater than the specified duration,
3495 as the generated audio is always cut at the end of a complete frame.
3496 If not specified, or the expressed duration is negative, the audio is
3497 supposed to be generated forever.
3498 Only used if plugin have zero inputs.
3502 @subsection Examples
3506 List all available plugins within amp (LADSPA example plugin) library:
3512 List all available controls and their valid ranges for @code{vcf_notch}
3513 plugin from @code{VCF} library:
3515 ladspa=f=vcf:p=vcf_notch:c=help
3519 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3522 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3526 Add reverberation to the audio using TAP-plugins
3527 (Tom's Audio Processing plugins):
3529 ladspa=file=tap_reverb:tap_reverb
3533 Generate white noise, with 0.2 amplitude:
3535 ladspa=file=cmt:noise_source_white:c=c0=.2
3539 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3540 @code{C* Audio Plugin Suite} (CAPS) library:
3542 ladspa=file=caps:Click:c=c1=20'
3546 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3548 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3552 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3553 @code{SWH Plugins} collection:
3555 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3559 Attenuate low frequencies using Multiband EQ from Steve Harris
3560 @code{SWH Plugins} collection:
3562 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3566 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3569 ladspa=caps:Narrower
3573 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3575 ladspa=caps:White:.2
3579 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3581 ladspa=caps:Fractal:c=c1=1
3585 Dynamic volume normalization using @code{VLevel} plugin:
3587 ladspa=vlevel-ladspa:vlevel_mono
3591 @subsection Commands
3593 This filter supports the following commands:
3596 Modify the @var{N}-th control value.
3598 If the specified value is not valid, it is ignored and prior one is kept.
3603 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3604 Support for both single pass (livestreams, files) and double pass (files) modes.
3605 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3606 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3607 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3609 The filter accepts the following options:
3613 Set integrated loudness target.
3614 Range is -70.0 - -5.0. Default value is -24.0.
3617 Set loudness range target.
3618 Range is 1.0 - 20.0. Default value is 7.0.
3621 Set maximum true peak.
3622 Range is -9.0 - +0.0. Default value is -2.0.
3624 @item measured_I, measured_i
3625 Measured IL of input file.
3626 Range is -99.0 - +0.0.
3628 @item measured_LRA, measured_lra
3629 Measured LRA of input file.
3630 Range is 0.0 - 99.0.
3632 @item measured_TP, measured_tp
3633 Measured true peak of input file.
3634 Range is -99.0 - +99.0.
3636 @item measured_thresh
3637 Measured threshold of input file.
3638 Range is -99.0 - +0.0.
3641 Set offset gain. Gain is applied before the true-peak limiter.
3642 Range is -99.0 - +99.0. Default is +0.0.
3645 Normalize linearly if possible.
3646 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3647 to be specified in order to use this mode.
3648 Options are true or false. Default is true.
3651 Treat mono input files as "dual-mono". If a mono file is intended for playback
3652 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3653 If set to @code{true}, this option will compensate for this effect.
3654 Multi-channel input files are not affected by this option.
3655 Options are true or false. Default is false.
3658 Set print format for stats. Options are summary, json, or none.
3659 Default value is none.
3664 Apply a low-pass filter with 3dB point frequency.
3665 The filter can be either single-pole or double-pole (the default).
3666 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3668 The filter accepts the following options:
3672 Set frequency in Hz. Default is 500.
3675 Set number of poles. Default is 2.
3678 Set method to specify band-width of filter.
3693 Specify the band-width of a filter in width_type units.
3694 Applies only to double-pole filter.
3695 The default is 0.707q and gives a Butterworth response.
3698 Specify which channels to filter, by default all available are filtered.
3701 @subsection Examples
3704 Lowpass only LFE channel, it LFE is not present it does nothing:
3710 @subsection Commands
3712 This filter supports the following commands:
3715 Change lowpass frequency.
3716 Syntax for the command is : "@var{frequency}"
3719 Change lowpass width_type.
3720 Syntax for the command is : "@var{width_type}"
3723 Change lowpass width.
3724 Syntax for the command is : "@var{width}"
3729 Load a LV2 (LADSPA Version 2) plugin.
3731 To enable compilation of this filter you need to configure FFmpeg with
3732 @code{--enable-lv2}.
3736 Specifies the plugin URI. You may need to escape ':'.
3739 Set the '|' separated list of controls which are zero or more floating point
3740 values that determine the behavior of the loaded plugin (for example delay,
3742 If @option{controls} is set to @code{help}, all available controls and
3743 their valid ranges are printed.
3745 @item sample_rate, s
3746 Specify the sample rate, default to 44100. Only used if plugin have
3750 Set the number of samples per channel per each output frame, default
3751 is 1024. Only used if plugin have zero inputs.
3754 Set the minimum duration of the sourced audio. See
3755 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3756 for the accepted syntax.
3757 Note that the resulting duration may be greater than the specified duration,
3758 as the generated audio is always cut at the end of a complete frame.
3759 If not specified, or the expressed duration is negative, the audio is
3760 supposed to be generated forever.
3761 Only used if plugin have zero inputs.
3764 @subsection Examples
3768 Apply bass enhancer plugin from Calf:
3770 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3774 Apply vinyl plugin from Calf:
3776 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3780 Apply bit crusher plugin from ArtyFX:
3782 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3787 Multiband Compress or expand the audio's dynamic range.
3789 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3790 This is akin to the crossover of a loudspeaker, and results in flat frequency
3791 response when absent compander action.
3793 It accepts the following parameters:
3797 This option syntax is:
3798 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3799 For explanation of each item refer to compand filter documentation.
3805 Mix channels with specific gain levels. The filter accepts the output
3806 channel layout followed by a set of channels definitions.
3808 This filter is also designed to efficiently remap the channels of an audio
3811 The filter accepts parameters of the form:
3812 "@var{l}|@var{outdef}|@var{outdef}|..."
3816 output channel layout or number of channels
3819 output channel specification, of the form:
3820 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3823 output channel to define, either a channel name (FL, FR, etc.) or a channel
3824 number (c0, c1, etc.)
3827 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3830 input channel to use, see out_name for details; it is not possible to mix
3831 named and numbered input channels
3834 If the `=' in a channel specification is replaced by `<', then the gains for
3835 that specification will be renormalized so that the total is 1, thus
3836 avoiding clipping noise.
3838 @subsection Mixing examples
3840 For example, if you want to down-mix from stereo to mono, but with a bigger
3841 factor for the left channel:
3843 pan=1c|c0=0.9*c0+0.1*c1
3846 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3847 7-channels surround:
3849 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3852 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3853 that should be preferred (see "-ac" option) unless you have very specific
3856 @subsection Remapping examples
3858 The channel remapping will be effective if, and only if:
3861 @item gain coefficients are zeroes or ones,
3862 @item only one input per channel output,
3865 If all these conditions are satisfied, the filter will notify the user ("Pure
3866 channel mapping detected"), and use an optimized and lossless method to do the
3869 For example, if you have a 5.1 source and want a stereo audio stream by
3870 dropping the extra channels:
3872 pan="stereo| c0=FL | c1=FR"
3875 Given the same source, you can also switch front left and front right channels
3876 and keep the input channel layout:
3878 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3881 If the input is a stereo audio stream, you can mute the front left channel (and
3882 still keep the stereo channel layout) with:
3887 Still with a stereo audio stream input, you can copy the right channel in both
3888 front left and right:
3890 pan="stereo| c0=FR | c1=FR"
3895 ReplayGain scanner filter. This filter takes an audio stream as an input and
3896 outputs it unchanged.
3897 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3901 Convert the audio sample format, sample rate and channel layout. It is
3902 not meant to be used directly.
3905 Apply time-stretching and pitch-shifting with librubberband.
3907 To enable compilation of this filter, you need to configure FFmpeg with
3908 @code{--enable-librubberband}.
3910 The filter accepts the following options:
3914 Set tempo scale factor.
3917 Set pitch scale factor.
3920 Set transients detector.
3921 Possible values are:
3930 Possible values are:
3939 Possible values are:
3946 Set processing window size.
3947 Possible values are:
3956 Possible values are:
3963 Enable formant preservation when shift pitching.
3964 Possible values are:
3972 Possible values are:
3981 Possible values are:
3988 @section sidechaincompress
3990 This filter acts like normal compressor but has the ability to compress
3991 detected signal using second input signal.
3992 It needs two input streams and returns one output stream.
3993 First input stream will be processed depending on second stream signal.
3994 The filtered signal then can be filtered with other filters in later stages of
3995 processing. See @ref{pan} and @ref{amerge} filter.
3997 The filter accepts the following options:
4001 Set input gain. Default is 1. Range is between 0.015625 and 64.
4004 If a signal of second stream raises above this level it will affect the gain
4005 reduction of first stream.
4006 By default is 0.125. Range is between 0.00097563 and 1.
4009 Set a ratio about which the signal is reduced. 1:2 means that if the level
4010 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4011 Default is 2. Range is between 1 and 20.
4014 Amount of milliseconds the signal has to rise above the threshold before gain
4015 reduction starts. Default is 20. Range is between 0.01 and 2000.
4018 Amount of milliseconds the signal has to fall below the threshold before
4019 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4022 Set the amount by how much signal will be amplified after processing.
4023 Default is 1. Range is from 1 to 64.
4026 Curve the sharp knee around the threshold to enter gain reduction more softly.
4027 Default is 2.82843. Range is between 1 and 8.
4030 Choose if the @code{average} level between all channels of side-chain stream
4031 or the louder(@code{maximum}) channel of side-chain stream affects the
4032 reduction. Default is @code{average}.
4035 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4036 of @code{rms}. Default is @code{rms} which is mainly smoother.
4039 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4042 How much to use compressed signal in output. Default is 1.
4043 Range is between 0 and 1.
4046 @subsection Examples
4050 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4051 depending on the signal of 2nd input and later compressed signal to be
4052 merged with 2nd input:
4054 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4058 @section sidechaingate
4060 A sidechain gate acts like a normal (wideband) gate but has the ability to
4061 filter the detected signal before sending it to the gain reduction stage.
4062 Normally a gate uses the full range signal to detect a level above the
4064 For example: If you cut all lower frequencies from your sidechain signal
4065 the gate will decrease the volume of your track only if not enough highs
4066 appear. With this technique you are able to reduce the resonation of a
4067 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4069 It needs two input streams and returns one output stream.
4070 First input stream will be processed depending on second stream signal.
4072 The filter accepts the following options:
4076 Set input level before filtering.
4077 Default is 1. Allowed range is from 0.015625 to 64.
4080 Set the level of gain reduction when the signal is below the threshold.
4081 Default is 0.06125. Allowed range is from 0 to 1.
4084 If a signal rises above this level the gain reduction is released.
4085 Default is 0.125. Allowed range is from 0 to 1.
4088 Set a ratio about which the signal is reduced.
4089 Default is 2. Allowed range is from 1 to 9000.
4092 Amount of milliseconds the signal has to rise above the threshold before gain
4094 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4097 Amount of milliseconds the signal has to fall below the threshold before the
4098 reduction is increased again. Default is 250 milliseconds.
4099 Allowed range is from 0.01 to 9000.
4102 Set amount of amplification of signal after processing.
4103 Default is 1. Allowed range is from 1 to 64.
4106 Curve the sharp knee around the threshold to enter gain reduction more softly.
4107 Default is 2.828427125. Allowed range is from 1 to 8.
4110 Choose if exact signal should be taken for detection or an RMS like one.
4111 Default is rms. Can be peak or rms.
4114 Choose if the average level between all channels or the louder channel affects
4116 Default is average. Can be average or maximum.
4119 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4122 @section silencedetect
4124 Detect silence in an audio stream.
4126 This filter logs a message when it detects that the input audio volume is less
4127 or equal to a noise tolerance value for a duration greater or equal to the
4128 minimum detected noise duration.
4130 The printed times and duration are expressed in seconds.
4132 The filter accepts the following options:
4136 Set silence duration until notification (default is 2 seconds).
4139 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4140 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4143 @subsection Examples
4147 Detect 5 seconds of silence with -50dB noise tolerance:
4149 silencedetect=n=-50dB:d=5
4153 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4154 tolerance in @file{silence.mp3}:
4156 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4160 @section silenceremove
4162 Remove silence from the beginning, middle or end of the audio.
4164 The filter accepts the following options:
4168 This value is used to indicate if audio should be trimmed at beginning of
4169 the audio. A value of zero indicates no silence should be trimmed from the
4170 beginning. When specifying a non-zero value, it trims audio up until it
4171 finds non-silence. Normally, when trimming silence from beginning of audio
4172 the @var{start_periods} will be @code{1} but it can be increased to higher
4173 values to trim all audio up to specific count of non-silence periods.
4174 Default value is @code{0}.
4176 @item start_duration
4177 Specify the amount of time that non-silence must be detected before it stops
4178 trimming audio. By increasing the duration, bursts of noises can be treated
4179 as silence and trimmed off. Default value is @code{0}.
4181 @item start_threshold
4182 This indicates what sample value should be treated as silence. For digital
4183 audio, a value of @code{0} may be fine but for audio recorded from analog,
4184 you may wish to increase the value to account for background noise.
4185 Can be specified in dB (in case "dB" is appended to the specified value)
4186 or amplitude ratio. Default value is @code{0}.
4189 Set the count for trimming silence from the end of audio.
4190 To remove silence from the middle of a file, specify a @var{stop_periods}
4191 that is negative. This value is then treated as a positive value and is
4192 used to indicate the effect should restart processing as specified by
4193 @var{start_periods}, making it suitable for removing periods of silence
4194 in the middle of the audio.
4195 Default value is @code{0}.
4198 Specify a duration of silence that must exist before audio is not copied any
4199 more. By specifying a higher duration, silence that is wanted can be left in
4201 Default value is @code{0}.
4203 @item stop_threshold
4204 This is the same as @option{start_threshold} but for trimming silence from
4206 Can be specified in dB (in case "dB" is appended to the specified value)
4207 or amplitude ratio. Default value is @code{0}.
4210 This indicates that @var{stop_duration} length of audio should be left intact
4211 at the beginning of each period of silence.
4212 For example, if you want to remove long pauses between words but do not want
4213 to remove the pauses completely. Default value is @code{0}.
4216 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4217 and works better with digital silence which is exactly 0.
4218 Default value is @code{rms}.
4221 Set ratio used to calculate size of window for detecting silence.
4222 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4225 @subsection Examples
4229 The following example shows how this filter can be used to start a recording
4230 that does not contain the delay at the start which usually occurs between
4231 pressing the record button and the start of the performance:
4233 silenceremove=1:5:0.02
4237 Trim all silence encountered from beginning to end where there is more than 1
4238 second of silence in audio:
4240 silenceremove=0:0:0:-1:1:-90dB
4246 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4247 loudspeakers around the user for binaural listening via headphones (audio
4248 formats up to 9 channels supported).
4249 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4250 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4251 Austrian Academy of Sciences.
4253 To enable compilation of this filter you need to configure FFmpeg with
4254 @code{--enable-libmysofa}.
4256 The filter accepts the following options:
4260 Set the SOFA file used for rendering.
4263 Set gain applied to audio. Value is in dB. Default is 0.
4266 Set rotation of virtual loudspeakers in deg. Default is 0.
4269 Set elevation of virtual speakers in deg. Default is 0.
4272 Set distance in meters between loudspeakers and the listener with near-field
4273 HRTFs. Default is 1.
4276 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4277 processing audio in time domain which is slow.
4278 @var{freq} is processing audio in frequency domain which is fast.
4279 Default is @var{freq}.
4282 Set custom positions of virtual loudspeakers. Syntax for this option is:
4283 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4284 Each virtual loudspeaker is described with short channel name following with
4285 azimuth and elevation in degrees.
4286 Each virtual loudspeaker description is separated by '|'.
4287 For example to override front left and front right channel positions use:
4288 'speakers=FL 45 15|FR 345 15'.
4289 Descriptions with unrecognised channel names are ignored.
4292 Set custom gain for LFE channels. Value is in dB. Default is 0.
4295 @subsection Examples
4299 Using ClubFritz6 sofa file:
4301 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4305 Using ClubFritz12 sofa file and bigger radius with small rotation:
4307 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4311 Similar as above but with custom speaker positions for front left, front right, back left and back right
4312 and also with custom gain:
4314 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4318 @section stereotools
4320 This filter has some handy utilities to manage stereo signals, for converting
4321 M/S stereo recordings to L/R signal while having control over the parameters
4322 or spreading the stereo image of master track.
4324 The filter accepts the following options:
4328 Set input level before filtering for both channels. Defaults is 1.
4329 Allowed range is from 0.015625 to 64.
4332 Set output level after filtering for both channels. Defaults is 1.
4333 Allowed range is from 0.015625 to 64.
4336 Set input balance between both channels. Default is 0.
4337 Allowed range is from -1 to 1.
4340 Set output balance between both channels. Default is 0.
4341 Allowed range is from -1 to 1.
4344 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4345 clipping. Disabled by default.
4348 Mute the left channel. Disabled by default.
4351 Mute the right channel. Disabled by default.
4354 Change the phase of the left channel. Disabled by default.
4357 Change the phase of the right channel. Disabled by default.
4360 Set stereo mode. Available values are:
4364 Left/Right to Left/Right, this is default.
4367 Left/Right to Mid/Side.
4370 Mid/Side to Left/Right.
4373 Left/Right to Left/Left.
4376 Left/Right to Right/Right.
4379 Left/Right to Left + Right.
4382 Left/Right to Right/Left.
4385 Mid/Side to Left/Left.
4388 Mid/Side to Right/Right.
4392 Set level of side signal. Default is 1.
4393 Allowed range is from 0.015625 to 64.
4396 Set balance of side signal. Default is 0.
4397 Allowed range is from -1 to 1.
4400 Set level of the middle signal. Default is 1.
4401 Allowed range is from 0.015625 to 64.
4404 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4407 Set stereo base between mono and inversed channels. Default is 0.
4408 Allowed range is from -1 to 1.
4411 Set delay in milliseconds how much to delay left from right channel and
4412 vice versa. Default is 0. Allowed range is from -20 to 20.
4415 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4418 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4420 @item bmode_in, bmode_out
4421 Set balance mode for balance_in/balance_out option.
4423 Can be one of the following:
4427 Classic balance mode. Attenuate one channel at time.
4428 Gain is raised up to 1.
4431 Similar as classic mode above but gain is raised up to 2.
4434 Equal power distribution, from -6dB to +6dB range.
4438 @subsection Examples
4442 Apply karaoke like effect:
4444 stereotools=mlev=0.015625
4448 Convert M/S signal to L/R:
4450 "stereotools=mode=ms>lr"
4454 @section stereowiden
4456 This filter enhance the stereo effect by suppressing signal common to both
4457 channels and by delaying the signal of left into right and vice versa,
4458 thereby widening the stereo effect.
4460 The filter accepts the following options:
4464 Time in milliseconds of the delay of left signal into right and vice versa.
4465 Default is 20 milliseconds.
4468 Amount of gain in delayed signal into right and vice versa. Gives a delay
4469 effect of left signal in right output and vice versa which gives widening
4470 effect. Default is 0.3.
4473 Cross feed of left into right with inverted phase. This helps in suppressing
4474 the mono. If the value is 1 it will cancel all the signal common to both
4475 channels. Default is 0.3.
4478 Set level of input signal of original channel. Default is 0.8.
4481 @section superequalizer
4482 Apply 18 band equalizer.
4484 The filter accepts the following options:
4491 Set 131Hz band gain.
4493 Set 185Hz band gain.
4495 Set 262Hz band gain.
4497 Set 370Hz band gain.
4499 Set 523Hz band gain.
4501 Set 740Hz band gain.
4503 Set 1047Hz band gain.
4505 Set 1480Hz band gain.
4507 Set 2093Hz band gain.
4509 Set 2960Hz band gain.
4511 Set 4186Hz band gain.
4513 Set 5920Hz band gain.
4515 Set 8372Hz band gain.
4517 Set 11840Hz band gain.
4519 Set 16744Hz band gain.
4521 Set 20000Hz band gain.
4525 Apply audio surround upmix filter.
4527 This filter allows to produce multichannel output from audio stream.
4529 The filter accepts the following options:
4533 Set output channel layout. By default, this is @var{5.1}.
4535 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4536 for the required syntax.
4539 Set input channel layout. By default, this is @var{stereo}.
4541 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4542 for the required syntax.
4545 Set input volume level. By default, this is @var{1}.
4548 Set output volume level. By default, this is @var{1}.
4551 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4554 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4557 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4560 Set front center input volume. By default, this is @var{1}.
4563 Set front center output volume. By default, this is @var{1}.
4566 Set LFE input volume. By default, this is @var{1}.
4569 Set LFE output volume. By default, this is @var{1}.
4572 @section treble, highshelf
4574 Boost or cut treble (upper) frequencies of the audio using a two-pole
4575 shelving filter with a response similar to that of a standard
4576 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4578 The filter accepts the following options:
4582 Give the gain at whichever is the lower of ~22 kHz and the
4583 Nyquist frequency. Its useful range is about -20 (for a large cut)
4584 to +20 (for a large boost). Beware of clipping when using a positive gain.
4587 Set the filter's central frequency and so can be used
4588 to extend or reduce the frequency range to be boosted or cut.
4589 The default value is @code{3000} Hz.
4592 Set method to specify band-width of filter.
4607 Determine how steep is the filter's shelf transition.
4610 Specify which channels to filter, by default all available are filtered.
4613 @subsection Commands
4615 This filter supports the following commands:
4618 Change treble frequency.
4619 Syntax for the command is : "@var{frequency}"
4622 Change treble width_type.
4623 Syntax for the command is : "@var{width_type}"
4626 Change treble width.
4627 Syntax for the command is : "@var{width}"
4631 Syntax for the command is : "@var{gain}"
4636 Sinusoidal amplitude modulation.
4638 The filter accepts the following options:
4642 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4643 (20 Hz or lower) will result in a tremolo effect.
4644 This filter may also be used as a ring modulator by specifying
4645 a modulation frequency higher than 20 Hz.
4646 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4649 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4650 Default value is 0.5.
4655 Sinusoidal phase modulation.
4657 The filter accepts the following options:
4661 Modulation frequency in Hertz.
4662 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4665 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4666 Default value is 0.5.
4671 Adjust the input audio volume.
4673 It accepts the following parameters:
4677 Set audio volume expression.
4679 Output values are clipped to the maximum value.
4681 The output audio volume is given by the relation:
4683 @var{output_volume} = @var{volume} * @var{input_volume}
4686 The default value for @var{volume} is "1.0".
4689 This parameter represents the mathematical precision.
4691 It determines which input sample formats will be allowed, which affects the
4692 precision of the volume scaling.
4696 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4698 32-bit floating-point; this limits input sample format to FLT. (default)
4700 64-bit floating-point; this limits input sample format to DBL.
4704 Choose the behaviour on encountering ReplayGain side data in input frames.
4708 Remove ReplayGain side data, ignoring its contents (the default).
4711 Ignore ReplayGain side data, but leave it in the frame.
4714 Prefer the track gain, if present.
4717 Prefer the album gain, if present.
4720 @item replaygain_preamp
4721 Pre-amplification gain in dB to apply to the selected replaygain gain.
4723 Default value for @var{replaygain_preamp} is 0.0.
4726 Set when the volume expression is evaluated.
4728 It accepts the following values:
4731 only evaluate expression once during the filter initialization, or
4732 when the @samp{volume} command is sent
4735 evaluate expression for each incoming frame
4738 Default value is @samp{once}.
4741 The volume expression can contain the following parameters.
4745 frame number (starting at zero)
4748 @item nb_consumed_samples
4749 number of samples consumed by the filter
4751 number of samples in the current frame
4753 original frame position in the file
4759 PTS at start of stream
4761 time at start of stream
4767 last set volume value
4770 Note that when @option{eval} is set to @samp{once} only the
4771 @var{sample_rate} and @var{tb} variables are available, all other
4772 variables will evaluate to NAN.
4774 @subsection Commands
4776 This filter supports the following commands:
4779 Modify the volume expression.
4780 The command accepts the same syntax of the corresponding option.
4782 If the specified expression is not valid, it is kept at its current
4784 @item replaygain_noclip
4785 Prevent clipping by limiting the gain applied.
4787 Default value for @var{replaygain_noclip} is 1.
4791 @subsection Examples
4795 Halve the input audio volume:
4799 volume=volume=-6.0206dB
4802 In all the above example the named key for @option{volume} can be
4803 omitted, for example like in:
4809 Increase input audio power by 6 decibels using fixed-point precision:
4811 volume=volume=6dB:precision=fixed
4815 Fade volume after time 10 with an annihilation period of 5 seconds:
4817 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4821 @section volumedetect
4823 Detect the volume of the input video.
4825 The filter has no parameters. The input is not modified. Statistics about
4826 the volume will be printed in the log when the input stream end is reached.
4828 In particular it will show the mean volume (root mean square), maximum
4829 volume (on a per-sample basis), and the beginning of a histogram of the
4830 registered volume values (from the maximum value to a cumulated 1/1000 of
4833 All volumes are in decibels relative to the maximum PCM value.
4835 @subsection Examples
4837 Here is an excerpt of the output:
4839 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4840 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4841 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4842 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4843 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4844 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4845 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4846 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4847 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4853 The mean square energy is approximately -27 dB, or 10^-2.7.
4855 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4857 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4860 In other words, raising the volume by +4 dB does not cause any clipping,
4861 raising it by +5 dB causes clipping for 6 samples, etc.
4863 @c man end AUDIO FILTERS
4865 @chapter Audio Sources
4866 @c man begin AUDIO SOURCES
4868 Below is a description of the currently available audio sources.
4872 Buffer audio frames, and make them available to the filter chain.
4874 This source is mainly intended for a programmatic use, in particular
4875 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4877 It accepts the following parameters:
4881 The timebase which will be used for timestamps of submitted frames. It must be
4882 either a floating-point number or in @var{numerator}/@var{denominator} form.
4885 The sample rate of the incoming audio buffers.
4888 The sample format of the incoming audio buffers.
4889 Either a sample format name or its corresponding integer representation from
4890 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4892 @item channel_layout
4893 The channel layout of the incoming audio buffers.
4894 Either a channel layout name from channel_layout_map in
4895 @file{libavutil/channel_layout.c} or its corresponding integer representation
4896 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4899 The number of channels of the incoming audio buffers.
4900 If both @var{channels} and @var{channel_layout} are specified, then they
4905 @subsection Examples
4908 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4911 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4912 Since the sample format with name "s16p" corresponds to the number
4913 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4916 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4921 Generate an audio signal specified by an expression.
4923 This source accepts in input one or more expressions (one for each
4924 channel), which are evaluated and used to generate a corresponding
4927 This source accepts the following options:
4931 Set the '|'-separated expressions list for each separate channel. In case the
4932 @option{channel_layout} option is not specified, the selected channel layout
4933 depends on the number of provided expressions. Otherwise the last
4934 specified expression is applied to the remaining output channels.
4936 @item channel_layout, c
4937 Set the channel layout. The number of channels in the specified layout
4938 must be equal to the number of specified expressions.
4941 Set the minimum duration of the sourced audio. See
4942 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4943 for the accepted syntax.
4944 Note that the resulting duration may be greater than the specified
4945 duration, as the generated audio is always cut at the end of a
4948 If not specified, or the expressed duration is negative, the audio is
4949 supposed to be generated forever.
4952 Set the number of samples per channel per each output frame,
4955 @item sample_rate, s
4956 Specify the sample rate, default to 44100.
4959 Each expression in @var{exprs} can contain the following constants:
4963 number of the evaluated sample, starting from 0
4966 time of the evaluated sample expressed in seconds, starting from 0
4973 @subsection Examples
4983 Generate a sin signal with frequency of 440 Hz, set sample rate to
4986 aevalsrc="sin(440*2*PI*t):s=8000"
4990 Generate a two channels signal, specify the channel layout (Front
4991 Center + Back Center) explicitly:
4993 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4997 Generate white noise:
4999 aevalsrc="-2+random(0)"
5003 Generate an amplitude modulated signal:
5005 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5009 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5011 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5018 The null audio source, return unprocessed audio frames. It is mainly useful
5019 as a template and to be employed in analysis / debugging tools, or as
5020 the source for filters which ignore the input data (for example the sox
5023 This source accepts the following options:
5027 @item channel_layout, cl
5029 Specifies the channel layout, and can be either an integer or a string
5030 representing a channel layout. The default value of @var{channel_layout}
5033 Check the channel_layout_map definition in
5034 @file{libavutil/channel_layout.c} for the mapping between strings and
5035 channel layout values.
5037 @item sample_rate, r
5038 Specifies the sample rate, and defaults to 44100.
5041 Set the number of samples per requested frames.
5045 @subsection Examples
5049 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5051 anullsrc=r=48000:cl=4
5055 Do the same operation with a more obvious syntax:
5057 anullsrc=r=48000:cl=mono
5061 All the parameters need to be explicitly defined.
5065 Synthesize a voice utterance using the libflite library.
5067 To enable compilation of this filter you need to configure FFmpeg with
5068 @code{--enable-libflite}.
5070 Note that versions of the flite library prior to 2.0 are not thread-safe.
5072 The filter accepts the following options:
5077 If set to 1, list the names of the available voices and exit
5078 immediately. Default value is 0.
5081 Set the maximum number of samples per frame. Default value is 512.
5084 Set the filename containing the text to speak.
5087 Set the text to speak.
5090 Set the voice to use for the speech synthesis. Default value is
5091 @code{kal}. See also the @var{list_voices} option.
5094 @subsection Examples
5098 Read from file @file{speech.txt}, and synthesize the text using the
5099 standard flite voice:
5101 flite=textfile=speech.txt
5105 Read the specified text selecting the @code{slt} voice:
5107 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5111 Input text to ffmpeg:
5113 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5117 Make @file{ffplay} speak the specified text, using @code{flite} and
5118 the @code{lavfi} device:
5120 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5124 For more information about libflite, check:
5125 @url{http://www.festvox.org/flite/}
5129 Generate a noise audio signal.
5131 The filter accepts the following options:
5134 @item sample_rate, r
5135 Specify the sample rate. Default value is 48000 Hz.
5138 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5142 Specify the duration of the generated audio stream. Not specifying this option
5143 results in noise with an infinite length.
5145 @item color, colour, c
5146 Specify the color of noise. Available noise colors are white, pink, brown,
5147 blue and violet. Default color is white.
5150 Specify a value used to seed the PRNG.
5153 Set the number of samples per each output frame, default is 1024.
5156 @subsection Examples
5161 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5163 anoisesrc=d=60:c=pink:r=44100:a=0.5
5169 Generate odd-tap Hilbert transform FIR coefficients.
5171 The resulting stream can be used with @ref{afir} filter for phase-shifting
5172 the signal by 90 degrees.
5174 This is used in many matrix coding schemes and for analytic signal generation.
5175 The process is often written as a multiplication by i (or j), the imaginary unit.
5177 The filter accepts the following options:
5181 @item sample_rate, s
5182 Set sample rate, default is 44100.
5185 Set length of FIR filter, default is 22051.
5188 Set number of samples per each frame.
5191 Set window function to be used when generating FIR coefficients.
5196 Generate an audio signal made of a sine wave with amplitude 1/8.
5198 The audio signal is bit-exact.
5200 The filter accepts the following options:
5205 Set the carrier frequency. Default is 440 Hz.
5207 @item beep_factor, b
5208 Enable a periodic beep every second with frequency @var{beep_factor} times
5209 the carrier frequency. Default is 0, meaning the beep is disabled.
5211 @item sample_rate, r
5212 Specify the sample rate, default is 44100.
5215 Specify the duration of the generated audio stream.
5217 @item samples_per_frame
5218 Set the number of samples per output frame.
5220 The expression can contain the following constants:
5224 The (sequential) number of the output audio frame, starting from 0.
5227 The PTS (Presentation TimeStamp) of the output audio frame,
5228 expressed in @var{TB} units.
5231 The PTS of the output audio frame, expressed in seconds.
5234 The timebase of the output audio frames.
5237 Default is @code{1024}.
5240 @subsection Examples
5245 Generate a simple 440 Hz sine wave:
5251 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5255 sine=frequency=220:beep_factor=4:duration=5
5259 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5262 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5266 @c man end AUDIO SOURCES
5268 @chapter Audio Sinks
5269 @c man begin AUDIO SINKS
5271 Below is a description of the currently available audio sinks.
5273 @section abuffersink
5275 Buffer audio frames, and make them available to the end of filter chain.
5277 This sink is mainly intended for programmatic use, in particular
5278 through the interface defined in @file{libavfilter/buffersink.h}
5279 or the options system.
5281 It accepts a pointer to an AVABufferSinkContext structure, which
5282 defines the incoming buffers' formats, to be passed as the opaque
5283 parameter to @code{avfilter_init_filter} for initialization.
5286 Null audio sink; do absolutely nothing with the input audio. It is
5287 mainly useful as a template and for use in analysis / debugging
5290 @c man end AUDIO SINKS
5292 @chapter Video Filters
5293 @c man begin VIDEO FILTERS
5295 When you configure your FFmpeg build, you can disable any of the
5296 existing filters using @code{--disable-filters}.
5297 The configure output will show the video filters included in your
5300 Below is a description of the currently available video filters.
5302 @section alphaextract
5304 Extract the alpha component from the input as a grayscale video. This
5305 is especially useful with the @var{alphamerge} filter.
5309 Add or replace the alpha component of the primary input with the
5310 grayscale value of a second input. This is intended for use with
5311 @var{alphaextract} to allow the transmission or storage of frame
5312 sequences that have alpha in a format that doesn't support an alpha
5315 For example, to reconstruct full frames from a normal YUV-encoded video
5316 and a separate video created with @var{alphaextract}, you might use:
5318 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5321 Since this filter is designed for reconstruction, it operates on frame
5322 sequences without considering timestamps, and terminates when either
5323 input reaches end of stream. This will cause problems if your encoding
5324 pipeline drops frames. If you're trying to apply an image as an
5325 overlay to a video stream, consider the @var{overlay} filter instead.
5329 Amplify differences between current pixel and pixels of adjacent frames in
5330 same pixel location.
5332 This filter accepts the following options:
5336 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5337 For example radius of 3 will instruct filter to calculate average of 7 frames.
5340 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5343 Set threshold for difference amplification. Any differrence greater or equal to
5344 this value will not alter source pixel. Default is 10.
5345 Allowed range is from 0 to 65535.
5348 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5349 This option controls maximum possible value that will decrease source pixel value.
5352 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5353 This option controls maximum possible value that will increase source pixel value.
5356 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5361 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5362 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5363 Substation Alpha) subtitles files.
5365 This filter accepts the following option in addition to the common options from
5366 the @ref{subtitles} filter:
5370 Set the shaping engine
5372 Available values are:
5375 The default libass shaping engine, which is the best available.
5377 Fast, font-agnostic shaper that can do only substitutions
5379 Slower shaper using OpenType for substitutions and positioning
5382 The default is @code{auto}.
5386 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5388 The filter accepts the following options:
5392 Set threshold A for 1st plane. Default is 0.02.
5393 Valid range is 0 to 0.3.
5396 Set threshold B for 1st plane. Default is 0.04.
5397 Valid range is 0 to 5.
5400 Set threshold A for 2nd plane. Default is 0.02.
5401 Valid range is 0 to 0.3.
5404 Set threshold B for 2nd plane. Default is 0.04.
5405 Valid range is 0 to 5.
5408 Set threshold A for 3rd plane. Default is 0.02.
5409 Valid range is 0 to 0.3.
5412 Set threshold B for 3rd plane. Default is 0.04.
5413 Valid range is 0 to 5.
5415 Threshold A is designed to react on abrupt changes in the input signal and
5416 threshold B is designed to react on continuous changes in the input signal.
5419 Set number of frames filter will use for averaging. Default is 9. Must be odd
5420 number in range [5, 129].
5423 Set what planes of frame filter will use for averaging. Default is all.
5428 Apply average blur filter.
5430 The filter accepts the following options:
5434 Set horizontal radius size.
5437 Set which planes to filter. By default all planes are filtered.
5440 Set vertical radius size, if zero it will be same as @code{sizeX}.
5441 Default is @code{0}.
5446 Compute the bounding box for the non-black pixels in the input frame
5449 This filter computes the bounding box containing all the pixels with a
5450 luminance value greater than the minimum allowed value.
5451 The parameters describing the bounding box are printed on the filter
5454 The filter accepts the following option:
5458 Set the minimal luminance value. Default is @code{16}.
5461 @section bitplanenoise
5463 Show and measure bit plane noise.
5465 The filter accepts the following options:
5469 Set which plane to analyze. Default is @code{1}.
5472 Filter out noisy pixels from @code{bitplane} set above.
5473 Default is disabled.
5476 @section blackdetect
5478 Detect video intervals that are (almost) completely black. Can be
5479 useful to detect chapter transitions, commercials, or invalid
5480 recordings. Output lines contains the time for the start, end and
5481 duration of the detected black interval expressed in seconds.
5483 In order to display the output lines, you need to set the loglevel at
5484 least to the AV_LOG_INFO value.
5486 The filter accepts the following options:
5489 @item black_min_duration, d
5490 Set the minimum detected black duration expressed in seconds. It must
5491 be a non-negative floating point number.
5493 Default value is 2.0.
5495 @item picture_black_ratio_th, pic_th
5496 Set the threshold for considering a picture "black".
5497 Express the minimum value for the ratio:
5499 @var{nb_black_pixels} / @var{nb_pixels}
5502 for which a picture is considered black.
5503 Default value is 0.98.
5505 @item pixel_black_th, pix_th
5506 Set the threshold for considering a pixel "black".
5508 The threshold expresses the maximum pixel luminance value for which a
5509 pixel is considered "black". The provided value is scaled according to
5510 the following equation:
5512 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5515 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5516 the input video format, the range is [0-255] for YUV full-range
5517 formats and [16-235] for YUV non full-range formats.
5519 Default value is 0.10.
5522 The following example sets the maximum pixel threshold to the minimum
5523 value, and detects only black intervals of 2 or more seconds:
5525 blackdetect=d=2:pix_th=0.00
5530 Detect frames that are (almost) completely black. Can be useful to
5531 detect chapter transitions or commercials. Output lines consist of
5532 the frame number of the detected frame, the percentage of blackness,
5533 the position in the file if known or -1 and the timestamp in seconds.
5535 In order to display the output lines, you need to set the loglevel at
5536 least to the AV_LOG_INFO value.
5538 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5539 The value represents the percentage of pixels in the picture that
5540 are below the threshold value.
5542 It accepts the following parameters:
5547 The percentage of the pixels that have to be below the threshold; it defaults to
5550 @item threshold, thresh
5551 The threshold below which a pixel value is considered black; it defaults to
5556 @section blend, tblend
5558 Blend two video frames into each other.
5560 The @code{blend} filter takes two input streams and outputs one
5561 stream, the first input is the "top" layer and second input is
5562 "bottom" layer. By default, the output terminates when the longest input terminates.
5564 The @code{tblend} (time blend) filter takes two consecutive frames
5565 from one single stream, and outputs the result obtained by blending
5566 the new frame on top of the old frame.
5568 A description of the accepted options follows.
5576 Set blend mode for specific pixel component or all pixel components in case
5577 of @var{all_mode}. Default value is @code{normal}.
5579 Available values for component modes are:
5621 Set blend opacity for specific pixel component or all pixel components in case
5622 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5629 Set blend expression for specific pixel component or all pixel components in case
5630 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5632 The expressions can use the following variables:
5636 The sequential number of the filtered frame, starting from @code{0}.
5640 the coordinates of the current sample
5644 the width and height of currently filtered plane
5648 Width and height scale for the plane being filtered. It is the
5649 ratio between the dimensions of the current plane to the luma plane,
5650 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
5651 the luma plane and @code{0.5,0.5} for the chroma planes.
5654 Time of the current frame, expressed in seconds.
5657 Value of pixel component at current location for first video frame (top layer).
5660 Value of pixel component at current location for second video frame (bottom layer).
5664 The @code{blend} filter also supports the @ref{framesync} options.
5666 @subsection Examples
5670 Apply transition from bottom layer to top layer in first 10 seconds:
5672 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5676 Apply linear horizontal transition from top layer to bottom layer:
5678 blend=all_expr='A*(X/W)+B*(1-X/W)'
5682 Apply 1x1 checkerboard effect:
5684 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5688 Apply uncover left effect:
5690 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5694 Apply uncover down effect:
5696 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5700 Apply uncover up-left effect:
5702 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5706 Split diagonally video and shows top and bottom layer on each side:
5708 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5712 Display differences between the current and the previous frame:
5714 tblend=all_mode=grainextract
5720 Apply a boxblur algorithm to the input video.
5722 It accepts the following parameters:
5726 @item luma_radius, lr
5727 @item luma_power, lp
5728 @item chroma_radius, cr
5729 @item chroma_power, cp
5730 @item alpha_radius, ar
5731 @item alpha_power, ap
5735 A description of the accepted options follows.
5738 @item luma_radius, lr
5739 @item chroma_radius, cr
5740 @item alpha_radius, ar
5741 Set an expression for the box radius in pixels used for blurring the
5742 corresponding input plane.
5744 The radius value must be a non-negative number, and must not be
5745 greater than the value of the expression @code{min(w,h)/2} for the
5746 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5749 Default value for @option{luma_radius} is "2". If not specified,
5750 @option{chroma_radius} and @option{alpha_radius} default to the
5751 corresponding value set for @option{luma_radius}.
5753 The expressions can contain the following constants:
5757 The input width and height in pixels.
5761 The input chroma image width and height in pixels.
5765 The horizontal and vertical chroma subsample values. For example, for the
5766 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5769 @item luma_power, lp
5770 @item chroma_power, cp
5771 @item alpha_power, ap
5772 Specify how many times the boxblur filter is applied to the
5773 corresponding plane.
5775 Default value for @option{luma_power} is 2. If not specified,
5776 @option{chroma_power} and @option{alpha_power} default to the
5777 corresponding value set for @option{luma_power}.
5779 A value of 0 will disable the effect.
5782 @subsection Examples
5786 Apply a boxblur filter with the luma, chroma, and alpha radii
5789 boxblur=luma_radius=2:luma_power=1
5794 Set the luma radius to 2, and alpha and chroma radius to 0:
5796 boxblur=2:1:cr=0:ar=0
5800 Set the luma and chroma radii to a fraction of the video dimension:
5802 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5808 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5809 Deinterlacing Filter").
5811 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5812 interpolation algorithms.
5813 It accepts the following parameters:
5817 The interlacing mode to adopt. It accepts one of the following values:
5821 Output one frame for each frame.
5823 Output one frame for each field.
5826 The default value is @code{send_field}.
5829 The picture field parity assumed for the input interlaced video. It accepts one
5830 of the following values:
5834 Assume the top field is first.
5836 Assume the bottom field is first.
5838 Enable automatic detection of field parity.
5841 The default value is @code{auto}.
5842 If the interlacing is unknown or the decoder does not export this information,
5843 top field first will be assumed.
5846 Specify which frames to deinterlace. Accept one of the following
5851 Deinterlace all frames.
5853 Only deinterlace frames marked as interlaced.
5856 The default value is @code{all}.
5860 YUV colorspace color/chroma keying.
5862 The filter accepts the following options:
5866 The color which will be replaced with transparency.
5869 Similarity percentage with the key color.
5871 0.01 matches only the exact key color, while 1.0 matches everything.
5876 0.0 makes pixels either fully transparent, or not transparent at all.
5878 Higher values result in semi-transparent pixels, with a higher transparency
5879 the more similar the pixels color is to the key color.
5882 Signals that the color passed is already in YUV instead of RGB.
5884 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5885 This can be used to pass exact YUV values as hexadecimal numbers.
5888 @subsection Examples
5892 Make every green pixel in the input image transparent:
5894 ffmpeg -i input.png -vf chromakey=green out.png
5898 Overlay a greenscreen-video on top of a static black background.
5900 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
5906 Display CIE color diagram with pixels overlaid onto it.
5908 The filter accepts the following options:
5923 @item uhdtv, rec2020
5936 Set what gamuts to draw.
5938 See @code{system} option for available values.
5941 Set ciescope size, by default set to 512.
5944 Set intensity used to map input pixel values to CIE diagram.
5947 Set contrast used to draw tongue colors that are out of active color system gamut.
5950 Correct gamma displayed on scope, by default enabled.
5953 Show white point on CIE diagram, by default disabled.
5956 Set input gamma. Used only with XYZ input color space.
5961 Visualize information exported by some codecs.
5963 Some codecs can export information through frames using side-data or other
5964 means. For example, some MPEG based codecs export motion vectors through the
5965 @var{export_mvs} flag in the codec @option{flags2} option.
5967 The filter accepts the following option:
5971 Set motion vectors to visualize.
5973 Available flags for @var{mv} are:
5977 forward predicted MVs of P-frames
5979 forward predicted MVs of B-frames
5981 backward predicted MVs of B-frames
5985 Display quantization parameters using the chroma planes.
5988 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5990 Available flags for @var{mv_type} are:
5994 forward predicted MVs
5996 backward predicted MVs
5999 @item frame_type, ft
6000 Set frame type to visualize motion vectors of.
6002 Available flags for @var{frame_type} are:
6006 intra-coded frames (I-frames)
6008 predicted frames (P-frames)
6010 bi-directionally predicted frames (B-frames)
6014 @subsection Examples
6018 Visualize forward predicted MVs of all frames using @command{ffplay}:
6020 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6024 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6026 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6030 @section colorbalance
6031 Modify intensity of primary colors (red, green and blue) of input frames.
6033 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6034 regions for the red-cyan, green-magenta or blue-yellow balance.
6036 A positive adjustment value shifts the balance towards the primary color, a negative
6037 value towards the complementary color.
6039 The filter accepts the following options:
6045 Adjust red, green and blue shadows (darkest pixels).
6050 Adjust red, green and blue midtones (medium pixels).
6055 Adjust red, green and blue highlights (brightest pixels).
6057 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6060 @subsection Examples
6064 Add red color cast to shadows:
6071 RGB colorspace color keying.
6073 The filter accepts the following options:
6077 The color which will be replaced with transparency.
6080 Similarity percentage with the key color.
6082 0.01 matches only the exact key color, while 1.0 matches everything.
6087 0.0 makes pixels either fully transparent, or not transparent at all.
6089 Higher values result in semi-transparent pixels, with a higher transparency
6090 the more similar the pixels color is to the key color.
6093 @subsection Examples
6097 Make every green pixel in the input image transparent:
6099 ffmpeg -i input.png -vf colorkey=green out.png
6103 Overlay a greenscreen-video on top of a static background image.
6105 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
6109 @section colorlevels
6111 Adjust video input frames using levels.
6113 The filter accepts the following options:
6120 Adjust red, green, blue and alpha input black point.
6121 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6127 Adjust red, green, blue and alpha input white point.
6128 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6130 Input levels are used to lighten highlights (bright tones), darken shadows
6131 (dark tones), change the balance of bright and dark tones.
6137 Adjust red, green, blue and alpha output black point.
6138 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6144 Adjust red, green, blue and alpha output white point.
6145 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6147 Output levels allows manual selection of a constrained output level range.
6150 @subsection Examples
6154 Make video output darker:
6156 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6162 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6166 Make video output lighter:
6168 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6172 Increase brightness:
6174 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6178 @section colorchannelmixer
6180 Adjust video input frames by re-mixing color channels.
6182 This filter modifies a color channel by adding the values associated to
6183 the other channels of the same pixels. For example if the value to
6184 modify is red, the output value will be:
6186 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6189 The filter accepts the following options:
6196 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6197 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6203 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6204 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6210 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6211 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6217 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6218 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6220 Allowed ranges for options are @code{[-2.0, 2.0]}.
6223 @subsection Examples
6227 Convert source to grayscale:
6229 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6232 Simulate sepia tones:
6234 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6238 @section colormatrix
6240 Convert color matrix.
6242 The filter accepts the following options:
6247 Specify the source and destination color matrix. Both values must be
6250 The accepted values are:
6278 For example to convert from BT.601 to SMPTE-240M, use the command:
6280 colormatrix=bt601:smpte240m
6285 Convert colorspace, transfer characteristics or color primaries.
6286 Input video needs to have an even size.
6288 The filter accepts the following options:
6293 Specify all color properties at once.
6295 The accepted values are:
6325 Specify output colorspace.
6327 The accepted values are:
6336 BT.470BG or BT.601-6 625
6339 SMPTE-170M or BT.601-6 525
6348 BT.2020 with non-constant luminance
6354 Specify output transfer characteristics.
6356 The accepted values are:
6368 Constant gamma of 2.2
6371 Constant gamma of 2.8
6374 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6392 BT.2020 for 10-bits content
6395 BT.2020 for 12-bits content
6401 Specify output color primaries.
6403 The accepted values are:
6412 BT.470BG or BT.601-6 625
6415 SMPTE-170M or BT.601-6 525
6439 Specify output color range.
6441 The accepted values are:
6444 TV (restricted) range
6447 MPEG (restricted) range
6458 Specify output color format.
6460 The accepted values are:
6463 YUV 4:2:0 planar 8-bits
6466 YUV 4:2:0 planar 10-bits
6469 YUV 4:2:0 planar 12-bits
6472 YUV 4:2:2 planar 8-bits
6475 YUV 4:2:2 planar 10-bits
6478 YUV 4:2:2 planar 12-bits
6481 YUV 4:4:4 planar 8-bits
6484 YUV 4:4:4 planar 10-bits
6487 YUV 4:4:4 planar 12-bits
6492 Do a fast conversion, which skips gamma/primary correction. This will take
6493 significantly less CPU, but will be mathematically incorrect. To get output
6494 compatible with that produced by the colormatrix filter, use fast=1.
6497 Specify dithering mode.
6499 The accepted values are:
6505 Floyd-Steinberg dithering
6509 Whitepoint adaptation mode.
6511 The accepted values are:
6514 Bradford whitepoint adaptation
6517 von Kries whitepoint adaptation
6520 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6524 Override all input properties at once. Same accepted values as @ref{all}.
6527 Override input colorspace. Same accepted values as @ref{space}.
6530 Override input color primaries. Same accepted values as @ref{primaries}.
6533 Override input transfer characteristics. Same accepted values as @ref{trc}.
6536 Override input color range. Same accepted values as @ref{range}.
6540 The filter converts the transfer characteristics, color space and color
6541 primaries to the specified user values. The output value, if not specified,
6542 is set to a default value based on the "all" property. If that property is
6543 also not specified, the filter will log an error. The output color range and
6544 format default to the same value as the input color range and format. The
6545 input transfer characteristics, color space, color primaries and color range
6546 should be set on the input data. If any of these are missing, the filter will
6547 log an error and no conversion will take place.
6549 For example to convert the input to SMPTE-240M, use the command:
6551 colorspace=smpte240m
6554 @section convolution
6556 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
6558 The filter accepts the following options:
6565 Set matrix for each plane.
6566 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
6567 and from 1 to 49 odd number of signed integers in @var{row} mode.
6573 Set multiplier for calculated value for each plane.
6574 If unset or 0, it will be sum of all matrix elements.
6580 Set bias for each plane. This value is added to the result of the multiplication.
6581 Useful for making the overall image brighter or darker. Default is 0.0.
6587 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
6588 Default is @var{square}.
6591 @subsection Examples
6597 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"
6603 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"
6609 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"
6615 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"
6619 Apply laplacian edge detector which includes diagonals:
6621 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"
6627 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"
6633 Apply 2D convolution of video stream in frequency domain using second stream
6636 The filter accepts the following options:
6640 Set which planes to process.
6643 Set which impulse video frames will be processed, can be @var{first}
6644 or @var{all}. Default is @var{all}.
6647 The @code{convolve} filter also supports the @ref{framesync} options.
6651 Copy the input video source unchanged to the output. This is mainly useful for
6656 Video filtering on GPU using Apple's CoreImage API on OSX.
6658 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6659 processed by video hardware. However, software-based OpenGL implementations
6660 exist which means there is no guarantee for hardware processing. It depends on
6663 There are many filters and image generators provided by Apple that come with a
6664 large variety of options. The filter has to be referenced by its name along
6667 The coreimage filter accepts the following options:
6670 List all available filters and generators along with all their respective
6671 options as well as possible minimum and maximum values along with the default
6678 Specify all filters by their respective name and options.
6679 Use @var{list_filters} to determine all valid filter names and options.
6680 Numerical options are specified by a float value and are automatically clamped
6681 to their respective value range. Vector and color options have to be specified
6682 by a list of space separated float values. Character escaping has to be done.
6683 A special option name @code{default} is available to use default options for a
6686 It is required to specify either @code{default} or at least one of the filter options.
6687 All omitted options are used with their default values.
6688 The syntax of the filter string is as follows:
6690 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6694 Specify a rectangle where the output of the filter chain is copied into the
6695 input image. It is given by a list of space separated float values:
6697 output_rect=x\ y\ width\ height
6699 If not given, the output rectangle equals the dimensions of the input image.
6700 The output rectangle is automatically cropped at the borders of the input
6701 image. Negative values are valid for each component.
6703 output_rect=25\ 25\ 100\ 100
6707 Several filters can be chained for successive processing without GPU-HOST
6708 transfers allowing for fast processing of complex filter chains.
6709 Currently, only filters with zero (generators) or exactly one (filters) input
6710 image and one output image are supported. Also, transition filters are not yet
6713 Some filters generate output images with additional padding depending on the
6714 respective filter kernel. The padding is automatically removed to ensure the
6715 filter output has the same size as the input image.
6717 For image generators, the size of the output image is determined by the
6718 previous output image of the filter chain or the input image of the whole
6719 filterchain, respectively. The generators do not use the pixel information of
6720 this image to generate their output. However, the generated output is
6721 blended onto this image, resulting in partial or complete coverage of the
6724 The @ref{coreimagesrc} video source can be used for generating input images
6725 which are directly fed into the filter chain. By using it, providing input
6726 images by another video source or an input video is not required.
6728 @subsection Examples
6733 List all filters available:
6735 coreimage=list_filters=true
6739 Use the CIBoxBlur filter with default options to blur an image:
6741 coreimage=filter=CIBoxBlur@@default
6745 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6746 its center at 100x100 and a radius of 50 pixels:
6748 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6752 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6753 given as complete and escaped command-line for Apple's standard bash shell:
6755 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6761 Crop the input video to given dimensions.
6763 It accepts the following parameters:
6767 The width of the output video. It defaults to @code{iw}.
6768 This expression is evaluated only once during the filter
6769 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6772 The height of the output video. It defaults to @code{ih}.
6773 This expression is evaluated only once during the filter
6774 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6777 The horizontal position, in the input video, of the left edge of the output
6778 video. It defaults to @code{(in_w-out_w)/2}.
6779 This expression is evaluated per-frame.
6782 The vertical position, in the input video, of the top edge of the output video.
6783 It defaults to @code{(in_h-out_h)/2}.
6784 This expression is evaluated per-frame.
6787 If set to 1 will force the output display aspect ratio
6788 to be the same of the input, by changing the output sample aspect
6789 ratio. It defaults to 0.
6792 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6793 width/height/x/y as specified and will not be rounded to nearest smaller value.
6797 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6798 expressions containing the following constants:
6803 The computed values for @var{x} and @var{y}. They are evaluated for
6808 The input width and height.
6812 These are the same as @var{in_w} and @var{in_h}.
6816 The output (cropped) width and height.
6820 These are the same as @var{out_w} and @var{out_h}.
6823 same as @var{iw} / @var{ih}
6826 input sample aspect ratio
6829 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6833 horizontal and vertical chroma subsample values. For example for the
6834 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6837 The number of the input frame, starting from 0.
6840 the position in the file of the input frame, NAN if unknown
6843 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6847 The expression for @var{out_w} may depend on the value of @var{out_h},
6848 and the expression for @var{out_h} may depend on @var{out_w}, but they
6849 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6850 evaluated after @var{out_w} and @var{out_h}.
6852 The @var{x} and @var{y} parameters specify the expressions for the
6853 position of the top-left corner of the output (non-cropped) area. They
6854 are evaluated for each frame. If the evaluated value is not valid, it
6855 is approximated to the nearest valid value.
6857 The expression for @var{x} may depend on @var{y}, and the expression
6858 for @var{y} may depend on @var{x}.
6860 @subsection Examples
6864 Crop area with size 100x100 at position (12,34).
6869 Using named options, the example above becomes:
6871 crop=w=100:h=100:x=12:y=34
6875 Crop the central input area with size 100x100:
6881 Crop the central input area with size 2/3 of the input video:
6883 crop=2/3*in_w:2/3*in_h
6887 Crop the input video central square:
6894 Delimit the rectangle with the top-left corner placed at position
6895 100:100 and the right-bottom corner corresponding to the right-bottom
6896 corner of the input image.
6898 crop=in_w-100:in_h-100:100:100
6902 Crop 10 pixels from the left and right borders, and 20 pixels from
6903 the top and bottom borders
6905 crop=in_w-2*10:in_h-2*20
6909 Keep only the bottom right quarter of the input image:
6911 crop=in_w/2:in_h/2:in_w/2:in_h/2
6915 Crop height for getting Greek harmony:
6917 crop=in_w:1/PHI*in_w
6921 Apply trembling effect:
6923 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)
6927 Apply erratic camera effect depending on timestamp:
6929 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)"
6933 Set x depending on the value of y:
6935 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6939 @subsection Commands
6941 This filter supports the following commands:
6947 Set width/height of the output video and the horizontal/vertical position
6949 The command accepts the same syntax of the corresponding option.
6951 If the specified expression is not valid, it is kept at its current
6957 Auto-detect the crop size.
6959 It calculates the necessary cropping parameters and prints the
6960 recommended parameters via the logging system. The detected dimensions
6961 correspond to the non-black area of the input video.
6963 It accepts the following parameters:
6968 Set higher black value threshold, which can be optionally specified
6969 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6970 value greater to the set value is considered non-black. It defaults to 24.
6971 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6972 on the bitdepth of the pixel format.
6975 The value which the width/height should be divisible by. It defaults to
6976 16. The offset is automatically adjusted to center the video. Use 2 to
6977 get only even dimensions (needed for 4:2:2 video). 16 is best when
6978 encoding to most video codecs.
6980 @item reset_count, reset
6981 Set the counter that determines after how many frames cropdetect will
6982 reset the previously detected largest video area and start over to
6983 detect the current optimal crop area. Default value is 0.
6985 This can be useful when channel logos distort the video area. 0
6986 indicates 'never reset', and returns the largest area encountered during
6993 Apply color adjustments using curves.
6995 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6996 component (red, green and blue) has its values defined by @var{N} key points
6997 tied from each other using a smooth curve. The x-axis represents the pixel
6998 values from the input frame, and the y-axis the new pixel values to be set for
7001 By default, a component curve is defined by the two points @var{(0;0)} and
7002 @var{(1;1)}. This creates a straight line where each original pixel value is
7003 "adjusted" to its own value, which means no change to the image.
7005 The filter allows you to redefine these two points and add some more. A new
7006 curve (using a natural cubic spline interpolation) will be define to pass
7007 smoothly through all these new coordinates. The new defined points needs to be
7008 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7009 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7010 the vector spaces, the values will be clipped accordingly.
7012 The filter accepts the following options:
7016 Select one of the available color presets. This option can be used in addition
7017 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7018 options takes priority on the preset values.
7019 Available presets are:
7022 @item color_negative
7025 @item increase_contrast
7027 @item linear_contrast
7028 @item medium_contrast
7030 @item strong_contrast
7033 Default is @code{none}.
7035 Set the master key points. These points will define a second pass mapping. It
7036 is sometimes called a "luminance" or "value" mapping. It can be used with
7037 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7038 post-processing LUT.
7040 Set the key points for the red component.
7042 Set the key points for the green component.
7044 Set the key points for the blue component.
7046 Set the key points for all components (not including master).
7047 Can be used in addition to the other key points component
7048 options. In this case, the unset component(s) will fallback on this
7049 @option{all} setting.
7051 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7053 Save Gnuplot script of the curves in specified file.
7056 To avoid some filtergraph syntax conflicts, each key points list need to be
7057 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7059 @subsection Examples
7063 Increase slightly the middle level of blue:
7065 curves=blue='0/0 0.5/0.58 1/1'
7071 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'
7073 Here we obtain the following coordinates for each components:
7076 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7078 @code{(0;0) (0.50;0.48) (1;1)}
7080 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7084 The previous example can also be achieved with the associated built-in preset:
7086 curves=preset=vintage
7096 Use a Photoshop preset and redefine the points of the green component:
7098 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7102 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7103 and @command{gnuplot}:
7105 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7106 gnuplot -p /tmp/curves.plt
7112 Video data analysis filter.
7114 This filter shows hexadecimal pixel values of part of video.
7116 The filter accepts the following options:
7120 Set output video size.
7123 Set x offset from where to pick pixels.
7126 Set y offset from where to pick pixels.
7129 Set scope mode, can be one of the following:
7132 Draw hexadecimal pixel values with white color on black background.
7135 Draw hexadecimal pixel values with input video pixel color on black
7139 Draw hexadecimal pixel values on color background picked from input video,
7140 the text color is picked in such way so its always visible.
7144 Draw rows and columns numbers on left and top of video.
7147 Set background opacity.
7152 Denoise frames using 2D DCT (frequency domain filtering).
7154 This filter is not designed for real time.
7156 The filter accepts the following options:
7160 Set the noise sigma constant.
7162 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7163 coefficient (absolute value) below this threshold with be dropped.
7165 If you need a more advanced filtering, see @option{expr}.
7167 Default is @code{0}.
7170 Set number overlapping pixels for each block. Since the filter can be slow, you
7171 may want to reduce this value, at the cost of a less effective filter and the
7172 risk of various artefacts.
7174 If the overlapping value doesn't permit processing the whole input width or
7175 height, a warning will be displayed and according borders won't be denoised.
7177 Default value is @var{blocksize}-1, which is the best possible setting.
7180 Set the coefficient factor expression.
7182 For each coefficient of a DCT block, this expression will be evaluated as a
7183 multiplier value for the coefficient.
7185 If this is option is set, the @option{sigma} option will be ignored.
7187 The absolute value of the coefficient can be accessed through the @var{c}
7191 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7192 @var{blocksize}, which is the width and height of the processed blocks.
7194 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7195 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7196 on the speed processing. Also, a larger block size does not necessarily means a
7200 @subsection Examples
7202 Apply a denoise with a @option{sigma} of @code{4.5}:
7207 The same operation can be achieved using the expression system:
7209 dctdnoiz=e='gte(c, 4.5*3)'
7212 Violent denoise using a block size of @code{16x16}:
7219 Remove banding artifacts from input video.
7220 It works by replacing banded pixels with average value of referenced pixels.
7222 The filter accepts the following options:
7229 Set banding detection threshold for each plane. Default is 0.02.
7230 Valid range is 0.00003 to 0.5.
7231 If difference between current pixel and reference pixel is less than threshold,
7232 it will be considered as banded.
7235 Banding detection range in pixels. Default is 16. If positive, random number
7236 in range 0 to set value will be used. If negative, exact absolute value
7238 The range defines square of four pixels around current pixel.
7241 Set direction in radians from which four pixel will be compared. If positive,
7242 random direction from 0 to set direction will be picked. If negative, exact of
7243 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7244 will pick only pixels on same row and -PI/2 will pick only pixels on same
7248 If enabled, current pixel is compared with average value of all four
7249 surrounding pixels. The default is enabled. If disabled current pixel is
7250 compared with all four surrounding pixels. The pixel is considered banded
7251 if only all four differences with surrounding pixels are less than threshold.
7254 If enabled, current pixel is changed if and only if all pixel components are banded,
7255 e.g. banding detection threshold is triggered for all color components.
7256 The default is disabled.
7261 Remove blocking artifacts from input video.
7263 The filter accepts the following options:
7267 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7268 This controls what kind of deblocking is applied.
7271 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7277 Set blocking detection thresholds. Allowed range is 0 to 1.
7278 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7279 Using higher threshold gives more deblocking strength.
7280 Setting @var{alpha} controls threshold detection at exact edge of block.
7281 Remaining options controls threshold detection near the edge. Each one for
7282 below/above or left/right. Setting any of those to @var{0} disables
7286 Set planes to filter. Default is to filter all available planes.
7289 @subsection Examples
7293 Deblock using weak filter and block size of 4 pixels.
7295 deblock=filter=weak:block=4
7299 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7300 deblocking more edges.
7302 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7306 Similar as above, but filter only first plane.
7308 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7312 Similar as above, but filter only second and third plane.
7314 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7321 Drop duplicated frames at regular intervals.
7323 The filter accepts the following options:
7327 Set the number of frames from which one will be dropped. Setting this to
7328 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7329 Default is @code{5}.
7332 Set the threshold for duplicate detection. If the difference metric for a frame
7333 is less than or equal to this value, then it is declared as duplicate. Default
7337 Set scene change threshold. Default is @code{15}.
7341 Set the size of the x and y-axis blocks used during metric calculations.
7342 Larger blocks give better noise suppression, but also give worse detection of
7343 small movements. Must be a power of two. Default is @code{32}.
7346 Mark main input as a pre-processed input and activate clean source input
7347 stream. This allows the input to be pre-processed with various filters to help
7348 the metrics calculation while keeping the frame selection lossless. When set to
7349 @code{1}, the first stream is for the pre-processed input, and the second
7350 stream is the clean source from where the kept frames are chosen. Default is
7354 Set whether or not chroma is considered in the metric calculations. Default is
7360 Apply 2D deconvolution of video stream in frequency domain using second stream
7363 The filter accepts the following options:
7367 Set which planes to process.
7370 Set which impulse video frames will be processed, can be @var{first}
7371 or @var{all}. Default is @var{all}.
7374 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7375 and height are not same and not power of 2 or if stream prior to convolving
7379 The @code{deconvolve} filter also supports the @ref{framesync} options.
7383 Apply deflate effect to the video.
7385 This filter replaces the pixel by the local(3x3) average by taking into account
7386 only values lower than the pixel.
7388 It accepts the following options:
7395 Limit the maximum change for each plane, default is 65535.
7396 If 0, plane will remain unchanged.
7401 Remove temporal frame luminance variations.
7403 It accepts the following options:
7407 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7410 Set averaging mode to smooth temporal luminance variations.
7412 Available values are:
7437 Do not actually modify frame. Useful when one only wants metadata.
7442 Remove judder produced by partially interlaced telecined content.
7444 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7445 source was partially telecined content then the output of @code{pullup,dejudder}
7446 will have a variable frame rate. May change the recorded frame rate of the
7447 container. Aside from that change, this filter will not affect constant frame
7450 The option available in this filter is:
7454 Specify the length of the window over which the judder repeats.
7456 Accepts any integer greater than 1. Useful values are:
7460 If the original was telecined from 24 to 30 fps (Film to NTSC).
7463 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7466 If a mixture of the two.
7469 The default is @samp{4}.
7474 Suppress a TV station logo by a simple interpolation of the surrounding
7475 pixels. Just set a rectangle covering the logo and watch it disappear
7476 (and sometimes something even uglier appear - your mileage may vary).
7478 It accepts the following parameters:
7483 Specify the top left corner coordinates of the logo. They must be
7488 Specify the width and height of the logo to clear. They must be
7492 Specify the thickness of the fuzzy edge of the rectangle (added to
7493 @var{w} and @var{h}). The default value is 1. This option is
7494 deprecated, setting higher values should no longer be necessary and
7498 When set to 1, a green rectangle is drawn on the screen to simplify
7499 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7500 The default value is 0.
7502 The rectangle is drawn on the outermost pixels which will be (partly)
7503 replaced with interpolated values. The values of the next pixels
7504 immediately outside this rectangle in each direction will be used to
7505 compute the interpolated pixel values inside the rectangle.
7509 @subsection Examples
7513 Set a rectangle covering the area with top left corner coordinates 0,0
7514 and size 100x77, and a band of size 10:
7516 delogo=x=0:y=0:w=100:h=77:band=10
7523 Attempt to fix small changes in horizontal and/or vertical shift. This
7524 filter helps remove camera shake from hand-holding a camera, bumping a
7525 tripod, moving on a vehicle, etc.
7527 The filter accepts the following options:
7535 Specify a rectangular area where to limit the search for motion
7537 If desired the search for motion vectors can be limited to a
7538 rectangular area of the frame defined by its top left corner, width
7539 and height. These parameters have the same meaning as the drawbox
7540 filter which can be used to visualise the position of the bounding
7543 This is useful when simultaneous movement of subjects within the frame
7544 might be confused for camera motion by the motion vector search.
7546 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7547 then the full frame is used. This allows later options to be set
7548 without specifying the bounding box for the motion vector search.
7550 Default - search the whole frame.
7554 Specify the maximum extent of movement in x and y directions in the
7555 range 0-64 pixels. Default 16.
7558 Specify how to generate pixels to fill blanks at the edge of the
7559 frame. Available values are:
7562 Fill zeroes at blank locations
7564 Original image at blank locations
7566 Extruded edge value at blank locations
7568 Mirrored edge at blank locations
7570 Default value is @samp{mirror}.
7573 Specify the blocksize to use for motion search. Range 4-128 pixels,
7577 Specify the contrast threshold for blocks. Only blocks with more than
7578 the specified contrast (difference between darkest and lightest
7579 pixels) will be considered. Range 1-255, default 125.
7582 Specify the search strategy. Available values are:
7585 Set exhaustive search
7587 Set less exhaustive search.
7589 Default value is @samp{exhaustive}.
7592 If set then a detailed log of the motion search is written to the
7599 Remove unwanted contamination of foreground colors, caused by reflected color of
7600 greenscreen or bluescreen.
7602 This filter accepts the following options:
7606 Set what type of despill to use.
7609 Set how spillmap will be generated.
7612 Set how much to get rid of still remaining spill.
7615 Controls amount of red in spill area.
7618 Controls amount of green in spill area.
7619 Should be -1 for greenscreen.
7622 Controls amount of blue in spill area.
7623 Should be -1 for bluescreen.
7626 Controls brightness of spill area, preserving colors.
7629 Modify alpha from generated spillmap.
7634 Apply an exact inverse of the telecine operation. It requires a predefined
7635 pattern specified using the pattern option which must be the same as that passed
7636 to the telecine filter.
7638 This filter accepts the following options:
7647 The default value is @code{top}.
7651 A string of numbers representing the pulldown pattern you wish to apply.
7652 The default value is @code{23}.
7655 A number representing position of the first frame with respect to the telecine
7656 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7661 Apply dilation effect to the video.
7663 This filter replaces the pixel by the local(3x3) maximum.
7665 It accepts the following options:
7672 Limit the maximum change for each plane, default is 65535.
7673 If 0, plane will remain unchanged.
7676 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7679 Flags to local 3x3 coordinates maps like this:
7688 Displace pixels as indicated by second and third input stream.
7690 It takes three input streams and outputs one stream, the first input is the
7691 source, and second and third input are displacement maps.
7693 The second input specifies how much to displace pixels along the
7694 x-axis, while the third input specifies how much to displace pixels
7696 If one of displacement map streams terminates, last frame from that
7697 displacement map will be used.
7699 Note that once generated, displacements maps can be reused over and over again.
7701 A description of the accepted options follows.
7705 Set displace behavior for pixels that are out of range.
7707 Available values are:
7710 Missing pixels are replaced by black pixels.
7713 Adjacent pixels will spread out to replace missing pixels.
7716 Out of range pixels are wrapped so they point to pixels of other side.
7719 Out of range pixels will be replaced with mirrored pixels.
7721 Default is @samp{smear}.
7725 @subsection Examples
7729 Add ripple effect to rgb input of video size hd720:
7731 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
7735 Add wave effect to rgb input of video size hd720:
7737 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
7743 Draw a colored box on the input image.
7745 It accepts the following parameters:
7750 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7754 The expressions which specify the width and height of the box; if 0 they are interpreted as
7755 the input width and height. It defaults to 0.
7758 Specify the color of the box to write. For the general syntax of this option,
7759 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7760 value @code{invert} is used, the box edge color is the same as the
7761 video with inverted luma.
7764 The expression which sets the thickness of the box edge.
7765 A value of @code{fill} will create a filled box. Default value is @code{3}.
7767 See below for the list of accepted constants.
7770 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7771 will overwrite the video's color and alpha pixels.
7772 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7775 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7776 following constants:
7780 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7784 horizontal and vertical chroma subsample values. For example for the
7785 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7789 The input width and height.
7792 The input sample aspect ratio.
7796 The x and y offset coordinates where the box is drawn.
7800 The width and height of the drawn box.
7803 The thickness of the drawn box.
7805 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7806 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7810 @subsection Examples
7814 Draw a black box around the edge of the input image:
7820 Draw a box with color red and an opacity of 50%:
7822 drawbox=10:20:200:60:red@@0.5
7825 The previous example can be specified as:
7827 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7831 Fill the box with pink color:
7833 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7837 Draw a 2-pixel red 2.40:1 mask:
7839 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
7845 Draw a grid on the input image.
7847 It accepts the following parameters:
7852 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7856 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7857 input width and height, respectively, minus @code{thickness}, so image gets
7858 framed. Default to 0.
7861 Specify the color of the grid. For the general syntax of this option,
7862 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7863 value @code{invert} is used, the grid color is the same as the
7864 video with inverted luma.
7867 The expression which sets the thickness of the grid line. Default value is @code{1}.
7869 See below for the list of accepted constants.
7872 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7873 will overwrite the video's color and alpha pixels.
7874 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7877 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7878 following constants:
7882 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7886 horizontal and vertical chroma subsample values. For example for the
7887 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7891 The input grid cell width and height.
7894 The input sample aspect ratio.
7898 The x and y coordinates of some point of grid intersection (meant to configure offset).
7902 The width and height of the drawn cell.
7905 The thickness of the drawn cell.
7907 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7908 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7912 @subsection Examples
7916 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7918 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7922 Draw a white 3x3 grid with an opacity of 50%:
7924 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7931 Draw a text string or text from a specified file on top of a video, using the
7932 libfreetype library.
7934 To enable compilation of this filter, you need to configure FFmpeg with
7935 @code{--enable-libfreetype}.
7936 To enable default font fallback and the @var{font} option you need to
7937 configure FFmpeg with @code{--enable-libfontconfig}.
7938 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7939 @code{--enable-libfribidi}.
7943 It accepts the following parameters:
7948 Used to draw a box around text using the background color.
7949 The value must be either 1 (enable) or 0 (disable).
7950 The default value of @var{box} is 0.
7953 Set the width of the border to be drawn around the box using @var{boxcolor}.
7954 The default value of @var{boxborderw} is 0.
7957 The color to be used for drawing box around text. For the syntax of this
7958 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7960 The default value of @var{boxcolor} is "white".
7963 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7964 The default value of @var{line_spacing} is 0.
7967 Set the width of the border to be drawn around the text using @var{bordercolor}.
7968 The default value of @var{borderw} is 0.
7971 Set the color to be used for drawing border around text. For the syntax of this
7972 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7974 The default value of @var{bordercolor} is "black".
7977 Select how the @var{text} is expanded. Can be either @code{none},
7978 @code{strftime} (deprecated) or
7979 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7983 Set a start time for the count. Value is in microseconds. Only applied
7984 in the deprecated strftime expansion mode. To emulate in normal expansion
7985 mode use the @code{pts} function, supplying the start time (in seconds)
7986 as the second argument.
7989 If true, check and fix text coords to avoid clipping.
7992 The color to be used for drawing fonts. For the syntax of this option, check
7993 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7995 The default value of @var{fontcolor} is "black".
7997 @item fontcolor_expr
7998 String which is expanded the same way as @var{text} to obtain dynamic
7999 @var{fontcolor} value. By default this option has empty value and is not
8000 processed. When this option is set, it overrides @var{fontcolor} option.
8003 The font family to be used for drawing text. By default Sans.
8006 The font file to be used for drawing text. The path must be included.
8007 This parameter is mandatory if the fontconfig support is disabled.
8010 Draw the text applying alpha blending. The value can
8011 be a number between 0.0 and 1.0.
8012 The expression accepts the same variables @var{x, y} as well.
8013 The default value is 1.
8014 Please see @var{fontcolor_expr}.
8017 The font size to be used for drawing text.
8018 The default value of @var{fontsize} is 16.
8021 If set to 1, attempt to shape the text (for example, reverse the order of
8022 right-to-left text and join Arabic characters) before drawing it.
8023 Otherwise, just draw the text exactly as given.
8024 By default 1 (if supported).
8027 The flags to be used for loading the fonts.
8029 The flags map the corresponding flags supported by libfreetype, and are
8030 a combination of the following values:
8037 @item vertical_layout
8038 @item force_autohint
8041 @item ignore_global_advance_width
8043 @item ignore_transform
8049 Default value is "default".
8051 For more information consult the documentation for the FT_LOAD_*
8055 The color to be used for drawing a shadow behind the drawn text. For the
8056 syntax of this option, check the @ref{color syntax,,"Color" section in the
8057 ffmpeg-utils manual,ffmpeg-utils}.
8059 The default value of @var{shadowcolor} is "black".
8063 The x and y offsets for the text shadow position with respect to the
8064 position of the text. They can be either positive or negative
8065 values. The default value for both is "0".
8068 The starting frame number for the n/frame_num variable. The default value
8072 The size in number of spaces to use for rendering the tab.
8076 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8077 format. It can be used with or without text parameter. @var{timecode_rate}
8078 option must be specified.
8080 @item timecode_rate, rate, r
8081 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8082 integer. Minimum value is "1".
8083 Drop-frame timecode is supported for frame rates 30 & 60.
8086 If set to 1, the output of the timecode option will wrap around at 24 hours.
8087 Default is 0 (disabled).
8090 The text string to be drawn. The text must be a sequence of UTF-8
8092 This parameter is mandatory if no file is specified with the parameter
8096 A text file containing text to be drawn. The text must be a sequence
8097 of UTF-8 encoded characters.
8099 This parameter is mandatory if no text string is specified with the
8100 parameter @var{text}.
8102 If both @var{text} and @var{textfile} are specified, an error is thrown.
8105 If set to 1, the @var{textfile} will be reloaded before each frame.
8106 Be sure to update it atomically, or it may be read partially, or even fail.
8110 The expressions which specify the offsets where text will be drawn
8111 within the video frame. They are relative to the top/left border of the
8114 The default value of @var{x} and @var{y} is "0".
8116 See below for the list of accepted constants and functions.
8119 The parameters for @var{x} and @var{y} are expressions containing the
8120 following constants and functions:
8124 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8128 horizontal and vertical chroma subsample values. For example for the
8129 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8132 the height of each text line
8140 @item max_glyph_a, ascent
8141 the maximum distance from the baseline to the highest/upper grid
8142 coordinate used to place a glyph outline point, for all the rendered
8144 It is a positive value, due to the grid's orientation with the Y axis
8147 @item max_glyph_d, descent
8148 the maximum distance from the baseline to the lowest grid coordinate
8149 used to place a glyph outline point, for all the rendered glyphs.
8150 This is a negative value, due to the grid's orientation, with the Y axis
8154 maximum glyph height, that is the maximum height for all the glyphs
8155 contained in the rendered text, it is equivalent to @var{ascent} -
8159 maximum glyph width, that is the maximum width for all the glyphs
8160 contained in the rendered text
8163 the number of input frame, starting from 0
8165 @item rand(min, max)
8166 return a random number included between @var{min} and @var{max}
8169 The input sample aspect ratio.
8172 timestamp expressed in seconds, NAN if the input timestamp is unknown
8175 the height of the rendered text
8178 the width of the rendered text
8182 the x and y offset coordinates where the text is drawn.
8184 These parameters allow the @var{x} and @var{y} expressions to refer
8185 each other, so you can for example specify @code{y=x/dar}.
8188 @anchor{drawtext_expansion}
8189 @subsection Text expansion
8191 If @option{expansion} is set to @code{strftime},
8192 the filter recognizes strftime() sequences in the provided text and
8193 expands them accordingly. Check the documentation of strftime(). This
8194 feature is deprecated.
8196 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8198 If @option{expansion} is set to @code{normal} (which is the default),
8199 the following expansion mechanism is used.
8201 The backslash character @samp{\}, followed by any character, always expands to
8202 the second character.
8204 Sequences of the form @code{%@{...@}} are expanded. The text between the
8205 braces is a function name, possibly followed by arguments separated by ':'.
8206 If the arguments contain special characters or delimiters (':' or '@}'),
8207 they should be escaped.
8209 Note that they probably must also be escaped as the value for the
8210 @option{text} option in the filter argument string and as the filter
8211 argument in the filtergraph description, and possibly also for the shell,
8212 that makes up to four levels of escaping; using a text file avoids these
8215 The following functions are available:
8220 The expression evaluation result.
8222 It must take one argument specifying the expression to be evaluated,
8223 which accepts the same constants and functions as the @var{x} and
8224 @var{y} values. Note that not all constants should be used, for
8225 example the text size is not known when evaluating the expression, so
8226 the constants @var{text_w} and @var{text_h} will have an undefined
8229 @item expr_int_format, eif
8230 Evaluate the expression's value and output as formatted integer.
8232 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8233 The second argument specifies the output format. Allowed values are @samp{x},
8234 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8235 @code{printf} function.
8236 The third parameter is optional and sets the number of positions taken by the output.
8237 It can be used to add padding with zeros from the left.
8240 The time at which the filter is running, expressed in UTC.
8241 It can accept an argument: a strftime() format string.
8244 The time at which the filter is running, expressed in the local time zone.
8245 It can accept an argument: a strftime() format string.
8248 Frame metadata. Takes one or two arguments.
8250 The first argument is mandatory and specifies the metadata key.
8252 The second argument is optional and specifies a default value, used when the
8253 metadata key is not found or empty.
8256 The frame number, starting from 0.
8259 A 1 character description of the current picture type.
8262 The timestamp of the current frame.
8263 It can take up to three arguments.
8265 The first argument is the format of the timestamp; it defaults to @code{flt}
8266 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8267 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8268 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8269 @code{localtime} stands for the timestamp of the frame formatted as
8270 local time zone time.
8272 The second argument is an offset added to the timestamp.
8274 If the format is set to @code{hms}, a third argument @code{24HH} may be
8275 supplied to present the hour part of the formatted timestamp in 24h format
8278 If the format is set to @code{localtime} or @code{gmtime},
8279 a third argument may be supplied: a strftime() format string.
8280 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8283 @subsection Examples
8287 Draw "Test Text" with font FreeSerif, using the default values for the
8288 optional parameters.
8291 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8295 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8296 and y=50 (counting from the top-left corner of the screen), text is
8297 yellow with a red box around it. Both the text and the box have an
8301 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8302 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8305 Note that the double quotes are not necessary if spaces are not used
8306 within the parameter list.
8309 Show the text at the center of the video frame:
8311 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8315 Show the text at a random position, switching to a new position every 30 seconds:
8317 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)"
8321 Show a text line sliding from right to left in the last row of the video
8322 frame. The file @file{LONG_LINE} is assumed to contain a single line
8325 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8329 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8331 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8335 Draw a single green letter "g", at the center of the input video.
8336 The glyph baseline is placed at half screen height.
8338 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8342 Show text for 1 second every 3 seconds:
8344 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8348 Use fontconfig to set the font. Note that the colons need to be escaped.
8350 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8354 Print the date of a real-time encoding (see strftime(3)):
8356 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8360 Show text fading in and out (appearing/disappearing):
8363 DS=1.0 # display start
8364 DE=10.0 # display end
8365 FID=1.5 # fade in duration
8366 FOD=5 # fade out duration
8367 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 @}"
8371 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8372 and the @option{fontsize} value are included in the @option{y} offset.
8374 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8375 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8380 For more information about libfreetype, check:
8381 @url{http://www.freetype.org/}.
8383 For more information about fontconfig, check:
8384 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8386 For more information about libfribidi, check:
8387 @url{http://fribidi.org/}.
8391 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8393 The filter accepts the following options:
8398 Set low and high threshold values used by the Canny thresholding
8401 The high threshold selects the "strong" edge pixels, which are then
8402 connected through 8-connectivity with the "weak" edge pixels selected
8403 by the low threshold.
8405 @var{low} and @var{high} threshold values must be chosen in the range
8406 [0,1], and @var{low} should be lesser or equal to @var{high}.
8408 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8412 Define the drawing mode.
8416 Draw white/gray wires on black background.
8419 Mix the colors to create a paint/cartoon effect.
8422 Apply Canny edge detector on all selected planes.
8424 Default value is @var{wires}.
8427 Select planes for filtering. By default all available planes are filtered.
8430 @subsection Examples
8434 Standard edge detection with custom values for the hysteresis thresholding:
8436 edgedetect=low=0.1:high=0.4
8440 Painting effect without thresholding:
8442 edgedetect=mode=colormix:high=0
8447 Set brightness, contrast, saturation and approximate gamma adjustment.
8449 The filter accepts the following options:
8453 Set the contrast expression. The value must be a float value in range
8454 @code{-2.0} to @code{2.0}. The default value is "1".
8457 Set the brightness expression. The value must be a float value in
8458 range @code{-1.0} to @code{1.0}. The default value is "0".
8461 Set the saturation expression. The value must be a float in
8462 range @code{0.0} to @code{3.0}. The default value is "1".
8465 Set the gamma expression. The value must be a float in range
8466 @code{0.1} to @code{10.0}. The default value is "1".
8469 Set the gamma expression for red. The value must be a float in
8470 range @code{0.1} to @code{10.0}. The default value is "1".
8473 Set the gamma expression for green. The value must be a float in range
8474 @code{0.1} to @code{10.0}. The default value is "1".
8477 Set the gamma expression for blue. The value must be a float in range
8478 @code{0.1} to @code{10.0}. The default value is "1".
8481 Set the gamma weight expression. It can be used to reduce the effect
8482 of a high gamma value on bright image areas, e.g. keep them from
8483 getting overamplified and just plain white. The value must be a float
8484 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8485 gamma correction all the way down while @code{1.0} leaves it at its
8486 full strength. Default is "1".
8489 Set when the expressions for brightness, contrast, saturation and
8490 gamma expressions are evaluated.
8492 It accepts the following values:
8495 only evaluate expressions once during the filter initialization or
8496 when a command is processed
8499 evaluate expressions for each incoming frame
8502 Default value is @samp{init}.
8505 The expressions accept the following parameters:
8508 frame count of the input frame starting from 0
8511 byte position of the corresponding packet in the input file, NAN if
8515 frame rate of the input video, NAN if the input frame rate is unknown
8518 timestamp expressed in seconds, NAN if the input timestamp is unknown
8521 @subsection Commands
8522 The filter supports the following commands:
8526 Set the contrast expression.
8529 Set the brightness expression.
8532 Set the saturation expression.
8535 Set the gamma expression.
8538 Set the gamma_r expression.
8541 Set gamma_g expression.
8544 Set gamma_b expression.
8547 Set gamma_weight expression.
8549 The command accepts the same syntax of the corresponding option.
8551 If the specified expression is not valid, it is kept at its current
8558 Apply erosion effect to the video.
8560 This filter replaces the pixel by the local(3x3) minimum.
8562 It accepts the following options:
8569 Limit the maximum change for each plane, default is 65535.
8570 If 0, plane will remain unchanged.
8573 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8576 Flags to local 3x3 coordinates maps like this:
8583 @section extractplanes
8585 Extract color channel components from input video stream into
8586 separate grayscale video streams.
8588 The filter accepts the following option:
8592 Set plane(s) to extract.
8594 Available values for planes are:
8605 Choosing planes not available in the input will result in an error.
8606 That means you cannot select @code{r}, @code{g}, @code{b} planes
8607 with @code{y}, @code{u}, @code{v} planes at same time.
8610 @subsection Examples
8614 Extract luma, u and v color channel component from input video frame
8615 into 3 grayscale outputs:
8617 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
8623 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8625 For each input image, the filter will compute the optimal mapping from
8626 the input to the output given the codebook length, that is the number
8627 of distinct output colors.
8629 This filter accepts the following options.
8632 @item codebook_length, l
8633 Set codebook length. The value must be a positive integer, and
8634 represents the number of distinct output colors. Default value is 256.
8637 Set the maximum number of iterations to apply for computing the optimal
8638 mapping. The higher the value the better the result and the higher the
8639 computation time. Default value is 1.
8642 Set a random seed, must be an integer included between 0 and
8643 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8644 will try to use a good random seed on a best effort basis.
8647 Set pal8 output pixel format. This option does not work with codebook
8648 length greater than 256.
8653 Measure graylevel entropy in histogram of color channels of video frames.
8655 It accepts the following parameters:
8659 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8661 @var{diff} mode measures entropy of histogram delta values, absolute differences
8662 between neighbour histogram values.
8667 Apply a fade-in/out effect to the input video.
8669 It accepts the following parameters:
8673 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8675 Default is @code{in}.
8677 @item start_frame, s
8678 Specify the number of the frame to start applying the fade
8679 effect at. Default is 0.
8682 The number of frames that the fade effect lasts. At the end of the
8683 fade-in effect, the output video will have the same intensity as the input video.
8684 At the end of the fade-out transition, the output video will be filled with the
8685 selected @option{color}.
8689 If set to 1, fade only alpha channel, if one exists on the input.
8692 @item start_time, st
8693 Specify the timestamp (in seconds) of the frame to start to apply the fade
8694 effect. If both start_frame and start_time are specified, the fade will start at
8695 whichever comes last. Default is 0.
8698 The number of seconds for which the fade effect has to last. At the end of the
8699 fade-in effect the output video will have the same intensity as the input video,
8700 at the end of the fade-out transition the output video will be filled with the
8701 selected @option{color}.
8702 If both duration and nb_frames are specified, duration is used. Default is 0
8703 (nb_frames is used by default).
8706 Specify the color of the fade. Default is "black".
8709 @subsection Examples
8713 Fade in the first 30 frames of video:
8718 The command above is equivalent to:
8724 Fade out the last 45 frames of a 200-frame video:
8727 fade=type=out:start_frame=155:nb_frames=45
8731 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8733 fade=in:0:25, fade=out:975:25
8737 Make the first 5 frames yellow, then fade in from frame 5-24:
8739 fade=in:5:20:color=yellow
8743 Fade in alpha over first 25 frames of video:
8745 fade=in:0:25:alpha=1
8749 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8751 fade=t=in:st=5.5:d=0.5
8757 Apply arbitrary expressions to samples in frequency domain
8761 Adjust the dc value (gain) of the luma plane of the image. The filter
8762 accepts an integer value in range @code{0} to @code{1000}. The default
8763 value is set to @code{0}.
8766 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8767 filter accepts an integer value in range @code{0} to @code{1000}. The
8768 default value is set to @code{0}.
8771 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8772 filter accepts an integer value in range @code{0} to @code{1000}. The
8773 default value is set to @code{0}.
8776 Set the frequency domain weight expression for the luma plane.
8779 Set the frequency domain weight expression for the 1st chroma plane.
8782 Set the frequency domain weight expression for the 2nd chroma plane.
8785 Set when the expressions are evaluated.
8787 It accepts the following values:
8790 Only evaluate expressions once during the filter initialization.
8793 Evaluate expressions for each incoming frame.
8796 Default value is @samp{init}.
8798 The filter accepts the following variables:
8801 The coordinates of the current sample.
8805 The width and height of the image.
8808 The number of input frame, starting from 0.
8811 @subsection Examples
8817 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8823 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8829 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8835 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8841 Denoise frames using 3D FFT (frequency domain filtering).
8843 The filter accepts the following options:
8847 Set the noise sigma constant. This sets denoising strength.
8848 Default value is 1. Allowed range is from 0 to 30.
8849 Using very high sigma with low overlap may give blocking artifacts.
8852 Set amount of denoising. By default all detected noise is reduced.
8853 Default value is 1. Allowed range is from 0 to 1.
8856 Set size of block, Default is 4, can be 3, 4, 5 or 6.
8857 Actual size of block in pixels is 2 to power of @var{block}, so by default
8858 block size in pixels is 2^4 which is 16.
8861 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
8864 Set number of previous frames to use for denoising. By default is set to 0.
8867 Set number of next frames to to use for denoising. By default is set to 0.
8870 Set planes which will be filtered, by default are all available filtered
8876 Extract a single field from an interlaced image using stride
8877 arithmetic to avoid wasting CPU time. The output frames are marked as
8880 The filter accepts the following options:
8884 Specify whether to extract the top (if the value is @code{0} or
8885 @code{top}) or the bottom field (if the value is @code{1} or
8891 Create new frames by copying the top and bottom fields from surrounding frames
8892 supplied as numbers by the hint file.
8896 Set file containing hints: absolute/relative frame numbers.
8898 There must be one line for each frame in a clip. Each line must contain two
8899 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8900 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8901 is current frame number for @code{absolute} mode or out of [-1, 1] range
8902 for @code{relative} mode. First number tells from which frame to pick up top
8903 field and second number tells from which frame to pick up bottom field.
8905 If optionally followed by @code{+} output frame will be marked as interlaced,
8906 else if followed by @code{-} output frame will be marked as progressive, else
8907 it will be marked same as input frame.
8908 If line starts with @code{#} or @code{;} that line is skipped.
8911 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8914 Example of first several lines of @code{hint} file for @code{relative} mode:
8917 1,0 - # second frame, use third's frame top field and second's frame bottom field
8918 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8935 Field matching filter for inverse telecine. It is meant to reconstruct the
8936 progressive frames from a telecined stream. The filter does not drop duplicated
8937 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8938 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8940 The separation of the field matching and the decimation is notably motivated by
8941 the possibility of inserting a de-interlacing filter fallback between the two.
8942 If the source has mixed telecined and real interlaced content,
8943 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8944 But these remaining combed frames will be marked as interlaced, and thus can be
8945 de-interlaced by a later filter such as @ref{yadif} before decimation.
8947 In addition to the various configuration options, @code{fieldmatch} can take an
8948 optional second stream, activated through the @option{ppsrc} option. If
8949 enabled, the frames reconstruction will be based on the fields and frames from
8950 this second stream. This allows the first input to be pre-processed in order to
8951 help the various algorithms of the filter, while keeping the output lossless
8952 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8953 or brightness/contrast adjustments can help.
8955 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8956 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8957 which @code{fieldmatch} is based on. While the semantic and usage are very
8958 close, some behaviour and options names can differ.
8960 The @ref{decimate} filter currently only works for constant frame rate input.
8961 If your input has mixed telecined (30fps) and progressive content with a lower
8962 framerate like 24fps use the following filterchain to produce the necessary cfr
8963 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8965 The filter accepts the following options:
8969 Specify the assumed field order of the input stream. Available values are:
8973 Auto detect parity (use FFmpeg's internal parity value).
8975 Assume bottom field first.
8977 Assume top field first.
8980 Note that it is sometimes recommended not to trust the parity announced by the
8983 Default value is @var{auto}.
8986 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8987 sense that it won't risk creating jerkiness due to duplicate frames when
8988 possible, but if there are bad edits or blended fields it will end up
8989 outputting combed frames when a good match might actually exist. On the other
8990 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8991 but will almost always find a good frame if there is one. The other values are
8992 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8993 jerkiness and creating duplicate frames versus finding good matches in sections
8994 with bad edits, orphaned fields, blended fields, etc.
8996 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8998 Available values are:
9002 2-way matching (p/c)
9004 2-way matching, and trying 3rd match if still combed (p/c + n)
9006 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9008 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9009 still combed (p/c + n + u/b)
9011 3-way matching (p/c/n)
9013 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9014 detected as combed (p/c/n + u/b)
9017 The parenthesis at the end indicate the matches that would be used for that
9018 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9021 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9024 Default value is @var{pc_n}.
9027 Mark the main input stream as a pre-processed input, and enable the secondary
9028 input stream as the clean source to pick the fields from. See the filter
9029 introduction for more details. It is similar to the @option{clip2} feature from
9032 Default value is @code{0} (disabled).
9035 Set the field to match from. It is recommended to set this to the same value as
9036 @option{order} unless you experience matching failures with that setting. In
9037 certain circumstances changing the field that is used to match from can have a
9038 large impact on matching performance. Available values are:
9042 Automatic (same value as @option{order}).
9044 Match from the bottom field.
9046 Match from the top field.
9049 Default value is @var{auto}.
9052 Set whether or not chroma is included during the match comparisons. In most
9053 cases it is recommended to leave this enabled. You should set this to @code{0}
9054 only if your clip has bad chroma problems such as heavy rainbowing or other
9055 artifacts. Setting this to @code{0} could also be used to speed things up at
9056 the cost of some accuracy.
9058 Default value is @code{1}.
9062 These define an exclusion band which excludes the lines between @option{y0} and
9063 @option{y1} from being included in the field matching decision. An exclusion
9064 band can be used to ignore subtitles, a logo, or other things that may
9065 interfere with the matching. @option{y0} sets the starting scan line and
9066 @option{y1} sets the ending line; all lines in between @option{y0} and
9067 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9068 @option{y0} and @option{y1} to the same value will disable the feature.
9069 @option{y0} and @option{y1} defaults to @code{0}.
9072 Set the scene change detection threshold as a percentage of maximum change on
9073 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9074 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9075 @option{scthresh} is @code{[0.0, 100.0]}.
9077 Default value is @code{12.0}.
9080 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9081 account the combed scores of matches when deciding what match to use as the
9082 final match. Available values are:
9086 No final matching based on combed scores.
9088 Combed scores are only used when a scene change is detected.
9090 Use combed scores all the time.
9093 Default is @var{sc}.
9096 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9097 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9098 Available values are:
9102 No forced calculation.
9104 Force p/c/n calculations.
9106 Force p/c/n/u/b calculations.
9109 Default value is @var{none}.
9112 This is the area combing threshold used for combed frame detection. This
9113 essentially controls how "strong" or "visible" combing must be to be detected.
9114 Larger values mean combing must be more visible and smaller values mean combing
9115 can be less visible or strong and still be detected. Valid settings are from
9116 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9117 be detected as combed). This is basically a pixel difference value. A good
9118 range is @code{[8, 12]}.
9120 Default value is @code{9}.
9123 Sets whether or not chroma is considered in the combed frame decision. Only
9124 disable this if your source has chroma problems (rainbowing, etc.) that are
9125 causing problems for the combed frame detection with chroma enabled. Actually,
9126 using @option{chroma}=@var{0} is usually more reliable, except for the case
9127 where there is chroma only combing in the source.
9129 Default value is @code{0}.
9133 Respectively set the x-axis and y-axis size of the window used during combed
9134 frame detection. This has to do with the size of the area in which
9135 @option{combpel} pixels are required to be detected as combed for a frame to be
9136 declared combed. See the @option{combpel} parameter description for more info.
9137 Possible values are any number that is a power of 2 starting at 4 and going up
9140 Default value is @code{16}.
9143 The number of combed pixels inside any of the @option{blocky} by
9144 @option{blockx} size blocks on the frame for the frame to be detected as
9145 combed. While @option{cthresh} controls how "visible" the combing must be, this
9146 setting controls "how much" combing there must be in any localized area (a
9147 window defined by the @option{blockx} and @option{blocky} settings) on the
9148 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9149 which point no frames will ever be detected as combed). This setting is known
9150 as @option{MI} in TFM/VFM vocabulary.
9152 Default value is @code{80}.
9155 @anchor{p/c/n/u/b meaning}
9156 @subsection p/c/n/u/b meaning
9158 @subsubsection p/c/n
9160 We assume the following telecined stream:
9163 Top fields: 1 2 2 3 4
9164 Bottom fields: 1 2 3 4 4
9167 The numbers correspond to the progressive frame the fields relate to. Here, the
9168 first two frames are progressive, the 3rd and 4th are combed, and so on.
9170 When @code{fieldmatch} is configured to run a matching from bottom
9171 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9176 B 1 2 3 4 4 <-- matching reference
9185 As a result of the field matching, we can see that some frames get duplicated.
9186 To perform a complete inverse telecine, you need to rely on a decimation filter
9187 after this operation. See for instance the @ref{decimate} filter.
9189 The same operation now matching from top fields (@option{field}=@var{top})
9194 T 1 2 2 3 4 <-- matching reference
9204 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9205 basically, they refer to the frame and field of the opposite parity:
9208 @item @var{p} matches the field of the opposite parity in the previous frame
9209 @item @var{c} matches the field of the opposite parity in the current frame
9210 @item @var{n} matches the field of the opposite parity in the next frame
9215 The @var{u} and @var{b} matching are a bit special in the sense that they match
9216 from the opposite parity flag. In the following examples, we assume that we are
9217 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9218 'x' is placed above and below each matched fields.
9220 With bottom matching (@option{field}=@var{bottom}):
9225 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9226 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9234 With top matching (@option{field}=@var{top}):
9239 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9240 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9248 @subsection Examples
9250 Simple IVTC of a top field first telecined stream:
9252 fieldmatch=order=tff:combmatch=none, decimate
9255 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9257 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9262 Transform the field order of the input video.
9264 It accepts the following parameters:
9269 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9270 for bottom field first.
9273 The default value is @samp{tff}.
9275 The transformation is done by shifting the picture content up or down
9276 by one line, and filling the remaining line with appropriate picture content.
9277 This method is consistent with most broadcast field order converters.
9279 If the input video is not flagged as being interlaced, or it is already
9280 flagged as being of the required output field order, then this filter does
9281 not alter the incoming video.
9283 It is very useful when converting to or from PAL DV material,
9284 which is bottom field first.
9288 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9291 @section fifo, afifo
9293 Buffer input images and send them when they are requested.
9295 It is mainly useful when auto-inserted by the libavfilter
9298 It does not take parameters.
9300 @section fillborders
9302 Fill borders of the input video, without changing video stream dimensions.
9303 Sometimes video can have garbage at the four edges and you may not want to
9304 crop video input to keep size multiple of some number.
9306 This filter accepts the following options:
9310 Number of pixels to fill from left border.
9313 Number of pixels to fill from right border.
9316 Number of pixels to fill from top border.
9319 Number of pixels to fill from bottom border.
9324 It accepts the following values:
9327 fill pixels using outermost pixels
9330 fill pixels using mirroring
9333 fill pixels with constant value
9336 Default is @var{smear}.
9339 Set color for pixels in fixed mode. Default is @var{black}.
9344 Find a rectangular object
9346 It accepts the following options:
9350 Filepath of the object image, needs to be in gray8.
9353 Detection threshold, default is 0.5.
9356 Number of mipmaps, default is 3.
9358 @item xmin, ymin, xmax, ymax
9359 Specifies the rectangle in which to search.
9362 @subsection Examples
9366 Generate a representative palette of a given video using @command{ffmpeg}:
9368 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9374 Cover a rectangular object
9376 It accepts the following options:
9380 Filepath of the optional cover image, needs to be in yuv420.
9385 It accepts the following values:
9388 cover it by the supplied image
9390 cover it by interpolating the surrounding pixels
9393 Default value is @var{blur}.
9396 @subsection Examples
9400 Generate a representative palette of a given video using @command{ffmpeg}:
9402 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9408 Flood area with values of same pixel components with another values.
9410 It accepts the following options:
9413 Set pixel x coordinate.
9416 Set pixel y coordinate.
9419 Set source #0 component value.
9422 Set source #1 component value.
9425 Set source #2 component value.
9428 Set source #3 component value.
9431 Set destination #0 component value.
9434 Set destination #1 component value.
9437 Set destination #2 component value.
9440 Set destination #3 component value.
9446 Convert the input video to one of the specified pixel formats.
9447 Libavfilter will try to pick one that is suitable as input to
9450 It accepts the following parameters:
9454 A '|'-separated list of pixel format names, such as
9455 "pix_fmts=yuv420p|monow|rgb24".
9459 @subsection Examples
9463 Convert the input video to the @var{yuv420p} format
9465 format=pix_fmts=yuv420p
9468 Convert the input video to any of the formats in the list
9470 format=pix_fmts=yuv420p|yuv444p|yuv410p
9477 Convert the video to specified constant frame rate by duplicating or dropping
9478 frames as necessary.
9480 It accepts the following parameters:
9484 The desired output frame rate. The default is @code{25}.
9487 Assume the first PTS should be the given value, in seconds. This allows for
9488 padding/trimming at the start of stream. By default, no assumption is made
9489 about the first frame's expected PTS, so no padding or trimming is done.
9490 For example, this could be set to 0 to pad the beginning with duplicates of
9491 the first frame if a video stream starts after the audio stream or to trim any
9492 frames with a negative PTS.
9495 Timestamp (PTS) rounding method.
9497 Possible values are:
9504 round towards -infinity
9506 round towards +infinity
9510 The default is @code{near}.
9513 Action performed when reading the last frame.
9515 Possible values are:
9518 Use same timestamp rounding method as used for other frames.
9520 Pass through last frame if input duration has not been reached yet.
9522 The default is @code{round}.
9526 Alternatively, the options can be specified as a flat string:
9527 @var{fps}[:@var{start_time}[:@var{round}]].
9529 See also the @ref{setpts} filter.
9531 @subsection Examples
9535 A typical usage in order to set the fps to 25:
9541 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9543 fps=fps=film:round=near
9549 Pack two different video streams into a stereoscopic video, setting proper
9550 metadata on supported codecs. The two views should have the same size and
9551 framerate and processing will stop when the shorter video ends. Please note
9552 that you may conveniently adjust view properties with the @ref{scale} and
9555 It accepts the following parameters:
9559 The desired packing format. Supported values are:
9564 The views are next to each other (default).
9567 The views are on top of each other.
9570 The views are packed by line.
9573 The views are packed by column.
9576 The views are temporally interleaved.
9585 # Convert left and right views into a frame-sequential video
9586 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9588 # Convert views into a side-by-side video with the same output resolution as the input
9589 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
9594 Change the frame rate by interpolating new video output frames from the source
9597 This filter is not designed to function correctly with interlaced media. If
9598 you wish to change the frame rate of interlaced media then you are required
9599 to deinterlace before this filter and re-interlace after this filter.
9601 A description of the accepted options follows.
9605 Specify the output frames per second. This option can also be specified
9606 as a value alone. The default is @code{50}.
9609 Specify the start of a range where the output frame will be created as a
9610 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9611 the default is @code{15}.
9614 Specify the end of a range where the output frame will be created as a
9615 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9616 the default is @code{240}.
9619 Specify the level at which a scene change is detected as a value between
9620 0 and 100 to indicate a new scene; a low value reflects a low
9621 probability for the current frame to introduce a new scene, while a higher
9622 value means the current frame is more likely to be one.
9623 The default is @code{8.2}.
9626 Specify flags influencing the filter process.
9628 Available value for @var{flags} is:
9631 @item scene_change_detect, scd
9632 Enable scene change detection using the value of the option @var{scene}.
9633 This flag is enabled by default.
9639 Select one frame every N-th frame.
9641 This filter accepts the following option:
9644 Select frame after every @code{step} frames.
9645 Allowed values are positive integers higher than 0. Default value is @code{1}.
9651 Apply a frei0r effect to the input video.
9653 To enable the compilation of this filter, you need to install the frei0r
9654 header and configure FFmpeg with @code{--enable-frei0r}.
9656 It accepts the following parameters:
9661 The name of the frei0r effect to load. If the environment variable
9662 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9663 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9664 Otherwise, the standard frei0r paths are searched, in this order:
9665 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9666 @file{/usr/lib/frei0r-1/}.
9669 A '|'-separated list of parameters to pass to the frei0r effect.
9673 A frei0r effect parameter can be a boolean (its value is either
9674 "y" or "n"), a double, a color (specified as
9675 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9676 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9677 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9678 a position (specified as @var{X}/@var{Y}, where
9679 @var{X} and @var{Y} are floating point numbers) and/or a string.
9681 The number and types of parameters depend on the loaded effect. If an
9682 effect parameter is not specified, the default value is set.
9684 @subsection Examples
9688 Apply the distort0r effect, setting the first two double parameters:
9690 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9694 Apply the colordistance effect, taking a color as the first parameter:
9696 frei0r=colordistance:0.2/0.3/0.4
9697 frei0r=colordistance:violet
9698 frei0r=colordistance:0x112233
9702 Apply the perspective effect, specifying the top left and top right image
9705 frei0r=perspective:0.2/0.2|0.8/0.2
9709 For more information, see
9710 @url{http://frei0r.dyne.org}
9714 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9716 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9717 processing filter, one of them is performed once per block, not per pixel.
9718 This allows for much higher speed.
9720 The filter accepts the following options:
9724 Set quality. This option defines the number of levels for averaging. It accepts
9725 an integer in the range 4-5. Default value is @code{4}.
9728 Force a constant quantization parameter. It accepts an integer in range 0-63.
9729 If not set, the filter will use the QP from the video stream (if available).
9732 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9733 more details but also more artifacts, while higher values make the image smoother
9734 but also blurrier. Default value is @code{0} − PSNR optimal.
9737 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9738 option may cause flicker since the B-Frames have often larger QP. Default is
9739 @code{0} (not enabled).
9745 Apply Gaussian blur filter.
9747 The filter accepts the following options:
9751 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9754 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9757 Set which planes to filter. By default all planes are filtered.
9760 Set vertical sigma, if negative it will be same as @code{sigma}.
9761 Default is @code{-1}.
9766 The filter accepts the following options:
9770 Set the luminance expression.
9772 Set the chrominance blue expression.
9774 Set the chrominance red expression.
9776 Set the alpha expression.
9778 Set the red expression.
9780 Set the green expression.
9782 Set the blue expression.
9785 The colorspace is selected according to the specified options. If one
9786 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9787 options is specified, the filter will automatically select a YCbCr
9788 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9789 @option{blue_expr} options is specified, it will select an RGB
9792 If one of the chrominance expression is not defined, it falls back on the other
9793 one. If no alpha expression is specified it will evaluate to opaque value.
9794 If none of chrominance expressions are specified, they will evaluate
9795 to the luminance expression.
9797 The expressions can use the following variables and functions:
9801 The sequential number of the filtered frame, starting from @code{0}.
9805 The coordinates of the current sample.
9809 The width and height of the image.
9813 Width and height scale depending on the currently filtered plane. It is the
9814 ratio between the corresponding luma plane number of pixels and the current
9815 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9816 @code{0.5,0.5} for chroma planes.
9819 Time of the current frame, expressed in seconds.
9822 Return the value of the pixel at location (@var{x},@var{y}) of the current
9826 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9830 Return the value of the pixel at location (@var{x},@var{y}) of the
9831 blue-difference chroma plane. Return 0 if there is no such plane.
9834 Return the value of the pixel at location (@var{x},@var{y}) of the
9835 red-difference chroma plane. Return 0 if there is no such plane.
9840 Return the value of the pixel at location (@var{x},@var{y}) of the
9841 red/green/blue component. Return 0 if there is no such component.
9844 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9845 plane. Return 0 if there is no such plane.
9848 For functions, if @var{x} and @var{y} are outside the area, the value will be
9849 automatically clipped to the closer edge.
9851 @subsection Examples
9855 Flip the image horizontally:
9861 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9862 wavelength of 100 pixels:
9864 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9868 Generate a fancy enigmatic moving light:
9870 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
9874 Generate a quick emboss effect:
9876 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9880 Modify RGB components depending on pixel position:
9882 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9886 Create a radial gradient that is the same size as the input (also see
9887 the @ref{vignette} filter):
9889 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9895 Fix the banding artifacts that are sometimes introduced into nearly flat
9896 regions by truncation to 8-bit color depth.
9897 Interpolate the gradients that should go where the bands are, and
9900 It is designed for playback only. Do not use it prior to
9901 lossy compression, because compression tends to lose the dither and
9902 bring back the bands.
9904 It accepts the following parameters:
9909 The maximum amount by which the filter will change any one pixel. This is also
9910 the threshold for detecting nearly flat regions. Acceptable values range from
9911 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9915 The neighborhood to fit the gradient to. A larger radius makes for smoother
9916 gradients, but also prevents the filter from modifying the pixels near detailed
9917 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9918 values will be clipped to the valid range.
9922 Alternatively, the options can be specified as a flat string:
9923 @var{strength}[:@var{radius}]
9925 @subsection Examples
9929 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9935 Specify radius, omitting the strength (which will fall-back to the default
9944 A color constancy variation filter which estimates scene illumination via grey edge algorithm
9945 and corrects the scene colors accordingly.
9947 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
9949 The filter accepts the following options:
9953 The order of differentiation to be applied on the scene. Must be chosen in the range
9954 [0,2] and default value is 1.
9957 The Minkowski parameter to be used for calculating the Minkowski distance. Must
9958 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
9959 max value instead of calculating Minkowski distance.
9962 The standard deviation of Gaussian blur to be applied on the scene. Must be
9963 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
9964 can't be euqal to 0 if @var{difford} is greater than 0.
9967 @subsection Examples
9973 greyedge=difford=1:minknorm=5:sigma=2
9979 greyedge=difford=1:minknorm=0:sigma=2
9987 Apply a Hald CLUT to a video stream.
9989 First input is the video stream to process, and second one is the Hald CLUT.
9990 The Hald CLUT input can be a simple picture or a complete video stream.
9992 The filter accepts the following options:
9996 Force termination when the shortest input terminates. Default is @code{0}.
9998 Continue applying the last CLUT after the end of the stream. A value of
9999 @code{0} disable the filter after the last frame of the CLUT is reached.
10000 Default is @code{1}.
10003 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10004 filters share the same internals).
10006 More information about the Hald CLUT can be found on Eskil Steenberg's website
10007 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10009 @subsection Workflow examples
10011 @subsubsection Hald CLUT video stream
10013 Generate an identity Hald CLUT stream altered with various effects:
10015 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
10018 Note: make sure you use a lossless codec.
10020 Then use it with @code{haldclut} to apply it on some random stream:
10022 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10025 The Hald CLUT will be applied to the 10 first seconds (duration of
10026 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10027 to the remaining frames of the @code{mandelbrot} stream.
10029 @subsubsection Hald CLUT with preview
10031 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10032 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10033 biggest possible square starting at the top left of the picture. The remaining
10034 padding pixels (bottom or right) will be ignored. This area can be used to add
10035 a preview of the Hald CLUT.
10037 Typically, the following generated Hald CLUT will be supported by the
10038 @code{haldclut} filter:
10041 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10042 pad=iw+320 [padded_clut];
10043 smptebars=s=320x256, split [a][b];
10044 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10045 [main][b] overlay=W-320" -frames:v 1 clut.png
10048 It contains the original and a preview of the effect of the CLUT: SMPTE color
10049 bars are displayed on the right-top, and below the same color bars processed by
10052 Then, the effect of this Hald CLUT can be visualized with:
10054 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10059 Flip the input video horizontally.
10061 For example, to horizontally flip the input video with @command{ffmpeg}:
10063 ffmpeg -i in.avi -vf "hflip" out.avi
10067 This filter applies a global color histogram equalization on a
10070 It can be used to correct video that has a compressed range of pixel
10071 intensities. The filter redistributes the pixel intensities to
10072 equalize their distribution across the intensity range. It may be
10073 viewed as an "automatically adjusting contrast filter". This filter is
10074 useful only for correcting degraded or poorly captured source
10077 The filter accepts the following options:
10081 Determine the amount of equalization to be applied. As the strength
10082 is reduced, the distribution of pixel intensities more-and-more
10083 approaches that of the input frame. The value must be a float number
10084 in the range [0,1] and defaults to 0.200.
10087 Set the maximum intensity that can generated and scale the output
10088 values appropriately. The strength should be set as desired and then
10089 the intensity can be limited if needed to avoid washing-out. The value
10090 must be a float number in the range [0,1] and defaults to 0.210.
10093 Set the antibanding level. If enabled the filter will randomly vary
10094 the luminance of output pixels by a small amount to avoid banding of
10095 the histogram. Possible values are @code{none}, @code{weak} or
10096 @code{strong}. It defaults to @code{none}.
10101 Compute and draw a color distribution histogram for the input video.
10103 The computed histogram is a representation of the color component
10104 distribution in an image.
10106 Standard histogram displays the color components distribution in an image.
10107 Displays color graph for each color component. Shows distribution of
10108 the Y, U, V, A or R, G, B components, depending on input format, in the
10109 current frame. Below each graph a color component scale meter is shown.
10111 The filter accepts the following options:
10115 Set height of level. Default value is @code{200}.
10116 Allowed range is [50, 2048].
10119 Set height of color scale. Default value is @code{12}.
10120 Allowed range is [0, 40].
10124 It accepts the following values:
10127 Per color component graphs are placed below each other.
10130 Per color component graphs are placed side by side.
10133 Presents information identical to that in the @code{parade}, except
10134 that the graphs representing color components are superimposed directly
10137 Default is @code{stack}.
10140 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10141 Default is @code{linear}.
10144 Set what color components to display.
10145 Default is @code{7}.
10148 Set foreground opacity. Default is @code{0.7}.
10151 Set background opacity. Default is @code{0.5}.
10154 @subsection Examples
10159 Calculate and draw histogram:
10161 ffplay -i input -vf histogram
10169 This is a high precision/quality 3d denoise filter. It aims to reduce
10170 image noise, producing smooth images and making still images really
10171 still. It should enhance compressibility.
10173 It accepts the following optional parameters:
10177 A non-negative floating point number which specifies spatial luma strength.
10178 It defaults to 4.0.
10180 @item chroma_spatial
10181 A non-negative floating point number which specifies spatial chroma strength.
10182 It defaults to 3.0*@var{luma_spatial}/4.0.
10185 A floating point number which specifies luma temporal strength. It defaults to
10186 6.0*@var{luma_spatial}/4.0.
10189 A floating point number which specifies chroma temporal strength. It defaults to
10190 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10193 @section hwdownload
10195 Download hardware frames to system memory.
10197 The input must be in hardware frames, and the output a non-hardware format.
10198 Not all formats will be supported on the output - it may be necessary to insert
10199 an additional @option{format} filter immediately following in the graph to get
10200 the output in a supported format.
10204 Map hardware frames to system memory or to another device.
10206 This filter has several different modes of operation; which one is used depends
10207 on the input and output formats:
10210 Hardware frame input, normal frame output
10212 Map the input frames to system memory and pass them to the output. If the
10213 original hardware frame is later required (for example, after overlaying
10214 something else on part of it), the @option{hwmap} filter can be used again
10215 in the next mode to retrieve it.
10217 Normal frame input, hardware frame output
10219 If the input is actually a software-mapped hardware frame, then unmap it -
10220 that is, return the original hardware frame.
10222 Otherwise, a device must be provided. Create new hardware surfaces on that
10223 device for the output, then map them back to the software format at the input
10224 and give those frames to the preceding filter. This will then act like the
10225 @option{hwupload} filter, but may be able to avoid an additional copy when
10226 the input is already in a compatible format.
10228 Hardware frame input and output
10230 A device must be supplied for the output, either directly or with the
10231 @option{derive_device} option. The input and output devices must be of
10232 different types and compatible - the exact meaning of this is
10233 system-dependent, but typically it means that they must refer to the same
10234 underlying hardware context (for example, refer to the same graphics card).
10236 If the input frames were originally created on the output device, then unmap
10237 to retrieve the original frames.
10239 Otherwise, map the frames to the output device - create new hardware frames
10240 on the output corresponding to the frames on the input.
10243 The following additional parameters are accepted:
10247 Set the frame mapping mode. Some combination of:
10250 The mapped frame should be readable.
10252 The mapped frame should be writeable.
10254 The mapping will always overwrite the entire frame.
10256 This may improve performance in some cases, as the original contents of the
10257 frame need not be loaded.
10259 The mapping must not involve any copying.
10261 Indirect mappings to copies of frames are created in some cases where either
10262 direct mapping is not possible or it would have unexpected properties.
10263 Setting this flag ensures that the mapping is direct and will fail if that is
10266 Defaults to @var{read+write} if not specified.
10268 @item derive_device @var{type}
10269 Rather than using the device supplied at initialisation, instead derive a new
10270 device of type @var{type} from the device the input frames exist on.
10273 In a hardware to hardware mapping, map in reverse - create frames in the sink
10274 and map them back to the source. This may be necessary in some cases where
10275 a mapping in one direction is required but only the opposite direction is
10276 supported by the devices being used.
10278 This option is dangerous - it may break the preceding filter in undefined
10279 ways if there are any additional constraints on that filter's output.
10280 Do not use it without fully understanding the implications of its use.
10285 Upload system memory frames to hardware surfaces.
10287 The device to upload to must be supplied when the filter is initialised. If
10288 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10291 @anchor{hwupload_cuda}
10292 @section hwupload_cuda
10294 Upload system memory frames to a CUDA device.
10296 It accepts the following optional parameters:
10300 The number of the CUDA device to use
10305 Apply a high-quality magnification filter designed for pixel art. This filter
10306 was originally created by Maxim Stepin.
10308 It accepts the following option:
10312 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10313 @code{hq3x} and @code{4} for @code{hq4x}.
10314 Default is @code{3}.
10318 Stack input videos horizontally.
10320 All streams must be of same pixel format and of same height.
10322 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10323 to create same output.
10325 The filter accept the following option:
10329 Set number of input streams. Default is 2.
10332 If set to 1, force the output to terminate when the shortest input
10333 terminates. Default value is 0.
10338 Modify the hue and/or the saturation of the input.
10340 It accepts the following parameters:
10344 Specify the hue angle as a number of degrees. It accepts an expression,
10345 and defaults to "0".
10348 Specify the saturation in the [-10,10] range. It accepts an expression and
10352 Specify the hue angle as a number of radians. It accepts an
10353 expression, and defaults to "0".
10356 Specify the brightness in the [-10,10] range. It accepts an expression and
10360 @option{h} and @option{H} are mutually exclusive, and can't be
10361 specified at the same time.
10363 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10364 expressions containing the following constants:
10368 frame count of the input frame starting from 0
10371 presentation timestamp of the input frame expressed in time base units
10374 frame rate of the input video, NAN if the input frame rate is unknown
10377 timestamp expressed in seconds, NAN if the input timestamp is unknown
10380 time base of the input video
10383 @subsection Examples
10387 Set the hue to 90 degrees and the saturation to 1.0:
10393 Same command but expressing the hue in radians:
10399 Rotate hue and make the saturation swing between 0
10400 and 2 over a period of 1 second:
10402 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10406 Apply a 3 seconds saturation fade-in effect starting at 0:
10408 hue="s=min(t/3\,1)"
10411 The general fade-in expression can be written as:
10413 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10417 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10419 hue="s=max(0\, min(1\, (8-t)/3))"
10422 The general fade-out expression can be written as:
10424 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10429 @subsection Commands
10431 This filter supports the following commands:
10437 Modify the hue and/or the saturation and/or brightness of the input video.
10438 The command accepts the same syntax of the corresponding option.
10440 If the specified expression is not valid, it is kept at its current
10444 @section hysteresis
10446 Grow first stream into second stream by connecting components.
10447 This makes it possible to build more robust edge masks.
10449 This filter accepts the following options:
10453 Set which planes will be processed as bitmap, unprocessed planes will be
10454 copied from first stream.
10455 By default value 0xf, all planes will be processed.
10458 Set threshold which is used in filtering. If pixel component value is higher than
10459 this value filter algorithm for connecting components is activated.
10460 By default value is 0.
10465 Detect video interlacing type.
10467 This filter tries to detect if the input frames are interlaced, progressive,
10468 top or bottom field first. It will also try to detect fields that are
10469 repeated between adjacent frames (a sign of telecine).
10471 Single frame detection considers only immediately adjacent frames when classifying each frame.
10472 Multiple frame detection incorporates the classification history of previous frames.
10474 The filter will log these metadata values:
10477 @item single.current_frame
10478 Detected type of current frame using single-frame detection. One of:
10479 ``tff'' (top field first), ``bff'' (bottom field first),
10480 ``progressive'', or ``undetermined''
10483 Cumulative number of frames detected as top field first using single-frame detection.
10486 Cumulative number of frames detected as top field first using multiple-frame detection.
10489 Cumulative number of frames detected as bottom field first using single-frame detection.
10491 @item multiple.current_frame
10492 Detected type of current frame using multiple-frame detection. One of:
10493 ``tff'' (top field first), ``bff'' (bottom field first),
10494 ``progressive'', or ``undetermined''
10497 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10499 @item single.progressive
10500 Cumulative number of frames detected as progressive using single-frame detection.
10502 @item multiple.progressive
10503 Cumulative number of frames detected as progressive using multiple-frame detection.
10505 @item single.undetermined
10506 Cumulative number of frames that could not be classified using single-frame detection.
10508 @item multiple.undetermined
10509 Cumulative number of frames that could not be classified using multiple-frame detection.
10511 @item repeated.current_frame
10512 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10514 @item repeated.neither
10515 Cumulative number of frames with no repeated field.
10518 Cumulative number of frames with the top field repeated from the previous frame's top field.
10520 @item repeated.bottom
10521 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10524 The filter accepts the following options:
10528 Set interlacing threshold.
10530 Set progressive threshold.
10532 Threshold for repeated field detection.
10534 Number of frames after which a given frame's contribution to the
10535 statistics is halved (i.e., it contributes only 0.5 to its
10536 classification). The default of 0 means that all frames seen are given
10537 full weight of 1.0 forever.
10538 @item analyze_interlaced_flag
10539 When this is not 0 then idet will use the specified number of frames to determine
10540 if the interlaced flag is accurate, it will not count undetermined frames.
10541 If the flag is found to be accurate it will be used without any further
10542 computations, if it is found to be inaccurate it will be cleared without any
10543 further computations. This allows inserting the idet filter as a low computational
10544 method to clean up the interlaced flag
10549 Deinterleave or interleave fields.
10551 This filter allows one to process interlaced images fields without
10552 deinterlacing them. Deinterleaving splits the input frame into 2
10553 fields (so called half pictures). Odd lines are moved to the top
10554 half of the output image, even lines to the bottom half.
10555 You can process (filter) them independently and then re-interleave them.
10557 The filter accepts the following options:
10561 @item chroma_mode, c
10562 @item alpha_mode, a
10563 Available values for @var{luma_mode}, @var{chroma_mode} and
10564 @var{alpha_mode} are:
10570 @item deinterleave, d
10571 Deinterleave fields, placing one above the other.
10573 @item interleave, i
10574 Interleave fields. Reverse the effect of deinterleaving.
10576 Default value is @code{none}.
10578 @item luma_swap, ls
10579 @item chroma_swap, cs
10580 @item alpha_swap, as
10581 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10586 Apply inflate effect to the video.
10588 This filter replaces the pixel by the local(3x3) average by taking into account
10589 only values higher than the pixel.
10591 It accepts the following options:
10598 Limit the maximum change for each plane, default is 65535.
10599 If 0, plane will remain unchanged.
10604 Simple interlacing filter from progressive contents. This interleaves upper (or
10605 lower) lines from odd frames with lower (or upper) lines from even frames,
10606 halving the frame rate and preserving image height.
10609 Original Original New Frame
10610 Frame 'j' Frame 'j+1' (tff)
10611 ========== =========== ==================
10612 Line 0 --------------------> Frame 'j' Line 0
10613 Line 1 Line 1 ----> Frame 'j+1' Line 1
10614 Line 2 ---------------------> Frame 'j' Line 2
10615 Line 3 Line 3 ----> Frame 'j+1' Line 3
10617 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10620 It accepts the following optional parameters:
10624 This determines whether the interlaced frame is taken from the even
10625 (tff - default) or odd (bff) lines of the progressive frame.
10628 Vertical lowpass filter to avoid twitter interlacing and
10629 reduce moire patterns.
10633 Disable vertical lowpass filter
10636 Enable linear filter (default)
10639 Enable complex filter. This will slightly less reduce twitter and moire
10640 but better retain detail and subjective sharpness impression.
10647 Deinterlace input video by applying Donald Graft's adaptive kernel
10648 deinterling. Work on interlaced parts of a video to produce
10649 progressive frames.
10651 The description of the accepted parameters follows.
10655 Set the threshold which affects the filter's tolerance when
10656 determining if a pixel line must be processed. It must be an integer
10657 in the range [0,255] and defaults to 10. A value of 0 will result in
10658 applying the process on every pixels.
10661 Paint pixels exceeding the threshold value to white if set to 1.
10665 Set the fields order. Swap fields if set to 1, leave fields alone if
10669 Enable additional sharpening if set to 1. Default is 0.
10672 Enable twoway sharpening if set to 1. Default is 0.
10675 @subsection Examples
10679 Apply default values:
10681 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10685 Enable additional sharpening:
10691 Paint processed pixels in white:
10697 @section lenscorrection
10699 Correct radial lens distortion
10701 This filter can be used to correct for radial distortion as can result from the use
10702 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10703 one can use tools available for example as part of opencv or simply trial-and-error.
10704 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10705 and extract the k1 and k2 coefficients from the resulting matrix.
10707 Note that effectively the same filter is available in the open-source tools Krita and
10708 Digikam from the KDE project.
10710 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10711 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10712 brightness distribution, so you may want to use both filters together in certain
10713 cases, though you will have to take care of ordering, i.e. whether vignetting should
10714 be applied before or after lens correction.
10716 @subsection Options
10718 The filter accepts the following options:
10722 Relative x-coordinate of the focal point of the image, and thereby the center of the
10723 distortion. This value has a range [0,1] and is expressed as fractions of the image
10724 width. Default is 0.5.
10726 Relative y-coordinate of the focal point of the image, and thereby the center of the
10727 distortion. This value has a range [0,1] and is expressed as fractions of the image
10728 height. Default is 0.5.
10730 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
10731 no correction. Default is 0.
10733 Coefficient of the double quadratic correction term. This value has a range [-1,1].
10734 0 means no correction. Default is 0.
10737 The formula that generates the correction is:
10739 @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)
10741 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10742 distances from the focal point in the source and target images, respectively.
10746 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
10748 The @code{lensfun} filter requires the camera make, camera model, and lens model
10749 to apply the lens correction. The filter will load the lensfun database and
10750 query it to find the corresponding camera and lens entries in the database. As
10751 long as these entries can be found with the given options, the filter can
10752 perform corrections on frames. Note that incomplete strings will result in the
10753 filter choosing the best match with the given options, and the filter will
10754 output the chosen camera and lens models (logged with level "info"). You must
10755 provide the make, camera model, and lens model as they are required.
10757 The filter accepts the following options:
10761 The make of the camera (for example, "Canon"). This option is required.
10764 The model of the camera (for example, "Canon EOS 100D"). This option is
10768 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
10769 option is required.
10772 The type of correction to apply. The following values are valid options:
10776 Enables fixing lens vignetting.
10779 Enables fixing lens geometry. This is the default.
10782 Enables fixing chromatic aberrations.
10785 Enables fixing lens vignetting and lens geometry.
10788 Enables fixing lens vignetting and chromatic aberrations.
10791 Enables fixing both lens geometry and chromatic aberrations.
10794 Enables all possible corrections.
10798 The focal length of the image/video (zoom; expected constant for video). For
10799 example, a 18--55mm lens has focal length range of [18--55], so a value in that
10800 range should be chosen when using that lens. Default 18.
10803 The aperture of the image/video (expected constant for video). Note that
10804 aperture is only used for vignetting correction. Default 3.5.
10806 @item focus_distance
10807 The focus distance of the image/video (expected constant for video). Note that
10808 focus distance is only used for vignetting and only slightly affects the
10809 vignetting correction process. If unknown, leave it at the default value (which
10812 @item target_geometry
10813 The target geometry of the output image/video. The following values are valid
10817 @item rectilinear (default)
10820 @item equirectangular
10821 @item fisheye_orthographic
10822 @item fisheye_stereographic
10823 @item fisheye_equisolid
10824 @item fisheye_thoby
10827 Apply the reverse of image correction (instead of correcting distortion, apply
10830 @item interpolation
10831 The type of interpolation used when correcting distortion. The following values
10836 @item linear (default)
10841 @subsection Examples
10845 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
10846 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
10850 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8 -c:v h264 -b:v 8000k output.mov
10854 Apply the same as before, but only for the first 5 seconds of video.
10857 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8:enable='lte(t\,5)' -c:v h264 -b:v 8000k output.mov
10864 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10865 score between two input videos.
10867 The obtained VMAF score is printed through the logging system.
10869 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10870 After installing the library it can be enabled using:
10871 @code{./configure --enable-libvmaf --enable-version3}.
10872 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10874 The filter has following options:
10878 Set the model path which is to be used for SVM.
10879 Default value: @code{"vmaf_v0.6.1.pkl"}
10882 Set the file path to be used to store logs.
10885 Set the format of the log file (xml or json).
10887 @item enable_transform
10888 Enables transform for computing vmaf.
10891 Invokes the phone model which will generate VMAF scores higher than in the
10892 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10895 Enables computing psnr along with vmaf.
10898 Enables computing ssim along with vmaf.
10901 Enables computing ms_ssim along with vmaf.
10904 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10907 Set number of threads to be used when computing vmaf.
10910 Set interval for frame subsampling used when computing vmaf.
10912 @item enable_conf_interval
10913 Enables confidence interval.
10916 This filter also supports the @ref{framesync} options.
10918 On the below examples the input file @file{main.mpg} being processed is
10919 compared with the reference file @file{ref.mpg}.
10922 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10925 Example with options:
10927 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10932 Limits the pixel components values to the specified range [min, max].
10934 The filter accepts the following options:
10938 Lower bound. Defaults to the lowest allowed value for the input.
10941 Upper bound. Defaults to the highest allowed value for the input.
10944 Specify which planes will be processed. Defaults to all available.
10951 The filter accepts the following options:
10955 Set the number of loops. Setting this value to -1 will result in infinite loops.
10959 Set maximal size in number of frames. Default is 0.
10962 Set first frame of loop. Default is 0.
10968 Apply a 3D LUT to an input video.
10970 The filter accepts the following options:
10974 Set the 3D LUT file name.
10976 Currently supported formats:
10988 Select interpolation mode.
10990 Available values are:
10994 Use values from the nearest defined point.
10996 Interpolate values using the 8 points defining a cube.
10998 Interpolate values using a tetrahedron.
11002 This filter also supports the @ref{framesync} options.
11006 Turn certain luma values into transparency.
11008 The filter accepts the following options:
11012 Set the luma which will be used as base for transparency.
11013 Default value is @code{0}.
11016 Set the range of luma values to be keyed out.
11017 Default value is @code{0}.
11020 Set the range of softness. Default value is @code{0}.
11021 Use this to control gradual transition from zero to full transparency.
11024 @section lut, lutrgb, lutyuv
11026 Compute a look-up table for binding each pixel component input value
11027 to an output value, and apply it to the input video.
11029 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11030 to an RGB input video.
11032 These filters accept the following parameters:
11035 set first pixel component expression
11037 set second pixel component expression
11039 set third pixel component expression
11041 set fourth pixel component expression, corresponds to the alpha component
11044 set red component expression
11046 set green component expression
11048 set blue component expression
11050 alpha component expression
11053 set Y/luminance component expression
11055 set U/Cb component expression
11057 set V/Cr component expression
11060 Each of them specifies the expression to use for computing the lookup table for
11061 the corresponding pixel component values.
11063 The exact component associated to each of the @var{c*} options depends on the
11066 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11067 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11069 The expressions can contain the following constants and functions:
11074 The input width and height.
11077 The input value for the pixel component.
11080 The input value, clipped to the @var{minval}-@var{maxval} range.
11083 The maximum value for the pixel component.
11086 The minimum value for the pixel component.
11089 The negated value for the pixel component value, clipped to the
11090 @var{minval}-@var{maxval} range; it corresponds to the expression
11091 "maxval-clipval+minval".
11094 The computed value in @var{val}, clipped to the
11095 @var{minval}-@var{maxval} range.
11097 @item gammaval(gamma)
11098 The computed gamma correction value of the pixel component value,
11099 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11101 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11105 All expressions default to "val".
11107 @subsection Examples
11111 Negate input video:
11113 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11114 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11117 The above is the same as:
11119 lutrgb="r=negval:g=negval:b=negval"
11120 lutyuv="y=negval:u=negval:v=negval"
11130 Remove chroma components, turning the video into a graytone image:
11132 lutyuv="u=128:v=128"
11136 Apply a luma burning effect:
11142 Remove green and blue components:
11148 Set a constant alpha channel value on input:
11150 format=rgba,lutrgb=a="maxval-minval/2"
11154 Correct luminance gamma by a factor of 0.5:
11156 lutyuv=y=gammaval(0.5)
11160 Discard least significant bits of luma:
11162 lutyuv=y='bitand(val, 128+64+32)'
11166 Technicolor like effect:
11168 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11172 @section lut2, tlut2
11174 The @code{lut2} filter takes two input streams and outputs one
11177 The @code{tlut2} (time lut2) filter takes two consecutive frames
11178 from one single stream.
11180 This filter accepts the following parameters:
11183 set first pixel component expression
11185 set second pixel component expression
11187 set third pixel component expression
11189 set fourth pixel component expression, corresponds to the alpha component
11192 Each of them specifies the expression to use for computing the lookup table for
11193 the corresponding pixel component values.
11195 The exact component associated to each of the @var{c*} options depends on the
11198 The expressions can contain the following constants:
11203 The input width and height.
11206 The first input value for the pixel component.
11209 The second input value for the pixel component.
11212 The first input video bit depth.
11215 The second input video bit depth.
11218 All expressions default to "x".
11220 @subsection Examples
11224 Highlight differences between two RGB video streams:
11226 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)'
11230 Highlight differences between two YUV video streams:
11232 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)'
11236 Show max difference between two video streams:
11238 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)))'
11242 @section maskedclamp
11244 Clamp the first input stream with the second input and third input stream.
11246 Returns the value of first stream to be between second input
11247 stream - @code{undershoot} and third input stream + @code{overshoot}.
11249 This filter accepts the following options:
11252 Default value is @code{0}.
11255 Default value is @code{0}.
11258 Set which planes will be processed as bitmap, unprocessed planes will be
11259 copied from first stream.
11260 By default value 0xf, all planes will be processed.
11263 @section maskedmerge
11265 Merge the first input stream with the second input stream using per pixel
11266 weights in the third input stream.
11268 A value of 0 in the third stream pixel component means that pixel component
11269 from first stream is returned unchanged, while maximum value (eg. 255 for
11270 8-bit videos) means that pixel component from second stream is returned
11271 unchanged. Intermediate values define the amount of merging between both
11272 input stream's pixel components.
11274 This filter accepts the following options:
11277 Set which planes will be processed as bitmap, unprocessed planes will be
11278 copied from first stream.
11279 By default value 0xf, all planes will be processed.
11284 Apply motion-compensation deinterlacing.
11286 It needs one field per frame as input and must thus be used together
11287 with yadif=1/3 or equivalent.
11289 This filter accepts the following options:
11292 Set the deinterlacing mode.
11294 It accepts one of the following values:
11299 use iterative motion estimation
11301 like @samp{slow}, but use multiple reference frames.
11303 Default value is @samp{fast}.
11306 Set the picture field parity assumed for the input video. It must be
11307 one of the following values:
11311 assume top field first
11313 assume bottom field first
11316 Default value is @samp{bff}.
11319 Set per-block quantization parameter (QP) used by the internal
11322 Higher values should result in a smoother motion vector field but less
11323 optimal individual vectors. Default value is 1.
11326 @section mergeplanes
11328 Merge color channel components from several video streams.
11330 The filter accepts up to 4 input streams, and merge selected input
11331 planes to the output video.
11333 This filter accepts the following options:
11336 Set input to output plane mapping. Default is @code{0}.
11338 The mappings is specified as a bitmap. It should be specified as a
11339 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
11340 mapping for the first plane of the output stream. 'A' sets the number of
11341 the input stream to use (from 0 to 3), and 'a' the plane number of the
11342 corresponding input to use (from 0 to 3). The rest of the mappings is
11343 similar, 'Bb' describes the mapping for the output stream second
11344 plane, 'Cc' describes the mapping for the output stream third plane and
11345 'Dd' describes the mapping for the output stream fourth plane.
11348 Set output pixel format. Default is @code{yuva444p}.
11351 @subsection Examples
11355 Merge three gray video streams of same width and height into single video stream:
11357 [a0][a1][a2]mergeplanes=0x001020:yuv444p
11361 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
11363 [a0][a1]mergeplanes=0x00010210:yuva444p
11367 Swap Y and A plane in yuva444p stream:
11369 format=yuva444p,mergeplanes=0x03010200:yuva444p
11373 Swap U and V plane in yuv420p stream:
11375 format=yuv420p,mergeplanes=0x000201:yuv420p
11379 Cast a rgb24 clip to yuv444p:
11381 format=rgb24,mergeplanes=0x000102:yuv444p
11387 Estimate and export motion vectors using block matching algorithms.
11388 Motion vectors are stored in frame side data to be used by other filters.
11390 This filter accepts the following options:
11393 Specify the motion estimation method. Accepts one of the following values:
11397 Exhaustive search algorithm.
11399 Three step search algorithm.
11401 Two dimensional logarithmic search algorithm.
11403 New three step search algorithm.
11405 Four step search algorithm.
11407 Diamond search algorithm.
11409 Hexagon-based search algorithm.
11411 Enhanced predictive zonal search algorithm.
11413 Uneven multi-hexagon search algorithm.
11415 Default value is @samp{esa}.
11418 Macroblock size. Default @code{16}.
11421 Search parameter. Default @code{7}.
11424 @section midequalizer
11426 Apply Midway Image Equalization effect using two video streams.
11428 Midway Image Equalization adjusts a pair of images to have the same
11429 histogram, while maintaining their dynamics as much as possible. It's
11430 useful for e.g. matching exposures from a pair of stereo cameras.
11432 This filter has two inputs and one output, which must be of same pixel format, but
11433 may be of different sizes. The output of filter is first input adjusted with
11434 midway histogram of both inputs.
11436 This filter accepts the following option:
11440 Set which planes to process. Default is @code{15}, which is all available planes.
11443 @section minterpolate
11445 Convert the video to specified frame rate using motion interpolation.
11447 This filter accepts the following options:
11450 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}.
11453 Motion interpolation mode. Following values are accepted:
11456 Duplicate previous or next frame for interpolating new ones.
11458 Blend source frames. Interpolated frame is mean of previous and next frames.
11460 Motion compensated interpolation. Following options are effective when this mode is selected:
11464 Motion compensation mode. Following values are accepted:
11467 Overlapped block motion compensation.
11469 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
11471 Default mode is @samp{obmc}.
11474 Motion estimation mode. Following values are accepted:
11477 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
11479 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11481 Default mode is @samp{bilat}.
11484 The algorithm to be used for motion estimation. Following values are accepted:
11487 Exhaustive search algorithm.
11489 Three step search algorithm.
11491 Two dimensional logarithmic search algorithm.
11493 New three step search algorithm.
11495 Four step search algorithm.
11497 Diamond search algorithm.
11499 Hexagon-based search algorithm.
11501 Enhanced predictive zonal search algorithm.
11503 Uneven multi-hexagon search algorithm.
11505 Default algorithm is @samp{epzs}.
11508 Macroblock size. Default @code{16}.
11511 Motion estimation search parameter. Default @code{32}.
11514 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).
11519 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:
11522 Disable scene change detection.
11524 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11526 Default method is @samp{fdiff}.
11528 @item scd_threshold
11529 Scene change detection threshold. Default is @code{5.0}.
11534 Mix several video input streams into one video stream.
11536 A description of the accepted options follows.
11540 The number of inputs. If unspecified, it defaults to 2.
11543 Specify weight of each input video stream as sequence.
11544 Each weight is separated by space. If number of weights
11545 is smaller than number of @var{frames} last specified
11546 weight will be used for all remaining unset weights.
11549 Specify scale, if it is set it will be multiplied with sum
11550 of each weight multiplied with pixel values to give final destination
11551 pixel value. By default @var{scale} is auto scaled to sum of weights.
11554 Specify how end of stream is determined.
11557 The duration of the longest input. (default)
11560 The duration of the shortest input.
11563 The duration of the first input.
11567 @section mpdecimate
11569 Drop frames that do not differ greatly from the previous frame in
11570 order to reduce frame rate.
11572 The main use of this filter is for very-low-bitrate encoding
11573 (e.g. streaming over dialup modem), but it could in theory be used for
11574 fixing movies that were inverse-telecined incorrectly.
11576 A description of the accepted options follows.
11580 Set the maximum number of consecutive frames which can be dropped (if
11581 positive), or the minimum interval between dropped frames (if
11582 negative). If the value is 0, the frame is dropped disregarding the
11583 number of previous sequentially dropped frames.
11585 Default value is 0.
11590 Set the dropping threshold values.
11592 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11593 represent actual pixel value differences, so a threshold of 64
11594 corresponds to 1 unit of difference for each pixel, or the same spread
11595 out differently over the block.
11597 A frame is a candidate for dropping if no 8x8 blocks differ by more
11598 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11599 meaning the whole image) differ by more than a threshold of @option{lo}.
11601 Default value for @option{hi} is 64*12, default value for @option{lo} is
11602 64*5, and default value for @option{frac} is 0.33.
11608 Negate (invert) the input video.
11610 It accepts the following option:
11615 With value 1, it negates the alpha component, if present. Default value is 0.
11620 Denoise frames using Non-Local Means algorithm.
11622 Each pixel is adjusted by looking for other pixels with similar contexts. This
11623 context similarity is defined by comparing their surrounding patches of size
11624 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11627 Note that the research area defines centers for patches, which means some
11628 patches will be made of pixels outside that research area.
11630 The filter accepts the following options.
11634 Set denoising strength.
11640 Same as @option{p} but for chroma planes.
11642 The default value is @var{0} and means automatic.
11648 Same as @option{r} but for chroma planes.
11650 The default value is @var{0} and means automatic.
11655 Deinterlace video using neural network edge directed interpolation.
11657 This filter accepts the following options:
11661 Mandatory option, without binary file filter can not work.
11662 Currently file can be found here:
11663 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11666 Set which frames to deinterlace, by default it is @code{all}.
11667 Can be @code{all} or @code{interlaced}.
11670 Set mode of operation.
11672 Can be one of the following:
11676 Use frame flags, both fields.
11678 Use frame flags, single field.
11680 Use top field only.
11682 Use bottom field only.
11684 Use both fields, top first.
11686 Use both fields, bottom first.
11690 Set which planes to process, by default filter process all frames.
11693 Set size of local neighborhood around each pixel, used by the predictor neural
11696 Can be one of the following:
11709 Set the number of neurons in predictor neural network.
11710 Can be one of the following:
11721 Controls the number of different neural network predictions that are blended
11722 together to compute the final output value. Can be @code{fast}, default or
11726 Set which set of weights to use in the predictor.
11727 Can be one of the following:
11731 weights trained to minimize absolute error
11733 weights trained to minimize squared error
11737 Controls whether or not the prescreener neural network is used to decide
11738 which pixels should be processed by the predictor neural network and which
11739 can be handled by simple cubic interpolation.
11740 The prescreener is trained to know whether cubic interpolation will be
11741 sufficient for a pixel or whether it should be predicted by the predictor nn.
11742 The computational complexity of the prescreener nn is much less than that of
11743 the predictor nn. Since most pixels can be handled by cubic interpolation,
11744 using the prescreener generally results in much faster processing.
11745 The prescreener is pretty accurate, so the difference between using it and not
11746 using it is almost always unnoticeable.
11748 Can be one of the following:
11756 Default is @code{new}.
11759 Set various debugging flags.
11764 Force libavfilter not to use any of the specified pixel formats for the
11765 input to the next filter.
11767 It accepts the following parameters:
11771 A '|'-separated list of pixel format names, such as
11772 pix_fmts=yuv420p|monow|rgb24".
11776 @subsection Examples
11780 Force libavfilter to use a format different from @var{yuv420p} for the
11781 input to the vflip filter:
11783 noformat=pix_fmts=yuv420p,vflip
11787 Convert the input video to any of the formats not contained in the list:
11789 noformat=yuv420p|yuv444p|yuv410p
11795 Add noise on video input frame.
11797 The filter accepts the following options:
11805 Set noise seed for specific pixel component or all pixel components in case
11806 of @var{all_seed}. Default value is @code{123457}.
11808 @item all_strength, alls
11809 @item c0_strength, c0s
11810 @item c1_strength, c1s
11811 @item c2_strength, c2s
11812 @item c3_strength, c3s
11813 Set noise strength for specific pixel component or all pixel components in case
11814 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11816 @item all_flags, allf
11817 @item c0_flags, c0f
11818 @item c1_flags, c1f
11819 @item c2_flags, c2f
11820 @item c3_flags, c3f
11821 Set pixel component flags or set flags for all components if @var{all_flags}.
11822 Available values for component flags are:
11825 averaged temporal noise (smoother)
11827 mix random noise with a (semi)regular pattern
11829 temporal noise (noise pattern changes between frames)
11831 uniform noise (gaussian otherwise)
11835 @subsection Examples
11837 Add temporal and uniform noise to input video:
11839 noise=alls=20:allf=t+u
11844 Normalize RGB video (aka histogram stretching, contrast stretching).
11845 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11847 For each channel of each frame, the filter computes the input range and maps
11848 it linearly to the user-specified output range. The output range defaults
11849 to the full dynamic range from pure black to pure white.
11851 Temporal smoothing can be used on the input range to reduce flickering (rapid
11852 changes in brightness) caused when small dark or bright objects enter or leave
11853 the scene. This is similar to the auto-exposure (automatic gain control) on a
11854 video camera, and, like a video camera, it may cause a period of over- or
11855 under-exposure of the video.
11857 The R,G,B channels can be normalized independently, which may cause some
11858 color shifting, or linked together as a single channel, which prevents
11859 color shifting. Linked normalization preserves hue. Independent normalization
11860 does not, so it can be used to remove some color casts. Independent and linked
11861 normalization can be combined in any ratio.
11863 The normalize filter accepts the following options:
11868 Colors which define the output range. The minimum input value is mapped to
11869 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11870 The defaults are black and white respectively. Specifying white for
11871 @var{blackpt} and black for @var{whitept} will give color-inverted,
11872 normalized video. Shades of grey can be used to reduce the dynamic range
11873 (contrast). Specifying saturated colors here can create some interesting
11877 The number of previous frames to use for temporal smoothing. The input range
11878 of each channel is smoothed using a rolling average over the current frame
11879 and the @var{smoothing} previous frames. The default is 0 (no temporal
11883 Controls the ratio of independent (color shifting) channel normalization to
11884 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11885 independent. Defaults to 1.0 (fully independent).
11888 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11889 expensive no-op. Defaults to 1.0 (full strength).
11893 @subsection Examples
11895 Stretch video contrast to use the full dynamic range, with no temporal
11896 smoothing; may flicker depending on the source content:
11898 normalize=blackpt=black:whitept=white:smoothing=0
11901 As above, but with 50 frames of temporal smoothing; flicker should be
11902 reduced, depending on the source content:
11904 normalize=blackpt=black:whitept=white:smoothing=50
11907 As above, but with hue-preserving linked channel normalization:
11909 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11912 As above, but with half strength:
11914 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11917 Map the darkest input color to red, the brightest input color to cyan:
11919 normalize=blackpt=red:whitept=cyan
11924 Pass the video source unchanged to the output.
11927 Optical Character Recognition
11929 This filter uses Tesseract for optical character recognition. To enable
11930 compilation of this filter, you need to configure FFmpeg with
11931 @code{--enable-libtesseract}.
11933 It accepts the following options:
11937 Set datapath to tesseract data. Default is to use whatever was
11938 set at installation.
11941 Set language, default is "eng".
11944 Set character whitelist.
11947 Set character blacklist.
11950 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11954 Apply a video transform using libopencv.
11956 To enable this filter, install the libopencv library and headers and
11957 configure FFmpeg with @code{--enable-libopencv}.
11959 It accepts the following parameters:
11964 The name of the libopencv filter to apply.
11966 @item filter_params
11967 The parameters to pass to the libopencv filter. If not specified, the default
11968 values are assumed.
11972 Refer to the official libopencv documentation for more precise
11974 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11976 Several libopencv filters are supported; see the following subsections.
11981 Dilate an image by using a specific structuring element.
11982 It corresponds to the libopencv function @code{cvDilate}.
11984 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11986 @var{struct_el} represents a structuring element, and has the syntax:
11987 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11989 @var{cols} and @var{rows} represent the number of columns and rows of
11990 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11991 point, and @var{shape} the shape for the structuring element. @var{shape}
11992 must be "rect", "cross", "ellipse", or "custom".
11994 If the value for @var{shape} is "custom", it must be followed by a
11995 string of the form "=@var{filename}". The file with name
11996 @var{filename} is assumed to represent a binary image, with each
11997 printable character corresponding to a bright pixel. When a custom
11998 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11999 or columns and rows of the read file are assumed instead.
12001 The default value for @var{struct_el} is "3x3+0x0/rect".
12003 @var{nb_iterations} specifies the number of times the transform is
12004 applied to the image, and defaults to 1.
12008 # Use the default values
12011 # Dilate using a structuring element with a 5x5 cross, iterating two times
12012 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12014 # Read the shape from the file diamond.shape, iterating two times.
12015 # The file diamond.shape may contain a pattern of characters like this
12021 # The specified columns and rows are ignored
12022 # but the anchor point coordinates are not
12023 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12028 Erode an image by using a specific structuring element.
12029 It corresponds to the libopencv function @code{cvErode}.
12031 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12032 with the same syntax and semantics as the @ref{dilate} filter.
12036 Smooth the input video.
12038 The filter takes the following parameters:
12039 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12041 @var{type} is the type of smooth filter to apply, and must be one of
12042 the following values: "blur", "blur_no_scale", "median", "gaussian",
12043 or "bilateral". The default value is "gaussian".
12045 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12046 depend on the smooth type. @var{param1} and
12047 @var{param2} accept integer positive values or 0. @var{param3} and
12048 @var{param4} accept floating point values.
12050 The default value for @var{param1} is 3. The default value for the
12051 other parameters is 0.
12053 These parameters correspond to the parameters assigned to the
12054 libopencv function @code{cvSmooth}.
12056 @section oscilloscope
12058 2D Video Oscilloscope.
12060 Useful to measure spatial impulse, step responses, chroma delays, etc.
12062 It accepts the following parameters:
12066 Set scope center x position.
12069 Set scope center y position.
12072 Set scope size, relative to frame diagonal.
12075 Set scope tilt/rotation.
12081 Set trace center x position.
12084 Set trace center y position.
12087 Set trace width, relative to width of frame.
12090 Set trace height, relative to height of frame.
12093 Set which components to trace. By default it traces first three components.
12096 Draw trace grid. By default is enabled.
12099 Draw some statistics. By default is enabled.
12102 Draw scope. By default is enabled.
12105 @subsection Examples
12109 Inspect full first row of video frame.
12111 oscilloscope=x=0.5:y=0:s=1
12115 Inspect full last row of video frame.
12117 oscilloscope=x=0.5:y=1:s=1
12121 Inspect full 5th line of video frame of height 1080.
12123 oscilloscope=x=0.5:y=5/1080:s=1
12127 Inspect full last column of video frame.
12129 oscilloscope=x=1:y=0.5:s=1:t=1
12137 Overlay one video on top of another.
12139 It takes two inputs and has one output. The first input is the "main"
12140 video on which the second input is overlaid.
12142 It accepts the following parameters:
12144 A description of the accepted options follows.
12149 Set the expression for the x and y coordinates of the overlaid video
12150 on the main video. Default value is "0" for both expressions. In case
12151 the expression is invalid, it is set to a huge value (meaning that the
12152 overlay will not be displayed within the output visible area).
12155 See @ref{framesync}.
12158 Set when the expressions for @option{x}, and @option{y} are evaluated.
12160 It accepts the following values:
12163 only evaluate expressions once during the filter initialization or
12164 when a command is processed
12167 evaluate expressions for each incoming frame
12170 Default value is @samp{frame}.
12173 See @ref{framesync}.
12176 Set the format for the output video.
12178 It accepts the following values:
12181 force YUV420 output
12184 force YUV422 output
12187 force YUV444 output
12190 force packed RGB output
12193 force planar RGB output
12196 automatically pick format
12199 Default value is @samp{yuv420}.
12202 See @ref{framesync}.
12205 Set format of alpha of the overlaid video, it can be @var{straight} or
12206 @var{premultiplied}. Default is @var{straight}.
12209 The @option{x}, and @option{y} expressions can contain the following
12215 The main input width and height.
12219 The overlay input width and height.
12223 The computed values for @var{x} and @var{y}. They are evaluated for
12228 horizontal and vertical chroma subsample values of the output
12229 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12233 the number of input frame, starting from 0
12236 the position in the file of the input frame, NAN if unknown
12239 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12243 This filter also supports the @ref{framesync} options.
12245 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12246 when evaluation is done @emph{per frame}, and will evaluate to NAN
12247 when @option{eval} is set to @samp{init}.
12249 Be aware that frames are taken from each input video in timestamp
12250 order, hence, if their initial timestamps differ, it is a good idea
12251 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12252 have them begin in the same zero timestamp, as the example for
12253 the @var{movie} filter does.
12255 You can chain together more overlays but you should test the
12256 efficiency of such approach.
12258 @subsection Commands
12260 This filter supports the following commands:
12264 Modify the x and y of the overlay input.
12265 The command accepts the same syntax of the corresponding option.
12267 If the specified expression is not valid, it is kept at its current
12271 @subsection Examples
12275 Draw the overlay at 10 pixels from the bottom right corner of the main
12278 overlay=main_w-overlay_w-10:main_h-overlay_h-10
12281 Using named options the example above becomes:
12283 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
12287 Insert a transparent PNG logo in the bottom left corner of the input,
12288 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12290 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12294 Insert 2 different transparent PNG logos (second logo on bottom
12295 right corner) using the @command{ffmpeg} tool:
12297 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
12301 Add a transparent color layer on top of the main video; @code{WxH}
12302 must specify the size of the main input to the overlay filter:
12304 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
12308 Play an original video and a filtered version (here with the deshake
12309 filter) side by side using the @command{ffplay} tool:
12311 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
12314 The above command is the same as:
12316 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
12320 Make a sliding overlay appearing from the left to the right top part of the
12321 screen starting since time 2:
12323 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
12327 Compose output by putting two input videos side to side:
12329 ffmpeg -i left.avi -i right.avi -filter_complex "
12330 nullsrc=size=200x100 [background];
12331 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
12332 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
12333 [background][left] overlay=shortest=1 [background+left];
12334 [background+left][right] overlay=shortest=1:x=100 [left+right]
12339 Mask 10-20 seconds of a video by applying the delogo filter to a section
12341 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
12342 -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]'
12347 Chain several overlays in cascade:
12349 nullsrc=s=200x200 [bg];
12350 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
12351 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
12352 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
12353 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
12354 [in3] null, [mid2] overlay=100:100 [out0]
12361 Apply Overcomplete Wavelet denoiser.
12363 The filter accepts the following options:
12369 Larger depth values will denoise lower frequency components more, but
12370 slow down filtering.
12372 Must be an int in the range 8-16, default is @code{8}.
12374 @item luma_strength, ls
12377 Must be a double value in the range 0-1000, default is @code{1.0}.
12379 @item chroma_strength, cs
12380 Set chroma strength.
12382 Must be a double value in the range 0-1000, default is @code{1.0}.
12388 Add paddings to the input image, and place the original input at the
12389 provided @var{x}, @var{y} coordinates.
12391 It accepts the following parameters:
12396 Specify an expression for the size of the output image with the
12397 paddings added. If the value for @var{width} or @var{height} is 0, the
12398 corresponding input size is used for the output.
12400 The @var{width} expression can reference the value set by the
12401 @var{height} expression, and vice versa.
12403 The default value of @var{width} and @var{height} is 0.
12407 Specify the offsets to place the input image at within the padded area,
12408 with respect to the top/left border of the output image.
12410 The @var{x} expression can reference the value set by the @var{y}
12411 expression, and vice versa.
12413 The default value of @var{x} and @var{y} is 0.
12415 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
12416 so the input image is centered on the padded area.
12419 Specify the color of the padded area. For the syntax of this option,
12420 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
12421 manual,ffmpeg-utils}.
12423 The default value of @var{color} is "black".
12426 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
12428 It accepts the following values:
12432 Only evaluate expressions once during the filter initialization or when
12433 a command is processed.
12436 Evaluate expressions for each incoming frame.
12440 Default value is @samp{init}.
12443 Pad to aspect instead to a resolution.
12447 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
12448 options are expressions containing the following constants:
12453 The input video width and height.
12457 These are the same as @var{in_w} and @var{in_h}.
12461 The output width and height (the size of the padded area), as
12462 specified by the @var{width} and @var{height} expressions.
12466 These are the same as @var{out_w} and @var{out_h}.
12470 The x and y offsets as specified by the @var{x} and @var{y}
12471 expressions, or NAN if not yet specified.
12474 same as @var{iw} / @var{ih}
12477 input sample aspect ratio
12480 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
12484 The horizontal and vertical chroma subsample values. For example for the
12485 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12488 @subsection Examples
12492 Add paddings with the color "violet" to the input video. The output video
12493 size is 640x480, and the top-left corner of the input video is placed at
12496 pad=640:480:0:40:violet
12499 The example above is equivalent to the following command:
12501 pad=width=640:height=480:x=0:y=40:color=violet
12505 Pad the input to get an output with dimensions increased by 3/2,
12506 and put the input video at the center of the padded area:
12508 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12512 Pad the input to get a squared output with size equal to the maximum
12513 value between the input width and height, and put the input video at
12514 the center of the padded area:
12516 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12520 Pad the input to get a final w/h ratio of 16:9:
12522 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12526 In case of anamorphic video, in order to set the output display aspect
12527 correctly, it is necessary to use @var{sar} in the expression,
12528 according to the relation:
12530 (ih * X / ih) * sar = output_dar
12531 X = output_dar / sar
12534 Thus the previous example needs to be modified to:
12536 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12540 Double the output size and put the input video in the bottom-right
12541 corner of the output padded area:
12543 pad="2*iw:2*ih:ow-iw:oh-ih"
12547 @anchor{palettegen}
12548 @section palettegen
12550 Generate one palette for a whole video stream.
12552 It accepts the following options:
12556 Set the maximum number of colors to quantize in the palette.
12557 Note: the palette will still contain 256 colors; the unused palette entries
12560 @item reserve_transparent
12561 Create a palette of 255 colors maximum and reserve the last one for
12562 transparency. Reserving the transparency color is useful for GIF optimization.
12563 If not set, the maximum of colors in the palette will be 256. You probably want
12564 to disable this option for a standalone image.
12567 @item transparency_color
12568 Set the color that will be used as background for transparency.
12571 Set statistics mode.
12573 It accepts the following values:
12576 Compute full frame histograms.
12578 Compute histograms only for the part that differs from previous frame. This
12579 might be relevant to give more importance to the moving part of your input if
12580 the background is static.
12582 Compute new histogram for each frame.
12585 Default value is @var{full}.
12588 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12589 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12590 color quantization of the palette. This information is also visible at
12591 @var{info} logging level.
12593 @subsection Examples
12597 Generate a representative palette of a given video using @command{ffmpeg}:
12599 ffmpeg -i input.mkv -vf palettegen palette.png
12603 @section paletteuse
12605 Use a palette to downsample an input video stream.
12607 The filter takes two inputs: one video stream and a palette. The palette must
12608 be a 256 pixels image.
12610 It accepts the following options:
12614 Select dithering mode. Available algorithms are:
12617 Ordered 8x8 bayer dithering (deterministic)
12619 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12620 Note: this dithering is sometimes considered "wrong" and is included as a
12622 @item floyd_steinberg
12623 Floyd and Steingberg dithering (error diffusion)
12625 Frankie Sierra dithering v2 (error diffusion)
12627 Frankie Sierra dithering v2 "Lite" (error diffusion)
12630 Default is @var{sierra2_4a}.
12633 When @var{bayer} dithering is selected, this option defines the scale of the
12634 pattern (how much the crosshatch pattern is visible). A low value means more
12635 visible pattern for less banding, and higher value means less visible pattern
12636 at the cost of more banding.
12638 The option must be an integer value in the range [0,5]. Default is @var{2}.
12641 If set, define the zone to process
12645 Only the changing rectangle will be reprocessed. This is similar to GIF
12646 cropping/offsetting compression mechanism. This option can be useful for speed
12647 if only a part of the image is changing, and has use cases such as limiting the
12648 scope of the error diffusal @option{dither} to the rectangle that bounds the
12649 moving scene (it leads to more deterministic output if the scene doesn't change
12650 much, and as a result less moving noise and better GIF compression).
12653 Default is @var{none}.
12656 Take new palette for each output frame.
12658 @item alpha_threshold
12659 Sets the alpha threshold for transparency. Alpha values above this threshold
12660 will be treated as completely opaque, and values below this threshold will be
12661 treated as completely transparent.
12663 The option must be an integer value in the range [0,255]. Default is @var{128}.
12666 @subsection Examples
12670 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12671 using @command{ffmpeg}:
12673 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12677 @section perspective
12679 Correct perspective of video not recorded perpendicular to the screen.
12681 A description of the accepted parameters follows.
12692 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12693 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12694 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12695 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12696 then the corners of the source will be sent to the specified coordinates.
12698 The expressions can use the following variables:
12703 the width and height of video frame.
12707 Output frame count.
12710 @item interpolation
12711 Set interpolation for perspective correction.
12713 It accepts the following values:
12719 Default value is @samp{linear}.
12722 Set interpretation of coordinate options.
12724 It accepts the following values:
12728 Send point in the source specified by the given coordinates to
12729 the corners of the destination.
12731 @item 1, destination
12733 Send the corners of the source to the point in the destination specified
12734 by the given coordinates.
12736 Default value is @samp{source}.
12740 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12742 It accepts the following values:
12745 only evaluate expressions once during the filter initialization or
12746 when a command is processed
12749 evaluate expressions for each incoming frame
12752 Default value is @samp{init}.
12757 Delay interlaced video by one field time so that the field order changes.
12759 The intended use is to fix PAL movies that have been captured with the
12760 opposite field order to the film-to-video transfer.
12762 A description of the accepted parameters follows.
12768 It accepts the following values:
12771 Capture field order top-first, transfer bottom-first.
12772 Filter will delay the bottom field.
12775 Capture field order bottom-first, transfer top-first.
12776 Filter will delay the top field.
12779 Capture and transfer with the same field order. This mode only exists
12780 for the documentation of the other options to refer to, but if you
12781 actually select it, the filter will faithfully do nothing.
12784 Capture field order determined automatically by field flags, transfer
12786 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12787 basis using field flags. If no field information is available,
12788 then this works just like @samp{u}.
12791 Capture unknown or varying, transfer opposite.
12792 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12793 analyzing the images and selecting the alternative that produces best
12794 match between the fields.
12797 Capture top-first, transfer unknown or varying.
12798 Filter selects among @samp{t} and @samp{p} using image analysis.
12801 Capture bottom-first, transfer unknown or varying.
12802 Filter selects among @samp{b} and @samp{p} using image analysis.
12805 Capture determined by field flags, transfer unknown or varying.
12806 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12807 image analysis. If no field information is available, then this works just
12808 like @samp{U}. This is the default mode.
12811 Both capture and transfer unknown or varying.
12812 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12816 @section pixdesctest
12818 Pixel format descriptor test filter, mainly useful for internal
12819 testing. The output video should be equal to the input video.
12823 format=monow, pixdesctest
12826 can be used to test the monowhite pixel format descriptor definition.
12830 Display sample values of color channels. Mainly useful for checking color
12831 and levels. Minimum supported resolution is 640x480.
12833 The filters accept the following options:
12837 Set scope X position, relative offset on X axis.
12840 Set scope Y position, relative offset on Y axis.
12849 Set window opacity. This window also holds statistics about pixel area.
12852 Set window X position, relative offset on X axis.
12855 Set window Y position, relative offset on Y axis.
12860 Enable the specified chain of postprocessing subfilters using libpostproc. This
12861 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12862 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12863 Each subfilter and some options have a short and a long name that can be used
12864 interchangeably, i.e. dr/dering are the same.
12866 The filters accept the following options:
12870 Set postprocessing subfilters string.
12873 All subfilters share common options to determine their scope:
12877 Honor the quality commands for this subfilter.
12880 Do chrominance filtering, too (default).
12883 Do luminance filtering only (no chrominance).
12886 Do chrominance filtering only (no luminance).
12889 These options can be appended after the subfilter name, separated by a '|'.
12891 Available subfilters are:
12894 @item hb/hdeblock[|difference[|flatness]]
12895 Horizontal deblocking filter
12898 Difference factor where higher values mean more deblocking (default: @code{32}).
12900 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12903 @item vb/vdeblock[|difference[|flatness]]
12904 Vertical deblocking filter
12907 Difference factor where higher values mean more deblocking (default: @code{32}).
12909 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12912 @item ha/hadeblock[|difference[|flatness]]
12913 Accurate horizontal deblocking filter
12916 Difference factor where higher values mean more deblocking (default: @code{32}).
12918 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12921 @item va/vadeblock[|difference[|flatness]]
12922 Accurate vertical deblocking filter
12925 Difference factor where higher values mean more deblocking (default: @code{32}).
12927 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12931 The horizontal and vertical deblocking filters share the difference and
12932 flatness values so you cannot set different horizontal and vertical
12936 @item h1/x1hdeblock
12937 Experimental horizontal deblocking filter
12939 @item v1/x1vdeblock
12940 Experimental vertical deblocking filter
12945 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12948 larger -> stronger filtering
12950 larger -> stronger filtering
12952 larger -> stronger filtering
12955 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12958 Stretch luminance to @code{0-255}.
12961 @item lb/linblenddeint
12962 Linear blend deinterlacing filter that deinterlaces the given block by
12963 filtering all lines with a @code{(1 2 1)} filter.
12965 @item li/linipoldeint
12966 Linear interpolating deinterlacing filter that deinterlaces the given block by
12967 linearly interpolating every second line.
12969 @item ci/cubicipoldeint
12970 Cubic interpolating deinterlacing filter deinterlaces the given block by
12971 cubically interpolating every second line.
12973 @item md/mediandeint
12974 Median deinterlacing filter that deinterlaces the given block by applying a
12975 median filter to every second line.
12977 @item fd/ffmpegdeint
12978 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12979 second line with a @code{(-1 4 2 4 -1)} filter.
12982 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12983 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12985 @item fq/forceQuant[|quantizer]
12986 Overrides the quantizer table from the input with the constant quantizer you
12994 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12997 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13000 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13003 @subsection Examples
13007 Apply horizontal and vertical deblocking, deringing and automatic
13008 brightness/contrast:
13014 Apply default filters without brightness/contrast correction:
13020 Apply default filters and temporal denoiser:
13022 pp=default/tmpnoise|1|2|3
13026 Apply deblocking on luminance only, and switch vertical deblocking on or off
13027 automatically depending on available CPU time:
13034 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13035 similar to spp = 6 with 7 point DCT, where only the center sample is
13038 The filter accepts the following options:
13042 Force a constant quantization parameter. It accepts an integer in range
13043 0 to 63. If not set, the filter will use the QP from the video stream
13047 Set thresholding mode. Available modes are:
13051 Set hard thresholding.
13053 Set soft thresholding (better de-ringing effect, but likely blurrier).
13055 Set medium thresholding (good results, default).
13059 @section premultiply
13060 Apply alpha premultiply effect to input video stream using first plane
13061 of second stream as alpha.
13063 Both streams must have same dimensions and same pixel format.
13065 The filter accepts the following option:
13069 Set which planes will be processed, unprocessed planes will be copied.
13070 By default value 0xf, all planes will be processed.
13073 Do not require 2nd input for processing, instead use alpha plane from input stream.
13077 Apply prewitt operator to input video stream.
13079 The filter accepts the following option:
13083 Set which planes will be processed, unprocessed planes will be copied.
13084 By default value 0xf, all planes will be processed.
13087 Set value which will be multiplied with filtered result.
13090 Set value which will be added to filtered result.
13093 @anchor{program_opencl}
13094 @section program_opencl
13096 Filter video using an OpenCL program.
13101 OpenCL program source file.
13104 Kernel name in program.
13107 Number of inputs to the filter. Defaults to 1.
13110 Size of output frames. Defaults to the same as the first input.
13114 The program source file must contain a kernel function with the given name,
13115 which will be run once for each plane of the output. Each run on a plane
13116 gets enqueued as a separate 2D global NDRange with one work-item for each
13117 pixel to be generated. The global ID offset for each work-item is therefore
13118 the coordinates of a pixel in the destination image.
13120 The kernel function needs to take the following arguments:
13123 Destination image, @var{__write_only image2d_t}.
13125 This image will become the output; the kernel should write all of it.
13127 Frame index, @var{unsigned int}.
13129 This is a counter starting from zero and increasing by one for each frame.
13131 Source images, @var{__read_only image2d_t}.
13133 These are the most recent images on each input. The kernel may read from
13134 them to generate the output, but they can't be written to.
13141 Copy the input to the output (output must be the same size as the input).
13143 __kernel void copy(__write_only image2d_t destination,
13144 unsigned int index,
13145 __read_only image2d_t source)
13147 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13149 int2 location = (int2)(get_global_id(0), get_global_id(1));
13151 float4 value = read_imagef(source, sampler, location);
13153 write_imagef(destination, location, value);
13158 Apply a simple transformation, rotating the input by an amount increasing
13159 with the index counter. Pixel values are linearly interpolated by the
13160 sampler, and the output need not have the same dimensions as the input.
13162 __kernel void rotate_image(__write_only image2d_t dst,
13163 unsigned int index,
13164 __read_only image2d_t src)
13166 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13167 CLK_FILTER_LINEAR);
13169 float angle = (float)index / 100.0f;
13171 float2 dst_dim = convert_float2(get_image_dim(dst));
13172 float2 src_dim = convert_float2(get_image_dim(src));
13174 float2 dst_cen = dst_dim / 2.0f;
13175 float2 src_cen = src_dim / 2.0f;
13177 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13179 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13181 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13182 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13184 src_pos = src_pos * src_dim / dst_dim;
13186 float2 src_loc = src_pos + src_cen;
13188 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13189 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13190 write_imagef(dst, dst_loc, 0.5f);
13192 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13197 Blend two inputs together, with the amount of each input used varying
13198 with the index counter.
13200 __kernel void blend_images(__write_only image2d_t dst,
13201 unsigned int index,
13202 __read_only image2d_t src1,
13203 __read_only image2d_t src2)
13205 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13206 CLK_FILTER_LINEAR);
13208 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13210 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13211 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13212 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13214 float4 val1 = read_imagef(src1, sampler, src1_loc);
13215 float4 val2 = read_imagef(src2, sampler, src2_loc);
13217 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13223 @section pseudocolor
13225 Alter frame colors in video with pseudocolors.
13227 This filter accept the following options:
13231 set pixel first component expression
13234 set pixel second component expression
13237 set pixel third component expression
13240 set pixel fourth component expression, corresponds to the alpha component
13243 set component to use as base for altering colors
13246 Each of them specifies the expression to use for computing the lookup table for
13247 the corresponding pixel component values.
13249 The expressions can contain the following constants and functions:
13254 The input width and height.
13257 The input value for the pixel component.
13259 @item ymin, umin, vmin, amin
13260 The minimum allowed component value.
13262 @item ymax, umax, vmax, amax
13263 The maximum allowed component value.
13266 All expressions default to "val".
13268 @subsection Examples
13272 Change too high luma values to gradient:
13274 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'"
13280 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
13281 Ratio) between two input videos.
13283 This filter takes in input two input videos, the first input is
13284 considered the "main" source and is passed unchanged to the
13285 output. The second input is used as a "reference" video for computing
13288 Both video inputs must have the same resolution and pixel format for
13289 this filter to work correctly. Also it assumes that both inputs
13290 have the same number of frames, which are compared one by one.
13292 The obtained average PSNR is printed through the logging system.
13294 The filter stores the accumulated MSE (mean squared error) of each
13295 frame, and at the end of the processing it is averaged across all frames
13296 equally, and the following formula is applied to obtain the PSNR:
13299 PSNR = 10*log10(MAX^2/MSE)
13302 Where MAX is the average of the maximum values of each component of the
13305 The description of the accepted parameters follows.
13308 @item stats_file, f
13309 If specified the filter will use the named file to save the PSNR of
13310 each individual frame. When filename equals "-" the data is sent to
13313 @item stats_version
13314 Specifies which version of the stats file format to use. Details of
13315 each format are written below.
13316 Default value is 1.
13318 @item stats_add_max
13319 Determines whether the max value is output to the stats log.
13320 Default value is 0.
13321 Requires stats_version >= 2. If this is set and stats_version < 2,
13322 the filter will return an error.
13325 This filter also supports the @ref{framesync} options.
13327 The file printed if @var{stats_file} is selected, contains a sequence of
13328 key/value pairs of the form @var{key}:@var{value} for each compared
13331 If a @var{stats_version} greater than 1 is specified, a header line precedes
13332 the list of per-frame-pair stats, with key value pairs following the frame
13333 format with the following parameters:
13336 @item psnr_log_version
13337 The version of the log file format. Will match @var{stats_version}.
13340 A comma separated list of the per-frame-pair parameters included in
13344 A description of each shown per-frame-pair parameter follows:
13348 sequential number of the input frame, starting from 1
13351 Mean Square Error pixel-by-pixel average difference of the compared
13352 frames, averaged over all the image components.
13354 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
13355 Mean Square Error pixel-by-pixel average difference of the compared
13356 frames for the component specified by the suffix.
13358 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
13359 Peak Signal to Noise ratio of the compared frames for the component
13360 specified by the suffix.
13362 @item max_avg, max_y, max_u, max_v
13363 Maximum allowed value for each channel, and average over all
13369 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13370 [main][ref] psnr="stats_file=stats.log" [out]
13373 On this example the input file being processed is compared with the
13374 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
13375 is stored in @file{stats.log}.
13380 Pulldown reversal (inverse telecine) filter, capable of handling mixed
13381 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
13384 The pullup filter is designed to take advantage of future context in making
13385 its decisions. This filter is stateless in the sense that it does not lock
13386 onto a pattern to follow, but it instead looks forward to the following
13387 fields in order to identify matches and rebuild progressive frames.
13389 To produce content with an even framerate, insert the fps filter after
13390 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
13391 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
13393 The filter accepts the following options:
13400 These options set the amount of "junk" to ignore at the left, right, top, and
13401 bottom of the image, respectively. Left and right are in units of 8 pixels,
13402 while top and bottom are in units of 2 lines.
13403 The default is 8 pixels on each side.
13406 Set the strict breaks. Setting this option to 1 will reduce the chances of
13407 filter generating an occasional mismatched frame, but it may also cause an
13408 excessive number of frames to be dropped during high motion sequences.
13409 Conversely, setting it to -1 will make filter match fields more easily.
13410 This may help processing of video where there is slight blurring between
13411 the fields, but may also cause there to be interlaced frames in the output.
13412 Default value is @code{0}.
13415 Set the metric plane to use. It accepts the following values:
13421 Use chroma blue plane.
13424 Use chroma red plane.
13427 This option may be set to use chroma plane instead of the default luma plane
13428 for doing filter's computations. This may improve accuracy on very clean
13429 source material, but more likely will decrease accuracy, especially if there
13430 is chroma noise (rainbow effect) or any grayscale video.
13431 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
13432 load and make pullup usable in realtime on slow machines.
13435 For best results (without duplicated frames in the output file) it is
13436 necessary to change the output frame rate. For example, to inverse
13437 telecine NTSC input:
13439 ffmpeg -i input -vf pullup -r 24000/1001 ...
13444 Change video quantization parameters (QP).
13446 The filter accepts the following option:
13450 Set expression for quantization parameter.
13453 The expression is evaluated through the eval API and can contain, among others,
13454 the following constants:
13458 1 if index is not 129, 0 otherwise.
13461 Sequential index starting from -129 to 128.
13464 @subsection Examples
13468 Some equation like:
13476 Flush video frames from internal cache of frames into a random order.
13477 No frame is discarded.
13478 Inspired by @ref{frei0r} nervous filter.
13482 Set size in number of frames of internal cache, in range from @code{2} to
13483 @code{512}. Default is @code{30}.
13486 Set seed for random number generator, must be an integer included between
13487 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
13488 less than @code{0}, the filter will try to use a good random seed on a
13492 @section readeia608
13494 Read closed captioning (EIA-608) information from the top lines of a video frame.
13496 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13497 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13498 with EIA-608 data (starting from 0). A description of each metadata value follows:
13501 @item lavfi.readeia608.X.cc
13502 The two bytes stored as EIA-608 data (printed in hexadecimal).
13504 @item lavfi.readeia608.X.line
13505 The number of the line on which the EIA-608 data was identified and read.
13508 This filter accepts the following options:
13512 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13515 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13518 Set minimal acceptable amplitude change for sync codes detection.
13519 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13522 Set the ratio of width reserved for sync code detection.
13523 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13526 Set the max peaks height difference for sync code detection.
13527 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13530 Set max peaks period difference for sync code detection.
13531 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13534 Set the first two max start code bits differences.
13535 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13538 Set the minimum ratio of bits height compared to 3rd start code bit.
13539 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13542 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13545 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13548 Enable checking the parity bit. In the event of a parity error, the filter will output
13549 @code{0x00} for that character. Default is false.
13552 @subsection Examples
13556 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13558 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
13564 Read vertical interval timecode (VITC) information from the top lines of a
13567 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13568 timecode value, if a valid timecode has been detected. Further metadata key
13569 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13570 timecode data has been found or not.
13572 This filter accepts the following options:
13576 Set the maximum number of lines to scan for VITC data. If the value is set to
13577 @code{-1} the full video frame is scanned. Default is @code{45}.
13580 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13581 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13584 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13585 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13588 @subsection Examples
13592 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13593 draw @code{--:--:--:--} as a placeholder:
13595 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13601 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13603 Destination pixel at position (X, Y) will be picked from source (x, y) position
13604 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13605 value for pixel will be used for destination pixel.
13607 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13608 will have Xmap/Ymap video stream dimensions.
13609 Xmap and Ymap input video streams are 16bit depth, single channel.
13611 @section removegrain
13613 The removegrain filter is a spatial denoiser for progressive video.
13617 Set mode for the first plane.
13620 Set mode for the second plane.
13623 Set mode for the third plane.
13626 Set mode for the fourth plane.
13629 Range of mode is from 0 to 24. Description of each mode follows:
13633 Leave input plane unchanged. Default.
13636 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13639 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13642 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13645 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13646 This is equivalent to a median filter.
13649 Line-sensitive clipping giving the minimal change.
13652 Line-sensitive clipping, intermediate.
13655 Line-sensitive clipping, intermediate.
13658 Line-sensitive clipping, intermediate.
13661 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13664 Replaces the target pixel with the closest neighbour.
13667 [1 2 1] horizontal and vertical kernel blur.
13673 Bob mode, interpolates top field from the line where the neighbours
13674 pixels are the closest.
13677 Bob mode, interpolates bottom field from the line where the neighbours
13678 pixels are the closest.
13681 Bob mode, interpolates top field. Same as 13 but with a more complicated
13682 interpolation formula.
13685 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13686 interpolation formula.
13689 Clips the pixel with the minimum and maximum of respectively the maximum and
13690 minimum of each pair of opposite neighbour pixels.
13693 Line-sensitive clipping using opposite neighbours whose greatest distance from
13694 the current pixel is minimal.
13697 Replaces the pixel with the average of its 8 neighbours.
13700 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13703 Clips pixels using the averages of opposite neighbour.
13706 Same as mode 21 but simpler and faster.
13709 Small edge and halo removal, but reputed useless.
13715 @section removelogo
13717 Suppress a TV station logo, using an image file to determine which
13718 pixels comprise the logo. It works by filling in the pixels that
13719 comprise the logo with neighboring pixels.
13721 The filter accepts the following options:
13725 Set the filter bitmap file, which can be any image format supported by
13726 libavformat. The width and height of the image file must match those of the
13727 video stream being processed.
13730 Pixels in the provided bitmap image with a value of zero are not
13731 considered part of the logo, non-zero pixels are considered part of
13732 the logo. If you use white (255) for the logo and black (0) for the
13733 rest, you will be safe. For making the filter bitmap, it is
13734 recommended to take a screen capture of a black frame with the logo
13735 visible, and then using a threshold filter followed by the erode
13736 filter once or twice.
13738 If needed, little splotches can be fixed manually. Remember that if
13739 logo pixels are not covered, the filter quality will be much
13740 reduced. Marking too many pixels as part of the logo does not hurt as
13741 much, but it will increase the amount of blurring needed to cover over
13742 the image and will destroy more information than necessary, and extra
13743 pixels will slow things down on a large logo.
13745 @section repeatfields
13747 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13748 fields based on its value.
13752 Reverse a video clip.
13754 Warning: This filter requires memory to buffer the entire clip, so trimming
13757 @subsection Examples
13761 Take the first 5 seconds of a clip, and reverse it.
13768 Apply roberts cross operator to input video stream.
13770 The filter accepts the following option:
13774 Set which planes will be processed, unprocessed planes will be copied.
13775 By default value 0xf, all planes will be processed.
13778 Set value which will be multiplied with filtered result.
13781 Set value which will be added to filtered result.
13786 Rotate video by an arbitrary angle expressed in radians.
13788 The filter accepts the following options:
13790 A description of the optional parameters follows.
13793 Set an expression for the angle by which to rotate the input video
13794 clockwise, expressed as a number of radians. A negative value will
13795 result in a counter-clockwise rotation. By default it is set to "0".
13797 This expression is evaluated for each frame.
13800 Set the output width expression, default value is "iw".
13801 This expression is evaluated just once during configuration.
13804 Set the output height expression, default value is "ih".
13805 This expression is evaluated just once during configuration.
13808 Enable bilinear interpolation if set to 1, a value of 0 disables
13809 it. Default value is 1.
13812 Set the color used to fill the output area not covered by the rotated
13813 image. For the general syntax of this option, check the
13814 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13815 If the special value "none" is selected then no
13816 background is printed (useful for example if the background is never shown).
13818 Default value is "black".
13821 The expressions for the angle and the output size can contain the
13822 following constants and functions:
13826 sequential number of the input frame, starting from 0. It is always NAN
13827 before the first frame is filtered.
13830 time in seconds of the input frame, it is set to 0 when the filter is
13831 configured. It is always NAN before the first frame is filtered.
13835 horizontal and vertical chroma subsample values. For example for the
13836 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13840 the input video width and height
13844 the output width and height, that is the size of the padded area as
13845 specified by the @var{width} and @var{height} expressions
13849 the minimal width/height required for completely containing the input
13850 video rotated by @var{a} radians.
13852 These are only available when computing the @option{out_w} and
13853 @option{out_h} expressions.
13856 @subsection Examples
13860 Rotate the input by PI/6 radians clockwise:
13866 Rotate the input by PI/6 radians counter-clockwise:
13872 Rotate the input by 45 degrees clockwise:
13878 Apply a constant rotation with period T, starting from an angle of PI/3:
13880 rotate=PI/3+2*PI*t/T
13884 Make the input video rotation oscillating with a period of T
13885 seconds and an amplitude of A radians:
13887 rotate=A*sin(2*PI/T*t)
13891 Rotate the video, output size is chosen so that the whole rotating
13892 input video is always completely contained in the output:
13894 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13898 Rotate the video, reduce the output size so that no background is ever
13901 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13905 @subsection Commands
13907 The filter supports the following commands:
13911 Set the angle expression.
13912 The command accepts the same syntax of the corresponding option.
13914 If the specified expression is not valid, it is kept at its current
13920 Apply Shape Adaptive Blur.
13922 The filter accepts the following options:
13925 @item luma_radius, lr
13926 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13927 value is 1.0. A greater value will result in a more blurred image, and
13928 in slower processing.
13930 @item luma_pre_filter_radius, lpfr
13931 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13934 @item luma_strength, ls
13935 Set luma maximum difference between pixels to still be considered, must
13936 be a value in the 0.1-100.0 range, default value is 1.0.
13938 @item chroma_radius, cr
13939 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13940 greater value will result in a more blurred image, and in slower
13943 @item chroma_pre_filter_radius, cpfr
13944 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13946 @item chroma_strength, cs
13947 Set chroma maximum difference between pixels to still be considered,
13948 must be a value in the -0.9-100.0 range.
13951 Each chroma option value, if not explicitly specified, is set to the
13952 corresponding luma option value.
13957 Scale (resize) the input video, using the libswscale library.
13959 The scale filter forces the output display aspect ratio to be the same
13960 of the input, by changing the output sample aspect ratio.
13962 If the input image format is different from the format requested by
13963 the next filter, the scale filter will convert the input to the
13966 @subsection Options
13967 The filter accepts the following options, or any of the options
13968 supported by the libswscale scaler.
13970 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13971 the complete list of scaler options.
13976 Set the output video dimension expression. Default value is the input
13979 If the @var{width} or @var{w} value is 0, the input width is used for
13980 the output. If the @var{height} or @var{h} value is 0, the input height
13981 is used for the output.
13983 If one and only one of the values is -n with n >= 1, the scale filter
13984 will use a value that maintains the aspect ratio of the input image,
13985 calculated from the other specified dimension. After that it will,
13986 however, make sure that the calculated dimension is divisible by n and
13987 adjust the value if necessary.
13989 If both values are -n with n >= 1, the behavior will be identical to
13990 both values being set to 0 as previously detailed.
13992 See below for the list of accepted constants for use in the dimension
13996 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14000 Only evaluate expressions once during the filter initialization or when a command is processed.
14003 Evaluate expressions for each incoming frame.
14007 Default value is @samp{init}.
14011 Set the interlacing mode. It accepts the following values:
14015 Force interlaced aware scaling.
14018 Do not apply interlaced scaling.
14021 Select interlaced aware scaling depending on whether the source frames
14022 are flagged as interlaced or not.
14025 Default value is @samp{0}.
14028 Set libswscale scaling flags. See
14029 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14030 complete list of values. If not explicitly specified the filter applies
14034 @item param0, param1
14035 Set libswscale input parameters for scaling algorithms that need them. See
14036 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14037 complete documentation. If not explicitly specified the filter applies
14043 Set the video size. For the syntax of this option, check the
14044 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14046 @item in_color_matrix
14047 @item out_color_matrix
14048 Set in/output YCbCr color space type.
14050 This allows the autodetected value to be overridden as well as allows forcing
14051 a specific value used for the output and encoder.
14053 If not specified, the color space type depends on the pixel format.
14059 Choose automatically.
14062 Format conforming to International Telecommunication Union (ITU)
14063 Recommendation BT.709.
14066 Set color space conforming to the United States Federal Communications
14067 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14070 Set color space conforming to:
14074 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14077 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14080 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14085 Set color space conforming to SMPTE ST 240:1999.
14090 Set in/output YCbCr sample range.
14092 This allows the autodetected value to be overridden as well as allows forcing
14093 a specific value used for the output and encoder. If not specified, the
14094 range depends on the pixel format. Possible values:
14098 Choose automatically.
14101 Set full range (0-255 in case of 8-bit luma).
14103 @item mpeg/limited/tv
14104 Set "MPEG" range (16-235 in case of 8-bit luma).
14107 @item force_original_aspect_ratio
14108 Enable decreasing or increasing output video width or height if necessary to
14109 keep the original aspect ratio. Possible values:
14113 Scale the video as specified and disable this feature.
14116 The output video dimensions will automatically be decreased if needed.
14119 The output video dimensions will automatically be increased if needed.
14123 One useful instance of this option is that when you know a specific device's
14124 maximum allowed resolution, you can use this to limit the output video to
14125 that, while retaining the aspect ratio. For example, device A allows
14126 1280x720 playback, and your video is 1920x800. Using this option (set it to
14127 decrease) and specifying 1280x720 to the command line makes the output
14130 Please note that this is a different thing than specifying -1 for @option{w}
14131 or @option{h}, you still need to specify the output resolution for this option
14136 The values of the @option{w} and @option{h} options are expressions
14137 containing the following constants:
14142 The input width and height
14146 These are the same as @var{in_w} and @var{in_h}.
14150 The output (scaled) width and height
14154 These are the same as @var{out_w} and @var{out_h}
14157 The same as @var{iw} / @var{ih}
14160 input sample aspect ratio
14163 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14167 horizontal and vertical input chroma subsample values. For example for the
14168 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14172 horizontal and vertical output chroma subsample values. For example for the
14173 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14176 @subsection Examples
14180 Scale the input video to a size of 200x100
14185 This is equivalent to:
14196 Specify a size abbreviation for the output size:
14201 which can also be written as:
14207 Scale the input to 2x:
14209 scale=w=2*iw:h=2*ih
14213 The above is the same as:
14215 scale=2*in_w:2*in_h
14219 Scale the input to 2x with forced interlaced scaling:
14221 scale=2*iw:2*ih:interl=1
14225 Scale the input to half size:
14227 scale=w=iw/2:h=ih/2
14231 Increase the width, and set the height to the same size:
14237 Seek Greek harmony:
14244 Increase the height, and set the width to 3/2 of the height:
14246 scale=w=3/2*oh:h=3/5*ih
14250 Increase the size, making the size a multiple of the chroma
14253 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
14257 Increase the width to a maximum of 500 pixels,
14258 keeping the same aspect ratio as the input:
14260 scale=w='min(500\, iw*3/2):h=-1'
14264 Make pixels square by combining scale and setsar:
14266 scale='trunc(ih*dar):ih',setsar=1/1
14270 Make pixels square by combining scale and setsar,
14271 making sure the resulting resolution is even (required by some codecs):
14273 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
14277 @subsection Commands
14279 This filter supports the following commands:
14283 Set the output video dimension expression.
14284 The command accepts the same syntax of the corresponding option.
14286 If the specified expression is not valid, it is kept at its current
14292 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
14293 format conversion on CUDA video frames. Setting the output width and height
14294 works in the same way as for the @var{scale} filter.
14296 The following additional options are accepted:
14299 The pixel format of the output CUDA frames. If set to the string "same" (the
14300 default), the input format will be kept. Note that automatic format negotiation
14301 and conversion is not yet supported for hardware frames
14304 The interpolation algorithm used for resizing. One of the following:
14311 @item cubic2p_bspline
14312 2-parameter cubic (B=1, C=0)
14314 @item cubic2p_catmullrom
14315 2-parameter cubic (B=0, C=1/2)
14317 @item cubic2p_b05c03
14318 2-parameter cubic (B=1/2, C=3/10)
14330 Scale (resize) the input video, based on a reference video.
14332 See the scale filter for available options, scale2ref supports the same but
14333 uses the reference video instead of the main input as basis. scale2ref also
14334 supports the following additional constants for the @option{w} and
14335 @option{h} options:
14340 The main input video's width and height
14343 The same as @var{main_w} / @var{main_h}
14346 The main input video's sample aspect ratio
14348 @item main_dar, mdar
14349 The main input video's display aspect ratio. Calculated from
14350 @code{(main_w / main_h) * main_sar}.
14354 The main input video's horizontal and vertical chroma subsample values.
14355 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
14359 @subsection Examples
14363 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
14365 'scale2ref[b][a];[a][b]overlay'
14369 @anchor{selectivecolor}
14370 @section selectivecolor
14372 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
14373 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
14374 by the "purity" of the color (that is, how saturated it already is).
14376 This filter is similar to the Adobe Photoshop Selective Color tool.
14378 The filter accepts the following options:
14381 @item correction_method
14382 Select color correction method.
14384 Available values are:
14387 Specified adjustments are applied "as-is" (added/subtracted to original pixel
14390 Specified adjustments are relative to the original component value.
14392 Default is @code{absolute}.
14394 Adjustments for red pixels (pixels where the red component is the maximum)
14396 Adjustments for yellow pixels (pixels where the blue component is the minimum)
14398 Adjustments for green pixels (pixels where the green component is the maximum)
14400 Adjustments for cyan pixels (pixels where the red component is the minimum)
14402 Adjustments for blue pixels (pixels where the blue component is the maximum)
14404 Adjustments for magenta pixels (pixels where the green component is the minimum)
14406 Adjustments for white pixels (pixels where all components are greater than 128)
14408 Adjustments for all pixels except pure black and pure white
14410 Adjustments for black pixels (pixels where all components are lesser than 128)
14412 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
14415 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
14416 4 space separated floating point adjustment values in the [-1,1] range,
14417 respectively to adjust the amount of cyan, magenta, yellow and black for the
14418 pixels of its range.
14420 @subsection Examples
14424 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
14425 increase magenta by 27% in blue areas:
14427 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
14431 Use a Photoshop selective color preset:
14433 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
14437 @anchor{separatefields}
14438 @section separatefields
14440 The @code{separatefields} takes a frame-based video input and splits
14441 each frame into its components fields, producing a new half height clip
14442 with twice the frame rate and twice the frame count.
14444 This filter use field-dominance information in frame to decide which
14445 of each pair of fields to place first in the output.
14446 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
14448 @section setdar, setsar
14450 The @code{setdar} filter sets the Display Aspect Ratio for the filter
14453 This is done by changing the specified Sample (aka Pixel) Aspect
14454 Ratio, according to the following equation:
14456 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
14459 Keep in mind that the @code{setdar} filter does not modify the pixel
14460 dimensions of the video frame. Also, the display aspect ratio set by
14461 this filter may be changed by later filters in the filterchain,
14462 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
14465 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
14466 the filter output video.
14468 Note that as a consequence of the application of this filter, the
14469 output display aspect ratio will change according to the equation
14472 Keep in mind that the sample aspect ratio set by the @code{setsar}
14473 filter may be changed by later filters in the filterchain, e.g. if
14474 another "setsar" or a "setdar" filter is applied.
14476 It accepts the following parameters:
14479 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
14480 Set the aspect ratio used by the filter.
14482 The parameter can be a floating point number string, an expression, or
14483 a string of the form @var{num}:@var{den}, where @var{num} and
14484 @var{den} are the numerator and denominator of the aspect ratio. If
14485 the parameter is not specified, it is assumed the value "0".
14486 In case the form "@var{num}:@var{den}" is used, the @code{:} character
14490 Set the maximum integer value to use for expressing numerator and
14491 denominator when reducing the expressed aspect ratio to a rational.
14492 Default value is @code{100}.
14496 The parameter @var{sar} is an expression containing
14497 the following constants:
14501 These are approximated values for the mathematical constants e
14502 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14505 The input width and height.
14508 These are the same as @var{w} / @var{h}.
14511 The input sample aspect ratio.
14514 The input display aspect ratio. It is the same as
14515 (@var{w} / @var{h}) * @var{sar}.
14518 Horizontal and vertical chroma subsample values. For example, for the
14519 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14522 @subsection Examples
14527 To change the display aspect ratio to 16:9, specify one of the following:
14534 To change the sample aspect ratio to 10:11, specify:
14540 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14541 1000 in the aspect ratio reduction, use the command:
14543 setdar=ratio=16/9:max=1000
14551 Force field for the output video frame.
14553 The @code{setfield} filter marks the interlace type field for the
14554 output frames. It does not change the input frame, but only sets the
14555 corresponding property, which affects how the frame is treated by
14556 following filters (e.g. @code{fieldorder} or @code{yadif}).
14558 The filter accepts the following options:
14563 Available values are:
14567 Keep the same field property.
14570 Mark the frame as bottom-field-first.
14573 Mark the frame as top-field-first.
14576 Mark the frame as progressive.
14582 Show a line containing various information for each input video frame.
14583 The input video is not modified.
14585 The shown line contains a sequence of key/value pairs of the form
14586 @var{key}:@var{value}.
14588 The following values are shown in the output:
14592 The (sequential) number of the input frame, starting from 0.
14595 The Presentation TimeStamp of the input frame, expressed as a number of
14596 time base units. The time base unit depends on the filter input pad.
14599 The Presentation TimeStamp of the input frame, expressed as a number of
14603 The position of the frame in the input stream, or -1 if this information is
14604 unavailable and/or meaningless (for example in case of synthetic video).
14607 The pixel format name.
14610 The sample aspect ratio of the input frame, expressed in the form
14611 @var{num}/@var{den}.
14614 The size of the input frame. For the syntax of this option, check the
14615 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14618 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14619 for bottom field first).
14622 This is 1 if the frame is a key frame, 0 otherwise.
14625 The picture type of the input frame ("I" for an I-frame, "P" for a
14626 P-frame, "B" for a B-frame, or "?" for an unknown type).
14627 Also refer to the documentation of the @code{AVPictureType} enum and of
14628 the @code{av_get_picture_type_char} function defined in
14629 @file{libavutil/avutil.h}.
14632 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14634 @item plane_checksum
14635 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14636 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14639 @section showpalette
14641 Displays the 256 colors palette of each frame. This filter is only relevant for
14642 @var{pal8} pixel format frames.
14644 It accepts the following option:
14648 Set the size of the box used to represent one palette color entry. Default is
14649 @code{30} (for a @code{30x30} pixel box).
14652 @section shuffleframes
14654 Reorder and/or duplicate and/or drop video frames.
14656 It accepts the following parameters:
14660 Set the destination indexes of input frames.
14661 This is space or '|' separated list of indexes that maps input frames to output
14662 frames. Number of indexes also sets maximal value that each index may have.
14663 '-1' index have special meaning and that is to drop frame.
14666 The first frame has the index 0. The default is to keep the input unchanged.
14668 @subsection Examples
14672 Swap second and third frame of every three frames of the input:
14674 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14678 Swap 10th and 1st frame of every ten frames of the input:
14680 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14684 @section shuffleplanes
14686 Reorder and/or duplicate video planes.
14688 It accepts the following parameters:
14693 The index of the input plane to be used as the first output plane.
14696 The index of the input plane to be used as the second output plane.
14699 The index of the input plane to be used as the third output plane.
14702 The index of the input plane to be used as the fourth output plane.
14706 The first plane has the index 0. The default is to keep the input unchanged.
14708 @subsection Examples
14712 Swap the second and third planes of the input:
14714 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14718 @anchor{signalstats}
14719 @section signalstats
14720 Evaluate various visual metrics that assist in determining issues associated
14721 with the digitization of analog video media.
14723 By default the filter will log these metadata values:
14727 Display the minimal Y value contained within the input frame. Expressed in
14731 Display the Y value at the 10% percentile within the input frame. Expressed in
14735 Display the average Y value within the input frame. Expressed in range of
14739 Display the Y value at the 90% percentile within the input frame. Expressed in
14743 Display the maximum Y value contained within the input frame. Expressed in
14747 Display the minimal U value contained within the input frame. Expressed in
14751 Display the U value at the 10% percentile within the input frame. Expressed in
14755 Display the average U value within the input frame. Expressed in range of
14759 Display the U value at the 90% percentile within the input frame. Expressed in
14763 Display the maximum U value contained within the input frame. Expressed in
14767 Display the minimal V value contained within the input frame. Expressed in
14771 Display the V value at the 10% percentile within the input frame. Expressed in
14775 Display the average V value within the input frame. Expressed in range of
14779 Display the V value at the 90% percentile within the input frame. Expressed in
14783 Display the maximum V value contained within the input frame. Expressed in
14787 Display the minimal saturation value contained within the input frame.
14788 Expressed in range of [0-~181.02].
14791 Display the saturation value at the 10% percentile within the input frame.
14792 Expressed in range of [0-~181.02].
14795 Display the average saturation value within the input frame. Expressed in range
14799 Display the saturation value at the 90% percentile within the input frame.
14800 Expressed in range of [0-~181.02].
14803 Display the maximum saturation value contained within the input frame.
14804 Expressed in range of [0-~181.02].
14807 Display the median value for hue within the input frame. Expressed in range of
14811 Display the average value for hue within the input frame. Expressed in range of
14815 Display the average of sample value difference between all values of the Y
14816 plane in the current frame and corresponding values of the previous input frame.
14817 Expressed in range of [0-255].
14820 Display the average of sample value difference between all values of the U
14821 plane in the current frame and corresponding values of the previous input frame.
14822 Expressed in range of [0-255].
14825 Display the average of sample value difference between all values of the V
14826 plane in the current frame and corresponding values of the previous input frame.
14827 Expressed in range of [0-255].
14830 Display bit depth of Y plane in current frame.
14831 Expressed in range of [0-16].
14834 Display bit depth of U plane in current frame.
14835 Expressed in range of [0-16].
14838 Display bit depth of V plane in current frame.
14839 Expressed in range of [0-16].
14842 The filter accepts the following options:
14848 @option{stat} specify an additional form of image analysis.
14849 @option{out} output video with the specified type of pixel highlighted.
14851 Both options accept the following values:
14855 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14856 unlike the neighboring pixels of the same field. Examples of temporal outliers
14857 include the results of video dropouts, head clogs, or tape tracking issues.
14860 Identify @var{vertical line repetition}. Vertical line repetition includes
14861 similar rows of pixels within a frame. In born-digital video vertical line
14862 repetition is common, but this pattern is uncommon in video digitized from an
14863 analog source. When it occurs in video that results from the digitization of an
14864 analog source it can indicate concealment from a dropout compensator.
14867 Identify pixels that fall outside of legal broadcast range.
14871 Set the highlight color for the @option{out} option. The default color is
14875 @subsection Examples
14879 Output data of various video metrics:
14881 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14885 Output specific data about the minimum and maximum values of the Y plane per frame:
14887 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14891 Playback video while highlighting pixels that are outside of broadcast range in red.
14893 ffplay example.mov -vf signalstats="out=brng:color=red"
14897 Playback video with signalstats metadata drawn over the frame.
14899 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14902 The contents of signalstat_drawtext.txt used in the command are:
14905 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14906 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14907 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14908 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14916 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14917 input. In this case the matching between the inputs can be calculated additionally.
14918 The filter always passes through the first input. The signature of each stream can
14919 be written into a file.
14921 It accepts the following options:
14925 Enable or disable the matching process.
14927 Available values are:
14931 Disable the calculation of a matching (default).
14933 Calculate the matching for the whole video and output whether the whole video
14934 matches or only parts.
14936 Calculate only until a matching is found or the video ends. Should be faster in
14941 Set the number of inputs. The option value must be a non negative integer.
14942 Default value is 1.
14945 Set the path to which the output is written. If there is more than one input,
14946 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14947 integer), that will be replaced with the input number. If no filename is
14948 specified, no output will be written. This is the default.
14951 Choose the output format.
14953 Available values are:
14957 Use the specified binary representation (default).
14959 Use the specified xml representation.
14963 Set threshold to detect one word as similar. The option value must be an integer
14964 greater than zero. The default value is 9000.
14967 Set threshold to detect all words as similar. The option value must be an integer
14968 greater than zero. The default value is 60000.
14971 Set threshold to detect frames as similar. The option value must be an integer
14972 greater than zero. The default value is 116.
14975 Set the minimum length of a sequence in frames to recognize it as matching
14976 sequence. The option value must be a non negative integer value.
14977 The default value is 0.
14980 Set the minimum relation, that matching frames to all frames must have.
14981 The option value must be a double value between 0 and 1. The default value is 0.5.
14984 @subsection Examples
14988 To calculate the signature of an input video and store it in signature.bin:
14990 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14994 To detect whether two videos match and store the signatures in XML format in
14995 signature0.xml and signature1.xml:
14997 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 -
15005 Blur the input video without impacting the outlines.
15007 It accepts the following options:
15010 @item luma_radius, lr
15011 Set the luma radius. The option value must be a float number in
15012 the range [0.1,5.0] that specifies the variance of the gaussian filter
15013 used to blur the image (slower if larger). Default value is 1.0.
15015 @item luma_strength, ls
15016 Set the luma strength. The option value must be a float number
15017 in the range [-1.0,1.0] that configures the blurring. A value included
15018 in [0.0,1.0] will blur the image whereas a value included in
15019 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15021 @item luma_threshold, lt
15022 Set the luma threshold used as a coefficient to determine
15023 whether a pixel should be blurred or not. The option value must be an
15024 integer in the range [-30,30]. A value of 0 will filter all the image,
15025 a value included in [0,30] will filter flat areas and a value included
15026 in [-30,0] will filter edges. Default value is 0.
15028 @item chroma_radius, cr
15029 Set the chroma radius. The option value must be a float number in
15030 the range [0.1,5.0] that specifies the variance of the gaussian filter
15031 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15033 @item chroma_strength, cs
15034 Set the chroma strength. The option value must be a float number
15035 in the range [-1.0,1.0] that configures the blurring. A value included
15036 in [0.0,1.0] will blur the image whereas a value included in
15037 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15039 @item chroma_threshold, ct
15040 Set the chroma threshold used as a coefficient to determine
15041 whether a pixel should be blurred or not. The option value must be an
15042 integer in the range [-30,30]. A value of 0 will filter all the image,
15043 a value included in [0,30] will filter flat areas and a value included
15044 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15047 If a chroma option is not explicitly set, the corresponding luma value
15052 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15054 This filter takes in input two input videos, the first input is
15055 considered the "main" source and is passed unchanged to the
15056 output. The second input is used as a "reference" video for computing
15059 Both video inputs must have the same resolution and pixel format for
15060 this filter to work correctly. Also it assumes that both inputs
15061 have the same number of frames, which are compared one by one.
15063 The filter stores the calculated SSIM of each frame.
15065 The description of the accepted parameters follows.
15068 @item stats_file, f
15069 If specified the filter will use the named file to save the SSIM of
15070 each individual frame. When filename equals "-" the data is sent to
15074 The file printed if @var{stats_file} is selected, contains a sequence of
15075 key/value pairs of the form @var{key}:@var{value} for each compared
15078 A description of each shown parameter follows:
15082 sequential number of the input frame, starting from 1
15084 @item Y, U, V, R, G, B
15085 SSIM of the compared frames for the component specified by the suffix.
15088 SSIM of the compared frames for the whole frame.
15091 Same as above but in dB representation.
15094 This filter also supports the @ref{framesync} options.
15098 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15099 [main][ref] ssim="stats_file=stats.log" [out]
15102 On this example the input file being processed is compared with the
15103 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15104 is stored in @file{stats.log}.
15106 Another example with both psnr and ssim at same time:
15108 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
15113 Convert between different stereoscopic image formats.
15115 The filters accept the following options:
15119 Set stereoscopic image format of input.
15121 Available values for input image formats are:
15124 side by side parallel (left eye left, right eye right)
15127 side by side crosseye (right eye left, left eye right)
15130 side by side parallel with half width resolution
15131 (left eye left, right eye right)
15134 side by side crosseye with half width resolution
15135 (right eye left, left eye right)
15138 above-below (left eye above, right eye below)
15141 above-below (right eye above, left eye below)
15144 above-below with half height resolution
15145 (left eye above, right eye below)
15148 above-below with half height resolution
15149 (right eye above, left eye below)
15152 alternating frames (left eye first, right eye second)
15155 alternating frames (right eye first, left eye second)
15158 interleaved rows (left eye has top row, right eye starts on next row)
15161 interleaved rows (right eye has top row, left eye starts on next row)
15164 interleaved columns, left eye first
15167 interleaved columns, right eye first
15169 Default value is @samp{sbsl}.
15173 Set stereoscopic image format of output.
15177 side by side parallel (left eye left, right eye right)
15180 side by side crosseye (right eye left, left eye right)
15183 side by side parallel with half width resolution
15184 (left eye left, right eye right)
15187 side by side crosseye with half width resolution
15188 (right eye left, left eye right)
15191 above-below (left eye above, right eye below)
15194 above-below (right eye above, left eye below)
15197 above-below with half height resolution
15198 (left eye above, right eye below)
15201 above-below with half height resolution
15202 (right eye above, left eye below)
15205 alternating frames (left eye first, right eye second)
15208 alternating frames (right eye first, left eye second)
15211 interleaved rows (left eye has top row, right eye starts on next row)
15214 interleaved rows (right eye has top row, left eye starts on next row)
15217 anaglyph red/blue gray
15218 (red filter on left eye, blue filter on right eye)
15221 anaglyph red/green gray
15222 (red filter on left eye, green filter on right eye)
15225 anaglyph red/cyan gray
15226 (red filter on left eye, cyan filter on right eye)
15229 anaglyph red/cyan half colored
15230 (red filter on left eye, cyan filter on right eye)
15233 anaglyph red/cyan color
15234 (red filter on left eye, cyan filter on right eye)
15237 anaglyph red/cyan color optimized with the least squares projection of dubois
15238 (red filter on left eye, cyan filter on right eye)
15241 anaglyph green/magenta gray
15242 (green filter on left eye, magenta filter on right eye)
15245 anaglyph green/magenta half colored
15246 (green filter on left eye, magenta filter on right eye)
15249 anaglyph green/magenta colored
15250 (green filter on left eye, magenta filter on right eye)
15253 anaglyph green/magenta color optimized with the least squares projection of dubois
15254 (green filter on left eye, magenta filter on right eye)
15257 anaglyph yellow/blue gray
15258 (yellow filter on left eye, blue filter on right eye)
15261 anaglyph yellow/blue half colored
15262 (yellow filter on left eye, blue filter on right eye)
15265 anaglyph yellow/blue colored
15266 (yellow filter on left eye, blue filter on right eye)
15269 anaglyph yellow/blue color optimized with the least squares projection of dubois
15270 (yellow filter on left eye, blue filter on right eye)
15273 mono output (left eye only)
15276 mono output (right eye only)
15279 checkerboard, left eye first
15282 checkerboard, right eye first
15285 interleaved columns, left eye first
15288 interleaved columns, right eye first
15294 Default value is @samp{arcd}.
15297 @subsection Examples
15301 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
15307 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
15313 @section streamselect, astreamselect
15314 Select video or audio streams.
15316 The filter accepts the following options:
15320 Set number of inputs. Default is 2.
15323 Set input indexes to remap to outputs.
15326 @subsection Commands
15328 The @code{streamselect} and @code{astreamselect} filter supports the following
15333 Set input indexes to remap to outputs.
15336 @subsection Examples
15340 Select first 5 seconds 1st stream and rest of time 2nd stream:
15342 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
15346 Same as above, but for audio:
15348 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
15353 Apply sobel operator to input video stream.
15355 The filter accepts the following option:
15359 Set which planes will be processed, unprocessed planes will be copied.
15360 By default value 0xf, all planes will be processed.
15363 Set value which will be multiplied with filtered result.
15366 Set value which will be added to filtered result.
15372 Apply a simple postprocessing filter that compresses and decompresses the image
15373 at several (or - in the case of @option{quality} level @code{6} - all) shifts
15374 and average the results.
15376 The filter accepts the following options:
15380 Set quality. This option defines the number of levels for averaging. It accepts
15381 an integer in the range 0-6. If set to @code{0}, the filter will have no
15382 effect. A value of @code{6} means the higher quality. For each increment of
15383 that value the speed drops by a factor of approximately 2. Default value is
15387 Force a constant quantization parameter. If not set, the filter will use the QP
15388 from the video stream (if available).
15391 Set thresholding mode. Available modes are:
15395 Set hard thresholding (default).
15397 Set soft thresholding (better de-ringing effect, but likely blurrier).
15400 @item use_bframe_qp
15401 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
15402 option may cause flicker since the B-Frames have often larger QP. Default is
15403 @code{0} (not enabled).
15408 Scale the input by applying one of the super-resolution methods based on
15409 convolutional neural networks.
15411 Training scripts as well as scripts for model generation are provided in
15412 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
15414 The filter accepts the following options:
15418 Specify which super-resolution model to use. This option accepts the following values:
15422 Super-Resolution Convolutional Neural Network model.
15423 See @url{https://arxiv.org/abs/1501.00092}.
15426 Efficient Sub-Pixel Convolutional Neural Network model.
15427 See @url{https://arxiv.org/abs/1609.05158}.
15431 Default value is @samp{srcnn}.
15434 Specify which DNN backend to use for model loading and execution. This option accepts
15435 the following values:
15439 Native implementation of DNN loading and execution.
15442 TensorFlow backend. To enable this backend you
15443 need to install the TensorFlow for C library (see
15444 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
15445 @code{--enable-libtensorflow}
15449 Default value is @samp{native}.
15452 Set scale factor for SRCNN model, for which custom model file was provided.
15453 Allowed values are @code{2}, @code{3} and @code{4}. Default value is @code{2}.
15454 Scale factor is necessary for SRCNN model, because it accepts input upscaled
15455 using bicubic upscaling with proper scale factor.
15457 @item model_filename
15458 Set path to model file specifying network architecture and its parameters.
15459 Note that different backends use different file formats. TensorFlow backend
15460 can load files for both formats, while native backend can load files for only
15468 Draw subtitles on top of input video using the libass library.
15470 To enable compilation of this filter you need to configure FFmpeg with
15471 @code{--enable-libass}. This filter also requires a build with libavcodec and
15472 libavformat to convert the passed subtitles file to ASS (Advanced Substation
15473 Alpha) subtitles format.
15475 The filter accepts the following options:
15479 Set the filename of the subtitle file to read. It must be specified.
15481 @item original_size
15482 Specify the size of the original video, the video for which the ASS file
15483 was composed. For the syntax of this option, check the
15484 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15485 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
15486 correctly scale the fonts if the aspect ratio has been changed.
15489 Set a directory path containing fonts that can be used by the filter.
15490 These fonts will be used in addition to whatever the font provider uses.
15493 Process alpha channel, by default alpha channel is untouched.
15496 Set subtitles input character encoding. @code{subtitles} filter only. Only
15497 useful if not UTF-8.
15499 @item stream_index, si
15500 Set subtitles stream index. @code{subtitles} filter only.
15503 Override default style or script info parameters of the subtitles. It accepts a
15504 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
15507 If the first key is not specified, it is assumed that the first value
15508 specifies the @option{filename}.
15510 For example, to render the file @file{sub.srt} on top of the input
15511 video, use the command:
15516 which is equivalent to:
15518 subtitles=filename=sub.srt
15521 To render the default subtitles stream from file @file{video.mkv}, use:
15523 subtitles=video.mkv
15526 To render the second subtitles stream from that file, use:
15528 subtitles=video.mkv:si=1
15531 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
15532 @code{DejaVu Serif}, use:
15534 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
15537 @section super2xsai
15539 Scale the input by 2x and smooth using the Super2xSaI (Scale and
15540 Interpolate) pixel art scaling algorithm.
15542 Useful for enlarging pixel art images without reducing sharpness.
15546 Swap two rectangular objects in video.
15548 This filter accepts the following options:
15558 Set 1st rect x coordinate.
15561 Set 1st rect y coordinate.
15564 Set 2nd rect x coordinate.
15567 Set 2nd rect y coordinate.
15569 All expressions are evaluated once for each frame.
15572 The all options are expressions containing the following constants:
15577 The input width and height.
15580 same as @var{w} / @var{h}
15583 input sample aspect ratio
15586 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15589 The number of the input frame, starting from 0.
15592 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15595 the position in the file of the input frame, NAN if unknown
15603 Apply telecine process to the video.
15605 This filter accepts the following options:
15614 The default value is @code{top}.
15618 A string of numbers representing the pulldown pattern you wish to apply.
15619 The default value is @code{23}.
15623 Some typical patterns:
15628 24p: 2332 (preferred)
15635 24p: 222222222223 ("Euro pulldown")
15642 Apply threshold effect to video stream.
15644 This filter needs four video streams to perform thresholding.
15645 First stream is stream we are filtering.
15646 Second stream is holding threshold values, third stream is holding min values,
15647 and last, fourth stream is holding max values.
15649 The filter accepts the following option:
15653 Set which planes will be processed, unprocessed planes will be copied.
15654 By default value 0xf, all planes will be processed.
15657 For example if first stream pixel's component value is less then threshold value
15658 of pixel component from 2nd threshold stream, third stream value will picked,
15659 otherwise fourth stream pixel component value will be picked.
15661 Using color source filter one can perform various types of thresholding:
15663 @subsection Examples
15667 Binary threshold, using gray color as threshold:
15669 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15673 Inverted binary threshold, using gray color as threshold:
15675 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15679 Truncate binary threshold, using gray color as threshold:
15681 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15685 Threshold to zero, using gray color as threshold:
15687 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15691 Inverted threshold to zero, using gray color as threshold:
15693 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15698 Select the most representative frame in a given sequence of consecutive frames.
15700 The filter accepts the following options:
15704 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15705 will pick one of them, and then handle the next batch of @var{n} frames until
15706 the end. Default is @code{100}.
15709 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15710 value will result in a higher memory usage, so a high value is not recommended.
15712 @subsection Examples
15716 Extract one picture each 50 frames:
15722 Complete example of a thumbnail creation with @command{ffmpeg}:
15724 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15730 Tile several successive frames together.
15732 The filter accepts the following options:
15737 Set the grid size (i.e. the number of lines and columns). For the syntax of
15738 this option, check the
15739 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15742 Set the maximum number of frames to render in the given area. It must be less
15743 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15744 the area will be used.
15747 Set the outer border margin in pixels.
15750 Set the inner border thickness (i.e. the number of pixels between frames). For
15751 more advanced padding options (such as having different values for the edges),
15752 refer to the pad video filter.
15755 Specify the color of the unused area. For the syntax of this option, check the
15756 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15757 The default value of @var{color} is "black".
15760 Set the number of frames to overlap when tiling several successive frames together.
15761 The value must be between @code{0} and @var{nb_frames - 1}.
15764 Set the number of frames to initially be empty before displaying first output frame.
15765 This controls how soon will one get first output frame.
15766 The value must be between @code{0} and @var{nb_frames - 1}.
15769 @subsection Examples
15773 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15775 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15777 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15778 duplicating each output frame to accommodate the originally detected frame
15782 Display @code{5} pictures in an area of @code{3x2} frames,
15783 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15784 mixed flat and named options:
15786 tile=3x2:nb_frames=5:padding=7:margin=2
15790 @section tinterlace
15792 Perform various types of temporal field interlacing.
15794 Frames are counted starting from 1, so the first input frame is
15797 The filter accepts the following options:
15802 Specify the mode of the interlacing. This option can also be specified
15803 as a value alone. See below for a list of values for this option.
15805 Available values are:
15809 Move odd frames into the upper field, even into the lower field,
15810 generating a double height frame at half frame rate.
15814 Frame 1 Frame 2 Frame 3 Frame 4
15816 11111 22222 33333 44444
15817 11111 22222 33333 44444
15818 11111 22222 33333 44444
15819 11111 22222 33333 44444
15833 Only output odd frames, even frames are dropped, generating a frame with
15834 unchanged height at half frame rate.
15839 Frame 1 Frame 2 Frame 3 Frame 4
15841 11111 22222 33333 44444
15842 11111 22222 33333 44444
15843 11111 22222 33333 44444
15844 11111 22222 33333 44444
15854 Only output even frames, odd frames are dropped, generating a frame with
15855 unchanged height at half frame rate.
15860 Frame 1 Frame 2 Frame 3 Frame 4
15862 11111 22222 33333 44444
15863 11111 22222 33333 44444
15864 11111 22222 33333 44444
15865 11111 22222 33333 44444
15875 Expand each frame to full height, but pad alternate lines with black,
15876 generating a frame with double height at the same input frame rate.
15881 Frame 1 Frame 2 Frame 3 Frame 4
15883 11111 22222 33333 44444
15884 11111 22222 33333 44444
15885 11111 22222 33333 44444
15886 11111 22222 33333 44444
15889 11111 ..... 33333 .....
15890 ..... 22222 ..... 44444
15891 11111 ..... 33333 .....
15892 ..... 22222 ..... 44444
15893 11111 ..... 33333 .....
15894 ..... 22222 ..... 44444
15895 11111 ..... 33333 .....
15896 ..... 22222 ..... 44444
15900 @item interleave_top, 4
15901 Interleave the upper field from odd frames with the lower field from
15902 even frames, generating a frame with unchanged height at half frame rate.
15907 Frame 1 Frame 2 Frame 3 Frame 4
15909 11111<- 22222 33333<- 44444
15910 11111 22222<- 33333 44444<-
15911 11111<- 22222 33333<- 44444
15912 11111 22222<- 33333 44444<-
15922 @item interleave_bottom, 5
15923 Interleave the lower field from odd frames with the upper field from
15924 even frames, generating a frame with unchanged height at half frame rate.
15929 Frame 1 Frame 2 Frame 3 Frame 4
15931 11111 22222<- 33333 44444<-
15932 11111<- 22222 33333<- 44444
15933 11111 22222<- 33333 44444<-
15934 11111<- 22222 33333<- 44444
15944 @item interlacex2, 6
15945 Double frame rate with unchanged height. Frames are inserted each
15946 containing the second temporal field from the previous input frame and
15947 the first temporal field from the next input frame. This mode relies on
15948 the top_field_first flag. Useful for interlaced video displays with no
15949 field synchronisation.
15954 Frame 1 Frame 2 Frame 3 Frame 4
15956 11111 22222 33333 44444
15957 11111 22222 33333 44444
15958 11111 22222 33333 44444
15959 11111 22222 33333 44444
15962 11111 22222 22222 33333 33333 44444 44444
15963 11111 11111 22222 22222 33333 33333 44444
15964 11111 22222 22222 33333 33333 44444 44444
15965 11111 11111 22222 22222 33333 33333 44444
15970 Move odd frames into the upper field, even into the lower field,
15971 generating a double height frame at same frame rate.
15976 Frame 1 Frame 2 Frame 3 Frame 4
15978 11111 22222 33333 44444
15979 11111 22222 33333 44444
15980 11111 22222 33333 44444
15981 11111 22222 33333 44444
15984 11111 33333 33333 55555
15985 22222 22222 44444 44444
15986 11111 33333 33333 55555
15987 22222 22222 44444 44444
15988 11111 33333 33333 55555
15989 22222 22222 44444 44444
15990 11111 33333 33333 55555
15991 22222 22222 44444 44444
15996 Numeric values are deprecated but are accepted for backward
15997 compatibility reasons.
15999 Default mode is @code{merge}.
16002 Specify flags influencing the filter process.
16004 Available value for @var{flags} is:
16007 @item low_pass_filter, vlfp
16008 Enable linear vertical low-pass filtering in the filter.
16009 Vertical low-pass filtering is required when creating an interlaced
16010 destination from a progressive source which contains high-frequency
16011 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16014 @item complex_filter, cvlfp
16015 Enable complex vertical low-pass filtering.
16016 This will slightly less reduce interlace 'twitter' and Moire
16017 patterning but better retain detail and subjective sharpness impression.
16021 Vertical low-pass filtering can only be enabled for @option{mode}
16022 @var{interleave_top} and @var{interleave_bottom}.
16028 Mix successive video frames.
16030 A description of the accepted options follows.
16034 The number of successive frames to mix. If unspecified, it defaults to 3.
16037 Specify weight of each input video frame.
16038 Each weight is separated by space. If number of weights is smaller than
16039 number of @var{frames} last specified weight will be used for all remaining
16043 Specify scale, if it is set it will be multiplied with sum
16044 of each weight multiplied with pixel values to give final destination
16045 pixel value. By default @var{scale} is auto scaled to sum of weights.
16048 @subsection Examples
16052 Average 7 successive frames:
16054 tmix=frames=7:weights="1 1 1 1 1 1 1"
16058 Apply simple temporal convolution:
16060 tmix=frames=3:weights="-1 3 -1"
16064 Similar as above but only showing temporal differences:
16066 tmix=frames=3:weights="-1 2 -1":scale=1
16071 Tone map colors from different dynamic ranges.
16073 This filter expects data in single precision floating point, as it needs to
16074 operate on (and can output) out-of-range values. Another filter, such as
16075 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16077 The tonemapping algorithms implemented only work on linear light, so input
16078 data should be linearized beforehand (and possibly correctly tagged).
16081 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16084 @subsection Options
16085 The filter accepts the following options.
16089 Set the tone map algorithm to use.
16091 Possible values are:
16094 Do not apply any tone map, only desaturate overbright pixels.
16097 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16098 in-range values, while distorting out-of-range values.
16101 Stretch the entire reference gamut to a linear multiple of the display.
16104 Fit a logarithmic transfer between the tone curves.
16107 Preserve overall image brightness with a simple curve, using nonlinear
16108 contrast, which results in flattening details and degrading color accuracy.
16111 Preserve both dark and bright details better than @var{reinhard}, at the cost
16112 of slightly darkening everything. Use it when detail preservation is more
16113 important than color and brightness accuracy.
16116 Smoothly map out-of-range values, while retaining contrast and colors for
16117 in-range material as much as possible. Use it when color accuracy is more
16118 important than detail preservation.
16124 Tune the tone mapping algorithm.
16126 This affects the following algorithms:
16132 Specifies the scale factor to use while stretching.
16136 Specifies the exponent of the function.
16140 Specify an extra linear coefficient to multiply into the signal before clipping.
16144 Specify the local contrast coefficient at the display peak.
16145 Default to 0.5, which means that in-gamut values will be about half as bright
16152 Specify the transition point from linear to mobius transform. Every value
16153 below this point is guaranteed to be mapped 1:1. The higher the value, the
16154 more accurate the result will be, at the cost of losing bright details.
16155 Default to 0.3, which due to the steep initial slope still preserves in-range
16156 colors fairly accurately.
16160 Apply desaturation for highlights that exceed this level of brightness. The
16161 higher the parameter, the more color information will be preserved. This
16162 setting helps prevent unnaturally blown-out colors for super-highlights, by
16163 (smoothly) turning into white instead. This makes images feel more natural,
16164 at the cost of reducing information about out-of-range colors.
16166 The default of 2.0 is somewhat conservative and will mostly just apply to
16167 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
16169 This option works only if the input frame has a supported color tag.
16172 Override signal/nominal/reference peak with this value. Useful when the
16173 embedded peak information in display metadata is not reliable or when tone
16174 mapping from a lower range to a higher range.
16179 Transpose rows with columns in the input video and optionally flip it.
16181 It accepts the following parameters:
16186 Specify the transposition direction.
16188 Can assume the following values:
16190 @item 0, 4, cclock_flip
16191 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
16199 Rotate by 90 degrees clockwise, that is:
16207 Rotate by 90 degrees counterclockwise, that is:
16214 @item 3, 7, clock_flip
16215 Rotate by 90 degrees clockwise and vertically flip, that is:
16223 For values between 4-7, the transposition is only done if the input
16224 video geometry is portrait and not landscape. These values are
16225 deprecated, the @code{passthrough} option should be used instead.
16227 Numerical values are deprecated, and should be dropped in favor of
16228 symbolic constants.
16231 Do not apply the transposition if the input geometry matches the one
16232 specified by the specified value. It accepts the following values:
16235 Always apply transposition.
16237 Preserve portrait geometry (when @var{height} >= @var{width}).
16239 Preserve landscape geometry (when @var{width} >= @var{height}).
16242 Default value is @code{none}.
16245 For example to rotate by 90 degrees clockwise and preserve portrait
16248 transpose=dir=1:passthrough=portrait
16251 The command above can also be specified as:
16253 transpose=1:portrait
16257 Trim the input so that the output contains one continuous subpart of the input.
16259 It accepts the following parameters:
16262 Specify the time of the start of the kept section, i.e. the frame with the
16263 timestamp @var{start} will be the first frame in the output.
16266 Specify the time of the first frame that will be dropped, i.e. the frame
16267 immediately preceding the one with the timestamp @var{end} will be the last
16268 frame in the output.
16271 This is the same as @var{start}, except this option sets the start timestamp
16272 in timebase units instead of seconds.
16275 This is the same as @var{end}, except this option sets the end timestamp
16276 in timebase units instead of seconds.
16279 The maximum duration of the output in seconds.
16282 The number of the first frame that should be passed to the output.
16285 The number of the first frame that should be dropped.
16288 @option{start}, @option{end}, and @option{duration} are expressed as time
16289 duration specifications; see
16290 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16291 for the accepted syntax.
16293 Note that the first two sets of the start/end options and the @option{duration}
16294 option look at the frame timestamp, while the _frame variants simply count the
16295 frames that pass through the filter. Also note that this filter does not modify
16296 the timestamps. If you wish for the output timestamps to start at zero, insert a
16297 setpts filter after the trim filter.
16299 If multiple start or end options are set, this filter tries to be greedy and
16300 keep all the frames that match at least one of the specified constraints. To keep
16301 only the part that matches all the constraints at once, chain multiple trim
16304 The defaults are such that all the input is kept. So it is possible to set e.g.
16305 just the end values to keep everything before the specified time.
16310 Drop everything except the second minute of input:
16312 ffmpeg -i INPUT -vf trim=60:120
16316 Keep only the first second:
16318 ffmpeg -i INPUT -vf trim=duration=1
16323 @section unpremultiply
16324 Apply alpha unpremultiply effect to input video stream using first plane
16325 of second stream as alpha.
16327 Both streams must have same dimensions and same pixel format.
16329 The filter accepts the following option:
16333 Set which planes will be processed, unprocessed planes will be copied.
16334 By default value 0xf, all planes will be processed.
16336 If the format has 1 or 2 components, then luma is bit 0.
16337 If the format has 3 or 4 components:
16338 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
16339 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
16340 If present, the alpha channel is always the last bit.
16343 Do not require 2nd input for processing, instead use alpha plane from input stream.
16349 Sharpen or blur the input video.
16351 It accepts the following parameters:
16354 @item luma_msize_x, lx
16355 Set the luma matrix horizontal size. It must be an odd integer between
16356 3 and 23. The default value is 5.
16358 @item luma_msize_y, ly
16359 Set the luma matrix vertical size. It must be an odd integer between 3
16360 and 23. The default value is 5.
16362 @item luma_amount, la
16363 Set the luma effect strength. It must be a floating point number, reasonable
16364 values lay between -1.5 and 1.5.
16366 Negative values will blur the input video, while positive values will
16367 sharpen it, a value of zero will disable the effect.
16369 Default value is 1.0.
16371 @item chroma_msize_x, cx
16372 Set the chroma matrix horizontal size. It must be an odd integer
16373 between 3 and 23. The default value is 5.
16375 @item chroma_msize_y, cy
16376 Set the chroma matrix vertical size. It must be an odd integer
16377 between 3 and 23. The default value is 5.
16379 @item chroma_amount, ca
16380 Set the chroma effect strength. It must be a floating point number, reasonable
16381 values lay between -1.5 and 1.5.
16383 Negative values will blur the input video, while positive values will
16384 sharpen it, a value of zero will disable the effect.
16386 Default value is 0.0.
16390 All parameters are optional and default to the equivalent of the
16391 string '5:5:1.0:5:5:0.0'.
16393 @subsection Examples
16397 Apply strong luma sharpen effect:
16399 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
16403 Apply a strong blur of both luma and chroma parameters:
16405 unsharp=7:7:-2:7:7:-2
16411 Apply ultra slow/simple postprocessing filter that compresses and decompresses
16412 the image at several (or - in the case of @option{quality} level @code{8} - all)
16413 shifts and average the results.
16415 The way this differs from the behavior of spp is that uspp actually encodes &
16416 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
16417 DCT similar to MJPEG.
16419 The filter accepts the following options:
16423 Set quality. This option defines the number of levels for averaging. It accepts
16424 an integer in the range 0-8. If set to @code{0}, the filter will have no
16425 effect. A value of @code{8} means the higher quality. For each increment of
16426 that value the speed drops by a factor of approximately 2. Default value is
16430 Force a constant quantization parameter. If not set, the filter will use the QP
16431 from the video stream (if available).
16434 @section vaguedenoiser
16436 Apply a wavelet based denoiser.
16438 It transforms each frame from the video input into the wavelet domain,
16439 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
16440 the obtained coefficients. It does an inverse wavelet transform after.
16441 Due to wavelet properties, it should give a nice smoothed result, and
16442 reduced noise, without blurring picture features.
16444 This filter accepts the following options:
16448 The filtering strength. The higher, the more filtered the video will be.
16449 Hard thresholding can use a higher threshold than soft thresholding
16450 before the video looks overfiltered. Default value is 2.
16453 The filtering method the filter will use.
16455 It accepts the following values:
16458 All values under the threshold will be zeroed.
16461 All values under the threshold will be zeroed. All values above will be
16462 reduced by the threshold.
16465 Scales or nullifies coefficients - intermediary between (more) soft and
16466 (less) hard thresholding.
16469 Default is garrote.
16472 Number of times, the wavelet will decompose the picture. Picture can't
16473 be decomposed beyond a particular point (typically, 8 for a 640x480
16474 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
16477 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
16480 A list of the planes to process. By default all planes are processed.
16483 @section vectorscope
16485 Display 2 color component values in the two dimensional graph (which is called
16488 This filter accepts the following options:
16492 Set vectorscope mode.
16494 It accepts the following values:
16497 Gray values are displayed on graph, higher brightness means more pixels have
16498 same component color value on location in graph. This is the default mode.
16501 Gray values are displayed on graph. Surrounding pixels values which are not
16502 present in video frame are drawn in gradient of 2 color components which are
16503 set by option @code{x} and @code{y}. The 3rd color component is static.
16506 Actual color components values present in video frame are displayed on graph.
16509 Similar as color2 but higher frequency of same values @code{x} and @code{y}
16510 on graph increases value of another color component, which is luminance by
16511 default values of @code{x} and @code{y}.
16514 Actual colors present in video frame are displayed on graph. If two different
16515 colors map to same position on graph then color with higher value of component
16516 not present in graph is picked.
16519 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
16520 component picked from radial gradient.
16524 Set which color component will be represented on X-axis. Default is @code{1}.
16527 Set which color component will be represented on Y-axis. Default is @code{2}.
16530 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
16531 of color component which represents frequency of (X, Y) location in graph.
16536 No envelope, this is default.
16539 Instant envelope, even darkest single pixel will be clearly highlighted.
16542 Hold maximum and minimum values presented in graph over time. This way you
16543 can still spot out of range values without constantly looking at vectorscope.
16546 Peak and instant envelope combined together.
16550 Set what kind of graticule to draw.
16558 Set graticule opacity.
16561 Set graticule flags.
16565 Draw graticule for white point.
16568 Draw graticule for black point.
16571 Draw color points short names.
16575 Set background opacity.
16577 @item lthreshold, l
16578 Set low threshold for color component not represented on X or Y axis.
16579 Values lower than this value will be ignored. Default is 0.
16580 Note this value is multiplied with actual max possible value one pixel component
16581 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
16584 @item hthreshold, h
16585 Set high threshold for color component not represented on X or Y axis.
16586 Values higher than this value will be ignored. Default is 1.
16587 Note this value is multiplied with actual max possible value one pixel component
16588 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
16589 is 0.9 * 255 = 230.
16591 @item colorspace, c
16592 Set what kind of colorspace to use when drawing graticule.
16601 @anchor{vidstabdetect}
16602 @section vidstabdetect
16604 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16605 @ref{vidstabtransform} for pass 2.
16607 This filter generates a file with relative translation and rotation
16608 transform information about subsequent frames, which is then used by
16609 the @ref{vidstabtransform} filter.
16611 To enable compilation of this filter you need to configure FFmpeg with
16612 @code{--enable-libvidstab}.
16614 This filter accepts the following options:
16618 Set the path to the file used to write the transforms information.
16619 Default value is @file{transforms.trf}.
16622 Set how shaky the video is and how quick the camera is. It accepts an
16623 integer in the range 1-10, a value of 1 means little shakiness, a
16624 value of 10 means strong shakiness. Default value is 5.
16627 Set the accuracy of the detection process. It must be a value in the
16628 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16629 accuracy. Default value is 15.
16632 Set stepsize of the search process. The region around minimum is
16633 scanned with 1 pixel resolution. Default value is 6.
16636 Set minimum contrast. Below this value a local measurement field is
16637 discarded. Must be a floating point value in the range 0-1. Default
16641 Set reference frame number for tripod mode.
16643 If enabled, the motion of the frames is compared to a reference frame
16644 in the filtered stream, identified by the specified number. The idea
16645 is to compensate all movements in a more-or-less static scene and keep
16646 the camera view absolutely still.
16648 If set to 0, it is disabled. The frames are counted starting from 1.
16651 Show fields and transforms in the resulting frames. It accepts an
16652 integer in the range 0-2. Default value is 0, which disables any
16656 @subsection Examples
16660 Use default values:
16666 Analyze strongly shaky movie and put the results in file
16667 @file{mytransforms.trf}:
16669 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16673 Visualize the result of internal transformations in the resulting
16676 vidstabdetect=show=1
16680 Analyze a video with medium shakiness using @command{ffmpeg}:
16682 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16686 @anchor{vidstabtransform}
16687 @section vidstabtransform
16689 Video stabilization/deshaking: pass 2 of 2,
16690 see @ref{vidstabdetect} for pass 1.
16692 Read a file with transform information for each frame and
16693 apply/compensate them. Together with the @ref{vidstabdetect}
16694 filter this can be used to deshake videos. See also
16695 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16696 the @ref{unsharp} filter, see below.
16698 To enable compilation of this filter you need to configure FFmpeg with
16699 @code{--enable-libvidstab}.
16701 @subsection Options
16705 Set path to the file used to read the transforms. Default value is
16706 @file{transforms.trf}.
16709 Set the number of frames (value*2 + 1) used for lowpass filtering the
16710 camera movements. Default value is 10.
16712 For example a number of 10 means that 21 frames are used (10 in the
16713 past and 10 in the future) to smoothen the motion in the video. A
16714 larger value leads to a smoother video, but limits the acceleration of
16715 the camera (pan/tilt movements). 0 is a special case where a static
16716 camera is simulated.
16719 Set the camera path optimization algorithm.
16721 Accepted values are:
16724 gaussian kernel low-pass filter on camera motion (default)
16726 averaging on transformations
16730 Set maximal number of pixels to translate frames. Default value is -1,
16734 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16735 value is -1, meaning no limit.
16738 Specify how to deal with borders that may be visible due to movement
16741 Available values are:
16744 keep image information from previous frame (default)
16746 fill the border black
16750 Invert transforms if set to 1. Default value is 0.
16753 Consider transforms as relative to previous frame if set to 1,
16754 absolute if set to 0. Default value is 0.
16757 Set percentage to zoom. A positive value will result in a zoom-in
16758 effect, a negative value in a zoom-out effect. Default value is 0 (no
16762 Set optimal zooming to avoid borders.
16764 Accepted values are:
16769 optimal static zoom value is determined (only very strong movements
16770 will lead to visible borders) (default)
16772 optimal adaptive zoom value is determined (no borders will be
16773 visible), see @option{zoomspeed}
16776 Note that the value given at zoom is added to the one calculated here.
16779 Set percent to zoom maximally each frame (enabled when
16780 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16784 Specify type of interpolation.
16786 Available values are:
16791 linear only horizontal
16793 linear in both directions (default)
16795 cubic in both directions (slow)
16799 Enable virtual tripod mode if set to 1, which is equivalent to
16800 @code{relative=0:smoothing=0}. Default value is 0.
16802 Use also @code{tripod} option of @ref{vidstabdetect}.
16805 Increase log verbosity if set to 1. Also the detected global motions
16806 are written to the temporary file @file{global_motions.trf}. Default
16810 @subsection Examples
16814 Use @command{ffmpeg} for a typical stabilization with default values:
16816 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16819 Note the use of the @ref{unsharp} filter which is always recommended.
16822 Zoom in a bit more and load transform data from a given file:
16824 vidstabtransform=zoom=5:input="mytransforms.trf"
16828 Smoothen the video even more:
16830 vidstabtransform=smoothing=30
16836 Flip the input video vertically.
16838 For example, to vertically flip a video with @command{ffmpeg}:
16840 ffmpeg -i in.avi -vf "vflip" out.avi
16845 Detect variable frame rate video.
16847 This filter tries to detect if the input is variable or constant frame rate.
16849 At end it will output number of frames detected as having variable delta pts,
16850 and ones with constant delta pts.
16851 If there was frames with variable delta, than it will also show min and max delta
16857 Make or reverse a natural vignetting effect.
16859 The filter accepts the following options:
16863 Set lens angle expression as a number of radians.
16865 The value is clipped in the @code{[0,PI/2]} range.
16867 Default value: @code{"PI/5"}
16871 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16875 Set forward/backward mode.
16877 Available modes are:
16880 The larger the distance from the central point, the darker the image becomes.
16883 The larger the distance from the central point, the brighter the image becomes.
16884 This can be used to reverse a vignette effect, though there is no automatic
16885 detection to extract the lens @option{angle} and other settings (yet). It can
16886 also be used to create a burning effect.
16889 Default value is @samp{forward}.
16892 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16894 It accepts the following values:
16897 Evaluate expressions only once during the filter initialization.
16900 Evaluate expressions for each incoming frame. This is way slower than the
16901 @samp{init} mode since it requires all the scalers to be re-computed, but it
16902 allows advanced dynamic expressions.
16905 Default value is @samp{init}.
16908 Set dithering to reduce the circular banding effects. Default is @code{1}
16912 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16913 Setting this value to the SAR of the input will make a rectangular vignetting
16914 following the dimensions of the video.
16916 Default is @code{1/1}.
16919 @subsection Expressions
16921 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16922 following parameters.
16927 input width and height
16930 the number of input frame, starting from 0
16933 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16934 @var{TB} units, NAN if undefined
16937 frame rate of the input video, NAN if the input frame rate is unknown
16940 the PTS (Presentation TimeStamp) of the filtered video frame,
16941 expressed in seconds, NAN if undefined
16944 time base of the input video
16948 @subsection Examples
16952 Apply simple strong vignetting effect:
16958 Make a flickering vignetting:
16960 vignette='PI/4+random(1)*PI/50':eval=frame
16965 @section vmafmotion
16967 Obtain the average vmaf motion score of a video.
16968 It is one of the component filters of VMAF.
16970 The obtained average motion score is printed through the logging system.
16972 In the below example the input file @file{ref.mpg} is being processed and score
16976 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16980 Stack input videos vertically.
16982 All streams must be of same pixel format and of same width.
16984 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16985 to create same output.
16987 The filter accept the following option:
16991 Set number of input streams. Default is 2.
16994 If set to 1, force the output to terminate when the shortest input
16995 terminates. Default value is 0.
17000 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17001 Deinterlacing Filter").
17003 Based on the process described by Martin Weston for BBC R&D, and
17004 implemented based on the de-interlace algorithm written by Jim
17005 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17006 uses filter coefficients calculated by BBC R&D.
17008 There are two sets of filter coefficients, so called "simple":
17009 and "complex". Which set of filter coefficients is used can
17010 be set by passing an optional parameter:
17014 Set the interlacing filter coefficients. Accepts one of the following values:
17018 Simple filter coefficient set.
17020 More-complex filter coefficient set.
17022 Default value is @samp{complex}.
17025 Specify which frames to deinterlace. Accept one of the following values:
17029 Deinterlace all frames,
17031 Only deinterlace frames marked as interlaced.
17034 Default value is @samp{all}.
17038 Video waveform monitor.
17040 The waveform monitor plots color component intensity. By default luminance
17041 only. Each column of the waveform corresponds to a column of pixels in the
17044 It accepts the following options:
17048 Can be either @code{row}, or @code{column}. Default is @code{column}.
17049 In row mode, the graph on the left side represents color component value 0 and
17050 the right side represents value = 255. In column mode, the top side represents
17051 color component value = 0 and bottom side represents value = 255.
17054 Set intensity. Smaller values are useful to find out how many values of the same
17055 luminance are distributed across input rows/columns.
17056 Default value is @code{0.04}. Allowed range is [0, 1].
17059 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
17060 In mirrored mode, higher values will be represented on the left
17061 side for @code{row} mode and at the top for @code{column} mode. Default is
17062 @code{1} (mirrored).
17066 It accepts the following values:
17069 Presents information identical to that in the @code{parade}, except
17070 that the graphs representing color components are superimposed directly
17073 This display mode makes it easier to spot relative differences or similarities
17074 in overlapping areas of the color components that are supposed to be identical,
17075 such as neutral whites, grays, or blacks.
17078 Display separate graph for the color components side by side in
17079 @code{row} mode or one below the other in @code{column} mode.
17082 Display separate graph for the color components side by side in
17083 @code{column} mode or one below the other in @code{row} mode.
17085 Using this display mode makes it easy to spot color casts in the highlights
17086 and shadows of an image, by comparing the contours of the top and the bottom
17087 graphs of each waveform. Since whites, grays, and blacks are characterized
17088 by exactly equal amounts of red, green, and blue, neutral areas of the picture
17089 should display three waveforms of roughly equal width/height. If not, the
17090 correction is easy to perform by making level adjustments the three waveforms.
17092 Default is @code{stack}.
17094 @item components, c
17095 Set which color components to display. Default is 1, which means only luminance
17096 or red color component if input is in RGB colorspace. If is set for example to
17097 7 it will display all 3 (if) available color components.
17102 No envelope, this is default.
17105 Instant envelope, minimum and maximum values presented in graph will be easily
17106 visible even with small @code{step} value.
17109 Hold minimum and maximum values presented in graph across time. This way you
17110 can still spot out of range values without constantly looking at waveforms.
17113 Peak and instant envelope combined together.
17119 No filtering, this is default.
17122 Luma and chroma combined together.
17125 Similar as above, but shows difference between blue and red chroma.
17128 Similar as above, but use different colors.
17131 Displays only chroma.
17134 Displays actual color value on waveform.
17137 Similar as above, but with luma showing frequency of chroma values.
17141 Set which graticule to display.
17145 Do not display graticule.
17148 Display green graticule showing legal broadcast ranges.
17151 Display orange graticule showing legal broadcast ranges.
17155 Set graticule opacity.
17158 Set graticule flags.
17162 Draw numbers above lines. By default enabled.
17165 Draw dots instead of lines.
17169 Set scale used for displaying graticule.
17176 Default is digital.
17179 Set background opacity.
17182 @section weave, doubleweave
17184 The @code{weave} takes a field-based video input and join
17185 each two sequential fields into single frame, producing a new double
17186 height clip with half the frame rate and half the frame count.
17188 The @code{doubleweave} works same as @code{weave} but without
17189 halving frame rate and frame count.
17191 It accepts the following option:
17195 Set first field. Available values are:
17199 Set the frame as top-field-first.
17202 Set the frame as bottom-field-first.
17206 @subsection Examples
17210 Interlace video using @ref{select} and @ref{separatefields} filter:
17212 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
17217 Apply the xBR high-quality magnification filter which is designed for pixel
17218 art. It follows a set of edge-detection rules, see
17219 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
17221 It accepts the following option:
17225 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
17226 @code{3xBR} and @code{4} for @code{4xBR}.
17227 Default is @code{3}.
17233 Deinterlace the input video ("yadif" means "yet another deinterlacing
17236 It accepts the following parameters:
17242 The interlacing mode to adopt. It accepts one of the following values:
17245 @item 0, send_frame
17246 Output one frame for each frame.
17247 @item 1, send_field
17248 Output one frame for each field.
17249 @item 2, send_frame_nospatial
17250 Like @code{send_frame}, but it skips the spatial interlacing check.
17251 @item 3, send_field_nospatial
17252 Like @code{send_field}, but it skips the spatial interlacing check.
17255 The default value is @code{send_frame}.
17258 The picture field parity assumed for the input interlaced video. It accepts one
17259 of the following values:
17263 Assume the top field is first.
17265 Assume the bottom field is first.
17267 Enable automatic detection of field parity.
17270 The default value is @code{auto}.
17271 If the interlacing is unknown or the decoder does not export this information,
17272 top field first will be assumed.
17275 Specify which frames to deinterlace. Accept one of the following
17280 Deinterlace all frames.
17281 @item 1, interlaced
17282 Only deinterlace frames marked as interlaced.
17285 The default value is @code{all}.
17290 Apply Zoom & Pan effect.
17292 This filter accepts the following options:
17296 Set the zoom expression. Default is 1.
17300 Set the x and y expression. Default is 0.
17303 Set the duration expression in number of frames.
17304 This sets for how many number of frames effect will last for
17305 single input image.
17308 Set the output image size, default is 'hd720'.
17311 Set the output frame rate, default is '25'.
17314 Each expression can contain the following constants:
17333 Output frame count.
17337 Last calculated 'x' and 'y' position from 'x' and 'y' expression
17338 for current input frame.
17342 'x' and 'y' of last output frame of previous input frame or 0 when there was
17343 not yet such frame (first input frame).
17346 Last calculated zoom from 'z' expression for current input frame.
17349 Last calculated zoom of last output frame of previous input frame.
17352 Number of output frames for current input frame. Calculated from 'd' expression
17353 for each input frame.
17356 number of output frames created for previous input frame
17359 Rational number: input width / input height
17362 sample aspect ratio
17365 display aspect ratio
17369 @subsection Examples
17373 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
17375 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
17379 Zoom-in up to 1.5 and pan always at center of picture:
17381 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17385 Same as above but without pausing:
17387 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
17393 Scale (resize) the input video, using the z.lib library:
17394 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
17395 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
17397 The zscale filter forces the output display aspect ratio to be the same
17398 as the input, by changing the output sample aspect ratio.
17400 If the input image format is different from the format requested by
17401 the next filter, the zscale filter will convert the input to the
17404 @subsection Options
17405 The filter accepts the following options.
17410 Set the output video dimension expression. Default value is the input
17413 If the @var{width} or @var{w} value is 0, the input width is used for
17414 the output. If the @var{height} or @var{h} value is 0, the input height
17415 is used for the output.
17417 If one and only one of the values is -n with n >= 1, the zscale filter
17418 will use a value that maintains the aspect ratio of the input image,
17419 calculated from the other specified dimension. After that it will,
17420 however, make sure that the calculated dimension is divisible by n and
17421 adjust the value if necessary.
17423 If both values are -n with n >= 1, the behavior will be identical to
17424 both values being set to 0 as previously detailed.
17426 See below for the list of accepted constants for use in the dimension
17430 Set the video size. For the syntax of this option, check the
17431 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17434 Set the dither type.
17436 Possible values are:
17441 @item error_diffusion
17447 Set the resize filter type.
17449 Possible values are:
17459 Default is bilinear.
17462 Set the color range.
17464 Possible values are:
17471 Default is same as input.
17474 Set the color primaries.
17476 Possible values are:
17486 Default is same as input.
17489 Set the transfer characteristics.
17491 Possible values are:
17505 Default is same as input.
17508 Set the colorspace matrix.
17510 Possible value are:
17521 Default is same as input.
17524 Set the input color range.
17526 Possible values are:
17533 Default is same as input.
17535 @item primariesin, pin
17536 Set the input color primaries.
17538 Possible values are:
17548 Default is same as input.
17550 @item transferin, tin
17551 Set the input transfer characteristics.
17553 Possible values are:
17564 Default is same as input.
17566 @item matrixin, min
17567 Set the input colorspace matrix.
17569 Possible value are:
17581 Set the output chroma location.
17583 Possible values are:
17594 @item chromalin, cin
17595 Set the input chroma location.
17597 Possible values are:
17609 Set the nominal peak luminance.
17612 The values of the @option{w} and @option{h} options are expressions
17613 containing the following constants:
17618 The input width and height
17622 These are the same as @var{in_w} and @var{in_h}.
17626 The output (scaled) width and height
17630 These are the same as @var{out_w} and @var{out_h}
17633 The same as @var{iw} / @var{ih}
17636 input sample aspect ratio
17639 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17643 horizontal and vertical input chroma subsample values. For example for the
17644 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17648 horizontal and vertical output chroma subsample values. For example for the
17649 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17655 @c man end VIDEO FILTERS
17657 @chapter Video Sources
17658 @c man begin VIDEO SOURCES
17660 Below is a description of the currently available video sources.
17664 Buffer video frames, and make them available to the filter chain.
17666 This source is mainly intended for a programmatic use, in particular
17667 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17669 It accepts the following parameters:
17674 Specify the size (width and height) of the buffered video frames. For the
17675 syntax of this option, check the
17676 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17679 The input video width.
17682 The input video height.
17685 A string representing the pixel format of the buffered video frames.
17686 It may be a number corresponding to a pixel format, or a pixel format
17690 Specify the timebase assumed by the timestamps of the buffered frames.
17693 Specify the frame rate expected for the video stream.
17695 @item pixel_aspect, sar
17696 The sample (pixel) aspect ratio of the input video.
17699 Specify the optional parameters to be used for the scale filter which
17700 is automatically inserted when an input change is detected in the
17701 input size or format.
17703 @item hw_frames_ctx
17704 When using a hardware pixel format, this should be a reference to an
17705 AVHWFramesContext describing input frames.
17710 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17713 will instruct the source to accept video frames with size 320x240 and
17714 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17715 square pixels (1:1 sample aspect ratio).
17716 Since the pixel format with name "yuv410p" corresponds to the number 6
17717 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17718 this example corresponds to:
17720 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17723 Alternatively, the options can be specified as a flat string, but this
17724 syntax is deprecated:
17726 @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}]
17730 Create a pattern generated by an elementary cellular automaton.
17732 The initial state of the cellular automaton can be defined through the
17733 @option{filename} and @option{pattern} options. If such options are
17734 not specified an initial state is created randomly.
17736 At each new frame a new row in the video is filled with the result of
17737 the cellular automaton next generation. The behavior when the whole
17738 frame is filled is defined by the @option{scroll} option.
17740 This source accepts the following options:
17744 Read the initial cellular automaton state, i.e. the starting row, from
17745 the specified file.
17746 In the file, each non-whitespace character is considered an alive
17747 cell, a newline will terminate the row, and further characters in the
17748 file will be ignored.
17751 Read the initial cellular automaton state, i.e. the starting row, from
17752 the specified string.
17754 Each non-whitespace character in the string is considered an alive
17755 cell, a newline will terminate the row, and further characters in the
17756 string will be ignored.
17759 Set the video rate, that is the number of frames generated per second.
17762 @item random_fill_ratio, ratio
17763 Set the random fill ratio for the initial cellular automaton row. It
17764 is a floating point number value ranging from 0 to 1, defaults to
17767 This option is ignored when a file or a pattern is specified.
17769 @item random_seed, seed
17770 Set the seed for filling randomly the initial row, must be an integer
17771 included between 0 and UINT32_MAX. If not specified, or if explicitly
17772 set to -1, the filter will try to use a good random seed on a best
17776 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17777 Default value is 110.
17780 Set the size of the output video. For the syntax of this option, check the
17781 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17783 If @option{filename} or @option{pattern} is specified, the size is set
17784 by default to the width of the specified initial state row, and the
17785 height is set to @var{width} * PHI.
17787 If @option{size} is set, it must contain the width of the specified
17788 pattern string, and the specified pattern will be centered in the
17791 If a filename or a pattern string is not specified, the size value
17792 defaults to "320x518" (used for a randomly generated initial state).
17795 If set to 1, scroll the output upward when all the rows in the output
17796 have been already filled. If set to 0, the new generated row will be
17797 written over the top row just after the bottom row is filled.
17800 @item start_full, full
17801 If set to 1, completely fill the output with generated rows before
17802 outputting the first frame.
17803 This is the default behavior, for disabling set the value to 0.
17806 If set to 1, stitch the left and right row edges together.
17807 This is the default behavior, for disabling set the value to 0.
17810 @subsection Examples
17814 Read the initial state from @file{pattern}, and specify an output of
17817 cellauto=f=pattern:s=200x400
17821 Generate a random initial row with a width of 200 cells, with a fill
17824 cellauto=ratio=2/3:s=200x200
17828 Create a pattern generated by rule 18 starting by a single alive cell
17829 centered on an initial row with width 100:
17831 cellauto=p=@@:s=100x400:full=0:rule=18
17835 Specify a more elaborated initial pattern:
17837 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17842 @anchor{coreimagesrc}
17843 @section coreimagesrc
17844 Video source generated on GPU using Apple's CoreImage API on OSX.
17846 This video source is a specialized version of the @ref{coreimage} video filter.
17847 Use a core image generator at the beginning of the applied filterchain to
17848 generate the content.
17850 The coreimagesrc video source accepts the following options:
17852 @item list_generators
17853 List all available generators along with all their respective options as well as
17854 possible minimum and maximum values along with the default values.
17856 list_generators=true
17860 Specify the size of the sourced video. For the syntax of this option, check the
17861 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17862 The default value is @code{320x240}.
17865 Specify the frame rate of the sourced video, as the number of frames
17866 generated per second. It has to be a string in the format
17867 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17868 number or a valid video frame rate abbreviation. The default value is
17872 Set the sample aspect ratio of the sourced video.
17875 Set the duration of the sourced video. See
17876 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17877 for the accepted syntax.
17879 If not specified, or the expressed duration is negative, the video is
17880 supposed to be generated forever.
17883 Additionally, all options of the @ref{coreimage} video filter are accepted.
17884 A complete filterchain can be used for further processing of the
17885 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17886 and examples for details.
17888 @subsection Examples
17893 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17894 given as complete and escaped command-line for Apple's standard bash shell:
17896 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17898 This example is equivalent to the QRCode example of @ref{coreimage} without the
17899 need for a nullsrc video source.
17903 @section mandelbrot
17905 Generate a Mandelbrot set fractal, and progressively zoom towards the
17906 point specified with @var{start_x} and @var{start_y}.
17908 This source accepts the following options:
17913 Set the terminal pts value. Default value is 400.
17916 Set the terminal scale value.
17917 Must be a floating point value. Default value is 0.3.
17920 Set the inner coloring mode, that is the algorithm used to draw the
17921 Mandelbrot fractal internal region.
17923 It shall assume one of the following values:
17928 Show time until convergence.
17930 Set color based on point closest to the origin of the iterations.
17935 Default value is @var{mincol}.
17938 Set the bailout value. Default value is 10.0.
17941 Set the maximum of iterations performed by the rendering
17942 algorithm. Default value is 7189.
17945 Set outer coloring mode.
17946 It shall assume one of following values:
17948 @item iteration_count
17949 Set iteration cound mode.
17950 @item normalized_iteration_count
17951 set normalized iteration count mode.
17953 Default value is @var{normalized_iteration_count}.
17956 Set frame rate, expressed as number of frames per second. Default
17960 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17961 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17964 Set the initial scale value. Default value is 3.0.
17967 Set the initial x position. Must be a floating point value between
17968 -100 and 100. Default value is -0.743643887037158704752191506114774.
17971 Set the initial y position. Must be a floating point value between
17972 -100 and 100. Default value is -0.131825904205311970493132056385139.
17977 Generate various test patterns, as generated by the MPlayer test filter.
17979 The size of the generated video is fixed, and is 256x256.
17980 This source is useful in particular for testing encoding features.
17982 This source accepts the following options:
17987 Specify the frame rate of the sourced video, as the number of frames
17988 generated per second. It has to be a string in the format
17989 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17990 number or a valid video frame rate abbreviation. The default value is
17994 Set the duration of the sourced video. See
17995 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17996 for the accepted syntax.
17998 If not specified, or the expressed duration is negative, the video is
17999 supposed to be generated forever.
18003 Set the number or the name of the test to perform. Supported tests are:
18019 Default value is "all", which will cycle through the list of all tests.
18024 mptestsrc=t=dc_luma
18027 will generate a "dc_luma" test pattern.
18029 @section frei0r_src
18031 Provide a frei0r source.
18033 To enable compilation of this filter you need to install the frei0r
18034 header and configure FFmpeg with @code{--enable-frei0r}.
18036 This source accepts the following parameters:
18041 The size of the video to generate. For the syntax of this option, check the
18042 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18045 The framerate of the generated video. It may be a string of the form
18046 @var{num}/@var{den} or a frame rate abbreviation.
18049 The name to the frei0r source to load. For more information regarding frei0r and
18050 how to set the parameters, read the @ref{frei0r} section in the video filters
18053 @item filter_params
18054 A '|'-separated list of parameters to pass to the frei0r source.
18058 For example, to generate a frei0r partik0l source with size 200x200
18059 and frame rate 10 which is overlaid on the overlay filter main input:
18061 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
18066 Generate a life pattern.
18068 This source is based on a generalization of John Conway's life game.
18070 The sourced input represents a life grid, each pixel represents a cell
18071 which can be in one of two possible states, alive or dead. Every cell
18072 interacts with its eight neighbours, which are the cells that are
18073 horizontally, vertically, or diagonally adjacent.
18075 At each interaction the grid evolves according to the adopted rule,
18076 which specifies the number of neighbor alive cells which will make a
18077 cell stay alive or born. The @option{rule} option allows one to specify
18080 This source accepts the following options:
18084 Set the file from which to read the initial grid state. In the file,
18085 each non-whitespace character is considered an alive cell, and newline
18086 is used to delimit the end of each row.
18088 If this option is not specified, the initial grid is generated
18092 Set the video rate, that is the number of frames generated per second.
18095 @item random_fill_ratio, ratio
18096 Set the random fill ratio for the initial random grid. It is a
18097 floating point number value ranging from 0 to 1, defaults to 1/PHI.
18098 It is ignored when a file is specified.
18100 @item random_seed, seed
18101 Set the seed for filling the initial random grid, must be an integer
18102 included between 0 and UINT32_MAX. If not specified, or if explicitly
18103 set to -1, the filter will try to use a good random seed on a best
18109 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
18110 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
18111 @var{NS} specifies the number of alive neighbor cells which make a
18112 live cell stay alive, and @var{NB} the number of alive neighbor cells
18113 which make a dead cell to become alive (i.e. to "born").
18114 "s" and "b" can be used in place of "S" and "B", respectively.
18116 Alternatively a rule can be specified by an 18-bits integer. The 9
18117 high order bits are used to encode the next cell state if it is alive
18118 for each number of neighbor alive cells, the low order bits specify
18119 the rule for "borning" new cells. Higher order bits encode for an
18120 higher number of neighbor cells.
18121 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
18122 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
18124 Default value is "S23/B3", which is the original Conway's game of life
18125 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
18126 cells, and will born a new cell if there are three alive cells around
18130 Set the size of the output video. For the syntax of this option, check the
18131 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18133 If @option{filename} is specified, the size is set by default to the
18134 same size of the input file. If @option{size} is set, it must contain
18135 the size specified in the input file, and the initial grid defined in
18136 that file is centered in the larger resulting area.
18138 If a filename is not specified, the size value defaults to "320x240"
18139 (used for a randomly generated initial grid).
18142 If set to 1, stitch the left and right grid edges together, and the
18143 top and bottom edges also. Defaults to 1.
18146 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
18147 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
18148 value from 0 to 255.
18151 Set the color of living (or new born) cells.
18154 Set the color of dead cells. If @option{mold} is set, this is the first color
18155 used to represent a dead cell.
18158 Set mold color, for definitely dead and moldy cells.
18160 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
18161 ffmpeg-utils manual,ffmpeg-utils}.
18164 @subsection Examples
18168 Read a grid from @file{pattern}, and center it on a grid of size
18171 life=f=pattern:s=300x300
18175 Generate a random grid of size 200x200, with a fill ratio of 2/3:
18177 life=ratio=2/3:s=200x200
18181 Specify a custom rule for evolving a randomly generated grid:
18187 Full example with slow death effect (mold) using @command{ffplay}:
18189 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
18196 @anchor{haldclutsrc}
18199 @anchor{pal100bars}
18200 @anchor{rgbtestsrc}
18202 @anchor{smptehdbars}
18205 @anchor{yuvtestsrc}
18206 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
18208 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
18210 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
18212 The @code{color} source provides an uniformly colored input.
18214 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
18215 @ref{haldclut} filter.
18217 The @code{nullsrc} source returns unprocessed video frames. It is
18218 mainly useful to be employed in analysis / debugging tools, or as the
18219 source for filters which ignore the input data.
18221 The @code{pal75bars} source generates a color bars pattern, based on
18222 EBU PAL recommendations with 75% color levels.
18224 The @code{pal100bars} source generates a color bars pattern, based on
18225 EBU PAL recommendations with 100% color levels.
18227 The @code{rgbtestsrc} source generates an RGB test pattern useful for
18228 detecting RGB vs BGR issues. You should see a red, green and blue
18229 stripe from top to bottom.
18231 The @code{smptebars} source generates a color bars pattern, based on
18232 the SMPTE Engineering Guideline EG 1-1990.
18234 The @code{smptehdbars} source generates a color bars pattern, based on
18235 the SMPTE RP 219-2002.
18237 The @code{testsrc} source generates a test video pattern, showing a
18238 color pattern, a scrolling gradient and a timestamp. This is mainly
18239 intended for testing purposes.
18241 The @code{testsrc2} source is similar to testsrc, but supports more
18242 pixel formats instead of just @code{rgb24}. This allows using it as an
18243 input for other tests without requiring a format conversion.
18245 The @code{yuvtestsrc} source generates an YUV test pattern. You should
18246 see a y, cb and cr stripe from top to bottom.
18248 The sources accept the following parameters:
18253 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
18254 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
18255 pixels to be used as identity matrix for 3D lookup tables. Each component is
18256 coded on a @code{1/(N*N)} scale.
18259 Specify the color of the source, only available in the @code{color}
18260 source. For the syntax of this option, check the
18261 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18264 Specify the size of the sourced video. For the syntax of this option, check the
18265 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18266 The default value is @code{320x240}.
18268 This option is not available with the @code{allrgb}, @code{allyuv}, and
18269 @code{haldclutsrc} filters.
18272 Specify the frame rate of the sourced video, as the number of frames
18273 generated per second. It has to be a string in the format
18274 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
18275 number or a valid video frame rate abbreviation. The default value is
18279 Set the duration of the sourced video. See
18280 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18281 for the accepted syntax.
18283 If not specified, or the expressed duration is negative, the video is
18284 supposed to be generated forever.
18287 Set the sample aspect ratio of the sourced video.
18290 Specify the alpha (opacity) of the background, only available in the
18291 @code{testsrc2} source. The value must be between 0 (fully transparent) and
18292 255 (fully opaque, the default).
18295 Set the number of decimals to show in the timestamp, only available in the
18296 @code{testsrc} source.
18298 The displayed timestamp value will correspond to the original
18299 timestamp value multiplied by the power of 10 of the specified
18300 value. Default value is 0.
18303 @subsection Examples
18307 Generate a video with a duration of 5.3 seconds, with size
18308 176x144 and a frame rate of 10 frames per second:
18310 testsrc=duration=5.3:size=qcif:rate=10
18314 The following graph description will generate a red source
18315 with an opacity of 0.2, with size "qcif" and a frame rate of 10
18318 color=c=red@@0.2:s=qcif:r=10
18322 If the input content is to be ignored, @code{nullsrc} can be used. The
18323 following command generates noise in the luminance plane by employing
18324 the @code{geq} filter:
18326 nullsrc=s=256x256, geq=random(1)*255:128:128
18330 @subsection Commands
18332 The @code{color} source supports the following commands:
18336 Set the color of the created image. Accepts the same syntax of the
18337 corresponding @option{color} option.
18342 Generate video using an OpenCL program.
18347 OpenCL program source file.
18350 Kernel name in program.
18353 Size of frames to generate. This must be set.
18356 Pixel format to use for the generated frames. This must be set.
18359 Number of frames generated every second. Default value is '25'.
18363 For details of how the program loading works, see the @ref{program_opencl}
18370 Generate a colour ramp by setting pixel values from the position of the pixel
18371 in the output image. (Note that this will work with all pixel formats, but
18372 the generated output will not be the same.)
18374 __kernel void ramp(__write_only image2d_t dst,
18375 unsigned int index)
18377 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18380 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
18382 write_imagef(dst, loc, val);
18387 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
18389 __kernel void sierpinski_carpet(__write_only image2d_t dst,
18390 unsigned int index)
18392 int2 loc = (int2)(get_global_id(0), get_global_id(1));
18394 float4 value = 0.0f;
18395 int x = loc.x + index;
18396 int y = loc.y + index;
18397 while (x > 0 || y > 0) {
18398 if (x % 3 == 1 && y % 3 == 1) {
18406 write_imagef(dst, loc, value);
18412 @c man end VIDEO SOURCES
18414 @chapter Video Sinks
18415 @c man begin VIDEO SINKS
18417 Below is a description of the currently available video sinks.
18419 @section buffersink
18421 Buffer video frames, and make them available to the end of the filter
18424 This sink is mainly intended for programmatic use, in particular
18425 through the interface defined in @file{libavfilter/buffersink.h}
18426 or the options system.
18428 It accepts a pointer to an AVBufferSinkContext structure, which
18429 defines the incoming buffers' formats, to be passed as the opaque
18430 parameter to @code{avfilter_init_filter} for initialization.
18434 Null video sink: do absolutely nothing with the input video. It is
18435 mainly useful as a template and for use in analysis / debugging
18438 @c man end VIDEO SINKS
18440 @chapter Multimedia Filters
18441 @c man begin MULTIMEDIA FILTERS
18443 Below is a description of the currently available multimedia filters.
18447 Convert input audio to a video output, displaying the audio bit scope.
18449 The filter accepts the following options:
18453 Set frame rate, expressed as number of frames per second. Default
18457 Specify the video size for the output. For the syntax of this option, check the
18458 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18459 Default value is @code{1024x256}.
18462 Specify list of colors separated by space or by '|' which will be used to
18463 draw channels. Unrecognized or missing colors will be replaced
18467 @section ahistogram
18469 Convert input audio to a video output, displaying the volume histogram.
18471 The filter accepts the following options:
18475 Specify how histogram is calculated.
18477 It accepts the following values:
18480 Use single histogram for all channels.
18482 Use separate histogram for each channel.
18484 Default is @code{single}.
18487 Set frame rate, expressed as number of frames per second. Default
18491 Specify the video size for the output. For the syntax of this option, check the
18492 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18493 Default value is @code{hd720}.
18498 It accepts the following values:
18509 reverse logarithmic
18511 Default is @code{log}.
18514 Set amplitude scale.
18516 It accepts the following values:
18523 Default is @code{log}.
18526 Set how much frames to accumulate in histogram.
18527 Defauls is 1. Setting this to -1 accumulates all frames.
18530 Set histogram ratio of window height.
18533 Set sonogram sliding.
18535 It accepts the following values:
18538 replace old rows with new ones.
18540 scroll from top to bottom.
18542 Default is @code{replace}.
18545 @section aphasemeter
18547 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
18548 representing mean phase of current audio frame. A video output can also be produced and is
18549 enabled by default. The audio is passed through as first output.
18551 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
18552 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
18553 and @code{1} means channels are in phase.
18555 The filter accepts the following options, all related to its video output:
18559 Set the output frame rate. Default value is @code{25}.
18562 Set the video size for the output. For the syntax of this option, check the
18563 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18564 Default value is @code{800x400}.
18569 Specify the red, green, blue contrast. Default values are @code{2},
18570 @code{7} and @code{1}.
18571 Allowed range is @code{[0, 255]}.
18574 Set color which will be used for drawing median phase. If color is
18575 @code{none} which is default, no median phase value will be drawn.
18578 Enable video output. Default is enabled.
18581 @section avectorscope
18583 Convert input audio to a video output, representing the audio vector
18586 The filter is used to measure the difference between channels of stereo
18587 audio stream. A monoaural signal, consisting of identical left and right
18588 signal, results in straight vertical line. Any stereo separation is visible
18589 as a deviation from this line, creating a Lissajous figure.
18590 If the straight (or deviation from it) but horizontal line appears this
18591 indicates that the left and right channels are out of phase.
18593 The filter accepts the following options:
18597 Set the vectorscope mode.
18599 Available values are:
18602 Lissajous rotated by 45 degrees.
18605 Same as above but not rotated.
18608 Shape resembling half of circle.
18611 Default value is @samp{lissajous}.
18614 Set the video size for the output. For the syntax of this option, check the
18615 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18616 Default value is @code{400x400}.
18619 Set the output frame rate. Default value is @code{25}.
18625 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18626 @code{160}, @code{80} and @code{255}.
18627 Allowed range is @code{[0, 255]}.
18633 Specify the red, green, blue and alpha fade. Default values are @code{15},
18634 @code{10}, @code{5} and @code{5}.
18635 Allowed range is @code{[0, 255]}.
18638 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18639 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18642 Set the vectorscope drawing mode.
18644 Available values are:
18647 Draw dot for each sample.
18650 Draw line between previous and current sample.
18653 Default value is @samp{dot}.
18656 Specify amplitude scale of audio samples.
18658 Available values are:
18674 Swap left channel axis with right channel axis.
18684 Mirror only x axis.
18687 Mirror only y axis.
18695 @subsection Examples
18699 Complete example using @command{ffplay}:
18701 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18702 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18706 @section bench, abench
18708 Benchmark part of a filtergraph.
18710 The filter accepts the following options:
18714 Start or stop a timer.
18716 Available values are:
18719 Get the current time, set it as frame metadata (using the key
18720 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18723 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18724 the input frame metadata to get the time difference. Time difference, average,
18725 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18726 @code{min}) are then printed. The timestamps are expressed in seconds.
18730 @subsection Examples
18734 Benchmark @ref{selectivecolor} filter:
18736 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18742 Concatenate audio and video streams, joining them together one after the
18745 The filter works on segments of synchronized video and audio streams. All
18746 segments must have the same number of streams of each type, and that will
18747 also be the number of streams at output.
18749 The filter accepts the following options:
18754 Set the number of segments. Default is 2.
18757 Set the number of output video streams, that is also the number of video
18758 streams in each segment. Default is 1.
18761 Set the number of output audio streams, that is also the number of audio
18762 streams in each segment. Default is 0.
18765 Activate unsafe mode: do not fail if segments have a different format.
18769 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18770 @var{a} audio outputs.
18772 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18773 segment, in the same order as the outputs, then the inputs for the second
18776 Related streams do not always have exactly the same duration, for various
18777 reasons including codec frame size or sloppy authoring. For that reason,
18778 related synchronized streams (e.g. a video and its audio track) should be
18779 concatenated at once. The concat filter will use the duration of the longest
18780 stream in each segment (except the last one), and if necessary pad shorter
18781 audio streams with silence.
18783 For this filter to work correctly, all segments must start at timestamp 0.
18785 All corresponding streams must have the same parameters in all segments; the
18786 filtering system will automatically select a common pixel format for video
18787 streams, and a common sample format, sample rate and channel layout for
18788 audio streams, but other settings, such as resolution, must be converted
18789 explicitly by the user.
18791 Different frame rates are acceptable but will result in variable frame rate
18792 at output; be sure to configure the output file to handle it.
18794 @subsection Examples
18798 Concatenate an opening, an episode and an ending, all in bilingual version
18799 (video in stream 0, audio in streams 1 and 2):
18801 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18802 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18803 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18804 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18808 Concatenate two parts, handling audio and video separately, using the
18809 (a)movie sources, and adjusting the resolution:
18811 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18812 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18813 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18815 Note that a desync will happen at the stitch if the audio and video streams
18816 do not have exactly the same duration in the first file.
18820 @subsection Commands
18822 This filter supports the following commands:
18825 Close the current segment and step to the next one
18828 @section drawgraph, adrawgraph
18830 Draw a graph using input video or audio metadata.
18832 It accepts the following parameters:
18836 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18839 Set 1st foreground color expression.
18842 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18845 Set 2nd foreground color expression.
18848 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18851 Set 3rd foreground color expression.
18854 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18857 Set 4th foreground color expression.
18860 Set minimal value of metadata value.
18863 Set maximal value of metadata value.
18866 Set graph background color. Default is white.
18871 Available values for mode is:
18878 Default is @code{line}.
18883 Available values for slide is:
18886 Draw new frame when right border is reached.
18889 Replace old columns with new ones.
18892 Scroll from right to left.
18895 Scroll from left to right.
18898 Draw single picture.
18901 Default is @code{frame}.
18904 Set size of graph video. For the syntax of this option, check the
18905 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18906 The default value is @code{900x256}.
18908 The foreground color expressions can use the following variables:
18911 Minimal value of metadata value.
18914 Maximal value of metadata value.
18917 Current metadata key value.
18920 The color is defined as 0xAABBGGRR.
18923 Example using metadata from @ref{signalstats} filter:
18925 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18928 Example using metadata from @ref{ebur128} filter:
18930 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18936 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18937 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18938 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18939 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18941 The filter also has a video output (see the @var{video} option) with a real
18942 time graph to observe the loudness evolution. The graphic contains the logged
18943 message mentioned above, so it is not printed anymore when this option is set,
18944 unless the verbose logging is set. The main graphing area contains the
18945 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18946 the momentary loudness (400 milliseconds).
18948 More information about the Loudness Recommendation EBU R128 on
18949 @url{http://tech.ebu.ch/loudness}.
18951 The filter accepts the following options:
18956 Activate the video output. The audio stream is passed unchanged whether this
18957 option is set or no. The video stream will be the first output stream if
18958 activated. Default is @code{0}.
18961 Set the video size. This option is for video only. For the syntax of this
18963 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18964 Default and minimum resolution is @code{640x480}.
18967 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18968 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18969 other integer value between this range is allowed.
18972 Set metadata injection. If set to @code{1}, the audio input will be segmented
18973 into 100ms output frames, each of them containing various loudness information
18974 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18976 Default is @code{0}.
18979 Force the frame logging level.
18981 Available values are:
18984 information logging level
18986 verbose logging level
18989 By default, the logging level is set to @var{info}. If the @option{video} or
18990 the @option{metadata} options are set, it switches to @var{verbose}.
18995 Available modes can be cumulated (the option is a @code{flag} type). Possible
18999 Disable any peak mode (default).
19001 Enable sample-peak mode.
19003 Simple peak mode looking for the higher sample value. It logs a message
19004 for sample-peak (identified by @code{SPK}).
19006 Enable true-peak mode.
19008 If enabled, the peak lookup is done on an over-sampled version of the input
19009 stream for better peak accuracy. It logs a message for true-peak.
19010 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
19011 This mode requires a build with @code{libswresample}.
19015 Treat mono input files as "dual mono". If a mono file is intended for playback
19016 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
19017 If set to @code{true}, this option will compensate for this effect.
19018 Multi-channel input files are not affected by this option.
19021 Set a specific pan law to be used for the measurement of dual mono files.
19022 This parameter is optional, and has a default value of -3.01dB.
19025 @subsection Examples
19029 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
19031 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
19035 Run an analysis with @command{ffmpeg}:
19037 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
19041 @section interleave, ainterleave
19043 Temporally interleave frames from several inputs.
19045 @code{interleave} works with video inputs, @code{ainterleave} with audio.
19047 These filters read frames from several inputs and send the oldest
19048 queued frame to the output.
19050 Input streams must have well defined, monotonically increasing frame
19053 In order to submit one frame to output, these filters need to enqueue
19054 at least one frame for each input, so they cannot work in case one
19055 input is not yet terminated and will not receive incoming frames.
19057 For example consider the case when one input is a @code{select} filter
19058 which always drops input frames. The @code{interleave} filter will keep
19059 reading from that input, but it will never be able to send new frames
19060 to output until the input sends an end-of-stream signal.
19062 Also, depending on inputs synchronization, the filters will drop
19063 frames in case one input receives more frames than the other ones, and
19064 the queue is already filled.
19066 These filters accept the following options:
19070 Set the number of different inputs, it is 2 by default.
19073 @subsection Examples
19077 Interleave frames belonging to different streams using @command{ffmpeg}:
19079 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
19083 Add flickering blur effect:
19085 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
19089 @section metadata, ametadata
19091 Manipulate frame metadata.
19093 This filter accepts the following options:
19097 Set mode of operation of the filter.
19099 Can be one of the following:
19103 If both @code{value} and @code{key} is set, select frames
19104 which have such metadata. If only @code{key} is set, select
19105 every frame that has such key in metadata.
19108 Add new metadata @code{key} and @code{value}. If key is already available
19112 Modify value of already present key.
19115 If @code{value} is set, delete only keys that have such value.
19116 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
19120 Print key and its value if metadata was found. If @code{key} is not set print all
19121 metadata values available in frame.
19125 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
19128 Set metadata value which will be used. This option is mandatory for
19129 @code{modify} and @code{add} mode.
19132 Which function to use when comparing metadata value and @code{value}.
19134 Can be one of following:
19138 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
19141 Values are interpreted as strings, returns true if metadata value starts with
19142 the @code{value} option string.
19145 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
19148 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
19151 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
19154 Values are interpreted as floats, returns true if expression from option @code{expr}
19159 Set expression which is used when @code{function} is set to @code{expr}.
19160 The expression is evaluated through the eval API and can contain the following
19165 Float representation of @code{value} from metadata key.
19168 Float representation of @code{value} as supplied by user in @code{value} option.
19172 If specified in @code{print} mode, output is written to the named file. Instead of
19173 plain filename any writable url can be specified. Filename ``-'' is a shorthand
19174 for standard output. If @code{file} option is not set, output is written to the log
19175 with AV_LOG_INFO loglevel.
19179 @subsection Examples
19183 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
19186 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
19189 Print silencedetect output to file @file{metadata.txt}.
19191 silencedetect,ametadata=mode=print:file=metadata.txt
19194 Direct all metadata to a pipe with file descriptor 4.
19196 metadata=mode=print:file='pipe\:4'
19200 @section perms, aperms
19202 Set read/write permissions for the output frames.
19204 These filters are mainly aimed at developers to test direct path in the
19205 following filter in the filtergraph.
19207 The filters accept the following options:
19211 Select the permissions mode.
19213 It accepts the following values:
19216 Do nothing. This is the default.
19218 Set all the output frames read-only.
19220 Set all the output frames directly writable.
19222 Make the frame read-only if writable, and writable if read-only.
19224 Set each output frame read-only or writable randomly.
19228 Set the seed for the @var{random} mode, must be an integer included between
19229 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
19230 @code{-1}, the filter will try to use a good random seed on a best effort
19234 Note: in case of auto-inserted filter between the permission filter and the
19235 following one, the permission might not be received as expected in that
19236 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
19237 perms/aperms filter can avoid this problem.
19239 @section realtime, arealtime
19241 Slow down filtering to match real time approximately.
19243 These filters will pause the filtering for a variable amount of time to
19244 match the output rate with the input timestamps.
19245 They are similar to the @option{re} option to @code{ffmpeg}.
19247 They accept the following options:
19251 Time limit for the pauses. Any pause longer than that will be considered
19252 a timestamp discontinuity and reset the timer. Default is 2 seconds.
19256 @section select, aselect
19258 Select frames to pass in output.
19260 This filter accepts the following options:
19265 Set expression, which is evaluated for each input frame.
19267 If the expression is evaluated to zero, the frame is discarded.
19269 If the evaluation result is negative or NaN, the frame is sent to the
19270 first output; otherwise it is sent to the output with index
19271 @code{ceil(val)-1}, assuming that the input index starts from 0.
19273 For example a value of @code{1.2} corresponds to the output with index
19274 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
19277 Set the number of outputs. The output to which to send the selected
19278 frame is based on the result of the evaluation. Default value is 1.
19281 The expression can contain the following constants:
19285 The (sequential) number of the filtered frame, starting from 0.
19288 The (sequential) number of the selected frame, starting from 0.
19290 @item prev_selected_n
19291 The sequential number of the last selected frame. It's NAN if undefined.
19294 The timebase of the input timestamps.
19297 The PTS (Presentation TimeStamp) of the filtered video frame,
19298 expressed in @var{TB} units. It's NAN if undefined.
19301 The PTS of the filtered video frame,
19302 expressed in seconds. It's NAN if undefined.
19305 The PTS of the previously filtered video frame. It's NAN if undefined.
19307 @item prev_selected_pts
19308 The PTS of the last previously filtered video frame. It's NAN if undefined.
19310 @item prev_selected_t
19311 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
19314 The PTS of the first video frame in the video. It's NAN if undefined.
19317 The time of the first video frame in the video. It's NAN if undefined.
19319 @item pict_type @emph{(video only)}
19320 The type of the filtered frame. It can assume one of the following
19332 @item interlace_type @emph{(video only)}
19333 The frame interlace type. It can assume one of the following values:
19336 The frame is progressive (not interlaced).
19338 The frame is top-field-first.
19340 The frame is bottom-field-first.
19343 @item consumed_sample_n @emph{(audio only)}
19344 the number of selected samples before the current frame
19346 @item samples_n @emph{(audio only)}
19347 the number of samples in the current frame
19349 @item sample_rate @emph{(audio only)}
19350 the input sample rate
19353 This is 1 if the filtered frame is a key-frame, 0 otherwise.
19356 the position in the file of the filtered frame, -1 if the information
19357 is not available (e.g. for synthetic video)
19359 @item scene @emph{(video only)}
19360 value between 0 and 1 to indicate a new scene; a low value reflects a low
19361 probability for the current frame to introduce a new scene, while a higher
19362 value means the current frame is more likely to be one (see the example below)
19364 @item concatdec_select
19365 The concat demuxer can select only part of a concat input file by setting an
19366 inpoint and an outpoint, but the output packets may not be entirely contained
19367 in the selected interval. By using this variable, it is possible to skip frames
19368 generated by the concat demuxer which are not exactly contained in the selected
19371 This works by comparing the frame pts against the @var{lavf.concat.start_time}
19372 and the @var{lavf.concat.duration} packet metadata values which are also
19373 present in the decoded frames.
19375 The @var{concatdec_select} variable is -1 if the frame pts is at least
19376 start_time and either the duration metadata is missing or the frame pts is less
19377 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
19380 That basically means that an input frame is selected if its pts is within the
19381 interval set by the concat demuxer.
19385 The default value of the select expression is "1".
19387 @subsection Examples
19391 Select all frames in input:
19396 The example above is the same as:
19408 Select only I-frames:
19410 select='eq(pict_type\,I)'
19414 Select one frame every 100:
19416 select='not(mod(n\,100))'
19420 Select only frames contained in the 10-20 time interval:
19422 select=between(t\,10\,20)
19426 Select only I-frames contained in the 10-20 time interval:
19428 select=between(t\,10\,20)*eq(pict_type\,I)
19432 Select frames with a minimum distance of 10 seconds:
19434 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
19438 Use aselect to select only audio frames with samples number > 100:
19440 aselect='gt(samples_n\,100)'
19444 Create a mosaic of the first scenes:
19446 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
19449 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
19453 Send even and odd frames to separate outputs, and compose them:
19455 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
19459 Select useful frames from an ffconcat file which is using inpoints and
19460 outpoints but where the source files are not intra frame only.
19462 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
19466 @section sendcmd, asendcmd
19468 Send commands to filters in the filtergraph.
19470 These filters read commands to be sent to other filters in the
19473 @code{sendcmd} must be inserted between two video filters,
19474 @code{asendcmd} must be inserted between two audio filters, but apart
19475 from that they act the same way.
19477 The specification of commands can be provided in the filter arguments
19478 with the @var{commands} option, or in a file specified by the
19479 @var{filename} option.
19481 These filters accept the following options:
19484 Set the commands to be read and sent to the other filters.
19486 Set the filename of the commands to be read and sent to the other
19490 @subsection Commands syntax
19492 A commands description consists of a sequence of interval
19493 specifications, comprising a list of commands to be executed when a
19494 particular event related to that interval occurs. The occurring event
19495 is typically the current frame time entering or leaving a given time
19498 An interval is specified by the following syntax:
19500 @var{START}[-@var{END}] @var{COMMANDS};
19503 The time interval is specified by the @var{START} and @var{END} times.
19504 @var{END} is optional and defaults to the maximum time.
19506 The current frame time is considered within the specified interval if
19507 it is included in the interval [@var{START}, @var{END}), that is when
19508 the time is greater or equal to @var{START} and is lesser than
19511 @var{COMMANDS} consists of a sequence of one or more command
19512 specifications, separated by ",", relating to that interval. The
19513 syntax of a command specification is given by:
19515 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
19518 @var{FLAGS} is optional and specifies the type of events relating to
19519 the time interval which enable sending the specified command, and must
19520 be a non-null sequence of identifier flags separated by "+" or "|" and
19521 enclosed between "[" and "]".
19523 The following flags are recognized:
19526 The command is sent when the current frame timestamp enters the
19527 specified interval. In other words, the command is sent when the
19528 previous frame timestamp was not in the given interval, and the
19532 The command is sent when the current frame timestamp leaves the
19533 specified interval. In other words, the command is sent when the
19534 previous frame timestamp was in the given interval, and the
19538 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
19541 @var{TARGET} specifies the target of the command, usually the name of
19542 the filter class or a specific filter instance name.
19544 @var{COMMAND} specifies the name of the command for the target filter.
19546 @var{ARG} is optional and specifies the optional list of argument for
19547 the given @var{COMMAND}.
19549 Between one interval specification and another, whitespaces, or
19550 sequences of characters starting with @code{#} until the end of line,
19551 are ignored and can be used to annotate comments.
19553 A simplified BNF description of the commands specification syntax
19556 @var{COMMAND_FLAG} ::= "enter" | "leave"
19557 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
19558 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
19559 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
19560 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
19561 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
19564 @subsection Examples
19568 Specify audio tempo change at second 4:
19570 asendcmd=c='4.0 atempo tempo 1.5',atempo
19574 Target a specific filter instance:
19576 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
19580 Specify a list of drawtext and hue commands in a file.
19582 # show text in the interval 5-10
19583 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
19584 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
19586 # desaturate the image in the interval 15-20
19587 15.0-20.0 [enter] hue s 0,
19588 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
19590 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
19592 # apply an exponential saturation fade-out effect, starting from time 25
19593 25 [enter] hue s exp(25-t)
19596 A filtergraph allowing to read and process the above command list
19597 stored in a file @file{test.cmd}, can be specified with:
19599 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
19604 @section setpts, asetpts
19606 Change the PTS (presentation timestamp) of the input frames.
19608 @code{setpts} works on video frames, @code{asetpts} on audio frames.
19610 This filter accepts the following options:
19615 The expression which is evaluated for each frame to construct its timestamp.
19619 The expression is evaluated through the eval API and can contain the following
19623 @item FRAME_RATE, FR
19624 frame rate, only defined for constant frame-rate video
19627 The presentation timestamp in input
19630 The count of the input frame for video or the number of consumed samples,
19631 not including the current frame for audio, starting from 0.
19633 @item NB_CONSUMED_SAMPLES
19634 The number of consumed samples, not including the current frame (only
19637 @item NB_SAMPLES, S
19638 The number of samples in the current frame (only audio)
19640 @item SAMPLE_RATE, SR
19641 The audio sample rate.
19644 The PTS of the first frame.
19647 the time in seconds of the first frame
19650 State whether the current frame is interlaced.
19653 the time in seconds of the current frame
19656 original position in the file of the frame, or undefined if undefined
19657 for the current frame
19660 The previous input PTS.
19663 previous input time in seconds
19666 The previous output PTS.
19669 previous output time in seconds
19672 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19676 The wallclock (RTC) time at the start of the movie in microseconds.
19679 The timebase of the input timestamps.
19683 @subsection Examples
19687 Start counting PTS from zero
19689 setpts=PTS-STARTPTS
19693 Apply fast motion effect:
19699 Apply slow motion effect:
19705 Set fixed rate of 25 frames per second:
19711 Set fixed rate 25 fps with some jitter:
19713 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19717 Apply an offset of 10 seconds to the input PTS:
19723 Generate timestamps from a "live source" and rebase onto the current timebase:
19725 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19729 Generate timestamps by counting samples:
19738 Force color range for the output video frame.
19740 The @code{setrange} filter marks the color range property for the
19741 output frames. It does not change the input frame, but only sets the
19742 corresponding property, which affects how the frame is treated by
19745 The filter accepts the following options:
19750 Available values are:
19754 Keep the same color range property.
19756 @item unspecified, unknown
19757 Set the color range as unspecified.
19759 @item limited, tv, mpeg
19760 Set the color range as limited.
19762 @item full, pc, jpeg
19763 Set the color range as full.
19767 @section settb, asettb
19769 Set the timebase to use for the output frames timestamps.
19770 It is mainly useful for testing timebase configuration.
19772 It accepts the following parameters:
19777 The expression which is evaluated into the output timebase.
19781 The value for @option{tb} is an arithmetic expression representing a
19782 rational. The expression can contain the constants "AVTB" (the default
19783 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19784 audio only). Default value is "intb".
19786 @subsection Examples
19790 Set the timebase to 1/25:
19796 Set the timebase to 1/10:
19802 Set the timebase to 1001/1000:
19808 Set the timebase to 2*intb:
19814 Set the default timebase value:
19821 Convert input audio to a video output representing frequency spectrum
19822 logarithmically using Brown-Puckette constant Q transform algorithm with
19823 direct frequency domain coefficient calculation (but the transform itself
19824 is not really constant Q, instead the Q factor is actually variable/clamped),
19825 with musical tone scale, from E0 to D#10.
19827 The filter accepts the following options:
19831 Specify the video size for the output. It must be even. For the syntax of this option,
19832 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19833 Default value is @code{1920x1080}.
19836 Set the output frame rate. Default value is @code{25}.
19839 Set the bargraph height. It must be even. Default value is @code{-1} which
19840 computes the bargraph height automatically.
19843 Set the axis height. It must be even. Default value is @code{-1} which computes
19844 the axis height automatically.
19847 Set the sonogram height. It must be even. Default value is @code{-1} which
19848 computes the sonogram height automatically.
19851 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19852 instead. Default value is @code{1}.
19854 @item sono_v, volume
19855 Specify the sonogram volume expression. It can contain variables:
19858 the @var{bar_v} evaluated expression
19859 @item frequency, freq, f
19860 the frequency where it is evaluated
19861 @item timeclamp, tc
19862 the value of @var{timeclamp} option
19866 @item a_weighting(f)
19867 A-weighting of equal loudness
19868 @item b_weighting(f)
19869 B-weighting of equal loudness
19870 @item c_weighting(f)
19871 C-weighting of equal loudness.
19873 Default value is @code{16}.
19875 @item bar_v, volume2
19876 Specify the bargraph volume expression. It can contain variables:
19879 the @var{sono_v} evaluated expression
19880 @item frequency, freq, f
19881 the frequency where it is evaluated
19882 @item timeclamp, tc
19883 the value of @var{timeclamp} option
19887 @item a_weighting(f)
19888 A-weighting of equal loudness
19889 @item b_weighting(f)
19890 B-weighting of equal loudness
19891 @item c_weighting(f)
19892 C-weighting of equal loudness.
19894 Default value is @code{sono_v}.
19896 @item sono_g, gamma
19897 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19898 higher gamma makes the spectrum having more range. Default value is @code{3}.
19899 Acceptable range is @code{[1, 7]}.
19901 @item bar_g, gamma2
19902 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19906 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19907 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19909 @item timeclamp, tc
19910 Specify the transform timeclamp. At low frequency, there is trade-off between
19911 accuracy in time domain and frequency domain. If timeclamp is lower,
19912 event in time domain is represented more accurately (such as fast bass drum),
19913 otherwise event in frequency domain is represented more accurately
19914 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19917 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19918 limits future samples by applying asymmetric windowing in time domain, useful
19919 when low latency is required. Accepted range is @code{[0, 1]}.
19922 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19923 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19926 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19927 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19930 This option is deprecated and ignored.
19933 Specify the transform length in time domain. Use this option to control accuracy
19934 trade-off between time domain and frequency domain at every frequency sample.
19935 It can contain variables:
19937 @item frequency, freq, f
19938 the frequency where it is evaluated
19939 @item timeclamp, tc
19940 the value of @var{timeclamp} option.
19942 Default value is @code{384*tc/(384+tc*f)}.
19945 Specify the transform count for every video frame. Default value is @code{6}.
19946 Acceptable range is @code{[1, 30]}.
19949 Specify the transform count for every single pixel. Default value is @code{0},
19950 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19953 Specify font file for use with freetype to draw the axis. If not specified,
19954 use embedded font. Note that drawing with font file or embedded font is not
19955 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19959 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19960 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19963 Specify font color expression. This is arithmetic expression that should return
19964 integer value 0xRRGGBB. It can contain variables:
19966 @item frequency, freq, f
19967 the frequency where it is evaluated
19968 @item timeclamp, tc
19969 the value of @var{timeclamp} option
19974 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19975 @item r(x), g(x), b(x)
19976 red, green, and blue value of intensity x.
19978 Default value is @code{st(0, (midi(f)-59.5)/12);
19979 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19980 r(1-ld(1)) + b(ld(1))}.
19983 Specify image file to draw the axis. This option override @var{fontfile} and
19984 @var{fontcolor} option.
19987 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19988 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19989 Default value is @code{1}.
19992 Set colorspace. The accepted values are:
19995 Unspecified (default)
20004 BT.470BG or BT.601-6 625
20007 SMPTE-170M or BT.601-6 525
20013 BT.2020 with non-constant luminance
20018 Set spectrogram color scheme. This is list of floating point values with format
20019 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
20020 The default is @code{1|0.5|0|0|0.5|1}.
20024 @subsection Examples
20028 Playing audio while showing the spectrum:
20030 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
20034 Same as above, but with frame rate 30 fps:
20036 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
20040 Playing at 1280x720:
20042 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
20046 Disable sonogram display:
20052 A1 and its harmonics: A1, A2, (near)E3, A3:
20054 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),
20055 asplit[a][out1]; [a] showcqt [out0]'
20059 Same as above, but with more accuracy in frequency domain:
20061 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),
20062 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
20068 bar_v=10:sono_v=bar_v*a_weighting(f)
20072 Custom gamma, now spectrum is linear to the amplitude.
20078 Custom tlength equation:
20080 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)))'
20084 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
20086 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
20090 Custom font using fontconfig:
20092 font='Courier New,Monospace,mono|bold'
20096 Custom frequency range with custom axis using image file:
20098 axisfile=myaxis.png:basefreq=40:endfreq=10000
20104 Convert input audio to video output representing the audio power spectrum.
20105 Audio amplitude is on Y-axis while frequency is on X-axis.
20107 The filter accepts the following options:
20111 Specify size of video. For the syntax of this option, check the
20112 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20113 Default is @code{1024x512}.
20117 This set how each frequency bin will be represented.
20119 It accepts the following values:
20125 Default is @code{bar}.
20128 Set amplitude scale.
20130 It accepts the following values:
20144 Default is @code{log}.
20147 Set frequency scale.
20149 It accepts the following values:
20158 Reverse logarithmic scale.
20160 Default is @code{lin}.
20165 It accepts the following values:
20181 Default is @code{w2048}
20184 Set windowing function.
20186 It accepts the following values:
20208 Default is @code{hanning}.
20211 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20212 which means optimal overlap for selected window function will be picked.
20215 Set time averaging. Setting this to 0 will display current maximal peaks.
20216 Default is @code{1}, which means time averaging is disabled.
20219 Specify list of colors separated by space or by '|' which will be used to
20220 draw channel frequencies. Unrecognized or missing colors will be replaced
20224 Set channel display mode.
20226 It accepts the following values:
20231 Default is @code{combined}.
20234 Set minimum amplitude used in @code{log} amplitude scaler.
20238 @anchor{showspectrum}
20239 @section showspectrum
20241 Convert input audio to a video output, representing the audio frequency
20244 The filter accepts the following options:
20248 Specify the video size for the output. For the syntax of this option, check the
20249 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20250 Default value is @code{640x512}.
20253 Specify how the spectrum should slide along the window.
20255 It accepts the following values:
20258 the samples start again on the left when they reach the right
20260 the samples scroll from right to left
20262 frames are only produced when the samples reach the right
20264 the samples scroll from left to right
20267 Default value is @code{replace}.
20270 Specify display mode.
20272 It accepts the following values:
20275 all channels are displayed in the same row
20277 all channels are displayed in separate rows
20280 Default value is @samp{combined}.
20283 Specify display color mode.
20285 It accepts the following values:
20288 each channel is displayed in a separate color
20290 each channel is displayed using the same color scheme
20292 each channel is displayed using the rainbow color scheme
20294 each channel is displayed using the moreland color scheme
20296 each channel is displayed using the nebulae color scheme
20298 each channel is displayed using the fire color scheme
20300 each channel is displayed using the fiery color scheme
20302 each channel is displayed using the fruit color scheme
20304 each channel is displayed using the cool color scheme
20307 Default value is @samp{channel}.
20310 Specify scale used for calculating intensity color values.
20312 It accepts the following values:
20317 square root, default
20328 Default value is @samp{sqrt}.
20331 Set saturation modifier for displayed colors. Negative values provide
20332 alternative color scheme. @code{0} is no saturation at all.
20333 Saturation must be in [-10.0, 10.0] range.
20334 Default value is @code{1}.
20337 Set window function.
20339 It accepts the following values:
20363 Default value is @code{hann}.
20366 Set orientation of time vs frequency axis. Can be @code{vertical} or
20367 @code{horizontal}. Default is @code{vertical}.
20370 Set ratio of overlap window. Default value is @code{0}.
20371 When value is @code{1} overlap is set to recommended size for specific
20372 window function currently used.
20375 Set scale gain for calculating intensity color values.
20376 Default value is @code{1}.
20379 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
20382 Set color rotation, must be in [-1.0, 1.0] range.
20383 Default value is @code{0}.
20386 The usage is very similar to the showwaves filter; see the examples in that
20389 @subsection Examples
20393 Large window with logarithmic color scaling:
20395 showspectrum=s=1280x480:scale=log
20399 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
20401 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20402 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
20406 @section showspectrumpic
20408 Convert input audio to a single video frame, representing the audio frequency
20411 The filter accepts the following options:
20415 Specify the video size for the output. For the syntax of this option, check the
20416 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20417 Default value is @code{4096x2048}.
20420 Specify display mode.
20422 It accepts the following values:
20425 all channels are displayed in the same row
20427 all channels are displayed in separate rows
20429 Default value is @samp{combined}.
20432 Specify display color mode.
20434 It accepts the following values:
20437 each channel is displayed in a separate color
20439 each channel is displayed using the same color scheme
20441 each channel is displayed using the rainbow color scheme
20443 each channel is displayed using the moreland color scheme
20445 each channel is displayed using the nebulae color scheme
20447 each channel is displayed using the fire color scheme
20449 each channel is displayed using the fiery color scheme
20451 each channel is displayed using the fruit color scheme
20453 each channel is displayed using the cool color scheme
20455 Default value is @samp{intensity}.
20458 Specify scale used for calculating intensity color values.
20460 It accepts the following values:
20465 square root, default
20475 Default value is @samp{log}.
20478 Set saturation modifier for displayed colors. Negative values provide
20479 alternative color scheme. @code{0} is no saturation at all.
20480 Saturation must be in [-10.0, 10.0] range.
20481 Default value is @code{1}.
20484 Set window function.
20486 It accepts the following values:
20509 Default value is @code{hann}.
20512 Set orientation of time vs frequency axis. Can be @code{vertical} or
20513 @code{horizontal}. Default is @code{vertical}.
20516 Set scale gain for calculating intensity color values.
20517 Default value is @code{1}.
20520 Draw time and frequency axes and legends. Default is enabled.
20523 Set color rotation, must be in [-1.0, 1.0] range.
20524 Default value is @code{0}.
20527 @subsection Examples
20531 Extract an audio spectrogram of a whole audio track
20532 in a 1024x1024 picture using @command{ffmpeg}:
20534 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
20538 @section showvolume
20540 Convert input audio volume to a video output.
20542 The filter accepts the following options:
20549 Set border width, allowed range is [0, 5]. Default is 1.
20552 Set channel width, allowed range is [80, 8192]. Default is 400.
20555 Set channel height, allowed range is [1, 900]. Default is 20.
20558 Set fade, allowed range is [0, 1]. Default is 0.95.
20561 Set volume color expression.
20563 The expression can use the following variables:
20567 Current max volume of channel in dB.
20573 Current channel number, starting from 0.
20577 If set, displays channel names. Default is enabled.
20580 If set, displays volume values. Default is enabled.
20583 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
20584 default is @code{h}.
20587 Set step size, allowed range is [0, 5]. Default is 0, which means
20591 Set background opacity, allowed range is [0, 1]. Default is 0.
20594 Set metering mode, can be peak: @code{p} or rms: @code{r},
20595 default is @code{p}.
20598 Set display scale, can be linear: @code{lin} or log: @code{log},
20599 default is @code{lin}.
20603 If set to > 0., display a line for the max level
20604 in the previous seconds.
20605 default is disabled: @code{0.}
20608 The color of the max line. Use when @code{dm} option is set to > 0.
20609 default is: @code{orange}
20614 Convert input audio to a video output, representing the samples waves.
20616 The filter accepts the following options:
20620 Specify the video size for the output. For the syntax of this option, check the
20621 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20622 Default value is @code{600x240}.
20627 Available values are:
20630 Draw a point for each sample.
20633 Draw a vertical line for each sample.
20636 Draw a point for each sample and a line between them.
20639 Draw a centered vertical line for each sample.
20642 Default value is @code{point}.
20645 Set the number of samples which are printed on the same column. A
20646 larger value will decrease the frame rate. Must be a positive
20647 integer. This option can be set only if the value for @var{rate}
20648 is not explicitly specified.
20651 Set the (approximate) output frame rate. This is done by setting the
20652 option @var{n}. Default value is "25".
20654 @item split_channels
20655 Set if channels should be drawn separately or overlap. Default value is 0.
20658 Set colors separated by '|' which are going to be used for drawing of each channel.
20661 Set amplitude scale.
20663 Available values are:
20681 Set the draw mode. This is mostly useful to set for high @var{n}.
20683 Available values are:
20686 Scale pixel values for each drawn sample.
20689 Draw every sample directly.
20692 Default value is @code{scale}.
20695 @subsection Examples
20699 Output the input file audio and the corresponding video representation
20702 amovie=a.mp3,asplit[out0],showwaves[out1]
20706 Create a synthetic signal and show it with showwaves, forcing a
20707 frame rate of 30 frames per second:
20709 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20713 @section showwavespic
20715 Convert input audio to a single video frame, representing the samples waves.
20717 The filter accepts the following options:
20721 Specify the video size for the output. For the syntax of this option, check the
20722 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20723 Default value is @code{600x240}.
20725 @item split_channels
20726 Set if channels should be drawn separately or overlap. Default value is 0.
20729 Set colors separated by '|' which are going to be used for drawing of each channel.
20732 Set amplitude scale.
20734 Available values are:
20752 @subsection Examples
20756 Extract a channel split representation of the wave form of a whole audio track
20757 in a 1024x800 picture using @command{ffmpeg}:
20759 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20763 @section sidedata, asidedata
20765 Delete frame side data, or select frames based on it.
20767 This filter accepts the following options:
20771 Set mode of operation of the filter.
20773 Can be one of the following:
20777 Select every frame with side data of @code{type}.
20780 Delete side data of @code{type}. If @code{type} is not set, delete all side
20786 Set side data type used with all modes. Must be set for @code{select} mode. For
20787 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20788 in @file{libavutil/frame.h}. For example, to choose
20789 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20793 @section spectrumsynth
20795 Sythesize audio from 2 input video spectrums, first input stream represents
20796 magnitude across time and second represents phase across time.
20797 The filter will transform from frequency domain as displayed in videos back
20798 to time domain as presented in audio output.
20800 This filter is primarily created for reversing processed @ref{showspectrum}
20801 filter outputs, but can synthesize sound from other spectrograms too.
20802 But in such case results are going to be poor if the phase data is not
20803 available, because in such cases phase data need to be recreated, usually
20804 its just recreated from random noise.
20805 For best results use gray only output (@code{channel} color mode in
20806 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20807 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20808 @code{data} option. Inputs videos should generally use @code{fullframe}
20809 slide mode as that saves resources needed for decoding video.
20811 The filter accepts the following options:
20815 Specify sample rate of output audio, the sample rate of audio from which
20816 spectrum was generated may differ.
20819 Set number of channels represented in input video spectrums.
20822 Set scale which was used when generating magnitude input spectrum.
20823 Can be @code{lin} or @code{log}. Default is @code{log}.
20826 Set slide which was used when generating inputs spectrums.
20827 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20828 Default is @code{fullframe}.
20831 Set window function used for resynthesis.
20834 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20835 which means optimal overlap for selected window function will be picked.
20838 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20839 Default is @code{vertical}.
20842 @subsection Examples
20846 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20847 then resynthesize videos back to audio with spectrumsynth:
20849 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
20850 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
20851 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20855 @section split, asplit
20857 Split input into several identical outputs.
20859 @code{asplit} works with audio input, @code{split} with video.
20861 The filter accepts a single parameter which specifies the number of outputs. If
20862 unspecified, it defaults to 2.
20864 @subsection Examples
20868 Create two separate outputs from the same input:
20870 [in] split [out0][out1]
20874 To create 3 or more outputs, you need to specify the number of
20877 [in] asplit=3 [out0][out1][out2]
20881 Create two separate outputs from the same input, one cropped and
20884 [in] split [splitout1][splitout2];
20885 [splitout1] crop=100:100:0:0 [cropout];
20886 [splitout2] pad=200:200:100:100 [padout];
20890 Create 5 copies of the input audio with @command{ffmpeg}:
20892 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20898 Receive commands sent through a libzmq client, and forward them to
20899 filters in the filtergraph.
20901 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20902 must be inserted between two video filters, @code{azmq} between two
20903 audio filters. Both are capable to send messages to any filter type.
20905 To enable these filters you need to install the libzmq library and
20906 headers and configure FFmpeg with @code{--enable-libzmq}.
20908 For more information about libzmq see:
20909 @url{http://www.zeromq.org/}
20911 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20912 receives messages sent through a network interface defined by the
20913 @option{bind_address} (or the abbreviation "@option{b}") option.
20914 Default value of this option is @file{tcp://localhost:5555}. You may
20915 want to alter this value to your needs, but do not forget to escape any
20916 ':' signs (see @ref{filtergraph escaping}).
20918 The received message must be in the form:
20920 @var{TARGET} @var{COMMAND} [@var{ARG}]
20923 @var{TARGET} specifies the target of the command, usually the name of
20924 the filter class or a specific filter instance name. The default
20925 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
20926 but you can override this by using the @samp{filter_name@@id} syntax
20927 (see @ref{Filtergraph syntax}).
20929 @var{COMMAND} specifies the name of the command for the target filter.
20931 @var{ARG} is optional and specifies the optional argument list for the
20932 given @var{COMMAND}.
20934 Upon reception, the message is processed and the corresponding command
20935 is injected into the filtergraph. Depending on the result, the filter
20936 will send a reply to the client, adopting the format:
20938 @var{ERROR_CODE} @var{ERROR_REASON}
20942 @var{MESSAGE} is optional.
20944 @subsection Examples
20946 Look at @file{tools/zmqsend} for an example of a zmq client which can
20947 be used to send commands processed by these filters.
20949 Consider the following filtergraph generated by @command{ffplay}.
20950 In this example the last overlay filter has an instance name. All other
20951 filters will have default instance names.
20954 ffplay -dumpgraph 1 -f lavfi "
20955 color=s=100x100:c=red [l];
20956 color=s=100x100:c=blue [r];
20957 nullsrc=s=200x100, zmq [bg];
20958 [bg][l] overlay [bg+l];
20959 [bg+l][r] overlay@@my=x=100 "
20962 To change the color of the left side of the video, the following
20963 command can be used:
20965 echo Parsed_color_0 c yellow | tools/zmqsend
20968 To change the right side:
20970 echo Parsed_color_1 c pink | tools/zmqsend
20973 To change the position of the right side:
20975 echo overlay@@my x 150 | tools/zmqsend
20979 @c man end MULTIMEDIA FILTERS
20981 @chapter Multimedia Sources
20982 @c man begin MULTIMEDIA SOURCES
20984 Below is a description of the currently available multimedia sources.
20988 This is the same as @ref{movie} source, except it selects an audio
20994 Read audio and/or video stream(s) from a movie container.
20996 It accepts the following parameters:
21000 The name of the resource to read (not necessarily a file; it can also be a
21001 device or a stream accessed through some protocol).
21003 @item format_name, f
21004 Specifies the format assumed for the movie to read, and can be either
21005 the name of a container or an input device. If not specified, the
21006 format is guessed from @var{movie_name} or by probing.
21008 @item seek_point, sp
21009 Specifies the seek point in seconds. The frames will be output
21010 starting from this seek point. The parameter is evaluated with
21011 @code{av_strtod}, so the numerical value may be suffixed by an IS
21012 postfix. The default value is "0".
21015 Specifies the streams to read. Several streams can be specified,
21016 separated by "+". The source will then have as many outputs, in the
21017 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
21018 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
21019 respectively the default (best suited) video and audio stream. Default
21020 is "dv", or "da" if the filter is called as "amovie".
21022 @item stream_index, si
21023 Specifies the index of the video stream to read. If the value is -1,
21024 the most suitable video stream will be automatically selected. The default
21025 value is "-1". Deprecated. If the filter is called "amovie", it will select
21026 audio instead of video.
21029 Specifies how many times to read the stream in sequence.
21030 If the value is 0, the stream will be looped infinitely.
21031 Default value is "1".
21033 Note that when the movie is looped the source timestamps are not
21034 changed, so it will generate non monotonically increasing timestamps.
21036 @item discontinuity
21037 Specifies the time difference between frames above which the point is
21038 considered a timestamp discontinuity which is removed by adjusting the later
21042 It allows overlaying a second video on top of the main input of
21043 a filtergraph, as shown in this graph:
21045 input -----------> deltapts0 --> overlay --> output
21048 movie --> scale--> deltapts1 -------+
21050 @subsection Examples
21054 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
21055 on top of the input labelled "in":
21057 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
21058 [in] setpts=PTS-STARTPTS [main];
21059 [main][over] overlay=16:16 [out]
21063 Read from a video4linux2 device, and overlay it on top of the input
21066 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
21067 [in] setpts=PTS-STARTPTS [main];
21068 [main][over] overlay=16:16 [out]
21072 Read the first video stream and the audio stream with id 0x81 from
21073 dvd.vob; the video is connected to the pad named "video" and the audio is
21074 connected to the pad named "audio":
21076 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
21080 @subsection Commands
21082 Both movie and amovie support the following commands:
21085 Perform seek using "av_seek_frame".
21086 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
21089 @var{stream_index}: If stream_index is -1, a default
21090 stream is selected, and @var{timestamp} is automatically converted
21091 from AV_TIME_BASE units to the stream specific time_base.
21093 @var{timestamp}: Timestamp in AVStream.time_base units
21094 or, if no stream is specified, in AV_TIME_BASE units.
21096 @var{flags}: Flags which select direction and seeking mode.
21100 Get movie duration in AV_TIME_BASE units.
21104 @c man end MULTIMEDIA SOURCES