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 Changing options at runtime with a command
318 Some options can be changed during the operation of the filter using
319 a command. These options are marked 'T' on the output of
320 @command{ffmpeg} @option{-h filter=<name of filter>}.
321 The name of the command is the name of the option and the argument is
325 @chapter Options for filters with several inputs (framesync)
326 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
328 Some filters with several inputs support a common set of options.
329 These options can only be set by name, not with the short notation.
333 The action to take when EOF is encountered on the secondary input; it accepts
334 one of the following values:
338 Repeat the last frame (the default).
342 Pass the main input through.
346 If set to 1, force the output to terminate when the shortest input
347 terminates. Default value is 0.
350 If set to 1, force the filter to extend the last frame of secondary streams
351 until the end of the primary stream. A value of 0 disables this behavior.
355 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
357 @chapter Audio Filters
358 @c man begin AUDIO FILTERS
360 When you configure your FFmpeg build, you can disable any of the
361 existing filters using @code{--disable-filters}.
362 The configure output will show the audio filters included in your
365 Below is a description of the currently available audio filters.
369 A compressor is mainly used to reduce the dynamic range of a signal.
370 Especially modern music is mostly compressed at a high ratio to
371 improve the overall loudness. It's done to get the highest attention
372 of a listener, "fatten" the sound and bring more "power" to the track.
373 If a signal is compressed too much it may sound dull or "dead"
374 afterwards or it may start to "pump" (which could be a powerful effect
375 but can also destroy a track completely).
376 The right compression is the key to reach a professional sound and is
377 the high art of mixing and mastering. Because of its complex settings
378 it may take a long time to get the right feeling for this kind of effect.
380 Compression is done by detecting the volume above a chosen level
381 @code{threshold} and dividing it by the factor set with @code{ratio}.
382 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
383 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
384 the signal would cause distortion of the waveform the reduction can be
385 levelled over the time. This is done by setting "Attack" and "Release".
386 @code{attack} determines how long the signal has to rise above the threshold
387 before any reduction will occur and @code{release} sets the time the signal
388 has to fall below the threshold to reduce the reduction again. Shorter signals
389 than the chosen attack time will be left untouched.
390 The overall reduction of the signal can be made up afterwards with the
391 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
392 raising the makeup to this level results in a signal twice as loud than the
393 source. To gain a softer entry in the compression the @code{knee} flattens the
394 hard edge at the threshold in the range of the chosen decibels.
396 The filter accepts the following options:
400 Set input gain. Default is 1. Range is between 0.015625 and 64.
403 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
404 Default is @code{downward}.
407 If a signal of stream rises above this level it will affect the gain
409 By default it is 0.125. Range is between 0.00097563 and 1.
412 Set a ratio by which the signal is reduced. 1:2 means that if the level
413 rose 4dB above the threshold, it will be only 2dB above after the reduction.
414 Default is 2. Range is between 1 and 20.
417 Amount of milliseconds the signal has to rise above the threshold before gain
418 reduction starts. Default is 20. Range is between 0.01 and 2000.
421 Amount of milliseconds the signal has to fall below the threshold before
422 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
425 Set the amount by how much signal will be amplified after processing.
426 Default is 1. Range is from 1 to 64.
429 Curve the sharp knee around the threshold to enter gain reduction more softly.
430 Default is 2.82843. Range is between 1 and 8.
433 Choose if the @code{average} level between all channels of input stream
434 or the louder(@code{maximum}) channel of input stream affects the
435 reduction. Default is @code{average}.
438 Should the exact signal be taken in case of @code{peak} or an RMS one in case
439 of @code{rms}. Default is @code{rms} which is mostly smoother.
442 How much to use compressed signal in output. Default is 1.
443 Range is between 0 and 1.
448 This filter supports the all above options as @ref{commands}.
451 Simple audio dynamic range compression/expansion filter.
453 The filter accepts the following options:
457 Set contrast. Default is 33. Allowed range is between 0 and 100.
462 Copy the input audio source unchanged to the output. This is mainly useful for
467 Apply cross fade from one input audio stream to another input audio stream.
468 The cross fade is applied for specified duration near the end of first stream.
470 The filter accepts the following options:
474 Specify the number of samples for which the cross fade effect has to last.
475 At the end of the cross fade effect the first input audio will be completely
476 silent. Default is 44100.
479 Specify the duration of the cross fade effect. See
480 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
481 for the accepted syntax.
482 By default the duration is determined by @var{nb_samples}.
483 If set this option is used instead of @var{nb_samples}.
486 Should first stream end overlap with second stream start. Default is enabled.
489 Set curve for cross fade transition for first stream.
492 Set curve for cross fade transition for second stream.
494 For description of available curve types see @ref{afade} filter description.
501 Cross fade from one input to another:
503 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
507 Cross fade from one input to another but without overlapping:
509 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
514 Split audio stream into several bands.
516 This filter splits audio stream into two or more frequency ranges.
517 Summing all streams back will give flat output.
519 The filter accepts the following options:
523 Set split frequencies. Those must be positive and increasing.
526 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
527 Default is @var{4th}.
532 Reduce audio bit resolution.
534 This filter is bit crusher with enhanced functionality. A bit crusher
535 is used to audibly reduce number of bits an audio signal is sampled
536 with. This doesn't change the bit depth at all, it just produces the
537 effect. Material reduced in bit depth sounds more harsh and "digital".
538 This filter is able to even round to continuous values instead of discrete
540 Additionally it has a D/C offset which results in different crushing of
541 the lower and the upper half of the signal.
542 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
544 Another feature of this filter is the logarithmic mode.
545 This setting switches from linear distances between bits to logarithmic ones.
546 The result is a much more "natural" sounding crusher which doesn't gate low
547 signals for example. The human ear has a logarithmic perception,
548 so this kind of crushing is much more pleasant.
549 Logarithmic crushing is also able to get anti-aliased.
551 The filter accepts the following options:
567 Can be linear: @code{lin} or logarithmic: @code{log}.
576 Set sample reduction.
579 Enable LFO. By default disabled.
590 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
594 Remove impulsive noise from input audio.
596 Samples detected as impulsive noise are replaced by interpolated samples using
597 autoregressive modelling.
601 Set window size, in milliseconds. Allowed range is from @code{10} to
602 @code{100}. Default value is @code{55} milliseconds.
603 This sets size of window which will be processed at once.
606 Set window overlap, in percentage of window size. Allowed range is from
607 @code{50} to @code{95}. Default value is @code{75} percent.
608 Setting this to a very high value increases impulsive noise removal but makes
609 whole process much slower.
612 Set autoregression order, in percentage of window size. Allowed range is from
613 @code{0} to @code{25}. Default value is @code{2} percent. This option also
614 controls quality of interpolated samples using neighbour good samples.
617 Set threshold value. Allowed range is from @code{1} to @code{100}.
618 Default value is @code{2}.
619 This controls the strength of impulsive noise which is going to be removed.
620 The lower value, the more samples will be detected as impulsive noise.
623 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
624 @code{10}. Default value is @code{2}.
625 If any two samples detected as noise are spaced less than this value then any
626 sample between those two samples will be also detected as noise.
631 It accepts the following values:
634 Select overlap-add method. Even not interpolated samples are slightly
635 changed with this method.
638 Select overlap-save method. Not interpolated samples remain unchanged.
641 Default value is @code{a}.
645 Remove clipped samples from input audio.
647 Samples detected as clipped are replaced by interpolated samples using
648 autoregressive modelling.
652 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
653 Default value is @code{55} milliseconds.
654 This sets size of window which will be processed at once.
657 Set window overlap, in percentage of window size. Allowed range is from @code{50}
658 to @code{95}. Default value is @code{75} percent.
661 Set autoregression order, in percentage of window size. Allowed range is from
662 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
663 quality of interpolated samples using neighbour good samples.
666 Set threshold value. Allowed range is from @code{1} to @code{100}.
667 Default value is @code{10}. Higher values make clip detection less aggressive.
670 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
671 Default value is @code{1000}. Higher values make clip detection less aggressive.
676 It accepts the following values:
679 Select overlap-add method. Even not interpolated samples are slightly changed
683 Select overlap-save method. Not interpolated samples remain unchanged.
686 Default value is @code{a}.
691 Delay one or more audio channels.
693 Samples in delayed channel are filled with silence.
695 The filter accepts the following option:
699 Set list of delays in milliseconds for each channel separated by '|'.
700 Unused delays will be silently ignored. If number of given delays is
701 smaller than number of channels all remaining channels will not be delayed.
702 If you want to delay exact number of samples, append 'S' to number.
703 If you want instead to delay in seconds, append 's' to number.
706 Use last set delay for all remaining channels. By default is disabled.
707 This option if enabled changes how option @code{delays} is interpreted.
714 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
715 the second channel (and any other channels that may be present) unchanged.
721 Delay second channel by 500 samples, the third channel by 700 samples and leave
722 the first channel (and any other channels that may be present) unchanged.
728 Delay all channels by same number of samples:
730 adelay=delays=64S:all=1
734 @section aderivative, aintegral
736 Compute derivative/integral of audio stream.
738 Applying both filters one after another produces original audio.
742 Apply echoing to the input audio.
744 Echoes are reflected sound and can occur naturally amongst mountains
745 (and sometimes large buildings) when talking or shouting; digital echo
746 effects emulate this behaviour and are often used to help fill out the
747 sound of a single instrument or vocal. The time difference between the
748 original signal and the reflection is the @code{delay}, and the
749 loudness of the reflected signal is the @code{decay}.
750 Multiple echoes can have different delays and decays.
752 A description of the accepted parameters follows.
756 Set input gain of reflected signal. Default is @code{0.6}.
759 Set output gain of reflected signal. Default is @code{0.3}.
762 Set list of time intervals in milliseconds between original signal and reflections
763 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
764 Default is @code{1000}.
767 Set list of loudness of reflected signals separated by '|'.
768 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
769 Default is @code{0.5}.
776 Make it sound as if there are twice as many instruments as are actually playing:
778 aecho=0.8:0.88:60:0.4
782 If delay is very short, then it sounds like a (metallic) robot playing music:
788 A longer delay will sound like an open air concert in the mountains:
790 aecho=0.8:0.9:1000:0.3
794 Same as above but with one more mountain:
796 aecho=0.8:0.9:1000|1800:0.3|0.25
801 Audio emphasis filter creates or restores material directly taken from LPs or
802 emphased CDs with different filter curves. E.g. to store music on vinyl the
803 signal has to be altered by a filter first to even out the disadvantages of
804 this recording medium.
805 Once the material is played back the inverse filter has to be applied to
806 restore the distortion of the frequency response.
808 The filter accepts the following options:
818 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
819 use @code{production} mode. Default is @code{reproduction} mode.
822 Set filter type. Selects medium. Can be one of the following:
834 select Compact Disc (CD).
840 select 50µs (FM-KF).
842 select 75µs (FM-KF).
848 Modify an audio signal according to the specified expressions.
850 This filter accepts one or more expressions (one for each channel),
851 which are evaluated and used to modify a corresponding audio signal.
853 It accepts the following parameters:
857 Set the '|'-separated expressions list for each separate channel. If
858 the number of input channels is greater than the number of
859 expressions, the last specified expression is used for the remaining
862 @item channel_layout, c
863 Set output channel layout. If not specified, the channel layout is
864 specified by the number of expressions. If set to @samp{same}, it will
865 use by default the same input channel layout.
868 Each expression in @var{exprs} can contain the following constants and functions:
872 channel number of the current expression
875 number of the evaluated sample, starting from 0
881 time of the evaluated sample expressed in seconds
884 @item nb_out_channels
885 input and output number of channels
888 the value of input channel with number @var{CH}
891 Note: this filter is slow. For faster processing you should use a
900 aeval=val(ch)/2:c=same
904 Invert phase of the second channel:
913 Apply fade-in/out effect to input audio.
915 A description of the accepted parameters follows.
919 Specify the effect type, can be either @code{in} for fade-in, or
920 @code{out} for a fade-out effect. Default is @code{in}.
922 @item start_sample, ss
923 Specify the number of the start sample for starting to apply the fade
924 effect. Default is 0.
927 Specify the number of samples for which the fade effect has to last. At
928 the end of the fade-in effect the output audio will have the same
929 volume as the input audio, at the end of the fade-out transition
930 the output audio will be silence. Default is 44100.
933 Specify the start time of the fade effect. Default is 0.
934 The value must be specified as a time duration; see
935 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
936 for the accepted syntax.
937 If set this option is used instead of @var{start_sample}.
940 Specify the duration of the fade effect. See
941 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
942 for the accepted syntax.
943 At the end of the fade-in effect the output audio will have the same
944 volume as the input audio, at the end of the fade-out transition
945 the output audio will be silence.
946 By default the duration is determined by @var{nb_samples}.
947 If set this option is used instead of @var{nb_samples}.
950 Set curve for fade transition.
952 It accepts the following values:
955 select triangular, linear slope (default)
957 select quarter of sine wave
959 select half of sine wave
961 select exponential sine wave
965 select inverted parabola
979 select inverted quarter of sine wave
981 select inverted half of sine wave
983 select double-exponential seat
985 select double-exponential sigmoid
987 select logistic sigmoid
997 Fade in first 15 seconds of audio:
1003 Fade out last 25 seconds of a 900 seconds audio:
1005 afade=t=out:st=875:d=25
1010 Denoise audio samples with FFT.
1012 A description of the accepted parameters follows.
1016 Set the noise reduction in dB, allowed range is 0.01 to 97.
1017 Default value is 12 dB.
1020 Set the noise floor in dB, allowed range is -80 to -20.
1021 Default value is -50 dB.
1026 It accepts the following values:
1035 Select shellac noise.
1038 Select custom noise, defined in @code{bn} option.
1040 Default value is white noise.
1044 Set custom band noise for every one of 15 bands.
1045 Bands are separated by ' ' or '|'.
1048 Set the residual floor in dB, allowed range is -80 to -20.
1049 Default value is -38 dB.
1052 Enable noise tracking. By default is disabled.
1053 With this enabled, noise floor is automatically adjusted.
1056 Enable residual tracking. By default is disabled.
1059 Set the output mode.
1061 It accepts the following values:
1064 Pass input unchanged.
1067 Pass noise filtered out.
1072 Default value is @var{o}.
1076 @subsection Commands
1078 This filter supports the following commands:
1080 @item sample_noise, sn
1081 Start or stop measuring noise profile.
1082 Syntax for the command is : "start" or "stop" string.
1083 After measuring noise profile is stopped it will be
1084 automatically applied in filtering.
1086 @item noise_reduction, nr
1087 Change noise reduction. Argument is single float number.
1088 Syntax for the command is : "@var{noise_reduction}"
1090 @item noise_floor, nf
1091 Change noise floor. Argument is single float number.
1092 Syntax for the command is : "@var{noise_floor}"
1094 @item output_mode, om
1095 Change output mode operation.
1096 Syntax for the command is : "i", "o" or "n" string.
1100 Apply arbitrary expressions to samples in frequency domain.
1104 Set frequency domain real expression for each separate channel separated
1105 by '|'. Default is "re".
1106 If the number of input channels is greater than the number of
1107 expressions, the last specified expression is used for the remaining
1111 Set frequency domain imaginary expression for each separate channel
1112 separated by '|'. Default is "im".
1114 Each expression in @var{real} and @var{imag} can contain the following
1115 constants and functions:
1122 current frequency bin number
1125 number of available bins
1128 channel number of the current expression
1137 current real part of frequency bin of current channel
1140 current imaginary part of frequency bin of current channel
1143 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1146 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1150 Set window size. Allowed range is from 16 to 131072.
1151 Default is @code{4096}
1154 Set window function. Default is @code{hann}.
1157 Set window overlap. If set to 1, the recommended overlap for selected
1158 window function will be picked. Default is @code{0.75}.
1161 @subsection Examples
1165 Leave almost only low frequencies in audio:
1167 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1171 Apply robotize effect:
1173 afftfilt="real='hypot(re,im)*sin(0)':imag='hypot(re,im)*cos(0)':win_size=512:overlap=0.75"
1177 Apply whisper effect:
1179 afftfilt="real='hypot(re,im)*cos((random(0)*2-1)*2*3.14)':imag='hypot(re,im)*sin((random(1)*2-1)*2*3.14)':win_size=128:overlap=0.8"
1186 Apply an arbitrary Finite Impulse Response filter.
1188 This filter is designed for applying long FIR filters,
1189 up to 60 seconds long.
1191 It can be used as component for digital crossover filters,
1192 room equalization, cross talk cancellation, wavefield synthesis,
1193 auralization, ambiophonics, ambisonics and spatialization.
1195 This filter uses the streams higher than first one as FIR coefficients.
1196 If the non-first stream holds a single channel, it will be used
1197 for all input channels in the first stream, otherwise
1198 the number of channels in the non-first stream must be same as
1199 the number of channels in the first stream.
1201 It accepts the following parameters:
1205 Set dry gain. This sets input gain.
1208 Set wet gain. This sets final output gain.
1211 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1214 Enable applying gain measured from power of IR.
1216 Set which approach to use for auto gain measurement.
1220 Do not apply any gain.
1223 select peak gain, very conservative approach. This is default value.
1226 select DC gain, limited application.
1229 select gain to noise approach, this is most popular one.
1233 Set gain to be applied to IR coefficients before filtering.
1234 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1237 Set format of IR stream. Can be @code{mono} or @code{input}.
1238 Default is @code{input}.
1241 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1242 Allowed range is 0.1 to 60 seconds.
1245 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1246 By default it is disabled.
1249 Set for which IR channel to display frequency response. By default is first channel
1250 displayed. This option is used only when @var{response} is enabled.
1253 Set video stream size. This option is used only when @var{response} is enabled.
1256 Set video stream frame rate. This option is used only when @var{response} is enabled.
1259 Set minimal partition size used for convolution. Default is @var{8192}.
1260 Allowed range is from @var{1} to @var{32768}.
1261 Lower values decreases latency at cost of higher CPU usage.
1264 Set maximal partition size used for convolution. Default is @var{8192}.
1265 Allowed range is from @var{8} to @var{32768}.
1266 Lower values may increase CPU usage.
1269 Set number of input impulse responses streams which will be switchable at runtime.
1270 Allowed range is from @var{1} to @var{32}. Default is @var{1}.
1273 Set IR stream which will be used for convolution, starting from @var{0}, should always be
1274 lower than supplied value by @code{nbirs} option. Default is @var{0}.
1275 This option can be changed at runtime via @ref{commands}.
1278 @subsection Examples
1282 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1284 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1291 Set output format constraints for the input audio. The framework will
1292 negotiate the most appropriate format to minimize conversions.
1294 It accepts the following parameters:
1297 @item sample_fmts, f
1298 A '|'-separated list of requested sample formats.
1300 @item sample_rates, r
1301 A '|'-separated list of requested sample rates.
1303 @item channel_layouts, cl
1304 A '|'-separated list of requested channel layouts.
1306 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1307 for the required syntax.
1310 If a parameter is omitted, all values are allowed.
1312 Force the output to either unsigned 8-bit or signed 16-bit stereo
1314 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1319 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1320 processing reduces disturbing noise between useful signals.
1322 Gating is done by detecting the volume below a chosen level @var{threshold}
1323 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1324 floor is set via @var{range}. Because an exact manipulation of the signal
1325 would cause distortion of the waveform the reduction can be levelled over
1326 time. This is done by setting @var{attack} and @var{release}.
1328 @var{attack} determines how long the signal has to fall below the threshold
1329 before any reduction will occur and @var{release} sets the time the signal
1330 has to rise above the threshold to reduce the reduction again.
1331 Shorter signals than the chosen attack time will be left untouched.
1335 Set input level before filtering.
1336 Default is 1. Allowed range is from 0.015625 to 64.
1339 Set the mode of operation. Can be @code{upward} or @code{downward}.
1340 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1341 will be amplified, expanding dynamic range in upward direction.
1342 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1345 Set the level of gain reduction when the signal is below the threshold.
1346 Default is 0.06125. Allowed range is from 0 to 1.
1347 Setting this to 0 disables reduction and then filter behaves like expander.
1350 If a signal rises above this level the gain reduction is released.
1351 Default is 0.125. Allowed range is from 0 to 1.
1354 Set a ratio by which the signal is reduced.
1355 Default is 2. Allowed range is from 1 to 9000.
1358 Amount of milliseconds the signal has to rise above the threshold before gain
1360 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1363 Amount of milliseconds the signal has to fall below the threshold before the
1364 reduction is increased again. Default is 250 milliseconds.
1365 Allowed range is from 0.01 to 9000.
1368 Set amount of amplification of signal after processing.
1369 Default is 1. Allowed range is from 1 to 64.
1372 Curve the sharp knee around the threshold to enter gain reduction more softly.
1373 Default is 2.828427125. Allowed range is from 1 to 8.
1376 Choose if exact signal should be taken for detection or an RMS like one.
1377 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1380 Choose if the average level between all channels or the louder channel affects
1382 Default is @code{average}. Can be @code{average} or @code{maximum}.
1387 Apply an arbitrary Infinite Impulse Response filter.
1389 It accepts the following parameters:
1393 Set numerator/zeros coefficients.
1396 Set denominator/poles coefficients.
1408 Set coefficients format.
1412 digital transfer function
1414 Z-plane zeros/poles, cartesian (default)
1416 Z-plane zeros/poles, polar radians
1418 Z-plane zeros/poles, polar degrees
1424 Set kind of processing.
1425 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1428 Set filtering precision.
1432 double-precision floating-point (default)
1434 single-precision floating-point
1442 Normalize filter coefficients, by default is enabled.
1443 Enabling it will normalize magnitude response at DC to 0dB.
1446 How much to use filtered signal in output. Default is 1.
1447 Range is between 0 and 1.
1450 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1451 By default it is disabled.
1454 Set for which IR channel to display frequency response. By default is first channel
1455 displayed. This option is used only when @var{response} is enabled.
1458 Set video stream size. This option is used only when @var{response} is enabled.
1461 Coefficients in @code{tf} format are separated by spaces and are in ascending
1464 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1465 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1468 Different coefficients and gains can be provided for every channel, in such case
1469 use '|' to separate coefficients or gains. Last provided coefficients will be
1470 used for all remaining channels.
1472 @subsection Examples
1476 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1478 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
1482 Same as above but in @code{zp} format:
1484 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
1490 The limiter prevents an input signal from rising over a desired threshold.
1491 This limiter uses lookahead technology to prevent your signal from distorting.
1492 It means that there is a small delay after the signal is processed. Keep in mind
1493 that the delay it produces is the attack time you set.
1495 The filter accepts the following options:
1499 Set input gain. Default is 1.
1502 Set output gain. Default is 1.
1505 Don't let signals above this level pass the limiter. Default is 1.
1508 The limiter will reach its attenuation level in this amount of time in
1509 milliseconds. Default is 5 milliseconds.
1512 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1513 Default is 50 milliseconds.
1516 When gain reduction is always needed ASC takes care of releasing to an
1517 average reduction level rather than reaching a reduction of 0 in the release
1521 Select how much the release time is affected by ASC, 0 means nearly no changes
1522 in release time while 1 produces higher release times.
1525 Auto level output signal. Default is enabled.
1526 This normalizes audio back to 0dB if enabled.
1529 Depending on picked setting it is recommended to upsample input 2x or 4x times
1530 with @ref{aresample} before applying this filter.
1534 Apply a two-pole all-pass filter with central frequency (in Hz)
1535 @var{frequency}, and filter-width @var{width}.
1536 An all-pass filter changes the audio's frequency to phase relationship
1537 without changing its frequency to amplitude relationship.
1539 The filter accepts the following options:
1543 Set frequency in Hz.
1546 Set method to specify band-width of filter.
1561 Specify the band-width of a filter in width_type units.
1564 How much to use filtered signal in output. Default is 1.
1565 Range is between 0 and 1.
1568 Specify which channels to filter, by default all available are filtered.
1571 Normalize biquad coefficients, by default is disabled.
1572 Enabling it will normalize magnitude response at DC to 0dB.
1575 Set the filter order, can be 1 or 2. Default is 2.
1578 @subsection Commands
1580 This filter supports the following commands:
1583 Change allpass frequency.
1584 Syntax for the command is : "@var{frequency}"
1587 Change allpass width_type.
1588 Syntax for the command is : "@var{width_type}"
1591 Change allpass width.
1592 Syntax for the command is : "@var{width}"
1596 Syntax for the command is : "@var{mix}"
1603 The filter accepts the following options:
1607 Set the number of loops. Setting this value to -1 will result in infinite loops.
1611 Set maximal number of samples. Default is 0.
1614 Set first sample of loop. Default is 0.
1620 Merge two or more audio streams into a single multi-channel stream.
1622 The filter accepts the following options:
1627 Set the number of inputs. Default is 2.
1631 If the channel layouts of the inputs are disjoint, and therefore compatible,
1632 the channel layout of the output will be set accordingly and the channels
1633 will be reordered as necessary. If the channel layouts of the inputs are not
1634 disjoint, the output will have all the channels of the first input then all
1635 the channels of the second input, in that order, and the channel layout of
1636 the output will be the default value corresponding to the total number of
1639 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1640 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1641 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1642 first input, b1 is the first channel of the second input).
1644 On the other hand, if both input are in stereo, the output channels will be
1645 in the default order: a1, a2, b1, b2, and the channel layout will be
1646 arbitrarily set to 4.0, which may or may not be the expected value.
1648 All inputs must have the same sample rate, and format.
1650 If inputs do not have the same duration, the output will stop with the
1653 @subsection Examples
1657 Merge two mono files into a stereo stream:
1659 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1663 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1665 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
1671 Mixes multiple audio inputs into a single output.
1673 Note that this filter only supports float samples (the @var{amerge}
1674 and @var{pan} audio filters support many formats). If the @var{amix}
1675 input has integer samples then @ref{aresample} will be automatically
1676 inserted to perform the conversion to float samples.
1680 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1682 will mix 3 input audio streams to a single output with the same duration as the
1683 first input and a dropout transition time of 3 seconds.
1685 It accepts the following parameters:
1689 The number of inputs. If unspecified, it defaults to 2.
1692 How to determine the end-of-stream.
1696 The duration of the longest input. (default)
1699 The duration of the shortest input.
1702 The duration of the first input.
1706 @item dropout_transition
1707 The transition time, in seconds, for volume renormalization when an input
1708 stream ends. The default value is 2 seconds.
1711 Specify weight of each input audio stream as sequence.
1712 Each weight is separated by space. By default all inputs have same weight.
1715 @subsection Commands
1717 This filter supports the following commands:
1720 Syntax is same as option with same name.
1725 Multiply first audio stream with second audio stream and store result
1726 in output audio stream. Multiplication is done by multiplying each
1727 sample from first stream with sample at same position from second stream.
1729 With this element-wise multiplication one can create amplitude fades and
1730 amplitude modulations.
1732 @section anequalizer
1734 High-order parametric multiband equalizer for each channel.
1736 It accepts the following parameters:
1740 This option string is in format:
1741 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1742 Each equalizer band is separated by '|'.
1746 Set channel number to which equalization will be applied.
1747 If input doesn't have that channel the entry is ignored.
1750 Set central frequency for band.
1751 If input doesn't have that frequency the entry is ignored.
1754 Set band width in hertz.
1757 Set band gain in dB.
1760 Set filter type for band, optional, can be:
1764 Butterworth, this is default.
1775 With this option activated frequency response of anequalizer is displayed
1779 Set video stream size. Only useful if curves option is activated.
1782 Set max gain that will be displayed. Only useful if curves option is activated.
1783 Setting this to a reasonable value makes it possible to display gain which is derived from
1784 neighbour bands which are too close to each other and thus produce higher gain
1785 when both are activated.
1788 Set frequency scale used to draw frequency response in video output.
1789 Can be linear or logarithmic. Default is logarithmic.
1792 Set color for each channel curve which is going to be displayed in video stream.
1793 This is list of color names separated by space or by '|'.
1794 Unrecognised or missing colors will be replaced by white color.
1797 @subsection Examples
1801 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1802 for first 2 channels using Chebyshev type 1 filter:
1804 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1808 @subsection Commands
1810 This filter supports the following commands:
1813 Alter existing filter parameters.
1814 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1816 @var{fN} is existing filter number, starting from 0, if no such filter is available
1818 @var{freq} set new frequency parameter.
1819 @var{width} set new width parameter in herz.
1820 @var{gain} set new gain parameter in dB.
1822 Full filter invocation with asendcmd may look like this:
1823 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1828 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1830 Each sample is adjusted by looking for other samples with similar contexts. This
1831 context similarity is defined by comparing their surrounding patches of size
1832 @option{p}. Patches are searched in an area of @option{r} around the sample.
1834 The filter accepts the following options:
1838 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1841 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1842 Default value is 2 milliseconds.
1845 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1846 Default value is 6 milliseconds.
1849 Set the output mode.
1851 It accepts the following values:
1854 Pass input unchanged.
1857 Pass noise filtered out.
1862 Default value is @var{o}.
1866 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
1869 @subsection Commands
1871 This filter supports the following commands:
1874 Change denoise strength. Argument is single float number.
1875 Syntax for the command is : "@var{s}"
1879 Syntax for the command is : "i", "o" or "n" string.
1883 Apply Normalized Least-Mean-Squares algorithm to the first audio stream using the second audio stream.
1885 This adaptive filter is used to mimic a desired filter by finding the filter coefficients that
1886 relate to producing the least mean square of the error signal (difference between the desired,
1887 2nd input audio stream and the actual signal, the 1st input audio stream).
1889 A description of the accepted options follows.
1902 Set the filter leakage.
1905 It accepts the following values:
1914 Pass filtered samples.
1917 Pass difference between desired and filtered samples.
1919 Default value is @var{o}.
1923 @subsection Examples
1927 One of many usages of this filter is noise reduction, input audio is filtered
1928 with same samples that are delayed by fixed amount, one such example for stereo audio is:
1930 asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
1934 @subsection Commands
1936 This filter supports the same commands as options, excluding option @code{order}.
1940 Pass the audio source unchanged to the output.
1944 Pad the end of an audio stream with silence.
1946 This can be used together with @command{ffmpeg} @option{-shortest} to
1947 extend audio streams to the same length as the video stream.
1949 A description of the accepted options follows.
1953 Set silence packet size. Default value is 4096.
1956 Set the number of samples of silence to add to the end. After the
1957 value is reached, the stream is terminated. This option is mutually
1958 exclusive with @option{whole_len}.
1961 Set the minimum total number of samples in the output audio stream. If
1962 the value is longer than the input audio length, silence is added to
1963 the end, until the value is reached. This option is mutually exclusive
1964 with @option{pad_len}.
1967 Specify the duration of samples of silence to add. See
1968 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1969 for the accepted syntax. Used only if set to non-zero value.
1972 Specify the minimum total duration in the output audio stream. See
1973 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1974 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1975 the input audio length, silence is added to the end, until the value is reached.
1976 This option is mutually exclusive with @option{pad_dur}
1979 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1980 nor @option{whole_dur} option is set, the filter will add silence to the end of
1981 the input stream indefinitely.
1983 @subsection Examples
1987 Add 1024 samples of silence to the end of the input:
1993 Make sure the audio output will contain at least 10000 samples, pad
1994 the input with silence if required:
1996 apad=whole_len=10000
2000 Use @command{ffmpeg} to pad the audio input with silence, so that the
2001 video stream will always result the shortest and will be converted
2002 until the end in the output file when using the @option{shortest}
2005 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
2010 Add a phasing effect to the input audio.
2012 A phaser filter creates series of peaks and troughs in the frequency spectrum.
2013 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
2015 A description of the accepted parameters follows.
2019 Set input gain. Default is 0.4.
2022 Set output gain. Default is 0.74
2025 Set delay in milliseconds. Default is 3.0.
2028 Set decay. Default is 0.4.
2031 Set modulation speed in Hz. Default is 0.5.
2034 Set modulation type. Default is triangular.
2036 It accepts the following values:
2045 Audio pulsator is something between an autopanner and a tremolo.
2046 But it can produce funny stereo effects as well. Pulsator changes the volume
2047 of the left and right channel based on a LFO (low frequency oscillator) with
2048 different waveforms and shifted phases.
2049 This filter have the ability to define an offset between left and right
2050 channel. An offset of 0 means that both LFO shapes match each other.
2051 The left and right channel are altered equally - a conventional tremolo.
2052 An offset of 50% means that the shape of the right channel is exactly shifted
2053 in phase (or moved backwards about half of the frequency) - pulsator acts as
2054 an autopanner. At 1 both curves match again. Every setting in between moves the
2055 phase shift gapless between all stages and produces some "bypassing" sounds with
2056 sine and triangle waveforms. The more you set the offset near 1 (starting from
2057 the 0.5) the faster the signal passes from the left to the right speaker.
2059 The filter accepts the following options:
2063 Set input gain. By default it is 1. Range is [0.015625 - 64].
2066 Set output gain. By default it is 1. Range is [0.015625 - 64].
2069 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
2070 sawup or sawdown. Default is sine.
2073 Set modulation. Define how much of original signal is affected by the LFO.
2076 Set left channel offset. Default is 0. Allowed range is [0 - 1].
2079 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
2082 Set pulse width. Default is 1. Allowed range is [0 - 2].
2085 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
2088 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
2092 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
2096 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
2097 if timing is set to hz.
2103 Resample the input audio to the specified parameters, using the
2104 libswresample library. If none are specified then the filter will
2105 automatically convert between its input and output.
2107 This filter is also able to stretch/squeeze the audio data to make it match
2108 the timestamps or to inject silence / cut out audio to make it match the
2109 timestamps, do a combination of both or do neither.
2111 The filter accepts the syntax
2112 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
2113 expresses a sample rate and @var{resampler_options} is a list of
2114 @var{key}=@var{value} pairs, separated by ":". See the
2115 @ref{Resampler Options,,"Resampler Options" section in the
2116 ffmpeg-resampler(1) manual,ffmpeg-resampler}
2117 for the complete list of supported options.
2119 @subsection Examples
2123 Resample the input audio to 44100Hz:
2129 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2130 samples per second compensation:
2132 aresample=async=1000
2138 Reverse an audio clip.
2140 Warning: This filter requires memory to buffer the entire clip, so trimming
2143 @subsection Examples
2147 Take the first 5 seconds of a clip, and reverse it.
2149 atrim=end=5,areverse
2155 Reduce noise from speech using Recurrent Neural Networks.
2157 This filter accepts the following options:
2161 Set train model file to load. This option is always required.
2164 @section asetnsamples
2166 Set the number of samples per each output audio frame.
2168 The last output packet may contain a different number of samples, as
2169 the filter will flush all the remaining samples when the input audio
2172 The filter accepts the following options:
2176 @item nb_out_samples, n
2177 Set the number of frames per each output audio frame. The number is
2178 intended as the number of samples @emph{per each channel}.
2179 Default value is 1024.
2182 If set to 1, the filter will pad the last audio frame with zeroes, so
2183 that the last frame will contain the same number of samples as the
2184 previous ones. Default value is 1.
2187 For example, to set the number of per-frame samples to 1234 and
2188 disable padding for the last frame, use:
2190 asetnsamples=n=1234:p=0
2195 Set the sample rate without altering the PCM data.
2196 This will result in a change of speed and pitch.
2198 The filter accepts the following options:
2201 @item sample_rate, r
2202 Set the output sample rate. Default is 44100 Hz.
2207 Show a line containing various information for each input audio frame.
2208 The input audio is not modified.
2210 The shown line contains a sequence of key/value pairs of the form
2211 @var{key}:@var{value}.
2213 The following values are shown in the output:
2217 The (sequential) number of the input frame, starting from 0.
2220 The presentation timestamp of the input frame, in time base units; the time base
2221 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2224 The presentation timestamp of the input frame in seconds.
2227 position of the frame in the input stream, -1 if this information in
2228 unavailable and/or meaningless (for example in case of synthetic audio)
2237 The sample rate for the audio frame.
2240 The number of samples (per channel) in the frame.
2243 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2244 audio, the data is treated as if all the planes were concatenated.
2246 @item plane_checksums
2247 A list of Adler-32 checksums for each data plane.
2251 Apply audio soft clipping.
2253 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2254 along a smooth curve, rather than the abrupt shape of hard-clipping.
2256 This filter accepts the following options:
2260 Set type of soft-clipping.
2262 It accepts the following values:
2274 Set additional parameter which controls sigmoid function.
2277 @subsection Commands
2279 This filter supports the all above options as @ref{commands}.
2282 Automatic Speech Recognition
2284 This filter uses PocketSphinx for speech recognition. To enable
2285 compilation of this filter, you need to configure FFmpeg with
2286 @code{--enable-pocketsphinx}.
2288 It accepts the following options:
2292 Set sampling rate of input audio. Defaults is @code{16000}.
2293 This need to match speech models, otherwise one will get poor results.
2296 Set dictionary containing acoustic model files.
2299 Set pronunciation dictionary.
2302 Set language model file.
2305 Set language model set.
2308 Set which language model to use.
2311 Set output for log messages.
2314 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2319 Display time domain statistical information about the audio channels.
2320 Statistics are calculated and displayed for each audio channel and,
2321 where applicable, an overall figure is also given.
2323 It accepts the following option:
2326 Short window length in seconds, used for peak and trough RMS measurement.
2327 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2331 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2332 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2335 Available keys for each channel are:
2381 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2382 this @code{lavfi.astats.Overall.Peak_count}.
2384 For description what each key means read below.
2387 Set number of frame after which stats are going to be recalculated.
2388 Default is disabled.
2390 @item measure_perchannel
2391 Select the entries which need to be measured per channel. The metadata keys can
2392 be used as flags, default is @option{all} which measures everything.
2393 @option{none} disables all per channel measurement.
2395 @item measure_overall
2396 Select the entries which need to be measured overall. The metadata keys can
2397 be used as flags, default is @option{all} which measures everything.
2398 @option{none} disables all overall measurement.
2402 A description of each shown parameter follows:
2406 Mean amplitude displacement from zero.
2409 Minimal sample level.
2412 Maximal sample level.
2414 @item Min difference
2415 Minimal difference between two consecutive samples.
2417 @item Max difference
2418 Maximal difference between two consecutive samples.
2420 @item Mean difference
2421 Mean difference between two consecutive samples.
2422 The average of each difference between two consecutive samples.
2424 @item RMS difference
2425 Root Mean Square difference between two consecutive samples.
2429 Standard peak and RMS level measured in dBFS.
2433 Peak and trough values for RMS level measured over a short window.
2436 Standard ratio of peak to RMS level (note: not in dB).
2439 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2440 (i.e. either @var{Min level} or @var{Max level}).
2443 Number of occasions (not the number of samples) that the signal attained either
2444 @var{Min level} or @var{Max level}.
2446 @item Noise floor dB
2447 Minimum local peak measured in dBFS over a short window.
2449 @item Noise floor count
2450 Number of occasions (not the number of samples) that the signal attained
2454 Overall bit depth of audio. Number of bits used for each sample.
2457 Measured dynamic range of audio in dB.
2459 @item Zero crossings
2460 Number of points where the waveform crosses the zero level axis.
2462 @item Zero crossings rate
2463 Rate of Zero crossings and number of audio samples.
2467 Boost subwoofer frequencies.
2469 The filter accepts the following options:
2473 Set dry gain, how much of original signal is kept. Allowed range is from 0 to 1.
2474 Default value is 0.5.
2477 Set wet gain, how much of filtered signal is kept. Allowed range is from 0 to 1.
2478 Default value is 0.8.
2481 Set delay line decay gain value. Allowed range is from 0 to 1.
2482 Default value is 0.7.
2485 Set delay line feedback gain value. Allowed range is from 0 to 1.
2486 Default value is 0.5.
2489 Set cutoff frequency in herz. Allowed range is 50 to 900.
2490 Default value is 100.
2493 Set slope amount for cutoff frequency. Allowed range is 0.0001 to 1.
2494 Default value is 0.5.
2497 Set delay. Allowed range is from 1 to 100.
2498 Default value is 20.
2501 @subsection Commands
2503 This filter supports the all above options as @ref{commands}.
2509 The filter accepts exactly one parameter, the audio tempo. If not
2510 specified then the filter will assume nominal 1.0 tempo. Tempo must
2511 be in the [0.5, 100.0] range.
2513 Note that tempo greater than 2 will skip some samples rather than
2514 blend them in. If for any reason this is a concern it is always
2515 possible to daisy-chain several instances of atempo to achieve the
2516 desired product tempo.
2518 @subsection Examples
2522 Slow down audio to 80% tempo:
2528 To speed up audio to 300% tempo:
2534 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2536 atempo=sqrt(3),atempo=sqrt(3)
2540 @subsection Commands
2542 This filter supports the following commands:
2545 Change filter tempo scale factor.
2546 Syntax for the command is : "@var{tempo}"
2551 Trim the input so that the output contains one continuous subpart of the input.
2553 It accepts the following parameters:
2556 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2557 sample with the timestamp @var{start} will be the first sample in the output.
2560 Specify time of the first audio sample that will be dropped, i.e. the
2561 audio sample immediately preceding the one with the timestamp @var{end} will be
2562 the last sample in the output.
2565 Same as @var{start}, except this option sets the start timestamp in samples
2569 Same as @var{end}, except this option sets the end timestamp in samples instead
2573 The maximum duration of the output in seconds.
2576 The number of the first sample that should be output.
2579 The number of the first sample that should be dropped.
2582 @option{start}, @option{end}, and @option{duration} are expressed as time
2583 duration specifications; see
2584 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2586 Note that the first two sets of the start/end options and the @option{duration}
2587 option look at the frame timestamp, while the _sample options simply count the
2588 samples that pass through the filter. So start/end_pts and start/end_sample will
2589 give different results when the timestamps are wrong, inexact or do not start at
2590 zero. Also note that this filter does not modify the timestamps. If you wish
2591 to have the output timestamps start at zero, insert the asetpts filter after the
2594 If multiple start or end options are set, this filter tries to be greedy and
2595 keep all samples that match at least one of the specified constraints. To keep
2596 only the part that matches all the constraints at once, chain multiple atrim
2599 The defaults are such that all the input is kept. So it is possible to set e.g.
2600 just the end values to keep everything before the specified time.
2605 Drop everything except the second minute of input:
2607 ffmpeg -i INPUT -af atrim=60:120
2611 Keep only the first 1000 samples:
2613 ffmpeg -i INPUT -af atrim=end_sample=1000
2618 @section axcorrelate
2619 Calculate normalized cross-correlation between two input audio streams.
2621 Resulted samples are always between -1 and 1 inclusive.
2622 If result is 1 it means two input samples are highly correlated in that selected segment.
2623 Result 0 means they are not correlated at all.
2624 If result is -1 it means two input samples are out of phase, which means they cancel each
2627 The filter accepts the following options:
2631 Set size of segment over which cross-correlation is calculated.
2632 Default is 256. Allowed range is from 2 to 131072.
2635 Set algorithm for cross-correlation. Can be @code{slow} or @code{fast}.
2636 Default is @code{slow}. Fast algorithm assumes mean values over any given segment
2637 are always zero and thus need much less calculations to make.
2638 This is generally not true, but is valid for typical audio streams.
2641 @subsection Examples
2645 Calculate correlation between channels in stereo audio stream:
2647 ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
2653 Apply a two-pole Butterworth band-pass filter with central
2654 frequency @var{frequency}, and (3dB-point) band-width width.
2655 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2656 instead of the default: constant 0dB peak gain.
2657 The filter roll off at 6dB per octave (20dB per decade).
2659 The filter accepts the following options:
2663 Set the filter's central frequency. Default is @code{3000}.
2666 Constant skirt gain if set to 1. Defaults to 0.
2669 Set method to specify band-width of filter.
2684 Specify the band-width of a filter in width_type units.
2687 How much to use filtered signal in output. Default is 1.
2688 Range is between 0 and 1.
2691 Specify which channels to filter, by default all available are filtered.
2694 Normalize biquad coefficients, by default is disabled.
2695 Enabling it will normalize magnitude response at DC to 0dB.
2698 @subsection Commands
2700 This filter supports the following commands:
2703 Change bandpass frequency.
2704 Syntax for the command is : "@var{frequency}"
2707 Change bandpass width_type.
2708 Syntax for the command is : "@var{width_type}"
2711 Change bandpass width.
2712 Syntax for the command is : "@var{width}"
2715 Change bandpass mix.
2716 Syntax for the command is : "@var{mix}"
2721 Apply a two-pole Butterworth band-reject filter with central
2722 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2723 The filter roll off at 6dB per octave (20dB per decade).
2725 The filter accepts the following options:
2729 Set the filter's central frequency. Default is @code{3000}.
2732 Set method to specify band-width of filter.
2747 Specify the band-width of a filter in width_type units.
2750 How much to use filtered signal in output. Default is 1.
2751 Range is between 0 and 1.
2754 Specify which channels to filter, by default all available are filtered.
2757 Normalize biquad coefficients, by default is disabled.
2758 Enabling it will normalize magnitude response at DC to 0dB.
2761 @subsection Commands
2763 This filter supports the following commands:
2766 Change bandreject frequency.
2767 Syntax for the command is : "@var{frequency}"
2770 Change bandreject width_type.
2771 Syntax for the command is : "@var{width_type}"
2774 Change bandreject width.
2775 Syntax for the command is : "@var{width}"
2778 Change bandreject mix.
2779 Syntax for the command is : "@var{mix}"
2782 @section bass, lowshelf
2784 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2785 shelving filter with a response similar to that of a standard
2786 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2788 The filter accepts the following options:
2792 Give the gain at 0 Hz. Its useful range is about -20
2793 (for a large cut) to +20 (for a large boost).
2794 Beware of clipping when using a positive gain.
2797 Set the filter's central frequency and so can be used
2798 to extend or reduce the frequency range to be boosted or cut.
2799 The default value is @code{100} Hz.
2802 Set method to specify band-width of filter.
2817 Determine how steep is the filter's shelf transition.
2820 How much to use filtered signal in output. Default is 1.
2821 Range is between 0 and 1.
2824 Specify which channels to filter, by default all available are filtered.
2827 Normalize biquad coefficients, by default is disabled.
2828 Enabling it will normalize magnitude response at DC to 0dB.
2831 @subsection Commands
2833 This filter supports the following commands:
2836 Change bass frequency.
2837 Syntax for the command is : "@var{frequency}"
2840 Change bass width_type.
2841 Syntax for the command is : "@var{width_type}"
2845 Syntax for the command is : "@var{width}"
2849 Syntax for the command is : "@var{gain}"
2853 Syntax for the command is : "@var{mix}"
2858 Apply a biquad IIR filter with the given coefficients.
2859 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2860 are the numerator and denominator coefficients respectively.
2861 and @var{channels}, @var{c} specify which channels to filter, by default all
2862 available are filtered.
2864 @subsection Commands
2866 This filter supports the following commands:
2874 Change biquad parameter.
2875 Syntax for the command is : "@var{value}"
2878 How much to use filtered signal in output. Default is 1.
2879 Range is between 0 and 1.
2882 Specify which channels to filter, by default all available are filtered.
2885 Normalize biquad coefficients, by default is disabled.
2886 Enabling it will normalize magnitude response at DC to 0dB.
2890 Bauer stereo to binaural transformation, which improves headphone listening of
2891 stereo audio records.
2893 To enable compilation of this filter you need to configure FFmpeg with
2894 @code{--enable-libbs2b}.
2896 It accepts the following parameters:
2900 Pre-defined crossfeed level.
2904 Default level (fcut=700, feed=50).
2907 Chu Moy circuit (fcut=700, feed=60).
2910 Jan Meier circuit (fcut=650, feed=95).
2915 Cut frequency (in Hz).
2924 Remap input channels to new locations.
2926 It accepts the following parameters:
2929 Map channels from input to output. The argument is a '|'-separated list of
2930 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2931 @var{in_channel} form. @var{in_channel} can be either the name of the input
2932 channel (e.g. FL for front left) or its index in the input channel layout.
2933 @var{out_channel} is the name of the output channel or its index in the output
2934 channel layout. If @var{out_channel} is not given then it is implicitly an
2935 index, starting with zero and increasing by one for each mapping.
2937 @item channel_layout
2938 The channel layout of the output stream.
2941 If no mapping is present, the filter will implicitly map input channels to
2942 output channels, preserving indices.
2944 @subsection Examples
2948 For example, assuming a 5.1+downmix input MOV file,
2950 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2952 will create an output WAV file tagged as stereo from the downmix channels of
2956 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2958 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2962 @section channelsplit
2964 Split each channel from an input audio stream into a separate output stream.
2966 It accepts the following parameters:
2968 @item channel_layout
2969 The channel layout of the input stream. The default is "stereo".
2971 A channel layout describing the channels to be extracted as separate output streams
2972 or "all" to extract each input channel as a separate stream. The default is "all".
2974 Choosing channels not present in channel layout in the input will result in an error.
2977 @subsection Examples
2981 For example, assuming a stereo input MP3 file,
2983 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2985 will create an output Matroska file with two audio streams, one containing only
2986 the left channel and the other the right channel.
2989 Split a 5.1 WAV file into per-channel files:
2991 ffmpeg -i in.wav -filter_complex
2992 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2993 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2994 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2999 Extract only LFE from a 5.1 WAV file:
3001 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
3002 -map '[LFE]' lfe.wav
3007 Add a chorus effect to the audio.
3009 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
3011 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
3012 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
3013 The modulation depth defines the range the modulated delay is played before or after
3014 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
3015 sound tuned around the original one, like in a chorus where some vocals are slightly
3018 It accepts the following parameters:
3021 Set input gain. Default is 0.4.
3024 Set output gain. Default is 0.4.
3027 Set delays. A typical delay is around 40ms to 60ms.
3039 @subsection Examples
3045 chorus=0.7:0.9:55:0.4:0.25:2
3051 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
3055 Fuller sounding chorus with three delays:
3057 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
3062 Compress or expand the audio's dynamic range.
3064 It accepts the following parameters:
3070 A list of times in seconds for each channel over which the instantaneous level
3071 of the input signal is averaged to determine its volume. @var{attacks} refers to
3072 increase of volume and @var{decays} refers to decrease of volume. For most
3073 situations, the attack time (response to the audio getting louder) should be
3074 shorter than the decay time, because the human ear is more sensitive to sudden
3075 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
3076 a typical value for decay is 0.8 seconds.
3077 If specified number of attacks & decays is lower than number of channels, the last
3078 set attack/decay will be used for all remaining channels.
3081 A list of points for the transfer function, specified in dB relative to the
3082 maximum possible signal amplitude. Each key points list must be defined using
3083 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
3084 @code{x0/y0 x1/y1 x2/y2 ....}
3086 The input values must be in strictly increasing order but the transfer function
3087 does not have to be monotonically rising. The point @code{0/0} is assumed but
3088 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
3089 function are @code{-70/-70|-60/-20|1/0}.
3092 Set the curve radius in dB for all joints. It defaults to 0.01.
3095 Set the additional gain in dB to be applied at all points on the transfer
3096 function. This allows for easy adjustment of the overall gain.
3100 Set an initial volume, in dB, to be assumed for each channel when filtering
3101 starts. This permits the user to supply a nominal level initially, so that, for
3102 example, a very large gain is not applied to initial signal levels before the
3103 companding has begun to operate. A typical value for audio which is initially
3104 quiet is -90 dB. It defaults to 0.
3107 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
3108 delayed before being fed to the volume adjuster. Specifying a delay
3109 approximately equal to the attack/decay times allows the filter to effectively
3110 operate in predictive rather than reactive mode. It defaults to 0.
3114 @subsection Examples
3118 Make music with both quiet and loud passages suitable for listening to in a
3121 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
3124 Another example for audio with whisper and explosion parts:
3126 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
3130 A noise gate for when the noise is at a lower level than the signal:
3132 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
3136 Here is another noise gate, this time for when the noise is at a higher level
3137 than the signal (making it, in some ways, similar to squelch):
3139 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
3143 2:1 compression starting at -6dB:
3145 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
3149 2:1 compression starting at -9dB:
3151 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
3155 2:1 compression starting at -12dB:
3157 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
3161 2:1 compression starting at -18dB:
3163 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
3167 3:1 compression starting at -15dB:
3169 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
3175 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
3181 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
3185 Hard limiter at -6dB:
3187 compand=attacks=0:points=-80/-80|-6/-6|20/-6
3191 Hard limiter at -12dB:
3193 compand=attacks=0:points=-80/-80|-12/-12|20/-12
3197 Hard noise gate at -35 dB:
3199 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
3205 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
3209 @section compensationdelay
3211 Compensation Delay Line is a metric based delay to compensate differing
3212 positions of microphones or speakers.
3214 For example, you have recorded guitar with two microphones placed in
3215 different locations. Because the front of sound wave has fixed speed in
3216 normal conditions, the phasing of microphones can vary and depends on
3217 their location and interposition. The best sound mix can be achieved when
3218 these microphones are in phase (synchronized). Note that a distance of
3219 ~30 cm between microphones makes one microphone capture the signal in
3220 antiphase to the other microphone. That makes the final mix sound moody.
3221 This filter helps to solve phasing problems by adding different delays
3222 to each microphone track and make them synchronized.
3224 The best result can be reached when you take one track as base and
3225 synchronize other tracks one by one with it.
3226 Remember that synchronization/delay tolerance depends on sample rate, too.
3227 Higher sample rates will give more tolerance.
3229 The filter accepts the following parameters:
3233 Set millimeters distance. This is compensation distance for fine tuning.
3237 Set cm distance. This is compensation distance for tightening distance setup.
3241 Set meters distance. This is compensation distance for hard distance setup.
3245 Set dry amount. Amount of unprocessed (dry) signal.
3249 Set wet amount. Amount of processed (wet) signal.
3253 Set temperature in degrees Celsius. This is the temperature of the environment.
3258 Apply headphone crossfeed filter.
3260 Crossfeed is the process of blending the left and right channels of stereo
3262 It is mainly used to reduce extreme stereo separation of low frequencies.
3264 The intent is to produce more speaker like sound to the listener.
3266 The filter accepts the following options:
3270 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3271 This sets gain of low shelf filter for side part of stereo image.
3272 Default is -6dB. Max allowed is -30db when strength is set to 1.
3275 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3276 This sets cut off frequency of low shelf filter. Default is cut off near
3277 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3280 Set curve slope of low shelf filter. Default is 0.5.
3281 Allowed range is from 0.01 to 1.
3284 Set input gain. Default is 0.9.
3287 Set output gain. Default is 1.
3290 @subsection Commands
3292 This filter supports the all above options as @ref{commands}.
3294 @section crystalizer
3295 Simple algorithm to expand audio dynamic range.
3297 The filter accepts the following options:
3301 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
3302 (unchanged sound) to 10.0 (maximum effect).
3305 Enable clipping. By default is enabled.
3308 @subsection Commands
3310 This filter supports the all above options as @ref{commands}.
3313 Apply a DC shift to the audio.
3315 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3316 in the recording chain) from the audio. The effect of a DC offset is reduced
3317 headroom and hence volume. The @ref{astats} filter can be used to determine if
3318 a signal has a DC offset.
3322 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3326 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3327 used to prevent clipping.
3332 Apply de-essing to the audio samples.
3336 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3340 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3344 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3348 Set the output mode.
3350 It accepts the following values:
3353 Pass input unchanged.
3356 Pass ess filtered out.
3361 Default value is @var{o}.
3367 Measure audio dynamic range.
3369 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3370 is found in transition material. And anything less that 8 have very poor dynamics
3371 and is very compressed.
3373 The filter accepts the following options:
3377 Set window length in seconds used to split audio into segments of equal length.
3378 Default is 3 seconds.
3382 Dynamic Audio Normalizer.
3384 This filter applies a certain amount of gain to the input audio in order
3385 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3386 contrast to more "simple" normalization algorithms, the Dynamic Audio
3387 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3388 This allows for applying extra gain to the "quiet" sections of the audio
3389 while avoiding distortions or clipping the "loud" sections. In other words:
3390 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3391 sections, in the sense that the volume of each section is brought to the
3392 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3393 this goal *without* applying "dynamic range compressing". It will retain 100%
3394 of the dynamic range *within* each section of the audio file.
3398 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3399 Default is 500 milliseconds.
3400 The Dynamic Audio Normalizer processes the input audio in small chunks,
3401 referred to as frames. This is required, because a peak magnitude has no
3402 meaning for just a single sample value. Instead, we need to determine the
3403 peak magnitude for a contiguous sequence of sample values. While a "standard"
3404 normalizer would simply use the peak magnitude of the complete file, the
3405 Dynamic Audio Normalizer determines the peak magnitude individually for each
3406 frame. The length of a frame is specified in milliseconds. By default, the
3407 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3408 been found to give good results with most files.
3409 Note that the exact frame length, in number of samples, will be determined
3410 automatically, based on the sampling rate of the individual input audio file.
3413 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3414 number. Default is 31.
3415 Probably the most important parameter of the Dynamic Audio Normalizer is the
3416 @code{window size} of the Gaussian smoothing filter. The filter's window size
3417 is specified in frames, centered around the current frame. For the sake of
3418 simplicity, this must be an odd number. Consequently, the default value of 31
3419 takes into account the current frame, as well as the 15 preceding frames and
3420 the 15 subsequent frames. Using a larger window results in a stronger
3421 smoothing effect and thus in less gain variation, i.e. slower gain
3422 adaptation. Conversely, using a smaller window results in a weaker smoothing
3423 effect and thus in more gain variation, i.e. faster gain adaptation.
3424 In other words, the more you increase this value, the more the Dynamic Audio
3425 Normalizer will behave like a "traditional" normalization filter. On the
3426 contrary, the more you decrease this value, the more the Dynamic Audio
3427 Normalizer will behave like a dynamic range compressor.
3430 Set the target peak value. This specifies the highest permissible magnitude
3431 level for the normalized audio input. This filter will try to approach the
3432 target peak magnitude as closely as possible, but at the same time it also
3433 makes sure that the normalized signal will never exceed the peak magnitude.
3434 A frame's maximum local gain factor is imposed directly by the target peak
3435 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3436 It is not recommended to go above this value.
3439 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3440 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3441 factor for each input frame, i.e. the maximum gain factor that does not
3442 result in clipping or distortion. The maximum gain factor is determined by
3443 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3444 additionally bounds the frame's maximum gain factor by a predetermined
3445 (global) maximum gain factor. This is done in order to avoid excessive gain
3446 factors in "silent" or almost silent frames. By default, the maximum gain
3447 factor is 10.0, For most inputs the default value should be sufficient and
3448 it usually is not recommended to increase this value. Though, for input
3449 with an extremely low overall volume level, it may be necessary to allow even
3450 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3451 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3452 Instead, a "sigmoid" threshold function will be applied. This way, the
3453 gain factors will smoothly approach the threshold value, but never exceed that
3457 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3458 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3459 This means that the maximum local gain factor for each frame is defined
3460 (only) by the frame's highest magnitude sample. This way, the samples can
3461 be amplified as much as possible without exceeding the maximum signal
3462 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3463 Normalizer can also take into account the frame's root mean square,
3464 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3465 determine the power of a time-varying signal. It is therefore considered
3466 that the RMS is a better approximation of the "perceived loudness" than
3467 just looking at the signal's peak magnitude. Consequently, by adjusting all
3468 frames to a constant RMS value, a uniform "perceived loudness" can be
3469 established. If a target RMS value has been specified, a frame's local gain
3470 factor is defined as the factor that would result in exactly that RMS value.
3471 Note, however, that the maximum local gain factor is still restricted by the
3472 frame's highest magnitude sample, in order to prevent clipping.
3475 Enable channels coupling. By default is enabled.
3476 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3477 amount. This means the same gain factor will be applied to all channels, i.e.
3478 the maximum possible gain factor is determined by the "loudest" channel.
3479 However, in some recordings, it may happen that the volume of the different
3480 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3481 In this case, this option can be used to disable the channel coupling. This way,
3482 the gain factor will be determined independently for each channel, depending
3483 only on the individual channel's highest magnitude sample. This allows for
3484 harmonizing the volume of the different channels.
3487 Enable DC bias correction. By default is disabled.
3488 An audio signal (in the time domain) is a sequence of sample values.
3489 In the Dynamic Audio Normalizer these sample values are represented in the
3490 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3491 audio signal, or "waveform", should be centered around the zero point.
3492 That means if we calculate the mean value of all samples in a file, or in a
3493 single frame, then the result should be 0.0 or at least very close to that
3494 value. If, however, there is a significant deviation of the mean value from
3495 0.0, in either positive or negative direction, this is referred to as a
3496 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3497 Audio Normalizer provides optional DC bias correction.
3498 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3499 the mean value, or "DC correction" offset, of each input frame and subtract
3500 that value from all of the frame's sample values which ensures those samples
3501 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3502 boundaries, the DC correction offset values will be interpolated smoothly
3503 between neighbouring frames.
3505 @item altboundary, b
3506 Enable alternative boundary mode. By default is disabled.
3507 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3508 around each frame. This includes the preceding frames as well as the
3509 subsequent frames. However, for the "boundary" frames, located at the very
3510 beginning and at the very end of the audio file, not all neighbouring
3511 frames are available. In particular, for the first few frames in the audio
3512 file, the preceding frames are not known. And, similarly, for the last few
3513 frames in the audio file, the subsequent frames are not known. Thus, the
3514 question arises which gain factors should be assumed for the missing frames
3515 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3516 to deal with this situation. The default boundary mode assumes a gain factor
3517 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3518 "fade out" at the beginning and at the end of the input, respectively.
3521 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3522 By default, the Dynamic Audio Normalizer does not apply "traditional"
3523 compression. This means that signal peaks will not be pruned and thus the
3524 full dynamic range will be retained within each local neighbourhood. However,
3525 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3526 normalization algorithm with a more "traditional" compression.
3527 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3528 (thresholding) function. If (and only if) the compression feature is enabled,
3529 all input frames will be processed by a soft knee thresholding function prior
3530 to the actual normalization process. Put simply, the thresholding function is
3531 going to prune all samples whose magnitude exceeds a certain threshold value.
3532 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3533 value. Instead, the threshold value will be adjusted for each individual
3535 In general, smaller parameters result in stronger compression, and vice versa.
3536 Values below 3.0 are not recommended, because audible distortion may appear.
3539 Set the target threshold value. This specifies the lowest permissible
3540 magnitude level for the audio input which will be normalized.
3541 If input frame volume is above this value frame will be normalized.
3542 Otherwise frame may not be normalized at all. The default value is set
3543 to 0, which means all input frames will be normalized.
3544 This option is mostly useful if digital noise is not wanted to be amplified.
3547 @subsection Commands
3549 This filter supports the all above options as @ref{commands}.
3553 Make audio easier to listen to on headphones.
3555 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3556 so that when listened to on headphones the stereo image is moved from
3557 inside your head (standard for headphones) to outside and in front of
3558 the listener (standard for speakers).
3564 Apply a two-pole peaking equalisation (EQ) filter. With this
3565 filter, the signal-level at and around a selected frequency can
3566 be increased or decreased, whilst (unlike bandpass and bandreject
3567 filters) that at all other frequencies is unchanged.
3569 In order to produce complex equalisation curves, this filter can
3570 be given several times, each with a different central frequency.
3572 The filter accepts the following options:
3576 Set the filter's central frequency in Hz.
3579 Set method to specify band-width of filter.
3594 Specify the band-width of a filter in width_type units.
3597 Set the required gain or attenuation in dB.
3598 Beware of clipping when using a positive gain.
3601 How much to use filtered signal in output. Default is 1.
3602 Range is between 0 and 1.
3605 Specify which channels to filter, by default all available are filtered.
3608 Normalize biquad coefficients, by default is disabled.
3609 Enabling it will normalize magnitude response at DC to 0dB.
3612 @subsection Examples
3615 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3617 equalizer=f=1000:t=h:width=200:g=-10
3621 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3623 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3627 @subsection Commands
3629 This filter supports the following commands:
3632 Change equalizer frequency.
3633 Syntax for the command is : "@var{frequency}"
3636 Change equalizer width_type.
3637 Syntax for the command is : "@var{width_type}"
3640 Change equalizer width.
3641 Syntax for the command is : "@var{width}"
3644 Change equalizer gain.
3645 Syntax for the command is : "@var{gain}"
3648 Change equalizer mix.
3649 Syntax for the command is : "@var{mix}"
3652 @section extrastereo
3654 Linearly increases the difference between left and right channels which
3655 adds some sort of "live" effect to playback.
3657 The filter accepts the following options:
3661 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3662 (average of both channels), with 1.0 sound will be unchanged, with
3663 -1.0 left and right channels will be swapped.
3666 Enable clipping. By default is enabled.
3669 @subsection Commands
3671 This filter supports the all above options as @ref{commands}.
3673 @section firequalizer
3674 Apply FIR Equalization using arbitrary frequency response.
3676 The filter accepts the following option:
3680 Set gain curve equation (in dB). The expression can contain variables:
3683 the evaluated frequency
3687 channel number, set to 0 when multichannels evaluation is disabled
3689 channel id, see libavutil/channel_layout.h, set to the first channel id when
3690 multichannels evaluation is disabled
3694 channel_layout, see libavutil/channel_layout.h
3699 @item gain_interpolate(f)
3700 interpolate gain on frequency f based on gain_entry
3701 @item cubic_interpolate(f)
3702 same as gain_interpolate, but smoother
3704 This option is also available as command. Default is @code{gain_interpolate(f)}.
3707 Set gain entry for gain_interpolate function. The expression can
3711 store gain entry at frequency f with value g
3713 This option is also available as command.
3716 Set filter delay in seconds. Higher value means more accurate.
3717 Default is @code{0.01}.
3720 Set filter accuracy in Hz. Lower value means more accurate.
3721 Default is @code{5}.
3724 Set window function. Acceptable values are:
3727 rectangular window, useful when gain curve is already smooth
3729 hann window (default)
3735 3-terms continuous 1st derivative nuttall window
3737 minimum 3-terms discontinuous nuttall window
3739 4-terms continuous 1st derivative nuttall window
3741 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3743 blackman-harris window
3749 If enabled, use fixed number of audio samples. This improves speed when
3750 filtering with large delay. Default is disabled.
3753 Enable multichannels evaluation on gain. Default is disabled.
3756 Enable zero phase mode by subtracting timestamp to compensate delay.
3757 Default is disabled.
3760 Set scale used by gain. Acceptable values are:
3763 linear frequency, linear gain
3765 linear frequency, logarithmic (in dB) gain (default)
3767 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3769 logarithmic frequency, logarithmic gain
3773 Set file for dumping, suitable for gnuplot.
3776 Set scale for dumpfile. Acceptable values are same with scale option.
3780 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3781 Default is disabled.
3784 Enable minimum phase impulse response. Default is disabled.
3787 @subsection Examples
3792 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3795 lowpass at 1000 Hz with gain_entry:
3797 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3800 custom equalization:
3802 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3805 higher delay with zero phase to compensate delay:
3807 firequalizer=delay=0.1:fixed=on:zero_phase=on
3810 lowpass on left channel, highpass on right channel:
3812 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3813 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3818 Apply a flanging effect to the audio.
3820 The filter accepts the following options:
3824 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3827 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3830 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3834 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3835 Default value is 71.
3838 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3841 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3842 Default value is @var{sinusoidal}.
3845 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3846 Default value is 25.
3849 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3850 Default is @var{linear}.
3854 Apply Haas effect to audio.
3856 Note that this makes most sense to apply on mono signals.
3857 With this filter applied to mono signals it give some directionality and
3858 stretches its stereo image.
3860 The filter accepts the following options:
3864 Set input level. By default is @var{1}, or 0dB
3867 Set output level. By default is @var{1}, or 0dB.
3870 Set gain applied to side part of signal. By default is @var{1}.
3873 Set kind of middle source. Can be one of the following:
3883 Pick middle part signal of stereo image.
3886 Pick side part signal of stereo image.
3890 Change middle phase. By default is disabled.
3893 Set left channel delay. By default is @var{2.05} milliseconds.
3896 Set left channel balance. By default is @var{-1}.
3899 Set left channel gain. By default is @var{1}.
3902 Change left phase. By default is disabled.
3905 Set right channel delay. By defaults is @var{2.12} milliseconds.
3908 Set right channel balance. By default is @var{1}.
3911 Set right channel gain. By default is @var{1}.
3914 Change right phase. By default is enabled.
3919 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3920 embedded HDCD codes is expanded into a 20-bit PCM stream.
3922 The filter supports the Peak Extend and Low-level Gain Adjustment features
3923 of HDCD, and detects the Transient Filter flag.
3926 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3929 When using the filter with wav, note the default encoding for wav is 16-bit,
3930 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3931 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3933 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3934 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3937 The filter accepts the following options:
3940 @item disable_autoconvert
3941 Disable any automatic format conversion or resampling in the filter graph.
3943 @item process_stereo
3944 Process the stereo channels together. If target_gain does not match between
3945 channels, consider it invalid and use the last valid target_gain.
3948 Set the code detect timer period in ms.
3951 Always extend peaks above -3dBFS even if PE isn't signaled.
3954 Replace audio with a solid tone and adjust the amplitude to signal some
3955 specific aspect of the decoding process. The output file can be loaded in
3956 an audio editor alongside the original to aid analysis.
3958 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3965 Gain adjustment level at each sample
3967 Samples where peak extend occurs
3969 Samples where the code detect timer is active
3971 Samples where the target gain does not match between channels
3977 Apply head-related transfer functions (HRTFs) to create virtual
3978 loudspeakers around the user for binaural listening via headphones.
3979 The HRIRs are provided via additional streams, for each channel
3980 one stereo input stream is needed.
3982 The filter accepts the following options:
3986 Set mapping of input streams for convolution.
3987 The argument is a '|'-separated list of channel names in order as they
3988 are given as additional stream inputs for filter.
3989 This also specify number of input streams. Number of input streams
3990 must be not less than number of channels in first stream plus one.
3993 Set gain applied to audio. Value is in dB. Default is 0.
3996 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3997 processing audio in time domain which is slow.
3998 @var{freq} is processing audio in frequency domain which is fast.
3999 Default is @var{freq}.
4002 Set custom gain for LFE channels. Value is in dB. Default is 0.
4005 Set size of frame in number of samples which will be processed at once.
4006 Default value is @var{1024}. Allowed range is from 1024 to 96000.
4009 Set format of hrir stream.
4010 Default value is @var{stereo}. Alternative value is @var{multich}.
4011 If value is set to @var{stereo}, number of additional streams should
4012 be greater or equal to number of input channels in first input stream.
4013 Also each additional stream should have stereo number of channels.
4014 If value is set to @var{multich}, number of additional streams should
4015 be exactly one. Also number of input channels of additional stream
4016 should be equal or greater than twice number of channels of first input
4020 @subsection Examples
4024 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4025 each amovie filter use stereo file with IR coefficients as input.
4026 The files give coefficients for each position of virtual loudspeaker:
4029 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
4034 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4035 but now in @var{multich} @var{hrir} format.
4037 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
4044 Apply a high-pass filter with 3dB point frequency.
4045 The filter can be either single-pole, or double-pole (the default).
4046 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4048 The filter accepts the following options:
4052 Set frequency in Hz. Default is 3000.
4055 Set number of poles. Default is 2.
4058 Set method to specify band-width of filter.
4073 Specify the band-width of a filter in width_type units.
4074 Applies only to double-pole filter.
4075 The default is 0.707q and gives a Butterworth response.
4078 How much to use filtered signal in output. Default is 1.
4079 Range is between 0 and 1.
4082 Specify which channels to filter, by default all available are filtered.
4085 Normalize biquad coefficients, by default is disabled.
4086 Enabling it will normalize magnitude response at DC to 0dB.
4089 @subsection Commands
4091 This filter supports the following commands:
4094 Change highpass frequency.
4095 Syntax for the command is : "@var{frequency}"
4098 Change highpass width_type.
4099 Syntax for the command is : "@var{width_type}"
4102 Change highpass width.
4103 Syntax for the command is : "@var{width}"
4106 Change highpass mix.
4107 Syntax for the command is : "@var{mix}"
4112 Join multiple input streams into one multi-channel stream.
4114 It accepts the following parameters:
4118 The number of input streams. It defaults to 2.
4120 @item channel_layout
4121 The desired output channel layout. It defaults to stereo.
4124 Map channels from inputs to output. The argument is a '|'-separated list of
4125 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
4126 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
4127 can be either the name of the input channel (e.g. FL for front left) or its
4128 index in the specified input stream. @var{out_channel} is the name of the output
4132 The filter will attempt to guess the mappings when they are not specified
4133 explicitly. It does so by first trying to find an unused matching input channel
4134 and if that fails it picks the first unused input channel.
4136 Join 3 inputs (with properly set channel layouts):
4138 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
4141 Build a 5.1 output from 6 single-channel streams:
4143 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
4144 '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'
4150 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
4152 To enable compilation of this filter you need to configure FFmpeg with
4153 @code{--enable-ladspa}.
4157 Specifies the name of LADSPA plugin library to load. If the environment
4158 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
4159 each one of the directories specified by the colon separated list in
4160 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
4161 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
4162 @file{/usr/lib/ladspa/}.
4165 Specifies the plugin within the library. Some libraries contain only
4166 one plugin, but others contain many of them. If this is not set filter
4167 will list all available plugins within the specified library.
4170 Set the '|' separated list of controls which are zero or more floating point
4171 values that determine the behavior of the loaded plugin (for example delay,
4173 Controls need to be defined using the following syntax:
4174 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
4175 @var{valuei} is the value set on the @var{i}-th control.
4176 Alternatively they can be also defined using the following syntax:
4177 @var{value0}|@var{value1}|@var{value2}|..., where
4178 @var{valuei} is the value set on the @var{i}-th control.
4179 If @option{controls} is set to @code{help}, all available controls and
4180 their valid ranges are printed.
4182 @item sample_rate, s
4183 Specify the sample rate, default to 44100. Only used if plugin have
4187 Set the number of samples per channel per each output frame, default
4188 is 1024. Only used if plugin have zero inputs.
4191 Set the minimum duration of the sourced audio. See
4192 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4193 for the accepted syntax.
4194 Note that the resulting duration may be greater than the specified duration,
4195 as the generated audio is always cut at the end of a complete frame.
4196 If not specified, or the expressed duration is negative, the audio is
4197 supposed to be generated forever.
4198 Only used if plugin have zero inputs.
4201 Enable latency compensation, by default is disabled.
4202 Only used if plugin have inputs.
4205 @subsection Examples
4209 List all available plugins within amp (LADSPA example plugin) library:
4215 List all available controls and their valid ranges for @code{vcf_notch}
4216 plugin from @code{VCF} library:
4218 ladspa=f=vcf:p=vcf_notch:c=help
4222 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
4225 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
4229 Add reverberation to the audio using TAP-plugins
4230 (Tom's Audio Processing plugins):
4232 ladspa=file=tap_reverb:tap_reverb
4236 Generate white noise, with 0.2 amplitude:
4238 ladspa=file=cmt:noise_source_white:c=c0=.2
4242 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
4243 @code{C* Audio Plugin Suite} (CAPS) library:
4245 ladspa=file=caps:Click:c=c1=20'
4249 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
4251 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
4255 Increase volume by 20dB using fast lookahead limiter from Steve Harris
4256 @code{SWH Plugins} collection:
4258 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
4262 Attenuate low frequencies using Multiband EQ from Steve Harris
4263 @code{SWH Plugins} collection:
4265 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
4269 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
4272 ladspa=caps:Narrower
4276 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
4278 ladspa=caps:White:.2
4282 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
4284 ladspa=caps:Fractal:c=c1=1
4288 Dynamic volume normalization using @code{VLevel} plugin:
4290 ladspa=vlevel-ladspa:vlevel_mono
4294 @subsection Commands
4296 This filter supports the following commands:
4299 Modify the @var{N}-th control value.
4301 If the specified value is not valid, it is ignored and prior one is kept.
4306 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4307 Support for both single pass (livestreams, files) and double pass (files) modes.
4308 This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately
4309 detect true peaks, the audio stream will be upsampled to 192 kHz.
4310 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4312 The filter accepts the following options:
4316 Set integrated loudness target.
4317 Range is -70.0 - -5.0. Default value is -24.0.
4320 Set loudness range target.
4321 Range is 1.0 - 20.0. Default value is 7.0.
4324 Set maximum true peak.
4325 Range is -9.0 - +0.0. Default value is -2.0.
4327 @item measured_I, measured_i
4328 Measured IL of input file.
4329 Range is -99.0 - +0.0.
4331 @item measured_LRA, measured_lra
4332 Measured LRA of input file.
4333 Range is 0.0 - 99.0.
4335 @item measured_TP, measured_tp
4336 Measured true peak of input file.
4337 Range is -99.0 - +99.0.
4339 @item measured_thresh
4340 Measured threshold of input file.
4341 Range is -99.0 - +0.0.
4344 Set offset gain. Gain is applied before the true-peak limiter.
4345 Range is -99.0 - +99.0. Default is +0.0.
4348 Normalize by linearly scaling the source audio.
4349 @code{measured_I}, @code{measured_LRA}, @code{measured_TP},
4350 and @code{measured_thresh} must all be specified. Target LRA shouldn't
4351 be lower than source LRA and the change in integrated loudness shouldn't
4352 result in a true peak which exceeds the target TP. If any of these
4353 conditions aren't met, normalization mode will revert to @var{dynamic}.
4354 Options are @code{true} or @code{false}. Default is @code{true}.
4357 Treat mono input files as "dual-mono". If a mono file is intended for playback
4358 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4359 If set to @code{true}, this option will compensate for this effect.
4360 Multi-channel input files are not affected by this option.
4361 Options are true or false. Default is false.
4364 Set print format for stats. Options are summary, json, or none.
4365 Default value is none.
4370 Apply a low-pass filter with 3dB point frequency.
4371 The filter can be either single-pole or double-pole (the default).
4372 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4374 The filter accepts the following options:
4378 Set frequency in Hz. Default is 500.
4381 Set number of poles. Default is 2.
4384 Set method to specify band-width of filter.
4399 Specify the band-width of a filter in width_type units.
4400 Applies only to double-pole filter.
4401 The default is 0.707q and gives a Butterworth response.
4404 How much to use filtered signal in output. Default is 1.
4405 Range is between 0 and 1.
4408 Specify which channels to filter, by default all available are filtered.
4411 Normalize biquad coefficients, by default is disabled.
4412 Enabling it will normalize magnitude response at DC to 0dB.
4415 @subsection Examples
4418 Lowpass only LFE channel, it LFE is not present it does nothing:
4424 @subsection Commands
4426 This filter supports the following commands:
4429 Change lowpass frequency.
4430 Syntax for the command is : "@var{frequency}"
4433 Change lowpass width_type.
4434 Syntax for the command is : "@var{width_type}"
4437 Change lowpass width.
4438 Syntax for the command is : "@var{width}"
4442 Syntax for the command is : "@var{mix}"
4447 Load a LV2 (LADSPA Version 2) plugin.
4449 To enable compilation of this filter you need to configure FFmpeg with
4450 @code{--enable-lv2}.
4454 Specifies the plugin URI. You may need to escape ':'.
4457 Set the '|' separated list of controls which are zero or more floating point
4458 values that determine the behavior of the loaded plugin (for example delay,
4460 If @option{controls} is set to @code{help}, all available controls and
4461 their valid ranges are printed.
4463 @item sample_rate, s
4464 Specify the sample rate, default to 44100. Only used if plugin have
4468 Set the number of samples per channel per each output frame, default
4469 is 1024. Only used if plugin have zero inputs.
4472 Set the minimum duration of the sourced audio. See
4473 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4474 for the accepted syntax.
4475 Note that the resulting duration may be greater than the specified duration,
4476 as the generated audio is always cut at the end of a complete frame.
4477 If not specified, or the expressed duration is negative, the audio is
4478 supposed to be generated forever.
4479 Only used if plugin have zero inputs.
4482 @subsection Examples
4486 Apply bass enhancer plugin from Calf:
4488 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4492 Apply vinyl plugin from Calf:
4494 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4498 Apply bit crusher plugin from ArtyFX:
4500 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4505 Multiband Compress or expand the audio's dynamic range.
4507 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4508 This is akin to the crossover of a loudspeaker, and results in flat frequency
4509 response when absent compander action.
4511 It accepts the following parameters:
4515 This option syntax is:
4516 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4517 For explanation of each item refer to compand filter documentation.
4523 Mix channels with specific gain levels. The filter accepts the output
4524 channel layout followed by a set of channels definitions.
4526 This filter is also designed to efficiently remap the channels of an audio
4529 The filter accepts parameters of the form:
4530 "@var{l}|@var{outdef}|@var{outdef}|..."
4534 output channel layout or number of channels
4537 output channel specification, of the form:
4538 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4541 output channel to define, either a channel name (FL, FR, etc.) or a channel
4542 number (c0, c1, etc.)
4545 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4548 input channel to use, see out_name for details; it is not possible to mix
4549 named and numbered input channels
4552 If the `=' in a channel specification is replaced by `<', then the gains for
4553 that specification will be renormalized so that the total is 1, thus
4554 avoiding clipping noise.
4556 @subsection Mixing examples
4558 For example, if you want to down-mix from stereo to mono, but with a bigger
4559 factor for the left channel:
4561 pan=1c|c0=0.9*c0+0.1*c1
4564 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4565 7-channels surround:
4567 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4570 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4571 that should be preferred (see "-ac" option) unless you have very specific
4574 @subsection Remapping examples
4576 The channel remapping will be effective if, and only if:
4579 @item gain coefficients are zeroes or ones,
4580 @item only one input per channel output,
4583 If all these conditions are satisfied, the filter will notify the user ("Pure
4584 channel mapping detected"), and use an optimized and lossless method to do the
4587 For example, if you have a 5.1 source and want a stereo audio stream by
4588 dropping the extra channels:
4590 pan="stereo| c0=FL | c1=FR"
4593 Given the same source, you can also switch front left and front right channels
4594 and keep the input channel layout:
4596 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4599 If the input is a stereo audio stream, you can mute the front left channel (and
4600 still keep the stereo channel layout) with:
4605 Still with a stereo audio stream input, you can copy the right channel in both
4606 front left and right:
4608 pan="stereo| c0=FR | c1=FR"
4613 ReplayGain scanner filter. This filter takes an audio stream as an input and
4614 outputs it unchanged.
4615 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4619 Convert the audio sample format, sample rate and channel layout. It is
4620 not meant to be used directly.
4623 Apply time-stretching and pitch-shifting with librubberband.
4625 To enable compilation of this filter, you need to configure FFmpeg with
4626 @code{--enable-librubberband}.
4628 The filter accepts the following options:
4632 Set tempo scale factor.
4635 Set pitch scale factor.
4638 Set transients detector.
4639 Possible values are:
4648 Possible values are:
4657 Possible values are:
4664 Set processing window size.
4665 Possible values are:
4674 Possible values are:
4681 Enable formant preservation when shift pitching.
4682 Possible values are:
4690 Possible values are:
4699 Possible values are:
4706 @subsection Commands
4708 This filter supports the following commands:
4711 Change filter tempo scale factor.
4712 Syntax for the command is : "@var{tempo}"
4715 Change filter pitch scale factor.
4716 Syntax for the command is : "@var{pitch}"
4719 @section sidechaincompress
4721 This filter acts like normal compressor but has the ability to compress
4722 detected signal using second input signal.
4723 It needs two input streams and returns one output stream.
4724 First input stream will be processed depending on second stream signal.
4725 The filtered signal then can be filtered with other filters in later stages of
4726 processing. See @ref{pan} and @ref{amerge} filter.
4728 The filter accepts the following options:
4732 Set input gain. Default is 1. Range is between 0.015625 and 64.
4735 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
4736 Default is @code{downward}.
4739 If a signal of second stream raises above this level it will affect the gain
4740 reduction of first stream.
4741 By default is 0.125. Range is between 0.00097563 and 1.
4744 Set a ratio about which the signal is reduced. 1:2 means that if the level
4745 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4746 Default is 2. Range is between 1 and 20.
4749 Amount of milliseconds the signal has to rise above the threshold before gain
4750 reduction starts. Default is 20. Range is between 0.01 and 2000.
4753 Amount of milliseconds the signal has to fall below the threshold before
4754 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4757 Set the amount by how much signal will be amplified after processing.
4758 Default is 1. Range is from 1 to 64.
4761 Curve the sharp knee around the threshold to enter gain reduction more softly.
4762 Default is 2.82843. Range is between 1 and 8.
4765 Choose if the @code{average} level between all channels of side-chain stream
4766 or the louder(@code{maximum}) channel of side-chain stream affects the
4767 reduction. Default is @code{average}.
4770 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4771 of @code{rms}. Default is @code{rms} which is mainly smoother.
4774 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4777 How much to use compressed signal in output. Default is 1.
4778 Range is between 0 and 1.
4781 @subsection Commands
4783 This filter supports the all above options as @ref{commands}.
4785 @subsection Examples
4789 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4790 depending on the signal of 2nd input and later compressed signal to be
4791 merged with 2nd input:
4793 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4797 @section sidechaingate
4799 A sidechain gate acts like a normal (wideband) gate but has the ability to
4800 filter the detected signal before sending it to the gain reduction stage.
4801 Normally a gate uses the full range signal to detect a level above the
4803 For example: If you cut all lower frequencies from your sidechain signal
4804 the gate will decrease the volume of your track only if not enough highs
4805 appear. With this technique you are able to reduce the resonation of a
4806 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4808 It needs two input streams and returns one output stream.
4809 First input stream will be processed depending on second stream signal.
4811 The filter accepts the following options:
4815 Set input level before filtering.
4816 Default is 1. Allowed range is from 0.015625 to 64.
4819 Set the mode of operation. Can be @code{upward} or @code{downward}.
4820 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
4821 will be amplified, expanding dynamic range in upward direction.
4822 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
4825 Set the level of gain reduction when the signal is below the threshold.
4826 Default is 0.06125. Allowed range is from 0 to 1.
4827 Setting this to 0 disables reduction and then filter behaves like expander.
4830 If a signal rises above this level the gain reduction is released.
4831 Default is 0.125. Allowed range is from 0 to 1.
4834 Set a ratio about which the signal is reduced.
4835 Default is 2. Allowed range is from 1 to 9000.
4838 Amount of milliseconds the signal has to rise above the threshold before gain
4840 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4843 Amount of milliseconds the signal has to fall below the threshold before the
4844 reduction is increased again. Default is 250 milliseconds.
4845 Allowed range is from 0.01 to 9000.
4848 Set amount of amplification of signal after processing.
4849 Default is 1. Allowed range is from 1 to 64.
4852 Curve the sharp knee around the threshold to enter gain reduction more softly.
4853 Default is 2.828427125. Allowed range is from 1 to 8.
4856 Choose if exact signal should be taken for detection or an RMS like one.
4857 Default is rms. Can be peak or rms.
4860 Choose if the average level between all channels or the louder channel affects
4862 Default is average. Can be average or maximum.
4865 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4868 @section silencedetect
4870 Detect silence in an audio stream.
4872 This filter logs a message when it detects that the input audio volume is less
4873 or equal to a noise tolerance value for a duration greater or equal to the
4874 minimum detected noise duration.
4876 The printed times and duration are expressed in seconds. The
4877 @code{lavfi.silence_start} or @code{lavfi.silence_start.X} metadata key
4878 is set on the first frame whose timestamp equals or exceeds the detection
4879 duration and it contains the timestamp of the first frame of the silence.
4881 The @code{lavfi.silence_duration} or @code{lavfi.silence_duration.X}
4882 and @code{lavfi.silence_end} or @code{lavfi.silence_end.X} metadata
4883 keys are set on the first frame after the silence. If @option{mono} is
4884 enabled, and each channel is evaluated separately, the @code{.X}
4885 suffixed keys are used, and @code{X} corresponds to the channel number.
4887 The filter accepts the following options:
4891 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4892 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4895 Set silence duration until notification (default is 2 seconds). See
4896 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4897 for the accepted syntax.
4900 Process each channel separately, instead of combined. By default is disabled.
4903 @subsection Examples
4907 Detect 5 seconds of silence with -50dB noise tolerance:
4909 silencedetect=n=-50dB:d=5
4913 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4914 tolerance in @file{silence.mp3}:
4916 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4920 @section silenceremove
4922 Remove silence from the beginning, middle or end of the audio.
4924 The filter accepts the following options:
4928 This value is used to indicate if audio should be trimmed at beginning of
4929 the audio. A value of zero indicates no silence should be trimmed from the
4930 beginning. When specifying a non-zero value, it trims audio up until it
4931 finds non-silence. Normally, when trimming silence from beginning of audio
4932 the @var{start_periods} will be @code{1} but it can be increased to higher
4933 values to trim all audio up to specific count of non-silence periods.
4934 Default value is @code{0}.
4936 @item start_duration
4937 Specify the amount of time that non-silence must be detected before it stops
4938 trimming audio. By increasing the duration, bursts of noises can be treated
4939 as silence and trimmed off. Default value is @code{0}.
4941 @item start_threshold
4942 This indicates what sample value should be treated as silence. For digital
4943 audio, a value of @code{0} may be fine but for audio recorded from analog,
4944 you may wish to increase the value to account for background noise.
4945 Can be specified in dB (in case "dB" is appended to the specified value)
4946 or amplitude ratio. Default value is @code{0}.
4949 Specify max duration of silence at beginning that will be kept after
4950 trimming. Default is 0, which is equal to trimming all samples detected
4954 Specify mode of detection of silence end in start of multi-channel audio.
4955 Can be @var{any} or @var{all}. Default is @var{any}.
4956 With @var{any}, any sample that is detected as non-silence will cause
4957 stopped trimming of silence.
4958 With @var{all}, only if all channels are detected as non-silence will cause
4959 stopped trimming of silence.
4962 Set the count for trimming silence from the end of audio.
4963 To remove silence from the middle of a file, specify a @var{stop_periods}
4964 that is negative. This value is then treated as a positive value and is
4965 used to indicate the effect should restart processing as specified by
4966 @var{start_periods}, making it suitable for removing periods of silence
4967 in the middle of the audio.
4968 Default value is @code{0}.
4971 Specify a duration of silence that must exist before audio is not copied any
4972 more. By specifying a higher duration, silence that is wanted can be left in
4974 Default value is @code{0}.
4976 @item stop_threshold
4977 This is the same as @option{start_threshold} but for trimming silence from
4979 Can be specified in dB (in case "dB" is appended to the specified value)
4980 or amplitude ratio. Default value is @code{0}.
4983 Specify max duration of silence at end that will be kept after
4984 trimming. Default is 0, which is equal to trimming all samples detected
4988 Specify mode of detection of silence start in end of multi-channel audio.
4989 Can be @var{any} or @var{all}. Default is @var{any}.
4990 With @var{any}, any sample that is detected as non-silence will cause
4991 stopped trimming of silence.
4992 With @var{all}, only if all channels are detected as non-silence will cause
4993 stopped trimming of silence.
4996 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4997 and works better with digital silence which is exactly 0.
4998 Default value is @code{rms}.
5001 Set duration in number of seconds used to calculate size of window in number
5002 of samples for detecting silence.
5003 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
5006 @subsection Examples
5010 The following example shows how this filter can be used to start a recording
5011 that does not contain the delay at the start which usually occurs between
5012 pressing the record button and the start of the performance:
5014 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
5018 Trim all silence encountered from beginning to end where there is more than 1
5019 second of silence in audio:
5021 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
5025 Trim all digital silence samples, using peak detection, from beginning to end
5026 where there is more than 0 samples of digital silence in audio and digital
5027 silence is detected in all channels at same positions in stream:
5029 silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
5035 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
5036 loudspeakers around the user for binaural listening via headphones (audio
5037 formats up to 9 channels supported).
5038 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
5039 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
5040 Austrian Academy of Sciences.
5042 To enable compilation of this filter you need to configure FFmpeg with
5043 @code{--enable-libmysofa}.
5045 The filter accepts the following options:
5049 Set the SOFA file used for rendering.
5052 Set gain applied to audio. Value is in dB. Default is 0.
5055 Set rotation of virtual loudspeakers in deg. Default is 0.
5058 Set elevation of virtual speakers in deg. Default is 0.
5061 Set distance in meters between loudspeakers and the listener with near-field
5062 HRTFs. Default is 1.
5065 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
5066 processing audio in time domain which is slow.
5067 @var{freq} is processing audio in frequency domain which is fast.
5068 Default is @var{freq}.
5071 Set custom positions of virtual loudspeakers. Syntax for this option is:
5072 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
5073 Each virtual loudspeaker is described with short channel name following with
5074 azimuth and elevation in degrees.
5075 Each virtual loudspeaker description is separated by '|'.
5076 For example to override front left and front right channel positions use:
5077 'speakers=FL 45 15|FR 345 15'.
5078 Descriptions with unrecognised channel names are ignored.
5081 Set custom gain for LFE channels. Value is in dB. Default is 0.
5084 Set custom frame size in number of samples. Default is 1024.
5085 Allowed range is from 1024 to 96000. Only used if option @samp{type}
5086 is set to @var{freq}.
5089 Should all IRs be normalized upon importing SOFA file.
5090 By default is enabled.
5093 Should nearest IRs be interpolated with neighbor IRs if exact position
5094 does not match. By default is disabled.
5097 Minphase all IRs upon loading of SOFA file. By default is disabled.
5100 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
5103 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
5106 @subsection Examples
5110 Using ClubFritz6 sofa file:
5112 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
5116 Using ClubFritz12 sofa file and bigger radius with small rotation:
5118 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
5122 Similar as above but with custom speaker positions for front left, front right, back left and back right
5123 and also with custom gain:
5125 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
5129 @section stereotools
5131 This filter has some handy utilities to manage stereo signals, for converting
5132 M/S stereo recordings to L/R signal while having control over the parameters
5133 or spreading the stereo image of master track.
5135 The filter accepts the following options:
5139 Set input level before filtering for both channels. Defaults is 1.
5140 Allowed range is from 0.015625 to 64.
5143 Set output level after filtering for both channels. Defaults is 1.
5144 Allowed range is from 0.015625 to 64.
5147 Set input balance between both channels. Default is 0.
5148 Allowed range is from -1 to 1.
5151 Set output balance between both channels. Default is 0.
5152 Allowed range is from -1 to 1.
5155 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
5156 clipping. Disabled by default.
5159 Mute the left channel. Disabled by default.
5162 Mute the right channel. Disabled by default.
5165 Change the phase of the left channel. Disabled by default.
5168 Change the phase of the right channel. Disabled by default.
5171 Set stereo mode. Available values are:
5175 Left/Right to Left/Right, this is default.
5178 Left/Right to Mid/Side.
5181 Mid/Side to Left/Right.
5184 Left/Right to Left/Left.
5187 Left/Right to Right/Right.
5190 Left/Right to Left + Right.
5193 Left/Right to Right/Left.
5196 Mid/Side to Left/Left.
5199 Mid/Side to Right/Right.
5203 Set level of side signal. Default is 1.
5204 Allowed range is from 0.015625 to 64.
5207 Set balance of side signal. Default is 0.
5208 Allowed range is from -1 to 1.
5211 Set level of the middle signal. Default is 1.
5212 Allowed range is from 0.015625 to 64.
5215 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
5218 Set stereo base between mono and inversed channels. Default is 0.
5219 Allowed range is from -1 to 1.
5222 Set delay in milliseconds how much to delay left from right channel and
5223 vice versa. Default is 0. Allowed range is from -20 to 20.
5226 Set S/C level. Default is 1. Allowed range is from 1 to 100.
5229 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
5231 @item bmode_in, bmode_out
5232 Set balance mode for balance_in/balance_out option.
5234 Can be one of the following:
5238 Classic balance mode. Attenuate one channel at time.
5239 Gain is raised up to 1.
5242 Similar as classic mode above but gain is raised up to 2.
5245 Equal power distribution, from -6dB to +6dB range.
5249 @subsection Examples
5253 Apply karaoke like effect:
5255 stereotools=mlev=0.015625
5259 Convert M/S signal to L/R:
5261 "stereotools=mode=ms>lr"
5265 @section stereowiden
5267 This filter enhance the stereo effect by suppressing signal common to both
5268 channels and by delaying the signal of left into right and vice versa,
5269 thereby widening the stereo effect.
5271 The filter accepts the following options:
5275 Time in milliseconds of the delay of left signal into right and vice versa.
5276 Default is 20 milliseconds.
5279 Amount of gain in delayed signal into right and vice versa. Gives a delay
5280 effect of left signal in right output and vice versa which gives widening
5281 effect. Default is 0.3.
5284 Cross feed of left into right with inverted phase. This helps in suppressing
5285 the mono. If the value is 1 it will cancel all the signal common to both
5286 channels. Default is 0.3.
5289 Set level of input signal of original channel. Default is 0.8.
5292 @subsection Commands
5294 This filter supports the all above options except @code{delay} as @ref{commands}.
5296 @section superequalizer
5297 Apply 18 band equalizer.
5299 The filter accepts the following options:
5306 Set 131Hz band gain.
5308 Set 185Hz band gain.
5310 Set 262Hz band gain.
5312 Set 370Hz band gain.
5314 Set 523Hz band gain.
5316 Set 740Hz band gain.
5318 Set 1047Hz band gain.
5320 Set 1480Hz band gain.
5322 Set 2093Hz band gain.
5324 Set 2960Hz band gain.
5326 Set 4186Hz band gain.
5328 Set 5920Hz band gain.
5330 Set 8372Hz band gain.
5332 Set 11840Hz band gain.
5334 Set 16744Hz band gain.
5336 Set 20000Hz band gain.
5340 Apply audio surround upmix filter.
5342 This filter allows to produce multichannel output from audio stream.
5344 The filter accepts the following options:
5348 Set output channel layout. By default, this is @var{5.1}.
5350 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5351 for the required syntax.
5354 Set input channel layout. By default, this is @var{stereo}.
5356 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5357 for the required syntax.
5360 Set input volume level. By default, this is @var{1}.
5363 Set output volume level. By default, this is @var{1}.
5366 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5369 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5372 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5375 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5376 In @var{add} mode, LFE channel is created from input audio and added to output.
5377 In @var{sub} mode, LFE channel is created from input audio and added to output but
5378 also all non-LFE output channels are subtracted with output LFE channel.
5381 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5382 Default is @var{90}.
5385 Set front center input volume. By default, this is @var{1}.
5388 Set front center output volume. By default, this is @var{1}.
5391 Set front left input volume. By default, this is @var{1}.
5394 Set front left output volume. By default, this is @var{1}.
5397 Set front right input volume. By default, this is @var{1}.
5400 Set front right output volume. By default, this is @var{1}.
5403 Set side left input volume. By default, this is @var{1}.
5406 Set side left output volume. By default, this is @var{1}.
5409 Set side right input volume. By default, this is @var{1}.
5412 Set side right output volume. By default, this is @var{1}.
5415 Set back left input volume. By default, this is @var{1}.
5418 Set back left output volume. By default, this is @var{1}.
5421 Set back right input volume. By default, this is @var{1}.
5424 Set back right output volume. By default, this is @var{1}.
5427 Set back center input volume. By default, this is @var{1}.
5430 Set back center output volume. By default, this is @var{1}.
5433 Set LFE input volume. By default, this is @var{1}.
5436 Set LFE output volume. By default, this is @var{1}.
5439 Set spread usage of stereo image across X axis for all channels.
5442 Set spread usage of stereo image across Y axis for all channels.
5444 @item fcx, flx, frx, blx, brx, slx, srx, bcx
5445 Set spread usage of stereo image across X axis for each channel.
5447 @item fcy, fly, fry, bly, bry, sly, sry, bcy
5448 Set spread usage of stereo image across Y axis for each channel.
5451 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
5454 Set window function.
5456 It accepts the following values:
5479 Default is @code{hann}.
5482 Set window overlap. If set to 1, the recommended overlap for selected
5483 window function will be picked. Default is @code{0.5}.
5486 @section treble, highshelf
5488 Boost or cut treble (upper) frequencies of the audio using a two-pole
5489 shelving filter with a response similar to that of a standard
5490 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
5492 The filter accepts the following options:
5496 Give the gain at whichever is the lower of ~22 kHz and the
5497 Nyquist frequency. Its useful range is about -20 (for a large cut)
5498 to +20 (for a large boost). Beware of clipping when using a positive gain.
5501 Set the filter's central frequency and so can be used
5502 to extend or reduce the frequency range to be boosted or cut.
5503 The default value is @code{3000} Hz.
5506 Set method to specify band-width of filter.
5521 Determine how steep is the filter's shelf transition.
5524 How much to use filtered signal in output. Default is 1.
5525 Range is between 0 and 1.
5528 Specify which channels to filter, by default all available are filtered.
5531 Normalize biquad coefficients, by default is disabled.
5532 Enabling it will normalize magnitude response at DC to 0dB.
5535 @subsection Commands
5537 This filter supports the following commands:
5540 Change treble frequency.
5541 Syntax for the command is : "@var{frequency}"
5544 Change treble width_type.
5545 Syntax for the command is : "@var{width_type}"
5548 Change treble width.
5549 Syntax for the command is : "@var{width}"
5553 Syntax for the command is : "@var{gain}"
5557 Syntax for the command is : "@var{mix}"
5562 Sinusoidal amplitude modulation.
5564 The filter accepts the following options:
5568 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
5569 (20 Hz or lower) will result in a tremolo effect.
5570 This filter may also be used as a ring modulator by specifying
5571 a modulation frequency higher than 20 Hz.
5572 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5575 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5576 Default value is 0.5.
5581 Sinusoidal phase modulation.
5583 The filter accepts the following options:
5587 Modulation frequency in Hertz.
5588 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5591 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5592 Default value is 0.5.
5597 Adjust the input audio volume.
5599 It accepts the following parameters:
5603 Set audio volume expression.
5605 Output values are clipped to the maximum value.
5607 The output audio volume is given by the relation:
5609 @var{output_volume} = @var{volume} * @var{input_volume}
5612 The default value for @var{volume} is "1.0".
5615 This parameter represents the mathematical precision.
5617 It determines which input sample formats will be allowed, which affects the
5618 precision of the volume scaling.
5622 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5624 32-bit floating-point; this limits input sample format to FLT. (default)
5626 64-bit floating-point; this limits input sample format to DBL.
5630 Choose the behaviour on encountering ReplayGain side data in input frames.
5634 Remove ReplayGain side data, ignoring its contents (the default).
5637 Ignore ReplayGain side data, but leave it in the frame.
5640 Prefer the track gain, if present.
5643 Prefer the album gain, if present.
5646 @item replaygain_preamp
5647 Pre-amplification gain in dB to apply to the selected replaygain gain.
5649 Default value for @var{replaygain_preamp} is 0.0.
5651 @item replaygain_noclip
5652 Prevent clipping by limiting the gain applied.
5654 Default value for @var{replaygain_noclip} is 1.
5657 Set when the volume expression is evaluated.
5659 It accepts the following values:
5662 only evaluate expression once during the filter initialization, or
5663 when the @samp{volume} command is sent
5666 evaluate expression for each incoming frame
5669 Default value is @samp{once}.
5672 The volume expression can contain the following parameters.
5676 frame number (starting at zero)
5679 @item nb_consumed_samples
5680 number of samples consumed by the filter
5682 number of samples in the current frame
5684 original frame position in the file
5690 PTS at start of stream
5692 time at start of stream
5698 last set volume value
5701 Note that when @option{eval} is set to @samp{once} only the
5702 @var{sample_rate} and @var{tb} variables are available, all other
5703 variables will evaluate to NAN.
5705 @subsection Commands
5707 This filter supports the following commands:
5710 Modify the volume expression.
5711 The command accepts the same syntax of the corresponding option.
5713 If the specified expression is not valid, it is kept at its current
5717 @subsection Examples
5721 Halve the input audio volume:
5725 volume=volume=-6.0206dB
5728 In all the above example the named key for @option{volume} can be
5729 omitted, for example like in:
5735 Increase input audio power by 6 decibels using fixed-point precision:
5737 volume=volume=6dB:precision=fixed
5741 Fade volume after time 10 with an annihilation period of 5 seconds:
5743 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5747 @section volumedetect
5749 Detect the volume of the input video.
5751 The filter has no parameters. The input is not modified. Statistics about
5752 the volume will be printed in the log when the input stream end is reached.
5754 In particular it will show the mean volume (root mean square), maximum
5755 volume (on a per-sample basis), and the beginning of a histogram of the
5756 registered volume values (from the maximum value to a cumulated 1/1000 of
5759 All volumes are in decibels relative to the maximum PCM value.
5761 @subsection Examples
5763 Here is an excerpt of the output:
5765 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5766 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5767 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5768 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5769 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5770 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5771 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5772 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5773 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5779 The mean square energy is approximately -27 dB, or 10^-2.7.
5781 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5783 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5786 In other words, raising the volume by +4 dB does not cause any clipping,
5787 raising it by +5 dB causes clipping for 6 samples, etc.
5789 @c man end AUDIO FILTERS
5791 @chapter Audio Sources
5792 @c man begin AUDIO SOURCES
5794 Below is a description of the currently available audio sources.
5798 Buffer audio frames, and make them available to the filter chain.
5800 This source is mainly intended for a programmatic use, in particular
5801 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5803 It accepts the following parameters:
5807 The timebase which will be used for timestamps of submitted frames. It must be
5808 either a floating-point number or in @var{numerator}/@var{denominator} form.
5811 The sample rate of the incoming audio buffers.
5814 The sample format of the incoming audio buffers.
5815 Either a sample format name or its corresponding integer representation from
5816 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5818 @item channel_layout
5819 The channel layout of the incoming audio buffers.
5820 Either a channel layout name from channel_layout_map in
5821 @file{libavutil/channel_layout.c} or its corresponding integer representation
5822 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5825 The number of channels of the incoming audio buffers.
5826 If both @var{channels} and @var{channel_layout} are specified, then they
5831 @subsection Examples
5834 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5837 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5838 Since the sample format with name "s16p" corresponds to the number
5839 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5842 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5847 Generate an audio signal specified by an expression.
5849 This source accepts in input one or more expressions (one for each
5850 channel), which are evaluated and used to generate a corresponding
5853 This source accepts the following options:
5857 Set the '|'-separated expressions list for each separate channel. In case the
5858 @option{channel_layout} option is not specified, the selected channel layout
5859 depends on the number of provided expressions. Otherwise the last
5860 specified expression is applied to the remaining output channels.
5862 @item channel_layout, c
5863 Set the channel layout. The number of channels in the specified layout
5864 must be equal to the number of specified expressions.
5867 Set the minimum duration of the sourced audio. See
5868 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5869 for the accepted syntax.
5870 Note that the resulting duration may be greater than the specified
5871 duration, as the generated audio is always cut at the end of a
5874 If not specified, or the expressed duration is negative, the audio is
5875 supposed to be generated forever.
5878 Set the number of samples per channel per each output frame,
5881 @item sample_rate, s
5882 Specify the sample rate, default to 44100.
5885 Each expression in @var{exprs} can contain the following constants:
5889 number of the evaluated sample, starting from 0
5892 time of the evaluated sample expressed in seconds, starting from 0
5899 @subsection Examples
5909 Generate a sin signal with frequency of 440 Hz, set sample rate to
5912 aevalsrc="sin(440*2*PI*t):s=8000"
5916 Generate a two channels signal, specify the channel layout (Front
5917 Center + Back Center) explicitly:
5919 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5923 Generate white noise:
5925 aevalsrc="-2+random(0)"
5929 Generate an amplitude modulated signal:
5931 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5935 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5937 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5944 Generate a FIR coefficients using frequency sampling method.
5946 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5948 The filter accepts the following options:
5952 Set number of filter coefficents in output audio stream.
5953 Default value is 1025.
5956 Set frequency points from where magnitude and phase are set.
5957 This must be in non decreasing order, and first element must be 0, while last element
5958 must be 1. Elements are separated by white spaces.
5961 Set magnitude value for every frequency point set by @option{frequency}.
5962 Number of values must be same as number of frequency points.
5963 Values are separated by white spaces.
5966 Set phase value for every frequency point set by @option{frequency}.
5967 Number of values must be same as number of frequency points.
5968 Values are separated by white spaces.
5970 @item sample_rate, r
5971 Set sample rate, default is 44100.
5974 Set number of samples per each frame. Default is 1024.
5977 Set window function. Default is blackman.
5982 The null audio source, return unprocessed audio frames. It is mainly useful
5983 as a template and to be employed in analysis / debugging tools, or as
5984 the source for filters which ignore the input data (for example the sox
5987 This source accepts the following options:
5991 @item channel_layout, cl
5993 Specifies the channel layout, and can be either an integer or a string
5994 representing a channel layout. The default value of @var{channel_layout}
5997 Check the channel_layout_map definition in
5998 @file{libavutil/channel_layout.c} for the mapping between strings and
5999 channel layout values.
6001 @item sample_rate, r
6002 Specifies the sample rate, and defaults to 44100.
6005 Set the number of samples per requested frames.
6009 @subsection Examples
6013 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
6015 anullsrc=r=48000:cl=4
6019 Do the same operation with a more obvious syntax:
6021 anullsrc=r=48000:cl=mono
6025 All the parameters need to be explicitly defined.
6029 Synthesize a voice utterance using the libflite library.
6031 To enable compilation of this filter you need to configure FFmpeg with
6032 @code{--enable-libflite}.
6034 Note that versions of the flite library prior to 2.0 are not thread-safe.
6036 The filter accepts the following options:
6041 If set to 1, list the names of the available voices and exit
6042 immediately. Default value is 0.
6045 Set the maximum number of samples per frame. Default value is 512.
6048 Set the filename containing the text to speak.
6051 Set the text to speak.
6054 Set the voice to use for the speech synthesis. Default value is
6055 @code{kal}. See also the @var{list_voices} option.
6058 @subsection Examples
6062 Read from file @file{speech.txt}, and synthesize the text using the
6063 standard flite voice:
6065 flite=textfile=speech.txt
6069 Read the specified text selecting the @code{slt} voice:
6071 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6075 Input text to ffmpeg:
6077 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6081 Make @file{ffplay} speak the specified text, using @code{flite} and
6082 the @code{lavfi} device:
6084 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
6088 For more information about libflite, check:
6089 @url{http://www.festvox.org/flite/}
6093 Generate a noise audio signal.
6095 The filter accepts the following options:
6098 @item sample_rate, r
6099 Specify the sample rate. Default value is 48000 Hz.
6102 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
6106 Specify the duration of the generated audio stream. Not specifying this option
6107 results in noise with an infinite length.
6109 @item color, colour, c
6110 Specify the color of noise. Available noise colors are white, pink, brown,
6111 blue, violet and velvet. Default color is white.
6114 Specify a value used to seed the PRNG.
6117 Set the number of samples per each output frame, default is 1024.
6120 @subsection Examples
6125 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
6127 anoisesrc=d=60:c=pink:r=44100:a=0.5
6133 Generate odd-tap Hilbert transform FIR coefficients.
6135 The resulting stream can be used with @ref{afir} filter for phase-shifting
6136 the signal by 90 degrees.
6138 This is used in many matrix coding schemes and for analytic signal generation.
6139 The process is often written as a multiplication by i (or j), the imaginary unit.
6141 The filter accepts the following options:
6145 @item sample_rate, s
6146 Set sample rate, default is 44100.
6149 Set length of FIR filter, default is 22051.
6152 Set number of samples per each frame.
6155 Set window function to be used when generating FIR coefficients.
6160 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
6162 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6164 The filter accepts the following options:
6167 @item sample_rate, r
6168 Set sample rate, default is 44100.
6171 Set number of samples per each frame. Default is 1024.
6174 Set high-pass frequency. Default is 0.
6177 Set low-pass frequency. Default is 0.
6178 If high-pass frequency is lower than low-pass frequency and low-pass frequency
6179 is higher than 0 then filter will create band-pass filter coefficients,
6180 otherwise band-reject filter coefficients.
6183 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
6186 Set Kaiser window beta.
6189 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
6192 Enable rounding, by default is disabled.
6195 Set number of taps for high-pass filter.
6198 Set number of taps for low-pass filter.
6203 Generate an audio signal made of a sine wave with amplitude 1/8.
6205 The audio signal is bit-exact.
6207 The filter accepts the following options:
6212 Set the carrier frequency. Default is 440 Hz.
6214 @item beep_factor, b
6215 Enable a periodic beep every second with frequency @var{beep_factor} times
6216 the carrier frequency. Default is 0, meaning the beep is disabled.
6218 @item sample_rate, r
6219 Specify the sample rate, default is 44100.
6222 Specify the duration of the generated audio stream.
6224 @item samples_per_frame
6225 Set the number of samples per output frame.
6227 The expression can contain the following constants:
6231 The (sequential) number of the output audio frame, starting from 0.
6234 The PTS (Presentation TimeStamp) of the output audio frame,
6235 expressed in @var{TB} units.
6238 The PTS of the output audio frame, expressed in seconds.
6241 The timebase of the output audio frames.
6244 Default is @code{1024}.
6247 @subsection Examples
6252 Generate a simple 440 Hz sine wave:
6258 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
6262 sine=frequency=220:beep_factor=4:duration=5
6266 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
6269 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
6273 @c man end AUDIO SOURCES
6275 @chapter Audio Sinks
6276 @c man begin AUDIO SINKS
6278 Below is a description of the currently available audio sinks.
6280 @section abuffersink
6282 Buffer audio frames, and make them available to the end of filter chain.
6284 This sink is mainly intended for programmatic use, in particular
6285 through the interface defined in @file{libavfilter/buffersink.h}
6286 or the options system.
6288 It accepts a pointer to an AVABufferSinkContext structure, which
6289 defines the incoming buffers' formats, to be passed as the opaque
6290 parameter to @code{avfilter_init_filter} for initialization.
6293 Null audio sink; do absolutely nothing with the input audio. It is
6294 mainly useful as a template and for use in analysis / debugging
6297 @c man end AUDIO SINKS
6299 @chapter Video Filters
6300 @c man begin VIDEO FILTERS
6302 When you configure your FFmpeg build, you can disable any of the
6303 existing filters using @code{--disable-filters}.
6304 The configure output will show the video filters included in your
6307 Below is a description of the currently available video filters.
6311 Mark a region of interest in a video frame.
6313 The frame data is passed through unchanged, but metadata is attached
6314 to the frame indicating regions of interest which can affect the
6315 behaviour of later encoding. Multiple regions can be marked by
6316 applying the filter multiple times.
6320 Region distance in pixels from the left edge of the frame.
6322 Region distance in pixels from the top edge of the frame.
6324 Region width in pixels.
6326 Region height in pixels.
6328 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
6329 and may contain the following variables:
6332 Width of the input frame.
6334 Height of the input frame.
6338 Quantisation offset to apply within the region.
6340 This must be a real value in the range -1 to +1. A value of zero
6341 indicates no quality change. A negative value asks for better quality
6342 (less quantisation), while a positive value asks for worse quality
6343 (greater quantisation).
6345 The range is calibrated so that the extreme values indicate the
6346 largest possible offset - if the rest of the frame is encoded with the
6347 worst possible quality, an offset of -1 indicates that this region
6348 should be encoded with the best possible quality anyway. Intermediate
6349 values are then interpolated in some codec-dependent way.
6351 For example, in 10-bit H.264 the quantisation parameter varies between
6352 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
6353 this region should be encoded with a QP around one-tenth of the full
6354 range better than the rest of the frame. So, if most of the frame
6355 were to be encoded with a QP of around 30, this region would get a QP
6356 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
6357 An extreme value of -1 would indicate that this region should be
6358 encoded with the best possible quality regardless of the treatment of
6359 the rest of the frame - that is, should be encoded at a QP of -12.
6361 If set to true, remove any existing regions of interest marked on the
6362 frame before adding the new one.
6365 @subsection Examples
6369 Mark the centre quarter of the frame as interesting.
6371 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
6374 Mark the 100-pixel-wide region on the left edge of the frame as very
6375 uninteresting (to be encoded at much lower quality than the rest of
6378 addroi=0:0:100:ih:+1/5
6382 @section alphaextract
6384 Extract the alpha component from the input as a grayscale video. This
6385 is especially useful with the @var{alphamerge} filter.
6389 Add or replace the alpha component of the primary input with the
6390 grayscale value of a second input. This is intended for use with
6391 @var{alphaextract} to allow the transmission or storage of frame
6392 sequences that have alpha in a format that doesn't support an alpha
6395 For example, to reconstruct full frames from a normal YUV-encoded video
6396 and a separate video created with @var{alphaextract}, you might use:
6398 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
6401 Since this filter is designed for reconstruction, it operates on frame
6402 sequences without considering timestamps, and terminates when either
6403 input reaches end of stream. This will cause problems if your encoding
6404 pipeline drops frames. If you're trying to apply an image as an
6405 overlay to a video stream, consider the @var{overlay} filter instead.
6409 Amplify differences between current pixel and pixels of adjacent frames in
6410 same pixel location.
6412 This filter accepts the following options:
6416 Set frame radius. Default is 2. Allowed range is from 1 to 63.
6417 For example radius of 3 will instruct filter to calculate average of 7 frames.
6420 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
6423 Set threshold for difference amplification. Any difference greater or equal to
6424 this value will not alter source pixel. Default is 10.
6425 Allowed range is from 0 to 65535.
6428 Set tolerance for difference amplification. Any difference lower to
6429 this value will not alter source pixel. Default is 0.
6430 Allowed range is from 0 to 65535.
6433 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
6434 This option controls maximum possible value that will decrease source pixel value.
6437 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
6438 This option controls maximum possible value that will increase source pixel value.
6441 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
6444 @subsection Commands
6446 This filter supports the following @ref{commands} that corresponds to option of same name:
6458 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
6459 and libavformat to work. On the other hand, it is limited to ASS (Advanced
6460 Substation Alpha) subtitles files.
6462 This filter accepts the following option in addition to the common options from
6463 the @ref{subtitles} filter:
6467 Set the shaping engine
6469 Available values are:
6472 The default libass shaping engine, which is the best available.
6474 Fast, font-agnostic shaper that can do only substitutions
6476 Slower shaper using OpenType for substitutions and positioning
6479 The default is @code{auto}.
6483 Apply an Adaptive Temporal Averaging Denoiser to the video input.
6485 The filter accepts the following options:
6489 Set threshold A for 1st plane. Default is 0.02.
6490 Valid range is 0 to 0.3.
6493 Set threshold B for 1st plane. Default is 0.04.
6494 Valid range is 0 to 5.
6497 Set threshold A for 2nd plane. Default is 0.02.
6498 Valid range is 0 to 0.3.
6501 Set threshold B for 2nd plane. Default is 0.04.
6502 Valid range is 0 to 5.
6505 Set threshold A for 3rd plane. Default is 0.02.
6506 Valid range is 0 to 0.3.
6509 Set threshold B for 3rd plane. Default is 0.04.
6510 Valid range is 0 to 5.
6512 Threshold A is designed to react on abrupt changes in the input signal and
6513 threshold B is designed to react on continuous changes in the input signal.
6516 Set number of frames filter will use for averaging. Default is 9. Must be odd
6517 number in range [5, 129].
6520 Set what planes of frame filter will use for averaging. Default is all.
6523 Set what variant of algorithm filter will use for averaging. Default is @code{p} parallel.
6524 Alternatively can be set to @code{s} serial.
6526 Parallel can be faster then serial, while other way around is never true.
6527 Parallel will abort early on first change being greater then thresholds, while serial
6528 will continue processing other side of frames if they are equal or bellow thresholds.
6531 @subsection Commands
6532 This filter supports same @ref{commands} as options except option @code{s}.
6533 The command accepts the same syntax of the corresponding option.
6537 Apply average blur filter.
6539 The filter accepts the following options:
6543 Set horizontal radius size.
6546 Set which planes to filter. By default all planes are filtered.
6549 Set vertical radius size, if zero it will be same as @code{sizeX}.
6550 Default is @code{0}.
6553 @subsection Commands
6554 This filter supports same commands as options.
6555 The command accepts the same syntax of the corresponding option.
6557 If the specified expression is not valid, it is kept at its current
6562 Compute the bounding box for the non-black pixels in the input frame
6565 This filter computes the bounding box containing all the pixels with a
6566 luminance value greater than the minimum allowed value.
6567 The parameters describing the bounding box are printed on the filter
6570 The filter accepts the following option:
6574 Set the minimal luminance value. Default is @code{16}.
6578 Apply bilateral filter, spatial smoothing while preserving edges.
6580 The filter accepts the following options:
6583 Set sigma of gaussian function to calculate spatial weight.
6584 Allowed range is 0 to 512. Default is 0.1.
6587 Set sigma of gaussian function to calculate range weight.
6588 Allowed range is 0 to 1. Default is 0.1.
6591 Set planes to filter. Default is first only.
6594 @section bitplanenoise
6596 Show and measure bit plane noise.
6598 The filter accepts the following options:
6602 Set which plane to analyze. Default is @code{1}.
6605 Filter out noisy pixels from @code{bitplane} set above.
6606 Default is disabled.
6609 @section blackdetect
6611 Detect video intervals that are (almost) completely black. Can be
6612 useful to detect chapter transitions, commercials, or invalid
6615 The filter outputs its detection analysis to both the log as well as
6616 frame metadata. If a black segment of at least the specified minimum
6617 duration is found, a line with the start and end timestamps as well
6618 as duration is printed to the log with level @code{info}. In addition,
6619 a log line with level @code{debug} is printed per frame showing the
6620 black amount detected for that frame.
6622 The filter also attaches metadata to the first frame of a black
6623 segment with key @code{lavfi.black_start} and to the first frame
6624 after the black segment ends with key @code{lavfi.black_end}. The
6625 value is the frame's timestamp. This metadata is added regardless
6626 of the minimum duration specified.
6628 The filter accepts the following options:
6631 @item black_min_duration, d
6632 Set the minimum detected black duration expressed in seconds. It must
6633 be a non-negative floating point number.
6635 Default value is 2.0.
6637 @item picture_black_ratio_th, pic_th
6638 Set the threshold for considering a picture "black".
6639 Express the minimum value for the ratio:
6641 @var{nb_black_pixels} / @var{nb_pixels}
6644 for which a picture is considered black.
6645 Default value is 0.98.
6647 @item pixel_black_th, pix_th
6648 Set the threshold for considering a pixel "black".
6650 The threshold expresses the maximum pixel luminance value for which a
6651 pixel is considered "black". The provided value is scaled according to
6652 the following equation:
6654 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
6657 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
6658 the input video format, the range is [0-255] for YUV full-range
6659 formats and [16-235] for YUV non full-range formats.
6661 Default value is 0.10.
6664 The following example sets the maximum pixel threshold to the minimum
6665 value, and detects only black intervals of 2 or more seconds:
6667 blackdetect=d=2:pix_th=0.00
6672 Detect frames that are (almost) completely black. Can be useful to
6673 detect chapter transitions or commercials. Output lines consist of
6674 the frame number of the detected frame, the percentage of blackness,
6675 the position in the file if known or -1 and the timestamp in seconds.
6677 In order to display the output lines, you need to set the loglevel at
6678 least to the AV_LOG_INFO value.
6680 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
6681 The value represents the percentage of pixels in the picture that
6682 are below the threshold value.
6684 It accepts the following parameters:
6689 The percentage of the pixels that have to be below the threshold; it defaults to
6692 @item threshold, thresh
6693 The threshold below which a pixel value is considered black; it defaults to
6701 Blend two video frames into each other.
6703 The @code{blend} filter takes two input streams and outputs one
6704 stream, the first input is the "top" layer and second input is
6705 "bottom" layer. By default, the output terminates when the longest input terminates.
6707 The @code{tblend} (time blend) filter takes two consecutive frames
6708 from one single stream, and outputs the result obtained by blending
6709 the new frame on top of the old frame.
6711 A description of the accepted options follows.
6719 Set blend mode for specific pixel component or all pixel components in case
6720 of @var{all_mode}. Default value is @code{normal}.
6722 Available values for component modes are:
6764 Set blend opacity for specific pixel component or all pixel components in case
6765 of @var{all_opacity}. Only used in combination with pixel component blend modes.
6772 Set blend expression for specific pixel component or all pixel components in case
6773 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6775 The expressions can use the following variables:
6779 The sequential number of the filtered frame, starting from @code{0}.
6783 the coordinates of the current sample
6787 the width and height of currently filtered plane
6791 Width and height scale for the plane being filtered. It is the
6792 ratio between the dimensions of the current plane to the luma plane,
6793 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6794 the luma plane and @code{0.5,0.5} for the chroma planes.
6797 Time of the current frame, expressed in seconds.
6800 Value of pixel component at current location for first video frame (top layer).
6803 Value of pixel component at current location for second video frame (bottom layer).
6807 The @code{blend} filter also supports the @ref{framesync} options.
6809 @subsection Examples
6813 Apply transition from bottom layer to top layer in first 10 seconds:
6815 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6819 Apply linear horizontal transition from top layer to bottom layer:
6821 blend=all_expr='A*(X/W)+B*(1-X/W)'
6825 Apply 1x1 checkerboard effect:
6827 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6831 Apply uncover left effect:
6833 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6837 Apply uncover down effect:
6839 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6843 Apply uncover up-left effect:
6845 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6849 Split diagonally video and shows top and bottom layer on each side:
6851 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6855 Display differences between the current and the previous frame:
6857 tblend=all_mode=grainextract
6863 Denoise frames using Block-Matching 3D algorithm.
6865 The filter accepts the following options.
6869 Set denoising strength. Default value is 1.
6870 Allowed range is from 0 to 999.9.
6871 The denoising algorithm is very sensitive to sigma, so adjust it
6872 according to the source.
6875 Set local patch size. This sets dimensions in 2D.
6878 Set sliding step for processing blocks. Default value is 4.
6879 Allowed range is from 1 to 64.
6880 Smaller values allows processing more reference blocks and is slower.
6883 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6884 When set to 1, no block matching is done. Larger values allows more blocks
6886 Allowed range is from 1 to 256.
6889 Set radius for search block matching. Default is 9.
6890 Allowed range is from 1 to INT32_MAX.
6893 Set step between two search locations for block matching. Default is 1.
6894 Allowed range is from 1 to 64. Smaller is slower.
6897 Set threshold of mean square error for block matching. Valid range is 0 to
6901 Set thresholding parameter for hard thresholding in 3D transformed domain.
6902 Larger values results in stronger hard-thresholding filtering in frequency
6906 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6907 Default is @code{basic}.
6910 If enabled, filter will use 2nd stream for block matching.
6911 Default is disabled for @code{basic} value of @var{estim} option,
6912 and always enabled if value of @var{estim} is @code{final}.
6915 Set planes to filter. Default is all available except alpha.
6918 @subsection Examples
6922 Basic filtering with bm3d:
6924 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6928 Same as above, but filtering only luma:
6930 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6934 Same as above, but with both estimation modes:
6936 split[a][b],[a]bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
6940 Same as above, but prefilter with @ref{nlmeans} filter instead:
6942 split[a][b],[a]nlmeans=s=3:r=7:p=3[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
6948 Apply a boxblur algorithm to the input video.
6950 It accepts the following parameters:
6954 @item luma_radius, lr
6955 @item luma_power, lp
6956 @item chroma_radius, cr
6957 @item chroma_power, cp
6958 @item alpha_radius, ar
6959 @item alpha_power, ap
6963 A description of the accepted options follows.
6966 @item luma_radius, lr
6967 @item chroma_radius, cr
6968 @item alpha_radius, ar
6969 Set an expression for the box radius in pixels used for blurring the
6970 corresponding input plane.
6972 The radius value must be a non-negative number, and must not be
6973 greater than the value of the expression @code{min(w,h)/2} for the
6974 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6977 Default value for @option{luma_radius} is "2". If not specified,
6978 @option{chroma_radius} and @option{alpha_radius} default to the
6979 corresponding value set for @option{luma_radius}.
6981 The expressions can contain the following constants:
6985 The input width and height in pixels.
6989 The input chroma image width and height in pixels.
6993 The horizontal and vertical chroma subsample values. For example, for the
6994 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6997 @item luma_power, lp
6998 @item chroma_power, cp
6999 @item alpha_power, ap
7000 Specify how many times the boxblur filter is applied to the
7001 corresponding plane.
7003 Default value for @option{luma_power} is 2. If not specified,
7004 @option{chroma_power} and @option{alpha_power} default to the
7005 corresponding value set for @option{luma_power}.
7007 A value of 0 will disable the effect.
7010 @subsection Examples
7014 Apply a boxblur filter with the luma, chroma, and alpha radii
7017 boxblur=luma_radius=2:luma_power=1
7022 Set the luma radius to 2, and alpha and chroma radius to 0:
7024 boxblur=2:1:cr=0:ar=0
7028 Set the luma and chroma radii to a fraction of the video dimension:
7030 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
7036 Deinterlace the input video ("bwdif" stands for "Bob Weaver
7037 Deinterlacing Filter").
7039 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
7040 interpolation algorithms.
7041 It accepts the following parameters:
7045 The interlacing mode to adopt. It accepts one of the following values:
7049 Output one frame for each frame.
7051 Output one frame for each field.
7054 The default value is @code{send_field}.
7057 The picture field parity assumed for the input interlaced video. It accepts one
7058 of the following values:
7062 Assume the top field is first.
7064 Assume the bottom field is first.
7066 Enable automatic detection of field parity.
7069 The default value is @code{auto}.
7070 If the interlacing is unknown or the decoder does not export this information,
7071 top field first will be assumed.
7074 Specify which frames to deinterlace. Accepts one of the following
7079 Deinterlace all frames.
7081 Only deinterlace frames marked as interlaced.
7084 The default value is @code{all}.
7089 Apply Contrast Adaptive Sharpen filter to video stream.
7091 The filter accepts the following options:
7095 Set the sharpening strength. Default value is 0.
7098 Set planes to filter. Default value is to filter all
7099 planes except alpha plane.
7103 Remove all color information for all colors except for certain one.
7105 The filter accepts the following options:
7109 The color which will not be replaced with neutral chroma.
7112 Similarity percentage with the above color.
7113 0.01 matches only the exact key color, while 1.0 matches everything.
7117 0.0 makes pixels either fully gray, or not gray at all.
7118 Higher values result in more preserved color.
7121 Signals that the color passed is already in YUV instead of RGB.
7123 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7124 This can be used to pass exact YUV values as hexadecimal numbers.
7127 @subsection Commands
7128 This filter supports same @ref{commands} as options.
7129 The command accepts the same syntax of the corresponding option.
7131 If the specified expression is not valid, it is kept at its current
7135 YUV colorspace color/chroma keying.
7137 The filter accepts the following options:
7141 The color which will be replaced with transparency.
7144 Similarity percentage with the key color.
7146 0.01 matches only the exact key color, while 1.0 matches everything.
7151 0.0 makes pixels either fully transparent, or not transparent at all.
7153 Higher values result in semi-transparent pixels, with a higher transparency
7154 the more similar the pixels color is to the key color.
7157 Signals that the color passed is already in YUV instead of RGB.
7159 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7160 This can be used to pass exact YUV values as hexadecimal numbers.
7163 @subsection Commands
7164 This filter supports same @ref{commands} as options.
7165 The command accepts the same syntax of the corresponding option.
7167 If the specified expression is not valid, it is kept at its current
7170 @subsection Examples
7174 Make every green pixel in the input image transparent:
7176 ffmpeg -i input.png -vf chromakey=green out.png
7180 Overlay a greenscreen-video on top of a static black background.
7182 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
7187 Reduce chrominance noise.
7189 The filter accepts the following options:
7193 Set threshold for averaging chrominance values.
7194 Sum of absolute difference of U and V pixel components or current
7195 pixel and neighbour pixels lower than this threshold will be used in
7196 averaging. Luma component is left unchanged and is copied to output.
7197 Default value is 30. Allowed range is from 1 to 200.
7200 Set horizontal radius of rectangle used for averaging.
7201 Allowed range is from 1 to 100. Default value is 5.
7204 Set vertical radius of rectangle used for averaging.
7205 Allowed range is from 1 to 100. Default value is 5.
7208 Set horizontal step when averaging. Default value is 1.
7209 Allowed range is from 1 to 50.
7210 Mostly useful to speed-up filtering.
7213 Set vertical step when averaging. Default value is 1.
7214 Allowed range is from 1 to 50.
7215 Mostly useful to speed-up filtering.
7218 @subsection Commands
7219 This filter supports same @ref{commands} as options.
7220 The command accepts the same syntax of the corresponding option.
7222 @section chromashift
7223 Shift chroma pixels horizontally and/or vertically.
7225 The filter accepts the following options:
7228 Set amount to shift chroma-blue horizontally.
7230 Set amount to shift chroma-blue vertically.
7232 Set amount to shift chroma-red horizontally.
7234 Set amount to shift chroma-red vertically.
7236 Set edge mode, can be @var{smear}, default, or @var{warp}.
7239 @subsection Commands
7241 This filter supports the all above options as @ref{commands}.
7245 Display CIE color diagram with pixels overlaid onto it.
7247 The filter accepts the following options:
7262 @item uhdtv, rec2020
7276 Set what gamuts to draw.
7278 See @code{system} option for available values.
7281 Set ciescope size, by default set to 512.
7284 Set intensity used to map input pixel values to CIE diagram.
7287 Set contrast used to draw tongue colors that are out of active color system gamut.
7290 Correct gamma displayed on scope, by default enabled.
7293 Show white point on CIE diagram, by default disabled.
7296 Set input gamma. Used only with XYZ input color space.
7301 Visualize information exported by some codecs.
7303 Some codecs can export information through frames using side-data or other
7304 means. For example, some MPEG based codecs export motion vectors through the
7305 @var{export_mvs} flag in the codec @option{flags2} option.
7307 The filter accepts the following option:
7311 Set motion vectors to visualize.
7313 Available flags for @var{mv} are:
7317 forward predicted MVs of P-frames
7319 forward predicted MVs of B-frames
7321 backward predicted MVs of B-frames
7325 Display quantization parameters using the chroma planes.
7328 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
7330 Available flags for @var{mv_type} are:
7334 forward predicted MVs
7336 backward predicted MVs
7339 @item frame_type, ft
7340 Set frame type to visualize motion vectors of.
7342 Available flags for @var{frame_type} are:
7346 intra-coded frames (I-frames)
7348 predicted frames (P-frames)
7350 bi-directionally predicted frames (B-frames)
7354 @subsection Examples
7358 Visualize forward predicted MVs of all frames using @command{ffplay}:
7360 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
7364 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
7366 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
7370 @section colorbalance
7371 Modify intensity of primary colors (red, green and blue) of input frames.
7373 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
7374 regions for the red-cyan, green-magenta or blue-yellow balance.
7376 A positive adjustment value shifts the balance towards the primary color, a negative
7377 value towards the complementary color.
7379 The filter accepts the following options:
7385 Adjust red, green and blue shadows (darkest pixels).
7390 Adjust red, green and blue midtones (medium pixels).
7395 Adjust red, green and blue highlights (brightest pixels).
7397 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
7400 Preserve lightness when changing color balance. Default is disabled.
7403 @subsection Examples
7407 Add red color cast to shadows:
7413 @subsection Commands
7415 This filter supports the all above options as @ref{commands}.
7417 @section colorchannelmixer
7419 Adjust video input frames by re-mixing color channels.
7421 This filter modifies a color channel by adding the values associated to
7422 the other channels of the same pixels. For example if the value to
7423 modify is red, the output value will be:
7425 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
7428 The filter accepts the following options:
7435 Adjust contribution of input red, green, blue and alpha channels for output red channel.
7436 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
7442 Adjust contribution of input red, green, blue and alpha channels for output green channel.
7443 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
7449 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
7450 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
7456 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
7457 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
7459 Allowed ranges for options are @code{[-2.0, 2.0]}.
7462 @subsection Examples
7466 Convert source to grayscale:
7468 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
7471 Simulate sepia tones:
7473 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
7477 @subsection Commands
7479 This filter supports the all above options as @ref{commands}.
7482 RGB colorspace color keying.
7484 The filter accepts the following options:
7488 The color which will be replaced with transparency.
7491 Similarity percentage with the key color.
7493 0.01 matches only the exact key color, while 1.0 matches everything.
7498 0.0 makes pixels either fully transparent, or not transparent at all.
7500 Higher values result in semi-transparent pixels, with a higher transparency
7501 the more similar the pixels color is to the key color.
7504 @subsection Examples
7508 Make every green pixel in the input image transparent:
7510 ffmpeg -i input.png -vf colorkey=green out.png
7514 Overlay a greenscreen-video on top of a static background image.
7516 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
7520 @subsection Commands
7521 This filter supports same @ref{commands} as options.
7522 The command accepts the same syntax of the corresponding option.
7524 If the specified expression is not valid, it is kept at its current
7528 Remove all color information for all RGB colors except for certain one.
7530 The filter accepts the following options:
7534 The color which will not be replaced with neutral gray.
7537 Similarity percentage with the above color.
7538 0.01 matches only the exact key color, while 1.0 matches everything.
7541 Blend percentage. 0.0 makes pixels fully gray.
7542 Higher values result in more preserved color.
7545 @subsection Commands
7546 This filter supports same @ref{commands} as options.
7547 The command accepts the same syntax of the corresponding option.
7549 If the specified expression is not valid, it is kept at its current
7552 @section colorlevels
7554 Adjust video input frames using levels.
7556 The filter accepts the following options:
7563 Adjust red, green, blue and alpha input black point.
7564 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
7570 Adjust red, green, blue and alpha input white point.
7571 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
7573 Input levels are used to lighten highlights (bright tones), darken shadows
7574 (dark tones), change the balance of bright and dark tones.
7580 Adjust red, green, blue and alpha output black point.
7581 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
7587 Adjust red, green, blue and alpha output white point.
7588 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
7590 Output levels allows manual selection of a constrained output level range.
7593 @subsection Examples
7597 Make video output darker:
7599 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
7605 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
7609 Make video output lighter:
7611 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
7615 Increase brightness:
7617 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
7621 @subsection Commands
7623 This filter supports the all above options as @ref{commands}.
7625 @section colormatrix
7627 Convert color matrix.
7629 The filter accepts the following options:
7634 Specify the source and destination color matrix. Both values must be
7637 The accepted values are:
7665 For example to convert from BT.601 to SMPTE-240M, use the command:
7667 colormatrix=bt601:smpte240m
7672 Convert colorspace, transfer characteristics or color primaries.
7673 Input video needs to have an even size.
7675 The filter accepts the following options:
7680 Specify all color properties at once.
7682 The accepted values are:
7712 Specify output colorspace.
7714 The accepted values are:
7723 BT.470BG or BT.601-6 625
7726 SMPTE-170M or BT.601-6 525
7735 BT.2020 with non-constant luminance
7741 Specify output transfer characteristics.
7743 The accepted values are:
7755 Constant gamma of 2.2
7758 Constant gamma of 2.8
7761 SMPTE-170M, BT.601-6 625 or BT.601-6 525
7779 BT.2020 for 10-bits content
7782 BT.2020 for 12-bits content
7788 Specify output color primaries.
7790 The accepted values are:
7799 BT.470BG or BT.601-6 625
7802 SMPTE-170M or BT.601-6 525
7826 Specify output color range.
7828 The accepted values are:
7831 TV (restricted) range
7834 MPEG (restricted) range
7845 Specify output color format.
7847 The accepted values are:
7850 YUV 4:2:0 planar 8-bits
7853 YUV 4:2:0 planar 10-bits
7856 YUV 4:2:0 planar 12-bits
7859 YUV 4:2:2 planar 8-bits
7862 YUV 4:2:2 planar 10-bits
7865 YUV 4:2:2 planar 12-bits
7868 YUV 4:4:4 planar 8-bits
7871 YUV 4:4:4 planar 10-bits
7874 YUV 4:4:4 planar 12-bits
7879 Do a fast conversion, which skips gamma/primary correction. This will take
7880 significantly less CPU, but will be mathematically incorrect. To get output
7881 compatible with that produced by the colormatrix filter, use fast=1.
7884 Specify dithering mode.
7886 The accepted values are:
7892 Floyd-Steinberg dithering
7896 Whitepoint adaptation mode.
7898 The accepted values are:
7901 Bradford whitepoint adaptation
7904 von Kries whitepoint adaptation
7907 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7911 Override all input properties at once. Same accepted values as @ref{all}.
7914 Override input colorspace. Same accepted values as @ref{space}.
7917 Override input color primaries. Same accepted values as @ref{primaries}.
7920 Override input transfer characteristics. Same accepted values as @ref{trc}.
7923 Override input color range. Same accepted values as @ref{range}.
7927 The filter converts the transfer characteristics, color space and color
7928 primaries to the specified user values. The output value, if not specified,
7929 is set to a default value based on the "all" property. If that property is
7930 also not specified, the filter will log an error. The output color range and
7931 format default to the same value as the input color range and format. The
7932 input transfer characteristics, color space, color primaries and color range
7933 should be set on the input data. If any of these are missing, the filter will
7934 log an error and no conversion will take place.
7936 For example to convert the input to SMPTE-240M, use the command:
7938 colorspace=smpte240m
7941 @section convolution
7943 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7945 The filter accepts the following options:
7952 Set matrix for each plane.
7953 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7954 and from 1 to 49 odd number of signed integers in @var{row} mode.
7960 Set multiplier for calculated value for each plane.
7961 If unset or 0, it will be sum of all matrix elements.
7967 Set bias for each plane. This value is added to the result of the multiplication.
7968 Useful for making the overall image brighter or darker. Default is 0.0.
7974 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7975 Default is @var{square}.
7978 @subsection Examples
7984 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"
7990 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"
7996 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"
8002 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"
8006 Apply laplacian edge detector which includes diagonals:
8008 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"
8014 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"
8020 Apply 2D convolution of video stream in frequency domain using second stream
8023 The filter accepts the following options:
8027 Set which planes to process.
8030 Set which impulse video frames will be processed, can be @var{first}
8031 or @var{all}. Default is @var{all}.
8034 The @code{convolve} filter also supports the @ref{framesync} options.
8038 Copy the input video source unchanged to the output. This is mainly useful for
8043 Video filtering on GPU using Apple's CoreImage API on OSX.
8045 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8046 processed by video hardware. However, software-based OpenGL implementations
8047 exist which means there is no guarantee for hardware processing. It depends on
8050 There are many filters and image generators provided by Apple that come with a
8051 large variety of options. The filter has to be referenced by its name along
8054 The coreimage filter accepts the following options:
8057 List all available filters and generators along with all their respective
8058 options as well as possible minimum and maximum values along with the default
8065 Specify all filters by their respective name and options.
8066 Use @var{list_filters} to determine all valid filter names and options.
8067 Numerical options are specified by a float value and are automatically clamped
8068 to their respective value range. Vector and color options have to be specified
8069 by a list of space separated float values. Character escaping has to be done.
8070 A special option name @code{default} is available to use default options for a
8073 It is required to specify either @code{default} or at least one of the filter options.
8074 All omitted options are used with their default values.
8075 The syntax of the filter string is as follows:
8077 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8081 Specify a rectangle where the output of the filter chain is copied into the
8082 input image. It is given by a list of space separated float values:
8084 output_rect=x\ y\ width\ height
8086 If not given, the output rectangle equals the dimensions of the input image.
8087 The output rectangle is automatically cropped at the borders of the input
8088 image. Negative values are valid for each component.
8090 output_rect=25\ 25\ 100\ 100
8094 Several filters can be chained for successive processing without GPU-HOST
8095 transfers allowing for fast processing of complex filter chains.
8096 Currently, only filters with zero (generators) or exactly one (filters) input
8097 image and one output image are supported. Also, transition filters are not yet
8100 Some filters generate output images with additional padding depending on the
8101 respective filter kernel. The padding is automatically removed to ensure the
8102 filter output has the same size as the input image.
8104 For image generators, the size of the output image is determined by the
8105 previous output image of the filter chain or the input image of the whole
8106 filterchain, respectively. The generators do not use the pixel information of
8107 this image to generate their output. However, the generated output is
8108 blended onto this image, resulting in partial or complete coverage of the
8111 The @ref{coreimagesrc} video source can be used for generating input images
8112 which are directly fed into the filter chain. By using it, providing input
8113 images by another video source or an input video is not required.
8115 @subsection Examples
8120 List all filters available:
8122 coreimage=list_filters=true
8126 Use the CIBoxBlur filter with default options to blur an image:
8128 coreimage=filter=CIBoxBlur@@default
8132 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8133 its center at 100x100 and a radius of 50 pixels:
8135 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8139 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8140 given as complete and escaped command-line for Apple's standard bash shell:
8142 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8148 Cover a rectangular object
8150 It accepts the following options:
8154 Filepath of the optional cover image, needs to be in yuv420.
8159 It accepts the following values:
8162 cover it by the supplied image
8164 cover it by interpolating the surrounding pixels
8167 Default value is @var{blur}.
8170 @subsection Examples
8174 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8176 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8182 Crop the input video to given dimensions.
8184 It accepts the following parameters:
8188 The width of the output video. It defaults to @code{iw}.
8189 This expression is evaluated only once during the filter
8190 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8193 The height of the output video. It defaults to @code{ih}.
8194 This expression is evaluated only once during the filter
8195 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8198 The horizontal position, in the input video, of the left edge of the output
8199 video. It defaults to @code{(in_w-out_w)/2}.
8200 This expression is evaluated per-frame.
8203 The vertical position, in the input video, of the top edge of the output video.
8204 It defaults to @code{(in_h-out_h)/2}.
8205 This expression is evaluated per-frame.
8208 If set to 1 will force the output display aspect ratio
8209 to be the same of the input, by changing the output sample aspect
8210 ratio. It defaults to 0.
8213 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8214 width/height/x/y as specified and will not be rounded to nearest smaller value.
8218 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8219 expressions containing the following constants:
8224 The computed values for @var{x} and @var{y}. They are evaluated for
8229 The input width and height.
8233 These are the same as @var{in_w} and @var{in_h}.
8237 The output (cropped) width and height.
8241 These are the same as @var{out_w} and @var{out_h}.
8244 same as @var{iw} / @var{ih}
8247 input sample aspect ratio
8250 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8254 horizontal and vertical chroma subsample values. For example for the
8255 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8258 The number of the input frame, starting from 0.
8261 the position in the file of the input frame, NAN if unknown
8264 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
8268 The expression for @var{out_w} may depend on the value of @var{out_h},
8269 and the expression for @var{out_h} may depend on @var{out_w}, but they
8270 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
8271 evaluated after @var{out_w} and @var{out_h}.
8273 The @var{x} and @var{y} parameters specify the expressions for the
8274 position of the top-left corner of the output (non-cropped) area. They
8275 are evaluated for each frame. If the evaluated value is not valid, it
8276 is approximated to the nearest valid value.
8278 The expression for @var{x} may depend on @var{y}, and the expression
8279 for @var{y} may depend on @var{x}.
8281 @subsection Examples
8285 Crop area with size 100x100 at position (12,34).
8290 Using named options, the example above becomes:
8292 crop=w=100:h=100:x=12:y=34
8296 Crop the central input area with size 100x100:
8302 Crop the central input area with size 2/3 of the input video:
8304 crop=2/3*in_w:2/3*in_h
8308 Crop the input video central square:
8315 Delimit the rectangle with the top-left corner placed at position
8316 100:100 and the right-bottom corner corresponding to the right-bottom
8317 corner of the input image.
8319 crop=in_w-100:in_h-100:100:100
8323 Crop 10 pixels from the left and right borders, and 20 pixels from
8324 the top and bottom borders
8326 crop=in_w-2*10:in_h-2*20
8330 Keep only the bottom right quarter of the input image:
8332 crop=in_w/2:in_h/2:in_w/2:in_h/2
8336 Crop height for getting Greek harmony:
8338 crop=in_w:1/PHI*in_w
8342 Apply trembling effect:
8344 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)
8348 Apply erratic camera effect depending on timestamp:
8350 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)"
8354 Set x depending on the value of y:
8356 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
8360 @subsection Commands
8362 This filter supports the following commands:
8368 Set width/height of the output video and the horizontal/vertical position
8370 The command accepts the same syntax of the corresponding option.
8372 If the specified expression is not valid, it is kept at its current
8378 Auto-detect the crop size.
8380 It calculates the necessary cropping parameters and prints the
8381 recommended parameters via the logging system. The detected dimensions
8382 correspond to the non-black area of the input video.
8384 It accepts the following parameters:
8389 Set higher black value threshold, which can be optionally specified
8390 from nothing (0) to everything (255 for 8-bit based formats). An intensity
8391 value greater to the set value is considered non-black. It defaults to 24.
8392 You can also specify a value between 0.0 and 1.0 which will be scaled depending
8393 on the bitdepth of the pixel format.
8396 The value which the width/height should be divisible by. It defaults to
8397 16. The offset is automatically adjusted to center the video. Use 2 to
8398 get only even dimensions (needed for 4:2:2 video). 16 is best when
8399 encoding to most video codecs.
8401 @item reset_count, reset
8402 Set the counter that determines after how many frames cropdetect will
8403 reset the previously detected largest video area and start over to
8404 detect the current optimal crop area. Default value is 0.
8406 This can be useful when channel logos distort the video area. 0
8407 indicates 'never reset', and returns the largest area encountered during
8414 Delay video filtering until a given wallclock timestamp. The filter first
8415 passes on @option{preroll} amount of frames, then it buffers at most
8416 @option{buffer} amount of frames and waits for the cue. After reaching the cue
8417 it forwards the buffered frames and also any subsequent frames coming in its
8420 The filter can be used synchronize the output of multiple ffmpeg processes for
8421 realtime output devices like decklink. By putting the delay in the filtering
8422 chain and pre-buffering frames the process can pass on data to output almost
8423 immediately after the target wallclock timestamp is reached.
8425 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
8431 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
8434 The duration of content to pass on as preroll expressed in seconds. Default is 0.
8437 The maximum duration of content to buffer before waiting for the cue expressed
8438 in seconds. Default is 0.
8445 Apply color adjustments using curves.
8447 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
8448 component (red, green and blue) has its values defined by @var{N} key points
8449 tied from each other using a smooth curve. The x-axis represents the pixel
8450 values from the input frame, and the y-axis the new pixel values to be set for
8453 By default, a component curve is defined by the two points @var{(0;0)} and
8454 @var{(1;1)}. This creates a straight line where each original pixel value is
8455 "adjusted" to its own value, which means no change to the image.
8457 The filter allows you to redefine these two points and add some more. A new
8458 curve (using a natural cubic spline interpolation) will be define to pass
8459 smoothly through all these new coordinates. The new defined points needs to be
8460 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
8461 be in the @var{[0;1]} interval. If the computed curves happened to go outside
8462 the vector spaces, the values will be clipped accordingly.
8464 The filter accepts the following options:
8468 Select one of the available color presets. This option can be used in addition
8469 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
8470 options takes priority on the preset values.
8471 Available presets are:
8474 @item color_negative
8477 @item increase_contrast
8479 @item linear_contrast
8480 @item medium_contrast
8482 @item strong_contrast
8485 Default is @code{none}.
8487 Set the master key points. These points will define a second pass mapping. It
8488 is sometimes called a "luminance" or "value" mapping. It can be used with
8489 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
8490 post-processing LUT.
8492 Set the key points for the red component.
8494 Set the key points for the green component.
8496 Set the key points for the blue component.
8498 Set the key points for all components (not including master).
8499 Can be used in addition to the other key points component
8500 options. In this case, the unset component(s) will fallback on this
8501 @option{all} setting.
8503 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
8505 Save Gnuplot script of the curves in specified file.
8508 To avoid some filtergraph syntax conflicts, each key points list need to be
8509 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
8511 @subsection Examples
8515 Increase slightly the middle level of blue:
8517 curves=blue='0/0 0.5/0.58 1/1'
8523 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'
8525 Here we obtain the following coordinates for each components:
8528 @code{(0;0.11) (0.42;0.51) (1;0.95)}
8530 @code{(0;0) (0.50;0.48) (1;1)}
8532 @code{(0;0.22) (0.49;0.44) (1;0.80)}
8536 The previous example can also be achieved with the associated built-in preset:
8538 curves=preset=vintage
8548 Use a Photoshop preset and redefine the points of the green component:
8550 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
8554 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
8555 and @command{gnuplot}:
8557 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
8558 gnuplot -p /tmp/curves.plt
8564 Video data analysis filter.
8566 This filter shows hexadecimal pixel values of part of video.
8568 The filter accepts the following options:
8572 Set output video size.
8575 Set x offset from where to pick pixels.
8578 Set y offset from where to pick pixels.
8581 Set scope mode, can be one of the following:
8584 Draw hexadecimal pixel values with white color on black background.
8587 Draw hexadecimal pixel values with input video pixel color on black
8591 Draw hexadecimal pixel values on color background picked from input video,
8592 the text color is picked in such way so its always visible.
8596 Draw rows and columns numbers on left and top of video.
8599 Set background opacity.
8602 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
8606 Apply Directional blur filter.
8608 The filter accepts the following options:
8612 Set angle of directional blur. Default is @code{45}.
8615 Set radius of directional blur. Default is @code{5}.
8618 Set which planes to filter. By default all planes are filtered.
8621 @subsection Commands
8622 This filter supports same @ref{commands} as options.
8623 The command accepts the same syntax of the corresponding option.
8625 If the specified expression is not valid, it is kept at its current
8630 Denoise frames using 2D DCT (frequency domain filtering).
8632 This filter is not designed for real time.
8634 The filter accepts the following options:
8638 Set the noise sigma constant.
8640 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
8641 coefficient (absolute value) below this threshold with be dropped.
8643 If you need a more advanced filtering, see @option{expr}.
8645 Default is @code{0}.
8648 Set number overlapping pixels for each block. Since the filter can be slow, you
8649 may want to reduce this value, at the cost of a less effective filter and the
8650 risk of various artefacts.
8652 If the overlapping value doesn't permit processing the whole input width or
8653 height, a warning will be displayed and according borders won't be denoised.
8655 Default value is @var{blocksize}-1, which is the best possible setting.
8658 Set the coefficient factor expression.
8660 For each coefficient of a DCT block, this expression will be evaluated as a
8661 multiplier value for the coefficient.
8663 If this is option is set, the @option{sigma} option will be ignored.
8665 The absolute value of the coefficient can be accessed through the @var{c}
8669 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
8670 @var{blocksize}, which is the width and height of the processed blocks.
8672 The default value is @var{3} (8x8) and can be raised to @var{4} for a
8673 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
8674 on the speed processing. Also, a larger block size does not necessarily means a
8678 @subsection Examples
8680 Apply a denoise with a @option{sigma} of @code{4.5}:
8685 The same operation can be achieved using the expression system:
8687 dctdnoiz=e='gte(c, 4.5*3)'
8690 Violent denoise using a block size of @code{16x16}:
8697 Remove banding artifacts from input video.
8698 It works by replacing banded pixels with average value of referenced pixels.
8700 The filter accepts the following options:
8707 Set banding detection threshold for each plane. Default is 0.02.
8708 Valid range is 0.00003 to 0.5.
8709 If difference between current pixel and reference pixel is less than threshold,
8710 it will be considered as banded.
8713 Banding detection range in pixels. Default is 16. If positive, random number
8714 in range 0 to set value will be used. If negative, exact absolute value
8716 The range defines square of four pixels around current pixel.
8719 Set direction in radians from which four pixel will be compared. If positive,
8720 random direction from 0 to set direction will be picked. If negative, exact of
8721 absolute value will be picked. For example direction 0, -PI or -2*PI radians
8722 will pick only pixels on same row and -PI/2 will pick only pixels on same
8726 If enabled, current pixel is compared with average value of all four
8727 surrounding pixels. The default is enabled. If disabled current pixel is
8728 compared with all four surrounding pixels. The pixel is considered banded
8729 if only all four differences with surrounding pixels are less than threshold.
8732 If enabled, current pixel is changed if and only if all pixel components are banded,
8733 e.g. banding detection threshold is triggered for all color components.
8734 The default is disabled.
8739 Remove blocking artifacts from input video.
8741 The filter accepts the following options:
8745 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
8746 This controls what kind of deblocking is applied.
8749 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
8755 Set blocking detection thresholds. Allowed range is 0 to 1.
8756 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
8757 Using higher threshold gives more deblocking strength.
8758 Setting @var{alpha} controls threshold detection at exact edge of block.
8759 Remaining options controls threshold detection near the edge. Each one for
8760 below/above or left/right. Setting any of those to @var{0} disables
8764 Set planes to filter. Default is to filter all available planes.
8767 @subsection Examples
8771 Deblock using weak filter and block size of 4 pixels.
8773 deblock=filter=weak:block=4
8777 Deblock using strong filter, block size of 4 pixels and custom thresholds for
8778 deblocking more edges.
8780 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
8784 Similar as above, but filter only first plane.
8786 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
8790 Similar as above, but filter only second and third plane.
8792 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
8799 Drop duplicated frames at regular intervals.
8801 The filter accepts the following options:
8805 Set the number of frames from which one will be dropped. Setting this to
8806 @var{N} means one frame in every batch of @var{N} frames will be dropped.
8807 Default is @code{5}.
8810 Set the threshold for duplicate detection. If the difference metric for a frame
8811 is less than or equal to this value, then it is declared as duplicate. Default
8815 Set scene change threshold. Default is @code{15}.
8819 Set the size of the x and y-axis blocks used during metric calculations.
8820 Larger blocks give better noise suppression, but also give worse detection of
8821 small movements. Must be a power of two. Default is @code{32}.
8824 Mark main input as a pre-processed input and activate clean source input
8825 stream. This allows the input to be pre-processed with various filters to help
8826 the metrics calculation while keeping the frame selection lossless. When set to
8827 @code{1}, the first stream is for the pre-processed input, and the second
8828 stream is the clean source from where the kept frames are chosen. Default is
8832 Set whether or not chroma is considered in the metric calculations. Default is
8838 Apply 2D deconvolution of video stream in frequency domain using second stream
8841 The filter accepts the following options:
8845 Set which planes to process.
8848 Set which impulse video frames will be processed, can be @var{first}
8849 or @var{all}. Default is @var{all}.
8852 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
8853 and height are not same and not power of 2 or if stream prior to convolving
8857 The @code{deconvolve} filter also supports the @ref{framesync} options.
8861 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
8863 It accepts the following options:
8867 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
8868 @var{rainbows} for cross-color reduction.
8871 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
8874 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
8877 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
8880 Set temporal chroma threshold. Lower values increases reduction of cross-color.
8885 Apply deflate effect to the video.
8887 This filter replaces the pixel by the local(3x3) average by taking into account
8888 only values lower than the pixel.
8890 It accepts the following options:
8897 Limit the maximum change for each plane, default is 65535.
8898 If 0, plane will remain unchanged.
8901 @subsection Commands
8903 This filter supports the all above options as @ref{commands}.
8907 Remove temporal frame luminance variations.
8909 It accepts the following options:
8913 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8916 Set averaging mode to smooth temporal luminance variations.
8918 Available values are:
8943 Do not actually modify frame. Useful when one only wants metadata.
8948 Remove judder produced by partially interlaced telecined content.
8950 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8951 source was partially telecined content then the output of @code{pullup,dejudder}
8952 will have a variable frame rate. May change the recorded frame rate of the
8953 container. Aside from that change, this filter will not affect constant frame
8956 The option available in this filter is:
8960 Specify the length of the window over which the judder repeats.
8962 Accepts any integer greater than 1. Useful values are:
8966 If the original was telecined from 24 to 30 fps (Film to NTSC).
8969 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8972 If a mixture of the two.
8975 The default is @samp{4}.
8980 Suppress a TV station logo by a simple interpolation of the surrounding
8981 pixels. Just set a rectangle covering the logo and watch it disappear
8982 (and sometimes something even uglier appear - your mileage may vary).
8984 It accepts the following parameters:
8989 Specify the top left corner coordinates of the logo. They must be
8994 Specify the width and height of the logo to clear. They must be
8998 Specify the thickness of the fuzzy edge of the rectangle (added to
8999 @var{w} and @var{h}). The default value is 1. This option is
9000 deprecated, setting higher values should no longer be necessary and
9004 When set to 1, a green rectangle is drawn on the screen to simplify
9005 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9006 The default value is 0.
9008 The rectangle is drawn on the outermost pixels which will be (partly)
9009 replaced with interpolated values. The values of the next pixels
9010 immediately outside this rectangle in each direction will be used to
9011 compute the interpolated pixel values inside the rectangle.
9015 @subsection Examples
9019 Set a rectangle covering the area with top left corner coordinates 0,0
9020 and size 100x77, and a band of size 10:
9022 delogo=x=0:y=0:w=100:h=77:band=10
9030 Remove the rain in the input image/video by applying the derain methods based on
9031 convolutional neural networks. Supported models:
9035 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9036 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9039 Training as well as model generation scripts are provided in
9040 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9042 Native model files (.model) can be generated from TensorFlow model
9043 files (.pb) by using tools/python/convert.py
9045 The filter accepts the following options:
9049 Specify which filter to use. This option accepts the following values:
9053 Derain filter. To conduct derain filter, you need to use a derain model.
9056 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9058 Default value is @samp{derain}.
9061 Specify which DNN backend to use for model loading and execution. This option accepts
9062 the following values:
9066 Native implementation of DNN loading and execution.
9069 TensorFlow backend. To enable this backend you
9070 need to install the TensorFlow for C library (see
9071 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9072 @code{--enable-libtensorflow}
9074 Default value is @samp{native}.
9077 Set path to model file specifying network architecture and its parameters.
9078 Note that different backends use different file formats. TensorFlow and native
9079 backend can load files for only its format.
9082 It can also be finished with @ref{dnn_processing} filter.
9086 Attempt to fix small changes in horizontal and/or vertical shift. This
9087 filter helps remove camera shake from hand-holding a camera, bumping a
9088 tripod, moving on a vehicle, etc.
9090 The filter accepts the following options:
9098 Specify a rectangular area where to limit the search for motion
9100 If desired the search for motion vectors can be limited to a
9101 rectangular area of the frame defined by its top left corner, width
9102 and height. These parameters have the same meaning as the drawbox
9103 filter which can be used to visualise the position of the bounding
9106 This is useful when simultaneous movement of subjects within the frame
9107 might be confused for camera motion by the motion vector search.
9109 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9110 then the full frame is used. This allows later options to be set
9111 without specifying the bounding box for the motion vector search.
9113 Default - search the whole frame.
9117 Specify the maximum extent of movement in x and y directions in the
9118 range 0-64 pixels. Default 16.
9121 Specify how to generate pixels to fill blanks at the edge of the
9122 frame. Available values are:
9125 Fill zeroes at blank locations
9127 Original image at blank locations
9129 Extruded edge value at blank locations
9131 Mirrored edge at blank locations
9133 Default value is @samp{mirror}.
9136 Specify the blocksize to use for motion search. Range 4-128 pixels,
9140 Specify the contrast threshold for blocks. Only blocks with more than
9141 the specified contrast (difference between darkest and lightest
9142 pixels) will be considered. Range 1-255, default 125.
9145 Specify the search strategy. Available values are:
9148 Set exhaustive search
9150 Set less exhaustive search.
9152 Default value is @samp{exhaustive}.
9155 If set then a detailed log of the motion search is written to the
9162 Remove unwanted contamination of foreground colors, caused by reflected color of
9163 greenscreen or bluescreen.
9165 This filter accepts the following options:
9169 Set what type of despill to use.
9172 Set how spillmap will be generated.
9175 Set how much to get rid of still remaining spill.
9178 Controls amount of red in spill area.
9181 Controls amount of green in spill area.
9182 Should be -1 for greenscreen.
9185 Controls amount of blue in spill area.
9186 Should be -1 for bluescreen.
9189 Controls brightness of spill area, preserving colors.
9192 Modify alpha from generated spillmap.
9197 Apply an exact inverse of the telecine operation. It requires a predefined
9198 pattern specified using the pattern option which must be the same as that passed
9199 to the telecine filter.
9201 This filter accepts the following options:
9210 The default value is @code{top}.
9214 A string of numbers representing the pulldown pattern you wish to apply.
9215 The default value is @code{23}.
9218 A number representing position of the first frame with respect to the telecine
9219 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9224 Apply dilation effect to the video.
9226 This filter replaces the pixel by the local(3x3) maximum.
9228 It accepts the following options:
9235 Limit the maximum change for each plane, default is 65535.
9236 If 0, plane will remain unchanged.
9239 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9242 Flags to local 3x3 coordinates maps like this:
9249 @subsection Commands
9251 This filter supports the all above options as @ref{commands}.
9255 Displace pixels as indicated by second and third input stream.
9257 It takes three input streams and outputs one stream, the first input is the
9258 source, and second and third input are displacement maps.
9260 The second input specifies how much to displace pixels along the
9261 x-axis, while the third input specifies how much to displace pixels
9263 If one of displacement map streams terminates, last frame from that
9264 displacement map will be used.
9266 Note that once generated, displacements maps can be reused over and over again.
9268 A description of the accepted options follows.
9272 Set displace behavior for pixels that are out of range.
9274 Available values are:
9277 Missing pixels are replaced by black pixels.
9280 Adjacent pixels will spread out to replace missing pixels.
9283 Out of range pixels are wrapped so they point to pixels of other side.
9286 Out of range pixels will be replaced with mirrored pixels.
9288 Default is @samp{smear}.
9292 @subsection Examples
9296 Add ripple effect to rgb input of video size hd720:
9298 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
9302 Add wave effect to rgb input of video size hd720:
9304 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
9308 @anchor{dnn_processing}
9309 @section dnn_processing
9311 Do image processing with deep neural networks. It works together with another filter
9312 which converts the pixel format of the Frame to what the dnn network requires.
9314 The filter accepts the following options:
9318 Specify which DNN backend to use for model loading and execution. This option accepts
9319 the following values:
9323 Native implementation of DNN loading and execution.
9326 TensorFlow backend. To enable this backend you
9327 need to install the TensorFlow for C library (see
9328 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9329 @code{--enable-libtensorflow}
9332 OpenVINO backend. To enable this backend you
9333 need to build and install the OpenVINO for C library (see
9334 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
9335 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
9336 be needed if the header files and libraries are not installed into system path)
9340 Default value is @samp{native}.
9343 Set path to model file specifying network architecture and its parameters.
9344 Note that different backends use different file formats. TensorFlow, OpenVINO and native
9345 backend can load files for only its format.
9347 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
9350 Set the input name of the dnn network.
9353 Set the output name of the dnn network.
9357 @subsection Examples
9361 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
9363 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
9367 Halve the pixel value of the frame with format gray32f:
9369 ffmpeg -i input.jpg -vf format=grayf32,dnn_processing=model=halve_gray_float.model:input=dnn_in:output=dnn_out:dnn_backend=native -y out.native.png
9373 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
9375 ./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
9379 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
9381 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
9388 Draw a colored box on the input image.
9390 It accepts the following parameters:
9395 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
9399 The expressions which specify the width and height of the box; if 0 they are interpreted as
9400 the input width and height. It defaults to 0.
9403 Specify the color of the box to write. For the general syntax of this option,
9404 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
9405 value @code{invert} is used, the box edge color is the same as the
9406 video with inverted luma.
9409 The expression which sets the thickness of the box edge.
9410 A value of @code{fill} will create a filled box. Default value is @code{3}.
9412 See below for the list of accepted constants.
9415 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
9416 will overwrite the video's color and alpha pixels.
9417 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
9420 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
9421 following constants:
9425 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
9429 horizontal and vertical chroma subsample values. For example for the
9430 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9434 The input width and height.
9437 The input sample aspect ratio.
9441 The x and y offset coordinates where the box is drawn.
9445 The width and height of the drawn box.
9448 The thickness of the drawn box.
9450 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
9451 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
9455 @subsection Examples
9459 Draw a black box around the edge of the input image:
9465 Draw a box with color red and an opacity of 50%:
9467 drawbox=10:20:200:60:red@@0.5
9470 The previous example can be specified as:
9472 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
9476 Fill the box with pink color:
9478 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
9482 Draw a 2-pixel red 2.40:1 mask:
9484 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
9488 @subsection Commands
9489 This filter supports same commands as options.
9490 The command accepts the same syntax of the corresponding option.
9492 If the specified expression is not valid, it is kept at its current
9497 Draw a graph using input video metadata.
9499 It accepts the following parameters:
9503 Set 1st frame metadata key from which metadata values will be used to draw a graph.
9506 Set 1st foreground color expression.
9509 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
9512 Set 2nd foreground color expression.
9515 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
9518 Set 3rd foreground color expression.
9521 Set 4th frame metadata key from which metadata values will be used to draw a graph.
9524 Set 4th foreground color expression.
9527 Set minimal value of metadata value.
9530 Set maximal value of metadata value.
9533 Set graph background color. Default is white.
9538 Available values for mode is:
9545 Default is @code{line}.
9550 Available values for slide is:
9553 Draw new frame when right border is reached.
9556 Replace old columns with new ones.
9559 Scroll from right to left.
9562 Scroll from left to right.
9565 Draw single picture.
9568 Default is @code{frame}.
9571 Set size of graph video. For the syntax of this option, check the
9572 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
9573 The default value is @code{900x256}.
9576 Set the output frame rate. Default value is @code{25}.
9578 The foreground color expressions can use the following variables:
9581 Minimal value of metadata value.
9584 Maximal value of metadata value.
9587 Current metadata key value.
9590 The color is defined as 0xAABBGGRR.
9593 Example using metadata from @ref{signalstats} filter:
9595 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
9598 Example using metadata from @ref{ebur128} filter:
9600 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
9605 Draw a grid on the input image.
9607 It accepts the following parameters:
9612 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
9616 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
9617 input width and height, respectively, minus @code{thickness}, so image gets
9618 framed. Default to 0.
9621 Specify the color of the grid. For the general syntax of this option,
9622 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
9623 value @code{invert} is used, the grid color is the same as the
9624 video with inverted luma.
9627 The expression which sets the thickness of the grid line. Default value is @code{1}.
9629 See below for the list of accepted constants.
9632 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
9633 will overwrite the video's color and alpha pixels.
9634 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
9637 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
9638 following constants:
9642 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
9646 horizontal and vertical chroma subsample values. For example for the
9647 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9651 The input grid cell width and height.
9654 The input sample aspect ratio.
9658 The x and y coordinates of some point of grid intersection (meant to configure offset).
9662 The width and height of the drawn cell.
9665 The thickness of the drawn cell.
9667 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
9668 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
9672 @subsection Examples
9676 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
9678 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
9682 Draw a white 3x3 grid with an opacity of 50%:
9684 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
9688 @subsection Commands
9689 This filter supports same commands as options.
9690 The command accepts the same syntax of the corresponding option.
9692 If the specified expression is not valid, it is kept at its current
9698 Draw a text string or text from a specified file on top of a video, using the
9699 libfreetype library.
9701 To enable compilation of this filter, you need to configure FFmpeg with
9702 @code{--enable-libfreetype}.
9703 To enable default font fallback and the @var{font} option you need to
9704 configure FFmpeg with @code{--enable-libfontconfig}.
9705 To enable the @var{text_shaping} option, you need to configure FFmpeg with
9706 @code{--enable-libfribidi}.
9710 It accepts the following parameters:
9715 Used to draw a box around text using the background color.
9716 The value must be either 1 (enable) or 0 (disable).
9717 The default value of @var{box} is 0.
9720 Set the width of the border to be drawn around the box using @var{boxcolor}.
9721 The default value of @var{boxborderw} is 0.
9724 The color to be used for drawing box around text. For the syntax of this
9725 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9727 The default value of @var{boxcolor} is "white".
9730 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
9731 The default value of @var{line_spacing} is 0.
9734 Set the width of the border to be drawn around the text using @var{bordercolor}.
9735 The default value of @var{borderw} is 0.
9738 Set the color to be used for drawing border around text. For the syntax of this
9739 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9741 The default value of @var{bordercolor} is "black".
9744 Select how the @var{text} is expanded. Can be either @code{none},
9745 @code{strftime} (deprecated) or
9746 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
9750 Set a start time for the count. Value is in microseconds. Only applied
9751 in the deprecated strftime expansion mode. To emulate in normal expansion
9752 mode use the @code{pts} function, supplying the start time (in seconds)
9753 as the second argument.
9756 If true, check and fix text coords to avoid clipping.
9759 The color to be used for drawing fonts. For the syntax of this option, check
9760 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9762 The default value of @var{fontcolor} is "black".
9764 @item fontcolor_expr
9765 String which is expanded the same way as @var{text} to obtain dynamic
9766 @var{fontcolor} value. By default this option has empty value and is not
9767 processed. When this option is set, it overrides @var{fontcolor} option.
9770 The font family to be used for drawing text. By default Sans.
9773 The font file to be used for drawing text. The path must be included.
9774 This parameter is mandatory if the fontconfig support is disabled.
9777 Draw the text applying alpha blending. The value can
9778 be a number between 0.0 and 1.0.
9779 The expression accepts the same variables @var{x, y} as well.
9780 The default value is 1.
9781 Please see @var{fontcolor_expr}.
9784 The font size to be used for drawing text.
9785 The default value of @var{fontsize} is 16.
9788 If set to 1, attempt to shape the text (for example, reverse the order of
9789 right-to-left text and join Arabic characters) before drawing it.
9790 Otherwise, just draw the text exactly as given.
9791 By default 1 (if supported).
9794 The flags to be used for loading the fonts.
9796 The flags map the corresponding flags supported by libfreetype, and are
9797 a combination of the following values:
9804 @item vertical_layout
9805 @item force_autohint
9808 @item ignore_global_advance_width
9810 @item ignore_transform
9816 Default value is "default".
9818 For more information consult the documentation for the FT_LOAD_*
9822 The color to be used for drawing a shadow behind the drawn text. For the
9823 syntax of this option, check the @ref{color syntax,,"Color" section in the
9824 ffmpeg-utils manual,ffmpeg-utils}.
9826 The default value of @var{shadowcolor} is "black".
9830 The x and y offsets for the text shadow position with respect to the
9831 position of the text. They can be either positive or negative
9832 values. The default value for both is "0".
9835 The starting frame number for the n/frame_num variable. The default value
9839 The size in number of spaces to use for rendering the tab.
9843 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
9844 format. It can be used with or without text parameter. @var{timecode_rate}
9845 option must be specified.
9847 @item timecode_rate, rate, r
9848 Set the timecode frame rate (timecode only). Value will be rounded to nearest
9849 integer. Minimum value is "1".
9850 Drop-frame timecode is supported for frame rates 30 & 60.
9853 If set to 1, the output of the timecode option will wrap around at 24 hours.
9854 Default is 0 (disabled).
9857 The text string to be drawn. The text must be a sequence of UTF-8
9859 This parameter is mandatory if no file is specified with the parameter
9863 A text file containing text to be drawn. The text must be a sequence
9864 of UTF-8 encoded characters.
9866 This parameter is mandatory if no text string is specified with the
9867 parameter @var{text}.
9869 If both @var{text} and @var{textfile} are specified, an error is thrown.
9872 If set to 1, the @var{textfile} will be reloaded before each frame.
9873 Be sure to update it atomically, or it may be read partially, or even fail.
9877 The expressions which specify the offsets where text will be drawn
9878 within the video frame. They are relative to the top/left border of the
9881 The default value of @var{x} and @var{y} is "0".
9883 See below for the list of accepted constants and functions.
9886 The parameters for @var{x} and @var{y} are expressions containing the
9887 following constants and functions:
9891 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
9895 horizontal and vertical chroma subsample values. For example for the
9896 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9899 the height of each text line
9907 @item max_glyph_a, ascent
9908 the maximum distance from the baseline to the highest/upper grid
9909 coordinate used to place a glyph outline point, for all the rendered
9911 It is a positive value, due to the grid's orientation with the Y axis
9914 @item max_glyph_d, descent
9915 the maximum distance from the baseline to the lowest grid coordinate
9916 used to place a glyph outline point, for all the rendered glyphs.
9917 This is a negative value, due to the grid's orientation, with the Y axis
9921 maximum glyph height, that is the maximum height for all the glyphs
9922 contained in the rendered text, it is equivalent to @var{ascent} -
9926 maximum glyph width, that is the maximum width for all the glyphs
9927 contained in the rendered text
9930 the number of input frame, starting from 0
9932 @item rand(min, max)
9933 return a random number included between @var{min} and @var{max}
9936 The input sample aspect ratio.
9939 timestamp expressed in seconds, NAN if the input timestamp is unknown
9942 the height of the rendered text
9945 the width of the rendered text
9949 the x and y offset coordinates where the text is drawn.
9951 These parameters allow the @var{x} and @var{y} expressions to refer
9952 to each other, so you can for example specify @code{y=x/dar}.
9955 A one character description of the current frame's picture type.
9958 The current packet's position in the input file or stream
9959 (in bytes, from the start of the input). A value of -1 indicates
9960 this info is not available.
9963 The current packet's duration, in seconds.
9966 The current packet's size (in bytes).
9969 @anchor{drawtext_expansion}
9970 @subsection Text expansion
9972 If @option{expansion} is set to @code{strftime},
9973 the filter recognizes strftime() sequences in the provided text and
9974 expands them accordingly. Check the documentation of strftime(). This
9975 feature is deprecated.
9977 If @option{expansion} is set to @code{none}, the text is printed verbatim.
9979 If @option{expansion} is set to @code{normal} (which is the default),
9980 the following expansion mechanism is used.
9982 The backslash character @samp{\}, followed by any character, always expands to
9983 the second character.
9985 Sequences of the form @code{%@{...@}} are expanded. The text between the
9986 braces is a function name, possibly followed by arguments separated by ':'.
9987 If the arguments contain special characters or delimiters (':' or '@}'),
9988 they should be escaped.
9990 Note that they probably must also be escaped as the value for the
9991 @option{text} option in the filter argument string and as the filter
9992 argument in the filtergraph description, and possibly also for the shell,
9993 that makes up to four levels of escaping; using a text file avoids these
9996 The following functions are available:
10001 The expression evaluation result.
10003 It must take one argument specifying the expression to be evaluated,
10004 which accepts the same constants and functions as the @var{x} and
10005 @var{y} values. Note that not all constants should be used, for
10006 example the text size is not known when evaluating the expression, so
10007 the constants @var{text_w} and @var{text_h} will have an undefined
10010 @item expr_int_format, eif
10011 Evaluate the expression's value and output as formatted integer.
10013 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10014 The second argument specifies the output format. Allowed values are @samp{x},
10015 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10016 @code{printf} function.
10017 The third parameter is optional and sets the number of positions taken by the output.
10018 It can be used to add padding with zeros from the left.
10021 The time at which the filter is running, expressed in UTC.
10022 It can accept an argument: a strftime() format string.
10025 The time at which the filter is running, expressed in the local time zone.
10026 It can accept an argument: a strftime() format string.
10029 Frame metadata. Takes one or two arguments.
10031 The first argument is mandatory and specifies the metadata key.
10033 The second argument is optional and specifies a default value, used when the
10034 metadata key is not found or empty.
10036 Available metadata can be identified by inspecting entries
10037 starting with TAG included within each frame section
10038 printed by running @code{ffprobe -show_frames}.
10040 String metadata generated in filters leading to
10041 the drawtext filter are also available.
10044 The frame number, starting from 0.
10047 A one character description of the current picture type.
10050 The timestamp of the current frame.
10051 It can take up to three arguments.
10053 The first argument is the format of the timestamp; it defaults to @code{flt}
10054 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10055 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10056 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10057 @code{localtime} stands for the timestamp of the frame formatted as
10058 local time zone time.
10060 The second argument is an offset added to the timestamp.
10062 If the format is set to @code{hms}, a third argument @code{24HH} may be
10063 supplied to present the hour part of the formatted timestamp in 24h format
10066 If the format is set to @code{localtime} or @code{gmtime},
10067 a third argument may be supplied: a strftime() format string.
10068 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10071 @subsection Commands
10073 This filter supports altering parameters via commands:
10076 Alter existing filter parameters.
10078 Syntax for the argument is the same as for filter invocation, e.g.
10081 fontsize=56:fontcolor=green:text='Hello World'
10084 Full filter invocation with sendcmd would look like this:
10087 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10091 If the entire argument can't be parsed or applied as valid values then the filter will
10092 continue with its existing parameters.
10094 @subsection Examples
10098 Draw "Test Text" with font FreeSerif, using the default values for the
10099 optional parameters.
10102 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10106 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10107 and y=50 (counting from the top-left corner of the screen), text is
10108 yellow with a red box around it. Both the text and the box have an
10112 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10113 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10116 Note that the double quotes are not necessary if spaces are not used
10117 within the parameter list.
10120 Show the text at the center of the video frame:
10122 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10126 Show the text at a random position, switching to a new position every 30 seconds:
10128 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)"
10132 Show a text line sliding from right to left in the last row of the video
10133 frame. The file @file{LONG_LINE} is assumed to contain a single line
10136 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10140 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10142 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10146 Draw a single green letter "g", at the center of the input video.
10147 The glyph baseline is placed at half screen height.
10149 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10153 Show text for 1 second every 3 seconds:
10155 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10159 Use fontconfig to set the font. Note that the colons need to be escaped.
10161 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10165 Print the date of a real-time encoding (see strftime(3)):
10167 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10171 Show text fading in and out (appearing/disappearing):
10174 DS=1.0 # display start
10175 DE=10.0 # display end
10176 FID=1.5 # fade in duration
10177 FOD=5 # fade out duration
10178 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 @}"
10182 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10183 and the @option{fontsize} value are included in the @option{y} offset.
10185 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10186 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10190 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10191 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10192 must have option @option{-export_path_metadata 1} for the special metadata fields
10193 to be available for filters.
10195 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10200 For more information about libfreetype, check:
10201 @url{http://www.freetype.org/}.
10203 For more information about fontconfig, check:
10204 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10206 For more information about libfribidi, check:
10207 @url{http://fribidi.org/}.
10209 @section edgedetect
10211 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10213 The filter accepts the following options:
10218 Set low and high threshold values used by the Canny thresholding
10221 The high threshold selects the "strong" edge pixels, which are then
10222 connected through 8-connectivity with the "weak" edge pixels selected
10223 by the low threshold.
10225 @var{low} and @var{high} threshold values must be chosen in the range
10226 [0,1], and @var{low} should be lesser or equal to @var{high}.
10228 Default value for @var{low} is @code{20/255}, and default value for @var{high}
10232 Define the drawing mode.
10236 Draw white/gray wires on black background.
10239 Mix the colors to create a paint/cartoon effect.
10242 Apply Canny edge detector on all selected planes.
10244 Default value is @var{wires}.
10247 Select planes for filtering. By default all available planes are filtered.
10250 @subsection Examples
10254 Standard edge detection with custom values for the hysteresis thresholding:
10256 edgedetect=low=0.1:high=0.4
10260 Painting effect without thresholding:
10262 edgedetect=mode=colormix:high=0
10268 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
10270 For each input image, the filter will compute the optimal mapping from
10271 the input to the output given the codebook length, that is the number
10272 of distinct output colors.
10274 This filter accepts the following options.
10277 @item codebook_length, l
10278 Set codebook length. The value must be a positive integer, and
10279 represents the number of distinct output colors. Default value is 256.
10282 Set the maximum number of iterations to apply for computing the optimal
10283 mapping. The higher the value the better the result and the higher the
10284 computation time. Default value is 1.
10287 Set a random seed, must be an integer included between 0 and
10288 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
10289 will try to use a good random seed on a best effort basis.
10292 Set pal8 output pixel format. This option does not work with codebook
10293 length greater than 256.
10298 Measure graylevel entropy in histogram of color channels of video frames.
10300 It accepts the following parameters:
10304 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
10306 @var{diff} mode measures entropy of histogram delta values, absolute differences
10307 between neighbour histogram values.
10311 Set brightness, contrast, saturation and approximate gamma adjustment.
10313 The filter accepts the following options:
10317 Set the contrast expression. The value must be a float value in range
10318 @code{-1000.0} to @code{1000.0}. The default value is "1".
10321 Set the brightness expression. The value must be a float value in
10322 range @code{-1.0} to @code{1.0}. The default value is "0".
10325 Set the saturation expression. The value must be a float in
10326 range @code{0.0} to @code{3.0}. The default value is "1".
10329 Set the gamma expression. The value must be a float in range
10330 @code{0.1} to @code{10.0}. The default value is "1".
10333 Set the gamma expression for red. The value must be a float in
10334 range @code{0.1} to @code{10.0}. The default value is "1".
10337 Set the gamma expression for green. The value must be a float in range
10338 @code{0.1} to @code{10.0}. The default value is "1".
10341 Set the gamma expression for blue. The value must be a float in range
10342 @code{0.1} to @code{10.0}. The default value is "1".
10345 Set the gamma weight expression. It can be used to reduce the effect
10346 of a high gamma value on bright image areas, e.g. keep them from
10347 getting overamplified and just plain white. The value must be a float
10348 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
10349 gamma correction all the way down while @code{1.0} leaves it at its
10350 full strength. Default is "1".
10353 Set when the expressions for brightness, contrast, saturation and
10354 gamma expressions are evaluated.
10356 It accepts the following values:
10359 only evaluate expressions once during the filter initialization or
10360 when a command is processed
10363 evaluate expressions for each incoming frame
10366 Default value is @samp{init}.
10369 The expressions accept the following parameters:
10372 frame count of the input frame starting from 0
10375 byte position of the corresponding packet in the input file, NAN if
10379 frame rate of the input video, NAN if the input frame rate is unknown
10382 timestamp expressed in seconds, NAN if the input timestamp is unknown
10385 @subsection Commands
10386 The filter supports the following commands:
10390 Set the contrast expression.
10393 Set the brightness expression.
10396 Set the saturation expression.
10399 Set the gamma expression.
10402 Set the gamma_r expression.
10405 Set gamma_g expression.
10408 Set gamma_b expression.
10411 Set gamma_weight expression.
10413 The command accepts the same syntax of the corresponding option.
10415 If the specified expression is not valid, it is kept at its current
10422 Apply erosion effect to the video.
10424 This filter replaces the pixel by the local(3x3) minimum.
10426 It accepts the following options:
10433 Limit the maximum change for each plane, default is 65535.
10434 If 0, plane will remain unchanged.
10437 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10440 Flags to local 3x3 coordinates maps like this:
10447 @subsection Commands
10449 This filter supports the all above options as @ref{commands}.
10451 @section extractplanes
10453 Extract color channel components from input video stream into
10454 separate grayscale video streams.
10456 The filter accepts the following option:
10460 Set plane(s) to extract.
10462 Available values for planes are:
10473 Choosing planes not available in the input will result in an error.
10474 That means you cannot select @code{r}, @code{g}, @code{b} planes
10475 with @code{y}, @code{u}, @code{v} planes at same time.
10478 @subsection Examples
10482 Extract luma, u and v color channel component from input video frame
10483 into 3 grayscale outputs:
10485 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
10491 Apply a fade-in/out effect to the input video.
10493 It accepts the following parameters:
10497 The effect type can be either "in" for a fade-in, or "out" for a fade-out
10499 Default is @code{in}.
10501 @item start_frame, s
10502 Specify the number of the frame to start applying the fade
10503 effect at. Default is 0.
10506 The number of frames that the fade effect lasts. At the end of the
10507 fade-in effect, the output video will have the same intensity as the input video.
10508 At the end of the fade-out transition, the output video will be filled with the
10509 selected @option{color}.
10513 If set to 1, fade only alpha channel, if one exists on the input.
10514 Default value is 0.
10516 @item start_time, st
10517 Specify the timestamp (in seconds) of the frame to start to apply the fade
10518 effect. If both start_frame and start_time are specified, the fade will start at
10519 whichever comes last. Default is 0.
10522 The number of seconds for which the fade effect has to last. At the end of the
10523 fade-in effect the output video will have the same intensity as the input video,
10524 at the end of the fade-out transition the output video will be filled with the
10525 selected @option{color}.
10526 If both duration and nb_frames are specified, duration is used. Default is 0
10527 (nb_frames is used by default).
10530 Specify the color of the fade. Default is "black".
10533 @subsection Examples
10537 Fade in the first 30 frames of video:
10542 The command above is equivalent to:
10548 Fade out the last 45 frames of a 200-frame video:
10551 fade=type=out:start_frame=155:nb_frames=45
10555 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
10557 fade=in:0:25, fade=out:975:25
10561 Make the first 5 frames yellow, then fade in from frame 5-24:
10563 fade=in:5:20:color=yellow
10567 Fade in alpha over first 25 frames of video:
10569 fade=in:0:25:alpha=1
10573 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
10575 fade=t=in:st=5.5:d=0.5
10581 Denoise frames using 3D FFT (frequency domain filtering).
10583 The filter accepts the following options:
10587 Set the noise sigma constant. This sets denoising strength.
10588 Default value is 1. Allowed range is from 0 to 30.
10589 Using very high sigma with low overlap may give blocking artifacts.
10592 Set amount of denoising. By default all detected noise is reduced.
10593 Default value is 1. Allowed range is from 0 to 1.
10596 Set size of block, Default is 4, can be 3, 4, 5 or 6.
10597 Actual size of block in pixels is 2 to power of @var{block}, so by default
10598 block size in pixels is 2^4 which is 16.
10601 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
10604 Set number of previous frames to use for denoising. By default is set to 0.
10607 Set number of next frames to to use for denoising. By default is set to 0.
10610 Set planes which will be filtered, by default are all available filtered
10615 Apply arbitrary expressions to samples in frequency domain
10619 Adjust the dc value (gain) of the luma plane of the image. The filter
10620 accepts an integer value in range @code{0} to @code{1000}. The default
10621 value is set to @code{0}.
10624 Adjust the dc value (gain) of the 1st chroma plane of the image. The
10625 filter accepts an integer value in range @code{0} to @code{1000}. The
10626 default value is set to @code{0}.
10629 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
10630 filter accepts an integer value in range @code{0} to @code{1000}. The
10631 default value is set to @code{0}.
10634 Set the frequency domain weight expression for the luma plane.
10637 Set the frequency domain weight expression for the 1st chroma plane.
10640 Set the frequency domain weight expression for the 2nd chroma plane.
10643 Set when the expressions are evaluated.
10645 It accepts the following values:
10648 Only evaluate expressions once during the filter initialization.
10651 Evaluate expressions for each incoming frame.
10654 Default value is @samp{init}.
10656 The filter accepts the following variables:
10659 The coordinates of the current sample.
10663 The width and height of the image.
10666 The number of input frame, starting from 0.
10669 @subsection Examples
10675 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
10681 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
10687 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
10693 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
10700 Extract a single field from an interlaced image using stride
10701 arithmetic to avoid wasting CPU time. The output frames are marked as
10704 The filter accepts the following options:
10708 Specify whether to extract the top (if the value is @code{0} or
10709 @code{top}) or the bottom field (if the value is @code{1} or
10715 Create new frames by copying the top and bottom fields from surrounding frames
10716 supplied as numbers by the hint file.
10720 Set file containing hints: absolute/relative frame numbers.
10722 There must be one line for each frame in a clip. Each line must contain two
10723 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
10724 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
10725 is current frame number for @code{absolute} mode or out of [-1, 1] range
10726 for @code{relative} mode. First number tells from which frame to pick up top
10727 field and second number tells from which frame to pick up bottom field.
10729 If optionally followed by @code{+} output frame will be marked as interlaced,
10730 else if followed by @code{-} output frame will be marked as progressive, else
10731 it will be marked same as input frame.
10732 If optionally followed by @code{t} output frame will use only top field, or in
10733 case of @code{b} it will use only bottom field.
10734 If line starts with @code{#} or @code{;} that line is skipped.
10737 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
10740 Example of first several lines of @code{hint} file for @code{relative} mode:
10742 0,0 - # first frame
10743 1,0 - # second frame, use third's frame top field and second's frame bottom field
10744 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
10759 @section fieldmatch
10761 Field matching filter for inverse telecine. It is meant to reconstruct the
10762 progressive frames from a telecined stream. The filter does not drop duplicated
10763 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
10764 followed by a decimation filter such as @ref{decimate} in the filtergraph.
10766 The separation of the field matching and the decimation is notably motivated by
10767 the possibility of inserting a de-interlacing filter fallback between the two.
10768 If the source has mixed telecined and real interlaced content,
10769 @code{fieldmatch} will not be able to match fields for the interlaced parts.
10770 But these remaining combed frames will be marked as interlaced, and thus can be
10771 de-interlaced by a later filter such as @ref{yadif} before decimation.
10773 In addition to the various configuration options, @code{fieldmatch} can take an
10774 optional second stream, activated through the @option{ppsrc} option. If
10775 enabled, the frames reconstruction will be based on the fields and frames from
10776 this second stream. This allows the first input to be pre-processed in order to
10777 help the various algorithms of the filter, while keeping the output lossless
10778 (assuming the fields are matched properly). Typically, a field-aware denoiser,
10779 or brightness/contrast adjustments can help.
10781 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
10782 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
10783 which @code{fieldmatch} is based on. While the semantic and usage are very
10784 close, some behaviour and options names can differ.
10786 The @ref{decimate} filter currently only works for constant frame rate input.
10787 If your input has mixed telecined (30fps) and progressive content with a lower
10788 framerate like 24fps use the following filterchain to produce the necessary cfr
10789 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
10791 The filter accepts the following options:
10795 Specify the assumed field order of the input stream. Available values are:
10799 Auto detect parity (use FFmpeg's internal parity value).
10801 Assume bottom field first.
10803 Assume top field first.
10806 Note that it is sometimes recommended not to trust the parity announced by the
10809 Default value is @var{auto}.
10812 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
10813 sense that it won't risk creating jerkiness due to duplicate frames when
10814 possible, but if there are bad edits or blended fields it will end up
10815 outputting combed frames when a good match might actually exist. On the other
10816 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
10817 but will almost always find a good frame if there is one. The other values are
10818 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
10819 jerkiness and creating duplicate frames versus finding good matches in sections
10820 with bad edits, orphaned fields, blended fields, etc.
10822 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
10824 Available values are:
10828 2-way matching (p/c)
10830 2-way matching, and trying 3rd match if still combed (p/c + n)
10832 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
10834 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
10835 still combed (p/c + n + u/b)
10837 3-way matching (p/c/n)
10839 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
10840 detected as combed (p/c/n + u/b)
10843 The parenthesis at the end indicate the matches that would be used for that
10844 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
10847 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
10850 Default value is @var{pc_n}.
10853 Mark the main input stream as a pre-processed input, and enable the secondary
10854 input stream as the clean source to pick the fields from. See the filter
10855 introduction for more details. It is similar to the @option{clip2} feature from
10858 Default value is @code{0} (disabled).
10861 Set the field to match from. It is recommended to set this to the same value as
10862 @option{order} unless you experience matching failures with that setting. In
10863 certain circumstances changing the field that is used to match from can have a
10864 large impact on matching performance. Available values are:
10868 Automatic (same value as @option{order}).
10870 Match from the bottom field.
10872 Match from the top field.
10875 Default value is @var{auto}.
10878 Set whether or not chroma is included during the match comparisons. In most
10879 cases it is recommended to leave this enabled. You should set this to @code{0}
10880 only if your clip has bad chroma problems such as heavy rainbowing or other
10881 artifacts. Setting this to @code{0} could also be used to speed things up at
10882 the cost of some accuracy.
10884 Default value is @code{1}.
10888 These define an exclusion band which excludes the lines between @option{y0} and
10889 @option{y1} from being included in the field matching decision. An exclusion
10890 band can be used to ignore subtitles, a logo, or other things that may
10891 interfere with the matching. @option{y0} sets the starting scan line and
10892 @option{y1} sets the ending line; all lines in between @option{y0} and
10893 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
10894 @option{y0} and @option{y1} to the same value will disable the feature.
10895 @option{y0} and @option{y1} defaults to @code{0}.
10898 Set the scene change detection threshold as a percentage of maximum change on
10899 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
10900 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
10901 @option{scthresh} is @code{[0.0, 100.0]}.
10903 Default value is @code{12.0}.
10906 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
10907 account the combed scores of matches when deciding what match to use as the
10908 final match. Available values are:
10912 No final matching based on combed scores.
10914 Combed scores are only used when a scene change is detected.
10916 Use combed scores all the time.
10919 Default is @var{sc}.
10922 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
10923 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
10924 Available values are:
10928 No forced calculation.
10930 Force p/c/n calculations.
10932 Force p/c/n/u/b calculations.
10935 Default value is @var{none}.
10938 This is the area combing threshold used for combed frame detection. This
10939 essentially controls how "strong" or "visible" combing must be to be detected.
10940 Larger values mean combing must be more visible and smaller values mean combing
10941 can be less visible or strong and still be detected. Valid settings are from
10942 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
10943 be detected as combed). This is basically a pixel difference value. A good
10944 range is @code{[8, 12]}.
10946 Default value is @code{9}.
10949 Sets whether or not chroma is considered in the combed frame decision. Only
10950 disable this if your source has chroma problems (rainbowing, etc.) that are
10951 causing problems for the combed frame detection with chroma enabled. Actually,
10952 using @option{chroma}=@var{0} is usually more reliable, except for the case
10953 where there is chroma only combing in the source.
10955 Default value is @code{0}.
10959 Respectively set the x-axis and y-axis size of the window used during combed
10960 frame detection. This has to do with the size of the area in which
10961 @option{combpel} pixels are required to be detected as combed for a frame to be
10962 declared combed. See the @option{combpel} parameter description for more info.
10963 Possible values are any number that is a power of 2 starting at 4 and going up
10966 Default value is @code{16}.
10969 The number of combed pixels inside any of the @option{blocky} by
10970 @option{blockx} size blocks on the frame for the frame to be detected as
10971 combed. While @option{cthresh} controls how "visible" the combing must be, this
10972 setting controls "how much" combing there must be in any localized area (a
10973 window defined by the @option{blockx} and @option{blocky} settings) on the
10974 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
10975 which point no frames will ever be detected as combed). This setting is known
10976 as @option{MI} in TFM/VFM vocabulary.
10978 Default value is @code{80}.
10981 @anchor{p/c/n/u/b meaning}
10982 @subsection p/c/n/u/b meaning
10984 @subsubsection p/c/n
10986 We assume the following telecined stream:
10989 Top fields: 1 2 2 3 4
10990 Bottom fields: 1 2 3 4 4
10993 The numbers correspond to the progressive frame the fields relate to. Here, the
10994 first two frames are progressive, the 3rd and 4th are combed, and so on.
10996 When @code{fieldmatch} is configured to run a matching from bottom
10997 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11002 B 1 2 3 4 4 <-- matching reference
11011 As a result of the field matching, we can see that some frames get duplicated.
11012 To perform a complete inverse telecine, you need to rely on a decimation filter
11013 after this operation. See for instance the @ref{decimate} filter.
11015 The same operation now matching from top fields (@option{field}=@var{top})
11020 T 1 2 2 3 4 <-- matching reference
11030 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11031 basically, they refer to the frame and field of the opposite parity:
11034 @item @var{p} matches the field of the opposite parity in the previous frame
11035 @item @var{c} matches the field of the opposite parity in the current frame
11036 @item @var{n} matches the field of the opposite parity in the next frame
11041 The @var{u} and @var{b} matching are a bit special in the sense that they match
11042 from the opposite parity flag. In the following examples, we assume that we are
11043 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11044 'x' is placed above and below each matched fields.
11046 With bottom matching (@option{field}=@var{bottom}):
11051 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11052 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11060 With top matching (@option{field}=@var{top}):
11065 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11066 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11074 @subsection Examples
11076 Simple IVTC of a top field first telecined stream:
11078 fieldmatch=order=tff:combmatch=none, decimate
11081 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11083 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11086 @section fieldorder
11088 Transform the field order of the input video.
11090 It accepts the following parameters:
11095 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11096 for bottom field first.
11099 The default value is @samp{tff}.
11101 The transformation is done by shifting the picture content up or down
11102 by one line, and filling the remaining line with appropriate picture content.
11103 This method is consistent with most broadcast field order converters.
11105 If the input video is not flagged as being interlaced, or it is already
11106 flagged as being of the required output field order, then this filter does
11107 not alter the incoming video.
11109 It is very useful when converting to or from PAL DV material,
11110 which is bottom field first.
11114 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11117 @section fifo, afifo
11119 Buffer input images and send them when they are requested.
11121 It is mainly useful when auto-inserted by the libavfilter
11124 It does not take parameters.
11126 @section fillborders
11128 Fill borders of the input video, without changing video stream dimensions.
11129 Sometimes video can have garbage at the four edges and you may not want to
11130 crop video input to keep size multiple of some number.
11132 This filter accepts the following options:
11136 Number of pixels to fill from left border.
11139 Number of pixels to fill from right border.
11142 Number of pixels to fill from top border.
11145 Number of pixels to fill from bottom border.
11150 It accepts the following values:
11153 fill pixels using outermost pixels
11156 fill pixels using mirroring
11159 fill pixels with constant value
11162 Default is @var{smear}.
11165 Set color for pixels in fixed mode. Default is @var{black}.
11168 @subsection Commands
11169 This filter supports same @ref{commands} as options.
11170 The command accepts the same syntax of the corresponding option.
11172 If the specified expression is not valid, it is kept at its current
11177 Find a rectangular object
11179 It accepts the following options:
11183 Filepath of the object image, needs to be in gray8.
11186 Detection threshold, default is 0.5.
11189 Number of mipmaps, default is 3.
11191 @item xmin, ymin, xmax, ymax
11192 Specifies the rectangle in which to search.
11195 @subsection Examples
11199 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
11201 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
11207 Flood area with values of same pixel components with another values.
11209 It accepts the following options:
11212 Set pixel x coordinate.
11215 Set pixel y coordinate.
11218 Set source #0 component value.
11221 Set source #1 component value.
11224 Set source #2 component value.
11227 Set source #3 component value.
11230 Set destination #0 component value.
11233 Set destination #1 component value.
11236 Set destination #2 component value.
11239 Set destination #3 component value.
11245 Convert the input video to one of the specified pixel formats.
11246 Libavfilter will try to pick one that is suitable as input to
11249 It accepts the following parameters:
11253 A '|'-separated list of pixel format names, such as
11254 "pix_fmts=yuv420p|monow|rgb24".
11258 @subsection Examples
11262 Convert the input video to the @var{yuv420p} format
11264 format=pix_fmts=yuv420p
11267 Convert the input video to any of the formats in the list
11269 format=pix_fmts=yuv420p|yuv444p|yuv410p
11276 Convert the video to specified constant frame rate by duplicating or dropping
11277 frames as necessary.
11279 It accepts the following parameters:
11283 The desired output frame rate. The default is @code{25}.
11286 Assume the first PTS should be the given value, in seconds. This allows for
11287 padding/trimming at the start of stream. By default, no assumption is made
11288 about the first frame's expected PTS, so no padding or trimming is done.
11289 For example, this could be set to 0 to pad the beginning with duplicates of
11290 the first frame if a video stream starts after the audio stream or to trim any
11291 frames with a negative PTS.
11294 Timestamp (PTS) rounding method.
11296 Possible values are:
11303 round towards -infinity
11305 round towards +infinity
11309 The default is @code{near}.
11312 Action performed when reading the last frame.
11314 Possible values are:
11317 Use same timestamp rounding method as used for other frames.
11319 Pass through last frame if input duration has not been reached yet.
11321 The default is @code{round}.
11325 Alternatively, the options can be specified as a flat string:
11326 @var{fps}[:@var{start_time}[:@var{round}]].
11328 See also the @ref{setpts} filter.
11330 @subsection Examples
11334 A typical usage in order to set the fps to 25:
11340 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
11342 fps=fps=film:round=near
11348 Pack two different video streams into a stereoscopic video, setting proper
11349 metadata on supported codecs. The two views should have the same size and
11350 framerate and processing will stop when the shorter video ends. Please note
11351 that you may conveniently adjust view properties with the @ref{scale} and
11354 It accepts the following parameters:
11358 The desired packing format. Supported values are:
11363 The views are next to each other (default).
11366 The views are on top of each other.
11369 The views are packed by line.
11372 The views are packed by column.
11375 The views are temporally interleaved.
11384 # Convert left and right views into a frame-sequential video
11385 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
11387 # Convert views into a side-by-side video with the same output resolution as the input
11388 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
11393 Change the frame rate by interpolating new video output frames from the source
11396 This filter is not designed to function correctly with interlaced media. If
11397 you wish to change the frame rate of interlaced media then you are required
11398 to deinterlace before this filter and re-interlace after this filter.
11400 A description of the accepted options follows.
11404 Specify the output frames per second. This option can also be specified
11405 as a value alone. The default is @code{50}.
11408 Specify the start of a range where the output frame will be created as a
11409 linear interpolation of two frames. The range is [@code{0}-@code{255}],
11410 the default is @code{15}.
11413 Specify the end of a range where the output frame will be created as a
11414 linear interpolation of two frames. The range is [@code{0}-@code{255}],
11415 the default is @code{240}.
11418 Specify the level at which a scene change is detected as a value between
11419 0 and 100 to indicate a new scene; a low value reflects a low
11420 probability for the current frame to introduce a new scene, while a higher
11421 value means the current frame is more likely to be one.
11422 The default is @code{8.2}.
11425 Specify flags influencing the filter process.
11427 Available value for @var{flags} is:
11430 @item scene_change_detect, scd
11431 Enable scene change detection using the value of the option @var{scene}.
11432 This flag is enabled by default.
11438 Select one frame every N-th frame.
11440 This filter accepts the following option:
11443 Select frame after every @code{step} frames.
11444 Allowed values are positive integers higher than 0. Default value is @code{1}.
11447 @section freezedetect
11449 Detect frozen video.
11451 This filter logs a message and sets frame metadata when it detects that the
11452 input video has no significant change in content during a specified duration.
11453 Video freeze detection calculates the mean average absolute difference of all
11454 the components of video frames and compares it to a noise floor.
11456 The printed times and duration are expressed in seconds. The
11457 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
11458 whose timestamp equals or exceeds the detection duration and it contains the
11459 timestamp of the first frame of the freeze. The
11460 @code{lavfi.freezedetect.freeze_duration} and
11461 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
11464 The filter accepts the following options:
11468 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
11469 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
11473 Set freeze duration until notification (default is 2 seconds).
11476 @section freezeframes
11478 Freeze video frames.
11480 This filter freezes video frames using frame from 2nd input.
11482 The filter accepts the following options:
11486 Set number of first frame from which to start freeze.
11489 Set number of last frame from which to end freeze.
11492 Set number of frame from 2nd input which will be used instead of replaced frames.
11498 Apply a frei0r effect to the input video.
11500 To enable the compilation of this filter, you need to install the frei0r
11501 header and configure FFmpeg with @code{--enable-frei0r}.
11503 It accepts the following parameters:
11508 The name of the frei0r effect to load. If the environment variable
11509 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
11510 directories specified by the colon-separated list in @env{FREI0R_PATH}.
11511 Otherwise, the standard frei0r paths are searched, in this order:
11512 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
11513 @file{/usr/lib/frei0r-1/}.
11515 @item filter_params
11516 A '|'-separated list of parameters to pass to the frei0r effect.
11520 A frei0r effect parameter can be a boolean (its value is either
11521 "y" or "n"), a double, a color (specified as
11522 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
11523 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
11524 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
11525 a position (specified as @var{X}/@var{Y}, where
11526 @var{X} and @var{Y} are floating point numbers) and/or a string.
11528 The number and types of parameters depend on the loaded effect. If an
11529 effect parameter is not specified, the default value is set.
11531 @subsection Examples
11535 Apply the distort0r effect, setting the first two double parameters:
11537 frei0r=filter_name=distort0r:filter_params=0.5|0.01
11541 Apply the colordistance effect, taking a color as the first parameter:
11543 frei0r=colordistance:0.2/0.3/0.4
11544 frei0r=colordistance:violet
11545 frei0r=colordistance:0x112233
11549 Apply the perspective effect, specifying the top left and top right image
11552 frei0r=perspective:0.2/0.2|0.8/0.2
11556 For more information, see
11557 @url{http://frei0r.dyne.org}
11561 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
11563 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
11564 processing filter, one of them is performed once per block, not per pixel.
11565 This allows for much higher speed.
11567 The filter accepts the following options:
11571 Set quality. This option defines the number of levels for averaging. It accepts
11572 an integer in the range 4-5. Default value is @code{4}.
11575 Force a constant quantization parameter. It accepts an integer in range 0-63.
11576 If not set, the filter will use the QP from the video stream (if available).
11579 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
11580 more details but also more artifacts, while higher values make the image smoother
11581 but also blurrier. Default value is @code{0} − PSNR optimal.
11583 @item use_bframe_qp
11584 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
11585 option may cause flicker since the B-Frames have often larger QP. Default is
11586 @code{0} (not enabled).
11592 Apply Gaussian blur filter.
11594 The filter accepts the following options:
11598 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
11601 Set number of steps for Gaussian approximation. Default is @code{1}.
11604 Set which planes to filter. By default all planes are filtered.
11607 Set vertical sigma, if negative it will be same as @code{sigma}.
11608 Default is @code{-1}.
11611 @subsection Commands
11612 This filter supports same commands as options.
11613 The command accepts the same syntax of the corresponding option.
11615 If the specified expression is not valid, it is kept at its current
11620 Apply generic equation to each pixel.
11622 The filter accepts the following options:
11625 @item lum_expr, lum
11626 Set the luminance expression.
11628 Set the chrominance blue expression.
11630 Set the chrominance red expression.
11631 @item alpha_expr, a
11632 Set the alpha expression.
11634 Set the red expression.
11635 @item green_expr, g
11636 Set the green expression.
11638 Set the blue expression.
11641 The colorspace is selected according to the specified options. If one
11642 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
11643 options is specified, the filter will automatically select a YCbCr
11644 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
11645 @option{blue_expr} options is specified, it will select an RGB
11648 If one of the chrominance expression is not defined, it falls back on the other
11649 one. If no alpha expression is specified it will evaluate to opaque value.
11650 If none of chrominance expressions are specified, they will evaluate
11651 to the luminance expression.
11653 The expressions can use the following variables and functions:
11657 The sequential number of the filtered frame, starting from @code{0}.
11661 The coordinates of the current sample.
11665 The width and height of the image.
11669 Width and height scale depending on the currently filtered plane. It is the
11670 ratio between the corresponding luma plane number of pixels and the current
11671 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
11672 @code{0.5,0.5} for chroma planes.
11675 Time of the current frame, expressed in seconds.
11678 Return the value of the pixel at location (@var{x},@var{y}) of the current
11682 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
11686 Return the value of the pixel at location (@var{x},@var{y}) of the
11687 blue-difference chroma plane. Return 0 if there is no such plane.
11690 Return the value of the pixel at location (@var{x},@var{y}) of the
11691 red-difference chroma plane. Return 0 if there is no such plane.
11696 Return the value of the pixel at location (@var{x},@var{y}) of the
11697 red/green/blue component. Return 0 if there is no such component.
11700 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
11701 plane. Return 0 if there is no such plane.
11703 @item psum(x,y), lumsum(x, y), cbsum(x,y), crsum(x,y), rsum(x,y), gsum(x,y), bsum(x,y), alphasum(x,y)
11704 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
11705 sums of samples within a rectangle. See the functions without the sum postfix.
11707 @item interpolation
11708 Set one of interpolation methods:
11713 Default is bilinear.
11716 For functions, if @var{x} and @var{y} are outside the area, the value will be
11717 automatically clipped to the closer edge.
11719 Please note that this filter can use multiple threads in which case each slice
11720 will have its own expression state. If you want to use only a single expression
11721 state because your expressions depend on previous state then you should limit
11722 the number of filter threads to 1.
11724 @subsection Examples
11728 Flip the image horizontally:
11734 Generate a bidimensional sine wave, with angle @code{PI/3} and a
11735 wavelength of 100 pixels:
11737 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
11741 Generate a fancy enigmatic moving light:
11743 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
11747 Generate a quick emboss effect:
11749 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
11753 Modify RGB components depending on pixel position:
11755 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
11759 Create a radial gradient that is the same size as the input (also see
11760 the @ref{vignette} filter):
11762 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
11768 Fix the banding artifacts that are sometimes introduced into nearly flat
11769 regions by truncation to 8-bit color depth.
11770 Interpolate the gradients that should go where the bands are, and
11773 It is designed for playback only. Do not use it prior to
11774 lossy compression, because compression tends to lose the dither and
11775 bring back the bands.
11777 It accepts the following parameters:
11782 The maximum amount by which the filter will change any one pixel. This is also
11783 the threshold for detecting nearly flat regions. Acceptable values range from
11784 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
11788 The neighborhood to fit the gradient to. A larger radius makes for smoother
11789 gradients, but also prevents the filter from modifying the pixels near detailed
11790 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
11791 values will be clipped to the valid range.
11795 Alternatively, the options can be specified as a flat string:
11796 @var{strength}[:@var{radius}]
11798 @subsection Examples
11802 Apply the filter with a @code{3.5} strength and radius of @code{8}:
11808 Specify radius, omitting the strength (which will fall-back to the default
11816 @anchor{graphmonitor}
11817 @section graphmonitor
11818 Show various filtergraph stats.
11820 With this filter one can debug complete filtergraph.
11821 Especially issues with links filling with queued frames.
11823 The filter accepts the following options:
11827 Set video output size. Default is @var{hd720}.
11830 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
11833 Set output mode, can be @var{fulll} or @var{compact}.
11834 In @var{compact} mode only filters with some queued frames have displayed stats.
11837 Set flags which enable which stats are shown in video.
11839 Available values for flags are:
11842 Display number of queued frames in each link.
11844 @item frame_count_in
11845 Display number of frames taken from filter.
11847 @item frame_count_out
11848 Display number of frames given out from filter.
11851 Display current filtered frame pts.
11854 Display current filtered frame time.
11857 Display time base for filter link.
11860 Display used format for filter link.
11863 Display video size or number of audio channels in case of audio used by filter link.
11866 Display video frame rate or sample rate in case of audio used by filter link.
11870 Set upper limit for video rate of output stream, Default value is @var{25}.
11871 This guarantee that output video frame rate will not be higher than this value.
11875 A color constancy variation filter which estimates scene illumination via grey edge algorithm
11876 and corrects the scene colors accordingly.
11878 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
11880 The filter accepts the following options:
11884 The order of differentiation to be applied on the scene. Must be chosen in the range
11885 [0,2] and default value is 1.
11888 The Minkowski parameter to be used for calculating the Minkowski distance. Must
11889 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
11890 max value instead of calculating Minkowski distance.
11893 The standard deviation of Gaussian blur to be applied on the scene. Must be
11894 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
11895 can't be equal to 0 if @var{difford} is greater than 0.
11898 @subsection Examples
11904 greyedge=difford=1:minknorm=5:sigma=2
11910 greyedge=difford=1:minknorm=0:sigma=2
11918 Apply a Hald CLUT to a video stream.
11920 First input is the video stream to process, and second one is the Hald CLUT.
11921 The Hald CLUT input can be a simple picture or a complete video stream.
11923 The filter accepts the following options:
11927 Force termination when the shortest input terminates. Default is @code{0}.
11929 Continue applying the last CLUT after the end of the stream. A value of
11930 @code{0} disable the filter after the last frame of the CLUT is reached.
11931 Default is @code{1}.
11934 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
11935 filters share the same internals).
11937 This filter also supports the @ref{framesync} options.
11939 More information about the Hald CLUT can be found on Eskil Steenberg's website
11940 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
11942 @subsection Workflow examples
11944 @subsubsection Hald CLUT video stream
11946 Generate an identity Hald CLUT stream altered with various effects:
11948 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
11951 Note: make sure you use a lossless codec.
11953 Then use it with @code{haldclut} to apply it on some random stream:
11955 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
11958 The Hald CLUT will be applied to the 10 first seconds (duration of
11959 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
11960 to the remaining frames of the @code{mandelbrot} stream.
11962 @subsubsection Hald CLUT with preview
11964 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
11965 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
11966 biggest possible square starting at the top left of the picture. The remaining
11967 padding pixels (bottom or right) will be ignored. This area can be used to add
11968 a preview of the Hald CLUT.
11970 Typically, the following generated Hald CLUT will be supported by the
11971 @code{haldclut} filter:
11974 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
11975 pad=iw+320 [padded_clut];
11976 smptebars=s=320x256, split [a][b];
11977 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
11978 [main][b] overlay=W-320" -frames:v 1 clut.png
11981 It contains the original and a preview of the effect of the CLUT: SMPTE color
11982 bars are displayed on the right-top, and below the same color bars processed by
11985 Then, the effect of this Hald CLUT can be visualized with:
11987 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
11992 Flip the input video horizontally.
11994 For example, to horizontally flip the input video with @command{ffmpeg}:
11996 ffmpeg -i in.avi -vf "hflip" out.avi
12000 This filter applies a global color histogram equalization on a
12003 It can be used to correct video that has a compressed range of pixel
12004 intensities. The filter redistributes the pixel intensities to
12005 equalize their distribution across the intensity range. It may be
12006 viewed as an "automatically adjusting contrast filter". This filter is
12007 useful only for correcting degraded or poorly captured source
12010 The filter accepts the following options:
12014 Determine the amount of equalization to be applied. As the strength
12015 is reduced, the distribution of pixel intensities more-and-more
12016 approaches that of the input frame. The value must be a float number
12017 in the range [0,1] and defaults to 0.200.
12020 Set the maximum intensity that can generated and scale the output
12021 values appropriately. The strength should be set as desired and then
12022 the intensity can be limited if needed to avoid washing-out. The value
12023 must be a float number in the range [0,1] and defaults to 0.210.
12026 Set the antibanding level. If enabled the filter will randomly vary
12027 the luminance of output pixels by a small amount to avoid banding of
12028 the histogram. Possible values are @code{none}, @code{weak} or
12029 @code{strong}. It defaults to @code{none}.
12035 Compute and draw a color distribution histogram for the input video.
12037 The computed histogram is a representation of the color component
12038 distribution in an image.
12040 Standard histogram displays the color components distribution in an image.
12041 Displays color graph for each color component. Shows distribution of
12042 the Y, U, V, A or R, G, B components, depending on input format, in the
12043 current frame. Below each graph a color component scale meter is shown.
12045 The filter accepts the following options:
12049 Set height of level. Default value is @code{200}.
12050 Allowed range is [50, 2048].
12053 Set height of color scale. Default value is @code{12}.
12054 Allowed range is [0, 40].
12058 It accepts the following values:
12061 Per color component graphs are placed below each other.
12064 Per color component graphs are placed side by side.
12067 Presents information identical to that in the @code{parade}, except
12068 that the graphs representing color components are superimposed directly
12071 Default is @code{stack}.
12074 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12075 Default is @code{linear}.
12078 Set what color components to display.
12079 Default is @code{7}.
12082 Set foreground opacity. Default is @code{0.7}.
12085 Set background opacity. Default is @code{0.5}.
12088 @subsection Examples
12093 Calculate and draw histogram:
12095 ffplay -i input -vf histogram
12103 This is a high precision/quality 3d denoise filter. It aims to reduce
12104 image noise, producing smooth images and making still images really
12105 still. It should enhance compressibility.
12107 It accepts the following optional parameters:
12111 A non-negative floating point number which specifies spatial luma strength.
12112 It defaults to 4.0.
12114 @item chroma_spatial
12115 A non-negative floating point number which specifies spatial chroma strength.
12116 It defaults to 3.0*@var{luma_spatial}/4.0.
12119 A floating point number which specifies luma temporal strength. It defaults to
12120 6.0*@var{luma_spatial}/4.0.
12123 A floating point number which specifies chroma temporal strength. It defaults to
12124 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
12127 @subsection Commands
12128 This filter supports same @ref{commands} as options.
12129 The command accepts the same syntax of the corresponding option.
12131 If the specified expression is not valid, it is kept at its current
12134 @anchor{hwdownload}
12135 @section hwdownload
12137 Download hardware frames to system memory.
12139 The input must be in hardware frames, and the output a non-hardware format.
12140 Not all formats will be supported on the output - it may be necessary to insert
12141 an additional @option{format} filter immediately following in the graph to get
12142 the output in a supported format.
12146 Map hardware frames to system memory or to another device.
12148 This filter has several different modes of operation; which one is used depends
12149 on the input and output formats:
12152 Hardware frame input, normal frame output
12154 Map the input frames to system memory and pass them to the output. If the
12155 original hardware frame is later required (for example, after overlaying
12156 something else on part of it), the @option{hwmap} filter can be used again
12157 in the next mode to retrieve it.
12159 Normal frame input, hardware frame output
12161 If the input is actually a software-mapped hardware frame, then unmap it -
12162 that is, return the original hardware frame.
12164 Otherwise, a device must be provided. Create new hardware surfaces on that
12165 device for the output, then map them back to the software format at the input
12166 and give those frames to the preceding filter. This will then act like the
12167 @option{hwupload} filter, but may be able to avoid an additional copy when
12168 the input is already in a compatible format.
12170 Hardware frame input and output
12172 A device must be supplied for the output, either directly or with the
12173 @option{derive_device} option. The input and output devices must be of
12174 different types and compatible - the exact meaning of this is
12175 system-dependent, but typically it means that they must refer to the same
12176 underlying hardware context (for example, refer to the same graphics card).
12178 If the input frames were originally created on the output device, then unmap
12179 to retrieve the original frames.
12181 Otherwise, map the frames to the output device - create new hardware frames
12182 on the output corresponding to the frames on the input.
12185 The following additional parameters are accepted:
12189 Set the frame mapping mode. Some combination of:
12192 The mapped frame should be readable.
12194 The mapped frame should be writeable.
12196 The mapping will always overwrite the entire frame.
12198 This may improve performance in some cases, as the original contents of the
12199 frame need not be loaded.
12201 The mapping must not involve any copying.
12203 Indirect mappings to copies of frames are created in some cases where either
12204 direct mapping is not possible or it would have unexpected properties.
12205 Setting this flag ensures that the mapping is direct and will fail if that is
12208 Defaults to @var{read+write} if not specified.
12210 @item derive_device @var{type}
12211 Rather than using the device supplied at initialisation, instead derive a new
12212 device of type @var{type} from the device the input frames exist on.
12215 In a hardware to hardware mapping, map in reverse - create frames in the sink
12216 and map them back to the source. This may be necessary in some cases where
12217 a mapping in one direction is required but only the opposite direction is
12218 supported by the devices being used.
12220 This option is dangerous - it may break the preceding filter in undefined
12221 ways if there are any additional constraints on that filter's output.
12222 Do not use it without fully understanding the implications of its use.
12228 Upload system memory frames to hardware surfaces.
12230 The device to upload to must be supplied when the filter is initialised. If
12231 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
12232 option or with the @option{derive_device} option. The input and output devices
12233 must be of different types and compatible - the exact meaning of this is
12234 system-dependent, but typically it means that they must refer to the same
12235 underlying hardware context (for example, refer to the same graphics card).
12237 The following additional parameters are accepted:
12240 @item derive_device @var{type}
12241 Rather than using the device supplied at initialisation, instead derive a new
12242 device of type @var{type} from the device the input frames exist on.
12245 @anchor{hwupload_cuda}
12246 @section hwupload_cuda
12248 Upload system memory frames to a CUDA device.
12250 It accepts the following optional parameters:
12254 The number of the CUDA device to use
12259 Apply a high-quality magnification filter designed for pixel art. This filter
12260 was originally created by Maxim Stepin.
12262 It accepts the following option:
12266 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
12267 @code{hq3x} and @code{4} for @code{hq4x}.
12268 Default is @code{3}.
12272 Stack input videos horizontally.
12274 All streams must be of same pixel format and of same height.
12276 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
12277 to create same output.
12279 The filter accepts the following option:
12283 Set number of input streams. Default is 2.
12286 If set to 1, force the output to terminate when the shortest input
12287 terminates. Default value is 0.
12292 Modify the hue and/or the saturation of the input.
12294 It accepts the following parameters:
12298 Specify the hue angle as a number of degrees. It accepts an expression,
12299 and defaults to "0".
12302 Specify the saturation in the [-10,10] range. It accepts an expression and
12306 Specify the hue angle as a number of radians. It accepts an
12307 expression, and defaults to "0".
12310 Specify the brightness in the [-10,10] range. It accepts an expression and
12314 @option{h} and @option{H} are mutually exclusive, and can't be
12315 specified at the same time.
12317 The @option{b}, @option{h}, @option{H} and @option{s} option values are
12318 expressions containing the following constants:
12322 frame count of the input frame starting from 0
12325 presentation timestamp of the input frame expressed in time base units
12328 frame rate of the input video, NAN if the input frame rate is unknown
12331 timestamp expressed in seconds, NAN if the input timestamp is unknown
12334 time base of the input video
12337 @subsection Examples
12341 Set the hue to 90 degrees and the saturation to 1.0:
12347 Same command but expressing the hue in radians:
12353 Rotate hue and make the saturation swing between 0
12354 and 2 over a period of 1 second:
12356 hue="H=2*PI*t: s=sin(2*PI*t)+1"
12360 Apply a 3 seconds saturation fade-in effect starting at 0:
12362 hue="s=min(t/3\,1)"
12365 The general fade-in expression can be written as:
12367 hue="s=min(0\, max((t-START)/DURATION\, 1))"
12371 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
12373 hue="s=max(0\, min(1\, (8-t)/3))"
12376 The general fade-out expression can be written as:
12378 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
12383 @subsection Commands
12385 This filter supports the following commands:
12391 Modify the hue and/or the saturation and/or brightness of the input video.
12392 The command accepts the same syntax of the corresponding option.
12394 If the specified expression is not valid, it is kept at its current
12398 @section hysteresis
12400 Grow first stream into second stream by connecting components.
12401 This makes it possible to build more robust edge masks.
12403 This filter accepts the following options:
12407 Set which planes will be processed as bitmap, unprocessed planes will be
12408 copied from first stream.
12409 By default value 0xf, all planes will be processed.
12412 Set threshold which is used in filtering. If pixel component value is higher than
12413 this value filter algorithm for connecting components is activated.
12414 By default value is 0.
12417 The @code{hysteresis} filter also supports the @ref{framesync} options.
12421 Detect video interlacing type.
12423 This filter tries to detect if the input frames are interlaced, progressive,
12424 top or bottom field first. It will also try to detect fields that are
12425 repeated between adjacent frames (a sign of telecine).
12427 Single frame detection considers only immediately adjacent frames when classifying each frame.
12428 Multiple frame detection incorporates the classification history of previous frames.
12430 The filter will log these metadata values:
12433 @item single.current_frame
12434 Detected type of current frame using single-frame detection. One of:
12435 ``tff'' (top field first), ``bff'' (bottom field first),
12436 ``progressive'', or ``undetermined''
12439 Cumulative number of frames detected as top field first using single-frame detection.
12442 Cumulative number of frames detected as top field first using multiple-frame detection.
12445 Cumulative number of frames detected as bottom field first using single-frame detection.
12447 @item multiple.current_frame
12448 Detected type of current frame using multiple-frame detection. One of:
12449 ``tff'' (top field first), ``bff'' (bottom field first),
12450 ``progressive'', or ``undetermined''
12453 Cumulative number of frames detected as bottom field first using multiple-frame detection.
12455 @item single.progressive
12456 Cumulative number of frames detected as progressive using single-frame detection.
12458 @item multiple.progressive
12459 Cumulative number of frames detected as progressive using multiple-frame detection.
12461 @item single.undetermined
12462 Cumulative number of frames that could not be classified using single-frame detection.
12464 @item multiple.undetermined
12465 Cumulative number of frames that could not be classified using multiple-frame detection.
12467 @item repeated.current_frame
12468 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
12470 @item repeated.neither
12471 Cumulative number of frames with no repeated field.
12474 Cumulative number of frames with the top field repeated from the previous frame's top field.
12476 @item repeated.bottom
12477 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
12480 The filter accepts the following options:
12484 Set interlacing threshold.
12486 Set progressive threshold.
12488 Threshold for repeated field detection.
12490 Number of frames after which a given frame's contribution to the
12491 statistics is halved (i.e., it contributes only 0.5 to its
12492 classification). The default of 0 means that all frames seen are given
12493 full weight of 1.0 forever.
12494 @item analyze_interlaced_flag
12495 When this is not 0 then idet will use the specified number of frames to determine
12496 if the interlaced flag is accurate, it will not count undetermined frames.
12497 If the flag is found to be accurate it will be used without any further
12498 computations, if it is found to be inaccurate it will be cleared without any
12499 further computations. This allows inserting the idet filter as a low computational
12500 method to clean up the interlaced flag
12505 Deinterleave or interleave fields.
12507 This filter allows one to process interlaced images fields without
12508 deinterlacing them. Deinterleaving splits the input frame into 2
12509 fields (so called half pictures). Odd lines are moved to the top
12510 half of the output image, even lines to the bottom half.
12511 You can process (filter) them independently and then re-interleave them.
12513 The filter accepts the following options:
12517 @item chroma_mode, c
12518 @item alpha_mode, a
12519 Available values for @var{luma_mode}, @var{chroma_mode} and
12520 @var{alpha_mode} are:
12526 @item deinterleave, d
12527 Deinterleave fields, placing one above the other.
12529 @item interleave, i
12530 Interleave fields. Reverse the effect of deinterleaving.
12532 Default value is @code{none}.
12534 @item luma_swap, ls
12535 @item chroma_swap, cs
12536 @item alpha_swap, as
12537 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
12540 @subsection Commands
12542 This filter supports the all above options as @ref{commands}.
12546 Apply inflate effect to the video.
12548 This filter replaces the pixel by the local(3x3) average by taking into account
12549 only values higher than the pixel.
12551 It accepts the following options:
12558 Limit the maximum change for each plane, default is 65535.
12559 If 0, plane will remain unchanged.
12562 @subsection Commands
12564 This filter supports the all above options as @ref{commands}.
12568 Simple interlacing filter from progressive contents. This interleaves upper (or
12569 lower) lines from odd frames with lower (or upper) lines from even frames,
12570 halving the frame rate and preserving image height.
12573 Original Original New Frame
12574 Frame 'j' Frame 'j+1' (tff)
12575 ========== =========== ==================
12576 Line 0 --------------------> Frame 'j' Line 0
12577 Line 1 Line 1 ----> Frame 'j+1' Line 1
12578 Line 2 ---------------------> Frame 'j' Line 2
12579 Line 3 Line 3 ----> Frame 'j+1' Line 3
12581 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
12584 It accepts the following optional parameters:
12588 This determines whether the interlaced frame is taken from the even
12589 (tff - default) or odd (bff) lines of the progressive frame.
12592 Vertical lowpass filter to avoid twitter interlacing and
12593 reduce moire patterns.
12597 Disable vertical lowpass filter
12600 Enable linear filter (default)
12603 Enable complex filter. This will slightly less reduce twitter and moire
12604 but better retain detail and subjective sharpness impression.
12611 Deinterlace input video by applying Donald Graft's adaptive kernel
12612 deinterling. Work on interlaced parts of a video to produce
12613 progressive frames.
12615 The description of the accepted parameters follows.
12619 Set the threshold which affects the filter's tolerance when
12620 determining if a pixel line must be processed. It must be an integer
12621 in the range [0,255] and defaults to 10. A value of 0 will result in
12622 applying the process on every pixels.
12625 Paint pixels exceeding the threshold value to white if set to 1.
12629 Set the fields order. Swap fields if set to 1, leave fields alone if
12633 Enable additional sharpening if set to 1. Default is 0.
12636 Enable twoway sharpening if set to 1. Default is 0.
12639 @subsection Examples
12643 Apply default values:
12645 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
12649 Enable additional sharpening:
12655 Paint processed pixels in white:
12663 Slowly update darker pixels.
12665 This filter makes short flashes of light appear longer.
12666 This filter accepts the following options:
12670 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
12673 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
12676 @section lenscorrection
12678 Correct radial lens distortion
12680 This filter can be used to correct for radial distortion as can result from the use
12681 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
12682 one can use tools available for example as part of opencv or simply trial-and-error.
12683 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
12684 and extract the k1 and k2 coefficients from the resulting matrix.
12686 Note that effectively the same filter is available in the open-source tools Krita and
12687 Digikam from the KDE project.
12689 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
12690 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
12691 brightness distribution, so you may want to use both filters together in certain
12692 cases, though you will have to take care of ordering, i.e. whether vignetting should
12693 be applied before or after lens correction.
12695 @subsection Options
12697 The filter accepts the following options:
12701 Relative x-coordinate of the focal point of the image, and thereby the center of the
12702 distortion. This value has a range [0,1] and is expressed as fractions of the image
12703 width. Default is 0.5.
12705 Relative y-coordinate of the focal point of the image, and thereby the center of the
12706 distortion. This value has a range [0,1] and is expressed as fractions of the image
12707 height. Default is 0.5.
12709 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
12710 no correction. Default is 0.
12712 Coefficient of the double quadratic correction term. This value has a range [-1,1].
12713 0 means no correction. Default is 0.
12716 The formula that generates the correction is:
12718 @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)
12720 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
12721 distances from the focal point in the source and target images, respectively.
12725 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
12727 The @code{lensfun} filter requires the camera make, camera model, and lens model
12728 to apply the lens correction. The filter will load the lensfun database and
12729 query it to find the corresponding camera and lens entries in the database. As
12730 long as these entries can be found with the given options, the filter can
12731 perform corrections on frames. Note that incomplete strings will result in the
12732 filter choosing the best match with the given options, and the filter will
12733 output the chosen camera and lens models (logged with level "info"). You must
12734 provide the make, camera model, and lens model as they are required.
12736 The filter accepts the following options:
12740 The make of the camera (for example, "Canon"). This option is required.
12743 The model of the camera (for example, "Canon EOS 100D"). This option is
12747 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
12748 option is required.
12751 The type of correction to apply. The following values are valid options:
12755 Enables fixing lens vignetting.
12758 Enables fixing lens geometry. This is the default.
12761 Enables fixing chromatic aberrations.
12764 Enables fixing lens vignetting and lens geometry.
12767 Enables fixing lens vignetting and chromatic aberrations.
12770 Enables fixing both lens geometry and chromatic aberrations.
12773 Enables all possible corrections.
12777 The focal length of the image/video (zoom; expected constant for video). For
12778 example, a 18--55mm lens has focal length range of [18--55], so a value in that
12779 range should be chosen when using that lens. Default 18.
12782 The aperture of the image/video (expected constant for video). Note that
12783 aperture is only used for vignetting correction. Default 3.5.
12785 @item focus_distance
12786 The focus distance of the image/video (expected constant for video). Note that
12787 focus distance is only used for vignetting and only slightly affects the
12788 vignetting correction process. If unknown, leave it at the default value (which
12792 The scale factor which is applied after transformation. After correction the
12793 video is no longer necessarily rectangular. This parameter controls how much of
12794 the resulting image is visible. The value 0 means that a value will be chosen
12795 automatically such that there is little or no unmapped area in the output
12796 image. 1.0 means that no additional scaling is done. Lower values may result
12797 in more of the corrected image being visible, while higher values may avoid
12798 unmapped areas in the output.
12800 @item target_geometry
12801 The target geometry of the output image/video. The following values are valid
12805 @item rectilinear (default)
12808 @item equirectangular
12809 @item fisheye_orthographic
12810 @item fisheye_stereographic
12811 @item fisheye_equisolid
12812 @item fisheye_thoby
12815 Apply the reverse of image correction (instead of correcting distortion, apply
12818 @item interpolation
12819 The type of interpolation used when correcting distortion. The following values
12824 @item linear (default)
12829 @subsection Examples
12833 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
12834 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
12838 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
12842 Apply the same as before, but only for the first 5 seconds of video.
12845 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
12852 Obtain the VMAF (Video Multi-Method Assessment Fusion)
12853 score between two input videos.
12855 The obtained VMAF score is printed through the logging system.
12857 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
12858 After installing the library it can be enabled using:
12859 @code{./configure --enable-libvmaf}.
12860 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
12862 The filter has following options:
12866 Set the model path which is to be used for SVM.
12867 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
12870 Set the file path to be used to store logs.
12873 Set the format of the log file (xml or json).
12875 @item enable_transform
12876 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
12877 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
12878 Default value: @code{false}
12881 Invokes the phone model which will generate VMAF scores higher than in the
12882 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
12883 Default value: @code{false}
12886 Enables computing psnr along with vmaf.
12887 Default value: @code{false}
12890 Enables computing ssim along with vmaf.
12891 Default value: @code{false}
12894 Enables computing ms_ssim along with vmaf.
12895 Default value: @code{false}
12898 Set the pool method to be used for computing vmaf.
12899 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
12902 Set number of threads to be used when computing vmaf.
12903 Default value: @code{0}, which makes use of all available logical processors.
12906 Set interval for frame subsampling used when computing vmaf.
12907 Default value: @code{1}
12909 @item enable_conf_interval
12910 Enables confidence interval.
12911 Default value: @code{false}
12914 This filter also supports the @ref{framesync} options.
12916 @subsection Examples
12919 On the below examples the input file @file{main.mpg} being processed is
12920 compared with the reference file @file{ref.mpg}.
12923 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
12927 Example with options:
12929 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
12933 Example with options and different containers:
12935 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]libvmaf=psnr=1:log_fmt=json" -f null -
12941 Limits the pixel components values to the specified range [min, max].
12943 The filter accepts the following options:
12947 Lower bound. Defaults to the lowest allowed value for the input.
12950 Upper bound. Defaults to the highest allowed value for the input.
12953 Specify which planes will be processed. Defaults to all available.
12960 The filter accepts the following options:
12964 Set the number of loops. Setting this value to -1 will result in infinite loops.
12968 Set maximal size in number of frames. Default is 0.
12971 Set first frame of loop. Default is 0.
12974 @subsection Examples
12978 Loop single first frame infinitely:
12980 loop=loop=-1:size=1:start=0
12984 Loop single first frame 10 times:
12986 loop=loop=10:size=1:start=0
12990 Loop 10 first frames 5 times:
12992 loop=loop=5:size=10:start=0
12998 Apply a 1D LUT to an input video.
13000 The filter accepts the following options:
13004 Set the 1D LUT file name.
13006 Currently supported formats:
13015 Select interpolation mode.
13017 Available values are:
13021 Use values from the nearest defined point.
13023 Interpolate values using the linear interpolation.
13025 Interpolate values using the cosine interpolation.
13027 Interpolate values using the cubic interpolation.
13029 Interpolate values using the spline interpolation.
13036 Apply a 3D LUT to an input video.
13038 The filter accepts the following options:
13042 Set the 3D LUT file name.
13044 Currently supported formats:
13058 Select interpolation mode.
13060 Available values are:
13064 Use values from the nearest defined point.
13066 Interpolate values using the 8 points defining a cube.
13068 Interpolate values using a tetrahedron.
13074 Turn certain luma values into transparency.
13076 The filter accepts the following options:
13080 Set the luma which will be used as base for transparency.
13081 Default value is @code{0}.
13084 Set the range of luma values to be keyed out.
13085 Default value is @code{0.01}.
13088 Set the range of softness. Default value is @code{0}.
13089 Use this to control gradual transition from zero to full transparency.
13092 @subsection Commands
13093 This filter supports same @ref{commands} as options.
13094 The command accepts the same syntax of the corresponding option.
13096 If the specified expression is not valid, it is kept at its current
13099 @section lut, lutrgb, lutyuv
13101 Compute a look-up table for binding each pixel component input value
13102 to an output value, and apply it to the input video.
13104 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
13105 to an RGB input video.
13107 These filters accept the following parameters:
13110 set first pixel component expression
13112 set second pixel component expression
13114 set third pixel component expression
13116 set fourth pixel component expression, corresponds to the alpha component
13119 set red component expression
13121 set green component expression
13123 set blue component expression
13125 alpha component expression
13128 set Y/luminance component expression
13130 set U/Cb component expression
13132 set V/Cr component expression
13135 Each of them specifies the expression to use for computing the lookup table for
13136 the corresponding pixel component values.
13138 The exact component associated to each of the @var{c*} options depends on the
13141 The @var{lut} filter requires either YUV or RGB pixel formats in input,
13142 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
13144 The expressions can contain the following constants and functions:
13149 The input width and height.
13152 The input value for the pixel component.
13155 The input value, clipped to the @var{minval}-@var{maxval} range.
13158 The maximum value for the pixel component.
13161 The minimum value for the pixel component.
13164 The negated value for the pixel component value, clipped to the
13165 @var{minval}-@var{maxval} range; it corresponds to the expression
13166 "maxval-clipval+minval".
13169 The computed value in @var{val}, clipped to the
13170 @var{minval}-@var{maxval} range.
13172 @item gammaval(gamma)
13173 The computed gamma correction value of the pixel component value,
13174 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
13176 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
13180 All expressions default to "val".
13182 @subsection Examples
13186 Negate input video:
13188 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
13189 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
13192 The above is the same as:
13194 lutrgb="r=negval:g=negval:b=negval"
13195 lutyuv="y=negval:u=negval:v=negval"
13205 Remove chroma components, turning the video into a graytone image:
13207 lutyuv="u=128:v=128"
13211 Apply a luma burning effect:
13217 Remove green and blue components:
13223 Set a constant alpha channel value on input:
13225 format=rgba,lutrgb=a="maxval-minval/2"
13229 Correct luminance gamma by a factor of 0.5:
13231 lutyuv=y=gammaval(0.5)
13235 Discard least significant bits of luma:
13237 lutyuv=y='bitand(val, 128+64+32)'
13241 Technicolor like effect:
13243 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
13247 @section lut2, tlut2
13249 The @code{lut2} filter takes two input streams and outputs one
13252 The @code{tlut2} (time lut2) filter takes two consecutive frames
13253 from one single stream.
13255 This filter accepts the following parameters:
13258 set first pixel component expression
13260 set second pixel component expression
13262 set third pixel component expression
13264 set fourth pixel component expression, corresponds to the alpha component
13267 set output bit depth, only available for @code{lut2} filter. By default is 0,
13268 which means bit depth is automatically picked from first input format.
13271 The @code{lut2} filter also supports the @ref{framesync} options.
13273 Each of them specifies the expression to use for computing the lookup table for
13274 the corresponding pixel component values.
13276 The exact component associated to each of the @var{c*} options depends on the
13279 The expressions can contain the following constants:
13284 The input width and height.
13287 The first input value for the pixel component.
13290 The second input value for the pixel component.
13293 The first input video bit depth.
13296 The second input video bit depth.
13299 All expressions default to "x".
13301 @subsection Examples
13305 Highlight differences between two RGB video streams:
13307 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)'
13311 Highlight differences between two YUV video streams:
13313 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)'
13317 Show max difference between two video streams:
13319 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)))'
13323 @section maskedclamp
13325 Clamp the first input stream with the second input and third input stream.
13327 Returns the value of first stream to be between second input
13328 stream - @code{undershoot} and third input stream + @code{overshoot}.
13330 This filter accepts the following options:
13333 Default value is @code{0}.
13336 Default value is @code{0}.
13339 Set which planes will be processed as bitmap, unprocessed planes will be
13340 copied from first stream.
13341 By default value 0xf, all planes will be processed.
13346 Merge the second and third input stream into output stream using absolute differences
13347 between second input stream and first input stream and absolute difference between
13348 third input stream and first input stream. The picked value will be from second input
13349 stream if second absolute difference is greater than first one or from third input stream
13352 This filter accepts the following options:
13355 Set which planes will be processed as bitmap, unprocessed planes will be
13356 copied from first stream.
13357 By default value 0xf, all planes will be processed.
13360 @section maskedmerge
13362 Merge the first input stream with the second input stream using per pixel
13363 weights in the third input stream.
13365 A value of 0 in the third stream pixel component means that pixel component
13366 from first stream is returned unchanged, while maximum value (eg. 255 for
13367 8-bit videos) means that pixel component from second stream is returned
13368 unchanged. Intermediate values define the amount of merging between both
13369 input stream's pixel components.
13371 This filter accepts the following options:
13374 Set which planes will be processed as bitmap, unprocessed planes will be
13375 copied from first stream.
13376 By default value 0xf, all planes will be processed.
13381 Merge the second and third input stream into output stream using absolute differences
13382 between second input stream and first input stream and absolute difference between
13383 third input stream and first input stream. The picked value will be from second input
13384 stream if second absolute difference is less than first one or from third input stream
13387 This filter accepts the following options:
13390 Set which planes will be processed as bitmap, unprocessed planes will be
13391 copied from first stream.
13392 By default value 0xf, all planes will be processed.
13395 @section maskedthreshold
13396 Pick pixels comparing absolute difference of two video streams with fixed
13399 If absolute difference between pixel component of first and second video
13400 stream is equal or lower than user supplied threshold than pixel component
13401 from first video stream is picked, otherwise pixel component from second
13402 video stream is picked.
13404 This filter accepts the following options:
13407 Set threshold used when picking pixels from absolute difference from two input
13411 Set which planes will be processed as bitmap, unprocessed planes will be
13412 copied from second stream.
13413 By default value 0xf, all planes will be processed.
13417 Create mask from input video.
13419 For example it is useful to create motion masks after @code{tblend} filter.
13421 This filter accepts the following options:
13425 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
13428 Set high threshold. Any pixel component higher than this value will be set to max value
13429 allowed for current pixel format.
13432 Set planes to filter, by default all available planes are filtered.
13435 Fill all frame pixels with this value.
13438 Set max average pixel value for frame. If sum of all pixel components is higher that this
13439 average, output frame will be completely filled with value set by @var{fill} option.
13440 Typically useful for scene changes when used in combination with @code{tblend} filter.
13445 Apply motion-compensation deinterlacing.
13447 It needs one field per frame as input and must thus be used together
13448 with yadif=1/3 or equivalent.
13450 This filter accepts the following options:
13453 Set the deinterlacing mode.
13455 It accepts one of the following values:
13460 use iterative motion estimation
13462 like @samp{slow}, but use multiple reference frames.
13464 Default value is @samp{fast}.
13467 Set the picture field parity assumed for the input video. It must be
13468 one of the following values:
13472 assume top field first
13474 assume bottom field first
13477 Default value is @samp{bff}.
13480 Set per-block quantization parameter (QP) used by the internal
13483 Higher values should result in a smoother motion vector field but less
13484 optimal individual vectors. Default value is 1.
13489 Pick median pixel from certain rectangle defined by radius.
13491 This filter accepts the following options:
13495 Set horizontal radius size. Default value is @code{1}.
13496 Allowed range is integer from 1 to 127.
13499 Set which planes to process. Default is @code{15}, which is all available planes.
13502 Set vertical radius size. Default value is @code{0}.
13503 Allowed range is integer from 0 to 127.
13504 If it is 0, value will be picked from horizontal @code{radius} option.
13507 Set median percentile. Default value is @code{0.5}.
13508 Default value of @code{0.5} will pick always median values, while @code{0} will pick
13509 minimum values, and @code{1} maximum values.
13512 @subsection Commands
13513 This filter supports same @ref{commands} as options.
13514 The command accepts the same syntax of the corresponding option.
13516 If the specified expression is not valid, it is kept at its current
13519 @section mergeplanes
13521 Merge color channel components from several video streams.
13523 The filter accepts up to 4 input streams, and merge selected input
13524 planes to the output video.
13526 This filter accepts the following options:
13529 Set input to output plane mapping. Default is @code{0}.
13531 The mappings is specified as a bitmap. It should be specified as a
13532 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
13533 mapping for the first plane of the output stream. 'A' sets the number of
13534 the input stream to use (from 0 to 3), and 'a' the plane number of the
13535 corresponding input to use (from 0 to 3). The rest of the mappings is
13536 similar, 'Bb' describes the mapping for the output stream second
13537 plane, 'Cc' describes the mapping for the output stream third plane and
13538 'Dd' describes the mapping for the output stream fourth plane.
13541 Set output pixel format. Default is @code{yuva444p}.
13544 @subsection Examples
13548 Merge three gray video streams of same width and height into single video stream:
13550 [a0][a1][a2]mergeplanes=0x001020:yuv444p
13554 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
13556 [a0][a1]mergeplanes=0x00010210:yuva444p
13560 Swap Y and A plane in yuva444p stream:
13562 format=yuva444p,mergeplanes=0x03010200:yuva444p
13566 Swap U and V plane in yuv420p stream:
13568 format=yuv420p,mergeplanes=0x000201:yuv420p
13572 Cast a rgb24 clip to yuv444p:
13574 format=rgb24,mergeplanes=0x000102:yuv444p
13580 Estimate and export motion vectors using block matching algorithms.
13581 Motion vectors are stored in frame side data to be used by other filters.
13583 This filter accepts the following options:
13586 Specify the motion estimation method. Accepts one of the following values:
13590 Exhaustive search algorithm.
13592 Three step search algorithm.
13594 Two dimensional logarithmic search algorithm.
13596 New three step search algorithm.
13598 Four step search algorithm.
13600 Diamond search algorithm.
13602 Hexagon-based search algorithm.
13604 Enhanced predictive zonal search algorithm.
13606 Uneven multi-hexagon search algorithm.
13608 Default value is @samp{esa}.
13611 Macroblock size. Default @code{16}.
13614 Search parameter. Default @code{7}.
13617 @section midequalizer
13619 Apply Midway Image Equalization effect using two video streams.
13621 Midway Image Equalization adjusts a pair of images to have the same
13622 histogram, while maintaining their dynamics as much as possible. It's
13623 useful for e.g. matching exposures from a pair of stereo cameras.
13625 This filter has two inputs and one output, which must be of same pixel format, but
13626 may be of different sizes. The output of filter is first input adjusted with
13627 midway histogram of both inputs.
13629 This filter accepts the following option:
13633 Set which planes to process. Default is @code{15}, which is all available planes.
13636 @section minterpolate
13638 Convert the video to specified frame rate using motion interpolation.
13640 This filter accepts the following options:
13643 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}.
13646 Motion interpolation mode. Following values are accepted:
13649 Duplicate previous or next frame for interpolating new ones.
13651 Blend source frames. Interpolated frame is mean of previous and next frames.
13653 Motion compensated interpolation. Following options are effective when this mode is selected:
13657 Motion compensation mode. Following values are accepted:
13660 Overlapped block motion compensation.
13662 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
13664 Default mode is @samp{obmc}.
13667 Motion estimation mode. Following values are accepted:
13670 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
13672 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
13674 Default mode is @samp{bilat}.
13677 The algorithm to be used for motion estimation. Following values are accepted:
13680 Exhaustive search algorithm.
13682 Three step search algorithm.
13684 Two dimensional logarithmic search algorithm.
13686 New three step search algorithm.
13688 Four step search algorithm.
13690 Diamond search algorithm.
13692 Hexagon-based search algorithm.
13694 Enhanced predictive zonal search algorithm.
13696 Uneven multi-hexagon search algorithm.
13698 Default algorithm is @samp{epzs}.
13701 Macroblock size. Default @code{16}.
13704 Motion estimation search parameter. Default @code{32}.
13707 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).
13712 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:
13715 Disable scene change detection.
13717 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
13719 Default method is @samp{fdiff}.
13721 @item scd_threshold
13722 Scene change detection threshold. Default is @code{10.}.
13727 Mix several video input streams into one video stream.
13729 A description of the accepted options follows.
13733 The number of inputs. If unspecified, it defaults to 2.
13736 Specify weight of each input video stream as sequence.
13737 Each weight is separated by space. If number of weights
13738 is smaller than number of @var{frames} last specified
13739 weight will be used for all remaining unset weights.
13742 Specify scale, if it is set it will be multiplied with sum
13743 of each weight multiplied with pixel values to give final destination
13744 pixel value. By default @var{scale} is auto scaled to sum of weights.
13747 Specify how end of stream is determined.
13750 The duration of the longest input. (default)
13753 The duration of the shortest input.
13756 The duration of the first input.
13760 @section mpdecimate
13762 Drop frames that do not differ greatly from the previous frame in
13763 order to reduce frame rate.
13765 The main use of this filter is for very-low-bitrate encoding
13766 (e.g. streaming over dialup modem), but it could in theory be used for
13767 fixing movies that were inverse-telecined incorrectly.
13769 A description of the accepted options follows.
13773 Set the maximum number of consecutive frames which can be dropped (if
13774 positive), or the minimum interval between dropped frames (if
13775 negative). If the value is 0, the frame is dropped disregarding the
13776 number of previous sequentially dropped frames.
13778 Default value is 0.
13783 Set the dropping threshold values.
13785 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
13786 represent actual pixel value differences, so a threshold of 64
13787 corresponds to 1 unit of difference for each pixel, or the same spread
13788 out differently over the block.
13790 A frame is a candidate for dropping if no 8x8 blocks differ by more
13791 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
13792 meaning the whole image) differ by more than a threshold of @option{lo}.
13794 Default value for @option{hi} is 64*12, default value for @option{lo} is
13795 64*5, and default value for @option{frac} is 0.33.
13801 Negate (invert) the input video.
13803 It accepts the following option:
13808 With value 1, it negates the alpha component, if present. Default value is 0.
13814 Denoise frames using Non-Local Means algorithm.
13816 Each pixel is adjusted by looking for other pixels with similar contexts. This
13817 context similarity is defined by comparing their surrounding patches of size
13818 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
13821 Note that the research area defines centers for patches, which means some
13822 patches will be made of pixels outside that research area.
13824 The filter accepts the following options.
13828 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
13831 Set patch size. Default is 7. Must be odd number in range [0, 99].
13834 Same as @option{p} but for chroma planes.
13836 The default value is @var{0} and means automatic.
13839 Set research size. Default is 15. Must be odd number in range [0, 99].
13842 Same as @option{r} but for chroma planes.
13844 The default value is @var{0} and means automatic.
13849 Deinterlace video using neural network edge directed interpolation.
13851 This filter accepts the following options:
13855 Mandatory option, without binary file filter can not work.
13856 Currently file can be found here:
13857 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
13860 Set which frames to deinterlace, by default it is @code{all}.
13861 Can be @code{all} or @code{interlaced}.
13864 Set mode of operation.
13866 Can be one of the following:
13870 Use frame flags, both fields.
13872 Use frame flags, single field.
13874 Use top field only.
13876 Use bottom field only.
13878 Use both fields, top first.
13880 Use both fields, bottom first.
13884 Set which planes to process, by default filter process all frames.
13887 Set size of local neighborhood around each pixel, used by the predictor neural
13890 Can be one of the following:
13903 Set the number of neurons in predictor neural network.
13904 Can be one of the following:
13915 Controls the number of different neural network predictions that are blended
13916 together to compute the final output value. Can be @code{fast}, default or
13920 Set which set of weights to use in the predictor.
13921 Can be one of the following:
13925 weights trained to minimize absolute error
13927 weights trained to minimize squared error
13931 Controls whether or not the prescreener neural network is used to decide
13932 which pixels should be processed by the predictor neural network and which
13933 can be handled by simple cubic interpolation.
13934 The prescreener is trained to know whether cubic interpolation will be
13935 sufficient for a pixel or whether it should be predicted by the predictor nn.
13936 The computational complexity of the prescreener nn is much less than that of
13937 the predictor nn. Since most pixels can be handled by cubic interpolation,
13938 using the prescreener generally results in much faster processing.
13939 The prescreener is pretty accurate, so the difference between using it and not
13940 using it is almost always unnoticeable.
13942 Can be one of the following:
13950 Default is @code{new}.
13953 Set various debugging flags.
13958 Force libavfilter not to use any of the specified pixel formats for the
13959 input to the next filter.
13961 It accepts the following parameters:
13965 A '|'-separated list of pixel format names, such as
13966 pix_fmts=yuv420p|monow|rgb24".
13970 @subsection Examples
13974 Force libavfilter to use a format different from @var{yuv420p} for the
13975 input to the vflip filter:
13977 noformat=pix_fmts=yuv420p,vflip
13981 Convert the input video to any of the formats not contained in the list:
13983 noformat=yuv420p|yuv444p|yuv410p
13989 Add noise on video input frame.
13991 The filter accepts the following options:
13999 Set noise seed for specific pixel component or all pixel components in case
14000 of @var{all_seed}. Default value is @code{123457}.
14002 @item all_strength, alls
14003 @item c0_strength, c0s
14004 @item c1_strength, c1s
14005 @item c2_strength, c2s
14006 @item c3_strength, c3s
14007 Set noise strength for specific pixel component or all pixel components in case
14008 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14010 @item all_flags, allf
14011 @item c0_flags, c0f
14012 @item c1_flags, c1f
14013 @item c2_flags, c2f
14014 @item c3_flags, c3f
14015 Set pixel component flags or set flags for all components if @var{all_flags}.
14016 Available values for component flags are:
14019 averaged temporal noise (smoother)
14021 mix random noise with a (semi)regular pattern
14023 temporal noise (noise pattern changes between frames)
14025 uniform noise (gaussian otherwise)
14029 @subsection Examples
14031 Add temporal and uniform noise to input video:
14033 noise=alls=20:allf=t+u
14038 Normalize RGB video (aka histogram stretching, contrast stretching).
14039 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
14041 For each channel of each frame, the filter computes the input range and maps
14042 it linearly to the user-specified output range. The output range defaults
14043 to the full dynamic range from pure black to pure white.
14045 Temporal smoothing can be used on the input range to reduce flickering (rapid
14046 changes in brightness) caused when small dark or bright objects enter or leave
14047 the scene. This is similar to the auto-exposure (automatic gain control) on a
14048 video camera, and, like a video camera, it may cause a period of over- or
14049 under-exposure of the video.
14051 The R,G,B channels can be normalized independently, which may cause some
14052 color shifting, or linked together as a single channel, which prevents
14053 color shifting. Linked normalization preserves hue. Independent normalization
14054 does not, so it can be used to remove some color casts. Independent and linked
14055 normalization can be combined in any ratio.
14057 The normalize filter accepts the following options:
14062 Colors which define the output range. The minimum input value is mapped to
14063 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
14064 The defaults are black and white respectively. Specifying white for
14065 @var{blackpt} and black for @var{whitept} will give color-inverted,
14066 normalized video. Shades of grey can be used to reduce the dynamic range
14067 (contrast). Specifying saturated colors here can create some interesting
14071 The number of previous frames to use for temporal smoothing. The input range
14072 of each channel is smoothed using a rolling average over the current frame
14073 and the @var{smoothing} previous frames. The default is 0 (no temporal
14077 Controls the ratio of independent (color shifting) channel normalization to
14078 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
14079 independent. Defaults to 1.0 (fully independent).
14082 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
14083 expensive no-op. Defaults to 1.0 (full strength).
14087 @subsection Commands
14088 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
14089 The command accepts the same syntax of the corresponding option.
14091 If the specified expression is not valid, it is kept at its current
14094 @subsection Examples
14096 Stretch video contrast to use the full dynamic range, with no temporal
14097 smoothing; may flicker depending on the source content:
14099 normalize=blackpt=black:whitept=white:smoothing=0
14102 As above, but with 50 frames of temporal smoothing; flicker should be
14103 reduced, depending on the source content:
14105 normalize=blackpt=black:whitept=white:smoothing=50
14108 As above, but with hue-preserving linked channel normalization:
14110 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
14113 As above, but with half strength:
14115 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
14118 Map the darkest input color to red, the brightest input color to cyan:
14120 normalize=blackpt=red:whitept=cyan
14125 Pass the video source unchanged to the output.
14128 Optical Character Recognition
14130 This filter uses Tesseract for optical character recognition. To enable
14131 compilation of this filter, you need to configure FFmpeg with
14132 @code{--enable-libtesseract}.
14134 It accepts the following options:
14138 Set datapath to tesseract data. Default is to use whatever was
14139 set at installation.
14142 Set language, default is "eng".
14145 Set character whitelist.
14148 Set character blacklist.
14151 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
14152 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
14156 Apply a video transform using libopencv.
14158 To enable this filter, install the libopencv library and headers and
14159 configure FFmpeg with @code{--enable-libopencv}.
14161 It accepts the following parameters:
14166 The name of the libopencv filter to apply.
14168 @item filter_params
14169 The parameters to pass to the libopencv filter. If not specified, the default
14170 values are assumed.
14174 Refer to the official libopencv documentation for more precise
14176 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
14178 Several libopencv filters are supported; see the following subsections.
14183 Dilate an image by using a specific structuring element.
14184 It corresponds to the libopencv function @code{cvDilate}.
14186 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
14188 @var{struct_el} represents a structuring element, and has the syntax:
14189 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
14191 @var{cols} and @var{rows} represent the number of columns and rows of
14192 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
14193 point, and @var{shape} the shape for the structuring element. @var{shape}
14194 must be "rect", "cross", "ellipse", or "custom".
14196 If the value for @var{shape} is "custom", it must be followed by a
14197 string of the form "=@var{filename}". The file with name
14198 @var{filename} is assumed to represent a binary image, with each
14199 printable character corresponding to a bright pixel. When a custom
14200 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
14201 or columns and rows of the read file are assumed instead.
14203 The default value for @var{struct_el} is "3x3+0x0/rect".
14205 @var{nb_iterations} specifies the number of times the transform is
14206 applied to the image, and defaults to 1.
14210 # Use the default values
14213 # Dilate using a structuring element with a 5x5 cross, iterating two times
14214 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
14216 # Read the shape from the file diamond.shape, iterating two times.
14217 # The file diamond.shape may contain a pattern of characters like this
14223 # The specified columns and rows are ignored
14224 # but the anchor point coordinates are not
14225 ocv=dilate:0x0+2x2/custom=diamond.shape|2
14230 Erode an image by using a specific structuring element.
14231 It corresponds to the libopencv function @code{cvErode}.
14233 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
14234 with the same syntax and semantics as the @ref{dilate} filter.
14238 Smooth the input video.
14240 The filter takes the following parameters:
14241 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
14243 @var{type} is the type of smooth filter to apply, and must be one of
14244 the following values: "blur", "blur_no_scale", "median", "gaussian",
14245 or "bilateral". The default value is "gaussian".
14247 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
14248 depends on the smooth type. @var{param1} and
14249 @var{param2} accept integer positive values or 0. @var{param3} and
14250 @var{param4} accept floating point values.
14252 The default value for @var{param1} is 3. The default value for the
14253 other parameters is 0.
14255 These parameters correspond to the parameters assigned to the
14256 libopencv function @code{cvSmooth}.
14258 @section oscilloscope
14260 2D Video Oscilloscope.
14262 Useful to measure spatial impulse, step responses, chroma delays, etc.
14264 It accepts the following parameters:
14268 Set scope center x position.
14271 Set scope center y position.
14274 Set scope size, relative to frame diagonal.
14277 Set scope tilt/rotation.
14283 Set trace center x position.
14286 Set trace center y position.
14289 Set trace width, relative to width of frame.
14292 Set trace height, relative to height of frame.
14295 Set which components to trace. By default it traces first three components.
14298 Draw trace grid. By default is enabled.
14301 Draw some statistics. By default is enabled.
14304 Draw scope. By default is enabled.
14307 @subsection Commands
14308 This filter supports same @ref{commands} as options.
14309 The command accepts the same syntax of the corresponding option.
14311 If the specified expression is not valid, it is kept at its current
14314 @subsection Examples
14318 Inspect full first row of video frame.
14320 oscilloscope=x=0.5:y=0:s=1
14324 Inspect full last row of video frame.
14326 oscilloscope=x=0.5:y=1:s=1
14330 Inspect full 5th line of video frame of height 1080.
14332 oscilloscope=x=0.5:y=5/1080:s=1
14336 Inspect full last column of video frame.
14338 oscilloscope=x=1:y=0.5:s=1:t=1
14346 Overlay one video on top of another.
14348 It takes two inputs and has one output. The first input is the "main"
14349 video on which the second input is overlaid.
14351 It accepts the following parameters:
14353 A description of the accepted options follows.
14358 Set the expression for the x and y coordinates of the overlaid video
14359 on the main video. Default value is "0" for both expressions. In case
14360 the expression is invalid, it is set to a huge value (meaning that the
14361 overlay will not be displayed within the output visible area).
14364 See @ref{framesync}.
14367 Set when the expressions for @option{x}, and @option{y} are evaluated.
14369 It accepts the following values:
14372 only evaluate expressions once during the filter initialization or
14373 when a command is processed
14376 evaluate expressions for each incoming frame
14379 Default value is @samp{frame}.
14382 See @ref{framesync}.
14385 Set the format for the output video.
14387 It accepts the following values:
14390 force YUV420 output
14393 force YUV420p10 output
14396 force YUV422 output
14399 force YUV422p10 output
14402 force YUV444 output
14405 force packed RGB output
14408 force planar RGB output
14411 automatically pick format
14414 Default value is @samp{yuv420}.
14417 See @ref{framesync}.
14420 Set format of alpha of the overlaid video, it can be @var{straight} or
14421 @var{premultiplied}. Default is @var{straight}.
14424 The @option{x}, and @option{y} expressions can contain the following
14430 The main input width and height.
14434 The overlay input width and height.
14438 The computed values for @var{x} and @var{y}. They are evaluated for
14443 horizontal and vertical chroma subsample values of the output
14444 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
14448 the number of input frame, starting from 0
14451 the position in the file of the input frame, NAN if unknown
14454 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
14458 This filter also supports the @ref{framesync} options.
14460 Note that the @var{n}, @var{pos}, @var{t} variables are available only
14461 when evaluation is done @emph{per frame}, and will evaluate to NAN
14462 when @option{eval} is set to @samp{init}.
14464 Be aware that frames are taken from each input video in timestamp
14465 order, hence, if their initial timestamps differ, it is a good idea
14466 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
14467 have them begin in the same zero timestamp, as the example for
14468 the @var{movie} filter does.
14470 You can chain together more overlays but you should test the
14471 efficiency of such approach.
14473 @subsection Commands
14475 This filter supports the following commands:
14479 Modify the x and y of the overlay input.
14480 The command accepts the same syntax of the corresponding option.
14482 If the specified expression is not valid, it is kept at its current
14486 @subsection Examples
14490 Draw the overlay at 10 pixels from the bottom right corner of the main
14493 overlay=main_w-overlay_w-10:main_h-overlay_h-10
14496 Using named options the example above becomes:
14498 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
14502 Insert a transparent PNG logo in the bottom left corner of the input,
14503 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
14505 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
14509 Insert 2 different transparent PNG logos (second logo on bottom
14510 right corner) using the @command{ffmpeg} tool:
14512 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
14516 Add a transparent color layer on top of the main video; @code{WxH}
14517 must specify the size of the main input to the overlay filter:
14519 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
14523 Play an original video and a filtered version (here with the deshake
14524 filter) side by side using the @command{ffplay} tool:
14526 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
14529 The above command is the same as:
14531 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
14535 Make a sliding overlay appearing from the left to the right top part of the
14536 screen starting since time 2:
14538 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
14542 Compose output by putting two input videos side to side:
14544 ffmpeg -i left.avi -i right.avi -filter_complex "
14545 nullsrc=size=200x100 [background];
14546 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
14547 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
14548 [background][left] overlay=shortest=1 [background+left];
14549 [background+left][right] overlay=shortest=1:x=100 [left+right]
14554 Mask 10-20 seconds of a video by applying the delogo filter to a section
14556 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
14557 -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]'
14562 Chain several overlays in cascade:
14564 nullsrc=s=200x200 [bg];
14565 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
14566 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
14567 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
14568 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
14569 [in3] null, [mid2] overlay=100:100 [out0]
14574 @anchor{overlay_cuda}
14575 @section overlay_cuda
14577 Overlay one video on top of another.
14579 This is the CUDA cariant of the @ref{overlay} filter.
14580 It only accepts CUDA frames. The underlying input pixel formats have to match.
14582 It takes two inputs and has one output. The first input is the "main"
14583 video on which the second input is overlaid.
14585 It accepts the following parameters:
14590 Set the x and y coordinates of the overlaid video on the main video.
14591 Default value is "0" for both expressions.
14594 See @ref{framesync}.
14597 See @ref{framesync}.
14600 See @ref{framesync}.
14604 This filter also supports the @ref{framesync} options.
14608 Apply Overcomplete Wavelet denoiser.
14610 The filter accepts the following options:
14616 Larger depth values will denoise lower frequency components more, but
14617 slow down filtering.
14619 Must be an int in the range 8-16, default is @code{8}.
14621 @item luma_strength, ls
14624 Must be a double value in the range 0-1000, default is @code{1.0}.
14626 @item chroma_strength, cs
14627 Set chroma strength.
14629 Must be a double value in the range 0-1000, default is @code{1.0}.
14635 Add paddings to the input image, and place the original input at the
14636 provided @var{x}, @var{y} coordinates.
14638 It accepts the following parameters:
14643 Specify an expression for the size of the output image with the
14644 paddings added. If the value for @var{width} or @var{height} is 0, the
14645 corresponding input size is used for the output.
14647 The @var{width} expression can reference the value set by the
14648 @var{height} expression, and vice versa.
14650 The default value of @var{width} and @var{height} is 0.
14654 Specify the offsets to place the input image at within the padded area,
14655 with respect to the top/left border of the output image.
14657 The @var{x} expression can reference the value set by the @var{y}
14658 expression, and vice versa.
14660 The default value of @var{x} and @var{y} is 0.
14662 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
14663 so the input image is centered on the padded area.
14666 Specify the color of the padded area. For the syntax of this option,
14667 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
14668 manual,ffmpeg-utils}.
14670 The default value of @var{color} is "black".
14673 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
14675 It accepts the following values:
14679 Only evaluate expressions once during the filter initialization or when
14680 a command is processed.
14683 Evaluate expressions for each incoming frame.
14687 Default value is @samp{init}.
14690 Pad to aspect instead to a resolution.
14694 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
14695 options are expressions containing the following constants:
14700 The input video width and height.
14704 These are the same as @var{in_w} and @var{in_h}.
14708 The output width and height (the size of the padded area), as
14709 specified by the @var{width} and @var{height} expressions.
14713 These are the same as @var{out_w} and @var{out_h}.
14717 The x and y offsets as specified by the @var{x} and @var{y}
14718 expressions, or NAN if not yet specified.
14721 same as @var{iw} / @var{ih}
14724 input sample aspect ratio
14727 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
14731 The horizontal and vertical chroma subsample values. For example for the
14732 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14735 @subsection Examples
14739 Add paddings with the color "violet" to the input video. The output video
14740 size is 640x480, and the top-left corner of the input video is placed at
14743 pad=640:480:0:40:violet
14746 The example above is equivalent to the following command:
14748 pad=width=640:height=480:x=0:y=40:color=violet
14752 Pad the input to get an output with dimensions increased by 3/2,
14753 and put the input video at the center of the padded area:
14755 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
14759 Pad the input to get a squared output with size equal to the maximum
14760 value between the input width and height, and put the input video at
14761 the center of the padded area:
14763 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
14767 Pad the input to get a final w/h ratio of 16:9:
14769 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
14773 In case of anamorphic video, in order to set the output display aspect
14774 correctly, it is necessary to use @var{sar} in the expression,
14775 according to the relation:
14777 (ih * X / ih) * sar = output_dar
14778 X = output_dar / sar
14781 Thus the previous example needs to be modified to:
14783 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
14787 Double the output size and put the input video in the bottom-right
14788 corner of the output padded area:
14790 pad="2*iw:2*ih:ow-iw:oh-ih"
14794 @anchor{palettegen}
14795 @section palettegen
14797 Generate one palette for a whole video stream.
14799 It accepts the following options:
14803 Set the maximum number of colors to quantize in the palette.
14804 Note: the palette will still contain 256 colors; the unused palette entries
14807 @item reserve_transparent
14808 Create a palette of 255 colors maximum and reserve the last one for
14809 transparency. Reserving the transparency color is useful for GIF optimization.
14810 If not set, the maximum of colors in the palette will be 256. You probably want
14811 to disable this option for a standalone image.
14814 @item transparency_color
14815 Set the color that will be used as background for transparency.
14818 Set statistics mode.
14820 It accepts the following values:
14823 Compute full frame histograms.
14825 Compute histograms only for the part that differs from previous frame. This
14826 might be relevant to give more importance to the moving part of your input if
14827 the background is static.
14829 Compute new histogram for each frame.
14832 Default value is @var{full}.
14835 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
14836 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
14837 color quantization of the palette. This information is also visible at
14838 @var{info} logging level.
14840 @subsection Examples
14844 Generate a representative palette of a given video using @command{ffmpeg}:
14846 ffmpeg -i input.mkv -vf palettegen palette.png
14850 @section paletteuse
14852 Use a palette to downsample an input video stream.
14854 The filter takes two inputs: one video stream and a palette. The palette must
14855 be a 256 pixels image.
14857 It accepts the following options:
14861 Select dithering mode. Available algorithms are:
14864 Ordered 8x8 bayer dithering (deterministic)
14866 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
14867 Note: this dithering is sometimes considered "wrong" and is included as a
14869 @item floyd_steinberg
14870 Floyd and Steingberg dithering (error diffusion)
14872 Frankie Sierra dithering v2 (error diffusion)
14874 Frankie Sierra dithering v2 "Lite" (error diffusion)
14877 Default is @var{sierra2_4a}.
14880 When @var{bayer} dithering is selected, this option defines the scale of the
14881 pattern (how much the crosshatch pattern is visible). A low value means more
14882 visible pattern for less banding, and higher value means less visible pattern
14883 at the cost of more banding.
14885 The option must be an integer value in the range [0,5]. Default is @var{2}.
14888 If set, define the zone to process
14892 Only the changing rectangle will be reprocessed. This is similar to GIF
14893 cropping/offsetting compression mechanism. This option can be useful for speed
14894 if only a part of the image is changing, and has use cases such as limiting the
14895 scope of the error diffusal @option{dither} to the rectangle that bounds the
14896 moving scene (it leads to more deterministic output if the scene doesn't change
14897 much, and as a result less moving noise and better GIF compression).
14900 Default is @var{none}.
14903 Take new palette for each output frame.
14905 @item alpha_threshold
14906 Sets the alpha threshold for transparency. Alpha values above this threshold
14907 will be treated as completely opaque, and values below this threshold will be
14908 treated as completely transparent.
14910 The option must be an integer value in the range [0,255]. Default is @var{128}.
14913 @subsection Examples
14917 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
14918 using @command{ffmpeg}:
14920 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
14924 @section perspective
14926 Correct perspective of video not recorded perpendicular to the screen.
14928 A description of the accepted parameters follows.
14939 Set coordinates expression for top left, top right, bottom left and bottom right corners.
14940 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
14941 If the @code{sense} option is set to @code{source}, then the specified points will be sent
14942 to the corners of the destination. If the @code{sense} option is set to @code{destination},
14943 then the corners of the source will be sent to the specified coordinates.
14945 The expressions can use the following variables:
14950 the width and height of video frame.
14954 Output frame count.
14957 @item interpolation
14958 Set interpolation for perspective correction.
14960 It accepts the following values:
14966 Default value is @samp{linear}.
14969 Set interpretation of coordinate options.
14971 It accepts the following values:
14975 Send point in the source specified by the given coordinates to
14976 the corners of the destination.
14978 @item 1, destination
14980 Send the corners of the source to the point in the destination specified
14981 by the given coordinates.
14983 Default value is @samp{source}.
14987 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
14989 It accepts the following values:
14992 only evaluate expressions once during the filter initialization or
14993 when a command is processed
14996 evaluate expressions for each incoming frame
14999 Default value is @samp{init}.
15004 Delay interlaced video by one field time so that the field order changes.
15006 The intended use is to fix PAL movies that have been captured with the
15007 opposite field order to the film-to-video transfer.
15009 A description of the accepted parameters follows.
15015 It accepts the following values:
15018 Capture field order top-first, transfer bottom-first.
15019 Filter will delay the bottom field.
15022 Capture field order bottom-first, transfer top-first.
15023 Filter will delay the top field.
15026 Capture and transfer with the same field order. This mode only exists
15027 for the documentation of the other options to refer to, but if you
15028 actually select it, the filter will faithfully do nothing.
15031 Capture field order determined automatically by field flags, transfer
15033 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
15034 basis using field flags. If no field information is available,
15035 then this works just like @samp{u}.
15038 Capture unknown or varying, transfer opposite.
15039 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
15040 analyzing the images and selecting the alternative that produces best
15041 match between the fields.
15044 Capture top-first, transfer unknown or varying.
15045 Filter selects among @samp{t} and @samp{p} using image analysis.
15048 Capture bottom-first, transfer unknown or varying.
15049 Filter selects among @samp{b} and @samp{p} using image analysis.
15052 Capture determined by field flags, transfer unknown or varying.
15053 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
15054 image analysis. If no field information is available, then this works just
15055 like @samp{U}. This is the default mode.
15058 Both capture and transfer unknown or varying.
15059 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
15063 @section photosensitivity
15064 Reduce various flashes in video, so to help users with epilepsy.
15066 It accepts the following options:
15069 Set how many frames to use when filtering. Default is 30.
15072 Set detection threshold factor. Default is 1.
15076 Set how many pixels to skip when sampling frames. Default is 1.
15077 Allowed range is from 1 to 1024.
15080 Leave frames unchanged. Default is disabled.
15083 @section pixdesctest
15085 Pixel format descriptor test filter, mainly useful for internal
15086 testing. The output video should be equal to the input video.
15090 format=monow, pixdesctest
15093 can be used to test the monowhite pixel format descriptor definition.
15097 Display sample values of color channels. Mainly useful for checking color
15098 and levels. Minimum supported resolution is 640x480.
15100 The filters accept the following options:
15104 Set scope X position, relative offset on X axis.
15107 Set scope Y position, relative offset on Y axis.
15116 Set window opacity. This window also holds statistics about pixel area.
15119 Set window X position, relative offset on X axis.
15122 Set window Y position, relative offset on Y axis.
15127 Enable the specified chain of postprocessing subfilters using libpostproc. This
15128 library should be automatically selected with a GPL build (@code{--enable-gpl}).
15129 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
15130 Each subfilter and some options have a short and a long name that can be used
15131 interchangeably, i.e. dr/dering are the same.
15133 The filters accept the following options:
15137 Set postprocessing subfilters string.
15140 All subfilters share common options to determine their scope:
15144 Honor the quality commands for this subfilter.
15147 Do chrominance filtering, too (default).
15150 Do luminance filtering only (no chrominance).
15153 Do chrominance filtering only (no luminance).
15156 These options can be appended after the subfilter name, separated by a '|'.
15158 Available subfilters are:
15161 @item hb/hdeblock[|difference[|flatness]]
15162 Horizontal deblocking filter
15165 Difference factor where higher values mean more deblocking (default: @code{32}).
15167 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15170 @item vb/vdeblock[|difference[|flatness]]
15171 Vertical deblocking filter
15174 Difference factor where higher values mean more deblocking (default: @code{32}).
15176 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15179 @item ha/hadeblock[|difference[|flatness]]
15180 Accurate horizontal deblocking filter
15183 Difference factor where higher values mean more deblocking (default: @code{32}).
15185 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15188 @item va/vadeblock[|difference[|flatness]]
15189 Accurate vertical deblocking filter
15192 Difference factor where higher values mean more deblocking (default: @code{32}).
15194 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15198 The horizontal and vertical deblocking filters share the difference and
15199 flatness values so you cannot set different horizontal and vertical
15203 @item h1/x1hdeblock
15204 Experimental horizontal deblocking filter
15206 @item v1/x1vdeblock
15207 Experimental vertical deblocking filter
15212 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
15215 larger -> stronger filtering
15217 larger -> stronger filtering
15219 larger -> stronger filtering
15222 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
15225 Stretch luminance to @code{0-255}.
15228 @item lb/linblenddeint
15229 Linear blend deinterlacing filter that deinterlaces the given block by
15230 filtering all lines with a @code{(1 2 1)} filter.
15232 @item li/linipoldeint
15233 Linear interpolating deinterlacing filter that deinterlaces the given block by
15234 linearly interpolating every second line.
15236 @item ci/cubicipoldeint
15237 Cubic interpolating deinterlacing filter deinterlaces the given block by
15238 cubically interpolating every second line.
15240 @item md/mediandeint
15241 Median deinterlacing filter that deinterlaces the given block by applying a
15242 median filter to every second line.
15244 @item fd/ffmpegdeint
15245 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
15246 second line with a @code{(-1 4 2 4 -1)} filter.
15249 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
15250 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
15252 @item fq/forceQuant[|quantizer]
15253 Overrides the quantizer table from the input with the constant quantizer you
15261 Default pp filter combination (@code{hb|a,vb|a,dr|a})
15264 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
15267 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
15270 @subsection Examples
15274 Apply horizontal and vertical deblocking, deringing and automatic
15275 brightness/contrast:
15281 Apply default filters without brightness/contrast correction:
15287 Apply default filters and temporal denoiser:
15289 pp=default/tmpnoise|1|2|3
15293 Apply deblocking on luminance only, and switch vertical deblocking on or off
15294 automatically depending on available CPU time:
15301 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
15302 similar to spp = 6 with 7 point DCT, where only the center sample is
15305 The filter accepts the following options:
15309 Force a constant quantization parameter. It accepts an integer in range
15310 0 to 63. If not set, the filter will use the QP from the video stream
15314 Set thresholding mode. Available modes are:
15318 Set hard thresholding.
15320 Set soft thresholding (better de-ringing effect, but likely blurrier).
15322 Set medium thresholding (good results, default).
15326 @section premultiply
15327 Apply alpha premultiply effect to input video stream using first plane
15328 of second stream as alpha.
15330 Both streams must have same dimensions and same pixel format.
15332 The filter accepts the following option:
15336 Set which planes will be processed, unprocessed planes will be copied.
15337 By default value 0xf, all planes will be processed.
15340 Do not require 2nd input for processing, instead use alpha plane from input stream.
15344 Apply prewitt operator to input video stream.
15346 The filter accepts the following option:
15350 Set which planes will be processed, unprocessed planes will be copied.
15351 By default value 0xf, all planes will be processed.
15354 Set value which will be multiplied with filtered result.
15357 Set value which will be added to filtered result.
15360 @section pseudocolor
15362 Alter frame colors in video with pseudocolors.
15364 This filter accepts the following options:
15368 set pixel first component expression
15371 set pixel second component expression
15374 set pixel third component expression
15377 set pixel fourth component expression, corresponds to the alpha component
15380 set component to use as base for altering colors
15383 Each of them specifies the expression to use for computing the lookup table for
15384 the corresponding pixel component values.
15386 The expressions can contain the following constants and functions:
15391 The input width and height.
15394 The input value for the pixel component.
15396 @item ymin, umin, vmin, amin
15397 The minimum allowed component value.
15399 @item ymax, umax, vmax, amax
15400 The maximum allowed component value.
15403 All expressions default to "val".
15405 @subsection Examples
15409 Change too high luma values to gradient:
15411 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'"
15417 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
15418 Ratio) between two input videos.
15420 This filter takes in input two input videos, the first input is
15421 considered the "main" source and is passed unchanged to the
15422 output. The second input is used as a "reference" video for computing
15425 Both video inputs must have the same resolution and pixel format for
15426 this filter to work correctly. Also it assumes that both inputs
15427 have the same number of frames, which are compared one by one.
15429 The obtained average PSNR is printed through the logging system.
15431 The filter stores the accumulated MSE (mean squared error) of each
15432 frame, and at the end of the processing it is averaged across all frames
15433 equally, and the following formula is applied to obtain the PSNR:
15436 PSNR = 10*log10(MAX^2/MSE)
15439 Where MAX is the average of the maximum values of each component of the
15442 The description of the accepted parameters follows.
15445 @item stats_file, f
15446 If specified the filter will use the named file to save the PSNR of
15447 each individual frame. When filename equals "-" the data is sent to
15450 @item stats_version
15451 Specifies which version of the stats file format to use. Details of
15452 each format are written below.
15453 Default value is 1.
15455 @item stats_add_max
15456 Determines whether the max value is output to the stats log.
15457 Default value is 0.
15458 Requires stats_version >= 2. If this is set and stats_version < 2,
15459 the filter will return an error.
15462 This filter also supports the @ref{framesync} options.
15464 The file printed if @var{stats_file} is selected, contains a sequence of
15465 key/value pairs of the form @var{key}:@var{value} for each compared
15468 If a @var{stats_version} greater than 1 is specified, a header line precedes
15469 the list of per-frame-pair stats, with key value pairs following the frame
15470 format with the following parameters:
15473 @item psnr_log_version
15474 The version of the log file format. Will match @var{stats_version}.
15477 A comma separated list of the per-frame-pair parameters included in
15481 A description of each shown per-frame-pair parameter follows:
15485 sequential number of the input frame, starting from 1
15488 Mean Square Error pixel-by-pixel average difference of the compared
15489 frames, averaged over all the image components.
15491 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
15492 Mean Square Error pixel-by-pixel average difference of the compared
15493 frames for the component specified by the suffix.
15495 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
15496 Peak Signal to Noise ratio of the compared frames for the component
15497 specified by the suffix.
15499 @item max_avg, max_y, max_u, max_v
15500 Maximum allowed value for each channel, and average over all
15504 @subsection Examples
15509 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15510 [main][ref] psnr="stats_file=stats.log" [out]
15513 On this example the input file being processed is compared with the
15514 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
15515 is stored in @file{stats.log}.
15518 Another example with different containers:
15520 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]psnr" -f null -
15527 Pulldown reversal (inverse telecine) filter, capable of handling mixed
15528 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
15531 The pullup filter is designed to take advantage of future context in making
15532 its decisions. This filter is stateless in the sense that it does not lock
15533 onto a pattern to follow, but it instead looks forward to the following
15534 fields in order to identify matches and rebuild progressive frames.
15536 To produce content with an even framerate, insert the fps filter after
15537 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
15538 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
15540 The filter accepts the following options:
15547 These options set the amount of "junk" to ignore at the left, right, top, and
15548 bottom of the image, respectively. Left and right are in units of 8 pixels,
15549 while top and bottom are in units of 2 lines.
15550 The default is 8 pixels on each side.
15553 Set the strict breaks. Setting this option to 1 will reduce the chances of
15554 filter generating an occasional mismatched frame, but it may also cause an
15555 excessive number of frames to be dropped during high motion sequences.
15556 Conversely, setting it to -1 will make filter match fields more easily.
15557 This may help processing of video where there is slight blurring between
15558 the fields, but may also cause there to be interlaced frames in the output.
15559 Default value is @code{0}.
15562 Set the metric plane to use. It accepts the following values:
15568 Use chroma blue plane.
15571 Use chroma red plane.
15574 This option may be set to use chroma plane instead of the default luma plane
15575 for doing filter's computations. This may improve accuracy on very clean
15576 source material, but more likely will decrease accuracy, especially if there
15577 is chroma noise (rainbow effect) or any grayscale video.
15578 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
15579 load and make pullup usable in realtime on slow machines.
15582 For best results (without duplicated frames in the output file) it is
15583 necessary to change the output frame rate. For example, to inverse
15584 telecine NTSC input:
15586 ffmpeg -i input -vf pullup -r 24000/1001 ...
15591 Change video quantization parameters (QP).
15593 The filter accepts the following option:
15597 Set expression for quantization parameter.
15600 The expression is evaluated through the eval API and can contain, among others,
15601 the following constants:
15605 1 if index is not 129, 0 otherwise.
15608 Sequential index starting from -129 to 128.
15611 @subsection Examples
15615 Some equation like:
15623 Flush video frames from internal cache of frames into a random order.
15624 No frame is discarded.
15625 Inspired by @ref{frei0r} nervous filter.
15629 Set size in number of frames of internal cache, in range from @code{2} to
15630 @code{512}. Default is @code{30}.
15633 Set seed for random number generator, must be an integer included between
15634 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15635 less than @code{0}, the filter will try to use a good random seed on a
15639 @section readeia608
15641 Read closed captioning (EIA-608) information from the top lines of a video frame.
15643 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
15644 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
15645 with EIA-608 data (starting from 0). A description of each metadata value follows:
15648 @item lavfi.readeia608.X.cc
15649 The two bytes stored as EIA-608 data (printed in hexadecimal).
15651 @item lavfi.readeia608.X.line
15652 The number of the line on which the EIA-608 data was identified and read.
15655 This filter accepts the following options:
15659 Set the line to start scanning for EIA-608 data. Default is @code{0}.
15662 Set the line to end scanning for EIA-608 data. Default is @code{29}.
15665 Set the ratio of width reserved for sync code detection.
15666 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
15669 Enable checking the parity bit. In the event of a parity error, the filter will output
15670 @code{0x00} for that character. Default is false.
15673 Lowpass lines prior to further processing. Default is enabled.
15676 @subsection Examples
15680 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
15682 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
15688 Read vertical interval timecode (VITC) information from the top lines of a
15691 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
15692 timecode value, if a valid timecode has been detected. Further metadata key
15693 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
15694 timecode data has been found or not.
15696 This filter accepts the following options:
15700 Set the maximum number of lines to scan for VITC data. If the value is set to
15701 @code{-1} the full video frame is scanned. Default is @code{45}.
15704 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
15705 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
15708 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
15709 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
15712 @subsection Examples
15716 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
15717 draw @code{--:--:--:--} as a placeholder:
15719 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
15725 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
15727 Destination pixel at position (X, Y) will be picked from source (x, y) position
15728 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
15729 value for pixel will be used for destination pixel.
15731 Xmap and Ymap input video streams must be of same dimensions. Output video stream
15732 will have Xmap/Ymap video stream dimensions.
15733 Xmap and Ymap input video streams are 16bit depth, single channel.
15737 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
15738 Default is @code{color}.
15741 Specify the color of the unmapped pixels. For the syntax of this option,
15742 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15743 manual,ffmpeg-utils}. Default color is @code{black}.
15746 @section removegrain
15748 The removegrain filter is a spatial denoiser for progressive video.
15752 Set mode for the first plane.
15755 Set mode for the second plane.
15758 Set mode for the third plane.
15761 Set mode for the fourth plane.
15764 Range of mode is from 0 to 24. Description of each mode follows:
15768 Leave input plane unchanged. Default.
15771 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
15774 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
15777 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
15780 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
15781 This is equivalent to a median filter.
15784 Line-sensitive clipping giving the minimal change.
15787 Line-sensitive clipping, intermediate.
15790 Line-sensitive clipping, intermediate.
15793 Line-sensitive clipping, intermediate.
15796 Line-sensitive clipping on a line where the neighbours pixels are the closest.
15799 Replaces the target pixel with the closest neighbour.
15802 [1 2 1] horizontal and vertical kernel blur.
15808 Bob mode, interpolates top field from the line where the neighbours
15809 pixels are the closest.
15812 Bob mode, interpolates bottom field from the line where the neighbours
15813 pixels are the closest.
15816 Bob mode, interpolates top field. Same as 13 but with a more complicated
15817 interpolation formula.
15820 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
15821 interpolation formula.
15824 Clips the pixel with the minimum and maximum of respectively the maximum and
15825 minimum of each pair of opposite neighbour pixels.
15828 Line-sensitive clipping using opposite neighbours whose greatest distance from
15829 the current pixel is minimal.
15832 Replaces the pixel with the average of its 8 neighbours.
15835 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
15838 Clips pixels using the averages of opposite neighbour.
15841 Same as mode 21 but simpler and faster.
15844 Small edge and halo removal, but reputed useless.
15850 @section removelogo
15852 Suppress a TV station logo, using an image file to determine which
15853 pixels comprise the logo. It works by filling in the pixels that
15854 comprise the logo with neighboring pixels.
15856 The filter accepts the following options:
15860 Set the filter bitmap file, which can be any image format supported by
15861 libavformat. The width and height of the image file must match those of the
15862 video stream being processed.
15865 Pixels in the provided bitmap image with a value of zero are not
15866 considered part of the logo, non-zero pixels are considered part of
15867 the logo. If you use white (255) for the logo and black (0) for the
15868 rest, you will be safe. For making the filter bitmap, it is
15869 recommended to take a screen capture of a black frame with the logo
15870 visible, and then using a threshold filter followed by the erode
15871 filter once or twice.
15873 If needed, little splotches can be fixed manually. Remember that if
15874 logo pixels are not covered, the filter quality will be much
15875 reduced. Marking too many pixels as part of the logo does not hurt as
15876 much, but it will increase the amount of blurring needed to cover over
15877 the image and will destroy more information than necessary, and extra
15878 pixels will slow things down on a large logo.
15880 @section repeatfields
15882 This filter uses the repeat_field flag from the Video ES headers and hard repeats
15883 fields based on its value.
15887 Reverse a video clip.
15889 Warning: This filter requires memory to buffer the entire clip, so trimming
15892 @subsection Examples
15896 Take the first 5 seconds of a clip, and reverse it.
15903 Shift R/G/B/A pixels horizontally and/or vertically.
15905 The filter accepts the following options:
15908 Set amount to shift red horizontally.
15910 Set amount to shift red vertically.
15912 Set amount to shift green horizontally.
15914 Set amount to shift green vertically.
15916 Set amount to shift blue horizontally.
15918 Set amount to shift blue vertically.
15920 Set amount to shift alpha horizontally.
15922 Set amount to shift alpha vertically.
15924 Set edge mode, can be @var{smear}, default, or @var{warp}.
15927 @subsection Commands
15929 This filter supports the all above options as @ref{commands}.
15932 Apply roberts cross operator to input video stream.
15934 The filter accepts the following option:
15938 Set which planes will be processed, unprocessed planes will be copied.
15939 By default value 0xf, all planes will be processed.
15942 Set value which will be multiplied with filtered result.
15945 Set value which will be added to filtered result.
15950 Rotate video by an arbitrary angle expressed in radians.
15952 The filter accepts the following options:
15954 A description of the optional parameters follows.
15957 Set an expression for the angle by which to rotate the input video
15958 clockwise, expressed as a number of radians. A negative value will
15959 result in a counter-clockwise rotation. By default it is set to "0".
15961 This expression is evaluated for each frame.
15964 Set the output width expression, default value is "iw".
15965 This expression is evaluated just once during configuration.
15968 Set the output height expression, default value is "ih".
15969 This expression is evaluated just once during configuration.
15972 Enable bilinear interpolation if set to 1, a value of 0 disables
15973 it. Default value is 1.
15976 Set the color used to fill the output area not covered by the rotated
15977 image. For the general syntax of this option, check the
15978 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15979 If the special value "none" is selected then no
15980 background is printed (useful for example if the background is never shown).
15982 Default value is "black".
15985 The expressions for the angle and the output size can contain the
15986 following constants and functions:
15990 sequential number of the input frame, starting from 0. It is always NAN
15991 before the first frame is filtered.
15994 time in seconds of the input frame, it is set to 0 when the filter is
15995 configured. It is always NAN before the first frame is filtered.
15999 horizontal and vertical chroma subsample values. For example for the
16000 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16004 the input video width and height
16008 the output width and height, that is the size of the padded area as
16009 specified by the @var{width} and @var{height} expressions
16013 the minimal width/height required for completely containing the input
16014 video rotated by @var{a} radians.
16016 These are only available when computing the @option{out_w} and
16017 @option{out_h} expressions.
16020 @subsection Examples
16024 Rotate the input by PI/6 radians clockwise:
16030 Rotate the input by PI/6 radians counter-clockwise:
16036 Rotate the input by 45 degrees clockwise:
16042 Apply a constant rotation with period T, starting from an angle of PI/3:
16044 rotate=PI/3+2*PI*t/T
16048 Make the input video rotation oscillating with a period of T
16049 seconds and an amplitude of A radians:
16051 rotate=A*sin(2*PI/T*t)
16055 Rotate the video, output size is chosen so that the whole rotating
16056 input video is always completely contained in the output:
16058 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
16062 Rotate the video, reduce the output size so that no background is ever
16065 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
16069 @subsection Commands
16071 The filter supports the following commands:
16075 Set the angle expression.
16076 The command accepts the same syntax of the corresponding option.
16078 If the specified expression is not valid, it is kept at its current
16084 Apply Shape Adaptive Blur.
16086 The filter accepts the following options:
16089 @item luma_radius, lr
16090 Set luma blur filter strength, must be a value in range 0.1-4.0, default
16091 value is 1.0. A greater value will result in a more blurred image, and
16092 in slower processing.
16094 @item luma_pre_filter_radius, lpfr
16095 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
16098 @item luma_strength, ls
16099 Set luma maximum difference between pixels to still be considered, must
16100 be a value in the 0.1-100.0 range, default value is 1.0.
16102 @item chroma_radius, cr
16103 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
16104 greater value will result in a more blurred image, and in slower
16107 @item chroma_pre_filter_radius, cpfr
16108 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
16110 @item chroma_strength, cs
16111 Set chroma maximum difference between pixels to still be considered,
16112 must be a value in the -0.9-100.0 range.
16115 Each chroma option value, if not explicitly specified, is set to the
16116 corresponding luma option value.
16121 Scale (resize) the input video, using the libswscale library.
16123 The scale filter forces the output display aspect ratio to be the same
16124 of the input, by changing the output sample aspect ratio.
16126 If the input image format is different from the format requested by
16127 the next filter, the scale filter will convert the input to the
16130 @subsection Options
16131 The filter accepts the following options, or any of the options
16132 supported by the libswscale scaler.
16134 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
16135 the complete list of scaler options.
16140 Set the output video dimension expression. Default value is the input
16143 If the @var{width} or @var{w} value is 0, the input width is used for
16144 the output. If the @var{height} or @var{h} value is 0, the input height
16145 is used for the output.
16147 If one and only one of the values is -n with n >= 1, the scale filter
16148 will use a value that maintains the aspect ratio of the input image,
16149 calculated from the other specified dimension. After that it will,
16150 however, make sure that the calculated dimension is divisible by n and
16151 adjust the value if necessary.
16153 If both values are -n with n >= 1, the behavior will be identical to
16154 both values being set to 0 as previously detailed.
16156 See below for the list of accepted constants for use in the dimension
16160 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
16164 Only evaluate expressions once during the filter initialization or when a command is processed.
16167 Evaluate expressions for each incoming frame.
16171 Default value is @samp{init}.
16175 Set the interlacing mode. It accepts the following values:
16179 Force interlaced aware scaling.
16182 Do not apply interlaced scaling.
16185 Select interlaced aware scaling depending on whether the source frames
16186 are flagged as interlaced or not.
16189 Default value is @samp{0}.
16192 Set libswscale scaling flags. See
16193 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
16194 complete list of values. If not explicitly specified the filter applies
16198 @item param0, param1
16199 Set libswscale input parameters for scaling algorithms that need them. See
16200 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
16201 complete documentation. If not explicitly specified the filter applies
16207 Set the video size. For the syntax of this option, check the
16208 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16210 @item in_color_matrix
16211 @item out_color_matrix
16212 Set in/output YCbCr color space type.
16214 This allows the autodetected value to be overridden as well as allows forcing
16215 a specific value used for the output and encoder.
16217 If not specified, the color space type depends on the pixel format.
16223 Choose automatically.
16226 Format conforming to International Telecommunication Union (ITU)
16227 Recommendation BT.709.
16230 Set color space conforming to the United States Federal Communications
16231 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
16236 Set color space conforming to:
16240 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
16243 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
16246 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
16251 Set color space conforming to SMPTE ST 240:1999.
16254 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
16259 Set in/output YCbCr sample range.
16261 This allows the autodetected value to be overridden as well as allows forcing
16262 a specific value used for the output and encoder. If not specified, the
16263 range depends on the pixel format. Possible values:
16267 Choose automatically.
16270 Set full range (0-255 in case of 8-bit luma).
16272 @item mpeg/limited/tv
16273 Set "MPEG" range (16-235 in case of 8-bit luma).
16276 @item force_original_aspect_ratio
16277 Enable decreasing or increasing output video width or height if necessary to
16278 keep the original aspect ratio. Possible values:
16282 Scale the video as specified and disable this feature.
16285 The output video dimensions will automatically be decreased if needed.
16288 The output video dimensions will automatically be increased if needed.
16292 One useful instance of this option is that when you know a specific device's
16293 maximum allowed resolution, you can use this to limit the output video to
16294 that, while retaining the aspect ratio. For example, device A allows
16295 1280x720 playback, and your video is 1920x800. Using this option (set it to
16296 decrease) and specifying 1280x720 to the command line makes the output
16299 Please note that this is a different thing than specifying -1 for @option{w}
16300 or @option{h}, you still need to specify the output resolution for this option
16303 @item force_divisible_by
16304 Ensures that both the output dimensions, width and height, are divisible by the
16305 given integer when used together with @option{force_original_aspect_ratio}. This
16306 works similar to using @code{-n} in the @option{w} and @option{h} options.
16308 This option respects the value set for @option{force_original_aspect_ratio},
16309 increasing or decreasing the resolution accordingly. The video's aspect ratio
16310 may be slightly modified.
16312 This option can be handy if you need to have a video fit within or exceed
16313 a defined resolution using @option{force_original_aspect_ratio} but also have
16314 encoder restrictions on width or height divisibility.
16318 The values of the @option{w} and @option{h} options are expressions
16319 containing the following constants:
16324 The input width and height
16328 These are the same as @var{in_w} and @var{in_h}.
16332 The output (scaled) width and height
16336 These are the same as @var{out_w} and @var{out_h}
16339 The same as @var{iw} / @var{ih}
16342 input sample aspect ratio
16345 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16349 horizontal and vertical input chroma subsample values. For example for the
16350 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16354 horizontal and vertical output chroma subsample values. For example for the
16355 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16358 The (sequential) number of the input frame, starting from 0.
16359 Only available with @code{eval=frame}.
16362 The presentation timestamp of the input frame, expressed as a number of
16363 seconds. Only available with @code{eval=frame}.
16366 The position (byte offset) of the frame in the input stream, or NaN if
16367 this information is unavailable and/or meaningless (for example in case of synthetic video).
16368 Only available with @code{eval=frame}.
16371 @subsection Examples
16375 Scale the input video to a size of 200x100
16380 This is equivalent to:
16391 Specify a size abbreviation for the output size:
16396 which can also be written as:
16402 Scale the input to 2x:
16404 scale=w=2*iw:h=2*ih
16408 The above is the same as:
16410 scale=2*in_w:2*in_h
16414 Scale the input to 2x with forced interlaced scaling:
16416 scale=2*iw:2*ih:interl=1
16420 Scale the input to half size:
16422 scale=w=iw/2:h=ih/2
16426 Increase the width, and set the height to the same size:
16432 Seek Greek harmony:
16439 Increase the height, and set the width to 3/2 of the height:
16441 scale=w=3/2*oh:h=3/5*ih
16445 Increase the size, making the size a multiple of the chroma
16448 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
16452 Increase the width to a maximum of 500 pixels,
16453 keeping the same aspect ratio as the input:
16455 scale=w='min(500\, iw*3/2):h=-1'
16459 Make pixels square by combining scale and setsar:
16461 scale='trunc(ih*dar):ih',setsar=1/1
16465 Make pixels square by combining scale and setsar,
16466 making sure the resulting resolution is even (required by some codecs):
16468 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
16472 @subsection Commands
16474 This filter supports the following commands:
16478 Set the output video dimension expression.
16479 The command accepts the same syntax of the corresponding option.
16481 If the specified expression is not valid, it is kept at its current
16487 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
16488 format conversion on CUDA video frames. Setting the output width and height
16489 works in the same way as for the @var{scale} filter.
16491 The following additional options are accepted:
16494 The pixel format of the output CUDA frames. If set to the string "same" (the
16495 default), the input format will be kept. Note that automatic format negotiation
16496 and conversion is not yet supported for hardware frames
16499 The interpolation algorithm used for resizing. One of the following:
16506 @item cubic2p_bspline
16507 2-parameter cubic (B=1, C=0)
16509 @item cubic2p_catmullrom
16510 2-parameter cubic (B=0, C=1/2)
16512 @item cubic2p_b05c03
16513 2-parameter cubic (B=1/2, C=3/10)
16521 @item force_original_aspect_ratio
16522 Enable decreasing or increasing output video width or height if necessary to
16523 keep the original aspect ratio. Possible values:
16527 Scale the video as specified and disable this feature.
16530 The output video dimensions will automatically be decreased if needed.
16533 The output video dimensions will automatically be increased if needed.
16537 One useful instance of this option is that when you know a specific device's
16538 maximum allowed resolution, you can use this to limit the output video to
16539 that, while retaining the aspect ratio. For example, device A allows
16540 1280x720 playback, and your video is 1920x800. Using this option (set it to
16541 decrease) and specifying 1280x720 to the command line makes the output
16544 Please note that this is a different thing than specifying -1 for @option{w}
16545 or @option{h}, you still need to specify the output resolution for this option
16548 @item force_divisible_by
16549 Ensures that both the output dimensions, width and height, are divisible by the
16550 given integer when used together with @option{force_original_aspect_ratio}. This
16551 works similar to using @code{-n} in the @option{w} and @option{h} options.
16553 This option respects the value set for @option{force_original_aspect_ratio},
16554 increasing or decreasing the resolution accordingly. The video's aspect ratio
16555 may be slightly modified.
16557 This option can be handy if you need to have a video fit within or exceed
16558 a defined resolution using @option{force_original_aspect_ratio} but also have
16559 encoder restrictions on width or height divisibility.
16565 Scale (resize) the input video, based on a reference video.
16567 See the scale filter for available options, scale2ref supports the same but
16568 uses the reference video instead of the main input as basis. scale2ref also
16569 supports the following additional constants for the @option{w} and
16570 @option{h} options:
16575 The main input video's width and height
16578 The same as @var{main_w} / @var{main_h}
16581 The main input video's sample aspect ratio
16583 @item main_dar, mdar
16584 The main input video's display aspect ratio. Calculated from
16585 @code{(main_w / main_h) * main_sar}.
16589 The main input video's horizontal and vertical chroma subsample values.
16590 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
16594 The (sequential) number of the main input frame, starting from 0.
16595 Only available with @code{eval=frame}.
16598 The presentation timestamp of the main input frame, expressed as a number of
16599 seconds. Only available with @code{eval=frame}.
16602 The position (byte offset) of the frame in the main input stream, or NaN if
16603 this information is unavailable and/or meaningless (for example in case of synthetic video).
16604 Only available with @code{eval=frame}.
16607 @subsection Examples
16611 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
16613 'scale2ref[b][a];[a][b]overlay'
16617 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
16619 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
16623 @subsection Commands
16625 This filter supports the following commands:
16629 Set the output video dimension expression.
16630 The command accepts the same syntax of the corresponding option.
16632 If the specified expression is not valid, it is kept at its current
16637 Scroll input video horizontally and/or vertically by constant speed.
16639 The filter accepts the following options:
16641 @item horizontal, h
16642 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
16643 Negative values changes scrolling direction.
16646 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
16647 Negative values changes scrolling direction.
16650 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
16653 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
16656 @subsection Commands
16658 This filter supports the following @ref{commands}:
16660 @item horizontal, h
16661 Set the horizontal scrolling speed.
16663 Set the vertical scrolling speed.
16669 Detect video scene change.
16671 This filter sets frame metadata with mafd between frame, the scene score, and
16672 forward the frame to the next filter, so they can use these metadata to detect
16673 scene change or others.
16675 In addition, this filter logs a message and sets frame metadata when it detects
16676 a scene change by @option{threshold}.
16678 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
16680 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
16681 to detect scene change.
16683 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
16684 detect scene change with @option{threshold}.
16686 The filter accepts the following options:
16690 Set the scene change detection threshold as a percentage of maximum change. Good
16691 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
16694 Default value is @code{10.}.
16697 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
16698 You can enable it if you want to get snapshot of scene change frames only.
16701 @anchor{selectivecolor}
16702 @section selectivecolor
16704 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
16705 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
16706 by the "purity" of the color (that is, how saturated it already is).
16708 This filter is similar to the Adobe Photoshop Selective Color tool.
16710 The filter accepts the following options:
16713 @item correction_method
16714 Select color correction method.
16716 Available values are:
16719 Specified adjustments are applied "as-is" (added/subtracted to original pixel
16722 Specified adjustments are relative to the original component value.
16724 Default is @code{absolute}.
16726 Adjustments for red pixels (pixels where the red component is the maximum)
16728 Adjustments for yellow pixels (pixels where the blue component is the minimum)
16730 Adjustments for green pixels (pixels where the green component is the maximum)
16732 Adjustments for cyan pixels (pixels where the red component is the minimum)
16734 Adjustments for blue pixels (pixels where the blue component is the maximum)
16736 Adjustments for magenta pixels (pixels where the green component is the minimum)
16738 Adjustments for white pixels (pixels where all components are greater than 128)
16740 Adjustments for all pixels except pure black and pure white
16742 Adjustments for black pixels (pixels where all components are lesser than 128)
16744 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
16747 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
16748 4 space separated floating point adjustment values in the [-1,1] range,
16749 respectively to adjust the amount of cyan, magenta, yellow and black for the
16750 pixels of its range.
16752 @subsection Examples
16756 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
16757 increase magenta by 27% in blue areas:
16759 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
16763 Use a Photoshop selective color preset:
16765 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
16769 @anchor{separatefields}
16770 @section separatefields
16772 The @code{separatefields} takes a frame-based video input and splits
16773 each frame into its components fields, producing a new half height clip
16774 with twice the frame rate and twice the frame count.
16776 This filter use field-dominance information in frame to decide which
16777 of each pair of fields to place first in the output.
16778 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
16780 @section setdar, setsar
16782 The @code{setdar} filter sets the Display Aspect Ratio for the filter
16785 This is done by changing the specified Sample (aka Pixel) Aspect
16786 Ratio, according to the following equation:
16788 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
16791 Keep in mind that the @code{setdar} filter does not modify the pixel
16792 dimensions of the video frame. Also, the display aspect ratio set by
16793 this filter may be changed by later filters in the filterchain,
16794 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
16797 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
16798 the filter output video.
16800 Note that as a consequence of the application of this filter, the
16801 output display aspect ratio will change according to the equation
16804 Keep in mind that the sample aspect ratio set by the @code{setsar}
16805 filter may be changed by later filters in the filterchain, e.g. if
16806 another "setsar" or a "setdar" filter is applied.
16808 It accepts the following parameters:
16811 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
16812 Set the aspect ratio used by the filter.
16814 The parameter can be a floating point number string, an expression, or
16815 a string of the form @var{num}:@var{den}, where @var{num} and
16816 @var{den} are the numerator and denominator of the aspect ratio. If
16817 the parameter is not specified, it is assumed the value "0".
16818 In case the form "@var{num}:@var{den}" is used, the @code{:} character
16822 Set the maximum integer value to use for expressing numerator and
16823 denominator when reducing the expressed aspect ratio to a rational.
16824 Default value is @code{100}.
16828 The parameter @var{sar} is an expression containing
16829 the following constants:
16833 These are approximated values for the mathematical constants e
16834 (Euler's number), pi (Greek pi), and phi (the golden ratio).
16837 The input width and height.
16840 These are the same as @var{w} / @var{h}.
16843 The input sample aspect ratio.
16846 The input display aspect ratio. It is the same as
16847 (@var{w} / @var{h}) * @var{sar}.
16850 Horizontal and vertical chroma subsample values. For example, for the
16851 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16854 @subsection Examples
16859 To change the display aspect ratio to 16:9, specify one of the following:
16866 To change the sample aspect ratio to 10:11, specify:
16872 To set a display aspect ratio of 16:9, and specify a maximum integer value of
16873 1000 in the aspect ratio reduction, use the command:
16875 setdar=ratio=16/9:max=1000
16883 Force field for the output video frame.
16885 The @code{setfield} filter marks the interlace type field for the
16886 output frames. It does not change the input frame, but only sets the
16887 corresponding property, which affects how the frame is treated by
16888 following filters (e.g. @code{fieldorder} or @code{yadif}).
16890 The filter accepts the following options:
16895 Available values are:
16899 Keep the same field property.
16902 Mark the frame as bottom-field-first.
16905 Mark the frame as top-field-first.
16908 Mark the frame as progressive.
16915 Force frame parameter for the output video frame.
16917 The @code{setparams} filter marks interlace and color range for the
16918 output frames. It does not change the input frame, but only sets the
16919 corresponding property, which affects how the frame is treated by
16924 Available values are:
16928 Keep the same field property (default).
16931 Mark the frame as bottom-field-first.
16934 Mark the frame as top-field-first.
16937 Mark the frame as progressive.
16941 Available values are:
16945 Keep the same color range property (default).
16947 @item unspecified, unknown
16948 Mark the frame as unspecified color range.
16950 @item limited, tv, mpeg
16951 Mark the frame as limited range.
16953 @item full, pc, jpeg
16954 Mark the frame as full range.
16957 @item color_primaries
16958 Set the color primaries.
16959 Available values are:
16963 Keep the same color primaries property (default).
16980 Set the color transfer.
16981 Available values are:
16985 Keep the same color trc property (default).
17007 Set the colorspace.
17008 Available values are:
17012 Keep the same colorspace property (default).
17025 @item chroma-derived-nc
17026 @item chroma-derived-c
17033 Show a line containing various information for each input video frame.
17034 The input video is not modified.
17036 This filter supports the following options:
17040 Calculate checksums of each plane. By default enabled.
17043 The shown line contains a sequence of key/value pairs of the form
17044 @var{key}:@var{value}.
17046 The following values are shown in the output:
17050 The (sequential) number of the input frame, starting from 0.
17053 The Presentation TimeStamp of the input frame, expressed as a number of
17054 time base units. The time base unit depends on the filter input pad.
17057 The Presentation TimeStamp of the input frame, expressed as a number of
17061 The position of the frame in the input stream, or -1 if this information is
17062 unavailable and/or meaningless (for example in case of synthetic video).
17065 The pixel format name.
17068 The sample aspect ratio of the input frame, expressed in the form
17069 @var{num}/@var{den}.
17072 The size of the input frame. For the syntax of this option, check the
17073 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17076 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
17077 for bottom field first).
17080 This is 1 if the frame is a key frame, 0 otherwise.
17083 The picture type of the input frame ("I" for an I-frame, "P" for a
17084 P-frame, "B" for a B-frame, or "?" for an unknown type).
17085 Also refer to the documentation of the @code{AVPictureType} enum and of
17086 the @code{av_get_picture_type_char} function defined in
17087 @file{libavutil/avutil.h}.
17090 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
17092 @item plane_checksum
17093 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
17094 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
17097 The mean value of pixels in each plane of the input frame, expressed in the form
17098 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
17101 The standard deviation of pixel values in each plane of the input frame, expressed
17102 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
17106 @section showpalette
17108 Displays the 256 colors palette of each frame. This filter is only relevant for
17109 @var{pal8} pixel format frames.
17111 It accepts the following option:
17115 Set the size of the box used to represent one palette color entry. Default is
17116 @code{30} (for a @code{30x30} pixel box).
17119 @section shuffleframes
17121 Reorder and/or duplicate and/or drop video frames.
17123 It accepts the following parameters:
17127 Set the destination indexes of input frames.
17128 This is space or '|' separated list of indexes that maps input frames to output
17129 frames. Number of indexes also sets maximal value that each index may have.
17130 '-1' index have special meaning and that is to drop frame.
17133 The first frame has the index 0. The default is to keep the input unchanged.
17135 @subsection Examples
17139 Swap second and third frame of every three frames of the input:
17141 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
17145 Swap 10th and 1st frame of every ten frames of the input:
17147 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
17151 @section shuffleplanes
17153 Reorder and/or duplicate video planes.
17155 It accepts the following parameters:
17160 The index of the input plane to be used as the first output plane.
17163 The index of the input plane to be used as the second output plane.
17166 The index of the input plane to be used as the third output plane.
17169 The index of the input plane to be used as the fourth output plane.
17173 The first plane has the index 0. The default is to keep the input unchanged.
17175 @subsection Examples
17179 Swap the second and third planes of the input:
17181 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
17185 @anchor{signalstats}
17186 @section signalstats
17187 Evaluate various visual metrics that assist in determining issues associated
17188 with the digitization of analog video media.
17190 By default the filter will log these metadata values:
17194 Display the minimal Y value contained within the input frame. Expressed in
17198 Display the Y value at the 10% percentile within the input frame. Expressed in
17202 Display the average Y value within the input frame. Expressed in range of
17206 Display the Y value at the 90% percentile within the input frame. Expressed in
17210 Display the maximum Y value contained within the input frame. Expressed in
17214 Display the minimal U value contained within the input frame. Expressed in
17218 Display the U value at the 10% percentile within the input frame. Expressed in
17222 Display the average U value within the input frame. Expressed in range of
17226 Display the U value at the 90% percentile within the input frame. Expressed in
17230 Display the maximum U value contained within the input frame. Expressed in
17234 Display the minimal V value contained within the input frame. Expressed in
17238 Display the V value at the 10% percentile within the input frame. Expressed in
17242 Display the average V value within the input frame. Expressed in range of
17246 Display the V value at the 90% percentile within the input frame. Expressed in
17250 Display the maximum V value contained within the input frame. Expressed in
17254 Display the minimal saturation value contained within the input frame.
17255 Expressed in range of [0-~181.02].
17258 Display the saturation value at the 10% percentile within the input frame.
17259 Expressed in range of [0-~181.02].
17262 Display the average saturation value within the input frame. Expressed in range
17266 Display the saturation value at the 90% percentile within the input frame.
17267 Expressed in range of [0-~181.02].
17270 Display the maximum saturation value contained within the input frame.
17271 Expressed in range of [0-~181.02].
17274 Display the median value for hue within the input frame. Expressed in range of
17278 Display the average value for hue within the input frame. Expressed in range of
17282 Display the average of sample value difference between all values of the Y
17283 plane in the current frame and corresponding values of the previous input frame.
17284 Expressed in range of [0-255].
17287 Display the average of sample value difference between all values of the U
17288 plane in the current frame and corresponding values of the previous input frame.
17289 Expressed in range of [0-255].
17292 Display the average of sample value difference between all values of the V
17293 plane in the current frame and corresponding values of the previous input frame.
17294 Expressed in range of [0-255].
17297 Display bit depth of Y plane in current frame.
17298 Expressed in range of [0-16].
17301 Display bit depth of U plane in current frame.
17302 Expressed in range of [0-16].
17305 Display bit depth of V plane in current frame.
17306 Expressed in range of [0-16].
17309 The filter accepts the following options:
17315 @option{stat} specify an additional form of image analysis.
17316 @option{out} output video with the specified type of pixel highlighted.
17318 Both options accept the following values:
17322 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
17323 unlike the neighboring pixels of the same field. Examples of temporal outliers
17324 include the results of video dropouts, head clogs, or tape tracking issues.
17327 Identify @var{vertical line repetition}. Vertical line repetition includes
17328 similar rows of pixels within a frame. In born-digital video vertical line
17329 repetition is common, but this pattern is uncommon in video digitized from an
17330 analog source. When it occurs in video that results from the digitization of an
17331 analog source it can indicate concealment from a dropout compensator.
17334 Identify pixels that fall outside of legal broadcast range.
17338 Set the highlight color for the @option{out} option. The default color is
17342 @subsection Examples
17346 Output data of various video metrics:
17348 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
17352 Output specific data about the minimum and maximum values of the Y plane per frame:
17354 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
17358 Playback video while highlighting pixels that are outside of broadcast range in red.
17360 ffplay example.mov -vf signalstats="out=brng:color=red"
17364 Playback video with signalstats metadata drawn over the frame.
17366 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
17369 The contents of signalstat_drawtext.txt used in the command are:
17372 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
17373 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
17374 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
17375 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
17383 Calculates the MPEG-7 Video Signature. The filter can handle more than one
17384 input. In this case the matching between the inputs can be calculated additionally.
17385 The filter always passes through the first input. The signature of each stream can
17386 be written into a file.
17388 It accepts the following options:
17392 Enable or disable the matching process.
17394 Available values are:
17398 Disable the calculation of a matching (default).
17400 Calculate the matching for the whole video and output whether the whole video
17401 matches or only parts.
17403 Calculate only until a matching is found or the video ends. Should be faster in
17408 Set the number of inputs. The option value must be a non negative integer.
17409 Default value is 1.
17412 Set the path to which the output is written. If there is more than one input,
17413 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
17414 integer), that will be replaced with the input number. If no filename is
17415 specified, no output will be written. This is the default.
17418 Choose the output format.
17420 Available values are:
17424 Use the specified binary representation (default).
17426 Use the specified xml representation.
17430 Set threshold to detect one word as similar. The option value must be an integer
17431 greater than zero. The default value is 9000.
17434 Set threshold to detect all words as similar. The option value must be an integer
17435 greater than zero. The default value is 60000.
17438 Set threshold to detect frames as similar. The option value must be an integer
17439 greater than zero. The default value is 116.
17442 Set the minimum length of a sequence in frames to recognize it as matching
17443 sequence. The option value must be a non negative integer value.
17444 The default value is 0.
17447 Set the minimum relation, that matching frames to all frames must have.
17448 The option value must be a double value between 0 and 1. The default value is 0.5.
17451 @subsection Examples
17455 To calculate the signature of an input video and store it in signature.bin:
17457 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
17461 To detect whether two videos match and store the signatures in XML format in
17462 signature0.xml and signature1.xml:
17464 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 -
17472 Blur the input video without impacting the outlines.
17474 It accepts the following options:
17477 @item luma_radius, lr
17478 Set the luma radius. The option value must be a float number in
17479 the range [0.1,5.0] that specifies the variance of the gaussian filter
17480 used to blur the image (slower if larger). Default value is 1.0.
17482 @item luma_strength, ls
17483 Set the luma strength. The option value must be a float number
17484 in the range [-1.0,1.0] that configures the blurring. A value included
17485 in [0.0,1.0] will blur the image whereas a value included in
17486 [-1.0,0.0] will sharpen the image. Default value is 1.0.
17488 @item luma_threshold, lt
17489 Set the luma threshold used as a coefficient to determine
17490 whether a pixel should be blurred or not. The option value must be an
17491 integer in the range [-30,30]. A value of 0 will filter all the image,
17492 a value included in [0,30] will filter flat areas and a value included
17493 in [-30,0] will filter edges. Default value is 0.
17495 @item chroma_radius, cr
17496 Set the chroma radius. The option value must be a float number in
17497 the range [0.1,5.0] that specifies the variance of the gaussian filter
17498 used to blur the image (slower if larger). Default value is @option{luma_radius}.
17500 @item chroma_strength, cs
17501 Set the chroma strength. The option value must be a float number
17502 in the range [-1.0,1.0] that configures the blurring. A value included
17503 in [0.0,1.0] will blur the image whereas a value included in
17504 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
17506 @item chroma_threshold, ct
17507 Set the chroma threshold used as a coefficient to determine
17508 whether a pixel should be blurred or not. The option value must be an
17509 integer in the range [-30,30]. A value of 0 will filter all the image,
17510 a value included in [0,30] will filter flat areas and a value included
17511 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
17514 If a chroma option is not explicitly set, the corresponding luma value
17518 Apply sobel operator to input video stream.
17520 The filter accepts the following option:
17524 Set which planes will be processed, unprocessed planes will be copied.
17525 By default value 0xf, all planes will be processed.
17528 Set value which will be multiplied with filtered result.
17531 Set value which will be added to filtered result.
17537 Apply a simple postprocessing filter that compresses and decompresses the image
17538 at several (or - in the case of @option{quality} level @code{6} - all) shifts
17539 and average the results.
17541 The filter accepts the following options:
17545 Set quality. This option defines the number of levels for averaging. It accepts
17546 an integer in the range 0-6. If set to @code{0}, the filter will have no
17547 effect. A value of @code{6} means the higher quality. For each increment of
17548 that value the speed drops by a factor of approximately 2. Default value is
17552 Force a constant quantization parameter. If not set, the filter will use the QP
17553 from the video stream (if available).
17556 Set thresholding mode. Available modes are:
17560 Set hard thresholding (default).
17562 Set soft thresholding (better de-ringing effect, but likely blurrier).
17565 @item use_bframe_qp
17566 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
17567 option may cause flicker since the B-Frames have often larger QP. Default is
17568 @code{0} (not enabled).
17571 @subsection Commands
17573 This filter supports the following commands:
17575 @item quality, level
17576 Set quality level. The value @code{max} can be used to set the maximum level,
17577 currently @code{6}.
17583 Scale the input by applying one of the super-resolution methods based on
17584 convolutional neural networks. Supported models:
17588 Super-Resolution Convolutional Neural Network model (SRCNN).
17589 See @url{https://arxiv.org/abs/1501.00092}.
17592 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
17593 See @url{https://arxiv.org/abs/1609.05158}.
17596 Training scripts as well as scripts for model file (.pb) saving can be found at
17597 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
17598 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
17600 Native model files (.model) can be generated from TensorFlow model
17601 files (.pb) by using tools/python/convert.py
17603 The filter accepts the following options:
17607 Specify which DNN backend to use for model loading and execution. This option accepts
17608 the following values:
17612 Native implementation of DNN loading and execution.
17615 TensorFlow backend. To enable this backend you
17616 need to install the TensorFlow for C library (see
17617 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
17618 @code{--enable-libtensorflow}
17621 Default value is @samp{native}.
17624 Set path to model file specifying network architecture and its parameters.
17625 Note that different backends use different file formats. TensorFlow backend
17626 can load files for both formats, while native backend can load files for only
17630 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
17631 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
17632 input upscaled using bicubic upscaling with proper scale factor.
17635 This feature can also be finished with @ref{dnn_processing} filter.
17639 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
17641 This filter takes in input two input videos, the first input is
17642 considered the "main" source and is passed unchanged to the
17643 output. The second input is used as a "reference" video for computing
17646 Both video inputs must have the same resolution and pixel format for
17647 this filter to work correctly. Also it assumes that both inputs
17648 have the same number of frames, which are compared one by one.
17650 The filter stores the calculated SSIM of each frame.
17652 The description of the accepted parameters follows.
17655 @item stats_file, f
17656 If specified the filter will use the named file to save the SSIM of
17657 each individual frame. When filename equals "-" the data is sent to
17661 The file printed if @var{stats_file} is selected, contains a sequence of
17662 key/value pairs of the form @var{key}:@var{value} for each compared
17665 A description of each shown parameter follows:
17669 sequential number of the input frame, starting from 1
17671 @item Y, U, V, R, G, B
17672 SSIM of the compared frames for the component specified by the suffix.
17675 SSIM of the compared frames for the whole frame.
17678 Same as above but in dB representation.
17681 This filter also supports the @ref{framesync} options.
17683 @subsection Examples
17688 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
17689 [main][ref] ssim="stats_file=stats.log" [out]
17692 On this example the input file being processed is compared with the
17693 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
17694 is stored in @file{stats.log}.
17697 Another example with both psnr and ssim at same time:
17699 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
17703 Another example with different containers:
17705 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]ssim" -f null -
17711 Convert between different stereoscopic image formats.
17713 The filters accept the following options:
17717 Set stereoscopic image format of input.
17719 Available values for input image formats are:
17722 side by side parallel (left eye left, right eye right)
17725 side by side crosseye (right eye left, left eye right)
17728 side by side parallel with half width resolution
17729 (left eye left, right eye right)
17732 side by side crosseye with half width resolution
17733 (right eye left, left eye right)
17737 above-below (left eye above, right eye below)
17741 above-below (right eye above, left eye below)
17745 above-below with half height resolution
17746 (left eye above, right eye below)
17750 above-below with half height resolution
17751 (right eye above, left eye below)
17754 alternating frames (left eye first, right eye second)
17757 alternating frames (right eye first, left eye second)
17760 interleaved rows (left eye has top row, right eye starts on next row)
17763 interleaved rows (right eye has top row, left eye starts on next row)
17766 interleaved columns, left eye first
17769 interleaved columns, right eye first
17771 Default value is @samp{sbsl}.
17775 Set stereoscopic image format of output.
17779 side by side parallel (left eye left, right eye right)
17782 side by side crosseye (right eye left, left eye right)
17785 side by side parallel with half width resolution
17786 (left eye left, right eye right)
17789 side by side crosseye with half width resolution
17790 (right eye left, left eye right)
17794 above-below (left eye above, right eye below)
17798 above-below (right eye above, left eye below)
17802 above-below with half height resolution
17803 (left eye above, right eye below)
17807 above-below with half height resolution
17808 (right eye above, left eye below)
17811 alternating frames (left eye first, right eye second)
17814 alternating frames (right eye first, left eye second)
17817 interleaved rows (left eye has top row, right eye starts on next row)
17820 interleaved rows (right eye has top row, left eye starts on next row)
17823 anaglyph red/blue gray
17824 (red filter on left eye, blue filter on right eye)
17827 anaglyph red/green gray
17828 (red filter on left eye, green filter on right eye)
17831 anaglyph red/cyan gray
17832 (red filter on left eye, cyan filter on right eye)
17835 anaglyph red/cyan half colored
17836 (red filter on left eye, cyan filter on right eye)
17839 anaglyph red/cyan color
17840 (red filter on left eye, cyan filter on right eye)
17843 anaglyph red/cyan color optimized with the least squares projection of dubois
17844 (red filter on left eye, cyan filter on right eye)
17847 anaglyph green/magenta gray
17848 (green filter on left eye, magenta filter on right eye)
17851 anaglyph green/magenta half colored
17852 (green filter on left eye, magenta filter on right eye)
17855 anaglyph green/magenta colored
17856 (green filter on left eye, magenta filter on right eye)
17859 anaglyph green/magenta color optimized with the least squares projection of dubois
17860 (green filter on left eye, magenta filter on right eye)
17863 anaglyph yellow/blue gray
17864 (yellow filter on left eye, blue filter on right eye)
17867 anaglyph yellow/blue half colored
17868 (yellow filter on left eye, blue filter on right eye)
17871 anaglyph yellow/blue colored
17872 (yellow filter on left eye, blue filter on right eye)
17875 anaglyph yellow/blue color optimized with the least squares projection of dubois
17876 (yellow filter on left eye, blue filter on right eye)
17879 mono output (left eye only)
17882 mono output (right eye only)
17885 checkerboard, left eye first
17888 checkerboard, right eye first
17891 interleaved columns, left eye first
17894 interleaved columns, right eye first
17900 Default value is @samp{arcd}.
17903 @subsection Examples
17907 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
17913 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
17919 @section streamselect, astreamselect
17920 Select video or audio streams.
17922 The filter accepts the following options:
17926 Set number of inputs. Default is 2.
17929 Set input indexes to remap to outputs.
17932 @subsection Commands
17934 The @code{streamselect} and @code{astreamselect} filter supports the following
17939 Set input indexes to remap to outputs.
17942 @subsection Examples
17946 Select first 5 seconds 1st stream and rest of time 2nd stream:
17948 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
17952 Same as above, but for audio:
17954 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
17961 Draw subtitles on top of input video using the libass library.
17963 To enable compilation of this filter you need to configure FFmpeg with
17964 @code{--enable-libass}. This filter also requires a build with libavcodec and
17965 libavformat to convert the passed subtitles file to ASS (Advanced Substation
17966 Alpha) subtitles format.
17968 The filter accepts the following options:
17972 Set the filename of the subtitle file to read. It must be specified.
17974 @item original_size
17975 Specify the size of the original video, the video for which the ASS file
17976 was composed. For the syntax of this option, check the
17977 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17978 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
17979 correctly scale the fonts if the aspect ratio has been changed.
17982 Set a directory path containing fonts that can be used by the filter.
17983 These fonts will be used in addition to whatever the font provider uses.
17986 Process alpha channel, by default alpha channel is untouched.
17989 Set subtitles input character encoding. @code{subtitles} filter only. Only
17990 useful if not UTF-8.
17992 @item stream_index, si
17993 Set subtitles stream index. @code{subtitles} filter only.
17996 Override default style or script info parameters of the subtitles. It accepts a
17997 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
18000 If the first key is not specified, it is assumed that the first value
18001 specifies the @option{filename}.
18003 For example, to render the file @file{sub.srt} on top of the input
18004 video, use the command:
18009 which is equivalent to:
18011 subtitles=filename=sub.srt
18014 To render the default subtitles stream from file @file{video.mkv}, use:
18016 subtitles=video.mkv
18019 To render the second subtitles stream from that file, use:
18021 subtitles=video.mkv:si=1
18024 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
18025 @code{DejaVu Serif}, use:
18027 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
18030 @section super2xsai
18032 Scale the input by 2x and smooth using the Super2xSaI (Scale and
18033 Interpolate) pixel art scaling algorithm.
18035 Useful for enlarging pixel art images without reducing sharpness.
18039 Swap two rectangular objects in video.
18041 This filter accepts the following options:
18051 Set 1st rect x coordinate.
18054 Set 1st rect y coordinate.
18057 Set 2nd rect x coordinate.
18060 Set 2nd rect y coordinate.
18062 All expressions are evaluated once for each frame.
18065 The all options are expressions containing the following constants:
18070 The input width and height.
18073 same as @var{w} / @var{h}
18076 input sample aspect ratio
18079 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
18082 The number of the input frame, starting from 0.
18085 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
18088 the position in the file of the input frame, NAN if unknown
18095 Blend successive video frames.
18101 Apply telecine process to the video.
18103 This filter accepts the following options:
18112 The default value is @code{top}.
18116 A string of numbers representing the pulldown pattern you wish to apply.
18117 The default value is @code{23}.
18121 Some typical patterns:
18126 24p: 2332 (preferred)
18133 24p: 222222222223 ("Euro pulldown")
18138 @section thistogram
18140 Compute and draw a color distribution histogram for the input video across time.
18142 Unlike @ref{histogram} video filter which only shows histogram of single input frame
18143 at certain time, this filter shows also past histograms of number of frames defined
18144 by @code{width} option.
18146 The computed histogram is a representation of the color component
18147 distribution in an image.
18149 The filter accepts the following options:
18153 Set width of single color component output. Default value is @code{0}.
18154 Value of @code{0} means width will be picked from input video.
18155 This also set number of passed histograms to keep.
18156 Allowed range is [0, 8192].
18158 @item display_mode, d
18160 It accepts the following values:
18163 Per color component graphs are placed below each other.
18166 Per color component graphs are placed side by side.
18169 Presents information identical to that in the @code{parade}, except
18170 that the graphs representing color components are superimposed directly
18173 Default is @code{stack}.
18175 @item levels_mode, m
18176 Set mode. Can be either @code{linear}, or @code{logarithmic}.
18177 Default is @code{linear}.
18179 @item components, c
18180 Set what color components to display.
18181 Default is @code{7}.
18184 Set background opacity. Default is @code{0.9}.
18187 Show envelope. Default is disabled.
18190 Set envelope color. Default is @code{gold}.
18195 Apply threshold effect to video stream.
18197 This filter needs four video streams to perform thresholding.
18198 First stream is stream we are filtering.
18199 Second stream is holding threshold values, third stream is holding min values,
18200 and last, fourth stream is holding max values.
18202 The filter accepts the following option:
18206 Set which planes will be processed, unprocessed planes will be copied.
18207 By default value 0xf, all planes will be processed.
18210 For example if first stream pixel's component value is less then threshold value
18211 of pixel component from 2nd threshold stream, third stream value will picked,
18212 otherwise fourth stream pixel component value will be picked.
18214 Using color source filter one can perform various types of thresholding:
18216 @subsection Examples
18220 Binary threshold, using gray color as threshold:
18222 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
18226 Inverted binary threshold, using gray color as threshold:
18228 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
18232 Truncate binary threshold, using gray color as threshold:
18234 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
18238 Threshold to zero, using gray color as threshold:
18240 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
18244 Inverted threshold to zero, using gray color as threshold:
18246 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
18251 Select the most representative frame in a given sequence of consecutive frames.
18253 The filter accepts the following options:
18257 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
18258 will pick one of them, and then handle the next batch of @var{n} frames until
18259 the end. Default is @code{100}.
18262 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
18263 value will result in a higher memory usage, so a high value is not recommended.
18265 @subsection Examples
18269 Extract one picture each 50 frames:
18275 Complete example of a thumbnail creation with @command{ffmpeg}:
18277 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
18284 Tile several successive frames together.
18286 The @ref{untile} filter can do the reverse.
18288 The filter accepts the following options:
18293 Set the grid size (i.e. the number of lines and columns). For the syntax of
18294 this option, check the
18295 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18298 Set the maximum number of frames to render in the given area. It must be less
18299 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
18300 the area will be used.
18303 Set the outer border margin in pixels.
18306 Set the inner border thickness (i.e. the number of pixels between frames). For
18307 more advanced padding options (such as having different values for the edges),
18308 refer to the pad video filter.
18311 Specify the color of the unused area. For the syntax of this option, check the
18312 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18313 The default value of @var{color} is "black".
18316 Set the number of frames to overlap when tiling several successive frames together.
18317 The value must be between @code{0} and @var{nb_frames - 1}.
18320 Set the number of frames to initially be empty before displaying first output frame.
18321 This controls how soon will one get first output frame.
18322 The value must be between @code{0} and @var{nb_frames - 1}.
18325 @subsection Examples
18329 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
18331 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
18333 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
18334 duplicating each output frame to accommodate the originally detected frame
18338 Display @code{5} pictures in an area of @code{3x2} frames,
18339 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
18340 mixed flat and named options:
18342 tile=3x2:nb_frames=5:padding=7:margin=2
18346 @section tinterlace
18348 Perform various types of temporal field interlacing.
18350 Frames are counted starting from 1, so the first input frame is
18353 The filter accepts the following options:
18358 Specify the mode of the interlacing. This option can also be specified
18359 as a value alone. See below for a list of values for this option.
18361 Available values are:
18365 Move odd frames into the upper field, even into the lower field,
18366 generating a double height frame at half frame rate.
18370 Frame 1 Frame 2 Frame 3 Frame 4
18372 11111 22222 33333 44444
18373 11111 22222 33333 44444
18374 11111 22222 33333 44444
18375 11111 22222 33333 44444
18389 Only output odd frames, even frames are dropped, generating a frame with
18390 unchanged height at half frame rate.
18395 Frame 1 Frame 2 Frame 3 Frame 4
18397 11111 22222 33333 44444
18398 11111 22222 33333 44444
18399 11111 22222 33333 44444
18400 11111 22222 33333 44444
18410 Only output even frames, odd frames are dropped, generating a frame with
18411 unchanged height at half frame rate.
18416 Frame 1 Frame 2 Frame 3 Frame 4
18418 11111 22222 33333 44444
18419 11111 22222 33333 44444
18420 11111 22222 33333 44444
18421 11111 22222 33333 44444
18431 Expand each frame to full height, but pad alternate lines with black,
18432 generating a frame with double height at the same input frame rate.
18437 Frame 1 Frame 2 Frame 3 Frame 4
18439 11111 22222 33333 44444
18440 11111 22222 33333 44444
18441 11111 22222 33333 44444
18442 11111 22222 33333 44444
18445 11111 ..... 33333 .....
18446 ..... 22222 ..... 44444
18447 11111 ..... 33333 .....
18448 ..... 22222 ..... 44444
18449 11111 ..... 33333 .....
18450 ..... 22222 ..... 44444
18451 11111 ..... 33333 .....
18452 ..... 22222 ..... 44444
18456 @item interleave_top, 4
18457 Interleave the upper field from odd frames with the lower field from
18458 even frames, generating a frame with unchanged height at half frame rate.
18463 Frame 1 Frame 2 Frame 3 Frame 4
18465 11111<- 22222 33333<- 44444
18466 11111 22222<- 33333 44444<-
18467 11111<- 22222 33333<- 44444
18468 11111 22222<- 33333 44444<-
18478 @item interleave_bottom, 5
18479 Interleave the lower field from odd frames with the upper field from
18480 even frames, generating a frame with unchanged height at half frame rate.
18485 Frame 1 Frame 2 Frame 3 Frame 4
18487 11111 22222<- 33333 44444<-
18488 11111<- 22222 33333<- 44444
18489 11111 22222<- 33333 44444<-
18490 11111<- 22222 33333<- 44444
18500 @item interlacex2, 6
18501 Double frame rate with unchanged height. Frames are inserted each
18502 containing the second temporal field from the previous input frame and
18503 the first temporal field from the next input frame. This mode relies on
18504 the top_field_first flag. Useful for interlaced video displays with no
18505 field synchronisation.
18510 Frame 1 Frame 2 Frame 3 Frame 4
18512 11111 22222 33333 44444
18513 11111 22222 33333 44444
18514 11111 22222 33333 44444
18515 11111 22222 33333 44444
18518 11111 22222 22222 33333 33333 44444 44444
18519 11111 11111 22222 22222 33333 33333 44444
18520 11111 22222 22222 33333 33333 44444 44444
18521 11111 11111 22222 22222 33333 33333 44444
18526 Move odd frames into the upper field, even into the lower field,
18527 generating a double height frame at same frame rate.
18532 Frame 1 Frame 2 Frame 3 Frame 4
18534 11111 22222 33333 44444
18535 11111 22222 33333 44444
18536 11111 22222 33333 44444
18537 11111 22222 33333 44444
18540 11111 33333 33333 55555
18541 22222 22222 44444 44444
18542 11111 33333 33333 55555
18543 22222 22222 44444 44444
18544 11111 33333 33333 55555
18545 22222 22222 44444 44444
18546 11111 33333 33333 55555
18547 22222 22222 44444 44444
18552 Numeric values are deprecated but are accepted for backward
18553 compatibility reasons.
18555 Default mode is @code{merge}.
18558 Specify flags influencing the filter process.
18560 Available value for @var{flags} is:
18563 @item low_pass_filter, vlpf
18564 Enable linear vertical low-pass filtering in the filter.
18565 Vertical low-pass filtering is required when creating an interlaced
18566 destination from a progressive source which contains high-frequency
18567 vertical detail. Filtering will reduce interlace 'twitter' and Moire
18570 @item complex_filter, cvlpf
18571 Enable complex vertical low-pass filtering.
18572 This will slightly less reduce interlace 'twitter' and Moire
18573 patterning but better retain detail and subjective sharpness impression.
18576 Bypass already interlaced frames, only adjust the frame rate.
18579 Vertical low-pass filtering and bypassing already interlaced frames can only be
18580 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
18585 Pick median pixels from several successive input video frames.
18587 The filter accepts the following options:
18591 Set radius of median filter.
18592 Default is 1. Allowed range is from 1 to 127.
18595 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
18598 Set median percentile. Default value is @code{0.5}.
18599 Default value of @code{0.5} will pick always median values, while @code{0} will pick
18600 minimum values, and @code{1} maximum values.
18605 Mix successive video frames.
18607 A description of the accepted options follows.
18611 The number of successive frames to mix. If unspecified, it defaults to 3.
18614 Specify weight of each input video frame.
18615 Each weight is separated by space. If number of weights is smaller than
18616 number of @var{frames} last specified weight will be used for all remaining
18620 Specify scale, if it is set it will be multiplied with sum
18621 of each weight multiplied with pixel values to give final destination
18622 pixel value. By default @var{scale} is auto scaled to sum of weights.
18625 @subsection Examples
18629 Average 7 successive frames:
18631 tmix=frames=7:weights="1 1 1 1 1 1 1"
18635 Apply simple temporal convolution:
18637 tmix=frames=3:weights="-1 3 -1"
18641 Similar as above but only showing temporal differences:
18643 tmix=frames=3:weights="-1 2 -1":scale=1
18649 Tone map colors from different dynamic ranges.
18651 This filter expects data in single precision floating point, as it needs to
18652 operate on (and can output) out-of-range values. Another filter, such as
18653 @ref{zscale}, is needed to convert the resulting frame to a usable format.
18655 The tonemapping algorithms implemented only work on linear light, so input
18656 data should be linearized beforehand (and possibly correctly tagged).
18659 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
18662 @subsection Options
18663 The filter accepts the following options.
18667 Set the tone map algorithm to use.
18669 Possible values are:
18672 Do not apply any tone map, only desaturate overbright pixels.
18675 Hard-clip any out-of-range values. Use it for perfect color accuracy for
18676 in-range values, while distorting out-of-range values.
18679 Stretch the entire reference gamut to a linear multiple of the display.
18682 Fit a logarithmic transfer between the tone curves.
18685 Preserve overall image brightness with a simple curve, using nonlinear
18686 contrast, which results in flattening details and degrading color accuracy.
18689 Preserve both dark and bright details better than @var{reinhard}, at the cost
18690 of slightly darkening everything. Use it when detail preservation is more
18691 important than color and brightness accuracy.
18694 Smoothly map out-of-range values, while retaining contrast and colors for
18695 in-range material as much as possible. Use it when color accuracy is more
18696 important than detail preservation.
18702 Tune the tone mapping algorithm.
18704 This affects the following algorithms:
18710 Specifies the scale factor to use while stretching.
18714 Specifies the exponent of the function.
18718 Specify an extra linear coefficient to multiply into the signal before clipping.
18722 Specify the local contrast coefficient at the display peak.
18723 Default to 0.5, which means that in-gamut values will be about half as bright
18730 Specify the transition point from linear to mobius transform. Every value
18731 below this point is guaranteed to be mapped 1:1. The higher the value, the
18732 more accurate the result will be, at the cost of losing bright details.
18733 Default to 0.3, which due to the steep initial slope still preserves in-range
18734 colors fairly accurately.
18738 Apply desaturation for highlights that exceed this level of brightness. The
18739 higher the parameter, the more color information will be preserved. This
18740 setting helps prevent unnaturally blown-out colors for super-highlights, by
18741 (smoothly) turning into white instead. This makes images feel more natural,
18742 at the cost of reducing information about out-of-range colors.
18744 The default of 2.0 is somewhat conservative and will mostly just apply to
18745 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
18747 This option works only if the input frame has a supported color tag.
18750 Override signal/nominal/reference peak with this value. Useful when the
18751 embedded peak information in display metadata is not reliable or when tone
18752 mapping from a lower range to a higher range.
18757 Temporarily pad video frames.
18759 The filter accepts the following options:
18763 Specify number of delay frames before input video stream. Default is 0.
18766 Specify number of padding frames after input video stream.
18767 Set to -1 to pad indefinitely. Default is 0.
18770 Set kind of frames added to beginning of stream.
18771 Can be either @var{add} or @var{clone}.
18772 With @var{add} frames of solid-color are added.
18773 With @var{clone} frames are clones of first frame.
18774 Default is @var{add}.
18777 Set kind of frames added to end of stream.
18778 Can be either @var{add} or @var{clone}.
18779 With @var{add} frames of solid-color are added.
18780 With @var{clone} frames are clones of last frame.
18781 Default is @var{add}.
18783 @item start_duration, stop_duration
18784 Specify the duration of the start/stop delay. See
18785 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18786 for the accepted syntax.
18787 These options override @var{start} and @var{stop}. Default is 0.
18790 Specify the color of the padded area. For the syntax of this option,
18791 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
18792 manual,ffmpeg-utils}.
18794 The default value of @var{color} is "black".
18800 Transpose rows with columns in the input video and optionally flip it.
18802 It accepts the following parameters:
18807 Specify the transposition direction.
18809 Can assume the following values:
18811 @item 0, 4, cclock_flip
18812 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
18820 Rotate by 90 degrees clockwise, that is:
18828 Rotate by 90 degrees counterclockwise, that is:
18835 @item 3, 7, clock_flip
18836 Rotate by 90 degrees clockwise and vertically flip, that is:
18844 For values between 4-7, the transposition is only done if the input
18845 video geometry is portrait and not landscape. These values are
18846 deprecated, the @code{passthrough} option should be used instead.
18848 Numerical values are deprecated, and should be dropped in favor of
18849 symbolic constants.
18852 Do not apply the transposition if the input geometry matches the one
18853 specified by the specified value. It accepts the following values:
18856 Always apply transposition.
18858 Preserve portrait geometry (when @var{height} >= @var{width}).
18860 Preserve landscape geometry (when @var{width} >= @var{height}).
18863 Default value is @code{none}.
18866 For example to rotate by 90 degrees clockwise and preserve portrait
18869 transpose=dir=1:passthrough=portrait
18872 The command above can also be specified as:
18874 transpose=1:portrait
18877 @section transpose_npp
18879 Transpose rows with columns in the input video and optionally flip it.
18880 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
18882 It accepts the following parameters:
18887 Specify the transposition direction.
18889 Can assume the following values:
18892 Rotate by 90 degrees counterclockwise and vertically flip. (default)
18895 Rotate by 90 degrees clockwise.
18898 Rotate by 90 degrees counterclockwise.
18901 Rotate by 90 degrees clockwise and vertically flip.
18905 Do not apply the transposition if the input geometry matches the one
18906 specified by the specified value. It accepts the following values:
18909 Always apply transposition. (default)
18911 Preserve portrait geometry (when @var{height} >= @var{width}).
18913 Preserve landscape geometry (when @var{width} >= @var{height}).
18919 Trim the input so that the output contains one continuous subpart of the input.
18921 It accepts the following parameters:
18924 Specify the time of the start of the kept section, i.e. the frame with the
18925 timestamp @var{start} will be the first frame in the output.
18928 Specify the time of the first frame that will be dropped, i.e. the frame
18929 immediately preceding the one with the timestamp @var{end} will be the last
18930 frame in the output.
18933 This is the same as @var{start}, except this option sets the start timestamp
18934 in timebase units instead of seconds.
18937 This is the same as @var{end}, except this option sets the end timestamp
18938 in timebase units instead of seconds.
18941 The maximum duration of the output in seconds.
18944 The number of the first frame that should be passed to the output.
18947 The number of the first frame that should be dropped.
18950 @option{start}, @option{end}, and @option{duration} are expressed as time
18951 duration specifications; see
18952 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18953 for the accepted syntax.
18955 Note that the first two sets of the start/end options and the @option{duration}
18956 option look at the frame timestamp, while the _frame variants simply count the
18957 frames that pass through the filter. Also note that this filter does not modify
18958 the timestamps. If you wish for the output timestamps to start at zero, insert a
18959 setpts filter after the trim filter.
18961 If multiple start or end options are set, this filter tries to be greedy and
18962 keep all the frames that match at least one of the specified constraints. To keep
18963 only the part that matches all the constraints at once, chain multiple trim
18966 The defaults are such that all the input is kept. So it is possible to set e.g.
18967 just the end values to keep everything before the specified time.
18972 Drop everything except the second minute of input:
18974 ffmpeg -i INPUT -vf trim=60:120
18978 Keep only the first second:
18980 ffmpeg -i INPUT -vf trim=duration=1
18985 @section unpremultiply
18986 Apply alpha unpremultiply effect to input video stream using first plane
18987 of second stream as alpha.
18989 Both streams must have same dimensions and same pixel format.
18991 The filter accepts the following option:
18995 Set which planes will be processed, unprocessed planes will be copied.
18996 By default value 0xf, all planes will be processed.
18998 If the format has 1 or 2 components, then luma is bit 0.
18999 If the format has 3 or 4 components:
19000 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
19001 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
19002 If present, the alpha channel is always the last bit.
19005 Do not require 2nd input for processing, instead use alpha plane from input stream.
19011 Sharpen or blur the input video.
19013 It accepts the following parameters:
19016 @item luma_msize_x, lx
19017 Set the luma matrix horizontal size. It must be an odd integer between
19018 3 and 23. The default value is 5.
19020 @item luma_msize_y, ly
19021 Set the luma matrix vertical size. It must be an odd integer between 3
19022 and 23. The default value is 5.
19024 @item luma_amount, la
19025 Set the luma effect strength. It must be a floating point number, reasonable
19026 values lay between -1.5 and 1.5.
19028 Negative values will blur the input video, while positive values will
19029 sharpen it, a value of zero will disable the effect.
19031 Default value is 1.0.
19033 @item chroma_msize_x, cx
19034 Set the chroma matrix horizontal size. It must be an odd integer
19035 between 3 and 23. The default value is 5.
19037 @item chroma_msize_y, cy
19038 Set the chroma matrix vertical size. It must be an odd integer
19039 between 3 and 23. The default value is 5.
19041 @item chroma_amount, ca
19042 Set the chroma effect strength. It must be a floating point number, reasonable
19043 values lay between -1.5 and 1.5.
19045 Negative values will blur the input video, while positive values will
19046 sharpen it, a value of zero will disable the effect.
19048 Default value is 0.0.
19052 All parameters are optional and default to the equivalent of the
19053 string '5:5:1.0:5:5:0.0'.
19055 @subsection Examples
19059 Apply strong luma sharpen effect:
19061 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
19065 Apply a strong blur of both luma and chroma parameters:
19067 unsharp=7:7:-2:7:7:-2
19074 Decompose a video made of tiled images into the individual images.
19076 The frame rate of the output video is the frame rate of the input video
19077 multiplied by the number of tiles.
19079 This filter does the reverse of @ref{tile}.
19081 The filter accepts the following options:
19086 Set the grid size (i.e. the number of lines and columns). For the syntax of
19087 this option, check the
19088 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19091 @subsection Examples
19095 Produce a 1-second video from a still image file made of 25 frames stacked
19096 vertically, like an analogic film reel:
19098 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
19104 Apply ultra slow/simple postprocessing filter that compresses and decompresses
19105 the image at several (or - in the case of @option{quality} level @code{8} - all)
19106 shifts and average the results.
19108 The way this differs from the behavior of spp is that uspp actually encodes &
19109 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
19110 DCT similar to MJPEG.
19112 The filter accepts the following options:
19116 Set quality. This option defines the number of levels for averaging. It accepts
19117 an integer in the range 0-8. If set to @code{0}, the filter will have no
19118 effect. A value of @code{8} means the higher quality. For each increment of
19119 that value the speed drops by a factor of approximately 2. Default value is
19123 Force a constant quantization parameter. If not set, the filter will use the QP
19124 from the video stream (if available).
19129 Convert 360 videos between various formats.
19131 The filter accepts the following options:
19137 Set format of the input/output video.
19145 Equirectangular projection.
19150 Cubemap with 3x2/6x1/1x6 layout.
19152 Format specific options:
19157 Set padding proportion for the input/output cubemap. Values in decimals.
19164 1% of face is padding. For example, with 1920x1280 resolution face size would be 640x640 and padding would be 3 pixels from each side. (640 * 0.01 = 6 pixels)
19167 Default value is @b{@samp{0}}.
19168 Maximum value is @b{@samp{0.1}}.
19172 Set fixed padding for the input/output cubemap. Values in pixels.
19174 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
19178 Set order of faces for the input/output cubemap. Choose one direction for each position.
19180 Designation of directions:
19196 Default value is @b{@samp{rludfb}}.
19200 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
19202 Designation of angles:
19205 0 degrees clockwise
19207 90 degrees clockwise
19209 180 degrees clockwise
19211 270 degrees clockwise
19214 Default value is @b{@samp{000000}}.
19218 Equi-Angular Cubemap.
19225 Format specific options:
19230 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19232 If diagonal field of view is set it overrides horizontal and vertical field of view.
19237 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19239 If diagonal field of view is set it overrides horizontal and vertical field of view.
19245 Format specific options:
19250 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19252 If diagonal field of view is set it overrides horizontal and vertical field of view.
19257 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19259 If diagonal field of view is set it overrides horizontal and vertical field of view.
19265 Facebook's 360 formats.
19268 Stereographic format.
19270 Format specific options:
19275 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19277 If diagonal field of view is set it overrides horizontal and vertical field of view.
19282 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19284 If diagonal field of view is set it overrides horizontal and vertical field of view.
19291 Ball format, gives significant distortion toward the back.
19294 Hammer-Aitoff map projection format.
19297 Sinusoidal map projection format.
19300 Fisheye projection.
19302 Format specific options:
19307 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19309 If diagonal field of view is set it overrides horizontal and vertical field of view.
19314 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19316 If diagonal field of view is set it overrides horizontal and vertical field of view.
19320 Pannini projection.
19322 Format specific options:
19325 Set output pannini parameter.
19328 Set input pannini parameter.
19332 Cylindrical projection.
19334 Format specific options:
19339 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19341 If diagonal field of view is set it overrides horizontal and vertical field of view.
19346 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19348 If diagonal field of view is set it overrides horizontal and vertical field of view.
19352 Perspective projection. @i{(output only)}
19354 Format specific options:
19357 Set perspective parameter.
19361 Tetrahedron projection.
19364 Truncated square pyramid projection.
19368 Half equirectangular projection.
19373 Format specific options:
19378 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19380 If diagonal field of view is set it overrides horizontal and vertical field of view.
19385 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19387 If diagonal field of view is set it overrides horizontal and vertical field of view.
19391 Orthographic format.
19393 Format specific options:
19398 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19400 If diagonal field of view is set it overrides horizontal and vertical field of view.
19405 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19407 If diagonal field of view is set it overrides horizontal and vertical field of view.
19412 Set interpolation method.@*
19413 @i{Note: more complex interpolation methods require much more memory to run.}
19423 Bilinear interpolation.
19425 Lagrange9 interpolation.
19428 Bicubic interpolation.
19431 Lanczos interpolation.
19434 Spline16 interpolation.
19437 Gaussian interpolation.
19440 Default value is @b{@samp{line}}.
19444 Set the output video resolution.
19446 Default resolution depends on formats.
19450 Set the input/output stereo format.
19461 Default value is @b{@samp{2d}} for input and output format.
19466 Set rotation for the output video. Values in degrees.
19469 Set rotation order for the output video. Choose one item for each position.
19480 Default value is @b{@samp{ypr}}.
19485 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
19489 Set if input video is flipped horizontally/vertically. Boolean values.
19492 Set if input video is transposed. Boolean value, by default disabled.
19495 Set if output video needs to be transposed. Boolean value, by default disabled.
19498 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
19501 @subsection Examples
19505 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
19507 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
19510 Extract back view of Equi-Angular Cubemap:
19512 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
19515 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
19517 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
19521 @subsection Commands
19523 This filter supports subset of above options as @ref{commands}.
19525 @section vaguedenoiser
19527 Apply a wavelet based denoiser.
19529 It transforms each frame from the video input into the wavelet domain,
19530 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
19531 the obtained coefficients. It does an inverse wavelet transform after.
19532 Due to wavelet properties, it should give a nice smoothed result, and
19533 reduced noise, without blurring picture features.
19535 This filter accepts the following options:
19539 The filtering strength. The higher, the more filtered the video will be.
19540 Hard thresholding can use a higher threshold than soft thresholding
19541 before the video looks overfiltered. Default value is 2.
19544 The filtering method the filter will use.
19546 It accepts the following values:
19549 All values under the threshold will be zeroed.
19552 All values under the threshold will be zeroed. All values above will be
19553 reduced by the threshold.
19556 Scales or nullifies coefficients - intermediary between (more) soft and
19557 (less) hard thresholding.
19560 Default is garrote.
19563 Number of times, the wavelet will decompose the picture. Picture can't
19564 be decomposed beyond a particular point (typically, 8 for a 640x480
19565 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
19568 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
19571 A list of the planes to process. By default all planes are processed.
19574 The threshold type the filter will use.
19576 It accepts the following values:
19579 Threshold used is same for all decompositions.
19582 Threshold used depends also on each decomposition coefficients.
19585 Default is universal.
19588 @section vectorscope
19590 Display 2 color component values in the two dimensional graph (which is called
19593 This filter accepts the following options:
19597 Set vectorscope mode.
19599 It accepts the following values:
19603 Gray values are displayed on graph, higher brightness means more pixels have
19604 same component color value on location in graph. This is the default mode.
19607 Gray values are displayed on graph. Surrounding pixels values which are not
19608 present in video frame are drawn in gradient of 2 color components which are
19609 set by option @code{x} and @code{y}. The 3rd color component is static.
19612 Actual color components values present in video frame are displayed on graph.
19615 Similar as color2 but higher frequency of same values @code{x} and @code{y}
19616 on graph increases value of another color component, which is luminance by
19617 default values of @code{x} and @code{y}.
19620 Actual colors present in video frame are displayed on graph. If two different
19621 colors map to same position on graph then color with higher value of component
19622 not present in graph is picked.
19625 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
19626 component picked from radial gradient.
19630 Set which color component will be represented on X-axis. Default is @code{1}.
19633 Set which color component will be represented on Y-axis. Default is @code{2}.
19636 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
19637 of color component which represents frequency of (X, Y) location in graph.
19642 No envelope, this is default.
19645 Instant envelope, even darkest single pixel will be clearly highlighted.
19648 Hold maximum and minimum values presented in graph over time. This way you
19649 can still spot out of range values without constantly looking at vectorscope.
19652 Peak and instant envelope combined together.
19656 Set what kind of graticule to draw.
19665 Set graticule opacity.
19668 Set graticule flags.
19672 Draw graticule for white point.
19675 Draw graticule for black point.
19678 Draw color points short names.
19682 Set background opacity.
19684 @item lthreshold, l
19685 Set low threshold for color component not represented on X or Y axis.
19686 Values lower than this value will be ignored. Default is 0.
19687 Note this value is multiplied with actual max possible value one pixel component
19688 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
19691 @item hthreshold, h
19692 Set high threshold for color component not represented on X or Y axis.
19693 Values higher than this value will be ignored. Default is 1.
19694 Note this value is multiplied with actual max possible value one pixel component
19695 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
19696 is 0.9 * 255 = 230.
19698 @item colorspace, c
19699 Set what kind of colorspace to use when drawing graticule.
19709 Set color tint for gray/tint vectorscope mode. By default both options are zero.
19710 This means no tint, and output will remain gray.
19713 @anchor{vidstabdetect}
19714 @section vidstabdetect
19716 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
19717 @ref{vidstabtransform} for pass 2.
19719 This filter generates a file with relative translation and rotation
19720 transform information about subsequent frames, which is then used by
19721 the @ref{vidstabtransform} filter.
19723 To enable compilation of this filter you need to configure FFmpeg with
19724 @code{--enable-libvidstab}.
19726 This filter accepts the following options:
19730 Set the path to the file used to write the transforms information.
19731 Default value is @file{transforms.trf}.
19734 Set how shaky the video is and how quick the camera is. It accepts an
19735 integer in the range 1-10, a value of 1 means little shakiness, a
19736 value of 10 means strong shakiness. Default value is 5.
19739 Set the accuracy of the detection process. It must be a value in the
19740 range 1-15. A value of 1 means low accuracy, a value of 15 means high
19741 accuracy. Default value is 15.
19744 Set stepsize of the search process. The region around minimum is
19745 scanned with 1 pixel resolution. Default value is 6.
19748 Set minimum contrast. Below this value a local measurement field is
19749 discarded. Must be a floating point value in the range 0-1. Default
19753 Set reference frame number for tripod mode.
19755 If enabled, the motion of the frames is compared to a reference frame
19756 in the filtered stream, identified by the specified number. The idea
19757 is to compensate all movements in a more-or-less static scene and keep
19758 the camera view absolutely still.
19760 If set to 0, it is disabled. The frames are counted starting from 1.
19763 Show fields and transforms in the resulting frames. It accepts an
19764 integer in the range 0-2. Default value is 0, which disables any
19768 @subsection Examples
19772 Use default values:
19778 Analyze strongly shaky movie and put the results in file
19779 @file{mytransforms.trf}:
19781 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
19785 Visualize the result of internal transformations in the resulting
19788 vidstabdetect=show=1
19792 Analyze a video with medium shakiness using @command{ffmpeg}:
19794 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
19798 @anchor{vidstabtransform}
19799 @section vidstabtransform
19801 Video stabilization/deshaking: pass 2 of 2,
19802 see @ref{vidstabdetect} for pass 1.
19804 Read a file with transform information for each frame and
19805 apply/compensate them. Together with the @ref{vidstabdetect}
19806 filter this can be used to deshake videos. See also
19807 @url{http://public.hronopik.de/vid.stab}. It is important to also use
19808 the @ref{unsharp} filter, see below.
19810 To enable compilation of this filter you need to configure FFmpeg with
19811 @code{--enable-libvidstab}.
19813 @subsection Options
19817 Set path to the file used to read the transforms. Default value is
19818 @file{transforms.trf}.
19821 Set the number of frames (value*2 + 1) used for lowpass filtering the
19822 camera movements. Default value is 10.
19824 For example a number of 10 means that 21 frames are used (10 in the
19825 past and 10 in the future) to smoothen the motion in the video. A
19826 larger value leads to a smoother video, but limits the acceleration of
19827 the camera (pan/tilt movements). 0 is a special case where a static
19828 camera is simulated.
19831 Set the camera path optimization algorithm.
19833 Accepted values are:
19836 gaussian kernel low-pass filter on camera motion (default)
19838 averaging on transformations
19842 Set maximal number of pixels to translate frames. Default value is -1,
19846 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
19847 value is -1, meaning no limit.
19850 Specify how to deal with borders that may be visible due to movement
19853 Available values are:
19856 keep image information from previous frame (default)
19858 fill the border black
19862 Invert transforms if set to 1. Default value is 0.
19865 Consider transforms as relative to previous frame if set to 1,
19866 absolute if set to 0. Default value is 0.
19869 Set percentage to zoom. A positive value will result in a zoom-in
19870 effect, a negative value in a zoom-out effect. Default value is 0 (no
19874 Set optimal zooming to avoid borders.
19876 Accepted values are:
19881 optimal static zoom value is determined (only very strong movements
19882 will lead to visible borders) (default)
19884 optimal adaptive zoom value is determined (no borders will be
19885 visible), see @option{zoomspeed}
19888 Note that the value given at zoom is added to the one calculated here.
19891 Set percent to zoom maximally each frame (enabled when
19892 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
19896 Specify type of interpolation.
19898 Available values are:
19903 linear only horizontal
19905 linear in both directions (default)
19907 cubic in both directions (slow)
19911 Enable virtual tripod mode if set to 1, which is equivalent to
19912 @code{relative=0:smoothing=0}. Default value is 0.
19914 Use also @code{tripod} option of @ref{vidstabdetect}.
19917 Increase log verbosity if set to 1. Also the detected global motions
19918 are written to the temporary file @file{global_motions.trf}. Default
19922 @subsection Examples
19926 Use @command{ffmpeg} for a typical stabilization with default values:
19928 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
19931 Note the use of the @ref{unsharp} filter which is always recommended.
19934 Zoom in a bit more and load transform data from a given file:
19936 vidstabtransform=zoom=5:input="mytransforms.trf"
19940 Smoothen the video even more:
19942 vidstabtransform=smoothing=30
19948 Flip the input video vertically.
19950 For example, to vertically flip a video with @command{ffmpeg}:
19952 ffmpeg -i in.avi -vf "vflip" out.avi
19957 Detect variable frame rate video.
19959 This filter tries to detect if the input is variable or constant frame rate.
19961 At end it will output number of frames detected as having variable delta pts,
19962 and ones with constant delta pts.
19963 If there was frames with variable delta, than it will also show min, max and
19964 average delta encountered.
19968 Boost or alter saturation.
19970 The filter accepts the following options:
19973 Set strength of boost if positive value or strength of alter if negative value.
19974 Default is 0. Allowed range is from -2 to 2.
19977 Set the red balance. Default is 1. Allowed range is from -10 to 10.
19980 Set the green balance. Default is 1. Allowed range is from -10 to 10.
19983 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
19986 Set the red luma coefficient.
19989 Set the green luma coefficient.
19992 Set the blue luma coefficient.
19995 If @code{intensity} is negative and this is set to 1, colors will change,
19996 otherwise colors will be less saturated, more towards gray.
19999 @subsection Commands
20001 This filter supports the all above options as @ref{commands}.
20006 Make or reverse a natural vignetting effect.
20008 The filter accepts the following options:
20012 Set lens angle expression as a number of radians.
20014 The value is clipped in the @code{[0,PI/2]} range.
20016 Default value: @code{"PI/5"}
20020 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
20024 Set forward/backward mode.
20026 Available modes are:
20029 The larger the distance from the central point, the darker the image becomes.
20032 The larger the distance from the central point, the brighter the image becomes.
20033 This can be used to reverse a vignette effect, though there is no automatic
20034 detection to extract the lens @option{angle} and other settings (yet). It can
20035 also be used to create a burning effect.
20038 Default value is @samp{forward}.
20041 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
20043 It accepts the following values:
20046 Evaluate expressions only once during the filter initialization.
20049 Evaluate expressions for each incoming frame. This is way slower than the
20050 @samp{init} mode since it requires all the scalers to be re-computed, but it
20051 allows advanced dynamic expressions.
20054 Default value is @samp{init}.
20057 Set dithering to reduce the circular banding effects. Default is @code{1}
20061 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
20062 Setting this value to the SAR of the input will make a rectangular vignetting
20063 following the dimensions of the video.
20065 Default is @code{1/1}.
20068 @subsection Expressions
20070 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
20071 following parameters.
20076 input width and height
20079 the number of input frame, starting from 0
20082 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
20083 @var{TB} units, NAN if undefined
20086 frame rate of the input video, NAN if the input frame rate is unknown
20089 the PTS (Presentation TimeStamp) of the filtered video frame,
20090 expressed in seconds, NAN if undefined
20093 time base of the input video
20097 @subsection Examples
20101 Apply simple strong vignetting effect:
20107 Make a flickering vignetting:
20109 vignette='PI/4+random(1)*PI/50':eval=frame
20114 @section vmafmotion
20116 Obtain the average VMAF motion score of a video.
20117 It is one of the component metrics of VMAF.
20119 The obtained average motion score is printed through the logging system.
20121 The filter accepts the following options:
20125 If specified, the filter will use the named file to save the motion score of
20126 each frame with respect to the previous frame.
20127 When filename equals "-" the data is sent to standard output.
20132 ffmpeg -i ref.mpg -vf vmafmotion -f null -
20136 Stack input videos vertically.
20138 All streams must be of same pixel format and of same width.
20140 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
20141 to create same output.
20143 The filter accepts the following options:
20147 Set number of input streams. Default is 2.
20150 If set to 1, force the output to terminate when the shortest input
20151 terminates. Default value is 0.
20156 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
20157 Deinterlacing Filter").
20159 Based on the process described by Martin Weston for BBC R&D, and
20160 implemented based on the de-interlace algorithm written by Jim
20161 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
20162 uses filter coefficients calculated by BBC R&D.
20164 This filter uses field-dominance information in frame to decide which
20165 of each pair of fields to place first in the output.
20166 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
20168 There are two sets of filter coefficients, so called "simple"
20169 and "complex". Which set of filter coefficients is used can
20170 be set by passing an optional parameter:
20174 Set the interlacing filter coefficients. Accepts one of the following values:
20178 Simple filter coefficient set.
20180 More-complex filter coefficient set.
20182 Default value is @samp{complex}.
20185 Specify which frames to deinterlace. Accepts one of the following values:
20189 Deinterlace all frames,
20191 Only deinterlace frames marked as interlaced.
20194 Default value is @samp{all}.
20198 Video waveform monitor.
20200 The waveform monitor plots color component intensity. By default luminance
20201 only. Each column of the waveform corresponds to a column of pixels in the
20204 It accepts the following options:
20208 Can be either @code{row}, or @code{column}. Default is @code{column}.
20209 In row mode, the graph on the left side represents color component value 0 and
20210 the right side represents value = 255. In column mode, the top side represents
20211 color component value = 0 and bottom side represents value = 255.
20214 Set intensity. Smaller values are useful to find out how many values of the same
20215 luminance are distributed across input rows/columns.
20216 Default value is @code{0.04}. Allowed range is [0, 1].
20219 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
20220 In mirrored mode, higher values will be represented on the left
20221 side for @code{row} mode and at the top for @code{column} mode. Default is
20222 @code{1} (mirrored).
20226 It accepts the following values:
20229 Presents information identical to that in the @code{parade}, except
20230 that the graphs representing color components are superimposed directly
20233 This display mode makes it easier to spot relative differences or similarities
20234 in overlapping areas of the color components that are supposed to be identical,
20235 such as neutral whites, grays, or blacks.
20238 Display separate graph for the color components side by side in
20239 @code{row} mode or one below the other in @code{column} mode.
20242 Display separate graph for the color components side by side in
20243 @code{column} mode or one below the other in @code{row} mode.
20245 Using this display mode makes it easy to spot color casts in the highlights
20246 and shadows of an image, by comparing the contours of the top and the bottom
20247 graphs of each waveform. Since whites, grays, and blacks are characterized
20248 by exactly equal amounts of red, green, and blue, neutral areas of the picture
20249 should display three waveforms of roughly equal width/height. If not, the
20250 correction is easy to perform by making level adjustments the three waveforms.
20252 Default is @code{stack}.
20254 @item components, c
20255 Set which color components to display. Default is 1, which means only luminance
20256 or red color component if input is in RGB colorspace. If is set for example to
20257 7 it will display all 3 (if) available color components.
20262 No envelope, this is default.
20265 Instant envelope, minimum and maximum values presented in graph will be easily
20266 visible even with small @code{step} value.
20269 Hold minimum and maximum values presented in graph across time. This way you
20270 can still spot out of range values without constantly looking at waveforms.
20273 Peak and instant envelope combined together.
20279 No filtering, this is default.
20282 Luma and chroma combined together.
20285 Similar as above, but shows difference between blue and red chroma.
20288 Similar as above, but use different colors.
20291 Similar as above, but again with different colors.
20294 Displays only chroma.
20297 Displays actual color value on waveform.
20300 Similar as above, but with luma showing frequency of chroma values.
20304 Set which graticule to display.
20308 Do not display graticule.
20311 Display green graticule showing legal broadcast ranges.
20314 Display orange graticule showing legal broadcast ranges.
20317 Display invert graticule showing legal broadcast ranges.
20321 Set graticule opacity.
20324 Set graticule flags.
20328 Draw numbers above lines. By default enabled.
20331 Draw dots instead of lines.
20335 Set scale used for displaying graticule.
20342 Default is digital.
20345 Set background opacity.
20349 Set tint for output.
20350 Only used with lowpass filter and when display is not overlay and input
20351 pixel formats are not RGB.
20354 @section weave, doubleweave
20356 The @code{weave} takes a field-based video input and join
20357 each two sequential fields into single frame, producing a new double
20358 height clip with half the frame rate and half the frame count.
20360 The @code{doubleweave} works same as @code{weave} but without
20361 halving frame rate and frame count.
20363 It accepts the following option:
20367 Set first field. Available values are:
20371 Set the frame as top-field-first.
20374 Set the frame as bottom-field-first.
20378 @subsection Examples
20382 Interlace video using @ref{select} and @ref{separatefields} filter:
20384 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
20389 Apply the xBR high-quality magnification filter which is designed for pixel
20390 art. It follows a set of edge-detection rules, see
20391 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
20393 It accepts the following option:
20397 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
20398 @code{3xBR} and @code{4} for @code{4xBR}.
20399 Default is @code{3}.
20404 Apply cross fade from one input video stream to another input video stream.
20405 The cross fade is applied for specified duration.
20407 The filter accepts the following options:
20411 Set one of available transition effects:
20451 Default transition effect is fade.
20454 Set cross fade duration in seconds.
20455 Default duration is 1 second.
20458 Set cross fade start relative to first input stream in seconds.
20459 Default offset is 0.
20462 Set expression for custom transition effect.
20464 The expressions can use the following variables and functions:
20469 The coordinates of the current sample.
20473 The width and height of the image.
20476 Progress of transition effect.
20479 Currently processed plane.
20482 Return value of first input at current location and plane.
20485 Return value of second input at current location and plane.
20491 Return the value of the pixel at location (@var{x},@var{y}) of the
20492 first/second/third/fourth component of first input.
20498 Return the value of the pixel at location (@var{x},@var{y}) of the
20499 first/second/third/fourth component of second input.
20503 @subsection Examples
20507 Cross fade from one input video to another input video, with fade transition and duration of transition
20508 of 2 seconds starting at offset of 5 seconds:
20510 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
20515 Pick median pixels from several input videos.
20517 The filter accepts the following options:
20521 Set number of inputs.
20522 Default is 3. Allowed range is from 3 to 255.
20523 If number of inputs is even number, than result will be mean value between two median values.
20526 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
20529 Set median percentile. Default value is @code{0.5}.
20530 Default value of @code{0.5} will pick always median values, while @code{0} will pick
20531 minimum values, and @code{1} maximum values.
20535 Stack video inputs into custom layout.
20537 All streams must be of same pixel format.
20539 The filter accepts the following options:
20543 Set number of input streams. Default is 2.
20546 Specify layout of inputs.
20547 This option requires the desired layout configuration to be explicitly set by the user.
20548 This sets position of each video input in output. Each input
20549 is separated by '|'.
20550 The first number represents the column, and the second number represents the row.
20551 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
20552 where X is video input from which to take width or height.
20553 Multiple values can be used when separated by '+'. In such
20554 case values are summed together.
20556 Note that if inputs are of different sizes gaps may appear, as not all of
20557 the output video frame will be filled. Similarly, videos can overlap each
20558 other if their position doesn't leave enough space for the full frame of
20561 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
20562 a layout must be set by the user.
20565 If set to 1, force the output to terminate when the shortest input
20566 terminates. Default value is 0.
20569 If set to valid color, all unused pixels will be filled with that color.
20570 By default fill is set to none, so it is disabled.
20573 @subsection Examples
20577 Display 4 inputs into 2x2 grid.
20581 input1(0, 0) | input3(w0, 0)
20582 input2(0, h0) | input4(w0, h0)
20586 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
20589 Note that if inputs are of different sizes, gaps or overlaps may occur.
20592 Display 4 inputs into 1x4 grid.
20599 input4(0, h0+h1+h2)
20603 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
20606 Note that if inputs are of different widths, unused space will appear.
20609 Display 9 inputs into 3x3 grid.
20613 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
20614 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
20615 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
20619 xstack=inputs=9:layout=0_0|0_h0|0_h0+h1|w0_0|w0_h0|w0_h0+h1|w0+w3_0|w0+w3_h0|w0+w3_h0+h1
20622 Note that if inputs are of different sizes, gaps or overlaps may occur.
20625 Display 16 inputs into 4x4 grid.
20629 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
20630 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
20631 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
20632 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
20636 xstack=inputs=16:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2|w0_0|w0_h0|w0_h0+h1|w0_h0+h1+h2|w0+w4_0|
20637 w0+w4_h0|w0+w4_h0+h1|w0+w4_h0+h1+h2|w0+w4+w8_0|w0+w4+w8_h0|w0+w4+w8_h0+h1|w0+w4+w8_h0+h1+h2
20640 Note that if inputs are of different sizes, gaps or overlaps may occur.
20647 Deinterlace the input video ("yadif" means "yet another deinterlacing
20650 It accepts the following parameters:
20656 The interlacing mode to adopt. It accepts one of the following values:
20659 @item 0, send_frame
20660 Output one frame for each frame.
20661 @item 1, send_field
20662 Output one frame for each field.
20663 @item 2, send_frame_nospatial
20664 Like @code{send_frame}, but it skips the spatial interlacing check.
20665 @item 3, send_field_nospatial
20666 Like @code{send_field}, but it skips the spatial interlacing check.
20669 The default value is @code{send_frame}.
20672 The picture field parity assumed for the input interlaced video. It accepts one
20673 of the following values:
20677 Assume the top field is first.
20679 Assume the bottom field is first.
20681 Enable automatic detection of field parity.
20684 The default value is @code{auto}.
20685 If the interlacing is unknown or the decoder does not export this information,
20686 top field first will be assumed.
20689 Specify which frames to deinterlace. Accepts one of the following
20694 Deinterlace all frames.
20695 @item 1, interlaced
20696 Only deinterlace frames marked as interlaced.
20699 The default value is @code{all}.
20702 @section yadif_cuda
20704 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
20705 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
20708 It accepts the following parameters:
20714 The interlacing mode to adopt. It accepts one of the following values:
20717 @item 0, send_frame
20718 Output one frame for each frame.
20719 @item 1, send_field
20720 Output one frame for each field.
20721 @item 2, send_frame_nospatial
20722 Like @code{send_frame}, but it skips the spatial interlacing check.
20723 @item 3, send_field_nospatial
20724 Like @code{send_field}, but it skips the spatial interlacing check.
20727 The default value is @code{send_frame}.
20730 The picture field parity assumed for the input interlaced video. It accepts one
20731 of the following values:
20735 Assume the top field is first.
20737 Assume the bottom field is first.
20739 Enable automatic detection of field parity.
20742 The default value is @code{auto}.
20743 If the interlacing is unknown or the decoder does not export this information,
20744 top field first will be assumed.
20747 Specify which frames to deinterlace. Accepts one of the following
20752 Deinterlace all frames.
20753 @item 1, interlaced
20754 Only deinterlace frames marked as interlaced.
20757 The default value is @code{all}.
20762 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
20763 The algorithm is described in
20764 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
20766 It accepts the following parameters:
20770 Set the window radius. Default value is 3.
20773 Set which planes to filter. Default is only the first plane.
20776 Set blur strength. Default value is 128.
20779 @subsection Commands
20780 This filter supports same @ref{commands} as options.
20784 Apply Zoom & Pan effect.
20786 This filter accepts the following options:
20790 Set the zoom expression. Range is 1-10. Default is 1.
20794 Set the x and y expression. Default is 0.
20797 Set the duration expression in number of frames.
20798 This sets for how many number of frames effect will last for
20799 single input image.
20802 Set the output image size, default is 'hd720'.
20805 Set the output frame rate, default is '25'.
20808 Each expression can contain the following constants:
20827 Output frame count.
20830 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
20832 @item out_time, time, ot
20833 The output timestamp expressed in seconds.
20837 Last calculated 'x' and 'y' position from 'x' and 'y' expression
20838 for current input frame.
20842 'x' and 'y' of last output frame of previous input frame or 0 when there was
20843 not yet such frame (first input frame).
20846 Last calculated zoom from 'z' expression for current input frame.
20849 Last calculated zoom of last output frame of previous input frame.
20852 Number of output frames for current input frame. Calculated from 'd' expression
20853 for each input frame.
20856 number of output frames created for previous input frame
20859 Rational number: input width / input height
20862 sample aspect ratio
20865 display aspect ratio
20869 @subsection Examples
20873 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
20875 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
20879 Zoom in up to 1.5x and pan always at center of picture:
20881 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20885 Same as above but without pausing:
20887 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20891 Zoom in 2x into center of picture only for the first second of the input video:
20893 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20900 Scale (resize) the input video, using the z.lib library:
20901 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
20902 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
20904 The zscale filter forces the output display aspect ratio to be the same
20905 as the input, by changing the output sample aspect ratio.
20907 If the input image format is different from the format requested by
20908 the next filter, the zscale filter will convert the input to the
20911 @subsection Options
20912 The filter accepts the following options.
20917 Set the output video dimension expression. Default value is the input
20920 If the @var{width} or @var{w} value is 0, the input width is used for
20921 the output. If the @var{height} or @var{h} value is 0, the input height
20922 is used for the output.
20924 If one and only one of the values is -n with n >= 1, the zscale filter
20925 will use a value that maintains the aspect ratio of the input image,
20926 calculated from the other specified dimension. After that it will,
20927 however, make sure that the calculated dimension is divisible by n and
20928 adjust the value if necessary.
20930 If both values are -n with n >= 1, the behavior will be identical to
20931 both values being set to 0 as previously detailed.
20933 See below for the list of accepted constants for use in the dimension
20937 Set the video size. For the syntax of this option, check the
20938 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20941 Set the dither type.
20943 Possible values are:
20948 @item error_diffusion
20954 Set the resize filter type.
20956 Possible values are:
20966 Default is bilinear.
20969 Set the color range.
20971 Possible values are:
20978 Default is same as input.
20981 Set the color primaries.
20983 Possible values are:
20993 Default is same as input.
20996 Set the transfer characteristics.
20998 Possible values are:
21012 Default is same as input.
21015 Set the colorspace matrix.
21017 Possible value are:
21028 Default is same as input.
21031 Set the input color range.
21033 Possible values are:
21040 Default is same as input.
21042 @item primariesin, pin
21043 Set the input color primaries.
21045 Possible values are:
21055 Default is same as input.
21057 @item transferin, tin
21058 Set the input transfer characteristics.
21060 Possible values are:
21071 Default is same as input.
21073 @item matrixin, min
21074 Set the input colorspace matrix.
21076 Possible value are:
21088 Set the output chroma location.
21090 Possible values are:
21101 @item chromalin, cin
21102 Set the input chroma location.
21104 Possible values are:
21116 Set the nominal peak luminance.
21119 The values of the @option{w} and @option{h} options are expressions
21120 containing the following constants:
21125 The input width and height
21129 These are the same as @var{in_w} and @var{in_h}.
21133 The output (scaled) width and height
21137 These are the same as @var{out_w} and @var{out_h}
21140 The same as @var{iw} / @var{ih}
21143 input sample aspect ratio
21146 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
21150 horizontal and vertical input chroma subsample values. For example for the
21151 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
21155 horizontal and vertical output chroma subsample values. For example for the
21156 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
21159 @subsection Commands
21161 This filter supports the following commands:
21165 Set the output video dimension expression.
21166 The command accepts the same syntax of the corresponding option.
21168 If the specified expression is not valid, it is kept at its current
21172 @c man end VIDEO FILTERS
21174 @chapter OpenCL Video Filters
21175 @c man begin OPENCL VIDEO FILTERS
21177 Below is a description of the currently available OpenCL video filters.
21179 To enable compilation of these filters you need to configure FFmpeg with
21180 @code{--enable-opencl}.
21182 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
21185 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
21186 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
21187 given device parameters.
21189 @item -filter_hw_device @var{name}
21190 Pass the hardware device called @var{name} to all filters in any filter graph.
21194 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
21198 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
21200 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
21204 Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(@ref{hwupload}) to hardware surfaces connected to the appropriate device before being used and then downloaded(@ref{hwdownload}) back to normal memory. Note that @ref{hwupload} will upload to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before to get the input into the right format and @ref{hwdownload} does not support all formats on the output - it may be necessary to insert an additional @ref{format} filter immediately following in the graph to get the output in a supported format.
21206 @section avgblur_opencl
21208 Apply average blur filter.
21210 The filter accepts the following options:
21214 Set horizontal radius size.
21215 Range is @code{[1, 1024]} and default value is @code{1}.
21218 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21221 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
21224 @subsection Example
21228 Apply average blur filter with horizontal and vertical size of 3, setting each pixel of the output to the average value of the 7x7 region centered on it in the input. For pixels on the edges of the image, the region does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
21230 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
21234 @section boxblur_opencl
21236 Apply a boxblur algorithm to the input video.
21238 It accepts the following parameters:
21242 @item luma_radius, lr
21243 @item luma_power, lp
21244 @item chroma_radius, cr
21245 @item chroma_power, cp
21246 @item alpha_radius, ar
21247 @item alpha_power, ap
21251 A description of the accepted options follows.
21254 @item luma_radius, lr
21255 @item chroma_radius, cr
21256 @item alpha_radius, ar
21257 Set an expression for the box radius in pixels used for blurring the
21258 corresponding input plane.
21260 The radius value must be a non-negative number, and must not be
21261 greater than the value of the expression @code{min(w,h)/2} for the
21262 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
21265 Default value for @option{luma_radius} is "2". If not specified,
21266 @option{chroma_radius} and @option{alpha_radius} default to the
21267 corresponding value set for @option{luma_radius}.
21269 The expressions can contain the following constants:
21273 The input width and height in pixels.
21277 The input chroma image width and height in pixels.
21281 The horizontal and vertical chroma subsample values. For example, for the
21282 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
21285 @item luma_power, lp
21286 @item chroma_power, cp
21287 @item alpha_power, ap
21288 Specify how many times the boxblur filter is applied to the
21289 corresponding plane.
21291 Default value for @option{luma_power} is 2. If not specified,
21292 @option{chroma_power} and @option{alpha_power} default to the
21293 corresponding value set for @option{luma_power}.
21295 A value of 0 will disable the effect.
21298 @subsection Examples
21300 Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses @var{luma_radius}, @var{chroma_radius}, @var{alpha_radius} for each plane respectively. The filter will apply @var{luma_power}, @var{chroma_power}, @var{alpha_power} times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
21304 Apply a boxblur filter with the luma, chroma, and alpha radius
21305 set to 2 and luma, chroma, and alpha power set to 3. The filter will run 3 times with box-radius set to 2 for every plane of the image.
21307 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
21308 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
21312 Apply a boxblur filter with luma radius set to 2, luma_power to 1, chroma_radius to 4, chroma_power to 5, alpha_radius to 3 and alpha_power to 7.
21314 For the luma plane, a 2x2 box radius will be run once.
21316 For the chroma plane, a 4x4 box radius will be run 5 times.
21318 For the alpha plane, a 3x3 box radius will be run 7 times.
21320 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
21324 @section colorkey_opencl
21325 RGB colorspace color keying.
21327 The filter accepts the following options:
21331 The color which will be replaced with transparency.
21334 Similarity percentage with the key color.
21336 0.01 matches only the exact key color, while 1.0 matches everything.
21341 0.0 makes pixels either fully transparent, or not transparent at all.
21343 Higher values result in semi-transparent pixels, with a higher transparency
21344 the more similar the pixels color is to the key color.
21347 @subsection Examples
21351 Make every semi-green pixel in the input transparent with some slight blending:
21353 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
21357 @section convolution_opencl
21359 Apply convolution of 3x3, 5x5, 7x7 matrix.
21361 The filter accepts the following options:
21368 Set matrix for each plane.
21369 Matrix is sequence of 9, 25 or 49 signed numbers.
21370 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
21376 Set multiplier for calculated value for each plane.
21377 If unset or 0, it will be sum of all matrix elements.
21378 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
21384 Set bias for each plane. This value is added to the result of the multiplication.
21385 Useful for making the overall image brighter or darker.
21386 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
21390 @subsection Examples
21396 -i INPUT -vf "hwupload, convolution_opencl=0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0, hwdownload" OUTPUT
21402 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9, hwdownload" OUTPUT
21406 Apply edge enhance:
21408 -i INPUT -vf "hwupload, convolution_opencl=0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128, hwdownload" OUTPUT
21414 -i INPUT -vf "hwupload, convolution_opencl=0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128, hwdownload" OUTPUT
21418 Apply laplacian edge detector which includes diagonals:
21420 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
21426 -i INPUT -vf "hwupload, convolution_opencl=-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2, hwdownload" OUTPUT
21430 @section erosion_opencl
21432 Apply erosion effect to the video.
21434 This filter replaces the pixel by the local(3x3) minimum.
21436 It accepts the following options:
21443 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
21444 If @code{0}, plane will remain unchanged.
21447 Flag which specifies the pixel to refer to.
21448 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
21450 Flags to local 3x3 coordinates region centered on @code{x}:
21459 @subsection Example
21463 Apply erosion filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local minimum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local minimum is more then threshold of the corresponding plane, output pixel will be set to input pixel - threshold of corresponding plane.
21465 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
21469 @section deshake_opencl
21470 Feature-point based video stabilization filter.
21472 The filter accepts the following options:
21476 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
21479 Whether or not additional debug info should be displayed, both in the processed output and in the console.
21481 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
21483 Viewing point matches in the output video is only supported for RGB input.
21485 Defaults to @code{0}.
21487 @item adaptive_crop
21488 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
21490 Defaults to @code{1}.
21492 @item refine_features
21493 Whether or not feature points should be refined at a sub-pixel level.
21495 This can be turned off for a slight performance gain at the cost of precision.
21497 Defaults to @code{1}.
21499 @item smooth_strength
21500 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
21502 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
21504 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
21506 Defaults to @code{0.0}.
21508 @item smooth_window_multiplier
21509 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
21511 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
21513 Acceptable values range from @code{0.1} to @code{10.0}.
21515 Larger values increase the amount of motion data available for determining how to smooth the camera path,
21516 potentially improving smoothness, but also increase latency and memory usage.
21518 Defaults to @code{2.0}.
21522 @subsection Examples
21526 Stabilize a video with a fixed, medium smoothing strength:
21528 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
21532 Stabilize a video with debugging (both in console and in rendered video):
21534 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
21538 @section dilation_opencl
21540 Apply dilation effect to the video.
21542 This filter replaces the pixel by the local(3x3) maximum.
21544 It accepts the following options:
21551 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
21552 If @code{0}, plane will remain unchanged.
21555 Flag which specifies the pixel to refer to.
21556 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
21558 Flags to local 3x3 coordinates region centered on @code{x}:
21567 @subsection Example
21571 Apply dilation filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local maximum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local maximum is more then threshold of the corresponding plane, output pixel will be set to input pixel + threshold of corresponding plane.
21573 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
21577 @section nlmeans_opencl
21579 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
21581 @section overlay_opencl
21583 Overlay one video on top of another.
21585 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
21586 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
21588 The filter accepts the following options:
21593 Set the x coordinate of the overlaid video on the main video.
21594 Default value is @code{0}.
21597 Set the y coordinate of the overlaid video on the main video.
21598 Default value is @code{0}.
21602 @subsection Examples
21606 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
21608 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
21611 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
21613 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
21618 @section pad_opencl
21620 Add paddings to the input image, and place the original input at the
21621 provided @var{x}, @var{y} coordinates.
21623 It accepts the following options:
21628 Specify an expression for the size of the output image with the
21629 paddings added. If the value for @var{width} or @var{height} is 0, the
21630 corresponding input size is used for the output.
21632 The @var{width} expression can reference the value set by the
21633 @var{height} expression, and vice versa.
21635 The default value of @var{width} and @var{height} is 0.
21639 Specify the offsets to place the input image at within the padded area,
21640 with respect to the top/left border of the output image.
21642 The @var{x} expression can reference the value set by the @var{y}
21643 expression, and vice versa.
21645 The default value of @var{x} and @var{y} is 0.
21647 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
21648 so the input image is centered on the padded area.
21651 Specify the color of the padded area. For the syntax of this option,
21652 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
21653 manual,ffmpeg-utils}.
21656 Pad to an aspect instead to a resolution.
21659 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
21660 options are expressions containing the following constants:
21665 The input video width and height.
21669 These are the same as @var{in_w} and @var{in_h}.
21673 The output width and height (the size of the padded area), as
21674 specified by the @var{width} and @var{height} expressions.
21678 These are the same as @var{out_w} and @var{out_h}.
21682 The x and y offsets as specified by the @var{x} and @var{y}
21683 expressions, or NAN if not yet specified.
21686 same as @var{iw} / @var{ih}
21689 input sample aspect ratio
21692 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
21695 @section prewitt_opencl
21697 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
21699 The filter accepts the following option:
21703 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21706 Set value which will be multiplied with filtered result.
21707 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21710 Set value which will be added to filtered result.
21711 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21714 @subsection Example
21718 Apply the Prewitt operator with scale set to 2 and delta set to 10.
21720 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
21724 @anchor{program_opencl}
21725 @section program_opencl
21727 Filter video using an OpenCL program.
21732 OpenCL program source file.
21735 Kernel name in program.
21738 Number of inputs to the filter. Defaults to 1.
21741 Size of output frames. Defaults to the same as the first input.
21745 The @code{program_opencl} filter also supports the @ref{framesync} options.
21747 The program source file must contain a kernel function with the given name,
21748 which will be run once for each plane of the output. Each run on a plane
21749 gets enqueued as a separate 2D global NDRange with one work-item for each
21750 pixel to be generated. The global ID offset for each work-item is therefore
21751 the coordinates of a pixel in the destination image.
21753 The kernel function needs to take the following arguments:
21756 Destination image, @var{__write_only image2d_t}.
21758 This image will become the output; the kernel should write all of it.
21760 Frame index, @var{unsigned int}.
21762 This is a counter starting from zero and increasing by one for each frame.
21764 Source images, @var{__read_only image2d_t}.
21766 These are the most recent images on each input. The kernel may read from
21767 them to generate the output, but they can't be written to.
21774 Copy the input to the output (output must be the same size as the input).
21776 __kernel void copy(__write_only image2d_t destination,
21777 unsigned int index,
21778 __read_only image2d_t source)
21780 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
21782 int2 location = (int2)(get_global_id(0), get_global_id(1));
21784 float4 value = read_imagef(source, sampler, location);
21786 write_imagef(destination, location, value);
21791 Apply a simple transformation, rotating the input by an amount increasing
21792 with the index counter. Pixel values are linearly interpolated by the
21793 sampler, and the output need not have the same dimensions as the input.
21795 __kernel void rotate_image(__write_only image2d_t dst,
21796 unsigned int index,
21797 __read_only image2d_t src)
21799 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
21800 CLK_FILTER_LINEAR);
21802 float angle = (float)index / 100.0f;
21804 float2 dst_dim = convert_float2(get_image_dim(dst));
21805 float2 src_dim = convert_float2(get_image_dim(src));
21807 float2 dst_cen = dst_dim / 2.0f;
21808 float2 src_cen = src_dim / 2.0f;
21810 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
21812 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
21814 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
21815 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
21817 src_pos = src_pos * src_dim / dst_dim;
21819 float2 src_loc = src_pos + src_cen;
21821 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
21822 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
21823 write_imagef(dst, dst_loc, 0.5f);
21825 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
21830 Blend two inputs together, with the amount of each input used varying
21831 with the index counter.
21833 __kernel void blend_images(__write_only image2d_t dst,
21834 unsigned int index,
21835 __read_only image2d_t src1,
21836 __read_only image2d_t src2)
21838 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
21839 CLK_FILTER_LINEAR);
21841 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
21843 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
21844 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
21845 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
21847 float4 val1 = read_imagef(src1, sampler, src1_loc);
21848 float4 val2 = read_imagef(src2, sampler, src2_loc);
21850 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
21856 @section roberts_opencl
21857 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
21859 The filter accepts the following option:
21863 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21866 Set value which will be multiplied with filtered result.
21867 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21870 Set value which will be added to filtered result.
21871 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21874 @subsection Example
21878 Apply the Roberts cross operator with scale set to 2 and delta set to 10
21880 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
21884 @section sobel_opencl
21886 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
21888 The filter accepts the following option:
21892 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21895 Set value which will be multiplied with filtered result.
21896 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21899 Set value which will be added to filtered result.
21900 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21903 @subsection Example
21907 Apply sobel operator with scale set to 2 and delta set to 10
21909 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
21913 @section tonemap_opencl
21915 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
21917 It accepts the following parameters:
21921 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
21924 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
21927 Apply desaturation for highlights that exceed this level of brightness. The
21928 higher the parameter, the more color information will be preserved. This
21929 setting helps prevent unnaturally blown-out colors for super-highlights, by
21930 (smoothly) turning into white instead. This makes images feel more natural,
21931 at the cost of reducing information about out-of-range colors.
21933 The default value is 0.5, and the algorithm here is a little different from
21934 the cpu version tonemap currently. A setting of 0.0 disables this option.
21937 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
21938 is used to detect whether the scene has changed or not. If the distance between
21939 the current frame average brightness and the current running average exceeds
21940 a threshold value, we would re-calculate scene average and peak brightness.
21941 The default value is 0.2.
21944 Specify the output pixel format.
21946 Currently supported formats are:
21953 Set the output color range.
21955 Possible values are:
21961 Default is same as input.
21964 Set the output color primaries.
21966 Possible values are:
21972 Default is same as input.
21975 Set the output transfer characteristics.
21977 Possible values are:
21986 Set the output colorspace matrix.
21988 Possible value are:
21994 Default is same as input.
21998 @subsection Example
22002 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
22004 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
22008 @section unsharp_opencl
22010 Sharpen or blur the input video.
22012 It accepts the following parameters:
22015 @item luma_msize_x, lx
22016 Set the luma matrix horizontal size.
22017 Range is @code{[1, 23]} and default value is @code{5}.
22019 @item luma_msize_y, ly
22020 Set the luma matrix vertical size.
22021 Range is @code{[1, 23]} and default value is @code{5}.
22023 @item luma_amount, la
22024 Set the luma effect strength.
22025 Range is @code{[-10, 10]} and default value is @code{1.0}.
22027 Negative values will blur the input video, while positive values will
22028 sharpen it, a value of zero will disable the effect.
22030 @item chroma_msize_x, cx
22031 Set the chroma matrix horizontal size.
22032 Range is @code{[1, 23]} and default value is @code{5}.
22034 @item chroma_msize_y, cy
22035 Set the chroma matrix vertical size.
22036 Range is @code{[1, 23]} and default value is @code{5}.
22038 @item chroma_amount, ca
22039 Set the chroma effect strength.
22040 Range is @code{[-10, 10]} and default value is @code{0.0}.
22042 Negative values will blur the input video, while positive values will
22043 sharpen it, a value of zero will disable the effect.
22047 All parameters are optional and default to the equivalent of the
22048 string '5:5:1.0:5:5:0.0'.
22050 @subsection Examples
22054 Apply strong luma sharpen effect:
22056 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
22060 Apply a strong blur of both luma and chroma parameters:
22062 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
22066 @section xfade_opencl
22068 Cross fade two videos with custom transition effect by using OpenCL.
22070 It accepts the following options:
22074 Set one of possible transition effects.
22078 Select custom transition effect, the actual transition description
22079 will be picked from source and kernel options.
22091 Default transition is fade.
22095 OpenCL program source file for custom transition.
22098 Set name of kernel to use for custom transition from program source file.
22101 Set duration of video transition.
22104 Set time of start of transition relative to first video.
22107 The program source file must contain a kernel function with the given name,
22108 which will be run once for each plane of the output. Each run on a plane
22109 gets enqueued as a separate 2D global NDRange with one work-item for each
22110 pixel to be generated. The global ID offset for each work-item is therefore
22111 the coordinates of a pixel in the destination image.
22113 The kernel function needs to take the following arguments:
22116 Destination image, @var{__write_only image2d_t}.
22118 This image will become the output; the kernel should write all of it.
22121 First Source image, @var{__read_only image2d_t}.
22122 Second Source image, @var{__read_only image2d_t}.
22124 These are the most recent images on each input. The kernel may read from
22125 them to generate the output, but they can't be written to.
22128 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
22135 Apply dots curtain transition effect:
22137 __kernel void blend_images(__write_only image2d_t dst,
22138 __read_only image2d_t src1,
22139 __read_only image2d_t src2,
22142 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22143 CLK_FILTER_LINEAR);
22144 int2 p = (int2)(get_global_id(0), get_global_id(1));
22145 float2 rp = (float2)(get_global_id(0), get_global_id(1));
22146 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
22149 float2 dots = (float2)(20.0, 20.0);
22150 float2 center = (float2)(0,0);
22153 float4 val1 = read_imagef(src1, sampler, p);
22154 float4 val2 = read_imagef(src2, sampler, p);
22155 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
22157 write_imagef(dst, p, next ? val1 : val2);
22163 @c man end OPENCL VIDEO FILTERS
22165 @chapter VAAPI Video Filters
22166 @c man begin VAAPI VIDEO FILTERS
22168 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
22170 To enable compilation of these filters you need to configure FFmpeg with
22171 @code{--enable-vaapi}.
22173 To use vaapi filters, you need to setup the vaapi device correctly. For more information, please read @url{https://trac.ffmpeg.org/wiki/Hardware/VAAPI}
22175 @section tonemap_vaapi
22177 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
22178 It maps the dynamic range of HDR10 content to the SDR content.
22179 It currently only accepts HDR10 as input.
22181 It accepts the following parameters:
22185 Specify the output pixel format.
22187 Currently supported formats are:
22196 Set the output color primaries.
22198 Default is same as input.
22201 Set the output transfer characteristics.
22206 Set the output colorspace matrix.
22208 Default is same as input.
22212 @subsection Example
22216 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
22218 tonemap_vaapi=format=p010:t=bt2020-10
22222 @c man end VAAPI VIDEO FILTERS
22224 @chapter Video Sources
22225 @c man begin VIDEO SOURCES
22227 Below is a description of the currently available video sources.
22231 Buffer video frames, and make them available to the filter chain.
22233 This source is mainly intended for a programmatic use, in particular
22234 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
22236 It accepts the following parameters:
22241 Specify the size (width and height) of the buffered video frames. For the
22242 syntax of this option, check the
22243 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22246 The input video width.
22249 The input video height.
22252 A string representing the pixel format of the buffered video frames.
22253 It may be a number corresponding to a pixel format, or a pixel format
22257 Specify the timebase assumed by the timestamps of the buffered frames.
22260 Specify the frame rate expected for the video stream.
22262 @item pixel_aspect, sar
22263 The sample (pixel) aspect ratio of the input video.
22266 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
22267 to the filtergraph description to specify swscale flags for automatically
22268 inserted scalers. See @ref{Filtergraph syntax}.
22270 @item hw_frames_ctx
22271 When using a hardware pixel format, this should be a reference to an
22272 AVHWFramesContext describing input frames.
22277 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
22280 will instruct the source to accept video frames with size 320x240 and
22281 with format "yuv410p", assuming 1/24 as the timestamps timebase and
22282 square pixels (1:1 sample aspect ratio).
22283 Since the pixel format with name "yuv410p" corresponds to the number 6
22284 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
22285 this example corresponds to:
22287 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
22290 Alternatively, the options can be specified as a flat string, but this
22291 syntax is deprecated:
22293 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
22297 Create a pattern generated by an elementary cellular automaton.
22299 The initial state of the cellular automaton can be defined through the
22300 @option{filename} and @option{pattern} options. If such options are
22301 not specified an initial state is created randomly.
22303 At each new frame a new row in the video is filled with the result of
22304 the cellular automaton next generation. The behavior when the whole
22305 frame is filled is defined by the @option{scroll} option.
22307 This source accepts the following options:
22311 Read the initial cellular automaton state, i.e. the starting row, from
22312 the specified file.
22313 In the file, each non-whitespace character is considered an alive
22314 cell, a newline will terminate the row, and further characters in the
22315 file will be ignored.
22318 Read the initial cellular automaton state, i.e. the starting row, from
22319 the specified string.
22321 Each non-whitespace character in the string is considered an alive
22322 cell, a newline will terminate the row, and further characters in the
22323 string will be ignored.
22326 Set the video rate, that is the number of frames generated per second.
22329 @item random_fill_ratio, ratio
22330 Set the random fill ratio for the initial cellular automaton row. It
22331 is a floating point number value ranging from 0 to 1, defaults to
22334 This option is ignored when a file or a pattern is specified.
22336 @item random_seed, seed
22337 Set the seed for filling randomly the initial row, must be an integer
22338 included between 0 and UINT32_MAX. If not specified, or if explicitly
22339 set to -1, the filter will try to use a good random seed on a best
22343 Set the cellular automaton rule, it is a number ranging from 0 to 255.
22344 Default value is 110.
22347 Set the size of the output video. For the syntax of this option, check the
22348 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22350 If @option{filename} or @option{pattern} is specified, the size is set
22351 by default to the width of the specified initial state row, and the
22352 height is set to @var{width} * PHI.
22354 If @option{size} is set, it must contain the width of the specified
22355 pattern string, and the specified pattern will be centered in the
22358 If a filename or a pattern string is not specified, the size value
22359 defaults to "320x518" (used for a randomly generated initial state).
22362 If set to 1, scroll the output upward when all the rows in the output
22363 have been already filled. If set to 0, the new generated row will be
22364 written over the top row just after the bottom row is filled.
22367 @item start_full, full
22368 If set to 1, completely fill the output with generated rows before
22369 outputting the first frame.
22370 This is the default behavior, for disabling set the value to 0.
22373 If set to 1, stitch the left and right row edges together.
22374 This is the default behavior, for disabling set the value to 0.
22377 @subsection Examples
22381 Read the initial state from @file{pattern}, and specify an output of
22384 cellauto=f=pattern:s=200x400
22388 Generate a random initial row with a width of 200 cells, with a fill
22391 cellauto=ratio=2/3:s=200x200
22395 Create a pattern generated by rule 18 starting by a single alive cell
22396 centered on an initial row with width 100:
22398 cellauto=p=@@:s=100x400:full=0:rule=18
22402 Specify a more elaborated initial pattern:
22404 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
22409 @anchor{coreimagesrc}
22410 @section coreimagesrc
22411 Video source generated on GPU using Apple's CoreImage API on OSX.
22413 This video source is a specialized version of the @ref{coreimage} video filter.
22414 Use a core image generator at the beginning of the applied filterchain to
22415 generate the content.
22417 The coreimagesrc video source accepts the following options:
22419 @item list_generators
22420 List all available generators along with all their respective options as well as
22421 possible minimum and maximum values along with the default values.
22423 list_generators=true
22427 Specify the size of the sourced video. For the syntax of this option, check the
22428 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22429 The default value is @code{320x240}.
22432 Specify the frame rate of the sourced video, as the number of frames
22433 generated per second. It has to be a string in the format
22434 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22435 number or a valid video frame rate abbreviation. The default value is
22439 Set the sample aspect ratio of the sourced video.
22442 Set the duration of the sourced video. See
22443 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22444 for the accepted syntax.
22446 If not specified, or the expressed duration is negative, the video is
22447 supposed to be generated forever.
22450 Additionally, all options of the @ref{coreimage} video filter are accepted.
22451 A complete filterchain can be used for further processing of the
22452 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
22453 and examples for details.
22455 @subsection Examples
22460 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
22461 given as complete and escaped command-line for Apple's standard bash shell:
22463 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
22465 This example is equivalent to the QRCode example of @ref{coreimage} without the
22466 need for a nullsrc video source.
22471 Generate several gradients.
22475 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
22476 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
22479 Set frame rate, expressed as number of frames per second. Default
22482 @item c0, c1, c2, c3, c4, c5, c6, c7
22483 Set 8 colors. Default values for colors is to pick random one.
22485 @item x0, y0, y0, y1
22486 Set gradient line source and destination points. If negative or out of range, random ones
22490 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
22493 Set seed for picking gradient line points.
22497 @section mandelbrot
22499 Generate a Mandelbrot set fractal, and progressively zoom towards the
22500 point specified with @var{start_x} and @var{start_y}.
22502 This source accepts the following options:
22507 Set the terminal pts value. Default value is 400.
22510 Set the terminal scale value.
22511 Must be a floating point value. Default value is 0.3.
22514 Set the inner coloring mode, that is the algorithm used to draw the
22515 Mandelbrot fractal internal region.
22517 It shall assume one of the following values:
22522 Show time until convergence.
22524 Set color based on point closest to the origin of the iterations.
22529 Default value is @var{mincol}.
22532 Set the bailout value. Default value is 10.0.
22535 Set the maximum of iterations performed by the rendering
22536 algorithm. Default value is 7189.
22539 Set outer coloring mode.
22540 It shall assume one of following values:
22542 @item iteration_count
22543 Set iteration count mode.
22544 @item normalized_iteration_count
22545 set normalized iteration count mode.
22547 Default value is @var{normalized_iteration_count}.
22550 Set frame rate, expressed as number of frames per second. Default
22554 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
22555 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
22558 Set the initial scale value. Default value is 3.0.
22561 Set the initial x position. Must be a floating point value between
22562 -100 and 100. Default value is -0.743643887037158704752191506114774.
22565 Set the initial y position. Must be a floating point value between
22566 -100 and 100. Default value is -0.131825904205311970493132056385139.
22571 Generate various test patterns, as generated by the MPlayer test filter.
22573 The size of the generated video is fixed, and is 256x256.
22574 This source is useful in particular for testing encoding features.
22576 This source accepts the following options:
22581 Specify the frame rate of the sourced video, as the number of frames
22582 generated per second. It has to be a string in the format
22583 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22584 number or a valid video frame rate abbreviation. The default value is
22588 Set the duration of the sourced video. See
22589 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22590 for the accepted syntax.
22592 If not specified, or the expressed duration is negative, the video is
22593 supposed to be generated forever.
22597 Set the number or the name of the test to perform. Supported tests are:
22611 @item max_frames, m
22612 Set the maximum number of frames generated for each test, default value is 30.
22616 Default value is "all", which will cycle through the list of all tests.
22621 mptestsrc=t=dc_luma
22624 will generate a "dc_luma" test pattern.
22626 @section frei0r_src
22628 Provide a frei0r source.
22630 To enable compilation of this filter you need to install the frei0r
22631 header and configure FFmpeg with @code{--enable-frei0r}.
22633 This source accepts the following parameters:
22638 The size of the video to generate. For the syntax of this option, check the
22639 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22642 The framerate of the generated video. It may be a string of the form
22643 @var{num}/@var{den} or a frame rate abbreviation.
22646 The name to the frei0r source to load. For more information regarding frei0r and
22647 how to set the parameters, read the @ref{frei0r} section in the video filters
22650 @item filter_params
22651 A '|'-separated list of parameters to pass to the frei0r source.
22655 For example, to generate a frei0r partik0l source with size 200x200
22656 and frame rate 10 which is overlaid on the overlay filter main input:
22658 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
22663 Generate a life pattern.
22665 This source is based on a generalization of John Conway's life game.
22667 The sourced input represents a life grid, each pixel represents a cell
22668 which can be in one of two possible states, alive or dead. Every cell
22669 interacts with its eight neighbours, which are the cells that are
22670 horizontally, vertically, or diagonally adjacent.
22672 At each interaction the grid evolves according to the adopted rule,
22673 which specifies the number of neighbor alive cells which will make a
22674 cell stay alive or born. The @option{rule} option allows one to specify
22677 This source accepts the following options:
22681 Set the file from which to read the initial grid state. In the file,
22682 each non-whitespace character is considered an alive cell, and newline
22683 is used to delimit the end of each row.
22685 If this option is not specified, the initial grid is generated
22689 Set the video rate, that is the number of frames generated per second.
22692 @item random_fill_ratio, ratio
22693 Set the random fill ratio for the initial random grid. It is a
22694 floating point number value ranging from 0 to 1, defaults to 1/PHI.
22695 It is ignored when a file is specified.
22697 @item random_seed, seed
22698 Set the seed for filling the initial random grid, must be an integer
22699 included between 0 and UINT32_MAX. If not specified, or if explicitly
22700 set to -1, the filter will try to use a good random seed on a best
22706 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
22707 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
22708 @var{NS} specifies the number of alive neighbor cells which make a
22709 live cell stay alive, and @var{NB} the number of alive neighbor cells
22710 which make a dead cell to become alive (i.e. to "born").
22711 "s" and "b" can be used in place of "S" and "B", respectively.
22713 Alternatively a rule can be specified by an 18-bits integer. The 9
22714 high order bits are used to encode the next cell state if it is alive
22715 for each number of neighbor alive cells, the low order bits specify
22716 the rule for "borning" new cells. Higher order bits encode for an
22717 higher number of neighbor cells.
22718 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
22719 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
22721 Default value is "S23/B3", which is the original Conway's game of life
22722 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
22723 cells, and will born a new cell if there are three alive cells around
22727 Set the size of the output video. For the syntax of this option, check the
22728 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22730 If @option{filename} is specified, the size is set by default to the
22731 same size of the input file. If @option{size} is set, it must contain
22732 the size specified in the input file, and the initial grid defined in
22733 that file is centered in the larger resulting area.
22735 If a filename is not specified, the size value defaults to "320x240"
22736 (used for a randomly generated initial grid).
22739 If set to 1, stitch the left and right grid edges together, and the
22740 top and bottom edges also. Defaults to 1.
22743 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
22744 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
22745 value from 0 to 255.
22748 Set the color of living (or new born) cells.
22751 Set the color of dead cells. If @option{mold} is set, this is the first color
22752 used to represent a dead cell.
22755 Set mold color, for definitely dead and moldy cells.
22757 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
22758 ffmpeg-utils manual,ffmpeg-utils}.
22761 @subsection Examples
22765 Read a grid from @file{pattern}, and center it on a grid of size
22768 life=f=pattern:s=300x300
22772 Generate a random grid of size 200x200, with a fill ratio of 2/3:
22774 life=ratio=2/3:s=200x200
22778 Specify a custom rule for evolving a randomly generated grid:
22784 Full example with slow death effect (mold) using @command{ffplay}:
22786 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
22793 @anchor{haldclutsrc}
22796 @anchor{pal100bars}
22797 @anchor{rgbtestsrc}
22799 @anchor{smptehdbars}
22802 @anchor{yuvtestsrc}
22803 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
22805 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
22807 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
22809 The @code{color} source provides an uniformly colored input.
22811 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
22812 @ref{haldclut} filter.
22814 The @code{nullsrc} source returns unprocessed video frames. It is
22815 mainly useful to be employed in analysis / debugging tools, or as the
22816 source for filters which ignore the input data.
22818 The @code{pal75bars} source generates a color bars pattern, based on
22819 EBU PAL recommendations with 75% color levels.
22821 The @code{pal100bars} source generates a color bars pattern, based on
22822 EBU PAL recommendations with 100% color levels.
22824 The @code{rgbtestsrc} source generates an RGB test pattern useful for
22825 detecting RGB vs BGR issues. You should see a red, green and blue
22826 stripe from top to bottom.
22828 The @code{smptebars} source generates a color bars pattern, based on
22829 the SMPTE Engineering Guideline EG 1-1990.
22831 The @code{smptehdbars} source generates a color bars pattern, based on
22832 the SMPTE RP 219-2002.
22834 The @code{testsrc} source generates a test video pattern, showing a
22835 color pattern, a scrolling gradient and a timestamp. This is mainly
22836 intended for testing purposes.
22838 The @code{testsrc2} source is similar to testsrc, but supports more
22839 pixel formats instead of just @code{rgb24}. This allows using it as an
22840 input for other tests without requiring a format conversion.
22842 The @code{yuvtestsrc} source generates an YUV test pattern. You should
22843 see a y, cb and cr stripe from top to bottom.
22845 The sources accept the following parameters:
22850 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
22851 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
22852 pixels to be used as identity matrix for 3D lookup tables. Each component is
22853 coded on a @code{1/(N*N)} scale.
22856 Specify the color of the source, only available in the @code{color}
22857 source. For the syntax of this option, check the
22858 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
22861 Specify the size of the sourced video. For the syntax of this option, check the
22862 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22863 The default value is @code{320x240}.
22865 This option is not available with the @code{allrgb}, @code{allyuv}, and
22866 @code{haldclutsrc} filters.
22869 Specify the frame rate of the sourced video, as the number of frames
22870 generated per second. It has to be a string in the format
22871 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22872 number or a valid video frame rate abbreviation. The default value is
22876 Set the duration of the sourced video. See
22877 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22878 for the accepted syntax.
22880 If not specified, or the expressed duration is negative, the video is
22881 supposed to be generated forever.
22884 Set the sample aspect ratio of the sourced video.
22887 Specify the alpha (opacity) of the background, only available in the
22888 @code{testsrc2} source. The value must be between 0 (fully transparent) and
22889 255 (fully opaque, the default).
22892 Set the number of decimals to show in the timestamp, only available in the
22893 @code{testsrc} source.
22895 The displayed timestamp value will correspond to the original
22896 timestamp value multiplied by the power of 10 of the specified
22897 value. Default value is 0.
22900 @subsection Examples
22904 Generate a video with a duration of 5.3 seconds, with size
22905 176x144 and a frame rate of 10 frames per second:
22907 testsrc=duration=5.3:size=qcif:rate=10
22911 The following graph description will generate a red source
22912 with an opacity of 0.2, with size "qcif" and a frame rate of 10
22915 color=c=red@@0.2:s=qcif:r=10
22919 If the input content is to be ignored, @code{nullsrc} can be used. The
22920 following command generates noise in the luminance plane by employing
22921 the @code{geq} filter:
22923 nullsrc=s=256x256, geq=random(1)*255:128:128
22927 @subsection Commands
22929 The @code{color} source supports the following commands:
22933 Set the color of the created image. Accepts the same syntax of the
22934 corresponding @option{color} option.
22939 Generate video using an OpenCL program.
22944 OpenCL program source file.
22947 Kernel name in program.
22950 Size of frames to generate. This must be set.
22953 Pixel format to use for the generated frames. This must be set.
22956 Number of frames generated every second. Default value is '25'.
22960 For details of how the program loading works, see the @ref{program_opencl}
22967 Generate a colour ramp by setting pixel values from the position of the pixel
22968 in the output image. (Note that this will work with all pixel formats, but
22969 the generated output will not be the same.)
22971 __kernel void ramp(__write_only image2d_t dst,
22972 unsigned int index)
22974 int2 loc = (int2)(get_global_id(0), get_global_id(1));
22977 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
22979 write_imagef(dst, loc, val);
22984 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
22986 __kernel void sierpinski_carpet(__write_only image2d_t dst,
22987 unsigned int index)
22989 int2 loc = (int2)(get_global_id(0), get_global_id(1));
22991 float4 value = 0.0f;
22992 int x = loc.x + index;
22993 int y = loc.y + index;
22994 while (x > 0 || y > 0) {
22995 if (x % 3 == 1 && y % 3 == 1) {
23003 write_imagef(dst, loc, value);
23009 @section sierpinski
23011 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
23013 This source accepts the following options:
23017 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23018 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23021 Set frame rate, expressed as number of frames per second. Default
23025 Set seed which is used for random panning.
23028 Set max jump for single pan destination. Allowed range is from 1 to 10000.
23031 Set fractal type, can be default @code{carpet} or @code{triangle}.
23034 @c man end VIDEO SOURCES
23036 @chapter Video Sinks
23037 @c man begin VIDEO SINKS
23039 Below is a description of the currently available video sinks.
23041 @section buffersink
23043 Buffer video frames, and make them available to the end of the filter
23046 This sink is mainly intended for programmatic use, in particular
23047 through the interface defined in @file{libavfilter/buffersink.h}
23048 or the options system.
23050 It accepts a pointer to an AVBufferSinkContext structure, which
23051 defines the incoming buffers' formats, to be passed as the opaque
23052 parameter to @code{avfilter_init_filter} for initialization.
23056 Null video sink: do absolutely nothing with the input video. It is
23057 mainly useful as a template and for use in analysis / debugging
23060 @c man end VIDEO SINKS
23062 @chapter Multimedia Filters
23063 @c man begin MULTIMEDIA FILTERS
23065 Below is a description of the currently available multimedia filters.
23069 Convert input audio to a video output, displaying the audio bit scope.
23071 The filter accepts the following options:
23075 Set frame rate, expressed as number of frames per second. Default
23079 Specify the video size for the output. For the syntax of this option, check the
23080 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23081 Default value is @code{1024x256}.
23084 Specify list of colors separated by space or by '|' which will be used to
23085 draw channels. Unrecognized or missing colors will be replaced
23089 @section adrawgraph
23090 Draw a graph using input audio metadata.
23092 See @ref{drawgraph}
23094 @section agraphmonitor
23096 See @ref{graphmonitor}.
23098 @section ahistogram
23100 Convert input audio to a video output, displaying the volume histogram.
23102 The filter accepts the following options:
23106 Specify how histogram is calculated.
23108 It accepts the following values:
23111 Use single histogram for all channels.
23113 Use separate histogram for each channel.
23115 Default is @code{single}.
23118 Set frame rate, expressed as number of frames per second. Default
23122 Specify the video size for the output. For the syntax of this option, check the
23123 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23124 Default value is @code{hd720}.
23129 It accepts the following values:
23140 reverse logarithmic
23142 Default is @code{log}.
23145 Set amplitude scale.
23147 It accepts the following values:
23154 Default is @code{log}.
23157 Set how much frames to accumulate in histogram.
23158 Default is 1. Setting this to -1 accumulates all frames.
23161 Set histogram ratio of window height.
23164 Set sonogram sliding.
23166 It accepts the following values:
23169 replace old rows with new ones.
23171 scroll from top to bottom.
23173 Default is @code{replace}.
23176 @section aphasemeter
23178 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
23179 representing mean phase of current audio frame. A video output can also be produced and is
23180 enabled by default. The audio is passed through as first output.
23182 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
23183 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
23184 and @code{1} means channels are in phase.
23186 The filter accepts the following options, all related to its video output:
23190 Set the output frame rate. Default value is @code{25}.
23193 Set the video size for the output. For the syntax of this option, check the
23194 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23195 Default value is @code{800x400}.
23200 Specify the red, green, blue contrast. Default values are @code{2},
23201 @code{7} and @code{1}.
23202 Allowed range is @code{[0, 255]}.
23205 Set color which will be used for drawing median phase. If color is
23206 @code{none} which is default, no median phase value will be drawn.
23209 Enable video output. Default is enabled.
23212 @section avectorscope
23214 Convert input audio to a video output, representing the audio vector
23217 The filter is used to measure the difference between channels of stereo
23218 audio stream. A monaural signal, consisting of identical left and right
23219 signal, results in straight vertical line. Any stereo separation is visible
23220 as a deviation from this line, creating a Lissajous figure.
23221 If the straight (or deviation from it) but horizontal line appears this
23222 indicates that the left and right channels are out of phase.
23224 The filter accepts the following options:
23228 Set the vectorscope mode.
23230 Available values are:
23233 Lissajous rotated by 45 degrees.
23236 Same as above but not rotated.
23239 Shape resembling half of circle.
23242 Default value is @samp{lissajous}.
23245 Set the video size for the output. For the syntax of this option, check the
23246 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23247 Default value is @code{400x400}.
23250 Set the output frame rate. Default value is @code{25}.
23256 Specify the red, green, blue and alpha contrast. Default values are @code{40},
23257 @code{160}, @code{80} and @code{255}.
23258 Allowed range is @code{[0, 255]}.
23264 Specify the red, green, blue and alpha fade. Default values are @code{15},
23265 @code{10}, @code{5} and @code{5}.
23266 Allowed range is @code{[0, 255]}.
23269 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
23270 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
23273 Set the vectorscope drawing mode.
23275 Available values are:
23278 Draw dot for each sample.
23281 Draw line between previous and current sample.
23284 Default value is @samp{dot}.
23287 Specify amplitude scale of audio samples.
23289 Available values are:
23305 Swap left channel axis with right channel axis.
23315 Mirror only x axis.
23318 Mirror only y axis.
23326 @subsection Examples
23330 Complete example using @command{ffplay}:
23332 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
23333 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
23337 @section bench, abench
23339 Benchmark part of a filtergraph.
23341 The filter accepts the following options:
23345 Start or stop a timer.
23347 Available values are:
23350 Get the current time, set it as frame metadata (using the key
23351 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
23354 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
23355 the input frame metadata to get the time difference. Time difference, average,
23356 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
23357 @code{min}) are then printed. The timestamps are expressed in seconds.
23361 @subsection Examples
23365 Benchmark @ref{selectivecolor} filter:
23367 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
23373 Concatenate audio and video streams, joining them together one after the
23376 The filter works on segments of synchronized video and audio streams. All
23377 segments must have the same number of streams of each type, and that will
23378 also be the number of streams at output.
23380 The filter accepts the following options:
23385 Set the number of segments. Default is 2.
23388 Set the number of output video streams, that is also the number of video
23389 streams in each segment. Default is 1.
23392 Set the number of output audio streams, that is also the number of audio
23393 streams in each segment. Default is 0.
23396 Activate unsafe mode: do not fail if segments have a different format.
23400 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
23401 @var{a} audio outputs.
23403 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
23404 segment, in the same order as the outputs, then the inputs for the second
23407 Related streams do not always have exactly the same duration, for various
23408 reasons including codec frame size or sloppy authoring. For that reason,
23409 related synchronized streams (e.g. a video and its audio track) should be
23410 concatenated at once. The concat filter will use the duration of the longest
23411 stream in each segment (except the last one), and if necessary pad shorter
23412 audio streams with silence.
23414 For this filter to work correctly, all segments must start at timestamp 0.
23416 All corresponding streams must have the same parameters in all segments; the
23417 filtering system will automatically select a common pixel format for video
23418 streams, and a common sample format, sample rate and channel layout for
23419 audio streams, but other settings, such as resolution, must be converted
23420 explicitly by the user.
23422 Different frame rates are acceptable but will result in variable frame rate
23423 at output; be sure to configure the output file to handle it.
23425 @subsection Examples
23429 Concatenate an opening, an episode and an ending, all in bilingual version
23430 (video in stream 0, audio in streams 1 and 2):
23432 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
23433 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
23434 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
23435 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
23439 Concatenate two parts, handling audio and video separately, using the
23440 (a)movie sources, and adjusting the resolution:
23442 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
23443 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
23444 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
23446 Note that a desync will happen at the stitch if the audio and video streams
23447 do not have exactly the same duration in the first file.
23451 @subsection Commands
23453 This filter supports the following commands:
23456 Close the current segment and step to the next one
23462 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
23463 level. By default, it logs a message at a frequency of 10Hz with the
23464 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
23465 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
23467 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
23468 sample format is double-precision floating point. The input stream will be converted to
23469 this specification, if needed. Users may need to insert aformat and/or aresample filters
23470 after this filter to obtain the original parameters.
23472 The filter also has a video output (see the @var{video} option) with a real
23473 time graph to observe the loudness evolution. The graphic contains the logged
23474 message mentioned above, so it is not printed anymore when this option is set,
23475 unless the verbose logging is set. The main graphing area contains the
23476 short-term loudness (3 seconds of analysis), and the gauge on the right is for
23477 the momentary loudness (400 milliseconds), but can optionally be configured
23478 to instead display short-term loudness (see @var{gauge}).
23480 The green area marks a +/- 1LU target range around the target loudness
23481 (-23LUFS by default, unless modified through @var{target}).
23483 More information about the Loudness Recommendation EBU R128 on
23484 @url{http://tech.ebu.ch/loudness}.
23486 The filter accepts the following options:
23491 Activate the video output. The audio stream is passed unchanged whether this
23492 option is set or no. The video stream will be the first output stream if
23493 activated. Default is @code{0}.
23496 Set the video size. This option is for video only. For the syntax of this
23498 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23499 Default and minimum resolution is @code{640x480}.
23502 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
23503 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
23504 other integer value between this range is allowed.
23507 Set metadata injection. If set to @code{1}, the audio input will be segmented
23508 into 100ms output frames, each of them containing various loudness information
23509 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
23511 Default is @code{0}.
23514 Force the frame logging level.
23516 Available values are:
23519 information logging level
23521 verbose logging level
23524 By default, the logging level is set to @var{info}. If the @option{video} or
23525 the @option{metadata} options are set, it switches to @var{verbose}.
23530 Available modes can be cumulated (the option is a @code{flag} type). Possible
23534 Disable any peak mode (default).
23536 Enable sample-peak mode.
23538 Simple peak mode looking for the higher sample value. It logs a message
23539 for sample-peak (identified by @code{SPK}).
23541 Enable true-peak mode.
23543 If enabled, the peak lookup is done on an over-sampled version of the input
23544 stream for better peak accuracy. It logs a message for true-peak.
23545 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
23546 This mode requires a build with @code{libswresample}.
23550 Treat mono input files as "dual mono". If a mono file is intended for playback
23551 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
23552 If set to @code{true}, this option will compensate for this effect.
23553 Multi-channel input files are not affected by this option.
23556 Set a specific pan law to be used for the measurement of dual mono files.
23557 This parameter is optional, and has a default value of -3.01dB.
23560 Set a specific target level (in LUFS) used as relative zero in the visualization.
23561 This parameter is optional and has a default value of -23LUFS as specified
23562 by EBU R128. However, material published online may prefer a level of -16LUFS
23563 (e.g. for use with podcasts or video platforms).
23566 Set the value displayed by the gauge. Valid values are @code{momentary} and s
23567 @code{shortterm}. By default the momentary value will be used, but in certain
23568 scenarios it may be more useful to observe the short term value instead (e.g.
23572 Sets the display scale for the loudness. Valid parameters are @code{absolute}
23573 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
23574 video output, not the summary or continuous log output.
23577 @subsection Examples
23581 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
23583 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
23587 Run an analysis with @command{ffmpeg}:
23589 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
23593 @section interleave, ainterleave
23595 Temporally interleave frames from several inputs.
23597 @code{interleave} works with video inputs, @code{ainterleave} with audio.
23599 These filters read frames from several inputs and send the oldest
23600 queued frame to the output.
23602 Input streams must have well defined, monotonically increasing frame
23605 In order to submit one frame to output, these filters need to enqueue
23606 at least one frame for each input, so they cannot work in case one
23607 input is not yet terminated and will not receive incoming frames.
23609 For example consider the case when one input is a @code{select} filter
23610 which always drops input frames. The @code{interleave} filter will keep
23611 reading from that input, but it will never be able to send new frames
23612 to output until the input sends an end-of-stream signal.
23614 Also, depending on inputs synchronization, the filters will drop
23615 frames in case one input receives more frames than the other ones, and
23616 the queue is already filled.
23618 These filters accept the following options:
23622 Set the number of different inputs, it is 2 by default.
23625 How to determine the end-of-stream.
23629 The duration of the longest input. (default)
23632 The duration of the shortest input.
23635 The duration of the first input.
23640 @subsection Examples
23644 Interleave frames belonging to different streams using @command{ffmpeg}:
23646 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
23650 Add flickering blur effect:
23652 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
23656 @section metadata, ametadata
23658 Manipulate frame metadata.
23660 This filter accepts the following options:
23664 Set mode of operation of the filter.
23666 Can be one of the following:
23670 If both @code{value} and @code{key} is set, select frames
23671 which have such metadata. If only @code{key} is set, select
23672 every frame that has such key in metadata.
23675 Add new metadata @code{key} and @code{value}. If key is already available
23679 Modify value of already present key.
23682 If @code{value} is set, delete only keys that have such value.
23683 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
23687 Print key and its value if metadata was found. If @code{key} is not set print all
23688 metadata values available in frame.
23692 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
23695 Set metadata value which will be used. This option is mandatory for
23696 @code{modify} and @code{add} mode.
23699 Which function to use when comparing metadata value and @code{value}.
23701 Can be one of following:
23705 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
23708 Values are interpreted as strings, returns true if metadata value starts with
23709 the @code{value} option string.
23712 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
23715 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
23718 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
23721 Values are interpreted as floats, returns true if expression from option @code{expr}
23725 Values are interpreted as strings, returns true if metadata value ends with
23726 the @code{value} option string.
23730 Set expression which is used when @code{function} is set to @code{expr}.
23731 The expression is evaluated through the eval API and can contain the following
23736 Float representation of @code{value} from metadata key.
23739 Float representation of @code{value} as supplied by user in @code{value} option.
23743 If specified in @code{print} mode, output is written to the named file. Instead of
23744 plain filename any writable url can be specified. Filename ``-'' is a shorthand
23745 for standard output. If @code{file} option is not set, output is written to the log
23746 with AV_LOG_INFO loglevel.
23749 Reduces buffering in print mode when output is written to a URL set using @var{file}.
23753 @subsection Examples
23757 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
23760 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
23763 Print silencedetect output to file @file{metadata.txt}.
23765 silencedetect,ametadata=mode=print:file=metadata.txt
23768 Direct all metadata to a pipe with file descriptor 4.
23770 metadata=mode=print:file='pipe\:4'
23774 @section perms, aperms
23776 Set read/write permissions for the output frames.
23778 These filters are mainly aimed at developers to test direct path in the
23779 following filter in the filtergraph.
23781 The filters accept the following options:
23785 Select the permissions mode.
23787 It accepts the following values:
23790 Do nothing. This is the default.
23792 Set all the output frames read-only.
23794 Set all the output frames directly writable.
23796 Make the frame read-only if writable, and writable if read-only.
23798 Set each output frame read-only or writable randomly.
23802 Set the seed for the @var{random} mode, must be an integer included between
23803 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
23804 @code{-1}, the filter will try to use a good random seed on a best effort
23808 Note: in case of auto-inserted filter between the permission filter and the
23809 following one, the permission might not be received as expected in that
23810 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
23811 perms/aperms filter can avoid this problem.
23813 @section realtime, arealtime
23815 Slow down filtering to match real time approximately.
23817 These filters will pause the filtering for a variable amount of time to
23818 match the output rate with the input timestamps.
23819 They are similar to the @option{re} option to @code{ffmpeg}.
23821 They accept the following options:
23825 Time limit for the pauses. Any pause longer than that will be considered
23826 a timestamp discontinuity and reset the timer. Default is 2 seconds.
23828 Speed factor for processing. The value must be a float larger than zero.
23829 Values larger than 1.0 will result in faster than realtime processing,
23830 smaller will slow processing down. The @var{limit} is automatically adapted
23831 accordingly. Default is 1.0.
23833 A processing speed faster than what is possible without these filters cannot
23838 @section select, aselect
23840 Select frames to pass in output.
23842 This filter accepts the following options:
23847 Set expression, which is evaluated for each input frame.
23849 If the expression is evaluated to zero, the frame is discarded.
23851 If the evaluation result is negative or NaN, the frame is sent to the
23852 first output; otherwise it is sent to the output with index
23853 @code{ceil(val)-1}, assuming that the input index starts from 0.
23855 For example a value of @code{1.2} corresponds to the output with index
23856 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
23859 Set the number of outputs. The output to which to send the selected
23860 frame is based on the result of the evaluation. Default value is 1.
23863 The expression can contain the following constants:
23867 The (sequential) number of the filtered frame, starting from 0.
23870 The (sequential) number of the selected frame, starting from 0.
23872 @item prev_selected_n
23873 The sequential number of the last selected frame. It's NAN if undefined.
23876 The timebase of the input timestamps.
23879 The PTS (Presentation TimeStamp) of the filtered video frame,
23880 expressed in @var{TB} units. It's NAN if undefined.
23883 The PTS of the filtered video frame,
23884 expressed in seconds. It's NAN if undefined.
23887 The PTS of the previously filtered video frame. It's NAN if undefined.
23889 @item prev_selected_pts
23890 The PTS of the last previously filtered video frame. It's NAN if undefined.
23892 @item prev_selected_t
23893 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
23896 The PTS of the first video frame in the video. It's NAN if undefined.
23899 The time of the first video frame in the video. It's NAN if undefined.
23901 @item pict_type @emph{(video only)}
23902 The type of the filtered frame. It can assume one of the following
23914 @item interlace_type @emph{(video only)}
23915 The frame interlace type. It can assume one of the following values:
23918 The frame is progressive (not interlaced).
23920 The frame is top-field-first.
23922 The frame is bottom-field-first.
23925 @item consumed_sample_n @emph{(audio only)}
23926 the number of selected samples before the current frame
23928 @item samples_n @emph{(audio only)}
23929 the number of samples in the current frame
23931 @item sample_rate @emph{(audio only)}
23932 the input sample rate
23935 This is 1 if the filtered frame is a key-frame, 0 otherwise.
23938 the position in the file of the filtered frame, -1 if the information
23939 is not available (e.g. for synthetic video)
23941 @item scene @emph{(video only)}
23942 value between 0 and 1 to indicate a new scene; a low value reflects a low
23943 probability for the current frame to introduce a new scene, while a higher
23944 value means the current frame is more likely to be one (see the example below)
23946 @item concatdec_select
23947 The concat demuxer can select only part of a concat input file by setting an
23948 inpoint and an outpoint, but the output packets may not be entirely contained
23949 in the selected interval. By using this variable, it is possible to skip frames
23950 generated by the concat demuxer which are not exactly contained in the selected
23953 This works by comparing the frame pts against the @var{lavf.concat.start_time}
23954 and the @var{lavf.concat.duration} packet metadata values which are also
23955 present in the decoded frames.
23957 The @var{concatdec_select} variable is -1 if the frame pts is at least
23958 start_time and either the duration metadata is missing or the frame pts is less
23959 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
23962 That basically means that an input frame is selected if its pts is within the
23963 interval set by the concat demuxer.
23967 The default value of the select expression is "1".
23969 @subsection Examples
23973 Select all frames in input:
23978 The example above is the same as:
23990 Select only I-frames:
23992 select='eq(pict_type\,I)'
23996 Select one frame every 100:
23998 select='not(mod(n\,100))'
24002 Select only frames contained in the 10-20 time interval:
24004 select=between(t\,10\,20)
24008 Select only I-frames contained in the 10-20 time interval:
24010 select=between(t\,10\,20)*eq(pict_type\,I)
24014 Select frames with a minimum distance of 10 seconds:
24016 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
24020 Use aselect to select only audio frames with samples number > 100:
24022 aselect='gt(samples_n\,100)'
24026 Create a mosaic of the first scenes:
24028 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
24031 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
24035 Send even and odd frames to separate outputs, and compose them:
24037 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
24041 Select useful frames from an ffconcat file which is using inpoints and
24042 outpoints but where the source files are not intra frame only.
24044 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
24048 @section sendcmd, asendcmd
24050 Send commands to filters in the filtergraph.
24052 These filters read commands to be sent to other filters in the
24055 @code{sendcmd} must be inserted between two video filters,
24056 @code{asendcmd} must be inserted between two audio filters, but apart
24057 from that they act the same way.
24059 The specification of commands can be provided in the filter arguments
24060 with the @var{commands} option, or in a file specified by the
24061 @var{filename} option.
24063 These filters accept the following options:
24066 Set the commands to be read and sent to the other filters.
24068 Set the filename of the commands to be read and sent to the other
24072 @subsection Commands syntax
24074 A commands description consists of a sequence of interval
24075 specifications, comprising a list of commands to be executed when a
24076 particular event related to that interval occurs. The occurring event
24077 is typically the current frame time entering or leaving a given time
24080 An interval is specified by the following syntax:
24082 @var{START}[-@var{END}] @var{COMMANDS};
24085 The time interval is specified by the @var{START} and @var{END} times.
24086 @var{END} is optional and defaults to the maximum time.
24088 The current frame time is considered within the specified interval if
24089 it is included in the interval [@var{START}, @var{END}), that is when
24090 the time is greater or equal to @var{START} and is lesser than
24093 @var{COMMANDS} consists of a sequence of one or more command
24094 specifications, separated by ",", relating to that interval. The
24095 syntax of a command specification is given by:
24097 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
24100 @var{FLAGS} is optional and specifies the type of events relating to
24101 the time interval which enable sending the specified command, and must
24102 be a non-null sequence of identifier flags separated by "+" or "|" and
24103 enclosed between "[" and "]".
24105 The following flags are recognized:
24108 The command is sent when the current frame timestamp enters the
24109 specified interval. In other words, the command is sent when the
24110 previous frame timestamp was not in the given interval, and the
24114 The command is sent when the current frame timestamp leaves the
24115 specified interval. In other words, the command is sent when the
24116 previous frame timestamp was in the given interval, and the
24120 The command @var{ARG} is interpreted as expression and result of
24121 expression is passed as @var{ARG}.
24123 The expression is evaluated through the eval API and can contain the following
24128 Original position in the file of the frame, or undefined if undefined
24129 for the current frame.
24132 The presentation timestamp in input.
24135 The count of the input frame for video or audio, starting from 0.
24138 The time in seconds of the current frame.
24141 The start time in seconds of the current command interval.
24144 The end time in seconds of the current command interval.
24147 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
24152 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
24155 @var{TARGET} specifies the target of the command, usually the name of
24156 the filter class or a specific filter instance name.
24158 @var{COMMAND} specifies the name of the command for the target filter.
24160 @var{ARG} is optional and specifies the optional list of argument for
24161 the given @var{COMMAND}.
24163 Between one interval specification and another, whitespaces, or
24164 sequences of characters starting with @code{#} until the end of line,
24165 are ignored and can be used to annotate comments.
24167 A simplified BNF description of the commands specification syntax
24170 @var{COMMAND_FLAG} ::= "enter" | "leave"
24171 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
24172 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
24173 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
24174 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
24175 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
24178 @subsection Examples
24182 Specify audio tempo change at second 4:
24184 asendcmd=c='4.0 atempo tempo 1.5',atempo
24188 Target a specific filter instance:
24190 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
24194 Specify a list of drawtext and hue commands in a file.
24196 # show text in the interval 5-10
24197 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
24198 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
24200 # desaturate the image in the interval 15-20
24201 15.0-20.0 [enter] hue s 0,
24202 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
24204 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
24206 # apply an exponential saturation fade-out effect, starting from time 25
24207 25 [enter] hue s exp(25-t)
24210 A filtergraph allowing to read and process the above command list
24211 stored in a file @file{test.cmd}, can be specified with:
24213 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
24218 @section setpts, asetpts
24220 Change the PTS (presentation timestamp) of the input frames.
24222 @code{setpts} works on video frames, @code{asetpts} on audio frames.
24224 This filter accepts the following options:
24229 The expression which is evaluated for each frame to construct its timestamp.
24233 The expression is evaluated through the eval API and can contain the following
24237 @item FRAME_RATE, FR
24238 frame rate, only defined for constant frame-rate video
24241 The presentation timestamp in input
24244 The count of the input frame for video or the number of consumed samples,
24245 not including the current frame for audio, starting from 0.
24247 @item NB_CONSUMED_SAMPLES
24248 The number of consumed samples, not including the current frame (only
24251 @item NB_SAMPLES, S
24252 The number of samples in the current frame (only audio)
24254 @item SAMPLE_RATE, SR
24255 The audio sample rate.
24258 The PTS of the first frame.
24261 the time in seconds of the first frame
24264 State whether the current frame is interlaced.
24267 the time in seconds of the current frame
24270 original position in the file of the frame, or undefined if undefined
24271 for the current frame
24274 The previous input PTS.
24277 previous input time in seconds
24280 The previous output PTS.
24283 previous output time in seconds
24286 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
24290 The wallclock (RTC) time at the start of the movie in microseconds.
24293 The timebase of the input timestamps.
24297 @subsection Examples
24301 Start counting PTS from zero
24303 setpts=PTS-STARTPTS
24307 Apply fast motion effect:
24313 Apply slow motion effect:
24319 Set fixed rate of 25 frames per second:
24325 Set fixed rate 25 fps with some jitter:
24327 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
24331 Apply an offset of 10 seconds to the input PTS:
24337 Generate timestamps from a "live source" and rebase onto the current timebase:
24339 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
24343 Generate timestamps by counting samples:
24352 Force color range for the output video frame.
24354 The @code{setrange} filter marks the color range property for the
24355 output frames. It does not change the input frame, but only sets the
24356 corresponding property, which affects how the frame is treated by
24359 The filter accepts the following options:
24364 Available values are:
24368 Keep the same color range property.
24370 @item unspecified, unknown
24371 Set the color range as unspecified.
24373 @item limited, tv, mpeg
24374 Set the color range as limited.
24376 @item full, pc, jpeg
24377 Set the color range as full.
24381 @section settb, asettb
24383 Set the timebase to use for the output frames timestamps.
24384 It is mainly useful for testing timebase configuration.
24386 It accepts the following parameters:
24391 The expression which is evaluated into the output timebase.
24395 The value for @option{tb} is an arithmetic expression representing a
24396 rational. The expression can contain the constants "AVTB" (the default
24397 timebase), "intb" (the input timebase) and "sr" (the sample rate,
24398 audio only). Default value is "intb".
24400 @subsection Examples
24404 Set the timebase to 1/25:
24410 Set the timebase to 1/10:
24416 Set the timebase to 1001/1000:
24422 Set the timebase to 2*intb:
24428 Set the default timebase value:
24435 Convert input audio to a video output representing frequency spectrum
24436 logarithmically using Brown-Puckette constant Q transform algorithm with
24437 direct frequency domain coefficient calculation (but the transform itself
24438 is not really constant Q, instead the Q factor is actually variable/clamped),
24439 with musical tone scale, from E0 to D#10.
24441 The filter accepts the following options:
24445 Specify the video size for the output. It must be even. For the syntax of this option,
24446 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24447 Default value is @code{1920x1080}.
24450 Set the output frame rate. Default value is @code{25}.
24453 Set the bargraph height. It must be even. Default value is @code{-1} which
24454 computes the bargraph height automatically.
24457 Set the axis height. It must be even. Default value is @code{-1} which computes
24458 the axis height automatically.
24461 Set the sonogram height. It must be even. Default value is @code{-1} which
24462 computes the sonogram height automatically.
24465 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
24466 instead. Default value is @code{1}.
24468 @item sono_v, volume
24469 Specify the sonogram volume expression. It can contain variables:
24472 the @var{bar_v} evaluated expression
24473 @item frequency, freq, f
24474 the frequency where it is evaluated
24475 @item timeclamp, tc
24476 the value of @var{timeclamp} option
24480 @item a_weighting(f)
24481 A-weighting of equal loudness
24482 @item b_weighting(f)
24483 B-weighting of equal loudness
24484 @item c_weighting(f)
24485 C-weighting of equal loudness.
24487 Default value is @code{16}.
24489 @item bar_v, volume2
24490 Specify the bargraph volume expression. It can contain variables:
24493 the @var{sono_v} evaluated expression
24494 @item frequency, freq, f
24495 the frequency where it is evaluated
24496 @item timeclamp, tc
24497 the value of @var{timeclamp} option
24501 @item a_weighting(f)
24502 A-weighting of equal loudness
24503 @item b_weighting(f)
24504 B-weighting of equal loudness
24505 @item c_weighting(f)
24506 C-weighting of equal loudness.
24508 Default value is @code{sono_v}.
24510 @item sono_g, gamma
24511 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
24512 higher gamma makes the spectrum having more range. Default value is @code{3}.
24513 Acceptable range is @code{[1, 7]}.
24515 @item bar_g, gamma2
24516 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
24520 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
24521 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
24523 @item timeclamp, tc
24524 Specify the transform timeclamp. At low frequency, there is trade-off between
24525 accuracy in time domain and frequency domain. If timeclamp is lower,
24526 event in time domain is represented more accurately (such as fast bass drum),
24527 otherwise event in frequency domain is represented more accurately
24528 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
24531 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
24532 limits future samples by applying asymmetric windowing in time domain, useful
24533 when low latency is required. Accepted range is @code{[0, 1]}.
24536 Specify the transform base frequency. Default value is @code{20.01523126408007475},
24537 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
24540 Specify the transform end frequency. Default value is @code{20495.59681441799654},
24541 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
24544 This option is deprecated and ignored.
24547 Specify the transform length in time domain. Use this option to control accuracy
24548 trade-off between time domain and frequency domain at every frequency sample.
24549 It can contain variables:
24551 @item frequency, freq, f
24552 the frequency where it is evaluated
24553 @item timeclamp, tc
24554 the value of @var{timeclamp} option.
24556 Default value is @code{384*tc/(384+tc*f)}.
24559 Specify the transform count for every video frame. Default value is @code{6}.
24560 Acceptable range is @code{[1, 30]}.
24563 Specify the transform count for every single pixel. Default value is @code{0},
24564 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
24567 Specify font file for use with freetype to draw the axis. If not specified,
24568 use embedded font. Note that drawing with font file or embedded font is not
24569 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
24573 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
24574 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
24578 Specify font color expression. This is arithmetic expression that should return
24579 integer value 0xRRGGBB. It can contain variables:
24581 @item frequency, freq, f
24582 the frequency where it is evaluated
24583 @item timeclamp, tc
24584 the value of @var{timeclamp} option
24589 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
24590 @item r(x), g(x), b(x)
24591 red, green, and blue value of intensity x.
24593 Default value is @code{st(0, (midi(f)-59.5)/12);
24594 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
24595 r(1-ld(1)) + b(ld(1))}.
24598 Specify image file to draw the axis. This option override @var{fontfile} and
24599 @var{fontcolor} option.
24602 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
24603 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
24604 Default value is @code{1}.
24607 Set colorspace. The accepted values are:
24610 Unspecified (default)
24619 BT.470BG or BT.601-6 625
24622 SMPTE-170M or BT.601-6 525
24628 BT.2020 with non-constant luminance
24633 Set spectrogram color scheme. This is list of floating point values with format
24634 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
24635 The default is @code{1|0.5|0|0|0.5|1}.
24639 @subsection Examples
24643 Playing audio while showing the spectrum:
24645 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
24649 Same as above, but with frame rate 30 fps:
24651 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
24655 Playing at 1280x720:
24657 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
24661 Disable sonogram display:
24667 A1 and its harmonics: A1, A2, (near)E3, A3:
24669 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),
24670 asplit[a][out1]; [a] showcqt [out0]'
24674 Same as above, but with more accuracy in frequency domain:
24676 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),
24677 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
24683 bar_v=10:sono_v=bar_v*a_weighting(f)
24687 Custom gamma, now spectrum is linear to the amplitude.
24693 Custom tlength equation:
24695 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)))'
24699 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
24701 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
24705 Custom font using fontconfig:
24707 font='Courier New,Monospace,mono|bold'
24711 Custom frequency range with custom axis using image file:
24713 axisfile=myaxis.png:basefreq=40:endfreq=10000
24719 Convert input audio to video output representing the audio power spectrum.
24720 Audio amplitude is on Y-axis while frequency is on X-axis.
24722 The filter accepts the following options:
24726 Specify size of video. For the syntax of this option, check the
24727 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24728 Default is @code{1024x512}.
24732 This set how each frequency bin will be represented.
24734 It accepts the following values:
24740 Default is @code{bar}.
24743 Set amplitude scale.
24745 It accepts the following values:
24759 Default is @code{log}.
24762 Set frequency scale.
24764 It accepts the following values:
24773 Reverse logarithmic scale.
24775 Default is @code{lin}.
24778 Set window size. Allowed range is from 16 to 65536.
24780 Default is @code{2048}
24783 Set windowing function.
24785 It accepts the following values:
24808 Default is @code{hanning}.
24811 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
24812 which means optimal overlap for selected window function will be picked.
24815 Set time averaging. Setting this to 0 will display current maximal peaks.
24816 Default is @code{1}, which means time averaging is disabled.
24819 Specify list of colors separated by space or by '|' which will be used to
24820 draw channel frequencies. Unrecognized or missing colors will be replaced
24824 Set channel display mode.
24826 It accepts the following values:
24831 Default is @code{combined}.
24834 Set minimum amplitude used in @code{log} amplitude scaler.
24838 @section showspatial
24840 Convert stereo input audio to a video output, representing the spatial relationship
24841 between two channels.
24843 The filter accepts the following options:
24847 Specify the video size for the output. For the syntax of this option, check the
24848 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24849 Default value is @code{512x512}.
24852 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
24855 Set window function.
24857 It accepts the following values:
24882 Default value is @code{hann}.
24885 Set ratio of overlap window. Default value is @code{0.5}.
24886 When value is @code{1} overlap is set to recommended size for specific
24887 window function currently used.
24890 @anchor{showspectrum}
24891 @section showspectrum
24893 Convert input audio to a video output, representing the audio frequency
24896 The filter accepts the following options:
24900 Specify the video size for the output. For the syntax of this option, check the
24901 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24902 Default value is @code{640x512}.
24905 Specify how the spectrum should slide along the window.
24907 It accepts the following values:
24910 the samples start again on the left when they reach the right
24912 the samples scroll from right to left
24914 frames are only produced when the samples reach the right
24916 the samples scroll from left to right
24919 Default value is @code{replace}.
24922 Specify display mode.
24924 It accepts the following values:
24927 all channels are displayed in the same row
24929 all channels are displayed in separate rows
24932 Default value is @samp{combined}.
24935 Specify display color mode.
24937 It accepts the following values:
24940 each channel is displayed in a separate color
24942 each channel is displayed using the same color scheme
24944 each channel is displayed using the rainbow color scheme
24946 each channel is displayed using the moreland color scheme
24948 each channel is displayed using the nebulae color scheme
24950 each channel is displayed using the fire color scheme
24952 each channel is displayed using the fiery color scheme
24954 each channel is displayed using the fruit color scheme
24956 each channel is displayed using the cool color scheme
24958 each channel is displayed using the magma color scheme
24960 each channel is displayed using the green color scheme
24962 each channel is displayed using the viridis color scheme
24964 each channel is displayed using the plasma color scheme
24966 each channel is displayed using the cividis color scheme
24968 each channel is displayed using the terrain color scheme
24971 Default value is @samp{channel}.
24974 Specify scale used for calculating intensity color values.
24976 It accepts the following values:
24981 square root, default
24992 Default value is @samp{sqrt}.
24995 Specify frequency scale.
24997 It accepts the following values:
25005 Default value is @samp{lin}.
25008 Set saturation modifier for displayed colors. Negative values provide
25009 alternative color scheme. @code{0} is no saturation at all.
25010 Saturation must be in [-10.0, 10.0] range.
25011 Default value is @code{1}.
25014 Set window function.
25016 It accepts the following values:
25041 Default value is @code{hann}.
25044 Set orientation of time vs frequency axis. Can be @code{vertical} or
25045 @code{horizontal}. Default is @code{vertical}.
25048 Set ratio of overlap window. Default value is @code{0}.
25049 When value is @code{1} overlap is set to recommended size for specific
25050 window function currently used.
25053 Set scale gain for calculating intensity color values.
25054 Default value is @code{1}.
25057 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
25060 Set color rotation, must be in [-1.0, 1.0] range.
25061 Default value is @code{0}.
25064 Set start frequency from which to display spectrogram. Default is @code{0}.
25067 Set stop frequency to which to display spectrogram. Default is @code{0}.
25070 Set upper frame rate limit. Default is @code{auto}, unlimited.
25073 Draw time and frequency axes and legends. Default is disabled.
25076 The usage is very similar to the showwaves filter; see the examples in that
25079 @subsection Examples
25083 Large window with logarithmic color scaling:
25085 showspectrum=s=1280x480:scale=log
25089 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
25091 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
25092 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
25096 @section showspectrumpic
25098 Convert input audio to a single video frame, representing the audio frequency
25101 The filter accepts the following options:
25105 Specify the video size for the output. For the syntax of this option, check the
25106 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25107 Default value is @code{4096x2048}.
25110 Specify display mode.
25112 It accepts the following values:
25115 all channels are displayed in the same row
25117 all channels are displayed in separate rows
25119 Default value is @samp{combined}.
25122 Specify display color mode.
25124 It accepts the following values:
25127 each channel is displayed in a separate color
25129 each channel is displayed using the same color scheme
25131 each channel is displayed using the rainbow color scheme
25133 each channel is displayed using the moreland color scheme
25135 each channel is displayed using the nebulae color scheme
25137 each channel is displayed using the fire color scheme
25139 each channel is displayed using the fiery color scheme
25141 each channel is displayed using the fruit color scheme
25143 each channel is displayed using the cool color scheme
25145 each channel is displayed using the magma color scheme
25147 each channel is displayed using the green color scheme
25149 each channel is displayed using the viridis color scheme
25151 each channel is displayed using the plasma color scheme
25153 each channel is displayed using the cividis color scheme
25155 each channel is displayed using the terrain color scheme
25157 Default value is @samp{intensity}.
25160 Specify scale used for calculating intensity color values.
25162 It accepts the following values:
25167 square root, default
25177 Default value is @samp{log}.
25180 Specify frequency scale.
25182 It accepts the following values:
25190 Default value is @samp{lin}.
25193 Set saturation modifier for displayed colors. Negative values provide
25194 alternative color scheme. @code{0} is no saturation at all.
25195 Saturation must be in [-10.0, 10.0] range.
25196 Default value is @code{1}.
25199 Set window function.
25201 It accepts the following values:
25225 Default value is @code{hann}.
25228 Set orientation of time vs frequency axis. Can be @code{vertical} or
25229 @code{horizontal}. Default is @code{vertical}.
25232 Set scale gain for calculating intensity color values.
25233 Default value is @code{1}.
25236 Draw time and frequency axes and legends. Default is enabled.
25239 Set color rotation, must be in [-1.0, 1.0] range.
25240 Default value is @code{0}.
25243 Set start frequency from which to display spectrogram. Default is @code{0}.
25246 Set stop frequency to which to display spectrogram. Default is @code{0}.
25249 @subsection Examples
25253 Extract an audio spectrogram of a whole audio track
25254 in a 1024x1024 picture using @command{ffmpeg}:
25256 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
25260 @section showvolume
25262 Convert input audio volume to a video output.
25264 The filter accepts the following options:
25271 Set border width, allowed range is [0, 5]. Default is 1.
25274 Set channel width, allowed range is [80, 8192]. Default is 400.
25277 Set channel height, allowed range is [1, 900]. Default is 20.
25280 Set fade, allowed range is [0, 1]. Default is 0.95.
25283 Set volume color expression.
25285 The expression can use the following variables:
25289 Current max volume of channel in dB.
25295 Current channel number, starting from 0.
25299 If set, displays channel names. Default is enabled.
25302 If set, displays volume values. Default is enabled.
25305 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
25306 default is @code{h}.
25309 Set step size, allowed range is [0, 5]. Default is 0, which means
25313 Set background opacity, allowed range is [0, 1]. Default is 0.
25316 Set metering mode, can be peak: @code{p} or rms: @code{r},
25317 default is @code{p}.
25320 Set display scale, can be linear: @code{lin} or log: @code{log},
25321 default is @code{lin}.
25325 If set to > 0., display a line for the max level
25326 in the previous seconds.
25327 default is disabled: @code{0.}
25330 The color of the max line. Use when @code{dm} option is set to > 0.
25331 default is: @code{orange}
25336 Convert input audio to a video output, representing the samples waves.
25338 The filter accepts the following options:
25342 Specify the video size for the output. For the syntax of this option, check the
25343 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25344 Default value is @code{600x240}.
25349 Available values are:
25352 Draw a point for each sample.
25355 Draw a vertical line for each sample.
25358 Draw a point for each sample and a line between them.
25361 Draw a centered vertical line for each sample.
25364 Default value is @code{point}.
25367 Set the number of samples which are printed on the same column. A
25368 larger value will decrease the frame rate. Must be a positive
25369 integer. This option can be set only if the value for @var{rate}
25370 is not explicitly specified.
25373 Set the (approximate) output frame rate. This is done by setting the
25374 option @var{n}. Default value is "25".
25376 @item split_channels
25377 Set if channels should be drawn separately or overlap. Default value is 0.
25380 Set colors separated by '|' which are going to be used for drawing of each channel.
25383 Set amplitude scale.
25385 Available values are:
25403 Set the draw mode. This is mostly useful to set for high @var{n}.
25405 Available values are:
25408 Scale pixel values for each drawn sample.
25411 Draw every sample directly.
25414 Default value is @code{scale}.
25417 @subsection Examples
25421 Output the input file audio and the corresponding video representation
25424 amovie=a.mp3,asplit[out0],showwaves[out1]
25428 Create a synthetic signal and show it with showwaves, forcing a
25429 frame rate of 30 frames per second:
25431 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
25435 @section showwavespic
25437 Convert input audio to a single video frame, representing the samples waves.
25439 The filter accepts the following options:
25443 Specify the video size for the output. For the syntax of this option, check the
25444 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25445 Default value is @code{600x240}.
25447 @item split_channels
25448 Set if channels should be drawn separately or overlap. Default value is 0.
25451 Set colors separated by '|' which are going to be used for drawing of each channel.
25454 Set amplitude scale.
25456 Available values are:
25476 Available values are:
25479 Scale pixel values for each drawn sample.
25482 Draw every sample directly.
25485 Default value is @code{scale}.
25488 Set the filter mode.
25490 Available values are:
25493 Use average samples values for each drawn sample.
25496 Use peak samples values for each drawn sample.
25499 Default value is @code{average}.
25502 @subsection Examples
25506 Extract a channel split representation of the wave form of a whole audio track
25507 in a 1024x800 picture using @command{ffmpeg}:
25509 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
25513 @section sidedata, asidedata
25515 Delete frame side data, or select frames based on it.
25517 This filter accepts the following options:
25521 Set mode of operation of the filter.
25523 Can be one of the following:
25527 Select every frame with side data of @code{type}.
25530 Delete side data of @code{type}. If @code{type} is not set, delete all side
25536 Set side data type used with all modes. Must be set for @code{select} mode. For
25537 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
25538 in @file{libavutil/frame.h}. For example, to choose
25539 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
25543 @section spectrumsynth
25545 Synthesize audio from 2 input video spectrums, first input stream represents
25546 magnitude across time and second represents phase across time.
25547 The filter will transform from frequency domain as displayed in videos back
25548 to time domain as presented in audio output.
25550 This filter is primarily created for reversing processed @ref{showspectrum}
25551 filter outputs, but can synthesize sound from other spectrograms too.
25552 But in such case results are going to be poor if the phase data is not
25553 available, because in such cases phase data need to be recreated, usually
25554 it's just recreated from random noise.
25555 For best results use gray only output (@code{channel} color mode in
25556 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
25557 @code{lin} scale for phase video. To produce phase, for 2nd video, use
25558 @code{data} option. Inputs videos should generally use @code{fullframe}
25559 slide mode as that saves resources needed for decoding video.
25561 The filter accepts the following options:
25565 Specify sample rate of output audio, the sample rate of audio from which
25566 spectrum was generated may differ.
25569 Set number of channels represented in input video spectrums.
25572 Set scale which was used when generating magnitude input spectrum.
25573 Can be @code{lin} or @code{log}. Default is @code{log}.
25576 Set slide which was used when generating inputs spectrums.
25577 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
25578 Default is @code{fullframe}.
25581 Set window function used for resynthesis.
25584 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
25585 which means optimal overlap for selected window function will be picked.
25588 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
25589 Default is @code{vertical}.
25592 @subsection Examples
25596 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
25597 then resynthesize videos back to audio with spectrumsynth:
25599 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
25600 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
25601 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
25605 @section split, asplit
25607 Split input into several identical outputs.
25609 @code{asplit} works with audio input, @code{split} with video.
25611 The filter accepts a single parameter which specifies the number of outputs. If
25612 unspecified, it defaults to 2.
25614 @subsection Examples
25618 Create two separate outputs from the same input:
25620 [in] split [out0][out1]
25624 To create 3 or more outputs, you need to specify the number of
25627 [in] asplit=3 [out0][out1][out2]
25631 Create two separate outputs from the same input, one cropped and
25634 [in] split [splitout1][splitout2];
25635 [splitout1] crop=100:100:0:0 [cropout];
25636 [splitout2] pad=200:200:100:100 [padout];
25640 Create 5 copies of the input audio with @command{ffmpeg}:
25642 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
25648 Receive commands sent through a libzmq client, and forward them to
25649 filters in the filtergraph.
25651 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
25652 must be inserted between two video filters, @code{azmq} between two
25653 audio filters. Both are capable to send messages to any filter type.
25655 To enable these filters you need to install the libzmq library and
25656 headers and configure FFmpeg with @code{--enable-libzmq}.
25658 For more information about libzmq see:
25659 @url{http://www.zeromq.org/}
25661 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
25662 receives messages sent through a network interface defined by the
25663 @option{bind_address} (or the abbreviation "@option{b}") option.
25664 Default value of this option is @file{tcp://localhost:5555}. You may
25665 want to alter this value to your needs, but do not forget to escape any
25666 ':' signs (see @ref{filtergraph escaping}).
25668 The received message must be in the form:
25670 @var{TARGET} @var{COMMAND} [@var{ARG}]
25673 @var{TARGET} specifies the target of the command, usually the name of
25674 the filter class or a specific filter instance name. The default
25675 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
25676 but you can override this by using the @samp{filter_name@@id} syntax
25677 (see @ref{Filtergraph syntax}).
25679 @var{COMMAND} specifies the name of the command for the target filter.
25681 @var{ARG} is optional and specifies the optional argument list for the
25682 given @var{COMMAND}.
25684 Upon reception, the message is processed and the corresponding command
25685 is injected into the filtergraph. Depending on the result, the filter
25686 will send a reply to the client, adopting the format:
25688 @var{ERROR_CODE} @var{ERROR_REASON}
25692 @var{MESSAGE} is optional.
25694 @subsection Examples
25696 Look at @file{tools/zmqsend} for an example of a zmq client which can
25697 be used to send commands processed by these filters.
25699 Consider the following filtergraph generated by @command{ffplay}.
25700 In this example the last overlay filter has an instance name. All other
25701 filters will have default instance names.
25704 ffplay -dumpgraph 1 -f lavfi "
25705 color=s=100x100:c=red [l];
25706 color=s=100x100:c=blue [r];
25707 nullsrc=s=200x100, zmq [bg];
25708 [bg][l] overlay [bg+l];
25709 [bg+l][r] overlay@@my=x=100 "
25712 To change the color of the left side of the video, the following
25713 command can be used:
25715 echo Parsed_color_0 c yellow | tools/zmqsend
25718 To change the right side:
25720 echo Parsed_color_1 c pink | tools/zmqsend
25723 To change the position of the right side:
25725 echo overlay@@my x 150 | tools/zmqsend
25729 @c man end MULTIMEDIA FILTERS
25731 @chapter Multimedia Sources
25732 @c man begin MULTIMEDIA SOURCES
25734 Below is a description of the currently available multimedia sources.
25738 This is the same as @ref{movie} source, except it selects an audio
25744 Read audio and/or video stream(s) from a movie container.
25746 It accepts the following parameters:
25750 The name of the resource to read (not necessarily a file; it can also be a
25751 device or a stream accessed through some protocol).
25753 @item format_name, f
25754 Specifies the format assumed for the movie to read, and can be either
25755 the name of a container or an input device. If not specified, the
25756 format is guessed from @var{movie_name} or by probing.
25758 @item seek_point, sp
25759 Specifies the seek point in seconds. The frames will be output
25760 starting from this seek point. The parameter is evaluated with
25761 @code{av_strtod}, so the numerical value may be suffixed by an IS
25762 postfix. The default value is "0".
25765 Specifies the streams to read. Several streams can be specified,
25766 separated by "+". The source will then have as many outputs, in the
25767 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
25768 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
25769 respectively the default (best suited) video and audio stream. Default
25770 is "dv", or "da" if the filter is called as "amovie".
25772 @item stream_index, si
25773 Specifies the index of the video stream to read. If the value is -1,
25774 the most suitable video stream will be automatically selected. The default
25775 value is "-1". Deprecated. If the filter is called "amovie", it will select
25776 audio instead of video.
25779 Specifies how many times to read the stream in sequence.
25780 If the value is 0, the stream will be looped infinitely.
25781 Default value is "1".
25783 Note that when the movie is looped the source timestamps are not
25784 changed, so it will generate non monotonically increasing timestamps.
25786 @item discontinuity
25787 Specifies the time difference between frames above which the point is
25788 considered a timestamp discontinuity which is removed by adjusting the later
25792 It allows overlaying a second video on top of the main input of
25793 a filtergraph, as shown in this graph:
25795 input -----------> deltapts0 --> overlay --> output
25798 movie --> scale--> deltapts1 -------+
25800 @subsection Examples
25804 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
25805 on top of the input labelled "in":
25807 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
25808 [in] setpts=PTS-STARTPTS [main];
25809 [main][over] overlay=16:16 [out]
25813 Read from a video4linux2 device, and overlay it on top of the input
25816 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
25817 [in] setpts=PTS-STARTPTS [main];
25818 [main][over] overlay=16:16 [out]
25822 Read the first video stream and the audio stream with id 0x81 from
25823 dvd.vob; the video is connected to the pad named "video" and the audio is
25824 connected to the pad named "audio":
25826 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
25830 @subsection Commands
25832 Both movie and amovie support the following commands:
25835 Perform seek using "av_seek_frame".
25836 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
25839 @var{stream_index}: If stream_index is -1, a default
25840 stream is selected, and @var{timestamp} is automatically converted
25841 from AV_TIME_BASE units to the stream specific time_base.
25843 @var{timestamp}: Timestamp in AVStream.time_base units
25844 or, if no stream is specified, in AV_TIME_BASE units.
25846 @var{flags}: Flags which select direction and seeking mode.
25850 Get movie duration in AV_TIME_BASE units.
25854 @c man end MULTIMEDIA SOURCES