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 10. 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
7186 @section chromashift
7187 Shift chroma pixels horizontally and/or vertically.
7189 The filter accepts the following options:
7192 Set amount to shift chroma-blue horizontally.
7194 Set amount to shift chroma-blue vertically.
7196 Set amount to shift chroma-red horizontally.
7198 Set amount to shift chroma-red vertically.
7200 Set edge mode, can be @var{smear}, default, or @var{warp}.
7203 @subsection Commands
7205 This filter supports the all above options as @ref{commands}.
7209 Display CIE color diagram with pixels overlaid onto it.
7211 The filter accepts the following options:
7226 @item uhdtv, rec2020
7240 Set what gamuts to draw.
7242 See @code{system} option for available values.
7245 Set ciescope size, by default set to 512.
7248 Set intensity used to map input pixel values to CIE diagram.
7251 Set contrast used to draw tongue colors that are out of active color system gamut.
7254 Correct gamma displayed on scope, by default enabled.
7257 Show white point on CIE diagram, by default disabled.
7260 Set input gamma. Used only with XYZ input color space.
7265 Visualize information exported by some codecs.
7267 Some codecs can export information through frames using side-data or other
7268 means. For example, some MPEG based codecs export motion vectors through the
7269 @var{export_mvs} flag in the codec @option{flags2} option.
7271 The filter accepts the following option:
7275 Set motion vectors to visualize.
7277 Available flags for @var{mv} are:
7281 forward predicted MVs of P-frames
7283 forward predicted MVs of B-frames
7285 backward predicted MVs of B-frames
7289 Display quantization parameters using the chroma planes.
7292 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
7294 Available flags for @var{mv_type} are:
7298 forward predicted MVs
7300 backward predicted MVs
7303 @item frame_type, ft
7304 Set frame type to visualize motion vectors of.
7306 Available flags for @var{frame_type} are:
7310 intra-coded frames (I-frames)
7312 predicted frames (P-frames)
7314 bi-directionally predicted frames (B-frames)
7318 @subsection Examples
7322 Visualize forward predicted MVs of all frames using @command{ffplay}:
7324 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
7328 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
7330 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
7334 @section colorbalance
7335 Modify intensity of primary colors (red, green and blue) of input frames.
7337 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
7338 regions for the red-cyan, green-magenta or blue-yellow balance.
7340 A positive adjustment value shifts the balance towards the primary color, a negative
7341 value towards the complementary color.
7343 The filter accepts the following options:
7349 Adjust red, green and blue shadows (darkest pixels).
7354 Adjust red, green and blue midtones (medium pixels).
7359 Adjust red, green and blue highlights (brightest pixels).
7361 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
7364 Preserve lightness when changing color balance. Default is disabled.
7367 @subsection Examples
7371 Add red color cast to shadows:
7377 @subsection Commands
7379 This filter supports the all above options as @ref{commands}.
7381 @section colorchannelmixer
7383 Adjust video input frames by re-mixing color channels.
7385 This filter modifies a color channel by adding the values associated to
7386 the other channels of the same pixels. For example if the value to
7387 modify is red, the output value will be:
7389 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
7392 The filter accepts the following options:
7399 Adjust contribution of input red, green, blue and alpha channels for output red channel.
7400 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
7406 Adjust contribution of input red, green, blue and alpha channels for output green channel.
7407 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
7413 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
7414 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
7420 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
7421 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
7423 Allowed ranges for options are @code{[-2.0, 2.0]}.
7426 @subsection Examples
7430 Convert source to grayscale:
7432 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
7435 Simulate sepia tones:
7437 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
7441 @subsection Commands
7443 This filter supports the all above options as @ref{commands}.
7446 RGB colorspace color keying.
7448 The filter accepts the following options:
7452 The color which will be replaced with transparency.
7455 Similarity percentage with the key color.
7457 0.01 matches only the exact key color, while 1.0 matches everything.
7462 0.0 makes pixels either fully transparent, or not transparent at all.
7464 Higher values result in semi-transparent pixels, with a higher transparency
7465 the more similar the pixels color is to the key color.
7468 @subsection Examples
7472 Make every green pixel in the input image transparent:
7474 ffmpeg -i input.png -vf colorkey=green out.png
7478 Overlay a greenscreen-video on top of a static background image.
7480 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
7484 @subsection Commands
7485 This filter supports same @ref{commands} as options.
7486 The command accepts the same syntax of the corresponding option.
7488 If the specified expression is not valid, it is kept at its current
7492 Remove all color information for all RGB colors except for certain one.
7494 The filter accepts the following options:
7498 The color which will not be replaced with neutral gray.
7501 Similarity percentage with the above color.
7502 0.01 matches only the exact key color, while 1.0 matches everything.
7505 Blend percentage. 0.0 makes pixels fully gray.
7506 Higher values result in more preserved color.
7509 @subsection Commands
7510 This filter supports same @ref{commands} as options.
7511 The command accepts the same syntax of the corresponding option.
7513 If the specified expression is not valid, it is kept at its current
7516 @section colorlevels
7518 Adjust video input frames using levels.
7520 The filter accepts the following options:
7527 Adjust red, green, blue and alpha input black point.
7528 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
7534 Adjust red, green, blue and alpha input white point.
7535 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
7537 Input levels are used to lighten highlights (bright tones), darken shadows
7538 (dark tones), change the balance of bright and dark tones.
7544 Adjust red, green, blue and alpha output black point.
7545 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
7551 Adjust red, green, blue and alpha output white point.
7552 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
7554 Output levels allows manual selection of a constrained output level range.
7557 @subsection Examples
7561 Make video output darker:
7563 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
7569 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
7573 Make video output lighter:
7575 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
7579 Increase brightness:
7581 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
7585 @subsection Commands
7587 This filter supports the all above options as @ref{commands}.
7589 @section colormatrix
7591 Convert color matrix.
7593 The filter accepts the following options:
7598 Specify the source and destination color matrix. Both values must be
7601 The accepted values are:
7629 For example to convert from BT.601 to SMPTE-240M, use the command:
7631 colormatrix=bt601:smpte240m
7636 Convert colorspace, transfer characteristics or color primaries.
7637 Input video needs to have an even size.
7639 The filter accepts the following options:
7644 Specify all color properties at once.
7646 The accepted values are:
7676 Specify output colorspace.
7678 The accepted values are:
7687 BT.470BG or BT.601-6 625
7690 SMPTE-170M or BT.601-6 525
7699 BT.2020 with non-constant luminance
7705 Specify output transfer characteristics.
7707 The accepted values are:
7719 Constant gamma of 2.2
7722 Constant gamma of 2.8
7725 SMPTE-170M, BT.601-6 625 or BT.601-6 525
7743 BT.2020 for 10-bits content
7746 BT.2020 for 12-bits content
7752 Specify output color primaries.
7754 The accepted values are:
7763 BT.470BG or BT.601-6 625
7766 SMPTE-170M or BT.601-6 525
7790 Specify output color range.
7792 The accepted values are:
7795 TV (restricted) range
7798 MPEG (restricted) range
7809 Specify output color format.
7811 The accepted values are:
7814 YUV 4:2:0 planar 8-bits
7817 YUV 4:2:0 planar 10-bits
7820 YUV 4:2:0 planar 12-bits
7823 YUV 4:2:2 planar 8-bits
7826 YUV 4:2:2 planar 10-bits
7829 YUV 4:2:2 planar 12-bits
7832 YUV 4:4:4 planar 8-bits
7835 YUV 4:4:4 planar 10-bits
7838 YUV 4:4:4 planar 12-bits
7843 Do a fast conversion, which skips gamma/primary correction. This will take
7844 significantly less CPU, but will be mathematically incorrect. To get output
7845 compatible with that produced by the colormatrix filter, use fast=1.
7848 Specify dithering mode.
7850 The accepted values are:
7856 Floyd-Steinberg dithering
7860 Whitepoint adaptation mode.
7862 The accepted values are:
7865 Bradford whitepoint adaptation
7868 von Kries whitepoint adaptation
7871 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7875 Override all input properties at once. Same accepted values as @ref{all}.
7878 Override input colorspace. Same accepted values as @ref{space}.
7881 Override input color primaries. Same accepted values as @ref{primaries}.
7884 Override input transfer characteristics. Same accepted values as @ref{trc}.
7887 Override input color range. Same accepted values as @ref{range}.
7891 The filter converts the transfer characteristics, color space and color
7892 primaries to the specified user values. The output value, if not specified,
7893 is set to a default value based on the "all" property. If that property is
7894 also not specified, the filter will log an error. The output color range and
7895 format default to the same value as the input color range and format. The
7896 input transfer characteristics, color space, color primaries and color range
7897 should be set on the input data. If any of these are missing, the filter will
7898 log an error and no conversion will take place.
7900 For example to convert the input to SMPTE-240M, use the command:
7902 colorspace=smpte240m
7905 @section convolution
7907 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7909 The filter accepts the following options:
7916 Set matrix for each plane.
7917 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7918 and from 1 to 49 odd number of signed integers in @var{row} mode.
7924 Set multiplier for calculated value for each plane.
7925 If unset or 0, it will be sum of all matrix elements.
7931 Set bias for each plane. This value is added to the result of the multiplication.
7932 Useful for making the overall image brighter or darker. Default is 0.0.
7938 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7939 Default is @var{square}.
7942 @subsection Examples
7948 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"
7954 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"
7960 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"
7966 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"
7970 Apply laplacian edge detector which includes diagonals:
7972 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"
7978 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"
7984 Apply 2D convolution of video stream in frequency domain using second stream
7987 The filter accepts the following options:
7991 Set which planes to process.
7994 Set which impulse video frames will be processed, can be @var{first}
7995 or @var{all}. Default is @var{all}.
7998 The @code{convolve} filter also supports the @ref{framesync} options.
8002 Copy the input video source unchanged to the output. This is mainly useful for
8007 Video filtering on GPU using Apple's CoreImage API on OSX.
8009 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8010 processed by video hardware. However, software-based OpenGL implementations
8011 exist which means there is no guarantee for hardware processing. It depends on
8014 There are many filters and image generators provided by Apple that come with a
8015 large variety of options. The filter has to be referenced by its name along
8018 The coreimage filter accepts the following options:
8021 List all available filters and generators along with all their respective
8022 options as well as possible minimum and maximum values along with the default
8029 Specify all filters by their respective name and options.
8030 Use @var{list_filters} to determine all valid filter names and options.
8031 Numerical options are specified by a float value and are automatically clamped
8032 to their respective value range. Vector and color options have to be specified
8033 by a list of space separated float values. Character escaping has to be done.
8034 A special option name @code{default} is available to use default options for a
8037 It is required to specify either @code{default} or at least one of the filter options.
8038 All omitted options are used with their default values.
8039 The syntax of the filter string is as follows:
8041 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8045 Specify a rectangle where the output of the filter chain is copied into the
8046 input image. It is given by a list of space separated float values:
8048 output_rect=x\ y\ width\ height
8050 If not given, the output rectangle equals the dimensions of the input image.
8051 The output rectangle is automatically cropped at the borders of the input
8052 image. Negative values are valid for each component.
8054 output_rect=25\ 25\ 100\ 100
8058 Several filters can be chained for successive processing without GPU-HOST
8059 transfers allowing for fast processing of complex filter chains.
8060 Currently, only filters with zero (generators) or exactly one (filters) input
8061 image and one output image are supported. Also, transition filters are not yet
8064 Some filters generate output images with additional padding depending on the
8065 respective filter kernel. The padding is automatically removed to ensure the
8066 filter output has the same size as the input image.
8068 For image generators, the size of the output image is determined by the
8069 previous output image of the filter chain or the input image of the whole
8070 filterchain, respectively. The generators do not use the pixel information of
8071 this image to generate their output. However, the generated output is
8072 blended onto this image, resulting in partial or complete coverage of the
8075 The @ref{coreimagesrc} video source can be used for generating input images
8076 which are directly fed into the filter chain. By using it, providing input
8077 images by another video source or an input video is not required.
8079 @subsection Examples
8084 List all filters available:
8086 coreimage=list_filters=true
8090 Use the CIBoxBlur filter with default options to blur an image:
8092 coreimage=filter=CIBoxBlur@@default
8096 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8097 its center at 100x100 and a radius of 50 pixels:
8099 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8103 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8104 given as complete and escaped command-line for Apple's standard bash shell:
8106 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8112 Cover a rectangular object
8114 It accepts the following options:
8118 Filepath of the optional cover image, needs to be in yuv420.
8123 It accepts the following values:
8126 cover it by the supplied image
8128 cover it by interpolating the surrounding pixels
8131 Default value is @var{blur}.
8134 @subsection Examples
8138 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8140 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8146 Crop the input video to given dimensions.
8148 It accepts the following parameters:
8152 The width of the output video. It defaults to @code{iw}.
8153 This expression is evaluated only once during the filter
8154 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8157 The height of the output video. It defaults to @code{ih}.
8158 This expression is evaluated only once during the filter
8159 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8162 The horizontal position, in the input video, of the left edge of the output
8163 video. It defaults to @code{(in_w-out_w)/2}.
8164 This expression is evaluated per-frame.
8167 The vertical position, in the input video, of the top edge of the output video.
8168 It defaults to @code{(in_h-out_h)/2}.
8169 This expression is evaluated per-frame.
8172 If set to 1 will force the output display aspect ratio
8173 to be the same of the input, by changing the output sample aspect
8174 ratio. It defaults to 0.
8177 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8178 width/height/x/y as specified and will not be rounded to nearest smaller value.
8182 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8183 expressions containing the following constants:
8188 The computed values for @var{x} and @var{y}. They are evaluated for
8193 The input width and height.
8197 These are the same as @var{in_w} and @var{in_h}.
8201 The output (cropped) width and height.
8205 These are the same as @var{out_w} and @var{out_h}.
8208 same as @var{iw} / @var{ih}
8211 input sample aspect ratio
8214 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8218 horizontal and vertical chroma subsample values. For example for the
8219 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8222 The number of the input frame, starting from 0.
8225 the position in the file of the input frame, NAN if unknown
8228 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
8232 The expression for @var{out_w} may depend on the value of @var{out_h},
8233 and the expression for @var{out_h} may depend on @var{out_w}, but they
8234 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
8235 evaluated after @var{out_w} and @var{out_h}.
8237 The @var{x} and @var{y} parameters specify the expressions for the
8238 position of the top-left corner of the output (non-cropped) area. They
8239 are evaluated for each frame. If the evaluated value is not valid, it
8240 is approximated to the nearest valid value.
8242 The expression for @var{x} may depend on @var{y}, and the expression
8243 for @var{y} may depend on @var{x}.
8245 @subsection Examples
8249 Crop area with size 100x100 at position (12,34).
8254 Using named options, the example above becomes:
8256 crop=w=100:h=100:x=12:y=34
8260 Crop the central input area with size 100x100:
8266 Crop the central input area with size 2/3 of the input video:
8268 crop=2/3*in_w:2/3*in_h
8272 Crop the input video central square:
8279 Delimit the rectangle with the top-left corner placed at position
8280 100:100 and the right-bottom corner corresponding to the right-bottom
8281 corner of the input image.
8283 crop=in_w-100:in_h-100:100:100
8287 Crop 10 pixels from the left and right borders, and 20 pixels from
8288 the top and bottom borders
8290 crop=in_w-2*10:in_h-2*20
8294 Keep only the bottom right quarter of the input image:
8296 crop=in_w/2:in_h/2:in_w/2:in_h/2
8300 Crop height for getting Greek harmony:
8302 crop=in_w:1/PHI*in_w
8306 Apply trembling effect:
8308 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)
8312 Apply erratic camera effect depending on timestamp:
8314 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)"
8318 Set x depending on the value of y:
8320 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
8324 @subsection Commands
8326 This filter supports the following commands:
8332 Set width/height of the output video and the horizontal/vertical position
8334 The command accepts the same syntax of the corresponding option.
8336 If the specified expression is not valid, it is kept at its current
8342 Auto-detect the crop size.
8344 It calculates the necessary cropping parameters and prints the
8345 recommended parameters via the logging system. The detected dimensions
8346 correspond to the non-black area of the input video.
8348 It accepts the following parameters:
8353 Set higher black value threshold, which can be optionally specified
8354 from nothing (0) to everything (255 for 8-bit based formats). An intensity
8355 value greater to the set value is considered non-black. It defaults to 24.
8356 You can also specify a value between 0.0 and 1.0 which will be scaled depending
8357 on the bitdepth of the pixel format.
8360 The value which the width/height should be divisible by. It defaults to
8361 16. The offset is automatically adjusted to center the video. Use 2 to
8362 get only even dimensions (needed for 4:2:2 video). 16 is best when
8363 encoding to most video codecs.
8365 @item reset_count, reset
8366 Set the counter that determines after how many frames cropdetect will
8367 reset the previously detected largest video area and start over to
8368 detect the current optimal crop area. Default value is 0.
8370 This can be useful when channel logos distort the video area. 0
8371 indicates 'never reset', and returns the largest area encountered during
8378 Delay video filtering until a given wallclock timestamp. The filter first
8379 passes on @option{preroll} amount of frames, then it buffers at most
8380 @option{buffer} amount of frames and waits for the cue. After reaching the cue
8381 it forwards the buffered frames and also any subsequent frames coming in its
8384 The filter can be used synchronize the output of multiple ffmpeg processes for
8385 realtime output devices like decklink. By putting the delay in the filtering
8386 chain and pre-buffering frames the process can pass on data to output almost
8387 immediately after the target wallclock timestamp is reached.
8389 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
8395 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
8398 The duration of content to pass on as preroll expressed in seconds. Default is 0.
8401 The maximum duration of content to buffer before waiting for the cue expressed
8402 in seconds. Default is 0.
8409 Apply color adjustments using curves.
8411 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
8412 component (red, green and blue) has its values defined by @var{N} key points
8413 tied from each other using a smooth curve. The x-axis represents the pixel
8414 values from the input frame, and the y-axis the new pixel values to be set for
8417 By default, a component curve is defined by the two points @var{(0;0)} and
8418 @var{(1;1)}. This creates a straight line where each original pixel value is
8419 "adjusted" to its own value, which means no change to the image.
8421 The filter allows you to redefine these two points and add some more. A new
8422 curve (using a natural cubic spline interpolation) will be define to pass
8423 smoothly through all these new coordinates. The new defined points needs to be
8424 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
8425 be in the @var{[0;1]} interval. If the computed curves happened to go outside
8426 the vector spaces, the values will be clipped accordingly.
8428 The filter accepts the following options:
8432 Select one of the available color presets. This option can be used in addition
8433 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
8434 options takes priority on the preset values.
8435 Available presets are:
8438 @item color_negative
8441 @item increase_contrast
8443 @item linear_contrast
8444 @item medium_contrast
8446 @item strong_contrast
8449 Default is @code{none}.
8451 Set the master key points. These points will define a second pass mapping. It
8452 is sometimes called a "luminance" or "value" mapping. It can be used with
8453 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
8454 post-processing LUT.
8456 Set the key points for the red component.
8458 Set the key points for the green component.
8460 Set the key points for the blue component.
8462 Set the key points for all components (not including master).
8463 Can be used in addition to the other key points component
8464 options. In this case, the unset component(s) will fallback on this
8465 @option{all} setting.
8467 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
8469 Save Gnuplot script of the curves in specified file.
8472 To avoid some filtergraph syntax conflicts, each key points list need to be
8473 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
8475 @subsection Examples
8479 Increase slightly the middle level of blue:
8481 curves=blue='0/0 0.5/0.58 1/1'
8487 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'
8489 Here we obtain the following coordinates for each components:
8492 @code{(0;0.11) (0.42;0.51) (1;0.95)}
8494 @code{(0;0) (0.50;0.48) (1;1)}
8496 @code{(0;0.22) (0.49;0.44) (1;0.80)}
8500 The previous example can also be achieved with the associated built-in preset:
8502 curves=preset=vintage
8512 Use a Photoshop preset and redefine the points of the green component:
8514 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
8518 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
8519 and @command{gnuplot}:
8521 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
8522 gnuplot -p /tmp/curves.plt
8528 Video data analysis filter.
8530 This filter shows hexadecimal pixel values of part of video.
8532 The filter accepts the following options:
8536 Set output video size.
8539 Set x offset from where to pick pixels.
8542 Set y offset from where to pick pixels.
8545 Set scope mode, can be one of the following:
8548 Draw hexadecimal pixel values with white color on black background.
8551 Draw hexadecimal pixel values with input video pixel color on black
8555 Draw hexadecimal pixel values on color background picked from input video,
8556 the text color is picked in such way so its always visible.
8560 Draw rows and columns numbers on left and top of video.
8563 Set background opacity.
8566 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
8570 Apply Directional blur filter.
8572 The filter accepts the following options:
8576 Set angle of directional blur. Default is @code{45}.
8579 Set radius of directional blur. Default is @code{5}.
8582 Set which planes to filter. By default all planes are filtered.
8585 @subsection Commands
8586 This filter supports same @ref{commands} as options.
8587 The command accepts the same syntax of the corresponding option.
8589 If the specified expression is not valid, it is kept at its current
8594 Denoise frames using 2D DCT (frequency domain filtering).
8596 This filter is not designed for real time.
8598 The filter accepts the following options:
8602 Set the noise sigma constant.
8604 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
8605 coefficient (absolute value) below this threshold with be dropped.
8607 If you need a more advanced filtering, see @option{expr}.
8609 Default is @code{0}.
8612 Set number overlapping pixels for each block. Since the filter can be slow, you
8613 may want to reduce this value, at the cost of a less effective filter and the
8614 risk of various artefacts.
8616 If the overlapping value doesn't permit processing the whole input width or
8617 height, a warning will be displayed and according borders won't be denoised.
8619 Default value is @var{blocksize}-1, which is the best possible setting.
8622 Set the coefficient factor expression.
8624 For each coefficient of a DCT block, this expression will be evaluated as a
8625 multiplier value for the coefficient.
8627 If this is option is set, the @option{sigma} option will be ignored.
8629 The absolute value of the coefficient can be accessed through the @var{c}
8633 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
8634 @var{blocksize}, which is the width and height of the processed blocks.
8636 The default value is @var{3} (8x8) and can be raised to @var{4} for a
8637 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
8638 on the speed processing. Also, a larger block size does not necessarily means a
8642 @subsection Examples
8644 Apply a denoise with a @option{sigma} of @code{4.5}:
8649 The same operation can be achieved using the expression system:
8651 dctdnoiz=e='gte(c, 4.5*3)'
8654 Violent denoise using a block size of @code{16x16}:
8661 Remove banding artifacts from input video.
8662 It works by replacing banded pixels with average value of referenced pixels.
8664 The filter accepts the following options:
8671 Set banding detection threshold for each plane. Default is 0.02.
8672 Valid range is 0.00003 to 0.5.
8673 If difference between current pixel and reference pixel is less than threshold,
8674 it will be considered as banded.
8677 Banding detection range in pixels. Default is 16. If positive, random number
8678 in range 0 to set value will be used. If negative, exact absolute value
8680 The range defines square of four pixels around current pixel.
8683 Set direction in radians from which four pixel will be compared. If positive,
8684 random direction from 0 to set direction will be picked. If negative, exact of
8685 absolute value will be picked. For example direction 0, -PI or -2*PI radians
8686 will pick only pixels on same row and -PI/2 will pick only pixels on same
8690 If enabled, current pixel is compared with average value of all four
8691 surrounding pixels. The default is enabled. If disabled current pixel is
8692 compared with all four surrounding pixels. The pixel is considered banded
8693 if only all four differences with surrounding pixels are less than threshold.
8696 If enabled, current pixel is changed if and only if all pixel components are banded,
8697 e.g. banding detection threshold is triggered for all color components.
8698 The default is disabled.
8703 Remove blocking artifacts from input video.
8705 The filter accepts the following options:
8709 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
8710 This controls what kind of deblocking is applied.
8713 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
8719 Set blocking detection thresholds. Allowed range is 0 to 1.
8720 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
8721 Using higher threshold gives more deblocking strength.
8722 Setting @var{alpha} controls threshold detection at exact edge of block.
8723 Remaining options controls threshold detection near the edge. Each one for
8724 below/above or left/right. Setting any of those to @var{0} disables
8728 Set planes to filter. Default is to filter all available planes.
8731 @subsection Examples
8735 Deblock using weak filter and block size of 4 pixels.
8737 deblock=filter=weak:block=4
8741 Deblock using strong filter, block size of 4 pixels and custom thresholds for
8742 deblocking more edges.
8744 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
8748 Similar as above, but filter only first plane.
8750 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
8754 Similar as above, but filter only second and third plane.
8756 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
8763 Drop duplicated frames at regular intervals.
8765 The filter accepts the following options:
8769 Set the number of frames from which one will be dropped. Setting this to
8770 @var{N} means one frame in every batch of @var{N} frames will be dropped.
8771 Default is @code{5}.
8774 Set the threshold for duplicate detection. If the difference metric for a frame
8775 is less than or equal to this value, then it is declared as duplicate. Default
8779 Set scene change threshold. Default is @code{15}.
8783 Set the size of the x and y-axis blocks used during metric calculations.
8784 Larger blocks give better noise suppression, but also give worse detection of
8785 small movements. Must be a power of two. Default is @code{32}.
8788 Mark main input as a pre-processed input and activate clean source input
8789 stream. This allows the input to be pre-processed with various filters to help
8790 the metrics calculation while keeping the frame selection lossless. When set to
8791 @code{1}, the first stream is for the pre-processed input, and the second
8792 stream is the clean source from where the kept frames are chosen. Default is
8796 Set whether or not chroma is considered in the metric calculations. Default is
8802 Apply 2D deconvolution of video stream in frequency domain using second stream
8805 The filter accepts the following options:
8809 Set which planes to process.
8812 Set which impulse video frames will be processed, can be @var{first}
8813 or @var{all}. Default is @var{all}.
8816 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
8817 and height are not same and not power of 2 or if stream prior to convolving
8821 The @code{deconvolve} filter also supports the @ref{framesync} options.
8825 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
8827 It accepts the following options:
8831 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
8832 @var{rainbows} for cross-color reduction.
8835 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
8838 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
8841 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
8844 Set temporal chroma threshold. Lower values increases reduction of cross-color.
8849 Apply deflate effect to the video.
8851 This filter replaces the pixel by the local(3x3) average by taking into account
8852 only values lower than the pixel.
8854 It accepts the following options:
8861 Limit the maximum change for each plane, default is 65535.
8862 If 0, plane will remain unchanged.
8865 @subsection Commands
8867 This filter supports the all above options as @ref{commands}.
8871 Remove temporal frame luminance variations.
8873 It accepts the following options:
8877 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8880 Set averaging mode to smooth temporal luminance variations.
8882 Available values are:
8907 Do not actually modify frame. Useful when one only wants metadata.
8912 Remove judder produced by partially interlaced telecined content.
8914 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8915 source was partially telecined content then the output of @code{pullup,dejudder}
8916 will have a variable frame rate. May change the recorded frame rate of the
8917 container. Aside from that change, this filter will not affect constant frame
8920 The option available in this filter is:
8924 Specify the length of the window over which the judder repeats.
8926 Accepts any integer greater than 1. Useful values are:
8930 If the original was telecined from 24 to 30 fps (Film to NTSC).
8933 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8936 If a mixture of the two.
8939 The default is @samp{4}.
8944 Suppress a TV station logo by a simple interpolation of the surrounding
8945 pixels. Just set a rectangle covering the logo and watch it disappear
8946 (and sometimes something even uglier appear - your mileage may vary).
8948 It accepts the following parameters:
8953 Specify the top left corner coordinates of the logo. They must be
8958 Specify the width and height of the logo to clear. They must be
8962 Specify the thickness of the fuzzy edge of the rectangle (added to
8963 @var{w} and @var{h}). The default value is 1. This option is
8964 deprecated, setting higher values should no longer be necessary and
8968 When set to 1, a green rectangle is drawn on the screen to simplify
8969 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8970 The default value is 0.
8972 The rectangle is drawn on the outermost pixels which will be (partly)
8973 replaced with interpolated values. The values of the next pixels
8974 immediately outside this rectangle in each direction will be used to
8975 compute the interpolated pixel values inside the rectangle.
8979 @subsection Examples
8983 Set a rectangle covering the area with top left corner coordinates 0,0
8984 and size 100x77, and a band of size 10:
8986 delogo=x=0:y=0:w=100:h=77:band=10
8994 Remove the rain in the input image/video by applying the derain methods based on
8995 convolutional neural networks. Supported models:
8999 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9000 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9003 Training as well as model generation scripts are provided in
9004 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9006 Native model files (.model) can be generated from TensorFlow model
9007 files (.pb) by using tools/python/convert.py
9009 The filter accepts the following options:
9013 Specify which filter to use. This option accepts the following values:
9017 Derain filter. To conduct derain filter, you need to use a derain model.
9020 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9022 Default value is @samp{derain}.
9025 Specify which DNN backend to use for model loading and execution. This option accepts
9026 the following values:
9030 Native implementation of DNN loading and execution.
9033 TensorFlow backend. To enable this backend you
9034 need to install the TensorFlow for C library (see
9035 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9036 @code{--enable-libtensorflow}
9038 Default value is @samp{native}.
9041 Set path to model file specifying network architecture and its parameters.
9042 Note that different backends use different file formats. TensorFlow and native
9043 backend can load files for only its format.
9046 It can also be finished with @ref{dnn_processing} filter.
9050 Attempt to fix small changes in horizontal and/or vertical shift. This
9051 filter helps remove camera shake from hand-holding a camera, bumping a
9052 tripod, moving on a vehicle, etc.
9054 The filter accepts the following options:
9062 Specify a rectangular area where to limit the search for motion
9064 If desired the search for motion vectors can be limited to a
9065 rectangular area of the frame defined by its top left corner, width
9066 and height. These parameters have the same meaning as the drawbox
9067 filter which can be used to visualise the position of the bounding
9070 This is useful when simultaneous movement of subjects within the frame
9071 might be confused for camera motion by the motion vector search.
9073 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9074 then the full frame is used. This allows later options to be set
9075 without specifying the bounding box for the motion vector search.
9077 Default - search the whole frame.
9081 Specify the maximum extent of movement in x and y directions in the
9082 range 0-64 pixels. Default 16.
9085 Specify how to generate pixels to fill blanks at the edge of the
9086 frame. Available values are:
9089 Fill zeroes at blank locations
9091 Original image at blank locations
9093 Extruded edge value at blank locations
9095 Mirrored edge at blank locations
9097 Default value is @samp{mirror}.
9100 Specify the blocksize to use for motion search. Range 4-128 pixels,
9104 Specify the contrast threshold for blocks. Only blocks with more than
9105 the specified contrast (difference between darkest and lightest
9106 pixels) will be considered. Range 1-255, default 125.
9109 Specify the search strategy. Available values are:
9112 Set exhaustive search
9114 Set less exhaustive search.
9116 Default value is @samp{exhaustive}.
9119 If set then a detailed log of the motion search is written to the
9126 Remove unwanted contamination of foreground colors, caused by reflected color of
9127 greenscreen or bluescreen.
9129 This filter accepts the following options:
9133 Set what type of despill to use.
9136 Set how spillmap will be generated.
9139 Set how much to get rid of still remaining spill.
9142 Controls amount of red in spill area.
9145 Controls amount of green in spill area.
9146 Should be -1 for greenscreen.
9149 Controls amount of blue in spill area.
9150 Should be -1 for bluescreen.
9153 Controls brightness of spill area, preserving colors.
9156 Modify alpha from generated spillmap.
9161 Apply an exact inverse of the telecine operation. It requires a predefined
9162 pattern specified using the pattern option which must be the same as that passed
9163 to the telecine filter.
9165 This filter accepts the following options:
9174 The default value is @code{top}.
9178 A string of numbers representing the pulldown pattern you wish to apply.
9179 The default value is @code{23}.
9182 A number representing position of the first frame with respect to the telecine
9183 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9188 Apply dilation effect to the video.
9190 This filter replaces the pixel by the local(3x3) maximum.
9192 It accepts the following options:
9199 Limit the maximum change for each plane, default is 65535.
9200 If 0, plane will remain unchanged.
9203 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9206 Flags to local 3x3 coordinates maps like this:
9213 @subsection Commands
9215 This filter supports the all above options as @ref{commands}.
9219 Displace pixels as indicated by second and third input stream.
9221 It takes three input streams and outputs one stream, the first input is the
9222 source, and second and third input are displacement maps.
9224 The second input specifies how much to displace pixels along the
9225 x-axis, while the third input specifies how much to displace pixels
9227 If one of displacement map streams terminates, last frame from that
9228 displacement map will be used.
9230 Note that once generated, displacements maps can be reused over and over again.
9232 A description of the accepted options follows.
9236 Set displace behavior for pixels that are out of range.
9238 Available values are:
9241 Missing pixels are replaced by black pixels.
9244 Adjacent pixels will spread out to replace missing pixels.
9247 Out of range pixels are wrapped so they point to pixels of other side.
9250 Out of range pixels will be replaced with mirrored pixels.
9252 Default is @samp{smear}.
9256 @subsection Examples
9260 Add ripple effect to rgb input of video size hd720:
9262 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
9266 Add wave effect to rgb input of video size hd720:
9268 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
9272 @anchor{dnn_processing}
9273 @section dnn_processing
9275 Do image processing with deep neural networks. It works together with another filter
9276 which converts the pixel format of the Frame to what the dnn network requires.
9278 The filter accepts the following options:
9282 Specify which DNN backend to use for model loading and execution. This option accepts
9283 the following values:
9287 Native implementation of DNN loading and execution.
9290 TensorFlow backend. To enable this backend you
9291 need to install the TensorFlow for C library (see
9292 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9293 @code{--enable-libtensorflow}
9296 OpenVINO backend. To enable this backend you
9297 need to build and install the OpenVINO for C library (see
9298 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
9299 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
9300 be needed if the header files and libraries are not installed into system path)
9304 Default value is @samp{native}.
9307 Set path to model file specifying network architecture and its parameters.
9308 Note that different backends use different file formats. TensorFlow, OpenVINO and native
9309 backend can load files for only its format.
9311 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
9314 Set the input name of the dnn network.
9317 Set the output name of the dnn network.
9321 @subsection Examples
9325 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
9327 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
9331 Halve the pixel value of the frame with format gray32f:
9333 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
9337 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
9339 ./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
9343 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
9345 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
9352 Draw a colored box on the input image.
9354 It accepts the following parameters:
9359 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
9363 The expressions which specify the width and height of the box; if 0 they are interpreted as
9364 the input width and height. It defaults to 0.
9367 Specify the color of the box to write. For the general syntax of this option,
9368 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
9369 value @code{invert} is used, the box edge color is the same as the
9370 video with inverted luma.
9373 The expression which sets the thickness of the box edge.
9374 A value of @code{fill} will create a filled box. Default value is @code{3}.
9376 See below for the list of accepted constants.
9379 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
9380 will overwrite the video's color and alpha pixels.
9381 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
9384 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
9385 following constants:
9389 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
9393 horizontal and vertical chroma subsample values. For example for the
9394 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9398 The input width and height.
9401 The input sample aspect ratio.
9405 The x and y offset coordinates where the box is drawn.
9409 The width and height of the drawn box.
9412 The thickness of the drawn box.
9414 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
9415 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
9419 @subsection Examples
9423 Draw a black box around the edge of the input image:
9429 Draw a box with color red and an opacity of 50%:
9431 drawbox=10:20:200:60:red@@0.5
9434 The previous example can be specified as:
9436 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
9440 Fill the box with pink color:
9442 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
9446 Draw a 2-pixel red 2.40:1 mask:
9448 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
9452 @subsection Commands
9453 This filter supports same commands as options.
9454 The command accepts the same syntax of the corresponding option.
9456 If the specified expression is not valid, it is kept at its current
9461 Draw a graph using input video metadata.
9463 It accepts the following parameters:
9467 Set 1st frame metadata key from which metadata values will be used to draw a graph.
9470 Set 1st foreground color expression.
9473 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
9476 Set 2nd foreground color expression.
9479 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
9482 Set 3rd foreground color expression.
9485 Set 4th frame metadata key from which metadata values will be used to draw a graph.
9488 Set 4th foreground color expression.
9491 Set minimal value of metadata value.
9494 Set maximal value of metadata value.
9497 Set graph background color. Default is white.
9502 Available values for mode is:
9509 Default is @code{line}.
9514 Available values for slide is:
9517 Draw new frame when right border is reached.
9520 Replace old columns with new ones.
9523 Scroll from right to left.
9526 Scroll from left to right.
9529 Draw single picture.
9532 Default is @code{frame}.
9535 Set size of graph video. For the syntax of this option, check the
9536 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
9537 The default value is @code{900x256}.
9540 Set the output frame rate. Default value is @code{25}.
9542 The foreground color expressions can use the following variables:
9545 Minimal value of metadata value.
9548 Maximal value of metadata value.
9551 Current metadata key value.
9554 The color is defined as 0xAABBGGRR.
9557 Example using metadata from @ref{signalstats} filter:
9559 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
9562 Example using metadata from @ref{ebur128} filter:
9564 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
9569 Draw a grid on the input image.
9571 It accepts the following parameters:
9576 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
9580 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
9581 input width and height, respectively, minus @code{thickness}, so image gets
9582 framed. Default to 0.
9585 Specify the color of the grid. For the general syntax of this option,
9586 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
9587 value @code{invert} is used, the grid color is the same as the
9588 video with inverted luma.
9591 The expression which sets the thickness of the grid line. Default value is @code{1}.
9593 See below for the list of accepted constants.
9596 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
9597 will overwrite the video's color and alpha pixels.
9598 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
9601 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
9602 following constants:
9606 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
9610 horizontal and vertical chroma subsample values. For example for the
9611 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9615 The input grid cell width and height.
9618 The input sample aspect ratio.
9622 The x and y coordinates of some point of grid intersection (meant to configure offset).
9626 The width and height of the drawn cell.
9629 The thickness of the drawn cell.
9631 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
9632 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
9636 @subsection Examples
9640 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
9642 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
9646 Draw a white 3x3 grid with an opacity of 50%:
9648 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
9652 @subsection Commands
9653 This filter supports same commands as options.
9654 The command accepts the same syntax of the corresponding option.
9656 If the specified expression is not valid, it is kept at its current
9662 Draw a text string or text from a specified file on top of a video, using the
9663 libfreetype library.
9665 To enable compilation of this filter, you need to configure FFmpeg with
9666 @code{--enable-libfreetype}.
9667 To enable default font fallback and the @var{font} option you need to
9668 configure FFmpeg with @code{--enable-libfontconfig}.
9669 To enable the @var{text_shaping} option, you need to configure FFmpeg with
9670 @code{--enable-libfribidi}.
9674 It accepts the following parameters:
9679 Used to draw a box around text using the background color.
9680 The value must be either 1 (enable) or 0 (disable).
9681 The default value of @var{box} is 0.
9684 Set the width of the border to be drawn around the box using @var{boxcolor}.
9685 The default value of @var{boxborderw} is 0.
9688 The color to be used for drawing box around text. For the syntax of this
9689 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9691 The default value of @var{boxcolor} is "white".
9694 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
9695 The default value of @var{line_spacing} is 0.
9698 Set the width of the border to be drawn around the text using @var{bordercolor}.
9699 The default value of @var{borderw} is 0.
9702 Set the color to be used for drawing border around text. For the syntax of this
9703 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9705 The default value of @var{bordercolor} is "black".
9708 Select how the @var{text} is expanded. Can be either @code{none},
9709 @code{strftime} (deprecated) or
9710 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
9714 Set a start time for the count. Value is in microseconds. Only applied
9715 in the deprecated strftime expansion mode. To emulate in normal expansion
9716 mode use the @code{pts} function, supplying the start time (in seconds)
9717 as the second argument.
9720 If true, check and fix text coords to avoid clipping.
9723 The color to be used for drawing fonts. For the syntax of this option, check
9724 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
9726 The default value of @var{fontcolor} is "black".
9728 @item fontcolor_expr
9729 String which is expanded the same way as @var{text} to obtain dynamic
9730 @var{fontcolor} value. By default this option has empty value and is not
9731 processed. When this option is set, it overrides @var{fontcolor} option.
9734 The font family to be used for drawing text. By default Sans.
9737 The font file to be used for drawing text. The path must be included.
9738 This parameter is mandatory if the fontconfig support is disabled.
9741 Draw the text applying alpha blending. The value can
9742 be a number between 0.0 and 1.0.
9743 The expression accepts the same variables @var{x, y} as well.
9744 The default value is 1.
9745 Please see @var{fontcolor_expr}.
9748 The font size to be used for drawing text.
9749 The default value of @var{fontsize} is 16.
9752 If set to 1, attempt to shape the text (for example, reverse the order of
9753 right-to-left text and join Arabic characters) before drawing it.
9754 Otherwise, just draw the text exactly as given.
9755 By default 1 (if supported).
9758 The flags to be used for loading the fonts.
9760 The flags map the corresponding flags supported by libfreetype, and are
9761 a combination of the following values:
9768 @item vertical_layout
9769 @item force_autohint
9772 @item ignore_global_advance_width
9774 @item ignore_transform
9780 Default value is "default".
9782 For more information consult the documentation for the FT_LOAD_*
9786 The color to be used for drawing a shadow behind the drawn text. For the
9787 syntax of this option, check the @ref{color syntax,,"Color" section in the
9788 ffmpeg-utils manual,ffmpeg-utils}.
9790 The default value of @var{shadowcolor} is "black".
9794 The x and y offsets for the text shadow position with respect to the
9795 position of the text. They can be either positive or negative
9796 values. The default value for both is "0".
9799 The starting frame number for the n/frame_num variable. The default value
9803 The size in number of spaces to use for rendering the tab.
9807 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
9808 format. It can be used with or without text parameter. @var{timecode_rate}
9809 option must be specified.
9811 @item timecode_rate, rate, r
9812 Set the timecode frame rate (timecode only). Value will be rounded to nearest
9813 integer. Minimum value is "1".
9814 Drop-frame timecode is supported for frame rates 30 & 60.
9817 If set to 1, the output of the timecode option will wrap around at 24 hours.
9818 Default is 0 (disabled).
9821 The text string to be drawn. The text must be a sequence of UTF-8
9823 This parameter is mandatory if no file is specified with the parameter
9827 A text file containing text to be drawn. The text must be a sequence
9828 of UTF-8 encoded characters.
9830 This parameter is mandatory if no text string is specified with the
9831 parameter @var{text}.
9833 If both @var{text} and @var{textfile} are specified, an error is thrown.
9836 If set to 1, the @var{textfile} will be reloaded before each frame.
9837 Be sure to update it atomically, or it may be read partially, or even fail.
9841 The expressions which specify the offsets where text will be drawn
9842 within the video frame. They are relative to the top/left border of the
9845 The default value of @var{x} and @var{y} is "0".
9847 See below for the list of accepted constants and functions.
9850 The parameters for @var{x} and @var{y} are expressions containing the
9851 following constants and functions:
9855 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
9859 horizontal and vertical chroma subsample values. For example for the
9860 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9863 the height of each text line
9871 @item max_glyph_a, ascent
9872 the maximum distance from the baseline to the highest/upper grid
9873 coordinate used to place a glyph outline point, for all the rendered
9875 It is a positive value, due to the grid's orientation with the Y axis
9878 @item max_glyph_d, descent
9879 the maximum distance from the baseline to the lowest grid coordinate
9880 used to place a glyph outline point, for all the rendered glyphs.
9881 This is a negative value, due to the grid's orientation, with the Y axis
9885 maximum glyph height, that is the maximum height for all the glyphs
9886 contained in the rendered text, it is equivalent to @var{ascent} -
9890 maximum glyph width, that is the maximum width for all the glyphs
9891 contained in the rendered text
9894 the number of input frame, starting from 0
9896 @item rand(min, max)
9897 return a random number included between @var{min} and @var{max}
9900 The input sample aspect ratio.
9903 timestamp expressed in seconds, NAN if the input timestamp is unknown
9906 the height of the rendered text
9909 the width of the rendered text
9913 the x and y offset coordinates where the text is drawn.
9915 These parameters allow the @var{x} and @var{y} expressions to refer
9916 to each other, so you can for example specify @code{y=x/dar}.
9919 A one character description of the current frame's picture type.
9922 The current packet's position in the input file or stream
9923 (in bytes, from the start of the input). A value of -1 indicates
9924 this info is not available.
9927 The current packet's duration, in seconds.
9930 The current packet's size (in bytes).
9933 @anchor{drawtext_expansion}
9934 @subsection Text expansion
9936 If @option{expansion} is set to @code{strftime},
9937 the filter recognizes strftime() sequences in the provided text and
9938 expands them accordingly. Check the documentation of strftime(). This
9939 feature is deprecated.
9941 If @option{expansion} is set to @code{none}, the text is printed verbatim.
9943 If @option{expansion} is set to @code{normal} (which is the default),
9944 the following expansion mechanism is used.
9946 The backslash character @samp{\}, followed by any character, always expands to
9947 the second character.
9949 Sequences of the form @code{%@{...@}} are expanded. The text between the
9950 braces is a function name, possibly followed by arguments separated by ':'.
9951 If the arguments contain special characters or delimiters (':' or '@}'),
9952 they should be escaped.
9954 Note that they probably must also be escaped as the value for the
9955 @option{text} option in the filter argument string and as the filter
9956 argument in the filtergraph description, and possibly also for the shell,
9957 that makes up to four levels of escaping; using a text file avoids these
9960 The following functions are available:
9965 The expression evaluation result.
9967 It must take one argument specifying the expression to be evaluated,
9968 which accepts the same constants and functions as the @var{x} and
9969 @var{y} values. Note that not all constants should be used, for
9970 example the text size is not known when evaluating the expression, so
9971 the constants @var{text_w} and @var{text_h} will have an undefined
9974 @item expr_int_format, eif
9975 Evaluate the expression's value and output as formatted integer.
9977 The first argument is the expression to be evaluated, just as for the @var{expr} function.
9978 The second argument specifies the output format. Allowed values are @samp{x},
9979 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
9980 @code{printf} function.
9981 The third parameter is optional and sets the number of positions taken by the output.
9982 It can be used to add padding with zeros from the left.
9985 The time at which the filter is running, expressed in UTC.
9986 It can accept an argument: a strftime() format string.
9989 The time at which the filter is running, expressed in the local time zone.
9990 It can accept an argument: a strftime() format string.
9993 Frame metadata. Takes one or two arguments.
9995 The first argument is mandatory and specifies the metadata key.
9997 The second argument is optional and specifies a default value, used when the
9998 metadata key is not found or empty.
10000 Available metadata can be identified by inspecting entries
10001 starting with TAG included within each frame section
10002 printed by running @code{ffprobe -show_frames}.
10004 String metadata generated in filters leading to
10005 the drawtext filter are also available.
10008 The frame number, starting from 0.
10011 A one character description of the current picture type.
10014 The timestamp of the current frame.
10015 It can take up to three arguments.
10017 The first argument is the format of the timestamp; it defaults to @code{flt}
10018 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10019 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10020 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10021 @code{localtime} stands for the timestamp of the frame formatted as
10022 local time zone time.
10024 The second argument is an offset added to the timestamp.
10026 If the format is set to @code{hms}, a third argument @code{24HH} may be
10027 supplied to present the hour part of the formatted timestamp in 24h format
10030 If the format is set to @code{localtime} or @code{gmtime},
10031 a third argument may be supplied: a strftime() format string.
10032 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10035 @subsection Commands
10037 This filter supports altering parameters via commands:
10040 Alter existing filter parameters.
10042 Syntax for the argument is the same as for filter invocation, e.g.
10045 fontsize=56:fontcolor=green:text='Hello World'
10048 Full filter invocation with sendcmd would look like this:
10051 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10055 If the entire argument can't be parsed or applied as valid values then the filter will
10056 continue with its existing parameters.
10058 @subsection Examples
10062 Draw "Test Text" with font FreeSerif, using the default values for the
10063 optional parameters.
10066 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10070 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10071 and y=50 (counting from the top-left corner of the screen), text is
10072 yellow with a red box around it. Both the text and the box have an
10076 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10077 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10080 Note that the double quotes are not necessary if spaces are not used
10081 within the parameter list.
10084 Show the text at the center of the video frame:
10086 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10090 Show the text at a random position, switching to a new position every 30 seconds:
10092 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)"
10096 Show a text line sliding from right to left in the last row of the video
10097 frame. The file @file{LONG_LINE} is assumed to contain a single line
10100 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10104 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10106 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10110 Draw a single green letter "g", at the center of the input video.
10111 The glyph baseline is placed at half screen height.
10113 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10117 Show text for 1 second every 3 seconds:
10119 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10123 Use fontconfig to set the font. Note that the colons need to be escaped.
10125 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10129 Print the date of a real-time encoding (see strftime(3)):
10131 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10135 Show text fading in and out (appearing/disappearing):
10138 DS=1.0 # display start
10139 DE=10.0 # display end
10140 FID=1.5 # fade in duration
10141 FOD=5 # fade out duration
10142 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 @}"
10146 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10147 and the @option{fontsize} value are included in the @option{y} offset.
10149 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10150 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10154 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10155 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10156 must have option @option{-export_path_metadata 1} for the special metadata fields
10157 to be available for filters.
10159 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10164 For more information about libfreetype, check:
10165 @url{http://www.freetype.org/}.
10167 For more information about fontconfig, check:
10168 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10170 For more information about libfribidi, check:
10171 @url{http://fribidi.org/}.
10173 @section edgedetect
10175 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10177 The filter accepts the following options:
10182 Set low and high threshold values used by the Canny thresholding
10185 The high threshold selects the "strong" edge pixels, which are then
10186 connected through 8-connectivity with the "weak" edge pixels selected
10187 by the low threshold.
10189 @var{low} and @var{high} threshold values must be chosen in the range
10190 [0,1], and @var{low} should be lesser or equal to @var{high}.
10192 Default value for @var{low} is @code{20/255}, and default value for @var{high}
10196 Define the drawing mode.
10200 Draw white/gray wires on black background.
10203 Mix the colors to create a paint/cartoon effect.
10206 Apply Canny edge detector on all selected planes.
10208 Default value is @var{wires}.
10211 Select planes for filtering. By default all available planes are filtered.
10214 @subsection Examples
10218 Standard edge detection with custom values for the hysteresis thresholding:
10220 edgedetect=low=0.1:high=0.4
10224 Painting effect without thresholding:
10226 edgedetect=mode=colormix:high=0
10232 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
10234 For each input image, the filter will compute the optimal mapping from
10235 the input to the output given the codebook length, that is the number
10236 of distinct output colors.
10238 This filter accepts the following options.
10241 @item codebook_length, l
10242 Set codebook length. The value must be a positive integer, and
10243 represents the number of distinct output colors. Default value is 256.
10246 Set the maximum number of iterations to apply for computing the optimal
10247 mapping. The higher the value the better the result and the higher the
10248 computation time. Default value is 1.
10251 Set a random seed, must be an integer included between 0 and
10252 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
10253 will try to use a good random seed on a best effort basis.
10256 Set pal8 output pixel format. This option does not work with codebook
10257 length greater than 256.
10262 Measure graylevel entropy in histogram of color channels of video frames.
10264 It accepts the following parameters:
10268 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
10270 @var{diff} mode measures entropy of histogram delta values, absolute differences
10271 between neighbour histogram values.
10275 Set brightness, contrast, saturation and approximate gamma adjustment.
10277 The filter accepts the following options:
10281 Set the contrast expression. The value must be a float value in range
10282 @code{-1000.0} to @code{1000.0}. The default value is "1".
10285 Set the brightness expression. The value must be a float value in
10286 range @code{-1.0} to @code{1.0}. The default value is "0".
10289 Set the saturation expression. The value must be a float in
10290 range @code{0.0} to @code{3.0}. The default value is "1".
10293 Set the gamma expression. The value must be a float in range
10294 @code{0.1} to @code{10.0}. The default value is "1".
10297 Set the gamma expression for red. The value must be a float in
10298 range @code{0.1} to @code{10.0}. The default value is "1".
10301 Set the gamma expression for green. The value must be a float in range
10302 @code{0.1} to @code{10.0}. The default value is "1".
10305 Set the gamma expression for blue. The value must be a float in range
10306 @code{0.1} to @code{10.0}. The default value is "1".
10309 Set the gamma weight expression. It can be used to reduce the effect
10310 of a high gamma value on bright image areas, e.g. keep them from
10311 getting overamplified and just plain white. The value must be a float
10312 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
10313 gamma correction all the way down while @code{1.0} leaves it at its
10314 full strength. Default is "1".
10317 Set when the expressions for brightness, contrast, saturation and
10318 gamma expressions are evaluated.
10320 It accepts the following values:
10323 only evaluate expressions once during the filter initialization or
10324 when a command is processed
10327 evaluate expressions for each incoming frame
10330 Default value is @samp{init}.
10333 The expressions accept the following parameters:
10336 frame count of the input frame starting from 0
10339 byte position of the corresponding packet in the input file, NAN if
10343 frame rate of the input video, NAN if the input frame rate is unknown
10346 timestamp expressed in seconds, NAN if the input timestamp is unknown
10349 @subsection Commands
10350 The filter supports the following commands:
10354 Set the contrast expression.
10357 Set the brightness expression.
10360 Set the saturation expression.
10363 Set the gamma expression.
10366 Set the gamma_r expression.
10369 Set gamma_g expression.
10372 Set gamma_b expression.
10375 Set gamma_weight expression.
10377 The command accepts the same syntax of the corresponding option.
10379 If the specified expression is not valid, it is kept at its current
10386 Apply erosion effect to the video.
10388 This filter replaces the pixel by the local(3x3) minimum.
10390 It accepts the following options:
10397 Limit the maximum change for each plane, default is 65535.
10398 If 0, plane will remain unchanged.
10401 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10404 Flags to local 3x3 coordinates maps like this:
10411 @subsection Commands
10413 This filter supports the all above options as @ref{commands}.
10415 @section extractplanes
10417 Extract color channel components from input video stream into
10418 separate grayscale video streams.
10420 The filter accepts the following option:
10424 Set plane(s) to extract.
10426 Available values for planes are:
10437 Choosing planes not available in the input will result in an error.
10438 That means you cannot select @code{r}, @code{g}, @code{b} planes
10439 with @code{y}, @code{u}, @code{v} planes at same time.
10442 @subsection Examples
10446 Extract luma, u and v color channel component from input video frame
10447 into 3 grayscale outputs:
10449 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
10455 Apply a fade-in/out effect to the input video.
10457 It accepts the following parameters:
10461 The effect type can be either "in" for a fade-in, or "out" for a fade-out
10463 Default is @code{in}.
10465 @item start_frame, s
10466 Specify the number of the frame to start applying the fade
10467 effect at. Default is 0.
10470 The number of frames that the fade effect lasts. At the end of the
10471 fade-in effect, the output video will have the same intensity as the input video.
10472 At the end of the fade-out transition, the output video will be filled with the
10473 selected @option{color}.
10477 If set to 1, fade only alpha channel, if one exists on the input.
10478 Default value is 0.
10480 @item start_time, st
10481 Specify the timestamp (in seconds) of the frame to start to apply the fade
10482 effect. If both start_frame and start_time are specified, the fade will start at
10483 whichever comes last. Default is 0.
10486 The number of seconds for which the fade effect has to last. At the end of the
10487 fade-in effect the output video will have the same intensity as the input video,
10488 at the end of the fade-out transition the output video will be filled with the
10489 selected @option{color}.
10490 If both duration and nb_frames are specified, duration is used. Default is 0
10491 (nb_frames is used by default).
10494 Specify the color of the fade. Default is "black".
10497 @subsection Examples
10501 Fade in the first 30 frames of video:
10506 The command above is equivalent to:
10512 Fade out the last 45 frames of a 200-frame video:
10515 fade=type=out:start_frame=155:nb_frames=45
10519 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
10521 fade=in:0:25, fade=out:975:25
10525 Make the first 5 frames yellow, then fade in from frame 5-24:
10527 fade=in:5:20:color=yellow
10531 Fade in alpha over first 25 frames of video:
10533 fade=in:0:25:alpha=1
10537 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
10539 fade=t=in:st=5.5:d=0.5
10545 Denoise frames using 3D FFT (frequency domain filtering).
10547 The filter accepts the following options:
10551 Set the noise sigma constant. This sets denoising strength.
10552 Default value is 1. Allowed range is from 0 to 30.
10553 Using very high sigma with low overlap may give blocking artifacts.
10556 Set amount of denoising. By default all detected noise is reduced.
10557 Default value is 1. Allowed range is from 0 to 1.
10560 Set size of block, Default is 4, can be 3, 4, 5 or 6.
10561 Actual size of block in pixels is 2 to power of @var{block}, so by default
10562 block size in pixels is 2^4 which is 16.
10565 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
10568 Set number of previous frames to use for denoising. By default is set to 0.
10571 Set number of next frames to to use for denoising. By default is set to 0.
10574 Set planes which will be filtered, by default are all available filtered
10579 Apply arbitrary expressions to samples in frequency domain
10583 Adjust the dc value (gain) of the luma plane of the image. The filter
10584 accepts an integer value in range @code{0} to @code{1000}. The default
10585 value is set to @code{0}.
10588 Adjust the dc value (gain) of the 1st chroma plane of the image. The
10589 filter accepts an integer value in range @code{0} to @code{1000}. The
10590 default value is set to @code{0}.
10593 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
10594 filter accepts an integer value in range @code{0} to @code{1000}. The
10595 default value is set to @code{0}.
10598 Set the frequency domain weight expression for the luma plane.
10601 Set the frequency domain weight expression for the 1st chroma plane.
10604 Set the frequency domain weight expression for the 2nd chroma plane.
10607 Set when the expressions are evaluated.
10609 It accepts the following values:
10612 Only evaluate expressions once during the filter initialization.
10615 Evaluate expressions for each incoming frame.
10618 Default value is @samp{init}.
10620 The filter accepts the following variables:
10623 The coordinates of the current sample.
10627 The width and height of the image.
10630 The number of input frame, starting from 0.
10633 @subsection Examples
10639 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
10645 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
10651 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
10657 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
10664 Extract a single field from an interlaced image using stride
10665 arithmetic to avoid wasting CPU time. The output frames are marked as
10668 The filter accepts the following options:
10672 Specify whether to extract the top (if the value is @code{0} or
10673 @code{top}) or the bottom field (if the value is @code{1} or
10679 Create new frames by copying the top and bottom fields from surrounding frames
10680 supplied as numbers by the hint file.
10684 Set file containing hints: absolute/relative frame numbers.
10686 There must be one line for each frame in a clip. Each line must contain two
10687 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
10688 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
10689 is current frame number for @code{absolute} mode or out of [-1, 1] range
10690 for @code{relative} mode. First number tells from which frame to pick up top
10691 field and second number tells from which frame to pick up bottom field.
10693 If optionally followed by @code{+} output frame will be marked as interlaced,
10694 else if followed by @code{-} output frame will be marked as progressive, else
10695 it will be marked same as input frame.
10696 If optionally followed by @code{t} output frame will use only top field, or in
10697 case of @code{b} it will use only bottom field.
10698 If line starts with @code{#} or @code{;} that line is skipped.
10701 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
10704 Example of first several lines of @code{hint} file for @code{relative} mode:
10706 0,0 - # first frame
10707 1,0 - # second frame, use third's frame top field and second's frame bottom field
10708 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
10723 @section fieldmatch
10725 Field matching filter for inverse telecine. It is meant to reconstruct the
10726 progressive frames from a telecined stream. The filter does not drop duplicated
10727 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
10728 followed by a decimation filter such as @ref{decimate} in the filtergraph.
10730 The separation of the field matching and the decimation is notably motivated by
10731 the possibility of inserting a de-interlacing filter fallback between the two.
10732 If the source has mixed telecined and real interlaced content,
10733 @code{fieldmatch} will not be able to match fields for the interlaced parts.
10734 But these remaining combed frames will be marked as interlaced, and thus can be
10735 de-interlaced by a later filter such as @ref{yadif} before decimation.
10737 In addition to the various configuration options, @code{fieldmatch} can take an
10738 optional second stream, activated through the @option{ppsrc} option. If
10739 enabled, the frames reconstruction will be based on the fields and frames from
10740 this second stream. This allows the first input to be pre-processed in order to
10741 help the various algorithms of the filter, while keeping the output lossless
10742 (assuming the fields are matched properly). Typically, a field-aware denoiser,
10743 or brightness/contrast adjustments can help.
10745 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
10746 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
10747 which @code{fieldmatch} is based on. While the semantic and usage are very
10748 close, some behaviour and options names can differ.
10750 The @ref{decimate} filter currently only works for constant frame rate input.
10751 If your input has mixed telecined (30fps) and progressive content with a lower
10752 framerate like 24fps use the following filterchain to produce the necessary cfr
10753 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
10755 The filter accepts the following options:
10759 Specify the assumed field order of the input stream. Available values are:
10763 Auto detect parity (use FFmpeg's internal parity value).
10765 Assume bottom field first.
10767 Assume top field first.
10770 Note that it is sometimes recommended not to trust the parity announced by the
10773 Default value is @var{auto}.
10776 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
10777 sense that it won't risk creating jerkiness due to duplicate frames when
10778 possible, but if there are bad edits or blended fields it will end up
10779 outputting combed frames when a good match might actually exist. On the other
10780 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
10781 but will almost always find a good frame if there is one. The other values are
10782 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
10783 jerkiness and creating duplicate frames versus finding good matches in sections
10784 with bad edits, orphaned fields, blended fields, etc.
10786 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
10788 Available values are:
10792 2-way matching (p/c)
10794 2-way matching, and trying 3rd match if still combed (p/c + n)
10796 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
10798 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
10799 still combed (p/c + n + u/b)
10801 3-way matching (p/c/n)
10803 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
10804 detected as combed (p/c/n + u/b)
10807 The parenthesis at the end indicate the matches that would be used for that
10808 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
10811 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
10814 Default value is @var{pc_n}.
10817 Mark the main input stream as a pre-processed input, and enable the secondary
10818 input stream as the clean source to pick the fields from. See the filter
10819 introduction for more details. It is similar to the @option{clip2} feature from
10822 Default value is @code{0} (disabled).
10825 Set the field to match from. It is recommended to set this to the same value as
10826 @option{order} unless you experience matching failures with that setting. In
10827 certain circumstances changing the field that is used to match from can have a
10828 large impact on matching performance. Available values are:
10832 Automatic (same value as @option{order}).
10834 Match from the bottom field.
10836 Match from the top field.
10839 Default value is @var{auto}.
10842 Set whether or not chroma is included during the match comparisons. In most
10843 cases it is recommended to leave this enabled. You should set this to @code{0}
10844 only if your clip has bad chroma problems such as heavy rainbowing or other
10845 artifacts. Setting this to @code{0} could also be used to speed things up at
10846 the cost of some accuracy.
10848 Default value is @code{1}.
10852 These define an exclusion band which excludes the lines between @option{y0} and
10853 @option{y1} from being included in the field matching decision. An exclusion
10854 band can be used to ignore subtitles, a logo, or other things that may
10855 interfere with the matching. @option{y0} sets the starting scan line and
10856 @option{y1} sets the ending line; all lines in between @option{y0} and
10857 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
10858 @option{y0} and @option{y1} to the same value will disable the feature.
10859 @option{y0} and @option{y1} defaults to @code{0}.
10862 Set the scene change detection threshold as a percentage of maximum change on
10863 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
10864 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
10865 @option{scthresh} is @code{[0.0, 100.0]}.
10867 Default value is @code{12.0}.
10870 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
10871 account the combed scores of matches when deciding what match to use as the
10872 final match. Available values are:
10876 No final matching based on combed scores.
10878 Combed scores are only used when a scene change is detected.
10880 Use combed scores all the time.
10883 Default is @var{sc}.
10886 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
10887 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
10888 Available values are:
10892 No forced calculation.
10894 Force p/c/n calculations.
10896 Force p/c/n/u/b calculations.
10899 Default value is @var{none}.
10902 This is the area combing threshold used for combed frame detection. This
10903 essentially controls how "strong" or "visible" combing must be to be detected.
10904 Larger values mean combing must be more visible and smaller values mean combing
10905 can be less visible or strong and still be detected. Valid settings are from
10906 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
10907 be detected as combed). This is basically a pixel difference value. A good
10908 range is @code{[8, 12]}.
10910 Default value is @code{9}.
10913 Sets whether or not chroma is considered in the combed frame decision. Only
10914 disable this if your source has chroma problems (rainbowing, etc.) that are
10915 causing problems for the combed frame detection with chroma enabled. Actually,
10916 using @option{chroma}=@var{0} is usually more reliable, except for the case
10917 where there is chroma only combing in the source.
10919 Default value is @code{0}.
10923 Respectively set the x-axis and y-axis size of the window used during combed
10924 frame detection. This has to do with the size of the area in which
10925 @option{combpel} pixels are required to be detected as combed for a frame to be
10926 declared combed. See the @option{combpel} parameter description for more info.
10927 Possible values are any number that is a power of 2 starting at 4 and going up
10930 Default value is @code{16}.
10933 The number of combed pixels inside any of the @option{blocky} by
10934 @option{blockx} size blocks on the frame for the frame to be detected as
10935 combed. While @option{cthresh} controls how "visible" the combing must be, this
10936 setting controls "how much" combing there must be in any localized area (a
10937 window defined by the @option{blockx} and @option{blocky} settings) on the
10938 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
10939 which point no frames will ever be detected as combed). This setting is known
10940 as @option{MI} in TFM/VFM vocabulary.
10942 Default value is @code{80}.
10945 @anchor{p/c/n/u/b meaning}
10946 @subsection p/c/n/u/b meaning
10948 @subsubsection p/c/n
10950 We assume the following telecined stream:
10953 Top fields: 1 2 2 3 4
10954 Bottom fields: 1 2 3 4 4
10957 The numbers correspond to the progressive frame the fields relate to. Here, the
10958 first two frames are progressive, the 3rd and 4th are combed, and so on.
10960 When @code{fieldmatch} is configured to run a matching from bottom
10961 (@option{field}=@var{bottom}) this is how this input stream get transformed:
10966 B 1 2 3 4 4 <-- matching reference
10975 As a result of the field matching, we can see that some frames get duplicated.
10976 To perform a complete inverse telecine, you need to rely on a decimation filter
10977 after this operation. See for instance the @ref{decimate} filter.
10979 The same operation now matching from top fields (@option{field}=@var{top})
10984 T 1 2 2 3 4 <-- matching reference
10994 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
10995 basically, they refer to the frame and field of the opposite parity:
10998 @item @var{p} matches the field of the opposite parity in the previous frame
10999 @item @var{c} matches the field of the opposite parity in the current frame
11000 @item @var{n} matches the field of the opposite parity in the next frame
11005 The @var{u} and @var{b} matching are a bit special in the sense that they match
11006 from the opposite parity flag. In the following examples, we assume that we are
11007 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11008 'x' is placed above and below each matched fields.
11010 With bottom matching (@option{field}=@var{bottom}):
11015 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11016 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11024 With top matching (@option{field}=@var{top}):
11029 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11030 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11038 @subsection Examples
11040 Simple IVTC of a top field first telecined stream:
11042 fieldmatch=order=tff:combmatch=none, decimate
11045 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11047 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11050 @section fieldorder
11052 Transform the field order of the input video.
11054 It accepts the following parameters:
11059 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11060 for bottom field first.
11063 The default value is @samp{tff}.
11065 The transformation is done by shifting the picture content up or down
11066 by one line, and filling the remaining line with appropriate picture content.
11067 This method is consistent with most broadcast field order converters.
11069 If the input video is not flagged as being interlaced, or it is already
11070 flagged as being of the required output field order, then this filter does
11071 not alter the incoming video.
11073 It is very useful when converting to or from PAL DV material,
11074 which is bottom field first.
11078 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11081 @section fifo, afifo
11083 Buffer input images and send them when they are requested.
11085 It is mainly useful when auto-inserted by the libavfilter
11088 It does not take parameters.
11090 @section fillborders
11092 Fill borders of the input video, without changing video stream dimensions.
11093 Sometimes video can have garbage at the four edges and you may not want to
11094 crop video input to keep size multiple of some number.
11096 This filter accepts the following options:
11100 Number of pixels to fill from left border.
11103 Number of pixels to fill from right border.
11106 Number of pixels to fill from top border.
11109 Number of pixels to fill from bottom border.
11114 It accepts the following values:
11117 fill pixels using outermost pixels
11120 fill pixels using mirroring
11123 fill pixels with constant value
11126 Default is @var{smear}.
11129 Set color for pixels in fixed mode. Default is @var{black}.
11132 @subsection Commands
11133 This filter supports same @ref{commands} as options.
11134 The command accepts the same syntax of the corresponding option.
11136 If the specified expression is not valid, it is kept at its current
11141 Find a rectangular object
11143 It accepts the following options:
11147 Filepath of the object image, needs to be in gray8.
11150 Detection threshold, default is 0.5.
11153 Number of mipmaps, default is 3.
11155 @item xmin, ymin, xmax, ymax
11156 Specifies the rectangle in which to search.
11159 @subsection Examples
11163 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
11165 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
11171 Flood area with values of same pixel components with another values.
11173 It accepts the following options:
11176 Set pixel x coordinate.
11179 Set pixel y coordinate.
11182 Set source #0 component value.
11185 Set source #1 component value.
11188 Set source #2 component value.
11191 Set source #3 component value.
11194 Set destination #0 component value.
11197 Set destination #1 component value.
11200 Set destination #2 component value.
11203 Set destination #3 component value.
11209 Convert the input video to one of the specified pixel formats.
11210 Libavfilter will try to pick one that is suitable as input to
11213 It accepts the following parameters:
11217 A '|'-separated list of pixel format names, such as
11218 "pix_fmts=yuv420p|monow|rgb24".
11222 @subsection Examples
11226 Convert the input video to the @var{yuv420p} format
11228 format=pix_fmts=yuv420p
11231 Convert the input video to any of the formats in the list
11233 format=pix_fmts=yuv420p|yuv444p|yuv410p
11240 Convert the video to specified constant frame rate by duplicating or dropping
11241 frames as necessary.
11243 It accepts the following parameters:
11247 The desired output frame rate. The default is @code{25}.
11250 Assume the first PTS should be the given value, in seconds. This allows for
11251 padding/trimming at the start of stream. By default, no assumption is made
11252 about the first frame's expected PTS, so no padding or trimming is done.
11253 For example, this could be set to 0 to pad the beginning with duplicates of
11254 the first frame if a video stream starts after the audio stream or to trim any
11255 frames with a negative PTS.
11258 Timestamp (PTS) rounding method.
11260 Possible values are:
11267 round towards -infinity
11269 round towards +infinity
11273 The default is @code{near}.
11276 Action performed when reading the last frame.
11278 Possible values are:
11281 Use same timestamp rounding method as used for other frames.
11283 Pass through last frame if input duration has not been reached yet.
11285 The default is @code{round}.
11289 Alternatively, the options can be specified as a flat string:
11290 @var{fps}[:@var{start_time}[:@var{round}]].
11292 See also the @ref{setpts} filter.
11294 @subsection Examples
11298 A typical usage in order to set the fps to 25:
11304 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
11306 fps=fps=film:round=near
11312 Pack two different video streams into a stereoscopic video, setting proper
11313 metadata on supported codecs. The two views should have the same size and
11314 framerate and processing will stop when the shorter video ends. Please note
11315 that you may conveniently adjust view properties with the @ref{scale} and
11318 It accepts the following parameters:
11322 The desired packing format. Supported values are:
11327 The views are next to each other (default).
11330 The views are on top of each other.
11333 The views are packed by line.
11336 The views are packed by column.
11339 The views are temporally interleaved.
11348 # Convert left and right views into a frame-sequential video
11349 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
11351 # Convert views into a side-by-side video with the same output resolution as the input
11352 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
11357 Change the frame rate by interpolating new video output frames from the source
11360 This filter is not designed to function correctly with interlaced media. If
11361 you wish to change the frame rate of interlaced media then you are required
11362 to deinterlace before this filter and re-interlace after this filter.
11364 A description of the accepted options follows.
11368 Specify the output frames per second. This option can also be specified
11369 as a value alone. The default is @code{50}.
11372 Specify the start of a range where the output frame will be created as a
11373 linear interpolation of two frames. The range is [@code{0}-@code{255}],
11374 the default is @code{15}.
11377 Specify the end of a range where the output frame will be created as a
11378 linear interpolation of two frames. The range is [@code{0}-@code{255}],
11379 the default is @code{240}.
11382 Specify the level at which a scene change is detected as a value between
11383 0 and 100 to indicate a new scene; a low value reflects a low
11384 probability for the current frame to introduce a new scene, while a higher
11385 value means the current frame is more likely to be one.
11386 The default is @code{8.2}.
11389 Specify flags influencing the filter process.
11391 Available value for @var{flags} is:
11394 @item scene_change_detect, scd
11395 Enable scene change detection using the value of the option @var{scene}.
11396 This flag is enabled by default.
11402 Select one frame every N-th frame.
11404 This filter accepts the following option:
11407 Select frame after every @code{step} frames.
11408 Allowed values are positive integers higher than 0. Default value is @code{1}.
11411 @section freezedetect
11413 Detect frozen video.
11415 This filter logs a message and sets frame metadata when it detects that the
11416 input video has no significant change in content during a specified duration.
11417 Video freeze detection calculates the mean average absolute difference of all
11418 the components of video frames and compares it to a noise floor.
11420 The printed times and duration are expressed in seconds. The
11421 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
11422 whose timestamp equals or exceeds the detection duration and it contains the
11423 timestamp of the first frame of the freeze. The
11424 @code{lavfi.freezedetect.freeze_duration} and
11425 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
11428 The filter accepts the following options:
11432 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
11433 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
11437 Set freeze duration until notification (default is 2 seconds).
11440 @section freezeframes
11442 Freeze video frames.
11444 This filter freezes video frames using frame from 2nd input.
11446 The filter accepts the following options:
11450 Set number of first frame from which to start freeze.
11453 Set number of last frame from which to end freeze.
11456 Set number of frame from 2nd input which will be used instead of replaced frames.
11462 Apply a frei0r effect to the input video.
11464 To enable the compilation of this filter, you need to install the frei0r
11465 header and configure FFmpeg with @code{--enable-frei0r}.
11467 It accepts the following parameters:
11472 The name of the frei0r effect to load. If the environment variable
11473 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
11474 directories specified by the colon-separated list in @env{FREI0R_PATH}.
11475 Otherwise, the standard frei0r paths are searched, in this order:
11476 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
11477 @file{/usr/lib/frei0r-1/}.
11479 @item filter_params
11480 A '|'-separated list of parameters to pass to the frei0r effect.
11484 A frei0r effect parameter can be a boolean (its value is either
11485 "y" or "n"), a double, a color (specified as
11486 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
11487 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
11488 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
11489 a position (specified as @var{X}/@var{Y}, where
11490 @var{X} and @var{Y} are floating point numbers) and/or a string.
11492 The number and types of parameters depend on the loaded effect. If an
11493 effect parameter is not specified, the default value is set.
11495 @subsection Examples
11499 Apply the distort0r effect, setting the first two double parameters:
11501 frei0r=filter_name=distort0r:filter_params=0.5|0.01
11505 Apply the colordistance effect, taking a color as the first parameter:
11507 frei0r=colordistance:0.2/0.3/0.4
11508 frei0r=colordistance:violet
11509 frei0r=colordistance:0x112233
11513 Apply the perspective effect, specifying the top left and top right image
11516 frei0r=perspective:0.2/0.2|0.8/0.2
11520 For more information, see
11521 @url{http://frei0r.dyne.org}
11525 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
11527 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
11528 processing filter, one of them is performed once per block, not per pixel.
11529 This allows for much higher speed.
11531 The filter accepts the following options:
11535 Set quality. This option defines the number of levels for averaging. It accepts
11536 an integer in the range 4-5. Default value is @code{4}.
11539 Force a constant quantization parameter. It accepts an integer in range 0-63.
11540 If not set, the filter will use the QP from the video stream (if available).
11543 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
11544 more details but also more artifacts, while higher values make the image smoother
11545 but also blurrier. Default value is @code{0} − PSNR optimal.
11547 @item use_bframe_qp
11548 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
11549 option may cause flicker since the B-Frames have often larger QP. Default is
11550 @code{0} (not enabled).
11556 Apply Gaussian blur filter.
11558 The filter accepts the following options:
11562 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
11565 Set number of steps for Gaussian approximation. Default is @code{1}.
11568 Set which planes to filter. By default all planes are filtered.
11571 Set vertical sigma, if negative it will be same as @code{sigma}.
11572 Default is @code{-1}.
11575 @subsection Commands
11576 This filter supports same commands as options.
11577 The command accepts the same syntax of the corresponding option.
11579 If the specified expression is not valid, it is kept at its current
11584 Apply generic equation to each pixel.
11586 The filter accepts the following options:
11589 @item lum_expr, lum
11590 Set the luminance expression.
11592 Set the chrominance blue expression.
11594 Set the chrominance red expression.
11595 @item alpha_expr, a
11596 Set the alpha expression.
11598 Set the red expression.
11599 @item green_expr, g
11600 Set the green expression.
11602 Set the blue expression.
11605 The colorspace is selected according to the specified options. If one
11606 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
11607 options is specified, the filter will automatically select a YCbCr
11608 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
11609 @option{blue_expr} options is specified, it will select an RGB
11612 If one of the chrominance expression is not defined, it falls back on the other
11613 one. If no alpha expression is specified it will evaluate to opaque value.
11614 If none of chrominance expressions are specified, they will evaluate
11615 to the luminance expression.
11617 The expressions can use the following variables and functions:
11621 The sequential number of the filtered frame, starting from @code{0}.
11625 The coordinates of the current sample.
11629 The width and height of the image.
11633 Width and height scale depending on the currently filtered plane. It is the
11634 ratio between the corresponding luma plane number of pixels and the current
11635 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
11636 @code{0.5,0.5} for chroma planes.
11639 Time of the current frame, expressed in seconds.
11642 Return the value of the pixel at location (@var{x},@var{y}) of the current
11646 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
11650 Return the value of the pixel at location (@var{x},@var{y}) of the
11651 blue-difference chroma plane. Return 0 if there is no such plane.
11654 Return the value of the pixel at location (@var{x},@var{y}) of the
11655 red-difference chroma plane. Return 0 if there is no such plane.
11660 Return the value of the pixel at location (@var{x},@var{y}) of the
11661 red/green/blue component. Return 0 if there is no such component.
11664 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
11665 plane. Return 0 if there is no such plane.
11667 @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)
11668 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
11669 sums of samples within a rectangle. See the functions without the sum postfix.
11671 @item interpolation
11672 Set one of interpolation methods:
11677 Default is bilinear.
11680 For functions, if @var{x} and @var{y} are outside the area, the value will be
11681 automatically clipped to the closer edge.
11683 Please note that this filter can use multiple threads in which case each slice
11684 will have its own expression state. If you want to use only a single expression
11685 state because your expressions depend on previous state then you should limit
11686 the number of filter threads to 1.
11688 @subsection Examples
11692 Flip the image horizontally:
11698 Generate a bidimensional sine wave, with angle @code{PI/3} and a
11699 wavelength of 100 pixels:
11701 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
11705 Generate a fancy enigmatic moving light:
11707 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
11711 Generate a quick emboss effect:
11713 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
11717 Modify RGB components depending on pixel position:
11719 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
11723 Create a radial gradient that is the same size as the input (also see
11724 the @ref{vignette} filter):
11726 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
11732 Fix the banding artifacts that are sometimes introduced into nearly flat
11733 regions by truncation to 8-bit color depth.
11734 Interpolate the gradients that should go where the bands are, and
11737 It is designed for playback only. Do not use it prior to
11738 lossy compression, because compression tends to lose the dither and
11739 bring back the bands.
11741 It accepts the following parameters:
11746 The maximum amount by which the filter will change any one pixel. This is also
11747 the threshold for detecting nearly flat regions. Acceptable values range from
11748 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
11752 The neighborhood to fit the gradient to. A larger radius makes for smoother
11753 gradients, but also prevents the filter from modifying the pixels near detailed
11754 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
11755 values will be clipped to the valid range.
11759 Alternatively, the options can be specified as a flat string:
11760 @var{strength}[:@var{radius}]
11762 @subsection Examples
11766 Apply the filter with a @code{3.5} strength and radius of @code{8}:
11772 Specify radius, omitting the strength (which will fall-back to the default
11780 @anchor{graphmonitor}
11781 @section graphmonitor
11782 Show various filtergraph stats.
11784 With this filter one can debug complete filtergraph.
11785 Especially issues with links filling with queued frames.
11787 The filter accepts the following options:
11791 Set video output size. Default is @var{hd720}.
11794 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
11797 Set output mode, can be @var{fulll} or @var{compact}.
11798 In @var{compact} mode only filters with some queued frames have displayed stats.
11801 Set flags which enable which stats are shown in video.
11803 Available values for flags are:
11806 Display number of queued frames in each link.
11808 @item frame_count_in
11809 Display number of frames taken from filter.
11811 @item frame_count_out
11812 Display number of frames given out from filter.
11815 Display current filtered frame pts.
11818 Display current filtered frame time.
11821 Display time base for filter link.
11824 Display used format for filter link.
11827 Display video size or number of audio channels in case of audio used by filter link.
11830 Display video frame rate or sample rate in case of audio used by filter link.
11834 Set upper limit for video rate of output stream, Default value is @var{25}.
11835 This guarantee that output video frame rate will not be higher than this value.
11839 A color constancy variation filter which estimates scene illumination via grey edge algorithm
11840 and corrects the scene colors accordingly.
11842 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
11844 The filter accepts the following options:
11848 The order of differentiation to be applied on the scene. Must be chosen in the range
11849 [0,2] and default value is 1.
11852 The Minkowski parameter to be used for calculating the Minkowski distance. Must
11853 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
11854 max value instead of calculating Minkowski distance.
11857 The standard deviation of Gaussian blur to be applied on the scene. Must be
11858 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
11859 can't be equal to 0 if @var{difford} is greater than 0.
11862 @subsection Examples
11868 greyedge=difford=1:minknorm=5:sigma=2
11874 greyedge=difford=1:minknorm=0:sigma=2
11882 Apply a Hald CLUT to a video stream.
11884 First input is the video stream to process, and second one is the Hald CLUT.
11885 The Hald CLUT input can be a simple picture or a complete video stream.
11887 The filter accepts the following options:
11891 Force termination when the shortest input terminates. Default is @code{0}.
11893 Continue applying the last CLUT after the end of the stream. A value of
11894 @code{0} disable the filter after the last frame of the CLUT is reached.
11895 Default is @code{1}.
11898 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
11899 filters share the same internals).
11901 This filter also supports the @ref{framesync} options.
11903 More information about the Hald CLUT can be found on Eskil Steenberg's website
11904 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
11906 @subsection Workflow examples
11908 @subsubsection Hald CLUT video stream
11910 Generate an identity Hald CLUT stream altered with various effects:
11912 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
11915 Note: make sure you use a lossless codec.
11917 Then use it with @code{haldclut} to apply it on some random stream:
11919 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
11922 The Hald CLUT will be applied to the 10 first seconds (duration of
11923 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
11924 to the remaining frames of the @code{mandelbrot} stream.
11926 @subsubsection Hald CLUT with preview
11928 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
11929 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
11930 biggest possible square starting at the top left of the picture. The remaining
11931 padding pixels (bottom or right) will be ignored. This area can be used to add
11932 a preview of the Hald CLUT.
11934 Typically, the following generated Hald CLUT will be supported by the
11935 @code{haldclut} filter:
11938 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
11939 pad=iw+320 [padded_clut];
11940 smptebars=s=320x256, split [a][b];
11941 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
11942 [main][b] overlay=W-320" -frames:v 1 clut.png
11945 It contains the original and a preview of the effect of the CLUT: SMPTE color
11946 bars are displayed on the right-top, and below the same color bars processed by
11949 Then, the effect of this Hald CLUT can be visualized with:
11951 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
11956 Flip the input video horizontally.
11958 For example, to horizontally flip the input video with @command{ffmpeg}:
11960 ffmpeg -i in.avi -vf "hflip" out.avi
11964 This filter applies a global color histogram equalization on a
11967 It can be used to correct video that has a compressed range of pixel
11968 intensities. The filter redistributes the pixel intensities to
11969 equalize their distribution across the intensity range. It may be
11970 viewed as an "automatically adjusting contrast filter". This filter is
11971 useful only for correcting degraded or poorly captured source
11974 The filter accepts the following options:
11978 Determine the amount of equalization to be applied. As the strength
11979 is reduced, the distribution of pixel intensities more-and-more
11980 approaches that of the input frame. The value must be a float number
11981 in the range [0,1] and defaults to 0.200.
11984 Set the maximum intensity that can generated and scale the output
11985 values appropriately. The strength should be set as desired and then
11986 the intensity can be limited if needed to avoid washing-out. The value
11987 must be a float number in the range [0,1] and defaults to 0.210.
11990 Set the antibanding level. If enabled the filter will randomly vary
11991 the luminance of output pixels by a small amount to avoid banding of
11992 the histogram. Possible values are @code{none}, @code{weak} or
11993 @code{strong}. It defaults to @code{none}.
11999 Compute and draw a color distribution histogram for the input video.
12001 The computed histogram is a representation of the color component
12002 distribution in an image.
12004 Standard histogram displays the color components distribution in an image.
12005 Displays color graph for each color component. Shows distribution of
12006 the Y, U, V, A or R, G, B components, depending on input format, in the
12007 current frame. Below each graph a color component scale meter is shown.
12009 The filter accepts the following options:
12013 Set height of level. Default value is @code{200}.
12014 Allowed range is [50, 2048].
12017 Set height of color scale. Default value is @code{12}.
12018 Allowed range is [0, 40].
12022 It accepts the following values:
12025 Per color component graphs are placed below each other.
12028 Per color component graphs are placed side by side.
12031 Presents information identical to that in the @code{parade}, except
12032 that the graphs representing color components are superimposed directly
12035 Default is @code{stack}.
12038 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12039 Default is @code{linear}.
12042 Set what color components to display.
12043 Default is @code{7}.
12046 Set foreground opacity. Default is @code{0.7}.
12049 Set background opacity. Default is @code{0.5}.
12052 @subsection Examples
12057 Calculate and draw histogram:
12059 ffplay -i input -vf histogram
12067 This is a high precision/quality 3d denoise filter. It aims to reduce
12068 image noise, producing smooth images and making still images really
12069 still. It should enhance compressibility.
12071 It accepts the following optional parameters:
12075 A non-negative floating point number which specifies spatial luma strength.
12076 It defaults to 4.0.
12078 @item chroma_spatial
12079 A non-negative floating point number which specifies spatial chroma strength.
12080 It defaults to 3.0*@var{luma_spatial}/4.0.
12083 A floating point number which specifies luma temporal strength. It defaults to
12084 6.0*@var{luma_spatial}/4.0.
12087 A floating point number which specifies chroma temporal strength. It defaults to
12088 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
12091 @subsection Commands
12092 This filter supports same @ref{commands} as options.
12093 The command accepts the same syntax of the corresponding option.
12095 If the specified expression is not valid, it is kept at its current
12098 @anchor{hwdownload}
12099 @section hwdownload
12101 Download hardware frames to system memory.
12103 The input must be in hardware frames, and the output a non-hardware format.
12104 Not all formats will be supported on the output - it may be necessary to insert
12105 an additional @option{format} filter immediately following in the graph to get
12106 the output in a supported format.
12110 Map hardware frames to system memory or to another device.
12112 This filter has several different modes of operation; which one is used depends
12113 on the input and output formats:
12116 Hardware frame input, normal frame output
12118 Map the input frames to system memory and pass them to the output. If the
12119 original hardware frame is later required (for example, after overlaying
12120 something else on part of it), the @option{hwmap} filter can be used again
12121 in the next mode to retrieve it.
12123 Normal frame input, hardware frame output
12125 If the input is actually a software-mapped hardware frame, then unmap it -
12126 that is, return the original hardware frame.
12128 Otherwise, a device must be provided. Create new hardware surfaces on that
12129 device for the output, then map them back to the software format at the input
12130 and give those frames to the preceding filter. This will then act like the
12131 @option{hwupload} filter, but may be able to avoid an additional copy when
12132 the input is already in a compatible format.
12134 Hardware frame input and output
12136 A device must be supplied for the output, either directly or with the
12137 @option{derive_device} option. The input and output devices must be of
12138 different types and compatible - the exact meaning of this is
12139 system-dependent, but typically it means that they must refer to the same
12140 underlying hardware context (for example, refer to the same graphics card).
12142 If the input frames were originally created on the output device, then unmap
12143 to retrieve the original frames.
12145 Otherwise, map the frames to the output device - create new hardware frames
12146 on the output corresponding to the frames on the input.
12149 The following additional parameters are accepted:
12153 Set the frame mapping mode. Some combination of:
12156 The mapped frame should be readable.
12158 The mapped frame should be writeable.
12160 The mapping will always overwrite the entire frame.
12162 This may improve performance in some cases, as the original contents of the
12163 frame need not be loaded.
12165 The mapping must not involve any copying.
12167 Indirect mappings to copies of frames are created in some cases where either
12168 direct mapping is not possible or it would have unexpected properties.
12169 Setting this flag ensures that the mapping is direct and will fail if that is
12172 Defaults to @var{read+write} if not specified.
12174 @item derive_device @var{type}
12175 Rather than using the device supplied at initialisation, instead derive a new
12176 device of type @var{type} from the device the input frames exist on.
12179 In a hardware to hardware mapping, map in reverse - create frames in the sink
12180 and map them back to the source. This may be necessary in some cases where
12181 a mapping in one direction is required but only the opposite direction is
12182 supported by the devices being used.
12184 This option is dangerous - it may break the preceding filter in undefined
12185 ways if there are any additional constraints on that filter's output.
12186 Do not use it without fully understanding the implications of its use.
12192 Upload system memory frames to hardware surfaces.
12194 The device to upload to must be supplied when the filter is initialised. If
12195 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
12196 option or with the @option{derive_device} option. The input and output devices
12197 must be of different types and compatible - the exact meaning of this is
12198 system-dependent, but typically it means that they must refer to the same
12199 underlying hardware context (for example, refer to the same graphics card).
12201 The following additional parameters are accepted:
12204 @item derive_device @var{type}
12205 Rather than using the device supplied at initialisation, instead derive a new
12206 device of type @var{type} from the device the input frames exist on.
12209 @anchor{hwupload_cuda}
12210 @section hwupload_cuda
12212 Upload system memory frames to a CUDA device.
12214 It accepts the following optional parameters:
12218 The number of the CUDA device to use
12223 Apply a high-quality magnification filter designed for pixel art. This filter
12224 was originally created by Maxim Stepin.
12226 It accepts the following option:
12230 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
12231 @code{hq3x} and @code{4} for @code{hq4x}.
12232 Default is @code{3}.
12236 Stack input videos horizontally.
12238 All streams must be of same pixel format and of same height.
12240 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
12241 to create same output.
12243 The filter accepts the following option:
12247 Set number of input streams. Default is 2.
12250 If set to 1, force the output to terminate when the shortest input
12251 terminates. Default value is 0.
12256 Modify the hue and/or the saturation of the input.
12258 It accepts the following parameters:
12262 Specify the hue angle as a number of degrees. It accepts an expression,
12263 and defaults to "0".
12266 Specify the saturation in the [-10,10] range. It accepts an expression and
12270 Specify the hue angle as a number of radians. It accepts an
12271 expression, and defaults to "0".
12274 Specify the brightness in the [-10,10] range. It accepts an expression and
12278 @option{h} and @option{H} are mutually exclusive, and can't be
12279 specified at the same time.
12281 The @option{b}, @option{h}, @option{H} and @option{s} option values are
12282 expressions containing the following constants:
12286 frame count of the input frame starting from 0
12289 presentation timestamp of the input frame expressed in time base units
12292 frame rate of the input video, NAN if the input frame rate is unknown
12295 timestamp expressed in seconds, NAN if the input timestamp is unknown
12298 time base of the input video
12301 @subsection Examples
12305 Set the hue to 90 degrees and the saturation to 1.0:
12311 Same command but expressing the hue in radians:
12317 Rotate hue and make the saturation swing between 0
12318 and 2 over a period of 1 second:
12320 hue="H=2*PI*t: s=sin(2*PI*t)+1"
12324 Apply a 3 seconds saturation fade-in effect starting at 0:
12326 hue="s=min(t/3\,1)"
12329 The general fade-in expression can be written as:
12331 hue="s=min(0\, max((t-START)/DURATION\, 1))"
12335 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
12337 hue="s=max(0\, min(1\, (8-t)/3))"
12340 The general fade-out expression can be written as:
12342 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
12347 @subsection Commands
12349 This filter supports the following commands:
12355 Modify the hue and/or the saturation and/or brightness of the input video.
12356 The command accepts the same syntax of the corresponding option.
12358 If the specified expression is not valid, it is kept at its current
12362 @section hysteresis
12364 Grow first stream into second stream by connecting components.
12365 This makes it possible to build more robust edge masks.
12367 This filter accepts the following options:
12371 Set which planes will be processed as bitmap, unprocessed planes will be
12372 copied from first stream.
12373 By default value 0xf, all planes will be processed.
12376 Set threshold which is used in filtering. If pixel component value is higher than
12377 this value filter algorithm for connecting components is activated.
12378 By default value is 0.
12381 The @code{hysteresis} filter also supports the @ref{framesync} options.
12385 Detect video interlacing type.
12387 This filter tries to detect if the input frames are interlaced, progressive,
12388 top or bottom field first. It will also try to detect fields that are
12389 repeated between adjacent frames (a sign of telecine).
12391 Single frame detection considers only immediately adjacent frames when classifying each frame.
12392 Multiple frame detection incorporates the classification history of previous frames.
12394 The filter will log these metadata values:
12397 @item single.current_frame
12398 Detected type of current frame using single-frame detection. One of:
12399 ``tff'' (top field first), ``bff'' (bottom field first),
12400 ``progressive'', or ``undetermined''
12403 Cumulative number of frames detected as top field first using single-frame detection.
12406 Cumulative number of frames detected as top field first using multiple-frame detection.
12409 Cumulative number of frames detected as bottom field first using single-frame detection.
12411 @item multiple.current_frame
12412 Detected type of current frame using multiple-frame detection. One of:
12413 ``tff'' (top field first), ``bff'' (bottom field first),
12414 ``progressive'', or ``undetermined''
12417 Cumulative number of frames detected as bottom field first using multiple-frame detection.
12419 @item single.progressive
12420 Cumulative number of frames detected as progressive using single-frame detection.
12422 @item multiple.progressive
12423 Cumulative number of frames detected as progressive using multiple-frame detection.
12425 @item single.undetermined
12426 Cumulative number of frames that could not be classified using single-frame detection.
12428 @item multiple.undetermined
12429 Cumulative number of frames that could not be classified using multiple-frame detection.
12431 @item repeated.current_frame
12432 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
12434 @item repeated.neither
12435 Cumulative number of frames with no repeated field.
12438 Cumulative number of frames with the top field repeated from the previous frame's top field.
12440 @item repeated.bottom
12441 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
12444 The filter accepts the following options:
12448 Set interlacing threshold.
12450 Set progressive threshold.
12452 Threshold for repeated field detection.
12454 Number of frames after which a given frame's contribution to the
12455 statistics is halved (i.e., it contributes only 0.5 to its
12456 classification). The default of 0 means that all frames seen are given
12457 full weight of 1.0 forever.
12458 @item analyze_interlaced_flag
12459 When this is not 0 then idet will use the specified number of frames to determine
12460 if the interlaced flag is accurate, it will not count undetermined frames.
12461 If the flag is found to be accurate it will be used without any further
12462 computations, if it is found to be inaccurate it will be cleared without any
12463 further computations. This allows inserting the idet filter as a low computational
12464 method to clean up the interlaced flag
12469 Deinterleave or interleave fields.
12471 This filter allows one to process interlaced images fields without
12472 deinterlacing them. Deinterleaving splits the input frame into 2
12473 fields (so called half pictures). Odd lines are moved to the top
12474 half of the output image, even lines to the bottom half.
12475 You can process (filter) them independently and then re-interleave them.
12477 The filter accepts the following options:
12481 @item chroma_mode, c
12482 @item alpha_mode, a
12483 Available values for @var{luma_mode}, @var{chroma_mode} and
12484 @var{alpha_mode} are:
12490 @item deinterleave, d
12491 Deinterleave fields, placing one above the other.
12493 @item interleave, i
12494 Interleave fields. Reverse the effect of deinterleaving.
12496 Default value is @code{none}.
12498 @item luma_swap, ls
12499 @item chroma_swap, cs
12500 @item alpha_swap, as
12501 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
12504 @subsection Commands
12506 This filter supports the all above options as @ref{commands}.
12510 Apply inflate effect to the video.
12512 This filter replaces the pixel by the local(3x3) average by taking into account
12513 only values higher than the pixel.
12515 It accepts the following options:
12522 Limit the maximum change for each plane, default is 65535.
12523 If 0, plane will remain unchanged.
12526 @subsection Commands
12528 This filter supports the all above options as @ref{commands}.
12532 Simple interlacing filter from progressive contents. This interleaves upper (or
12533 lower) lines from odd frames with lower (or upper) lines from even frames,
12534 halving the frame rate and preserving image height.
12537 Original Original New Frame
12538 Frame 'j' Frame 'j+1' (tff)
12539 ========== =========== ==================
12540 Line 0 --------------------> Frame 'j' Line 0
12541 Line 1 Line 1 ----> Frame 'j+1' Line 1
12542 Line 2 ---------------------> Frame 'j' Line 2
12543 Line 3 Line 3 ----> Frame 'j+1' Line 3
12545 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
12548 It accepts the following optional parameters:
12552 This determines whether the interlaced frame is taken from the even
12553 (tff - default) or odd (bff) lines of the progressive frame.
12556 Vertical lowpass filter to avoid twitter interlacing and
12557 reduce moire patterns.
12561 Disable vertical lowpass filter
12564 Enable linear filter (default)
12567 Enable complex filter. This will slightly less reduce twitter and moire
12568 but better retain detail and subjective sharpness impression.
12575 Deinterlace input video by applying Donald Graft's adaptive kernel
12576 deinterling. Work on interlaced parts of a video to produce
12577 progressive frames.
12579 The description of the accepted parameters follows.
12583 Set the threshold which affects the filter's tolerance when
12584 determining if a pixel line must be processed. It must be an integer
12585 in the range [0,255] and defaults to 10. A value of 0 will result in
12586 applying the process on every pixels.
12589 Paint pixels exceeding the threshold value to white if set to 1.
12593 Set the fields order. Swap fields if set to 1, leave fields alone if
12597 Enable additional sharpening if set to 1. Default is 0.
12600 Enable twoway sharpening if set to 1. Default is 0.
12603 @subsection Examples
12607 Apply default values:
12609 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
12613 Enable additional sharpening:
12619 Paint processed pixels in white:
12627 Slowly update darker pixels.
12629 This filter makes short flashes of light appear longer.
12630 This filter accepts the following options:
12634 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
12637 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
12640 @section lenscorrection
12642 Correct radial lens distortion
12644 This filter can be used to correct for radial distortion as can result from the use
12645 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
12646 one can use tools available for example as part of opencv or simply trial-and-error.
12647 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
12648 and extract the k1 and k2 coefficients from the resulting matrix.
12650 Note that effectively the same filter is available in the open-source tools Krita and
12651 Digikam from the KDE project.
12653 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
12654 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
12655 brightness distribution, so you may want to use both filters together in certain
12656 cases, though you will have to take care of ordering, i.e. whether vignetting should
12657 be applied before or after lens correction.
12659 @subsection Options
12661 The filter accepts the following options:
12665 Relative x-coordinate of the focal point of the image, and thereby the center of the
12666 distortion. This value has a range [0,1] and is expressed as fractions of the image
12667 width. Default is 0.5.
12669 Relative y-coordinate of the focal point of the image, and thereby the center of the
12670 distortion. This value has a range [0,1] and is expressed as fractions of the image
12671 height. Default is 0.5.
12673 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
12674 no correction. Default is 0.
12676 Coefficient of the double quadratic correction term. This value has a range [-1,1].
12677 0 means no correction. Default is 0.
12680 The formula that generates the correction is:
12682 @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)
12684 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
12685 distances from the focal point in the source and target images, respectively.
12689 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
12691 The @code{lensfun} filter requires the camera make, camera model, and lens model
12692 to apply the lens correction. The filter will load the lensfun database and
12693 query it to find the corresponding camera and lens entries in the database. As
12694 long as these entries can be found with the given options, the filter can
12695 perform corrections on frames. Note that incomplete strings will result in the
12696 filter choosing the best match with the given options, and the filter will
12697 output the chosen camera and lens models (logged with level "info"). You must
12698 provide the make, camera model, and lens model as they are required.
12700 The filter accepts the following options:
12704 The make of the camera (for example, "Canon"). This option is required.
12707 The model of the camera (for example, "Canon EOS 100D"). This option is
12711 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
12712 option is required.
12715 The type of correction to apply. The following values are valid options:
12719 Enables fixing lens vignetting.
12722 Enables fixing lens geometry. This is the default.
12725 Enables fixing chromatic aberrations.
12728 Enables fixing lens vignetting and lens geometry.
12731 Enables fixing lens vignetting and chromatic aberrations.
12734 Enables fixing both lens geometry and chromatic aberrations.
12737 Enables all possible corrections.
12741 The focal length of the image/video (zoom; expected constant for video). For
12742 example, a 18--55mm lens has focal length range of [18--55], so a value in that
12743 range should be chosen when using that lens. Default 18.
12746 The aperture of the image/video (expected constant for video). Note that
12747 aperture is only used for vignetting correction. Default 3.5.
12749 @item focus_distance
12750 The focus distance of the image/video (expected constant for video). Note that
12751 focus distance is only used for vignetting and only slightly affects the
12752 vignetting correction process. If unknown, leave it at the default value (which
12756 The scale factor which is applied after transformation. After correction the
12757 video is no longer necessarily rectangular. This parameter controls how much of
12758 the resulting image is visible. The value 0 means that a value will be chosen
12759 automatically such that there is little or no unmapped area in the output
12760 image. 1.0 means that no additional scaling is done. Lower values may result
12761 in more of the corrected image being visible, while higher values may avoid
12762 unmapped areas in the output.
12764 @item target_geometry
12765 The target geometry of the output image/video. The following values are valid
12769 @item rectilinear (default)
12772 @item equirectangular
12773 @item fisheye_orthographic
12774 @item fisheye_stereographic
12775 @item fisheye_equisolid
12776 @item fisheye_thoby
12779 Apply the reverse of image correction (instead of correcting distortion, apply
12782 @item interpolation
12783 The type of interpolation used when correcting distortion. The following values
12788 @item linear (default)
12793 @subsection Examples
12797 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
12798 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
12802 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
12806 Apply the same as before, but only for the first 5 seconds of video.
12809 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
12816 Obtain the VMAF (Video Multi-Method Assessment Fusion)
12817 score between two input videos.
12819 The obtained VMAF score is printed through the logging system.
12821 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
12822 After installing the library it can be enabled using:
12823 @code{./configure --enable-libvmaf --enable-version3}.
12824 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
12826 The filter has following options:
12830 Set the model path which is to be used for SVM.
12831 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
12834 Set the file path to be used to store logs.
12837 Set the format of the log file (xml or json).
12839 @item enable_transform
12840 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
12841 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
12842 Default value: @code{false}
12845 Invokes the phone model which will generate VMAF scores higher than in the
12846 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
12847 Default value: @code{false}
12850 Enables computing psnr along with vmaf.
12851 Default value: @code{false}
12854 Enables computing ssim along with vmaf.
12855 Default value: @code{false}
12858 Enables computing ms_ssim along with vmaf.
12859 Default value: @code{false}
12862 Set the pool method to be used for computing vmaf.
12863 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
12866 Set number of threads to be used when computing vmaf.
12867 Default value: @code{0}, which makes use of all available logical processors.
12870 Set interval for frame subsampling used when computing vmaf.
12871 Default value: @code{1}
12873 @item enable_conf_interval
12874 Enables confidence interval.
12875 Default value: @code{false}
12878 This filter also supports the @ref{framesync} options.
12880 @subsection Examples
12883 On the below examples the input file @file{main.mpg} being processed is
12884 compared with the reference file @file{ref.mpg}.
12887 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
12891 Example with options:
12893 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
12897 Example with options and different containers:
12899 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 -
12905 Limits the pixel components values to the specified range [min, max].
12907 The filter accepts the following options:
12911 Lower bound. Defaults to the lowest allowed value for the input.
12914 Upper bound. Defaults to the highest allowed value for the input.
12917 Specify which planes will be processed. Defaults to all available.
12924 The filter accepts the following options:
12928 Set the number of loops. Setting this value to -1 will result in infinite loops.
12932 Set maximal size in number of frames. Default is 0.
12935 Set first frame of loop. Default is 0.
12938 @subsection Examples
12942 Loop single first frame infinitely:
12944 loop=loop=-1:size=1:start=0
12948 Loop single first frame 10 times:
12950 loop=loop=10:size=1:start=0
12954 Loop 10 first frames 5 times:
12956 loop=loop=5:size=10:start=0
12962 Apply a 1D LUT to an input video.
12964 The filter accepts the following options:
12968 Set the 1D LUT file name.
12970 Currently supported formats:
12979 Select interpolation mode.
12981 Available values are:
12985 Use values from the nearest defined point.
12987 Interpolate values using the linear interpolation.
12989 Interpolate values using the cosine interpolation.
12991 Interpolate values using the cubic interpolation.
12993 Interpolate values using the spline interpolation.
13000 Apply a 3D LUT to an input video.
13002 The filter accepts the following options:
13006 Set the 3D LUT file name.
13008 Currently supported formats:
13022 Select interpolation mode.
13024 Available values are:
13028 Use values from the nearest defined point.
13030 Interpolate values using the 8 points defining a cube.
13032 Interpolate values using a tetrahedron.
13038 Turn certain luma values into transparency.
13040 The filter accepts the following options:
13044 Set the luma which will be used as base for transparency.
13045 Default value is @code{0}.
13048 Set the range of luma values to be keyed out.
13049 Default value is @code{0.01}.
13052 Set the range of softness. Default value is @code{0}.
13053 Use this to control gradual transition from zero to full transparency.
13056 @subsection Commands
13057 This filter supports same @ref{commands} as options.
13058 The command accepts the same syntax of the corresponding option.
13060 If the specified expression is not valid, it is kept at its current
13063 @section lut, lutrgb, lutyuv
13065 Compute a look-up table for binding each pixel component input value
13066 to an output value, and apply it to the input video.
13068 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
13069 to an RGB input video.
13071 These filters accept the following parameters:
13074 set first pixel component expression
13076 set second pixel component expression
13078 set third pixel component expression
13080 set fourth pixel component expression, corresponds to the alpha component
13083 set red component expression
13085 set green component expression
13087 set blue component expression
13089 alpha component expression
13092 set Y/luminance component expression
13094 set U/Cb component expression
13096 set V/Cr component expression
13099 Each of them specifies the expression to use for computing the lookup table for
13100 the corresponding pixel component values.
13102 The exact component associated to each of the @var{c*} options depends on the
13105 The @var{lut} filter requires either YUV or RGB pixel formats in input,
13106 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
13108 The expressions can contain the following constants and functions:
13113 The input width and height.
13116 The input value for the pixel component.
13119 The input value, clipped to the @var{minval}-@var{maxval} range.
13122 The maximum value for the pixel component.
13125 The minimum value for the pixel component.
13128 The negated value for the pixel component value, clipped to the
13129 @var{minval}-@var{maxval} range; it corresponds to the expression
13130 "maxval-clipval+minval".
13133 The computed value in @var{val}, clipped to the
13134 @var{minval}-@var{maxval} range.
13136 @item gammaval(gamma)
13137 The computed gamma correction value of the pixel component value,
13138 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
13140 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
13144 All expressions default to "val".
13146 @subsection Examples
13150 Negate input video:
13152 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
13153 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
13156 The above is the same as:
13158 lutrgb="r=negval:g=negval:b=negval"
13159 lutyuv="y=negval:u=negval:v=negval"
13169 Remove chroma components, turning the video into a graytone image:
13171 lutyuv="u=128:v=128"
13175 Apply a luma burning effect:
13181 Remove green and blue components:
13187 Set a constant alpha channel value on input:
13189 format=rgba,lutrgb=a="maxval-minval/2"
13193 Correct luminance gamma by a factor of 0.5:
13195 lutyuv=y=gammaval(0.5)
13199 Discard least significant bits of luma:
13201 lutyuv=y='bitand(val, 128+64+32)'
13205 Technicolor like effect:
13207 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
13211 @section lut2, tlut2
13213 The @code{lut2} filter takes two input streams and outputs one
13216 The @code{tlut2} (time lut2) filter takes two consecutive frames
13217 from one single stream.
13219 This filter accepts the following parameters:
13222 set first pixel component expression
13224 set second pixel component expression
13226 set third pixel component expression
13228 set fourth pixel component expression, corresponds to the alpha component
13231 set output bit depth, only available for @code{lut2} filter. By default is 0,
13232 which means bit depth is automatically picked from first input format.
13235 The @code{lut2} filter also supports the @ref{framesync} options.
13237 Each of them specifies the expression to use for computing the lookup table for
13238 the corresponding pixel component values.
13240 The exact component associated to each of the @var{c*} options depends on the
13243 The expressions can contain the following constants:
13248 The input width and height.
13251 The first input value for the pixel component.
13254 The second input value for the pixel component.
13257 The first input video bit depth.
13260 The second input video bit depth.
13263 All expressions default to "x".
13265 @subsection Examples
13269 Highlight differences between two RGB video streams:
13271 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)'
13275 Highlight differences between two YUV video streams:
13277 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)'
13281 Show max difference between two video streams:
13283 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)))'
13287 @section maskedclamp
13289 Clamp the first input stream with the second input and third input stream.
13291 Returns the value of first stream to be between second input
13292 stream - @code{undershoot} and third input stream + @code{overshoot}.
13294 This filter accepts the following options:
13297 Default value is @code{0}.
13300 Default value is @code{0}.
13303 Set which planes will be processed as bitmap, unprocessed planes will be
13304 copied from first stream.
13305 By default value 0xf, all planes will be processed.
13310 Merge the second and third input stream into output stream using absolute differences
13311 between second input stream and first input stream and absolute difference between
13312 third input stream and first input stream. The picked value will be from second input
13313 stream if second absolute difference is greater than first one or from third input stream
13316 This filter accepts the following options:
13319 Set which planes will be processed as bitmap, unprocessed planes will be
13320 copied from first stream.
13321 By default value 0xf, all planes will be processed.
13324 @section maskedmerge
13326 Merge the first input stream with the second input stream using per pixel
13327 weights in the third input stream.
13329 A value of 0 in the third stream pixel component means that pixel component
13330 from first stream is returned unchanged, while maximum value (eg. 255 for
13331 8-bit videos) means that pixel component from second stream is returned
13332 unchanged. Intermediate values define the amount of merging between both
13333 input stream's pixel components.
13335 This filter accepts the following options:
13338 Set which planes will be processed as bitmap, unprocessed planes will be
13339 copied from first stream.
13340 By default value 0xf, all planes will be processed.
13345 Merge the second and third input stream into output stream using absolute differences
13346 between second input stream and first input stream and absolute difference between
13347 third input stream and first input stream. The picked value will be from second input
13348 stream if second absolute difference is less than first one or from third input stream
13351 This filter accepts the following options:
13354 Set which planes will be processed as bitmap, unprocessed planes will be
13355 copied from first stream.
13356 By default value 0xf, all planes will be processed.
13359 @section maskedthreshold
13360 Pick pixels comparing absolute difference of two video streams with fixed
13363 If absolute difference between pixel component of first and second video
13364 stream is equal or lower than user supplied threshold than pixel component
13365 from first video stream is picked, otherwise pixel component from second
13366 video stream is picked.
13368 This filter accepts the following options:
13371 Set threshold used when picking pixels from absolute difference from two input
13375 Set which planes will be processed as bitmap, unprocessed planes will be
13376 copied from second stream.
13377 By default value 0xf, all planes will be processed.
13381 Create mask from input video.
13383 For example it is useful to create motion masks after @code{tblend} filter.
13385 This filter accepts the following options:
13389 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
13392 Set high threshold. Any pixel component higher than this value will be set to max value
13393 allowed for current pixel format.
13396 Set planes to filter, by default all available planes are filtered.
13399 Fill all frame pixels with this value.
13402 Set max average pixel value for frame. If sum of all pixel components is higher that this
13403 average, output frame will be completely filled with value set by @var{fill} option.
13404 Typically useful for scene changes when used in combination with @code{tblend} filter.
13409 Apply motion-compensation deinterlacing.
13411 It needs one field per frame as input and must thus be used together
13412 with yadif=1/3 or equivalent.
13414 This filter accepts the following options:
13417 Set the deinterlacing mode.
13419 It accepts one of the following values:
13424 use iterative motion estimation
13426 like @samp{slow}, but use multiple reference frames.
13428 Default value is @samp{fast}.
13431 Set the picture field parity assumed for the input video. It must be
13432 one of the following values:
13436 assume top field first
13438 assume bottom field first
13441 Default value is @samp{bff}.
13444 Set per-block quantization parameter (QP) used by the internal
13447 Higher values should result in a smoother motion vector field but less
13448 optimal individual vectors. Default value is 1.
13453 Pick median pixel from certain rectangle defined by radius.
13455 This filter accepts the following options:
13459 Set horizontal radius size. Default value is @code{1}.
13460 Allowed range is integer from 1 to 127.
13463 Set which planes to process. Default is @code{15}, which is all available planes.
13466 Set vertical radius size. Default value is @code{0}.
13467 Allowed range is integer from 0 to 127.
13468 If it is 0, value will be picked from horizontal @code{radius} option.
13471 Set median percentile. Default value is @code{0.5}.
13472 Default value of @code{0.5} will pick always median values, while @code{0} will pick
13473 minimum values, and @code{1} maximum values.
13476 @subsection Commands
13477 This filter supports same @ref{commands} as options.
13478 The command accepts the same syntax of the corresponding option.
13480 If the specified expression is not valid, it is kept at its current
13483 @section mergeplanes
13485 Merge color channel components from several video streams.
13487 The filter accepts up to 4 input streams, and merge selected input
13488 planes to the output video.
13490 This filter accepts the following options:
13493 Set input to output plane mapping. Default is @code{0}.
13495 The mappings is specified as a bitmap. It should be specified as a
13496 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
13497 mapping for the first plane of the output stream. 'A' sets the number of
13498 the input stream to use (from 0 to 3), and 'a' the plane number of the
13499 corresponding input to use (from 0 to 3). The rest of the mappings is
13500 similar, 'Bb' describes the mapping for the output stream second
13501 plane, 'Cc' describes the mapping for the output stream third plane and
13502 'Dd' describes the mapping for the output stream fourth plane.
13505 Set output pixel format. Default is @code{yuva444p}.
13508 @subsection Examples
13512 Merge three gray video streams of same width and height into single video stream:
13514 [a0][a1][a2]mergeplanes=0x001020:yuv444p
13518 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
13520 [a0][a1]mergeplanes=0x00010210:yuva444p
13524 Swap Y and A plane in yuva444p stream:
13526 format=yuva444p,mergeplanes=0x03010200:yuva444p
13530 Swap U and V plane in yuv420p stream:
13532 format=yuv420p,mergeplanes=0x000201:yuv420p
13536 Cast a rgb24 clip to yuv444p:
13538 format=rgb24,mergeplanes=0x000102:yuv444p
13544 Estimate and export motion vectors using block matching algorithms.
13545 Motion vectors are stored in frame side data to be used by other filters.
13547 This filter accepts the following options:
13550 Specify the motion estimation method. Accepts one of the following values:
13554 Exhaustive search algorithm.
13556 Three step search algorithm.
13558 Two dimensional logarithmic search algorithm.
13560 New three step search algorithm.
13562 Four step search algorithm.
13564 Diamond search algorithm.
13566 Hexagon-based search algorithm.
13568 Enhanced predictive zonal search algorithm.
13570 Uneven multi-hexagon search algorithm.
13572 Default value is @samp{esa}.
13575 Macroblock size. Default @code{16}.
13578 Search parameter. Default @code{7}.
13581 @section midequalizer
13583 Apply Midway Image Equalization effect using two video streams.
13585 Midway Image Equalization adjusts a pair of images to have the same
13586 histogram, while maintaining their dynamics as much as possible. It's
13587 useful for e.g. matching exposures from a pair of stereo cameras.
13589 This filter has two inputs and one output, which must be of same pixel format, but
13590 may be of different sizes. The output of filter is first input adjusted with
13591 midway histogram of both inputs.
13593 This filter accepts the following option:
13597 Set which planes to process. Default is @code{15}, which is all available planes.
13600 @section minterpolate
13602 Convert the video to specified frame rate using motion interpolation.
13604 This filter accepts the following options:
13607 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}.
13610 Motion interpolation mode. Following values are accepted:
13613 Duplicate previous or next frame for interpolating new ones.
13615 Blend source frames. Interpolated frame is mean of previous and next frames.
13617 Motion compensated interpolation. Following options are effective when this mode is selected:
13621 Motion compensation mode. Following values are accepted:
13624 Overlapped block motion compensation.
13626 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
13628 Default mode is @samp{obmc}.
13631 Motion estimation mode. Following values are accepted:
13634 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
13636 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
13638 Default mode is @samp{bilat}.
13641 The algorithm to be used for motion estimation. Following values are accepted:
13644 Exhaustive search algorithm.
13646 Three step search algorithm.
13648 Two dimensional logarithmic search algorithm.
13650 New three step search algorithm.
13652 Four step search algorithm.
13654 Diamond search algorithm.
13656 Hexagon-based search algorithm.
13658 Enhanced predictive zonal search algorithm.
13660 Uneven multi-hexagon search algorithm.
13662 Default algorithm is @samp{epzs}.
13665 Macroblock size. Default @code{16}.
13668 Motion estimation search parameter. Default @code{32}.
13671 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).
13676 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:
13679 Disable scene change detection.
13681 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
13683 Default method is @samp{fdiff}.
13685 @item scd_threshold
13686 Scene change detection threshold. Default is @code{10.}.
13691 Mix several video input streams into one video stream.
13693 A description of the accepted options follows.
13697 The number of inputs. If unspecified, it defaults to 2.
13700 Specify weight of each input video stream as sequence.
13701 Each weight is separated by space. If number of weights
13702 is smaller than number of @var{frames} last specified
13703 weight will be used for all remaining unset weights.
13706 Specify scale, if it is set it will be multiplied with sum
13707 of each weight multiplied with pixel values to give final destination
13708 pixel value. By default @var{scale} is auto scaled to sum of weights.
13711 Specify how end of stream is determined.
13714 The duration of the longest input. (default)
13717 The duration of the shortest input.
13720 The duration of the first input.
13724 @section mpdecimate
13726 Drop frames that do not differ greatly from the previous frame in
13727 order to reduce frame rate.
13729 The main use of this filter is for very-low-bitrate encoding
13730 (e.g. streaming over dialup modem), but it could in theory be used for
13731 fixing movies that were inverse-telecined incorrectly.
13733 A description of the accepted options follows.
13737 Set the maximum number of consecutive frames which can be dropped (if
13738 positive), or the minimum interval between dropped frames (if
13739 negative). If the value is 0, the frame is dropped disregarding the
13740 number of previous sequentially dropped frames.
13742 Default value is 0.
13747 Set the dropping threshold values.
13749 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
13750 represent actual pixel value differences, so a threshold of 64
13751 corresponds to 1 unit of difference for each pixel, or the same spread
13752 out differently over the block.
13754 A frame is a candidate for dropping if no 8x8 blocks differ by more
13755 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
13756 meaning the whole image) differ by more than a threshold of @option{lo}.
13758 Default value for @option{hi} is 64*12, default value for @option{lo} is
13759 64*5, and default value for @option{frac} is 0.33.
13765 Negate (invert) the input video.
13767 It accepts the following option:
13772 With value 1, it negates the alpha component, if present. Default value is 0.
13778 Denoise frames using Non-Local Means algorithm.
13780 Each pixel is adjusted by looking for other pixels with similar contexts. This
13781 context similarity is defined by comparing their surrounding patches of size
13782 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
13785 Note that the research area defines centers for patches, which means some
13786 patches will be made of pixels outside that research area.
13788 The filter accepts the following options.
13792 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
13795 Set patch size. Default is 7. Must be odd number in range [0, 99].
13798 Same as @option{p} but for chroma planes.
13800 The default value is @var{0} and means automatic.
13803 Set research size. Default is 15. Must be odd number in range [0, 99].
13806 Same as @option{r} but for chroma planes.
13808 The default value is @var{0} and means automatic.
13813 Deinterlace video using neural network edge directed interpolation.
13815 This filter accepts the following options:
13819 Mandatory option, without binary file filter can not work.
13820 Currently file can be found here:
13821 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
13824 Set which frames to deinterlace, by default it is @code{all}.
13825 Can be @code{all} or @code{interlaced}.
13828 Set mode of operation.
13830 Can be one of the following:
13834 Use frame flags, both fields.
13836 Use frame flags, single field.
13838 Use top field only.
13840 Use bottom field only.
13842 Use both fields, top first.
13844 Use both fields, bottom first.
13848 Set which planes to process, by default filter process all frames.
13851 Set size of local neighborhood around each pixel, used by the predictor neural
13854 Can be one of the following:
13867 Set the number of neurons in predictor neural network.
13868 Can be one of the following:
13879 Controls the number of different neural network predictions that are blended
13880 together to compute the final output value. Can be @code{fast}, default or
13884 Set which set of weights to use in the predictor.
13885 Can be one of the following:
13889 weights trained to minimize absolute error
13891 weights trained to minimize squared error
13895 Controls whether or not the prescreener neural network is used to decide
13896 which pixels should be processed by the predictor neural network and which
13897 can be handled by simple cubic interpolation.
13898 The prescreener is trained to know whether cubic interpolation will be
13899 sufficient for a pixel or whether it should be predicted by the predictor nn.
13900 The computational complexity of the prescreener nn is much less than that of
13901 the predictor nn. Since most pixels can be handled by cubic interpolation,
13902 using the prescreener generally results in much faster processing.
13903 The prescreener is pretty accurate, so the difference between using it and not
13904 using it is almost always unnoticeable.
13906 Can be one of the following:
13914 Default is @code{new}.
13917 Set various debugging flags.
13922 Force libavfilter not to use any of the specified pixel formats for the
13923 input to the next filter.
13925 It accepts the following parameters:
13929 A '|'-separated list of pixel format names, such as
13930 pix_fmts=yuv420p|monow|rgb24".
13934 @subsection Examples
13938 Force libavfilter to use a format different from @var{yuv420p} for the
13939 input to the vflip filter:
13941 noformat=pix_fmts=yuv420p,vflip
13945 Convert the input video to any of the formats not contained in the list:
13947 noformat=yuv420p|yuv444p|yuv410p
13953 Add noise on video input frame.
13955 The filter accepts the following options:
13963 Set noise seed for specific pixel component or all pixel components in case
13964 of @var{all_seed}. Default value is @code{123457}.
13966 @item all_strength, alls
13967 @item c0_strength, c0s
13968 @item c1_strength, c1s
13969 @item c2_strength, c2s
13970 @item c3_strength, c3s
13971 Set noise strength for specific pixel component or all pixel components in case
13972 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
13974 @item all_flags, allf
13975 @item c0_flags, c0f
13976 @item c1_flags, c1f
13977 @item c2_flags, c2f
13978 @item c3_flags, c3f
13979 Set pixel component flags or set flags for all components if @var{all_flags}.
13980 Available values for component flags are:
13983 averaged temporal noise (smoother)
13985 mix random noise with a (semi)regular pattern
13987 temporal noise (noise pattern changes between frames)
13989 uniform noise (gaussian otherwise)
13993 @subsection Examples
13995 Add temporal and uniform noise to input video:
13997 noise=alls=20:allf=t+u
14002 Normalize RGB video (aka histogram stretching, contrast stretching).
14003 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
14005 For each channel of each frame, the filter computes the input range and maps
14006 it linearly to the user-specified output range. The output range defaults
14007 to the full dynamic range from pure black to pure white.
14009 Temporal smoothing can be used on the input range to reduce flickering (rapid
14010 changes in brightness) caused when small dark or bright objects enter or leave
14011 the scene. This is similar to the auto-exposure (automatic gain control) on a
14012 video camera, and, like a video camera, it may cause a period of over- or
14013 under-exposure of the video.
14015 The R,G,B channels can be normalized independently, which may cause some
14016 color shifting, or linked together as a single channel, which prevents
14017 color shifting. Linked normalization preserves hue. Independent normalization
14018 does not, so it can be used to remove some color casts. Independent and linked
14019 normalization can be combined in any ratio.
14021 The normalize filter accepts the following options:
14026 Colors which define the output range. The minimum input value is mapped to
14027 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
14028 The defaults are black and white respectively. Specifying white for
14029 @var{blackpt} and black for @var{whitept} will give color-inverted,
14030 normalized video. Shades of grey can be used to reduce the dynamic range
14031 (contrast). Specifying saturated colors here can create some interesting
14035 The number of previous frames to use for temporal smoothing. The input range
14036 of each channel is smoothed using a rolling average over the current frame
14037 and the @var{smoothing} previous frames. The default is 0 (no temporal
14041 Controls the ratio of independent (color shifting) channel normalization to
14042 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
14043 independent. Defaults to 1.0 (fully independent).
14046 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
14047 expensive no-op. Defaults to 1.0 (full strength).
14051 @subsection Commands
14052 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
14053 The command accepts the same syntax of the corresponding option.
14055 If the specified expression is not valid, it is kept at its current
14058 @subsection Examples
14060 Stretch video contrast to use the full dynamic range, with no temporal
14061 smoothing; may flicker depending on the source content:
14063 normalize=blackpt=black:whitept=white:smoothing=0
14066 As above, but with 50 frames of temporal smoothing; flicker should be
14067 reduced, depending on the source content:
14069 normalize=blackpt=black:whitept=white:smoothing=50
14072 As above, but with hue-preserving linked channel normalization:
14074 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
14077 As above, but with half strength:
14079 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
14082 Map the darkest input color to red, the brightest input color to cyan:
14084 normalize=blackpt=red:whitept=cyan
14089 Pass the video source unchanged to the output.
14092 Optical Character Recognition
14094 This filter uses Tesseract for optical character recognition. To enable
14095 compilation of this filter, you need to configure FFmpeg with
14096 @code{--enable-libtesseract}.
14098 It accepts the following options:
14102 Set datapath to tesseract data. Default is to use whatever was
14103 set at installation.
14106 Set language, default is "eng".
14109 Set character whitelist.
14112 Set character blacklist.
14115 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
14116 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
14120 Apply a video transform using libopencv.
14122 To enable this filter, install the libopencv library and headers and
14123 configure FFmpeg with @code{--enable-libopencv}.
14125 It accepts the following parameters:
14130 The name of the libopencv filter to apply.
14132 @item filter_params
14133 The parameters to pass to the libopencv filter. If not specified, the default
14134 values are assumed.
14138 Refer to the official libopencv documentation for more precise
14140 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
14142 Several libopencv filters are supported; see the following subsections.
14147 Dilate an image by using a specific structuring element.
14148 It corresponds to the libopencv function @code{cvDilate}.
14150 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
14152 @var{struct_el} represents a structuring element, and has the syntax:
14153 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
14155 @var{cols} and @var{rows} represent the number of columns and rows of
14156 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
14157 point, and @var{shape} the shape for the structuring element. @var{shape}
14158 must be "rect", "cross", "ellipse", or "custom".
14160 If the value for @var{shape} is "custom", it must be followed by a
14161 string of the form "=@var{filename}". The file with name
14162 @var{filename} is assumed to represent a binary image, with each
14163 printable character corresponding to a bright pixel. When a custom
14164 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
14165 or columns and rows of the read file are assumed instead.
14167 The default value for @var{struct_el} is "3x3+0x0/rect".
14169 @var{nb_iterations} specifies the number of times the transform is
14170 applied to the image, and defaults to 1.
14174 # Use the default values
14177 # Dilate using a structuring element with a 5x5 cross, iterating two times
14178 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
14180 # Read the shape from the file diamond.shape, iterating two times.
14181 # The file diamond.shape may contain a pattern of characters like this
14187 # The specified columns and rows are ignored
14188 # but the anchor point coordinates are not
14189 ocv=dilate:0x0+2x2/custom=diamond.shape|2
14194 Erode an image by using a specific structuring element.
14195 It corresponds to the libopencv function @code{cvErode}.
14197 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
14198 with the same syntax and semantics as the @ref{dilate} filter.
14202 Smooth the input video.
14204 The filter takes the following parameters:
14205 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
14207 @var{type} is the type of smooth filter to apply, and must be one of
14208 the following values: "blur", "blur_no_scale", "median", "gaussian",
14209 or "bilateral". The default value is "gaussian".
14211 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
14212 depends on the smooth type. @var{param1} and
14213 @var{param2} accept integer positive values or 0. @var{param3} and
14214 @var{param4} accept floating point values.
14216 The default value for @var{param1} is 3. The default value for the
14217 other parameters is 0.
14219 These parameters correspond to the parameters assigned to the
14220 libopencv function @code{cvSmooth}.
14222 @section oscilloscope
14224 2D Video Oscilloscope.
14226 Useful to measure spatial impulse, step responses, chroma delays, etc.
14228 It accepts the following parameters:
14232 Set scope center x position.
14235 Set scope center y position.
14238 Set scope size, relative to frame diagonal.
14241 Set scope tilt/rotation.
14247 Set trace center x position.
14250 Set trace center y position.
14253 Set trace width, relative to width of frame.
14256 Set trace height, relative to height of frame.
14259 Set which components to trace. By default it traces first three components.
14262 Draw trace grid. By default is enabled.
14265 Draw some statistics. By default is enabled.
14268 Draw scope. By default is enabled.
14271 @subsection Commands
14272 This filter supports same @ref{commands} as options.
14273 The command accepts the same syntax of the corresponding option.
14275 If the specified expression is not valid, it is kept at its current
14278 @subsection Examples
14282 Inspect full first row of video frame.
14284 oscilloscope=x=0.5:y=0:s=1
14288 Inspect full last row of video frame.
14290 oscilloscope=x=0.5:y=1:s=1
14294 Inspect full 5th line of video frame of height 1080.
14296 oscilloscope=x=0.5:y=5/1080:s=1
14300 Inspect full last column of video frame.
14302 oscilloscope=x=1:y=0.5:s=1:t=1
14310 Overlay one video on top of another.
14312 It takes two inputs and has one output. The first input is the "main"
14313 video on which the second input is overlaid.
14315 It accepts the following parameters:
14317 A description of the accepted options follows.
14322 Set the expression for the x and y coordinates of the overlaid video
14323 on the main video. Default value is "0" for both expressions. In case
14324 the expression is invalid, it is set to a huge value (meaning that the
14325 overlay will not be displayed within the output visible area).
14328 See @ref{framesync}.
14331 Set when the expressions for @option{x}, and @option{y} are evaluated.
14333 It accepts the following values:
14336 only evaluate expressions once during the filter initialization or
14337 when a command is processed
14340 evaluate expressions for each incoming frame
14343 Default value is @samp{frame}.
14346 See @ref{framesync}.
14349 Set the format for the output video.
14351 It accepts the following values:
14354 force YUV420 output
14357 force YUV420p10 output
14360 force YUV422 output
14363 force YUV422p10 output
14366 force YUV444 output
14369 force packed RGB output
14372 force planar RGB output
14375 automatically pick format
14378 Default value is @samp{yuv420}.
14381 See @ref{framesync}.
14384 Set format of alpha of the overlaid video, it can be @var{straight} or
14385 @var{premultiplied}. Default is @var{straight}.
14388 The @option{x}, and @option{y} expressions can contain the following
14394 The main input width and height.
14398 The overlay input width and height.
14402 The computed values for @var{x} and @var{y}. They are evaluated for
14407 horizontal and vertical chroma subsample values of the output
14408 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
14412 the number of input frame, starting from 0
14415 the position in the file of the input frame, NAN if unknown
14418 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
14422 This filter also supports the @ref{framesync} options.
14424 Note that the @var{n}, @var{pos}, @var{t} variables are available only
14425 when evaluation is done @emph{per frame}, and will evaluate to NAN
14426 when @option{eval} is set to @samp{init}.
14428 Be aware that frames are taken from each input video in timestamp
14429 order, hence, if their initial timestamps differ, it is a good idea
14430 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
14431 have them begin in the same zero timestamp, as the example for
14432 the @var{movie} filter does.
14434 You can chain together more overlays but you should test the
14435 efficiency of such approach.
14437 @subsection Commands
14439 This filter supports the following commands:
14443 Modify the x and y of the overlay input.
14444 The command accepts the same syntax of the corresponding option.
14446 If the specified expression is not valid, it is kept at its current
14450 @subsection Examples
14454 Draw the overlay at 10 pixels from the bottom right corner of the main
14457 overlay=main_w-overlay_w-10:main_h-overlay_h-10
14460 Using named options the example above becomes:
14462 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
14466 Insert a transparent PNG logo in the bottom left corner of the input,
14467 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
14469 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
14473 Insert 2 different transparent PNG logos (second logo on bottom
14474 right corner) using the @command{ffmpeg} tool:
14476 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
14480 Add a transparent color layer on top of the main video; @code{WxH}
14481 must specify the size of the main input to the overlay filter:
14483 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
14487 Play an original video and a filtered version (here with the deshake
14488 filter) side by side using the @command{ffplay} tool:
14490 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
14493 The above command is the same as:
14495 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
14499 Make a sliding overlay appearing from the left to the right top part of the
14500 screen starting since time 2:
14502 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
14506 Compose output by putting two input videos side to side:
14508 ffmpeg -i left.avi -i right.avi -filter_complex "
14509 nullsrc=size=200x100 [background];
14510 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
14511 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
14512 [background][left] overlay=shortest=1 [background+left];
14513 [background+left][right] overlay=shortest=1:x=100 [left+right]
14518 Mask 10-20 seconds of a video by applying the delogo filter to a section
14520 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
14521 -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]'
14526 Chain several overlays in cascade:
14528 nullsrc=s=200x200 [bg];
14529 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
14530 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
14531 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
14532 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
14533 [in3] null, [mid2] overlay=100:100 [out0]
14538 @anchor{overlay_cuda}
14539 @section overlay_cuda
14541 Overlay one video on top of another.
14543 This is the CUDA cariant of the @ref{overlay} filter.
14544 It only accepts CUDA frames. The underlying input pixel formats have to match.
14546 It takes two inputs and has one output. The first input is the "main"
14547 video on which the second input is overlaid.
14549 It accepts the following parameters:
14554 Set the x and y coordinates of the overlaid video on the main video.
14555 Default value is "0" for both expressions.
14558 See @ref{framesync}.
14561 See @ref{framesync}.
14564 See @ref{framesync}.
14568 This filter also supports the @ref{framesync} options.
14572 Apply Overcomplete Wavelet denoiser.
14574 The filter accepts the following options:
14580 Larger depth values will denoise lower frequency components more, but
14581 slow down filtering.
14583 Must be an int in the range 8-16, default is @code{8}.
14585 @item luma_strength, ls
14588 Must be a double value in the range 0-1000, default is @code{1.0}.
14590 @item chroma_strength, cs
14591 Set chroma strength.
14593 Must be a double value in the range 0-1000, default is @code{1.0}.
14599 Add paddings to the input image, and place the original input at the
14600 provided @var{x}, @var{y} coordinates.
14602 It accepts the following parameters:
14607 Specify an expression for the size of the output image with the
14608 paddings added. If the value for @var{width} or @var{height} is 0, the
14609 corresponding input size is used for the output.
14611 The @var{width} expression can reference the value set by the
14612 @var{height} expression, and vice versa.
14614 The default value of @var{width} and @var{height} is 0.
14618 Specify the offsets to place the input image at within the padded area,
14619 with respect to the top/left border of the output image.
14621 The @var{x} expression can reference the value set by the @var{y}
14622 expression, and vice versa.
14624 The default value of @var{x} and @var{y} is 0.
14626 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
14627 so the input image is centered on the padded area.
14630 Specify the color of the padded area. For the syntax of this option,
14631 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
14632 manual,ffmpeg-utils}.
14634 The default value of @var{color} is "black".
14637 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
14639 It accepts the following values:
14643 Only evaluate expressions once during the filter initialization or when
14644 a command is processed.
14647 Evaluate expressions for each incoming frame.
14651 Default value is @samp{init}.
14654 Pad to aspect instead to a resolution.
14658 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
14659 options are expressions containing the following constants:
14664 The input video width and height.
14668 These are the same as @var{in_w} and @var{in_h}.
14672 The output width and height (the size of the padded area), as
14673 specified by the @var{width} and @var{height} expressions.
14677 These are the same as @var{out_w} and @var{out_h}.
14681 The x and y offsets as specified by the @var{x} and @var{y}
14682 expressions, or NAN if not yet specified.
14685 same as @var{iw} / @var{ih}
14688 input sample aspect ratio
14691 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
14695 The horizontal and vertical chroma subsample values. For example for the
14696 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14699 @subsection Examples
14703 Add paddings with the color "violet" to the input video. The output video
14704 size is 640x480, and the top-left corner of the input video is placed at
14707 pad=640:480:0:40:violet
14710 The example above is equivalent to the following command:
14712 pad=width=640:height=480:x=0:y=40:color=violet
14716 Pad the input to get an output with dimensions increased by 3/2,
14717 and put the input video at the center of the padded area:
14719 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
14723 Pad the input to get a squared output with size equal to the maximum
14724 value between the input width and height, and put the input video at
14725 the center of the padded area:
14727 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
14731 Pad the input to get a final w/h ratio of 16:9:
14733 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
14737 In case of anamorphic video, in order to set the output display aspect
14738 correctly, it is necessary to use @var{sar} in the expression,
14739 according to the relation:
14741 (ih * X / ih) * sar = output_dar
14742 X = output_dar / sar
14745 Thus the previous example needs to be modified to:
14747 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
14751 Double the output size and put the input video in the bottom-right
14752 corner of the output padded area:
14754 pad="2*iw:2*ih:ow-iw:oh-ih"
14758 @anchor{palettegen}
14759 @section palettegen
14761 Generate one palette for a whole video stream.
14763 It accepts the following options:
14767 Set the maximum number of colors to quantize in the palette.
14768 Note: the palette will still contain 256 colors; the unused palette entries
14771 @item reserve_transparent
14772 Create a palette of 255 colors maximum and reserve the last one for
14773 transparency. Reserving the transparency color is useful for GIF optimization.
14774 If not set, the maximum of colors in the palette will be 256. You probably want
14775 to disable this option for a standalone image.
14778 @item transparency_color
14779 Set the color that will be used as background for transparency.
14782 Set statistics mode.
14784 It accepts the following values:
14787 Compute full frame histograms.
14789 Compute histograms only for the part that differs from previous frame. This
14790 might be relevant to give more importance to the moving part of your input if
14791 the background is static.
14793 Compute new histogram for each frame.
14796 Default value is @var{full}.
14799 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
14800 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
14801 color quantization of the palette. This information is also visible at
14802 @var{info} logging level.
14804 @subsection Examples
14808 Generate a representative palette of a given video using @command{ffmpeg}:
14810 ffmpeg -i input.mkv -vf palettegen palette.png
14814 @section paletteuse
14816 Use a palette to downsample an input video stream.
14818 The filter takes two inputs: one video stream and a palette. The palette must
14819 be a 256 pixels image.
14821 It accepts the following options:
14825 Select dithering mode. Available algorithms are:
14828 Ordered 8x8 bayer dithering (deterministic)
14830 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
14831 Note: this dithering is sometimes considered "wrong" and is included as a
14833 @item floyd_steinberg
14834 Floyd and Steingberg dithering (error diffusion)
14836 Frankie Sierra dithering v2 (error diffusion)
14838 Frankie Sierra dithering v2 "Lite" (error diffusion)
14841 Default is @var{sierra2_4a}.
14844 When @var{bayer} dithering is selected, this option defines the scale of the
14845 pattern (how much the crosshatch pattern is visible). A low value means more
14846 visible pattern for less banding, and higher value means less visible pattern
14847 at the cost of more banding.
14849 The option must be an integer value in the range [0,5]. Default is @var{2}.
14852 If set, define the zone to process
14856 Only the changing rectangle will be reprocessed. This is similar to GIF
14857 cropping/offsetting compression mechanism. This option can be useful for speed
14858 if only a part of the image is changing, and has use cases such as limiting the
14859 scope of the error diffusal @option{dither} to the rectangle that bounds the
14860 moving scene (it leads to more deterministic output if the scene doesn't change
14861 much, and as a result less moving noise and better GIF compression).
14864 Default is @var{none}.
14867 Take new palette for each output frame.
14869 @item alpha_threshold
14870 Sets the alpha threshold for transparency. Alpha values above this threshold
14871 will be treated as completely opaque, and values below this threshold will be
14872 treated as completely transparent.
14874 The option must be an integer value in the range [0,255]. Default is @var{128}.
14877 @subsection Examples
14881 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
14882 using @command{ffmpeg}:
14884 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
14888 @section perspective
14890 Correct perspective of video not recorded perpendicular to the screen.
14892 A description of the accepted parameters follows.
14903 Set coordinates expression for top left, top right, bottom left and bottom right corners.
14904 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
14905 If the @code{sense} option is set to @code{source}, then the specified points will be sent
14906 to the corners of the destination. If the @code{sense} option is set to @code{destination},
14907 then the corners of the source will be sent to the specified coordinates.
14909 The expressions can use the following variables:
14914 the width and height of video frame.
14918 Output frame count.
14921 @item interpolation
14922 Set interpolation for perspective correction.
14924 It accepts the following values:
14930 Default value is @samp{linear}.
14933 Set interpretation of coordinate options.
14935 It accepts the following values:
14939 Send point in the source specified by the given coordinates to
14940 the corners of the destination.
14942 @item 1, destination
14944 Send the corners of the source to the point in the destination specified
14945 by the given coordinates.
14947 Default value is @samp{source}.
14951 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
14953 It accepts the following values:
14956 only evaluate expressions once during the filter initialization or
14957 when a command is processed
14960 evaluate expressions for each incoming frame
14963 Default value is @samp{init}.
14968 Delay interlaced video by one field time so that the field order changes.
14970 The intended use is to fix PAL movies that have been captured with the
14971 opposite field order to the film-to-video transfer.
14973 A description of the accepted parameters follows.
14979 It accepts the following values:
14982 Capture field order top-first, transfer bottom-first.
14983 Filter will delay the bottom field.
14986 Capture field order bottom-first, transfer top-first.
14987 Filter will delay the top field.
14990 Capture and transfer with the same field order. This mode only exists
14991 for the documentation of the other options to refer to, but if you
14992 actually select it, the filter will faithfully do nothing.
14995 Capture field order determined automatically by field flags, transfer
14997 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
14998 basis using field flags. If no field information is available,
14999 then this works just like @samp{u}.
15002 Capture unknown or varying, transfer opposite.
15003 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
15004 analyzing the images and selecting the alternative that produces best
15005 match between the fields.
15008 Capture top-first, transfer unknown or varying.
15009 Filter selects among @samp{t} and @samp{p} using image analysis.
15012 Capture bottom-first, transfer unknown or varying.
15013 Filter selects among @samp{b} and @samp{p} using image analysis.
15016 Capture determined by field flags, transfer unknown or varying.
15017 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
15018 image analysis. If no field information is available, then this works just
15019 like @samp{U}. This is the default mode.
15022 Both capture and transfer unknown or varying.
15023 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
15027 @section photosensitivity
15028 Reduce various flashes in video, so to help users with epilepsy.
15030 It accepts the following options:
15033 Set how many frames to use when filtering. Default is 30.
15036 Set detection threshold factor. Default is 1.
15040 Set how many pixels to skip when sampling frames. Default is 1.
15041 Allowed range is from 1 to 1024.
15044 Leave frames unchanged. Default is disabled.
15047 @section pixdesctest
15049 Pixel format descriptor test filter, mainly useful for internal
15050 testing. The output video should be equal to the input video.
15054 format=monow, pixdesctest
15057 can be used to test the monowhite pixel format descriptor definition.
15061 Display sample values of color channels. Mainly useful for checking color
15062 and levels. Minimum supported resolution is 640x480.
15064 The filters accept the following options:
15068 Set scope X position, relative offset on X axis.
15071 Set scope Y position, relative offset on Y axis.
15080 Set window opacity. This window also holds statistics about pixel area.
15083 Set window X position, relative offset on X axis.
15086 Set window Y position, relative offset on Y axis.
15091 Enable the specified chain of postprocessing subfilters using libpostproc. This
15092 library should be automatically selected with a GPL build (@code{--enable-gpl}).
15093 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
15094 Each subfilter and some options have a short and a long name that can be used
15095 interchangeably, i.e. dr/dering are the same.
15097 The filters accept the following options:
15101 Set postprocessing subfilters string.
15104 All subfilters share common options to determine their scope:
15108 Honor the quality commands for this subfilter.
15111 Do chrominance filtering, too (default).
15114 Do luminance filtering only (no chrominance).
15117 Do chrominance filtering only (no luminance).
15120 These options can be appended after the subfilter name, separated by a '|'.
15122 Available subfilters are:
15125 @item hb/hdeblock[|difference[|flatness]]
15126 Horizontal deblocking filter
15129 Difference factor where higher values mean more deblocking (default: @code{32}).
15131 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15134 @item vb/vdeblock[|difference[|flatness]]
15135 Vertical deblocking filter
15138 Difference factor where higher values mean more deblocking (default: @code{32}).
15140 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15143 @item ha/hadeblock[|difference[|flatness]]
15144 Accurate horizontal deblocking filter
15147 Difference factor where higher values mean more deblocking (default: @code{32}).
15149 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15152 @item va/vadeblock[|difference[|flatness]]
15153 Accurate vertical deblocking filter
15156 Difference factor where higher values mean more deblocking (default: @code{32}).
15158 Flatness threshold where lower values mean more deblocking (default: @code{39}).
15162 The horizontal and vertical deblocking filters share the difference and
15163 flatness values so you cannot set different horizontal and vertical
15167 @item h1/x1hdeblock
15168 Experimental horizontal deblocking filter
15170 @item v1/x1vdeblock
15171 Experimental vertical deblocking filter
15176 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
15179 larger -> stronger filtering
15181 larger -> stronger filtering
15183 larger -> stronger filtering
15186 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
15189 Stretch luminance to @code{0-255}.
15192 @item lb/linblenddeint
15193 Linear blend deinterlacing filter that deinterlaces the given block by
15194 filtering all lines with a @code{(1 2 1)} filter.
15196 @item li/linipoldeint
15197 Linear interpolating deinterlacing filter that deinterlaces the given block by
15198 linearly interpolating every second line.
15200 @item ci/cubicipoldeint
15201 Cubic interpolating deinterlacing filter deinterlaces the given block by
15202 cubically interpolating every second line.
15204 @item md/mediandeint
15205 Median deinterlacing filter that deinterlaces the given block by applying a
15206 median filter to every second line.
15208 @item fd/ffmpegdeint
15209 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
15210 second line with a @code{(-1 4 2 4 -1)} filter.
15213 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
15214 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
15216 @item fq/forceQuant[|quantizer]
15217 Overrides the quantizer table from the input with the constant quantizer you
15225 Default pp filter combination (@code{hb|a,vb|a,dr|a})
15228 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
15231 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
15234 @subsection Examples
15238 Apply horizontal and vertical deblocking, deringing and automatic
15239 brightness/contrast:
15245 Apply default filters without brightness/contrast correction:
15251 Apply default filters and temporal denoiser:
15253 pp=default/tmpnoise|1|2|3
15257 Apply deblocking on luminance only, and switch vertical deblocking on or off
15258 automatically depending on available CPU time:
15265 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
15266 similar to spp = 6 with 7 point DCT, where only the center sample is
15269 The filter accepts the following options:
15273 Force a constant quantization parameter. It accepts an integer in range
15274 0 to 63. If not set, the filter will use the QP from the video stream
15278 Set thresholding mode. Available modes are:
15282 Set hard thresholding.
15284 Set soft thresholding (better de-ringing effect, but likely blurrier).
15286 Set medium thresholding (good results, default).
15290 @section premultiply
15291 Apply alpha premultiply effect to input video stream using first plane
15292 of second stream as alpha.
15294 Both streams must have same dimensions and same pixel format.
15296 The filter accepts the following option:
15300 Set which planes will be processed, unprocessed planes will be copied.
15301 By default value 0xf, all planes will be processed.
15304 Do not require 2nd input for processing, instead use alpha plane from input stream.
15308 Apply prewitt operator to input video stream.
15310 The filter accepts the following option:
15314 Set which planes will be processed, unprocessed planes will be copied.
15315 By default value 0xf, all planes will be processed.
15318 Set value which will be multiplied with filtered result.
15321 Set value which will be added to filtered result.
15324 @section pseudocolor
15326 Alter frame colors in video with pseudocolors.
15328 This filter accepts the following options:
15332 set pixel first component expression
15335 set pixel second component expression
15338 set pixel third component expression
15341 set pixel fourth component expression, corresponds to the alpha component
15344 set component to use as base for altering colors
15347 Each of them specifies the expression to use for computing the lookup table for
15348 the corresponding pixel component values.
15350 The expressions can contain the following constants and functions:
15355 The input width and height.
15358 The input value for the pixel component.
15360 @item ymin, umin, vmin, amin
15361 The minimum allowed component value.
15363 @item ymax, umax, vmax, amax
15364 The maximum allowed component value.
15367 All expressions default to "val".
15369 @subsection Examples
15373 Change too high luma values to gradient:
15375 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'"
15381 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
15382 Ratio) between two input videos.
15384 This filter takes in input two input videos, the first input is
15385 considered the "main" source and is passed unchanged to the
15386 output. The second input is used as a "reference" video for computing
15389 Both video inputs must have the same resolution and pixel format for
15390 this filter to work correctly. Also it assumes that both inputs
15391 have the same number of frames, which are compared one by one.
15393 The obtained average PSNR is printed through the logging system.
15395 The filter stores the accumulated MSE (mean squared error) of each
15396 frame, and at the end of the processing it is averaged across all frames
15397 equally, and the following formula is applied to obtain the PSNR:
15400 PSNR = 10*log10(MAX^2/MSE)
15403 Where MAX is the average of the maximum values of each component of the
15406 The description of the accepted parameters follows.
15409 @item stats_file, f
15410 If specified the filter will use the named file to save the PSNR of
15411 each individual frame. When filename equals "-" the data is sent to
15414 @item stats_version
15415 Specifies which version of the stats file format to use. Details of
15416 each format are written below.
15417 Default value is 1.
15419 @item stats_add_max
15420 Determines whether the max value is output to the stats log.
15421 Default value is 0.
15422 Requires stats_version >= 2. If this is set and stats_version < 2,
15423 the filter will return an error.
15426 This filter also supports the @ref{framesync} options.
15428 The file printed if @var{stats_file} is selected, contains a sequence of
15429 key/value pairs of the form @var{key}:@var{value} for each compared
15432 If a @var{stats_version} greater than 1 is specified, a header line precedes
15433 the list of per-frame-pair stats, with key value pairs following the frame
15434 format with the following parameters:
15437 @item psnr_log_version
15438 The version of the log file format. Will match @var{stats_version}.
15441 A comma separated list of the per-frame-pair parameters included in
15445 A description of each shown per-frame-pair parameter follows:
15449 sequential number of the input frame, starting from 1
15452 Mean Square Error pixel-by-pixel average difference of the compared
15453 frames, averaged over all the image components.
15455 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
15456 Mean Square Error pixel-by-pixel average difference of the compared
15457 frames for the component specified by the suffix.
15459 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
15460 Peak Signal to Noise ratio of the compared frames for the component
15461 specified by the suffix.
15463 @item max_avg, max_y, max_u, max_v
15464 Maximum allowed value for each channel, and average over all
15468 @subsection Examples
15473 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15474 [main][ref] psnr="stats_file=stats.log" [out]
15477 On this example the input file being processed is compared with the
15478 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
15479 is stored in @file{stats.log}.
15482 Another example with different containers:
15484 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 -
15491 Pulldown reversal (inverse telecine) filter, capable of handling mixed
15492 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
15495 The pullup filter is designed to take advantage of future context in making
15496 its decisions. This filter is stateless in the sense that it does not lock
15497 onto a pattern to follow, but it instead looks forward to the following
15498 fields in order to identify matches and rebuild progressive frames.
15500 To produce content with an even framerate, insert the fps filter after
15501 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
15502 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
15504 The filter accepts the following options:
15511 These options set the amount of "junk" to ignore at the left, right, top, and
15512 bottom of the image, respectively. Left and right are in units of 8 pixels,
15513 while top and bottom are in units of 2 lines.
15514 The default is 8 pixels on each side.
15517 Set the strict breaks. Setting this option to 1 will reduce the chances of
15518 filter generating an occasional mismatched frame, but it may also cause an
15519 excessive number of frames to be dropped during high motion sequences.
15520 Conversely, setting it to -1 will make filter match fields more easily.
15521 This may help processing of video where there is slight blurring between
15522 the fields, but may also cause there to be interlaced frames in the output.
15523 Default value is @code{0}.
15526 Set the metric plane to use. It accepts the following values:
15532 Use chroma blue plane.
15535 Use chroma red plane.
15538 This option may be set to use chroma plane instead of the default luma plane
15539 for doing filter's computations. This may improve accuracy on very clean
15540 source material, but more likely will decrease accuracy, especially if there
15541 is chroma noise (rainbow effect) or any grayscale video.
15542 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
15543 load and make pullup usable in realtime on slow machines.
15546 For best results (without duplicated frames in the output file) it is
15547 necessary to change the output frame rate. For example, to inverse
15548 telecine NTSC input:
15550 ffmpeg -i input -vf pullup -r 24000/1001 ...
15555 Change video quantization parameters (QP).
15557 The filter accepts the following option:
15561 Set expression for quantization parameter.
15564 The expression is evaluated through the eval API and can contain, among others,
15565 the following constants:
15569 1 if index is not 129, 0 otherwise.
15572 Sequential index starting from -129 to 128.
15575 @subsection Examples
15579 Some equation like:
15587 Flush video frames from internal cache of frames into a random order.
15588 No frame is discarded.
15589 Inspired by @ref{frei0r} nervous filter.
15593 Set size in number of frames of internal cache, in range from @code{2} to
15594 @code{512}. Default is @code{30}.
15597 Set seed for random number generator, must be an integer included between
15598 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15599 less than @code{0}, the filter will try to use a good random seed on a
15603 @section readeia608
15605 Read closed captioning (EIA-608) information from the top lines of a video frame.
15607 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
15608 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
15609 with EIA-608 data (starting from 0). A description of each metadata value follows:
15612 @item lavfi.readeia608.X.cc
15613 The two bytes stored as EIA-608 data (printed in hexadecimal).
15615 @item lavfi.readeia608.X.line
15616 The number of the line on which the EIA-608 data was identified and read.
15619 This filter accepts the following options:
15623 Set the line to start scanning for EIA-608 data. Default is @code{0}.
15626 Set the line to end scanning for EIA-608 data. Default is @code{29}.
15629 Set the ratio of width reserved for sync code detection.
15630 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
15633 Enable checking the parity bit. In the event of a parity error, the filter will output
15634 @code{0x00} for that character. Default is false.
15637 Lowpass lines prior to further processing. Default is enabled.
15640 @subsection Examples
15644 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
15646 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
15652 Read vertical interval timecode (VITC) information from the top lines of a
15655 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
15656 timecode value, if a valid timecode has been detected. Further metadata key
15657 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
15658 timecode data has been found or not.
15660 This filter accepts the following options:
15664 Set the maximum number of lines to scan for VITC data. If the value is set to
15665 @code{-1} the full video frame is scanned. Default is @code{45}.
15668 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
15669 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
15672 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
15673 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
15676 @subsection Examples
15680 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
15681 draw @code{--:--:--:--} as a placeholder:
15683 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
15689 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
15691 Destination pixel at position (X, Y) will be picked from source (x, y) position
15692 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
15693 value for pixel will be used for destination pixel.
15695 Xmap and Ymap input video streams must be of same dimensions. Output video stream
15696 will have Xmap/Ymap video stream dimensions.
15697 Xmap and Ymap input video streams are 16bit depth, single channel.
15701 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
15702 Default is @code{color}.
15705 Specify the color of the unmapped pixels. For the syntax of this option,
15706 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15707 manual,ffmpeg-utils}. Default color is @code{black}.
15710 @section removegrain
15712 The removegrain filter is a spatial denoiser for progressive video.
15716 Set mode for the first plane.
15719 Set mode for the second plane.
15722 Set mode for the third plane.
15725 Set mode for the fourth plane.
15728 Range of mode is from 0 to 24. Description of each mode follows:
15732 Leave input plane unchanged. Default.
15735 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
15738 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
15741 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
15744 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
15745 This is equivalent to a median filter.
15748 Line-sensitive clipping giving the minimal change.
15751 Line-sensitive clipping, intermediate.
15754 Line-sensitive clipping, intermediate.
15757 Line-sensitive clipping, intermediate.
15760 Line-sensitive clipping on a line where the neighbours pixels are the closest.
15763 Replaces the target pixel with the closest neighbour.
15766 [1 2 1] horizontal and vertical kernel blur.
15772 Bob mode, interpolates top field from the line where the neighbours
15773 pixels are the closest.
15776 Bob mode, interpolates bottom field from the line where the neighbours
15777 pixels are the closest.
15780 Bob mode, interpolates top field. Same as 13 but with a more complicated
15781 interpolation formula.
15784 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
15785 interpolation formula.
15788 Clips the pixel with the minimum and maximum of respectively the maximum and
15789 minimum of each pair of opposite neighbour pixels.
15792 Line-sensitive clipping using opposite neighbours whose greatest distance from
15793 the current pixel is minimal.
15796 Replaces the pixel with the average of its 8 neighbours.
15799 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
15802 Clips pixels using the averages of opposite neighbour.
15805 Same as mode 21 but simpler and faster.
15808 Small edge and halo removal, but reputed useless.
15814 @section removelogo
15816 Suppress a TV station logo, using an image file to determine which
15817 pixels comprise the logo. It works by filling in the pixels that
15818 comprise the logo with neighboring pixels.
15820 The filter accepts the following options:
15824 Set the filter bitmap file, which can be any image format supported by
15825 libavformat. The width and height of the image file must match those of the
15826 video stream being processed.
15829 Pixels in the provided bitmap image with a value of zero are not
15830 considered part of the logo, non-zero pixels are considered part of
15831 the logo. If you use white (255) for the logo and black (0) for the
15832 rest, you will be safe. For making the filter bitmap, it is
15833 recommended to take a screen capture of a black frame with the logo
15834 visible, and then using a threshold filter followed by the erode
15835 filter once or twice.
15837 If needed, little splotches can be fixed manually. Remember that if
15838 logo pixels are not covered, the filter quality will be much
15839 reduced. Marking too many pixels as part of the logo does not hurt as
15840 much, but it will increase the amount of blurring needed to cover over
15841 the image and will destroy more information than necessary, and extra
15842 pixels will slow things down on a large logo.
15844 @section repeatfields
15846 This filter uses the repeat_field flag from the Video ES headers and hard repeats
15847 fields based on its value.
15851 Reverse a video clip.
15853 Warning: This filter requires memory to buffer the entire clip, so trimming
15856 @subsection Examples
15860 Take the first 5 seconds of a clip, and reverse it.
15867 Shift R/G/B/A pixels horizontally and/or vertically.
15869 The filter accepts the following options:
15872 Set amount to shift red horizontally.
15874 Set amount to shift red vertically.
15876 Set amount to shift green horizontally.
15878 Set amount to shift green vertically.
15880 Set amount to shift blue horizontally.
15882 Set amount to shift blue vertically.
15884 Set amount to shift alpha horizontally.
15886 Set amount to shift alpha vertically.
15888 Set edge mode, can be @var{smear}, default, or @var{warp}.
15891 @subsection Commands
15893 This filter supports the all above options as @ref{commands}.
15896 Apply roberts cross operator to input video stream.
15898 The filter accepts the following option:
15902 Set which planes will be processed, unprocessed planes will be copied.
15903 By default value 0xf, all planes will be processed.
15906 Set value which will be multiplied with filtered result.
15909 Set value which will be added to filtered result.
15914 Rotate video by an arbitrary angle expressed in radians.
15916 The filter accepts the following options:
15918 A description of the optional parameters follows.
15921 Set an expression for the angle by which to rotate the input video
15922 clockwise, expressed as a number of radians. A negative value will
15923 result in a counter-clockwise rotation. By default it is set to "0".
15925 This expression is evaluated for each frame.
15928 Set the output width expression, default value is "iw".
15929 This expression is evaluated just once during configuration.
15932 Set the output height expression, default value is "ih".
15933 This expression is evaluated just once during configuration.
15936 Enable bilinear interpolation if set to 1, a value of 0 disables
15937 it. Default value is 1.
15940 Set the color used to fill the output area not covered by the rotated
15941 image. For the general syntax of this option, check the
15942 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15943 If the special value "none" is selected then no
15944 background is printed (useful for example if the background is never shown).
15946 Default value is "black".
15949 The expressions for the angle and the output size can contain the
15950 following constants and functions:
15954 sequential number of the input frame, starting from 0. It is always NAN
15955 before the first frame is filtered.
15958 time in seconds of the input frame, it is set to 0 when the filter is
15959 configured. It is always NAN before the first frame is filtered.
15963 horizontal and vertical chroma subsample values. For example for the
15964 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15968 the input video width and height
15972 the output width and height, that is the size of the padded area as
15973 specified by the @var{width} and @var{height} expressions
15977 the minimal width/height required for completely containing the input
15978 video rotated by @var{a} radians.
15980 These are only available when computing the @option{out_w} and
15981 @option{out_h} expressions.
15984 @subsection Examples
15988 Rotate the input by PI/6 radians clockwise:
15994 Rotate the input by PI/6 radians counter-clockwise:
16000 Rotate the input by 45 degrees clockwise:
16006 Apply a constant rotation with period T, starting from an angle of PI/3:
16008 rotate=PI/3+2*PI*t/T
16012 Make the input video rotation oscillating with a period of T
16013 seconds and an amplitude of A radians:
16015 rotate=A*sin(2*PI/T*t)
16019 Rotate the video, output size is chosen so that the whole rotating
16020 input video is always completely contained in the output:
16022 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
16026 Rotate the video, reduce the output size so that no background is ever
16029 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
16033 @subsection Commands
16035 The filter supports the following commands:
16039 Set the angle expression.
16040 The command accepts the same syntax of the corresponding option.
16042 If the specified expression is not valid, it is kept at its current
16048 Apply Shape Adaptive Blur.
16050 The filter accepts the following options:
16053 @item luma_radius, lr
16054 Set luma blur filter strength, must be a value in range 0.1-4.0, default
16055 value is 1.0. A greater value will result in a more blurred image, and
16056 in slower processing.
16058 @item luma_pre_filter_radius, lpfr
16059 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
16062 @item luma_strength, ls
16063 Set luma maximum difference between pixels to still be considered, must
16064 be a value in the 0.1-100.0 range, default value is 1.0.
16066 @item chroma_radius, cr
16067 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
16068 greater value will result in a more blurred image, and in slower
16071 @item chroma_pre_filter_radius, cpfr
16072 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
16074 @item chroma_strength, cs
16075 Set chroma maximum difference between pixels to still be considered,
16076 must be a value in the -0.9-100.0 range.
16079 Each chroma option value, if not explicitly specified, is set to the
16080 corresponding luma option value.
16085 Scale (resize) the input video, using the libswscale library.
16087 The scale filter forces the output display aspect ratio to be the same
16088 of the input, by changing the output sample aspect ratio.
16090 If the input image format is different from the format requested by
16091 the next filter, the scale filter will convert the input to the
16094 @subsection Options
16095 The filter accepts the following options, or any of the options
16096 supported by the libswscale scaler.
16098 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
16099 the complete list of scaler options.
16104 Set the output video dimension expression. Default value is the input
16107 If the @var{width} or @var{w} value is 0, the input width is used for
16108 the output. If the @var{height} or @var{h} value is 0, the input height
16109 is used for the output.
16111 If one and only one of the values is -n with n >= 1, the scale filter
16112 will use a value that maintains the aspect ratio of the input image,
16113 calculated from the other specified dimension. After that it will,
16114 however, make sure that the calculated dimension is divisible by n and
16115 adjust the value if necessary.
16117 If both values are -n with n >= 1, the behavior will be identical to
16118 both values being set to 0 as previously detailed.
16120 See below for the list of accepted constants for use in the dimension
16124 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
16128 Only evaluate expressions once during the filter initialization or when a command is processed.
16131 Evaluate expressions for each incoming frame.
16135 Default value is @samp{init}.
16139 Set the interlacing mode. It accepts the following values:
16143 Force interlaced aware scaling.
16146 Do not apply interlaced scaling.
16149 Select interlaced aware scaling depending on whether the source frames
16150 are flagged as interlaced or not.
16153 Default value is @samp{0}.
16156 Set libswscale scaling flags. See
16157 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
16158 complete list of values. If not explicitly specified the filter applies
16162 @item param0, param1
16163 Set libswscale input parameters for scaling algorithms that need them. See
16164 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
16165 complete documentation. If not explicitly specified the filter applies
16171 Set the video size. For the syntax of this option, check the
16172 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16174 @item in_color_matrix
16175 @item out_color_matrix
16176 Set in/output YCbCr color space type.
16178 This allows the autodetected value to be overridden as well as allows forcing
16179 a specific value used for the output and encoder.
16181 If not specified, the color space type depends on the pixel format.
16187 Choose automatically.
16190 Format conforming to International Telecommunication Union (ITU)
16191 Recommendation BT.709.
16194 Set color space conforming to the United States Federal Communications
16195 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
16200 Set color space conforming to:
16204 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
16207 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
16210 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
16215 Set color space conforming to SMPTE ST 240:1999.
16218 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
16223 Set in/output YCbCr sample range.
16225 This allows the autodetected value to be overridden as well as allows forcing
16226 a specific value used for the output and encoder. If not specified, the
16227 range depends on the pixel format. Possible values:
16231 Choose automatically.
16234 Set full range (0-255 in case of 8-bit luma).
16236 @item mpeg/limited/tv
16237 Set "MPEG" range (16-235 in case of 8-bit luma).
16240 @item force_original_aspect_ratio
16241 Enable decreasing or increasing output video width or height if necessary to
16242 keep the original aspect ratio. Possible values:
16246 Scale the video as specified and disable this feature.
16249 The output video dimensions will automatically be decreased if needed.
16252 The output video dimensions will automatically be increased if needed.
16256 One useful instance of this option is that when you know a specific device's
16257 maximum allowed resolution, you can use this to limit the output video to
16258 that, while retaining the aspect ratio. For example, device A allows
16259 1280x720 playback, and your video is 1920x800. Using this option (set it to
16260 decrease) and specifying 1280x720 to the command line makes the output
16263 Please note that this is a different thing than specifying -1 for @option{w}
16264 or @option{h}, you still need to specify the output resolution for this option
16267 @item force_divisible_by
16268 Ensures that both the output dimensions, width and height, are divisible by the
16269 given integer when used together with @option{force_original_aspect_ratio}. This
16270 works similar to using @code{-n} in the @option{w} and @option{h} options.
16272 This option respects the value set for @option{force_original_aspect_ratio},
16273 increasing or decreasing the resolution accordingly. The video's aspect ratio
16274 may be slightly modified.
16276 This option can be handy if you need to have a video fit within or exceed
16277 a defined resolution using @option{force_original_aspect_ratio} but also have
16278 encoder restrictions on width or height divisibility.
16282 The values of the @option{w} and @option{h} options are expressions
16283 containing the following constants:
16288 The input width and height
16292 These are the same as @var{in_w} and @var{in_h}.
16296 The output (scaled) width and height
16300 These are the same as @var{out_w} and @var{out_h}
16303 The same as @var{iw} / @var{ih}
16306 input sample aspect ratio
16309 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16313 horizontal and vertical input chroma subsample values. For example for the
16314 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16318 horizontal and vertical output chroma subsample values. For example for the
16319 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16322 The (sequential) number of the input frame, starting from 0.
16323 Only available with @code{eval=frame}.
16326 The presentation timestamp of the input frame, expressed as a number of
16327 seconds. Only available with @code{eval=frame}.
16330 The position (byte offset) of the frame in the input stream, or NaN if
16331 this information is unavailable and/or meaningless (for example in case of synthetic video).
16332 Only available with @code{eval=frame}.
16335 @subsection Examples
16339 Scale the input video to a size of 200x100
16344 This is equivalent to:
16355 Specify a size abbreviation for the output size:
16360 which can also be written as:
16366 Scale the input to 2x:
16368 scale=w=2*iw:h=2*ih
16372 The above is the same as:
16374 scale=2*in_w:2*in_h
16378 Scale the input to 2x with forced interlaced scaling:
16380 scale=2*iw:2*ih:interl=1
16384 Scale the input to half size:
16386 scale=w=iw/2:h=ih/2
16390 Increase the width, and set the height to the same size:
16396 Seek Greek harmony:
16403 Increase the height, and set the width to 3/2 of the height:
16405 scale=w=3/2*oh:h=3/5*ih
16409 Increase the size, making the size a multiple of the chroma
16412 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
16416 Increase the width to a maximum of 500 pixels,
16417 keeping the same aspect ratio as the input:
16419 scale=w='min(500\, iw*3/2):h=-1'
16423 Make pixels square by combining scale and setsar:
16425 scale='trunc(ih*dar):ih',setsar=1/1
16429 Make pixels square by combining scale and setsar,
16430 making sure the resulting resolution is even (required by some codecs):
16432 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
16436 @subsection Commands
16438 This filter supports the following commands:
16442 Set the output video dimension expression.
16443 The command accepts the same syntax of the corresponding option.
16445 If the specified expression is not valid, it is kept at its current
16451 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
16452 format conversion on CUDA video frames. Setting the output width and height
16453 works in the same way as for the @var{scale} filter.
16455 The following additional options are accepted:
16458 The pixel format of the output CUDA frames. If set to the string "same" (the
16459 default), the input format will be kept. Note that automatic format negotiation
16460 and conversion is not yet supported for hardware frames
16463 The interpolation algorithm used for resizing. One of the following:
16470 @item cubic2p_bspline
16471 2-parameter cubic (B=1, C=0)
16473 @item cubic2p_catmullrom
16474 2-parameter cubic (B=0, C=1/2)
16476 @item cubic2p_b05c03
16477 2-parameter cubic (B=1/2, C=3/10)
16485 @item force_original_aspect_ratio
16486 Enable decreasing or increasing output video width or height if necessary to
16487 keep the original aspect ratio. Possible values:
16491 Scale the video as specified and disable this feature.
16494 The output video dimensions will automatically be decreased if needed.
16497 The output video dimensions will automatically be increased if needed.
16501 One useful instance of this option is that when you know a specific device's
16502 maximum allowed resolution, you can use this to limit the output video to
16503 that, while retaining the aspect ratio. For example, device A allows
16504 1280x720 playback, and your video is 1920x800. Using this option (set it to
16505 decrease) and specifying 1280x720 to the command line makes the output
16508 Please note that this is a different thing than specifying -1 for @option{w}
16509 or @option{h}, you still need to specify the output resolution for this option
16512 @item force_divisible_by
16513 Ensures that both the output dimensions, width and height, are divisible by the
16514 given integer when used together with @option{force_original_aspect_ratio}. This
16515 works similar to using @code{-n} in the @option{w} and @option{h} options.
16517 This option respects the value set for @option{force_original_aspect_ratio},
16518 increasing or decreasing the resolution accordingly. The video's aspect ratio
16519 may be slightly modified.
16521 This option can be handy if you need to have a video fit within or exceed
16522 a defined resolution using @option{force_original_aspect_ratio} but also have
16523 encoder restrictions on width or height divisibility.
16529 Scale (resize) the input video, based on a reference video.
16531 See the scale filter for available options, scale2ref supports the same but
16532 uses the reference video instead of the main input as basis. scale2ref also
16533 supports the following additional constants for the @option{w} and
16534 @option{h} options:
16539 The main input video's width and height
16542 The same as @var{main_w} / @var{main_h}
16545 The main input video's sample aspect ratio
16547 @item main_dar, mdar
16548 The main input video's display aspect ratio. Calculated from
16549 @code{(main_w / main_h) * main_sar}.
16553 The main input video's horizontal and vertical chroma subsample values.
16554 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
16558 The (sequential) number of the main input frame, starting from 0.
16559 Only available with @code{eval=frame}.
16562 The presentation timestamp of the main input frame, expressed as a number of
16563 seconds. Only available with @code{eval=frame}.
16566 The position (byte offset) of the frame in the main input stream, or NaN if
16567 this information is unavailable and/or meaningless (for example in case of synthetic video).
16568 Only available with @code{eval=frame}.
16571 @subsection Examples
16575 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
16577 'scale2ref[b][a];[a][b]overlay'
16581 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
16583 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
16587 @subsection Commands
16589 This filter supports the following commands:
16593 Set the output video dimension expression.
16594 The command accepts the same syntax of the corresponding option.
16596 If the specified expression is not valid, it is kept at its current
16601 Scroll input video horizontally and/or vertically by constant speed.
16603 The filter accepts the following options:
16605 @item horizontal, h
16606 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
16607 Negative values changes scrolling direction.
16610 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
16611 Negative values changes scrolling direction.
16614 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
16617 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
16620 @subsection Commands
16622 This filter supports the following @ref{commands}:
16624 @item horizontal, h
16625 Set the horizontal scrolling speed.
16627 Set the vertical scrolling speed.
16633 Detect video scene change.
16635 This filter sets frame metadata with mafd between frame, the scene score, and
16636 forward the frame to the next filter, so they can use these metadata to detect
16637 scene change or others.
16639 In addition, this filter logs a message and sets frame metadata when it detects
16640 a scene change by @option{threshold}.
16642 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
16644 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
16645 to detect scene change.
16647 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
16648 detect scene change with @option{threshold}.
16650 The filter accepts the following options:
16654 Set the scene change detection threshold as a percentage of maximum change. Good
16655 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
16658 Default value is @code{10.}.
16661 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
16662 You can enable it if you want to get snapshot of scene change frames only.
16665 @anchor{selectivecolor}
16666 @section selectivecolor
16668 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
16669 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
16670 by the "purity" of the color (that is, how saturated it already is).
16672 This filter is similar to the Adobe Photoshop Selective Color tool.
16674 The filter accepts the following options:
16677 @item correction_method
16678 Select color correction method.
16680 Available values are:
16683 Specified adjustments are applied "as-is" (added/subtracted to original pixel
16686 Specified adjustments are relative to the original component value.
16688 Default is @code{absolute}.
16690 Adjustments for red pixels (pixels where the red component is the maximum)
16692 Adjustments for yellow pixels (pixels where the blue component is the minimum)
16694 Adjustments for green pixels (pixels where the green component is the maximum)
16696 Adjustments for cyan pixels (pixels where the red component is the minimum)
16698 Adjustments for blue pixels (pixels where the blue component is the maximum)
16700 Adjustments for magenta pixels (pixels where the green component is the minimum)
16702 Adjustments for white pixels (pixels where all components are greater than 128)
16704 Adjustments for all pixels except pure black and pure white
16706 Adjustments for black pixels (pixels where all components are lesser than 128)
16708 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
16711 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
16712 4 space separated floating point adjustment values in the [-1,1] range,
16713 respectively to adjust the amount of cyan, magenta, yellow and black for the
16714 pixels of its range.
16716 @subsection Examples
16720 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
16721 increase magenta by 27% in blue areas:
16723 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
16727 Use a Photoshop selective color preset:
16729 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
16733 @anchor{separatefields}
16734 @section separatefields
16736 The @code{separatefields} takes a frame-based video input and splits
16737 each frame into its components fields, producing a new half height clip
16738 with twice the frame rate and twice the frame count.
16740 This filter use field-dominance information in frame to decide which
16741 of each pair of fields to place first in the output.
16742 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
16744 @section setdar, setsar
16746 The @code{setdar} filter sets the Display Aspect Ratio for the filter
16749 This is done by changing the specified Sample (aka Pixel) Aspect
16750 Ratio, according to the following equation:
16752 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
16755 Keep in mind that the @code{setdar} filter does not modify the pixel
16756 dimensions of the video frame. Also, the display aspect ratio set by
16757 this filter may be changed by later filters in the filterchain,
16758 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
16761 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
16762 the filter output video.
16764 Note that as a consequence of the application of this filter, the
16765 output display aspect ratio will change according to the equation
16768 Keep in mind that the sample aspect ratio set by the @code{setsar}
16769 filter may be changed by later filters in the filterchain, e.g. if
16770 another "setsar" or a "setdar" filter is applied.
16772 It accepts the following parameters:
16775 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
16776 Set the aspect ratio used by the filter.
16778 The parameter can be a floating point number string, an expression, or
16779 a string of the form @var{num}:@var{den}, where @var{num} and
16780 @var{den} are the numerator and denominator of the aspect ratio. If
16781 the parameter is not specified, it is assumed the value "0".
16782 In case the form "@var{num}:@var{den}" is used, the @code{:} character
16786 Set the maximum integer value to use for expressing numerator and
16787 denominator when reducing the expressed aspect ratio to a rational.
16788 Default value is @code{100}.
16792 The parameter @var{sar} is an expression containing
16793 the following constants:
16797 These are approximated values for the mathematical constants e
16798 (Euler's number), pi (Greek pi), and phi (the golden ratio).
16801 The input width and height.
16804 These are the same as @var{w} / @var{h}.
16807 The input sample aspect ratio.
16810 The input display aspect ratio. It is the same as
16811 (@var{w} / @var{h}) * @var{sar}.
16814 Horizontal and vertical chroma subsample values. For example, for the
16815 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16818 @subsection Examples
16823 To change the display aspect ratio to 16:9, specify one of the following:
16830 To change the sample aspect ratio to 10:11, specify:
16836 To set a display aspect ratio of 16:9, and specify a maximum integer value of
16837 1000 in the aspect ratio reduction, use the command:
16839 setdar=ratio=16/9:max=1000
16847 Force field for the output video frame.
16849 The @code{setfield} filter marks the interlace type field for the
16850 output frames. It does not change the input frame, but only sets the
16851 corresponding property, which affects how the frame is treated by
16852 following filters (e.g. @code{fieldorder} or @code{yadif}).
16854 The filter accepts the following options:
16859 Available values are:
16863 Keep the same field property.
16866 Mark the frame as bottom-field-first.
16869 Mark the frame as top-field-first.
16872 Mark the frame as progressive.
16879 Force frame parameter for the output video frame.
16881 The @code{setparams} filter marks interlace and color range for the
16882 output frames. It does not change the input frame, but only sets the
16883 corresponding property, which affects how the frame is treated by
16888 Available values are:
16892 Keep the same field property (default).
16895 Mark the frame as bottom-field-first.
16898 Mark the frame as top-field-first.
16901 Mark the frame as progressive.
16905 Available values are:
16909 Keep the same color range property (default).
16911 @item unspecified, unknown
16912 Mark the frame as unspecified color range.
16914 @item limited, tv, mpeg
16915 Mark the frame as limited range.
16917 @item full, pc, jpeg
16918 Mark the frame as full range.
16921 @item color_primaries
16922 Set the color primaries.
16923 Available values are:
16927 Keep the same color primaries property (default).
16944 Set the color transfer.
16945 Available values are:
16949 Keep the same color trc property (default).
16971 Set the colorspace.
16972 Available values are:
16976 Keep the same colorspace property (default).
16989 @item chroma-derived-nc
16990 @item chroma-derived-c
16997 Show a line containing various information for each input video frame.
16998 The input video is not modified.
17000 This filter supports the following options:
17004 Calculate checksums of each plane. By default enabled.
17007 The shown line contains a sequence of key/value pairs of the form
17008 @var{key}:@var{value}.
17010 The following values are shown in the output:
17014 The (sequential) number of the input frame, starting from 0.
17017 The Presentation TimeStamp of the input frame, expressed as a number of
17018 time base units. The time base unit depends on the filter input pad.
17021 The Presentation TimeStamp of the input frame, expressed as a number of
17025 The position of the frame in the input stream, or -1 if this information is
17026 unavailable and/or meaningless (for example in case of synthetic video).
17029 The pixel format name.
17032 The sample aspect ratio of the input frame, expressed in the form
17033 @var{num}/@var{den}.
17036 The size of the input frame. For the syntax of this option, check the
17037 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17040 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
17041 for bottom field first).
17044 This is 1 if the frame is a key frame, 0 otherwise.
17047 The picture type of the input frame ("I" for an I-frame, "P" for a
17048 P-frame, "B" for a B-frame, or "?" for an unknown type).
17049 Also refer to the documentation of the @code{AVPictureType} enum and of
17050 the @code{av_get_picture_type_char} function defined in
17051 @file{libavutil/avutil.h}.
17054 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
17056 @item plane_checksum
17057 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
17058 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
17061 The mean value of pixels in each plane of the input frame, expressed in the form
17062 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
17065 The standard deviation of pixel values in each plane of the input frame, expressed
17066 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
17070 @section showpalette
17072 Displays the 256 colors palette of each frame. This filter is only relevant for
17073 @var{pal8} pixel format frames.
17075 It accepts the following option:
17079 Set the size of the box used to represent one palette color entry. Default is
17080 @code{30} (for a @code{30x30} pixel box).
17083 @section shuffleframes
17085 Reorder and/or duplicate and/or drop video frames.
17087 It accepts the following parameters:
17091 Set the destination indexes of input frames.
17092 This is space or '|' separated list of indexes that maps input frames to output
17093 frames. Number of indexes also sets maximal value that each index may have.
17094 '-1' index have special meaning and that is to drop frame.
17097 The first frame has the index 0. The default is to keep the input unchanged.
17099 @subsection Examples
17103 Swap second and third frame of every three frames of the input:
17105 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
17109 Swap 10th and 1st frame of every ten frames of the input:
17111 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
17115 @section shuffleplanes
17117 Reorder and/or duplicate video planes.
17119 It accepts the following parameters:
17124 The index of the input plane to be used as the first output plane.
17127 The index of the input plane to be used as the second output plane.
17130 The index of the input plane to be used as the third output plane.
17133 The index of the input plane to be used as the fourth output plane.
17137 The first plane has the index 0. The default is to keep the input unchanged.
17139 @subsection Examples
17143 Swap the second and third planes of the input:
17145 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
17149 @anchor{signalstats}
17150 @section signalstats
17151 Evaluate various visual metrics that assist in determining issues associated
17152 with the digitization of analog video media.
17154 By default the filter will log these metadata values:
17158 Display the minimal Y value contained within the input frame. Expressed in
17162 Display the Y value at the 10% percentile within the input frame. Expressed in
17166 Display the average Y value within the input frame. Expressed in range of
17170 Display the Y value at the 90% percentile within the input frame. Expressed in
17174 Display the maximum Y value contained within the input frame. Expressed in
17178 Display the minimal U value contained within the input frame. Expressed in
17182 Display the U value at the 10% percentile within the input frame. Expressed in
17186 Display the average U value within the input frame. Expressed in range of
17190 Display the U value at the 90% percentile within the input frame. Expressed in
17194 Display the maximum U value contained within the input frame. Expressed in
17198 Display the minimal V value contained within the input frame. Expressed in
17202 Display the V value at the 10% percentile within the input frame. Expressed in
17206 Display the average V value within the input frame. Expressed in range of
17210 Display the V value at the 90% percentile within the input frame. Expressed in
17214 Display the maximum V value contained within the input frame. Expressed in
17218 Display the minimal saturation value contained within the input frame.
17219 Expressed in range of [0-~181.02].
17222 Display the saturation value at the 10% percentile within the input frame.
17223 Expressed in range of [0-~181.02].
17226 Display the average saturation value within the input frame. Expressed in range
17230 Display the saturation value at the 90% percentile within the input frame.
17231 Expressed in range of [0-~181.02].
17234 Display the maximum saturation value contained within the input frame.
17235 Expressed in range of [0-~181.02].
17238 Display the median value for hue within the input frame. Expressed in range of
17242 Display the average value for hue within the input frame. Expressed in range of
17246 Display the average of sample value difference between all values of the Y
17247 plane in the current frame and corresponding values of the previous input frame.
17248 Expressed in range of [0-255].
17251 Display the average of sample value difference between all values of the U
17252 plane in the current frame and corresponding values of the previous input frame.
17253 Expressed in range of [0-255].
17256 Display the average of sample value difference between all values of the V
17257 plane in the current frame and corresponding values of the previous input frame.
17258 Expressed in range of [0-255].
17261 Display bit depth of Y plane in current frame.
17262 Expressed in range of [0-16].
17265 Display bit depth of U plane in current frame.
17266 Expressed in range of [0-16].
17269 Display bit depth of V plane in current frame.
17270 Expressed in range of [0-16].
17273 The filter accepts the following options:
17279 @option{stat} specify an additional form of image analysis.
17280 @option{out} output video with the specified type of pixel highlighted.
17282 Both options accept the following values:
17286 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
17287 unlike the neighboring pixels of the same field. Examples of temporal outliers
17288 include the results of video dropouts, head clogs, or tape tracking issues.
17291 Identify @var{vertical line repetition}. Vertical line repetition includes
17292 similar rows of pixels within a frame. In born-digital video vertical line
17293 repetition is common, but this pattern is uncommon in video digitized from an
17294 analog source. When it occurs in video that results from the digitization of an
17295 analog source it can indicate concealment from a dropout compensator.
17298 Identify pixels that fall outside of legal broadcast range.
17302 Set the highlight color for the @option{out} option. The default color is
17306 @subsection Examples
17310 Output data of various video metrics:
17312 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
17316 Output specific data about the minimum and maximum values of the Y plane per frame:
17318 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
17322 Playback video while highlighting pixels that are outside of broadcast range in red.
17324 ffplay example.mov -vf signalstats="out=brng:color=red"
17328 Playback video with signalstats metadata drawn over the frame.
17330 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
17333 The contents of signalstat_drawtext.txt used in the command are:
17336 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
17337 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
17338 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
17339 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
17347 Calculates the MPEG-7 Video Signature. The filter can handle more than one
17348 input. In this case the matching between the inputs can be calculated additionally.
17349 The filter always passes through the first input. The signature of each stream can
17350 be written into a file.
17352 It accepts the following options:
17356 Enable or disable the matching process.
17358 Available values are:
17362 Disable the calculation of a matching (default).
17364 Calculate the matching for the whole video and output whether the whole video
17365 matches or only parts.
17367 Calculate only until a matching is found or the video ends. Should be faster in
17372 Set the number of inputs. The option value must be a non negative integer.
17373 Default value is 1.
17376 Set the path to which the output is written. If there is more than one input,
17377 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
17378 integer), that will be replaced with the input number. If no filename is
17379 specified, no output will be written. This is the default.
17382 Choose the output format.
17384 Available values are:
17388 Use the specified binary representation (default).
17390 Use the specified xml representation.
17394 Set threshold to detect one word as similar. The option value must be an integer
17395 greater than zero. The default value is 9000.
17398 Set threshold to detect all words as similar. The option value must be an integer
17399 greater than zero. The default value is 60000.
17402 Set threshold to detect frames as similar. The option value must be an integer
17403 greater than zero. The default value is 116.
17406 Set the minimum length of a sequence in frames to recognize it as matching
17407 sequence. The option value must be a non negative integer value.
17408 The default value is 0.
17411 Set the minimum relation, that matching frames to all frames must have.
17412 The option value must be a double value between 0 and 1. The default value is 0.5.
17415 @subsection Examples
17419 To calculate the signature of an input video and store it in signature.bin:
17421 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
17425 To detect whether two videos match and store the signatures in XML format in
17426 signature0.xml and signature1.xml:
17428 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 -
17436 Blur the input video without impacting the outlines.
17438 It accepts the following options:
17441 @item luma_radius, lr
17442 Set the luma radius. The option value must be a float number in
17443 the range [0.1,5.0] that specifies the variance of the gaussian filter
17444 used to blur the image (slower if larger). Default value is 1.0.
17446 @item luma_strength, ls
17447 Set the luma strength. The option value must be a float number
17448 in the range [-1.0,1.0] that configures the blurring. A value included
17449 in [0.0,1.0] will blur the image whereas a value included in
17450 [-1.0,0.0] will sharpen the image. Default value is 1.0.
17452 @item luma_threshold, lt
17453 Set the luma threshold used as a coefficient to determine
17454 whether a pixel should be blurred or not. The option value must be an
17455 integer in the range [-30,30]. A value of 0 will filter all the image,
17456 a value included in [0,30] will filter flat areas and a value included
17457 in [-30,0] will filter edges. Default value is 0.
17459 @item chroma_radius, cr
17460 Set the chroma radius. The option value must be a float number in
17461 the range [0.1,5.0] that specifies the variance of the gaussian filter
17462 used to blur the image (slower if larger). Default value is @option{luma_radius}.
17464 @item chroma_strength, cs
17465 Set the chroma strength. The option value must be a float number
17466 in the range [-1.0,1.0] that configures the blurring. A value included
17467 in [0.0,1.0] will blur the image whereas a value included in
17468 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
17470 @item chroma_threshold, ct
17471 Set the chroma threshold used as a coefficient to determine
17472 whether a pixel should be blurred or not. The option value must be an
17473 integer in the range [-30,30]. A value of 0 will filter all the image,
17474 a value included in [0,30] will filter flat areas and a value included
17475 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
17478 If a chroma option is not explicitly set, the corresponding luma value
17482 Apply sobel operator to input video stream.
17484 The filter accepts the following option:
17488 Set which planes will be processed, unprocessed planes will be copied.
17489 By default value 0xf, all planes will be processed.
17492 Set value which will be multiplied with filtered result.
17495 Set value which will be added to filtered result.
17501 Apply a simple postprocessing filter that compresses and decompresses the image
17502 at several (or - in the case of @option{quality} level @code{6} - all) shifts
17503 and average the results.
17505 The filter accepts the following options:
17509 Set quality. This option defines the number of levels for averaging. It accepts
17510 an integer in the range 0-6. If set to @code{0}, the filter will have no
17511 effect. A value of @code{6} means the higher quality. For each increment of
17512 that value the speed drops by a factor of approximately 2. Default value is
17516 Force a constant quantization parameter. If not set, the filter will use the QP
17517 from the video stream (if available).
17520 Set thresholding mode. Available modes are:
17524 Set hard thresholding (default).
17526 Set soft thresholding (better de-ringing effect, but likely blurrier).
17529 @item use_bframe_qp
17530 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
17531 option may cause flicker since the B-Frames have often larger QP. Default is
17532 @code{0} (not enabled).
17535 @subsection Commands
17537 This filter supports the following commands:
17539 @item quality, level
17540 Set quality level. The value @code{max} can be used to set the maximum level,
17541 currently @code{6}.
17547 Scale the input by applying one of the super-resolution methods based on
17548 convolutional neural networks. Supported models:
17552 Super-Resolution Convolutional Neural Network model (SRCNN).
17553 See @url{https://arxiv.org/abs/1501.00092}.
17556 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
17557 See @url{https://arxiv.org/abs/1609.05158}.
17560 Training scripts as well as scripts for model file (.pb) saving can be found at
17561 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
17562 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
17564 Native model files (.model) can be generated from TensorFlow model
17565 files (.pb) by using tools/python/convert.py
17567 The filter accepts the following options:
17571 Specify which DNN backend to use for model loading and execution. This option accepts
17572 the following values:
17576 Native implementation of DNN loading and execution.
17579 TensorFlow backend. To enable this backend you
17580 need to install the TensorFlow for C library (see
17581 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
17582 @code{--enable-libtensorflow}
17585 Default value is @samp{native}.
17588 Set path to model file specifying network architecture and its parameters.
17589 Note that different backends use different file formats. TensorFlow backend
17590 can load files for both formats, while native backend can load files for only
17594 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
17595 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
17596 input upscaled using bicubic upscaling with proper scale factor.
17599 This feature can also be finished with @ref{dnn_processing} filter.
17603 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
17605 This filter takes in input two input videos, the first input is
17606 considered the "main" source and is passed unchanged to the
17607 output. The second input is used as a "reference" video for computing
17610 Both video inputs must have the same resolution and pixel format for
17611 this filter to work correctly. Also it assumes that both inputs
17612 have the same number of frames, which are compared one by one.
17614 The filter stores the calculated SSIM of each frame.
17616 The description of the accepted parameters follows.
17619 @item stats_file, f
17620 If specified the filter will use the named file to save the SSIM of
17621 each individual frame. When filename equals "-" the data is sent to
17625 The file printed if @var{stats_file} is selected, contains a sequence of
17626 key/value pairs of the form @var{key}:@var{value} for each compared
17629 A description of each shown parameter follows:
17633 sequential number of the input frame, starting from 1
17635 @item Y, U, V, R, G, B
17636 SSIM of the compared frames for the component specified by the suffix.
17639 SSIM of the compared frames for the whole frame.
17642 Same as above but in dB representation.
17645 This filter also supports the @ref{framesync} options.
17647 @subsection Examples
17652 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
17653 [main][ref] ssim="stats_file=stats.log" [out]
17656 On this example the input file being processed is compared with the
17657 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
17658 is stored in @file{stats.log}.
17661 Another example with both psnr and ssim at same time:
17663 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
17667 Another example with different containers:
17669 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 -
17675 Convert between different stereoscopic image formats.
17677 The filters accept the following options:
17681 Set stereoscopic image format of input.
17683 Available values for input image formats are:
17686 side by side parallel (left eye left, right eye right)
17689 side by side crosseye (right eye left, left eye right)
17692 side by side parallel with half width resolution
17693 (left eye left, right eye right)
17696 side by side crosseye with half width resolution
17697 (right eye left, left eye right)
17701 above-below (left eye above, right eye below)
17705 above-below (right eye above, left eye below)
17709 above-below with half height resolution
17710 (left eye above, right eye below)
17714 above-below with half height resolution
17715 (right eye above, left eye below)
17718 alternating frames (left eye first, right eye second)
17721 alternating frames (right eye first, left eye second)
17724 interleaved rows (left eye has top row, right eye starts on next row)
17727 interleaved rows (right eye has top row, left eye starts on next row)
17730 interleaved columns, left eye first
17733 interleaved columns, right eye first
17735 Default value is @samp{sbsl}.
17739 Set stereoscopic image format of output.
17743 side by side parallel (left eye left, right eye right)
17746 side by side crosseye (right eye left, left eye right)
17749 side by side parallel with half width resolution
17750 (left eye left, right eye right)
17753 side by side crosseye with half width resolution
17754 (right eye left, left eye right)
17758 above-below (left eye above, right eye below)
17762 above-below (right eye above, left eye below)
17766 above-below with half height resolution
17767 (left eye above, right eye below)
17771 above-below with half height resolution
17772 (right eye above, left eye below)
17775 alternating frames (left eye first, right eye second)
17778 alternating frames (right eye first, left eye second)
17781 interleaved rows (left eye has top row, right eye starts on next row)
17784 interleaved rows (right eye has top row, left eye starts on next row)
17787 anaglyph red/blue gray
17788 (red filter on left eye, blue filter on right eye)
17791 anaglyph red/green gray
17792 (red filter on left eye, green filter on right eye)
17795 anaglyph red/cyan gray
17796 (red filter on left eye, cyan filter on right eye)
17799 anaglyph red/cyan half colored
17800 (red filter on left eye, cyan filter on right eye)
17803 anaglyph red/cyan color
17804 (red filter on left eye, cyan filter on right eye)
17807 anaglyph red/cyan color optimized with the least squares projection of dubois
17808 (red filter on left eye, cyan filter on right eye)
17811 anaglyph green/magenta gray
17812 (green filter on left eye, magenta filter on right eye)
17815 anaglyph green/magenta half colored
17816 (green filter on left eye, magenta filter on right eye)
17819 anaglyph green/magenta colored
17820 (green filter on left eye, magenta filter on right eye)
17823 anaglyph green/magenta color optimized with the least squares projection of dubois
17824 (green filter on left eye, magenta filter on right eye)
17827 anaglyph yellow/blue gray
17828 (yellow filter on left eye, blue filter on right eye)
17831 anaglyph yellow/blue half colored
17832 (yellow filter on left eye, blue filter on right eye)
17835 anaglyph yellow/blue colored
17836 (yellow filter on left eye, blue filter on right eye)
17839 anaglyph yellow/blue color optimized with the least squares projection of dubois
17840 (yellow filter on left eye, blue filter on right eye)
17843 mono output (left eye only)
17846 mono output (right eye only)
17849 checkerboard, left eye first
17852 checkerboard, right eye first
17855 interleaved columns, left eye first
17858 interleaved columns, right eye first
17864 Default value is @samp{arcd}.
17867 @subsection Examples
17871 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
17877 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
17883 @section streamselect, astreamselect
17884 Select video or audio streams.
17886 The filter accepts the following options:
17890 Set number of inputs. Default is 2.
17893 Set input indexes to remap to outputs.
17896 @subsection Commands
17898 The @code{streamselect} and @code{astreamselect} filter supports the following
17903 Set input indexes to remap to outputs.
17906 @subsection Examples
17910 Select first 5 seconds 1st stream and rest of time 2nd stream:
17912 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
17916 Same as above, but for audio:
17918 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
17925 Draw subtitles on top of input video using the libass library.
17927 To enable compilation of this filter you need to configure FFmpeg with
17928 @code{--enable-libass}. This filter also requires a build with libavcodec and
17929 libavformat to convert the passed subtitles file to ASS (Advanced Substation
17930 Alpha) subtitles format.
17932 The filter accepts the following options:
17936 Set the filename of the subtitle file to read. It must be specified.
17938 @item original_size
17939 Specify the size of the original video, the video for which the ASS file
17940 was composed. For the syntax of this option, check the
17941 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17942 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
17943 correctly scale the fonts if the aspect ratio has been changed.
17946 Set a directory path containing fonts that can be used by the filter.
17947 These fonts will be used in addition to whatever the font provider uses.
17950 Process alpha channel, by default alpha channel is untouched.
17953 Set subtitles input character encoding. @code{subtitles} filter only. Only
17954 useful if not UTF-8.
17956 @item stream_index, si
17957 Set subtitles stream index. @code{subtitles} filter only.
17960 Override default style or script info parameters of the subtitles. It accepts a
17961 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
17964 If the first key is not specified, it is assumed that the first value
17965 specifies the @option{filename}.
17967 For example, to render the file @file{sub.srt} on top of the input
17968 video, use the command:
17973 which is equivalent to:
17975 subtitles=filename=sub.srt
17978 To render the default subtitles stream from file @file{video.mkv}, use:
17980 subtitles=video.mkv
17983 To render the second subtitles stream from that file, use:
17985 subtitles=video.mkv:si=1
17988 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
17989 @code{DejaVu Serif}, use:
17991 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
17994 @section super2xsai
17996 Scale the input by 2x and smooth using the Super2xSaI (Scale and
17997 Interpolate) pixel art scaling algorithm.
17999 Useful for enlarging pixel art images without reducing sharpness.
18003 Swap two rectangular objects in video.
18005 This filter accepts the following options:
18015 Set 1st rect x coordinate.
18018 Set 1st rect y coordinate.
18021 Set 2nd rect x coordinate.
18024 Set 2nd rect y coordinate.
18026 All expressions are evaluated once for each frame.
18029 The all options are expressions containing the following constants:
18034 The input width and height.
18037 same as @var{w} / @var{h}
18040 input sample aspect ratio
18043 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
18046 The number of the input frame, starting from 0.
18049 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
18052 the position in the file of the input frame, NAN if unknown
18059 Blend successive video frames.
18065 Apply telecine process to the video.
18067 This filter accepts the following options:
18076 The default value is @code{top}.
18080 A string of numbers representing the pulldown pattern you wish to apply.
18081 The default value is @code{23}.
18085 Some typical patterns:
18090 24p: 2332 (preferred)
18097 24p: 222222222223 ("Euro pulldown")
18102 @section thistogram
18104 Compute and draw a color distribution histogram for the input video across time.
18106 Unlike @ref{histogram} video filter which only shows histogram of single input frame
18107 at certain time, this filter shows also past histograms of number of frames defined
18108 by @code{width} option.
18110 The computed histogram is a representation of the color component
18111 distribution in an image.
18113 The filter accepts the following options:
18117 Set width of single color component output. Default value is @code{0}.
18118 Value of @code{0} means width will be picked from input video.
18119 This also set number of passed histograms to keep.
18120 Allowed range is [0, 8192].
18122 @item display_mode, d
18124 It accepts the following values:
18127 Per color component graphs are placed below each other.
18130 Per color component graphs are placed side by side.
18133 Presents information identical to that in the @code{parade}, except
18134 that the graphs representing color components are superimposed directly
18137 Default is @code{stack}.
18139 @item levels_mode, m
18140 Set mode. Can be either @code{linear}, or @code{logarithmic}.
18141 Default is @code{linear}.
18143 @item components, c
18144 Set what color components to display.
18145 Default is @code{7}.
18148 Set background opacity. Default is @code{0.9}.
18151 Show envelope. Default is disabled.
18154 Set envelope color. Default is @code{gold}.
18159 Apply threshold effect to video stream.
18161 This filter needs four video streams to perform thresholding.
18162 First stream is stream we are filtering.
18163 Second stream is holding threshold values, third stream is holding min values,
18164 and last, fourth stream is holding max values.
18166 The filter accepts the following option:
18170 Set which planes will be processed, unprocessed planes will be copied.
18171 By default value 0xf, all planes will be processed.
18174 For example if first stream pixel's component value is less then threshold value
18175 of pixel component from 2nd threshold stream, third stream value will picked,
18176 otherwise fourth stream pixel component value will be picked.
18178 Using color source filter one can perform various types of thresholding:
18180 @subsection Examples
18184 Binary threshold, using gray color as threshold:
18186 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
18190 Inverted binary threshold, using gray color as threshold:
18192 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
18196 Truncate binary threshold, using gray color as threshold:
18198 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
18202 Threshold to zero, using gray color as threshold:
18204 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
18208 Inverted threshold to zero, using gray color as threshold:
18210 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
18215 Select the most representative frame in a given sequence of consecutive frames.
18217 The filter accepts the following options:
18221 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
18222 will pick one of them, and then handle the next batch of @var{n} frames until
18223 the end. Default is @code{100}.
18226 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
18227 value will result in a higher memory usage, so a high value is not recommended.
18229 @subsection Examples
18233 Extract one picture each 50 frames:
18239 Complete example of a thumbnail creation with @command{ffmpeg}:
18241 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
18248 Tile several successive frames together.
18250 The @ref{untile} filter can do the reverse.
18252 The filter accepts the following options:
18257 Set the grid size (i.e. the number of lines and columns). For the syntax of
18258 this option, check the
18259 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18262 Set the maximum number of frames to render in the given area. It must be less
18263 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
18264 the area will be used.
18267 Set the outer border margin in pixels.
18270 Set the inner border thickness (i.e. the number of pixels between frames). For
18271 more advanced padding options (such as having different values for the edges),
18272 refer to the pad video filter.
18275 Specify the color of the unused area. For the syntax of this option, check the
18276 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18277 The default value of @var{color} is "black".
18280 Set the number of frames to overlap when tiling several successive frames together.
18281 The value must be between @code{0} and @var{nb_frames - 1}.
18284 Set the number of frames to initially be empty before displaying first output frame.
18285 This controls how soon will one get first output frame.
18286 The value must be between @code{0} and @var{nb_frames - 1}.
18289 @subsection Examples
18293 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
18295 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
18297 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
18298 duplicating each output frame to accommodate the originally detected frame
18302 Display @code{5} pictures in an area of @code{3x2} frames,
18303 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
18304 mixed flat and named options:
18306 tile=3x2:nb_frames=5:padding=7:margin=2
18310 @section tinterlace
18312 Perform various types of temporal field interlacing.
18314 Frames are counted starting from 1, so the first input frame is
18317 The filter accepts the following options:
18322 Specify the mode of the interlacing. This option can also be specified
18323 as a value alone. See below for a list of values for this option.
18325 Available values are:
18329 Move odd frames into the upper field, even into the lower field,
18330 generating a double height frame at half frame rate.
18334 Frame 1 Frame 2 Frame 3 Frame 4
18336 11111 22222 33333 44444
18337 11111 22222 33333 44444
18338 11111 22222 33333 44444
18339 11111 22222 33333 44444
18353 Only output odd frames, even frames are dropped, generating a frame with
18354 unchanged height at half frame rate.
18359 Frame 1 Frame 2 Frame 3 Frame 4
18361 11111 22222 33333 44444
18362 11111 22222 33333 44444
18363 11111 22222 33333 44444
18364 11111 22222 33333 44444
18374 Only output even frames, odd frames are dropped, generating a frame with
18375 unchanged height at half frame rate.
18380 Frame 1 Frame 2 Frame 3 Frame 4
18382 11111 22222 33333 44444
18383 11111 22222 33333 44444
18384 11111 22222 33333 44444
18385 11111 22222 33333 44444
18395 Expand each frame to full height, but pad alternate lines with black,
18396 generating a frame with double height at the same input frame rate.
18401 Frame 1 Frame 2 Frame 3 Frame 4
18403 11111 22222 33333 44444
18404 11111 22222 33333 44444
18405 11111 22222 33333 44444
18406 11111 22222 33333 44444
18409 11111 ..... 33333 .....
18410 ..... 22222 ..... 44444
18411 11111 ..... 33333 .....
18412 ..... 22222 ..... 44444
18413 11111 ..... 33333 .....
18414 ..... 22222 ..... 44444
18415 11111 ..... 33333 .....
18416 ..... 22222 ..... 44444
18420 @item interleave_top, 4
18421 Interleave the upper field from odd frames with the lower field from
18422 even frames, generating a frame with unchanged height at half frame rate.
18427 Frame 1 Frame 2 Frame 3 Frame 4
18429 11111<- 22222 33333<- 44444
18430 11111 22222<- 33333 44444<-
18431 11111<- 22222 33333<- 44444
18432 11111 22222<- 33333 44444<-
18442 @item interleave_bottom, 5
18443 Interleave the lower field from odd frames with the upper field from
18444 even frames, generating a frame with unchanged height at half frame rate.
18449 Frame 1 Frame 2 Frame 3 Frame 4
18451 11111 22222<- 33333 44444<-
18452 11111<- 22222 33333<- 44444
18453 11111 22222<- 33333 44444<-
18454 11111<- 22222 33333<- 44444
18464 @item interlacex2, 6
18465 Double frame rate with unchanged height. Frames are inserted each
18466 containing the second temporal field from the previous input frame and
18467 the first temporal field from the next input frame. This mode relies on
18468 the top_field_first flag. Useful for interlaced video displays with no
18469 field synchronisation.
18474 Frame 1 Frame 2 Frame 3 Frame 4
18476 11111 22222 33333 44444
18477 11111 22222 33333 44444
18478 11111 22222 33333 44444
18479 11111 22222 33333 44444
18482 11111 22222 22222 33333 33333 44444 44444
18483 11111 11111 22222 22222 33333 33333 44444
18484 11111 22222 22222 33333 33333 44444 44444
18485 11111 11111 22222 22222 33333 33333 44444
18490 Move odd frames into the upper field, even into the lower field,
18491 generating a double height frame at same frame rate.
18496 Frame 1 Frame 2 Frame 3 Frame 4
18498 11111 22222 33333 44444
18499 11111 22222 33333 44444
18500 11111 22222 33333 44444
18501 11111 22222 33333 44444
18504 11111 33333 33333 55555
18505 22222 22222 44444 44444
18506 11111 33333 33333 55555
18507 22222 22222 44444 44444
18508 11111 33333 33333 55555
18509 22222 22222 44444 44444
18510 11111 33333 33333 55555
18511 22222 22222 44444 44444
18516 Numeric values are deprecated but are accepted for backward
18517 compatibility reasons.
18519 Default mode is @code{merge}.
18522 Specify flags influencing the filter process.
18524 Available value for @var{flags} is:
18527 @item low_pass_filter, vlpf
18528 Enable linear vertical low-pass filtering in the filter.
18529 Vertical low-pass filtering is required when creating an interlaced
18530 destination from a progressive source which contains high-frequency
18531 vertical detail. Filtering will reduce interlace 'twitter' and Moire
18534 @item complex_filter, cvlpf
18535 Enable complex vertical low-pass filtering.
18536 This will slightly less reduce interlace 'twitter' and Moire
18537 patterning but better retain detail and subjective sharpness impression.
18540 Bypass already interlaced frames, only adjust the frame rate.
18543 Vertical low-pass filtering and bypassing already interlaced frames can only be
18544 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
18549 Pick median pixels from several successive input video frames.
18551 The filter accepts the following options:
18555 Set radius of median filter.
18556 Default is 1. Allowed range is from 1 to 127.
18559 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
18562 Set median percentile. Default value is @code{0.5}.
18563 Default value of @code{0.5} will pick always median values, while @code{0} will pick
18564 minimum values, and @code{1} maximum values.
18569 Mix successive video frames.
18571 A description of the accepted options follows.
18575 The number of successive frames to mix. If unspecified, it defaults to 3.
18578 Specify weight of each input video frame.
18579 Each weight is separated by space. If number of weights is smaller than
18580 number of @var{frames} last specified weight will be used for all remaining
18584 Specify scale, if it is set it will be multiplied with sum
18585 of each weight multiplied with pixel values to give final destination
18586 pixel value. By default @var{scale} is auto scaled to sum of weights.
18589 @subsection Examples
18593 Average 7 successive frames:
18595 tmix=frames=7:weights="1 1 1 1 1 1 1"
18599 Apply simple temporal convolution:
18601 tmix=frames=3:weights="-1 3 -1"
18605 Similar as above but only showing temporal differences:
18607 tmix=frames=3:weights="-1 2 -1":scale=1
18613 Tone map colors from different dynamic ranges.
18615 This filter expects data in single precision floating point, as it needs to
18616 operate on (and can output) out-of-range values. Another filter, such as
18617 @ref{zscale}, is needed to convert the resulting frame to a usable format.
18619 The tonemapping algorithms implemented only work on linear light, so input
18620 data should be linearized beforehand (and possibly correctly tagged).
18623 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
18626 @subsection Options
18627 The filter accepts the following options.
18631 Set the tone map algorithm to use.
18633 Possible values are:
18636 Do not apply any tone map, only desaturate overbright pixels.
18639 Hard-clip any out-of-range values. Use it for perfect color accuracy for
18640 in-range values, while distorting out-of-range values.
18643 Stretch the entire reference gamut to a linear multiple of the display.
18646 Fit a logarithmic transfer between the tone curves.
18649 Preserve overall image brightness with a simple curve, using nonlinear
18650 contrast, which results in flattening details and degrading color accuracy.
18653 Preserve both dark and bright details better than @var{reinhard}, at the cost
18654 of slightly darkening everything. Use it when detail preservation is more
18655 important than color and brightness accuracy.
18658 Smoothly map out-of-range values, while retaining contrast and colors for
18659 in-range material as much as possible. Use it when color accuracy is more
18660 important than detail preservation.
18666 Tune the tone mapping algorithm.
18668 This affects the following algorithms:
18674 Specifies the scale factor to use while stretching.
18678 Specifies the exponent of the function.
18682 Specify an extra linear coefficient to multiply into the signal before clipping.
18686 Specify the local contrast coefficient at the display peak.
18687 Default to 0.5, which means that in-gamut values will be about half as bright
18694 Specify the transition point from linear to mobius transform. Every value
18695 below this point is guaranteed to be mapped 1:1. The higher the value, the
18696 more accurate the result will be, at the cost of losing bright details.
18697 Default to 0.3, which due to the steep initial slope still preserves in-range
18698 colors fairly accurately.
18702 Apply desaturation for highlights that exceed this level of brightness. The
18703 higher the parameter, the more color information will be preserved. This
18704 setting helps prevent unnaturally blown-out colors for super-highlights, by
18705 (smoothly) turning into white instead. This makes images feel more natural,
18706 at the cost of reducing information about out-of-range colors.
18708 The default of 2.0 is somewhat conservative and will mostly just apply to
18709 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
18711 This option works only if the input frame has a supported color tag.
18714 Override signal/nominal/reference peak with this value. Useful when the
18715 embedded peak information in display metadata is not reliable or when tone
18716 mapping from a lower range to a higher range.
18721 Temporarily pad video frames.
18723 The filter accepts the following options:
18727 Specify number of delay frames before input video stream. Default is 0.
18730 Specify number of padding frames after input video stream.
18731 Set to -1 to pad indefinitely. Default is 0.
18734 Set kind of frames added to beginning of stream.
18735 Can be either @var{add} or @var{clone}.
18736 With @var{add} frames of solid-color are added.
18737 With @var{clone} frames are clones of first frame.
18738 Default is @var{add}.
18741 Set kind of frames added to end of stream.
18742 Can be either @var{add} or @var{clone}.
18743 With @var{add} frames of solid-color are added.
18744 With @var{clone} frames are clones of last frame.
18745 Default is @var{add}.
18747 @item start_duration, stop_duration
18748 Specify the duration of the start/stop delay. See
18749 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18750 for the accepted syntax.
18751 These options override @var{start} and @var{stop}. Default is 0.
18754 Specify the color of the padded area. For the syntax of this option,
18755 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
18756 manual,ffmpeg-utils}.
18758 The default value of @var{color} is "black".
18764 Transpose rows with columns in the input video and optionally flip it.
18766 It accepts the following parameters:
18771 Specify the transposition direction.
18773 Can assume the following values:
18775 @item 0, 4, cclock_flip
18776 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
18784 Rotate by 90 degrees clockwise, that is:
18792 Rotate by 90 degrees counterclockwise, that is:
18799 @item 3, 7, clock_flip
18800 Rotate by 90 degrees clockwise and vertically flip, that is:
18808 For values between 4-7, the transposition is only done if the input
18809 video geometry is portrait and not landscape. These values are
18810 deprecated, the @code{passthrough} option should be used instead.
18812 Numerical values are deprecated, and should be dropped in favor of
18813 symbolic constants.
18816 Do not apply the transposition if the input geometry matches the one
18817 specified by the specified value. It accepts the following values:
18820 Always apply transposition.
18822 Preserve portrait geometry (when @var{height} >= @var{width}).
18824 Preserve landscape geometry (when @var{width} >= @var{height}).
18827 Default value is @code{none}.
18830 For example to rotate by 90 degrees clockwise and preserve portrait
18833 transpose=dir=1:passthrough=portrait
18836 The command above can also be specified as:
18838 transpose=1:portrait
18841 @section transpose_npp
18843 Transpose rows with columns in the input video and optionally flip it.
18844 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
18846 It accepts the following parameters:
18851 Specify the transposition direction.
18853 Can assume the following values:
18856 Rotate by 90 degrees counterclockwise and vertically flip. (default)
18859 Rotate by 90 degrees clockwise.
18862 Rotate by 90 degrees counterclockwise.
18865 Rotate by 90 degrees clockwise and vertically flip.
18869 Do not apply the transposition if the input geometry matches the one
18870 specified by the specified value. It accepts the following values:
18873 Always apply transposition. (default)
18875 Preserve portrait geometry (when @var{height} >= @var{width}).
18877 Preserve landscape geometry (when @var{width} >= @var{height}).
18883 Trim the input so that the output contains one continuous subpart of the input.
18885 It accepts the following parameters:
18888 Specify the time of the start of the kept section, i.e. the frame with the
18889 timestamp @var{start} will be the first frame in the output.
18892 Specify the time of the first frame that will be dropped, i.e. the frame
18893 immediately preceding the one with the timestamp @var{end} will be the last
18894 frame in the output.
18897 This is the same as @var{start}, except this option sets the start timestamp
18898 in timebase units instead of seconds.
18901 This is the same as @var{end}, except this option sets the end timestamp
18902 in timebase units instead of seconds.
18905 The maximum duration of the output in seconds.
18908 The number of the first frame that should be passed to the output.
18911 The number of the first frame that should be dropped.
18914 @option{start}, @option{end}, and @option{duration} are expressed as time
18915 duration specifications; see
18916 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
18917 for the accepted syntax.
18919 Note that the first two sets of the start/end options and the @option{duration}
18920 option look at the frame timestamp, while the _frame variants simply count the
18921 frames that pass through the filter. Also note that this filter does not modify
18922 the timestamps. If you wish for the output timestamps to start at zero, insert a
18923 setpts filter after the trim filter.
18925 If multiple start or end options are set, this filter tries to be greedy and
18926 keep all the frames that match at least one of the specified constraints. To keep
18927 only the part that matches all the constraints at once, chain multiple trim
18930 The defaults are such that all the input is kept. So it is possible to set e.g.
18931 just the end values to keep everything before the specified time.
18936 Drop everything except the second minute of input:
18938 ffmpeg -i INPUT -vf trim=60:120
18942 Keep only the first second:
18944 ffmpeg -i INPUT -vf trim=duration=1
18949 @section unpremultiply
18950 Apply alpha unpremultiply effect to input video stream using first plane
18951 of second stream as alpha.
18953 Both streams must have same dimensions and same pixel format.
18955 The filter accepts the following option:
18959 Set which planes will be processed, unprocessed planes will be copied.
18960 By default value 0xf, all planes will be processed.
18962 If the format has 1 or 2 components, then luma is bit 0.
18963 If the format has 3 or 4 components:
18964 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
18965 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
18966 If present, the alpha channel is always the last bit.
18969 Do not require 2nd input for processing, instead use alpha plane from input stream.
18975 Sharpen or blur the input video.
18977 It accepts the following parameters:
18980 @item luma_msize_x, lx
18981 Set the luma matrix horizontal size. It must be an odd integer between
18982 3 and 23. The default value is 5.
18984 @item luma_msize_y, ly
18985 Set the luma matrix vertical size. It must be an odd integer between 3
18986 and 23. The default value is 5.
18988 @item luma_amount, la
18989 Set the luma effect strength. It must be a floating point number, reasonable
18990 values lay between -1.5 and 1.5.
18992 Negative values will blur the input video, while positive values will
18993 sharpen it, a value of zero will disable the effect.
18995 Default value is 1.0.
18997 @item chroma_msize_x, cx
18998 Set the chroma matrix horizontal size. It must be an odd integer
18999 between 3 and 23. The default value is 5.
19001 @item chroma_msize_y, cy
19002 Set the chroma matrix vertical size. It must be an odd integer
19003 between 3 and 23. The default value is 5.
19005 @item chroma_amount, ca
19006 Set the chroma effect strength. It must be a floating point number, reasonable
19007 values lay between -1.5 and 1.5.
19009 Negative values will blur the input video, while positive values will
19010 sharpen it, a value of zero will disable the effect.
19012 Default value is 0.0.
19016 All parameters are optional and default to the equivalent of the
19017 string '5:5:1.0:5:5:0.0'.
19019 @subsection Examples
19023 Apply strong luma sharpen effect:
19025 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
19029 Apply a strong blur of both luma and chroma parameters:
19031 unsharp=7:7:-2:7:7:-2
19038 Decompose a video made of tiled images into the individual images.
19040 The frame rate of the output video is the frame rate of the input video
19041 multiplied by the number of tiles.
19043 This filter does the reverse of @ref{tile}.
19045 The filter accepts the following options:
19050 Set the grid size (i.e. the number of lines and columns). For the syntax of
19051 this option, check the
19052 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19055 @subsection Examples
19059 Produce a 1-second video from a still image file made of 25 frames stacked
19060 vertically, like an analogic film reel:
19062 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
19068 Apply ultra slow/simple postprocessing filter that compresses and decompresses
19069 the image at several (or - in the case of @option{quality} level @code{8} - all)
19070 shifts and average the results.
19072 The way this differs from the behavior of spp is that uspp actually encodes &
19073 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
19074 DCT similar to MJPEG.
19076 The filter accepts the following options:
19080 Set quality. This option defines the number of levels for averaging. It accepts
19081 an integer in the range 0-8. If set to @code{0}, the filter will have no
19082 effect. A value of @code{8} means the higher quality. For each increment of
19083 that value the speed drops by a factor of approximately 2. Default value is
19087 Force a constant quantization parameter. If not set, the filter will use the QP
19088 from the video stream (if available).
19093 Convert 360 videos between various formats.
19095 The filter accepts the following options:
19101 Set format of the input/output video.
19109 Equirectangular projection.
19114 Cubemap with 3x2/6x1/1x6 layout.
19116 Format specific options:
19121 Set padding proportion for the input/output cubemap. Values in decimals.
19128 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)
19131 Default value is @b{@samp{0}}.
19132 Maximum value is @b{@samp{0.1}}.
19136 Set fixed padding for the input/output cubemap. Values in pixels.
19138 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
19142 Set order of faces for the input/output cubemap. Choose one direction for each position.
19144 Designation of directions:
19160 Default value is @b{@samp{rludfb}}.
19164 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
19166 Designation of angles:
19169 0 degrees clockwise
19171 90 degrees clockwise
19173 180 degrees clockwise
19175 270 degrees clockwise
19178 Default value is @b{@samp{000000}}.
19182 Equi-Angular Cubemap.
19189 Format specific options:
19194 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19196 If diagonal field of view is set it overrides horizontal and vertical field of view.
19201 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19203 If diagonal field of view is set it overrides horizontal and vertical field of view.
19209 Format specific options:
19214 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19216 If diagonal field of view is set it overrides horizontal and vertical field of view.
19221 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19223 If diagonal field of view is set it overrides horizontal and vertical field of view.
19229 Facebook's 360 formats.
19232 Stereographic format.
19234 Format specific options:
19239 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19241 If diagonal field of view is set it overrides horizontal and vertical field of view.
19246 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19248 If diagonal field of view is set it overrides horizontal and vertical field of view.
19255 Ball format, gives significant distortion toward the back.
19258 Hammer-Aitoff map projection format.
19261 Sinusoidal map projection format.
19264 Fisheye projection.
19266 Format specific options:
19271 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19273 If diagonal field of view is set it overrides horizontal and vertical field of view.
19278 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19280 If diagonal field of view is set it overrides horizontal and vertical field of view.
19284 Pannini projection.
19286 Format specific options:
19289 Set output pannini parameter.
19292 Set input pannini parameter.
19296 Cylindrical projection.
19298 Format specific options:
19303 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19305 If diagonal field of view is set it overrides horizontal and vertical field of view.
19310 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19312 If diagonal field of view is set it overrides horizontal and vertical field of view.
19316 Perspective projection. @i{(output only)}
19318 Format specific options:
19321 Set perspective parameter.
19325 Tetrahedron projection.
19328 Truncated square pyramid projection.
19332 Half equirectangular projection.
19337 Format specific options:
19342 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19344 If diagonal field of view is set it overrides horizontal and vertical field of view.
19349 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19351 If diagonal field of view is set it overrides horizontal and vertical field of view.
19355 Orthographic format.
19357 Format specific options:
19362 Set output horizontal/vertical/diagonal field of view. Values in degrees.
19364 If diagonal field of view is set it overrides horizontal and vertical field of view.
19369 Set input horizontal/vertical/diagonal field of view. Values in degrees.
19371 If diagonal field of view is set it overrides horizontal and vertical field of view.
19376 Set interpolation method.@*
19377 @i{Note: more complex interpolation methods require much more memory to run.}
19387 Bilinear interpolation.
19389 Lagrange9 interpolation.
19392 Bicubic interpolation.
19395 Lanczos interpolation.
19398 Spline16 interpolation.
19401 Gaussian interpolation.
19404 Default value is @b{@samp{line}}.
19408 Set the output video resolution.
19410 Default resolution depends on formats.
19414 Set the input/output stereo format.
19425 Default value is @b{@samp{2d}} for input and output format.
19430 Set rotation for the output video. Values in degrees.
19433 Set rotation order for the output video. Choose one item for each position.
19444 Default value is @b{@samp{ypr}}.
19449 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
19453 Set if input video is flipped horizontally/vertically. Boolean values.
19456 Set if input video is transposed. Boolean value, by default disabled.
19459 Set if output video needs to be transposed. Boolean value, by default disabled.
19462 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
19465 @subsection Examples
19469 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
19471 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
19474 Extract back view of Equi-Angular Cubemap:
19476 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
19479 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
19481 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
19485 @subsection Commands
19487 This filter supports subset of above options as @ref{commands}.
19489 @section vaguedenoiser
19491 Apply a wavelet based denoiser.
19493 It transforms each frame from the video input into the wavelet domain,
19494 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
19495 the obtained coefficients. It does an inverse wavelet transform after.
19496 Due to wavelet properties, it should give a nice smoothed result, and
19497 reduced noise, without blurring picture features.
19499 This filter accepts the following options:
19503 The filtering strength. The higher, the more filtered the video will be.
19504 Hard thresholding can use a higher threshold than soft thresholding
19505 before the video looks overfiltered. Default value is 2.
19508 The filtering method the filter will use.
19510 It accepts the following values:
19513 All values under the threshold will be zeroed.
19516 All values under the threshold will be zeroed. All values above will be
19517 reduced by the threshold.
19520 Scales or nullifies coefficients - intermediary between (more) soft and
19521 (less) hard thresholding.
19524 Default is garrote.
19527 Number of times, the wavelet will decompose the picture. Picture can't
19528 be decomposed beyond a particular point (typically, 8 for a 640x480
19529 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
19532 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
19535 A list of the planes to process. By default all planes are processed.
19538 The threshold type the filter will use.
19540 It accepts the following values:
19543 Threshold used is same for all decompositions.
19546 Threshold used depends also on each decomposition coefficients.
19549 Default is universal.
19552 @section vectorscope
19554 Display 2 color component values in the two dimensional graph (which is called
19557 This filter accepts the following options:
19561 Set vectorscope mode.
19563 It accepts the following values:
19567 Gray values are displayed on graph, higher brightness means more pixels have
19568 same component color value on location in graph. This is the default mode.
19571 Gray values are displayed on graph. Surrounding pixels values which are not
19572 present in video frame are drawn in gradient of 2 color components which are
19573 set by option @code{x} and @code{y}. The 3rd color component is static.
19576 Actual color components values present in video frame are displayed on graph.
19579 Similar as color2 but higher frequency of same values @code{x} and @code{y}
19580 on graph increases value of another color component, which is luminance by
19581 default values of @code{x} and @code{y}.
19584 Actual colors present in video frame are displayed on graph. If two different
19585 colors map to same position on graph then color with higher value of component
19586 not present in graph is picked.
19589 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
19590 component picked from radial gradient.
19594 Set which color component will be represented on X-axis. Default is @code{1}.
19597 Set which color component will be represented on Y-axis. Default is @code{2}.
19600 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
19601 of color component which represents frequency of (X, Y) location in graph.
19606 No envelope, this is default.
19609 Instant envelope, even darkest single pixel will be clearly highlighted.
19612 Hold maximum and minimum values presented in graph over time. This way you
19613 can still spot out of range values without constantly looking at vectorscope.
19616 Peak and instant envelope combined together.
19620 Set what kind of graticule to draw.
19629 Set graticule opacity.
19632 Set graticule flags.
19636 Draw graticule for white point.
19639 Draw graticule for black point.
19642 Draw color points short names.
19646 Set background opacity.
19648 @item lthreshold, l
19649 Set low threshold for color component not represented on X or Y axis.
19650 Values lower than this value will be ignored. Default is 0.
19651 Note this value is multiplied with actual max possible value one pixel component
19652 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
19655 @item hthreshold, h
19656 Set high threshold for color component not represented on X or Y axis.
19657 Values higher than this value will be ignored. Default is 1.
19658 Note this value is multiplied with actual max possible value one pixel component
19659 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
19660 is 0.9 * 255 = 230.
19662 @item colorspace, c
19663 Set what kind of colorspace to use when drawing graticule.
19673 Set color tint for gray/tint vectorscope mode. By default both options are zero.
19674 This means no tint, and output will remain gray.
19677 @anchor{vidstabdetect}
19678 @section vidstabdetect
19680 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
19681 @ref{vidstabtransform} for pass 2.
19683 This filter generates a file with relative translation and rotation
19684 transform information about subsequent frames, which is then used by
19685 the @ref{vidstabtransform} filter.
19687 To enable compilation of this filter you need to configure FFmpeg with
19688 @code{--enable-libvidstab}.
19690 This filter accepts the following options:
19694 Set the path to the file used to write the transforms information.
19695 Default value is @file{transforms.trf}.
19698 Set how shaky the video is and how quick the camera is. It accepts an
19699 integer in the range 1-10, a value of 1 means little shakiness, a
19700 value of 10 means strong shakiness. Default value is 5.
19703 Set the accuracy of the detection process. It must be a value in the
19704 range 1-15. A value of 1 means low accuracy, a value of 15 means high
19705 accuracy. Default value is 15.
19708 Set stepsize of the search process. The region around minimum is
19709 scanned with 1 pixel resolution. Default value is 6.
19712 Set minimum contrast. Below this value a local measurement field is
19713 discarded. Must be a floating point value in the range 0-1. Default
19717 Set reference frame number for tripod mode.
19719 If enabled, the motion of the frames is compared to a reference frame
19720 in the filtered stream, identified by the specified number. The idea
19721 is to compensate all movements in a more-or-less static scene and keep
19722 the camera view absolutely still.
19724 If set to 0, it is disabled. The frames are counted starting from 1.
19727 Show fields and transforms in the resulting frames. It accepts an
19728 integer in the range 0-2. Default value is 0, which disables any
19732 @subsection Examples
19736 Use default values:
19742 Analyze strongly shaky movie and put the results in file
19743 @file{mytransforms.trf}:
19745 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
19749 Visualize the result of internal transformations in the resulting
19752 vidstabdetect=show=1
19756 Analyze a video with medium shakiness using @command{ffmpeg}:
19758 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
19762 @anchor{vidstabtransform}
19763 @section vidstabtransform
19765 Video stabilization/deshaking: pass 2 of 2,
19766 see @ref{vidstabdetect} for pass 1.
19768 Read a file with transform information for each frame and
19769 apply/compensate them. Together with the @ref{vidstabdetect}
19770 filter this can be used to deshake videos. See also
19771 @url{http://public.hronopik.de/vid.stab}. It is important to also use
19772 the @ref{unsharp} filter, see below.
19774 To enable compilation of this filter you need to configure FFmpeg with
19775 @code{--enable-libvidstab}.
19777 @subsection Options
19781 Set path to the file used to read the transforms. Default value is
19782 @file{transforms.trf}.
19785 Set the number of frames (value*2 + 1) used for lowpass filtering the
19786 camera movements. Default value is 10.
19788 For example a number of 10 means that 21 frames are used (10 in the
19789 past and 10 in the future) to smoothen the motion in the video. A
19790 larger value leads to a smoother video, but limits the acceleration of
19791 the camera (pan/tilt movements). 0 is a special case where a static
19792 camera is simulated.
19795 Set the camera path optimization algorithm.
19797 Accepted values are:
19800 gaussian kernel low-pass filter on camera motion (default)
19802 averaging on transformations
19806 Set maximal number of pixels to translate frames. Default value is -1,
19810 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
19811 value is -1, meaning no limit.
19814 Specify how to deal with borders that may be visible due to movement
19817 Available values are:
19820 keep image information from previous frame (default)
19822 fill the border black
19826 Invert transforms if set to 1. Default value is 0.
19829 Consider transforms as relative to previous frame if set to 1,
19830 absolute if set to 0. Default value is 0.
19833 Set percentage to zoom. A positive value will result in a zoom-in
19834 effect, a negative value in a zoom-out effect. Default value is 0 (no
19838 Set optimal zooming to avoid borders.
19840 Accepted values are:
19845 optimal static zoom value is determined (only very strong movements
19846 will lead to visible borders) (default)
19848 optimal adaptive zoom value is determined (no borders will be
19849 visible), see @option{zoomspeed}
19852 Note that the value given at zoom is added to the one calculated here.
19855 Set percent to zoom maximally each frame (enabled when
19856 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
19860 Specify type of interpolation.
19862 Available values are:
19867 linear only horizontal
19869 linear in both directions (default)
19871 cubic in both directions (slow)
19875 Enable virtual tripod mode if set to 1, which is equivalent to
19876 @code{relative=0:smoothing=0}. Default value is 0.
19878 Use also @code{tripod} option of @ref{vidstabdetect}.
19881 Increase log verbosity if set to 1. Also the detected global motions
19882 are written to the temporary file @file{global_motions.trf}. Default
19886 @subsection Examples
19890 Use @command{ffmpeg} for a typical stabilization with default values:
19892 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
19895 Note the use of the @ref{unsharp} filter which is always recommended.
19898 Zoom in a bit more and load transform data from a given file:
19900 vidstabtransform=zoom=5:input="mytransforms.trf"
19904 Smoothen the video even more:
19906 vidstabtransform=smoothing=30
19912 Flip the input video vertically.
19914 For example, to vertically flip a video with @command{ffmpeg}:
19916 ffmpeg -i in.avi -vf "vflip" out.avi
19921 Detect variable frame rate video.
19923 This filter tries to detect if the input is variable or constant frame rate.
19925 At end it will output number of frames detected as having variable delta pts,
19926 and ones with constant delta pts.
19927 If there was frames with variable delta, than it will also show min, max and
19928 average delta encountered.
19932 Boost or alter saturation.
19934 The filter accepts the following options:
19937 Set strength of boost if positive value or strength of alter if negative value.
19938 Default is 0. Allowed range is from -2 to 2.
19941 Set the red balance. Default is 1. Allowed range is from -10 to 10.
19944 Set the green balance. Default is 1. Allowed range is from -10 to 10.
19947 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
19950 Set the red luma coefficient.
19953 Set the green luma coefficient.
19956 Set the blue luma coefficient.
19959 If @code{intensity} is negative and this is set to 1, colors will change,
19960 otherwise colors will be less saturated, more towards gray.
19963 @subsection Commands
19965 This filter supports the all above options as @ref{commands}.
19970 Make or reverse a natural vignetting effect.
19972 The filter accepts the following options:
19976 Set lens angle expression as a number of radians.
19978 The value is clipped in the @code{[0,PI/2]} range.
19980 Default value: @code{"PI/5"}
19984 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
19988 Set forward/backward mode.
19990 Available modes are:
19993 The larger the distance from the central point, the darker the image becomes.
19996 The larger the distance from the central point, the brighter the image becomes.
19997 This can be used to reverse a vignette effect, though there is no automatic
19998 detection to extract the lens @option{angle} and other settings (yet). It can
19999 also be used to create a burning effect.
20002 Default value is @samp{forward}.
20005 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
20007 It accepts the following values:
20010 Evaluate expressions only once during the filter initialization.
20013 Evaluate expressions for each incoming frame. This is way slower than the
20014 @samp{init} mode since it requires all the scalers to be re-computed, but it
20015 allows advanced dynamic expressions.
20018 Default value is @samp{init}.
20021 Set dithering to reduce the circular banding effects. Default is @code{1}
20025 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
20026 Setting this value to the SAR of the input will make a rectangular vignetting
20027 following the dimensions of the video.
20029 Default is @code{1/1}.
20032 @subsection Expressions
20034 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
20035 following parameters.
20040 input width and height
20043 the number of input frame, starting from 0
20046 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
20047 @var{TB} units, NAN if undefined
20050 frame rate of the input video, NAN if the input frame rate is unknown
20053 the PTS (Presentation TimeStamp) of the filtered video frame,
20054 expressed in seconds, NAN if undefined
20057 time base of the input video
20061 @subsection Examples
20065 Apply simple strong vignetting effect:
20071 Make a flickering vignetting:
20073 vignette='PI/4+random(1)*PI/50':eval=frame
20078 @section vmafmotion
20080 Obtain the average VMAF motion score of a video.
20081 It is one of the component metrics of VMAF.
20083 The obtained average motion score is printed through the logging system.
20085 The filter accepts the following options:
20089 If specified, the filter will use the named file to save the motion score of
20090 each frame with respect to the previous frame.
20091 When filename equals "-" the data is sent to standard output.
20096 ffmpeg -i ref.mpg -vf vmafmotion -f null -
20100 Stack input videos vertically.
20102 All streams must be of same pixel format and of same width.
20104 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
20105 to create same output.
20107 The filter accepts the following options:
20111 Set number of input streams. Default is 2.
20114 If set to 1, force the output to terminate when the shortest input
20115 terminates. Default value is 0.
20120 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
20121 Deinterlacing Filter").
20123 Based on the process described by Martin Weston for BBC R&D, and
20124 implemented based on the de-interlace algorithm written by Jim
20125 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
20126 uses filter coefficients calculated by BBC R&D.
20128 This filter uses field-dominance information in frame to decide which
20129 of each pair of fields to place first in the output.
20130 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
20132 There are two sets of filter coefficients, so called "simple"
20133 and "complex". Which set of filter coefficients is used can
20134 be set by passing an optional parameter:
20138 Set the interlacing filter coefficients. Accepts one of the following values:
20142 Simple filter coefficient set.
20144 More-complex filter coefficient set.
20146 Default value is @samp{complex}.
20149 Specify which frames to deinterlace. Accepts one of the following values:
20153 Deinterlace all frames,
20155 Only deinterlace frames marked as interlaced.
20158 Default value is @samp{all}.
20162 Video waveform monitor.
20164 The waveform monitor plots color component intensity. By default luminance
20165 only. Each column of the waveform corresponds to a column of pixels in the
20168 It accepts the following options:
20172 Can be either @code{row}, or @code{column}. Default is @code{column}.
20173 In row mode, the graph on the left side represents color component value 0 and
20174 the right side represents value = 255. In column mode, the top side represents
20175 color component value = 0 and bottom side represents value = 255.
20178 Set intensity. Smaller values are useful to find out how many values of the same
20179 luminance are distributed across input rows/columns.
20180 Default value is @code{0.04}. Allowed range is [0, 1].
20183 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
20184 In mirrored mode, higher values will be represented on the left
20185 side for @code{row} mode and at the top for @code{column} mode. Default is
20186 @code{1} (mirrored).
20190 It accepts the following values:
20193 Presents information identical to that in the @code{parade}, except
20194 that the graphs representing color components are superimposed directly
20197 This display mode makes it easier to spot relative differences or similarities
20198 in overlapping areas of the color components that are supposed to be identical,
20199 such as neutral whites, grays, or blacks.
20202 Display separate graph for the color components side by side in
20203 @code{row} mode or one below the other in @code{column} mode.
20206 Display separate graph for the color components side by side in
20207 @code{column} mode or one below the other in @code{row} mode.
20209 Using this display mode makes it easy to spot color casts in the highlights
20210 and shadows of an image, by comparing the contours of the top and the bottom
20211 graphs of each waveform. Since whites, grays, and blacks are characterized
20212 by exactly equal amounts of red, green, and blue, neutral areas of the picture
20213 should display three waveforms of roughly equal width/height. If not, the
20214 correction is easy to perform by making level adjustments the three waveforms.
20216 Default is @code{stack}.
20218 @item components, c
20219 Set which color components to display. Default is 1, which means only luminance
20220 or red color component if input is in RGB colorspace. If is set for example to
20221 7 it will display all 3 (if) available color components.
20226 No envelope, this is default.
20229 Instant envelope, minimum and maximum values presented in graph will be easily
20230 visible even with small @code{step} value.
20233 Hold minimum and maximum values presented in graph across time. This way you
20234 can still spot out of range values without constantly looking at waveforms.
20237 Peak and instant envelope combined together.
20243 No filtering, this is default.
20246 Luma and chroma combined together.
20249 Similar as above, but shows difference between blue and red chroma.
20252 Similar as above, but use different colors.
20255 Similar as above, but again with different colors.
20258 Displays only chroma.
20261 Displays actual color value on waveform.
20264 Similar as above, but with luma showing frequency of chroma values.
20268 Set which graticule to display.
20272 Do not display graticule.
20275 Display green graticule showing legal broadcast ranges.
20278 Display orange graticule showing legal broadcast ranges.
20281 Display invert graticule showing legal broadcast ranges.
20285 Set graticule opacity.
20288 Set graticule flags.
20292 Draw numbers above lines. By default enabled.
20295 Draw dots instead of lines.
20299 Set scale used for displaying graticule.
20306 Default is digital.
20309 Set background opacity.
20313 Set tint for output.
20314 Only used with lowpass filter and when display is not overlay and input
20315 pixel formats are not RGB.
20318 @section weave, doubleweave
20320 The @code{weave} takes a field-based video input and join
20321 each two sequential fields into single frame, producing a new double
20322 height clip with half the frame rate and half the frame count.
20324 The @code{doubleweave} works same as @code{weave} but without
20325 halving frame rate and frame count.
20327 It accepts the following option:
20331 Set first field. Available values are:
20335 Set the frame as top-field-first.
20338 Set the frame as bottom-field-first.
20342 @subsection Examples
20346 Interlace video using @ref{select} and @ref{separatefields} filter:
20348 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
20353 Apply the xBR high-quality magnification filter which is designed for pixel
20354 art. It follows a set of edge-detection rules, see
20355 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
20357 It accepts the following option:
20361 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
20362 @code{3xBR} and @code{4} for @code{4xBR}.
20363 Default is @code{3}.
20368 Apply cross fade from one input video stream to another input video stream.
20369 The cross fade is applied for specified duration.
20371 The filter accepts the following options:
20375 Set one of available transition effects:
20415 Default transition effect is fade.
20418 Set cross fade duration in seconds.
20419 Default duration is 1 second.
20422 Set cross fade start relative to first input stream in seconds.
20423 Default offset is 0.
20426 Set expression for custom transition effect.
20428 The expressions can use the following variables and functions:
20433 The coordinates of the current sample.
20437 The width and height of the image.
20440 Progress of transition effect.
20443 Currently processed plane.
20446 Return value of first input at current location and plane.
20449 Return value of second input at current location and plane.
20455 Return the value of the pixel at location (@var{x},@var{y}) of the
20456 first/second/third/fourth component of first input.
20462 Return the value of the pixel at location (@var{x},@var{y}) of the
20463 first/second/third/fourth component of second input.
20467 @subsection Examples
20471 Cross fade from one input video to another input video, with fade transition and duration of transition
20472 of 2 seconds starting at offset of 5 seconds:
20474 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
20479 Pick median pixels from several input videos.
20481 The filter accepts the following options:
20485 Set number of inputs.
20486 Default is 3. Allowed range is from 3 to 255.
20487 If number of inputs is even number, than result will be mean value between two median values.
20490 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
20493 Set median percentile. Default value is @code{0.5}.
20494 Default value of @code{0.5} will pick always median values, while @code{0} will pick
20495 minimum values, and @code{1} maximum values.
20499 Stack video inputs into custom layout.
20501 All streams must be of same pixel format.
20503 The filter accepts the following options:
20507 Set number of input streams. Default is 2.
20510 Specify layout of inputs.
20511 This option requires the desired layout configuration to be explicitly set by the user.
20512 This sets position of each video input in output. Each input
20513 is separated by '|'.
20514 The first number represents the column, and the second number represents the row.
20515 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
20516 where X is video input from which to take width or height.
20517 Multiple values can be used when separated by '+'. In such
20518 case values are summed together.
20520 Note that if inputs are of different sizes gaps may appear, as not all of
20521 the output video frame will be filled. Similarly, videos can overlap each
20522 other if their position doesn't leave enough space for the full frame of
20525 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
20526 a layout must be set by the user.
20529 If set to 1, force the output to terminate when the shortest input
20530 terminates. Default value is 0.
20533 If set to valid color, all unused pixels will be filled with that color.
20534 By default fill is set to none, so it is disabled.
20537 @subsection Examples
20541 Display 4 inputs into 2x2 grid.
20545 input1(0, 0) | input3(w0, 0)
20546 input2(0, h0) | input4(w0, h0)
20550 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
20553 Note that if inputs are of different sizes, gaps or overlaps may occur.
20556 Display 4 inputs into 1x4 grid.
20563 input4(0, h0+h1+h2)
20567 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
20570 Note that if inputs are of different widths, unused space will appear.
20573 Display 9 inputs into 3x3 grid.
20577 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
20578 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
20579 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
20583 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
20586 Note that if inputs are of different sizes, gaps or overlaps may occur.
20589 Display 16 inputs into 4x4 grid.
20593 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
20594 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
20595 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
20596 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
20600 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|
20601 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
20604 Note that if inputs are of different sizes, gaps or overlaps may occur.
20611 Deinterlace the input video ("yadif" means "yet another deinterlacing
20614 It accepts the following parameters:
20620 The interlacing mode to adopt. It accepts one of the following values:
20623 @item 0, send_frame
20624 Output one frame for each frame.
20625 @item 1, send_field
20626 Output one frame for each field.
20627 @item 2, send_frame_nospatial
20628 Like @code{send_frame}, but it skips the spatial interlacing check.
20629 @item 3, send_field_nospatial
20630 Like @code{send_field}, but it skips the spatial interlacing check.
20633 The default value is @code{send_frame}.
20636 The picture field parity assumed for the input interlaced video. It accepts one
20637 of the following values:
20641 Assume the top field is first.
20643 Assume the bottom field is first.
20645 Enable automatic detection of field parity.
20648 The default value is @code{auto}.
20649 If the interlacing is unknown or the decoder does not export this information,
20650 top field first will be assumed.
20653 Specify which frames to deinterlace. Accepts one of the following
20658 Deinterlace all frames.
20659 @item 1, interlaced
20660 Only deinterlace frames marked as interlaced.
20663 The default value is @code{all}.
20666 @section yadif_cuda
20668 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
20669 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
20672 It accepts the following parameters:
20678 The interlacing mode to adopt. It accepts one of the following values:
20681 @item 0, send_frame
20682 Output one frame for each frame.
20683 @item 1, send_field
20684 Output one frame for each field.
20685 @item 2, send_frame_nospatial
20686 Like @code{send_frame}, but it skips the spatial interlacing check.
20687 @item 3, send_field_nospatial
20688 Like @code{send_field}, but it skips the spatial interlacing check.
20691 The default value is @code{send_frame}.
20694 The picture field parity assumed for the input interlaced video. It accepts one
20695 of the following values:
20699 Assume the top field is first.
20701 Assume the bottom field is first.
20703 Enable automatic detection of field parity.
20706 The default value is @code{auto}.
20707 If the interlacing is unknown or the decoder does not export this information,
20708 top field first will be assumed.
20711 Specify which frames to deinterlace. Accepts one of the following
20716 Deinterlace all frames.
20717 @item 1, interlaced
20718 Only deinterlace frames marked as interlaced.
20721 The default value is @code{all}.
20726 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
20727 The algorithm is described in
20728 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
20730 It accepts the following parameters:
20734 Set the window radius. Default value is 3.
20737 Set which planes to filter. Default is only the first plane.
20740 Set blur strength. Default value is 128.
20743 @subsection Commands
20744 This filter supports same @ref{commands} as options.
20748 Apply Zoom & Pan effect.
20750 This filter accepts the following options:
20754 Set the zoom expression. Range is 1-10. Default is 1.
20758 Set the x and y expression. Default is 0.
20761 Set the duration expression in number of frames.
20762 This sets for how many number of frames effect will last for
20763 single input image.
20766 Set the output image size, default is 'hd720'.
20769 Set the output frame rate, default is '25'.
20772 Each expression can contain the following constants:
20791 Output frame count.
20794 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
20796 @item out_time, time, ot
20797 The output timestamp expressed in seconds.
20801 Last calculated 'x' and 'y' position from 'x' and 'y' expression
20802 for current input frame.
20806 'x' and 'y' of last output frame of previous input frame or 0 when there was
20807 not yet such frame (first input frame).
20810 Last calculated zoom from 'z' expression for current input frame.
20813 Last calculated zoom of last output frame of previous input frame.
20816 Number of output frames for current input frame. Calculated from 'd' expression
20817 for each input frame.
20820 number of output frames created for previous input frame
20823 Rational number: input width / input height
20826 sample aspect ratio
20829 display aspect ratio
20833 @subsection Examples
20837 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
20839 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
20843 Zoom in up to 1.5x and pan always at center of picture:
20845 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20849 Same as above but without pausing:
20851 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20855 Zoom in 2x into center of picture only for the first second of the input video:
20857 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
20864 Scale (resize) the input video, using the z.lib library:
20865 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
20866 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
20868 The zscale filter forces the output display aspect ratio to be the same
20869 as the input, by changing the output sample aspect ratio.
20871 If the input image format is different from the format requested by
20872 the next filter, the zscale filter will convert the input to the
20875 @subsection Options
20876 The filter accepts the following options.
20881 Set the output video dimension expression. Default value is the input
20884 If the @var{width} or @var{w} value is 0, the input width is used for
20885 the output. If the @var{height} or @var{h} value is 0, the input height
20886 is used for the output.
20888 If one and only one of the values is -n with n >= 1, the zscale filter
20889 will use a value that maintains the aspect ratio of the input image,
20890 calculated from the other specified dimension. After that it will,
20891 however, make sure that the calculated dimension is divisible by n and
20892 adjust the value if necessary.
20894 If both values are -n with n >= 1, the behavior will be identical to
20895 both values being set to 0 as previously detailed.
20897 See below for the list of accepted constants for use in the dimension
20901 Set the video size. For the syntax of this option, check the
20902 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20905 Set the dither type.
20907 Possible values are:
20912 @item error_diffusion
20918 Set the resize filter type.
20920 Possible values are:
20930 Default is bilinear.
20933 Set the color range.
20935 Possible values are:
20942 Default is same as input.
20945 Set the color primaries.
20947 Possible values are:
20957 Default is same as input.
20960 Set the transfer characteristics.
20962 Possible values are:
20976 Default is same as input.
20979 Set the colorspace matrix.
20981 Possible value are:
20992 Default is same as input.
20995 Set the input color range.
20997 Possible values are:
21004 Default is same as input.
21006 @item primariesin, pin
21007 Set the input color primaries.
21009 Possible values are:
21019 Default is same as input.
21021 @item transferin, tin
21022 Set the input transfer characteristics.
21024 Possible values are:
21035 Default is same as input.
21037 @item matrixin, min
21038 Set the input colorspace matrix.
21040 Possible value are:
21052 Set the output chroma location.
21054 Possible values are:
21065 @item chromalin, cin
21066 Set the input chroma location.
21068 Possible values are:
21080 Set the nominal peak luminance.
21083 The values of the @option{w} and @option{h} options are expressions
21084 containing the following constants:
21089 The input width and height
21093 These are the same as @var{in_w} and @var{in_h}.
21097 The output (scaled) width and height
21101 These are the same as @var{out_w} and @var{out_h}
21104 The same as @var{iw} / @var{ih}
21107 input sample aspect ratio
21110 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
21114 horizontal and vertical input chroma subsample values. For example for the
21115 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
21119 horizontal and vertical output chroma subsample values. For example for the
21120 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
21123 @subsection Commands
21125 This filter supports the following commands:
21129 Set the output video dimension expression.
21130 The command accepts the same syntax of the corresponding option.
21132 If the specified expression is not valid, it is kept at its current
21136 @c man end VIDEO FILTERS
21138 @chapter OpenCL Video Filters
21139 @c man begin OPENCL VIDEO FILTERS
21141 Below is a description of the currently available OpenCL video filters.
21143 To enable compilation of these filters you need to configure FFmpeg with
21144 @code{--enable-opencl}.
21146 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
21149 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
21150 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
21151 given device parameters.
21153 @item -filter_hw_device @var{name}
21154 Pass the hardware device called @var{name} to all filters in any filter graph.
21158 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
21162 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
21164 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
21168 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.
21170 @section avgblur_opencl
21172 Apply average blur filter.
21174 The filter accepts the following options:
21178 Set horizontal radius size.
21179 Range is @code{[1, 1024]} and default value is @code{1}.
21182 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21185 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
21188 @subsection Example
21192 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.
21194 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
21198 @section boxblur_opencl
21200 Apply a boxblur algorithm to the input video.
21202 It accepts the following parameters:
21206 @item luma_radius, lr
21207 @item luma_power, lp
21208 @item chroma_radius, cr
21209 @item chroma_power, cp
21210 @item alpha_radius, ar
21211 @item alpha_power, ap
21215 A description of the accepted options follows.
21218 @item luma_radius, lr
21219 @item chroma_radius, cr
21220 @item alpha_radius, ar
21221 Set an expression for the box radius in pixels used for blurring the
21222 corresponding input plane.
21224 The radius value must be a non-negative number, and must not be
21225 greater than the value of the expression @code{min(w,h)/2} for the
21226 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
21229 Default value for @option{luma_radius} is "2". If not specified,
21230 @option{chroma_radius} and @option{alpha_radius} default to the
21231 corresponding value set for @option{luma_radius}.
21233 The expressions can contain the following constants:
21237 The input width and height in pixels.
21241 The input chroma image width and height in pixels.
21245 The horizontal and vertical chroma subsample values. For example, for the
21246 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
21249 @item luma_power, lp
21250 @item chroma_power, cp
21251 @item alpha_power, ap
21252 Specify how many times the boxblur filter is applied to the
21253 corresponding plane.
21255 Default value for @option{luma_power} is 2. If not specified,
21256 @option{chroma_power} and @option{alpha_power} default to the
21257 corresponding value set for @option{luma_power}.
21259 A value of 0 will disable the effect.
21262 @subsection Examples
21264 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.
21268 Apply a boxblur filter with the luma, chroma, and alpha radius
21269 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.
21271 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
21272 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
21276 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.
21278 For the luma plane, a 2x2 box radius will be run once.
21280 For the chroma plane, a 4x4 box radius will be run 5 times.
21282 For the alpha plane, a 3x3 box radius will be run 7 times.
21284 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
21288 @section colorkey_opencl
21289 RGB colorspace color keying.
21291 The filter accepts the following options:
21295 The color which will be replaced with transparency.
21298 Similarity percentage with the key color.
21300 0.01 matches only the exact key color, while 1.0 matches everything.
21305 0.0 makes pixels either fully transparent, or not transparent at all.
21307 Higher values result in semi-transparent pixels, with a higher transparency
21308 the more similar the pixels color is to the key color.
21311 @subsection Examples
21315 Make every semi-green pixel in the input transparent with some slight blending:
21317 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
21321 @section convolution_opencl
21323 Apply convolution of 3x3, 5x5, 7x7 matrix.
21325 The filter accepts the following options:
21332 Set matrix for each plane.
21333 Matrix is sequence of 9, 25 or 49 signed numbers.
21334 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
21340 Set multiplier for calculated value for each plane.
21341 If unset or 0, it will be sum of all matrix elements.
21342 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
21348 Set bias for each plane. This value is added to the result of the multiplication.
21349 Useful for making the overall image brighter or darker.
21350 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
21354 @subsection Examples
21360 -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
21366 -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
21370 Apply edge enhance:
21372 -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
21378 -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
21382 Apply laplacian edge detector which includes diagonals:
21384 -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
21390 -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
21394 @section erosion_opencl
21396 Apply erosion effect to the video.
21398 This filter replaces the pixel by the local(3x3) minimum.
21400 It accepts the following options:
21407 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
21408 If @code{0}, plane will remain unchanged.
21411 Flag which specifies the pixel to refer to.
21412 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
21414 Flags to local 3x3 coordinates region centered on @code{x}:
21423 @subsection Example
21427 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.
21429 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
21433 @section deshake_opencl
21434 Feature-point based video stabilization filter.
21436 The filter accepts the following options:
21440 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
21443 Whether or not additional debug info should be displayed, both in the processed output and in the console.
21445 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
21447 Viewing point matches in the output video is only supported for RGB input.
21449 Defaults to @code{0}.
21451 @item adaptive_crop
21452 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
21454 Defaults to @code{1}.
21456 @item refine_features
21457 Whether or not feature points should be refined at a sub-pixel level.
21459 This can be turned off for a slight performance gain at the cost of precision.
21461 Defaults to @code{1}.
21463 @item smooth_strength
21464 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
21466 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
21468 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
21470 Defaults to @code{0.0}.
21472 @item smooth_window_multiplier
21473 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
21475 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
21477 Acceptable values range from @code{0.1} to @code{10.0}.
21479 Larger values increase the amount of motion data available for determining how to smooth the camera path,
21480 potentially improving smoothness, but also increase latency and memory usage.
21482 Defaults to @code{2.0}.
21486 @subsection Examples
21490 Stabilize a video with a fixed, medium smoothing strength:
21492 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
21496 Stabilize a video with debugging (both in console and in rendered video):
21498 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
21502 @section dilation_opencl
21504 Apply dilation effect to the video.
21506 This filter replaces the pixel by the local(3x3) maximum.
21508 It accepts the following options:
21515 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
21516 If @code{0}, plane will remain unchanged.
21519 Flag which specifies the pixel to refer to.
21520 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
21522 Flags to local 3x3 coordinates region centered on @code{x}:
21531 @subsection Example
21535 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.
21537 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
21541 @section nlmeans_opencl
21543 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
21545 @section overlay_opencl
21547 Overlay one video on top of another.
21549 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
21550 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
21552 The filter accepts the following options:
21557 Set the x coordinate of the overlaid video on the main video.
21558 Default value is @code{0}.
21561 Set the y coordinate of the overlaid video on the main video.
21562 Default value is @code{0}.
21566 @subsection Examples
21570 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
21572 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
21575 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
21577 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
21582 @section pad_opencl
21584 Add paddings to the input image, and place the original input at the
21585 provided @var{x}, @var{y} coordinates.
21587 It accepts the following options:
21592 Specify an expression for the size of the output image with the
21593 paddings added. If the value for @var{width} or @var{height} is 0, the
21594 corresponding input size is used for the output.
21596 The @var{width} expression can reference the value set by the
21597 @var{height} expression, and vice versa.
21599 The default value of @var{width} and @var{height} is 0.
21603 Specify the offsets to place the input image at within the padded area,
21604 with respect to the top/left border of the output image.
21606 The @var{x} expression can reference the value set by the @var{y}
21607 expression, and vice versa.
21609 The default value of @var{x} and @var{y} is 0.
21611 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
21612 so the input image is centered on the padded area.
21615 Specify the color of the padded area. For the syntax of this option,
21616 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
21617 manual,ffmpeg-utils}.
21620 Pad to an aspect instead to a resolution.
21623 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
21624 options are expressions containing the following constants:
21629 The input video width and height.
21633 These are the same as @var{in_w} and @var{in_h}.
21637 The output width and height (the size of the padded area), as
21638 specified by the @var{width} and @var{height} expressions.
21642 These are the same as @var{out_w} and @var{out_h}.
21646 The x and y offsets as specified by the @var{x} and @var{y}
21647 expressions, or NAN if not yet specified.
21650 same as @var{iw} / @var{ih}
21653 input sample aspect ratio
21656 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
21659 @section prewitt_opencl
21661 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
21663 The filter accepts the following option:
21667 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21670 Set value which will be multiplied with filtered result.
21671 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21674 Set value which will be added to filtered result.
21675 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21678 @subsection Example
21682 Apply the Prewitt operator with scale set to 2 and delta set to 10.
21684 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
21688 @anchor{program_opencl}
21689 @section program_opencl
21691 Filter video using an OpenCL program.
21696 OpenCL program source file.
21699 Kernel name in program.
21702 Number of inputs to the filter. Defaults to 1.
21705 Size of output frames. Defaults to the same as the first input.
21709 The @code{program_opencl} filter also supports the @ref{framesync} options.
21711 The program source file must contain a kernel function with the given name,
21712 which will be run once for each plane of the output. Each run on a plane
21713 gets enqueued as a separate 2D global NDRange with one work-item for each
21714 pixel to be generated. The global ID offset for each work-item is therefore
21715 the coordinates of a pixel in the destination image.
21717 The kernel function needs to take the following arguments:
21720 Destination image, @var{__write_only image2d_t}.
21722 This image will become the output; the kernel should write all of it.
21724 Frame index, @var{unsigned int}.
21726 This is a counter starting from zero and increasing by one for each frame.
21728 Source images, @var{__read_only image2d_t}.
21730 These are the most recent images on each input. The kernel may read from
21731 them to generate the output, but they can't be written to.
21738 Copy the input to the output (output must be the same size as the input).
21740 __kernel void copy(__write_only image2d_t destination,
21741 unsigned int index,
21742 __read_only image2d_t source)
21744 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
21746 int2 location = (int2)(get_global_id(0), get_global_id(1));
21748 float4 value = read_imagef(source, sampler, location);
21750 write_imagef(destination, location, value);
21755 Apply a simple transformation, rotating the input by an amount increasing
21756 with the index counter. Pixel values are linearly interpolated by the
21757 sampler, and the output need not have the same dimensions as the input.
21759 __kernel void rotate_image(__write_only image2d_t dst,
21760 unsigned int index,
21761 __read_only image2d_t src)
21763 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
21764 CLK_FILTER_LINEAR);
21766 float angle = (float)index / 100.0f;
21768 float2 dst_dim = convert_float2(get_image_dim(dst));
21769 float2 src_dim = convert_float2(get_image_dim(src));
21771 float2 dst_cen = dst_dim / 2.0f;
21772 float2 src_cen = src_dim / 2.0f;
21774 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
21776 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
21778 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
21779 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
21781 src_pos = src_pos * src_dim / dst_dim;
21783 float2 src_loc = src_pos + src_cen;
21785 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
21786 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
21787 write_imagef(dst, dst_loc, 0.5f);
21789 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
21794 Blend two inputs together, with the amount of each input used varying
21795 with the index counter.
21797 __kernel void blend_images(__write_only image2d_t dst,
21798 unsigned int index,
21799 __read_only image2d_t src1,
21800 __read_only image2d_t src2)
21802 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
21803 CLK_FILTER_LINEAR);
21805 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
21807 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
21808 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
21809 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
21811 float4 val1 = read_imagef(src1, sampler, src1_loc);
21812 float4 val2 = read_imagef(src2, sampler, src2_loc);
21814 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
21820 @section roberts_opencl
21821 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
21823 The filter accepts the following option:
21827 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21830 Set value which will be multiplied with filtered result.
21831 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21834 Set value which will be added to filtered result.
21835 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21838 @subsection Example
21842 Apply the Roberts cross operator with scale set to 2 and delta set to 10
21844 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
21848 @section sobel_opencl
21850 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
21852 The filter accepts the following option:
21856 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
21859 Set value which will be multiplied with filtered result.
21860 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
21863 Set value which will be added to filtered result.
21864 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
21867 @subsection Example
21871 Apply sobel operator with scale set to 2 and delta set to 10
21873 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
21877 @section tonemap_opencl
21879 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
21881 It accepts the following parameters:
21885 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
21888 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
21891 Apply desaturation for highlights that exceed this level of brightness. The
21892 higher the parameter, the more color information will be preserved. This
21893 setting helps prevent unnaturally blown-out colors for super-highlights, by
21894 (smoothly) turning into white instead. This makes images feel more natural,
21895 at the cost of reducing information about out-of-range colors.
21897 The default value is 0.5, and the algorithm here is a little different from
21898 the cpu version tonemap currently. A setting of 0.0 disables this option.
21901 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
21902 is used to detect whether the scene has changed or not. If the distance between
21903 the current frame average brightness and the current running average exceeds
21904 a threshold value, we would re-calculate scene average and peak brightness.
21905 The default value is 0.2.
21908 Specify the output pixel format.
21910 Currently supported formats are:
21917 Set the output color range.
21919 Possible values are:
21925 Default is same as input.
21928 Set the output color primaries.
21930 Possible values are:
21936 Default is same as input.
21939 Set the output transfer characteristics.
21941 Possible values are:
21950 Set the output colorspace matrix.
21952 Possible value are:
21958 Default is same as input.
21962 @subsection Example
21966 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
21968 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
21972 @section unsharp_opencl
21974 Sharpen or blur the input video.
21976 It accepts the following parameters:
21979 @item luma_msize_x, lx
21980 Set the luma matrix horizontal size.
21981 Range is @code{[1, 23]} and default value is @code{5}.
21983 @item luma_msize_y, ly
21984 Set the luma matrix vertical size.
21985 Range is @code{[1, 23]} and default value is @code{5}.
21987 @item luma_amount, la
21988 Set the luma effect strength.
21989 Range is @code{[-10, 10]} and default value is @code{1.0}.
21991 Negative values will blur the input video, while positive values will
21992 sharpen it, a value of zero will disable the effect.
21994 @item chroma_msize_x, cx
21995 Set the chroma matrix horizontal size.
21996 Range is @code{[1, 23]} and default value is @code{5}.
21998 @item chroma_msize_y, cy
21999 Set the chroma matrix vertical size.
22000 Range is @code{[1, 23]} and default value is @code{5}.
22002 @item chroma_amount, ca
22003 Set the chroma effect strength.
22004 Range is @code{[-10, 10]} and default value is @code{0.0}.
22006 Negative values will blur the input video, while positive values will
22007 sharpen it, a value of zero will disable the effect.
22011 All parameters are optional and default to the equivalent of the
22012 string '5:5:1.0:5:5:0.0'.
22014 @subsection Examples
22018 Apply strong luma sharpen effect:
22020 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
22024 Apply a strong blur of both luma and chroma parameters:
22026 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
22030 @section xfade_opencl
22032 Cross fade two videos with custom transition effect by using OpenCL.
22034 It accepts the following options:
22038 Set one of possible transition effects.
22042 Select custom transition effect, the actual transition description
22043 will be picked from source and kernel options.
22055 Default transition is fade.
22059 OpenCL program source file for custom transition.
22062 Set name of kernel to use for custom transition from program source file.
22065 Set duration of video transition.
22068 Set time of start of transition relative to first video.
22071 The program source file must contain a kernel function with the given name,
22072 which will be run once for each plane of the output. Each run on a plane
22073 gets enqueued as a separate 2D global NDRange with one work-item for each
22074 pixel to be generated. The global ID offset for each work-item is therefore
22075 the coordinates of a pixel in the destination image.
22077 The kernel function needs to take the following arguments:
22080 Destination image, @var{__write_only image2d_t}.
22082 This image will become the output; the kernel should write all of it.
22085 First Source image, @var{__read_only image2d_t}.
22086 Second Source image, @var{__read_only image2d_t}.
22088 These are the most recent images on each input. The kernel may read from
22089 them to generate the output, but they can't be written to.
22092 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
22099 Apply dots curtain transition effect:
22101 __kernel void blend_images(__write_only image2d_t dst,
22102 __read_only image2d_t src1,
22103 __read_only image2d_t src2,
22106 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22107 CLK_FILTER_LINEAR);
22108 int2 p = (int2)(get_global_id(0), get_global_id(1));
22109 float2 rp = (float2)(get_global_id(0), get_global_id(1));
22110 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
22113 float2 dots = (float2)(20.0, 20.0);
22114 float2 center = (float2)(0,0);
22117 float4 val1 = read_imagef(src1, sampler, p);
22118 float4 val2 = read_imagef(src2, sampler, p);
22119 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
22121 write_imagef(dst, p, next ? val1 : val2);
22127 @c man end OPENCL VIDEO FILTERS
22129 @chapter VAAPI Video Filters
22130 @c man begin VAAPI VIDEO FILTERS
22132 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
22134 To enable compilation of these filters you need to configure FFmpeg with
22135 @code{--enable-vaapi}.
22137 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}
22139 @section tonemap_vaapi
22141 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
22142 It maps the dynamic range of HDR10 content to the SDR content.
22143 It currently only accepts HDR10 as input.
22145 It accepts the following parameters:
22149 Specify the output pixel format.
22151 Currently supported formats are:
22160 Set the output color primaries.
22162 Default is same as input.
22165 Set the output transfer characteristics.
22170 Set the output colorspace matrix.
22172 Default is same as input.
22176 @subsection Example
22180 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
22182 tonemap_vaapi=format=p010:t=bt2020-10
22186 @c man end VAAPI VIDEO FILTERS
22188 @chapter Video Sources
22189 @c man begin VIDEO SOURCES
22191 Below is a description of the currently available video sources.
22195 Buffer video frames, and make them available to the filter chain.
22197 This source is mainly intended for a programmatic use, in particular
22198 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
22200 It accepts the following parameters:
22205 Specify the size (width and height) of the buffered video frames. For the
22206 syntax of this option, check the
22207 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22210 The input video width.
22213 The input video height.
22216 A string representing the pixel format of the buffered video frames.
22217 It may be a number corresponding to a pixel format, or a pixel format
22221 Specify the timebase assumed by the timestamps of the buffered frames.
22224 Specify the frame rate expected for the video stream.
22226 @item pixel_aspect, sar
22227 The sample (pixel) aspect ratio of the input video.
22230 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
22231 to the filtergraph description to specify swscale flags for automatically
22232 inserted scalers. See @ref{Filtergraph syntax}.
22234 @item hw_frames_ctx
22235 When using a hardware pixel format, this should be a reference to an
22236 AVHWFramesContext describing input frames.
22241 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
22244 will instruct the source to accept video frames with size 320x240 and
22245 with format "yuv410p", assuming 1/24 as the timestamps timebase and
22246 square pixels (1:1 sample aspect ratio).
22247 Since the pixel format with name "yuv410p" corresponds to the number 6
22248 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
22249 this example corresponds to:
22251 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
22254 Alternatively, the options can be specified as a flat string, but this
22255 syntax is deprecated:
22257 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
22261 Create a pattern generated by an elementary cellular automaton.
22263 The initial state of the cellular automaton can be defined through the
22264 @option{filename} and @option{pattern} options. If such options are
22265 not specified an initial state is created randomly.
22267 At each new frame a new row in the video is filled with the result of
22268 the cellular automaton next generation. The behavior when the whole
22269 frame is filled is defined by the @option{scroll} option.
22271 This source accepts the following options:
22275 Read the initial cellular automaton state, i.e. the starting row, from
22276 the specified file.
22277 In the file, each non-whitespace character is considered an alive
22278 cell, a newline will terminate the row, and further characters in the
22279 file will be ignored.
22282 Read the initial cellular automaton state, i.e. the starting row, from
22283 the specified string.
22285 Each non-whitespace character in the string is considered an alive
22286 cell, a newline will terminate the row, and further characters in the
22287 string will be ignored.
22290 Set the video rate, that is the number of frames generated per second.
22293 @item random_fill_ratio, ratio
22294 Set the random fill ratio for the initial cellular automaton row. It
22295 is a floating point number value ranging from 0 to 1, defaults to
22298 This option is ignored when a file or a pattern is specified.
22300 @item random_seed, seed
22301 Set the seed for filling randomly the initial row, must be an integer
22302 included between 0 and UINT32_MAX. If not specified, or if explicitly
22303 set to -1, the filter will try to use a good random seed on a best
22307 Set the cellular automaton rule, it is a number ranging from 0 to 255.
22308 Default value is 110.
22311 Set the size of the output video. For the syntax of this option, check the
22312 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22314 If @option{filename} or @option{pattern} is specified, the size is set
22315 by default to the width of the specified initial state row, and the
22316 height is set to @var{width} * PHI.
22318 If @option{size} is set, it must contain the width of the specified
22319 pattern string, and the specified pattern will be centered in the
22322 If a filename or a pattern string is not specified, the size value
22323 defaults to "320x518" (used for a randomly generated initial state).
22326 If set to 1, scroll the output upward when all the rows in the output
22327 have been already filled. If set to 0, the new generated row will be
22328 written over the top row just after the bottom row is filled.
22331 @item start_full, full
22332 If set to 1, completely fill the output with generated rows before
22333 outputting the first frame.
22334 This is the default behavior, for disabling set the value to 0.
22337 If set to 1, stitch the left and right row edges together.
22338 This is the default behavior, for disabling set the value to 0.
22341 @subsection Examples
22345 Read the initial state from @file{pattern}, and specify an output of
22348 cellauto=f=pattern:s=200x400
22352 Generate a random initial row with a width of 200 cells, with a fill
22355 cellauto=ratio=2/3:s=200x200
22359 Create a pattern generated by rule 18 starting by a single alive cell
22360 centered on an initial row with width 100:
22362 cellauto=p=@@:s=100x400:full=0:rule=18
22366 Specify a more elaborated initial pattern:
22368 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
22373 @anchor{coreimagesrc}
22374 @section coreimagesrc
22375 Video source generated on GPU using Apple's CoreImage API on OSX.
22377 This video source is a specialized version of the @ref{coreimage} video filter.
22378 Use a core image generator at the beginning of the applied filterchain to
22379 generate the content.
22381 The coreimagesrc video source accepts the following options:
22383 @item list_generators
22384 List all available generators along with all their respective options as well as
22385 possible minimum and maximum values along with the default values.
22387 list_generators=true
22391 Specify the size of the sourced video. For the syntax of this option, check the
22392 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22393 The default value is @code{320x240}.
22396 Specify the frame rate of the sourced video, as the number of frames
22397 generated per second. It has to be a string in the format
22398 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22399 number or a valid video frame rate abbreviation. The default value is
22403 Set the sample aspect ratio of the sourced video.
22406 Set the duration of the sourced video. See
22407 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22408 for the accepted syntax.
22410 If not specified, or the expressed duration is negative, the video is
22411 supposed to be generated forever.
22414 Additionally, all options of the @ref{coreimage} video filter are accepted.
22415 A complete filterchain can be used for further processing of the
22416 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
22417 and examples for details.
22419 @subsection Examples
22424 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
22425 given as complete and escaped command-line for Apple's standard bash shell:
22427 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
22429 This example is equivalent to the QRCode example of @ref{coreimage} without the
22430 need for a nullsrc video source.
22435 Generate several gradients.
22439 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
22440 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
22443 Set frame rate, expressed as number of frames per second. Default
22446 @item c0, c1, c2, c3, c4, c5, c6, c7
22447 Set 8 colors. Default values for colors is to pick random one.
22449 @item x0, y0, y0, y1
22450 Set gradient line source and destination points. If negative or out of range, random ones
22454 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
22457 Set seed for picking gradient line points.
22461 @section mandelbrot
22463 Generate a Mandelbrot set fractal, and progressively zoom towards the
22464 point specified with @var{start_x} and @var{start_y}.
22466 This source accepts the following options:
22471 Set the terminal pts value. Default value is 400.
22474 Set the terminal scale value.
22475 Must be a floating point value. Default value is 0.3.
22478 Set the inner coloring mode, that is the algorithm used to draw the
22479 Mandelbrot fractal internal region.
22481 It shall assume one of the following values:
22486 Show time until convergence.
22488 Set color based on point closest to the origin of the iterations.
22493 Default value is @var{mincol}.
22496 Set the bailout value. Default value is 10.0.
22499 Set the maximum of iterations performed by the rendering
22500 algorithm. Default value is 7189.
22503 Set outer coloring mode.
22504 It shall assume one of following values:
22506 @item iteration_count
22507 Set iteration count mode.
22508 @item normalized_iteration_count
22509 set normalized iteration count mode.
22511 Default value is @var{normalized_iteration_count}.
22514 Set frame rate, expressed as number of frames per second. Default
22518 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
22519 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
22522 Set the initial scale value. Default value is 3.0.
22525 Set the initial x position. Must be a floating point value between
22526 -100 and 100. Default value is -0.743643887037158704752191506114774.
22529 Set the initial y position. Must be a floating point value between
22530 -100 and 100. Default value is -0.131825904205311970493132056385139.
22535 Generate various test patterns, as generated by the MPlayer test filter.
22537 The size of the generated video is fixed, and is 256x256.
22538 This source is useful in particular for testing encoding features.
22540 This source accepts the following options:
22545 Specify the frame rate of the sourced video, as the number of frames
22546 generated per second. It has to be a string in the format
22547 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22548 number or a valid video frame rate abbreviation. The default value is
22552 Set the duration of the sourced video. See
22553 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22554 for the accepted syntax.
22556 If not specified, or the expressed duration is negative, the video is
22557 supposed to be generated forever.
22561 Set the number or the name of the test to perform. Supported tests are:
22575 @item max_frames, m
22576 Set the maximum number of frames generated for each test, default value is 30.
22580 Default value is "all", which will cycle through the list of all tests.
22585 mptestsrc=t=dc_luma
22588 will generate a "dc_luma" test pattern.
22590 @section frei0r_src
22592 Provide a frei0r source.
22594 To enable compilation of this filter you need to install the frei0r
22595 header and configure FFmpeg with @code{--enable-frei0r}.
22597 This source accepts the following parameters:
22602 The size of the video to generate. For the syntax of this option, check the
22603 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22606 The framerate of the generated video. It may be a string of the form
22607 @var{num}/@var{den} or a frame rate abbreviation.
22610 The name to the frei0r source to load. For more information regarding frei0r and
22611 how to set the parameters, read the @ref{frei0r} section in the video filters
22614 @item filter_params
22615 A '|'-separated list of parameters to pass to the frei0r source.
22619 For example, to generate a frei0r partik0l source with size 200x200
22620 and frame rate 10 which is overlaid on the overlay filter main input:
22622 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
22627 Generate a life pattern.
22629 This source is based on a generalization of John Conway's life game.
22631 The sourced input represents a life grid, each pixel represents a cell
22632 which can be in one of two possible states, alive or dead. Every cell
22633 interacts with its eight neighbours, which are the cells that are
22634 horizontally, vertically, or diagonally adjacent.
22636 At each interaction the grid evolves according to the adopted rule,
22637 which specifies the number of neighbor alive cells which will make a
22638 cell stay alive or born. The @option{rule} option allows one to specify
22641 This source accepts the following options:
22645 Set the file from which to read the initial grid state. In the file,
22646 each non-whitespace character is considered an alive cell, and newline
22647 is used to delimit the end of each row.
22649 If this option is not specified, the initial grid is generated
22653 Set the video rate, that is the number of frames generated per second.
22656 @item random_fill_ratio, ratio
22657 Set the random fill ratio for the initial random grid. It is a
22658 floating point number value ranging from 0 to 1, defaults to 1/PHI.
22659 It is ignored when a file is specified.
22661 @item random_seed, seed
22662 Set the seed for filling the initial random grid, must be an integer
22663 included between 0 and UINT32_MAX. If not specified, or if explicitly
22664 set to -1, the filter will try to use a good random seed on a best
22670 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
22671 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
22672 @var{NS} specifies the number of alive neighbor cells which make a
22673 live cell stay alive, and @var{NB} the number of alive neighbor cells
22674 which make a dead cell to become alive (i.e. to "born").
22675 "s" and "b" can be used in place of "S" and "B", respectively.
22677 Alternatively a rule can be specified by an 18-bits integer. The 9
22678 high order bits are used to encode the next cell state if it is alive
22679 for each number of neighbor alive cells, the low order bits specify
22680 the rule for "borning" new cells. Higher order bits encode for an
22681 higher number of neighbor cells.
22682 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
22683 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
22685 Default value is "S23/B3", which is the original Conway's game of life
22686 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
22687 cells, and will born a new cell if there are three alive cells around
22691 Set the size of the output video. For the syntax of this option, check the
22692 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22694 If @option{filename} is specified, the size is set by default to the
22695 same size of the input file. If @option{size} is set, it must contain
22696 the size specified in the input file, and the initial grid defined in
22697 that file is centered in the larger resulting area.
22699 If a filename is not specified, the size value defaults to "320x240"
22700 (used for a randomly generated initial grid).
22703 If set to 1, stitch the left and right grid edges together, and the
22704 top and bottom edges also. Defaults to 1.
22707 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
22708 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
22709 value from 0 to 255.
22712 Set the color of living (or new born) cells.
22715 Set the color of dead cells. If @option{mold} is set, this is the first color
22716 used to represent a dead cell.
22719 Set mold color, for definitely dead and moldy cells.
22721 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
22722 ffmpeg-utils manual,ffmpeg-utils}.
22725 @subsection Examples
22729 Read a grid from @file{pattern}, and center it on a grid of size
22732 life=f=pattern:s=300x300
22736 Generate a random grid of size 200x200, with a fill ratio of 2/3:
22738 life=ratio=2/3:s=200x200
22742 Specify a custom rule for evolving a randomly generated grid:
22748 Full example with slow death effect (mold) using @command{ffplay}:
22750 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
22757 @anchor{haldclutsrc}
22760 @anchor{pal100bars}
22761 @anchor{rgbtestsrc}
22763 @anchor{smptehdbars}
22766 @anchor{yuvtestsrc}
22767 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
22769 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
22771 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
22773 The @code{color} source provides an uniformly colored input.
22775 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
22776 @ref{haldclut} filter.
22778 The @code{nullsrc} source returns unprocessed video frames. It is
22779 mainly useful to be employed in analysis / debugging tools, or as the
22780 source for filters which ignore the input data.
22782 The @code{pal75bars} source generates a color bars pattern, based on
22783 EBU PAL recommendations with 75% color levels.
22785 The @code{pal100bars} source generates a color bars pattern, based on
22786 EBU PAL recommendations with 100% color levels.
22788 The @code{rgbtestsrc} source generates an RGB test pattern useful for
22789 detecting RGB vs BGR issues. You should see a red, green and blue
22790 stripe from top to bottom.
22792 The @code{smptebars} source generates a color bars pattern, based on
22793 the SMPTE Engineering Guideline EG 1-1990.
22795 The @code{smptehdbars} source generates a color bars pattern, based on
22796 the SMPTE RP 219-2002.
22798 The @code{testsrc} source generates a test video pattern, showing a
22799 color pattern, a scrolling gradient and a timestamp. This is mainly
22800 intended for testing purposes.
22802 The @code{testsrc2} source is similar to testsrc, but supports more
22803 pixel formats instead of just @code{rgb24}. This allows using it as an
22804 input for other tests without requiring a format conversion.
22806 The @code{yuvtestsrc} source generates an YUV test pattern. You should
22807 see a y, cb and cr stripe from top to bottom.
22809 The sources accept the following parameters:
22814 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
22815 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
22816 pixels to be used as identity matrix for 3D lookup tables. Each component is
22817 coded on a @code{1/(N*N)} scale.
22820 Specify the color of the source, only available in the @code{color}
22821 source. For the syntax of this option, check the
22822 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
22825 Specify the size of the sourced video. For the syntax of this option, check the
22826 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22827 The default value is @code{320x240}.
22829 This option is not available with the @code{allrgb}, @code{allyuv}, and
22830 @code{haldclutsrc} filters.
22833 Specify the frame rate of the sourced video, as the number of frames
22834 generated per second. It has to be a string in the format
22835 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
22836 number or a valid video frame rate abbreviation. The default value is
22840 Set the duration of the sourced video. See
22841 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
22842 for the accepted syntax.
22844 If not specified, or the expressed duration is negative, the video is
22845 supposed to be generated forever.
22848 Set the sample aspect ratio of the sourced video.
22851 Specify the alpha (opacity) of the background, only available in the
22852 @code{testsrc2} source. The value must be between 0 (fully transparent) and
22853 255 (fully opaque, the default).
22856 Set the number of decimals to show in the timestamp, only available in the
22857 @code{testsrc} source.
22859 The displayed timestamp value will correspond to the original
22860 timestamp value multiplied by the power of 10 of the specified
22861 value. Default value is 0.
22864 @subsection Examples
22868 Generate a video with a duration of 5.3 seconds, with size
22869 176x144 and a frame rate of 10 frames per second:
22871 testsrc=duration=5.3:size=qcif:rate=10
22875 The following graph description will generate a red source
22876 with an opacity of 0.2, with size "qcif" and a frame rate of 10
22879 color=c=red@@0.2:s=qcif:r=10
22883 If the input content is to be ignored, @code{nullsrc} can be used. The
22884 following command generates noise in the luminance plane by employing
22885 the @code{geq} filter:
22887 nullsrc=s=256x256, geq=random(1)*255:128:128
22891 @subsection Commands
22893 The @code{color} source supports the following commands:
22897 Set the color of the created image. Accepts the same syntax of the
22898 corresponding @option{color} option.
22903 Generate video using an OpenCL program.
22908 OpenCL program source file.
22911 Kernel name in program.
22914 Size of frames to generate. This must be set.
22917 Pixel format to use for the generated frames. This must be set.
22920 Number of frames generated every second. Default value is '25'.
22924 For details of how the program loading works, see the @ref{program_opencl}
22931 Generate a colour ramp by setting pixel values from the position of the pixel
22932 in the output image. (Note that this will work with all pixel formats, but
22933 the generated output will not be the same.)
22935 __kernel void ramp(__write_only image2d_t dst,
22936 unsigned int index)
22938 int2 loc = (int2)(get_global_id(0), get_global_id(1));
22941 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
22943 write_imagef(dst, loc, val);
22948 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
22950 __kernel void sierpinski_carpet(__write_only image2d_t dst,
22951 unsigned int index)
22953 int2 loc = (int2)(get_global_id(0), get_global_id(1));
22955 float4 value = 0.0f;
22956 int x = loc.x + index;
22957 int y = loc.y + index;
22958 while (x > 0 || y > 0) {
22959 if (x % 3 == 1 && y % 3 == 1) {
22967 write_imagef(dst, loc, value);
22973 @section sierpinski
22975 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
22977 This source accepts the following options:
22981 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
22982 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
22985 Set frame rate, expressed as number of frames per second. Default
22989 Set seed which is used for random panning.
22992 Set max jump for single pan destination. Allowed range is from 1 to 10000.
22995 Set fractal type, can be default @code{carpet} or @code{triangle}.
22998 @c man end VIDEO SOURCES
23000 @chapter Video Sinks
23001 @c man begin VIDEO SINKS
23003 Below is a description of the currently available video sinks.
23005 @section buffersink
23007 Buffer video frames, and make them available to the end of the filter
23010 This sink is mainly intended for programmatic use, in particular
23011 through the interface defined in @file{libavfilter/buffersink.h}
23012 or the options system.
23014 It accepts a pointer to an AVBufferSinkContext structure, which
23015 defines the incoming buffers' formats, to be passed as the opaque
23016 parameter to @code{avfilter_init_filter} for initialization.
23020 Null video sink: do absolutely nothing with the input video. It is
23021 mainly useful as a template and for use in analysis / debugging
23024 @c man end VIDEO SINKS
23026 @chapter Multimedia Filters
23027 @c man begin MULTIMEDIA FILTERS
23029 Below is a description of the currently available multimedia filters.
23033 Convert input audio to a video output, displaying the audio bit scope.
23035 The filter accepts the following options:
23039 Set frame rate, expressed as number of frames per second. Default
23043 Specify the video size for the output. For the syntax of this option, check the
23044 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23045 Default value is @code{1024x256}.
23048 Specify list of colors separated by space or by '|' which will be used to
23049 draw channels. Unrecognized or missing colors will be replaced
23053 @section adrawgraph
23054 Draw a graph using input audio metadata.
23056 See @ref{drawgraph}
23058 @section agraphmonitor
23060 See @ref{graphmonitor}.
23062 @section ahistogram
23064 Convert input audio to a video output, displaying the volume histogram.
23066 The filter accepts the following options:
23070 Specify how histogram is calculated.
23072 It accepts the following values:
23075 Use single histogram for all channels.
23077 Use separate histogram for each channel.
23079 Default is @code{single}.
23082 Set frame rate, expressed as number of frames per second. Default
23086 Specify the video size for the output. For the syntax of this option, check the
23087 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23088 Default value is @code{hd720}.
23093 It accepts the following values:
23104 reverse logarithmic
23106 Default is @code{log}.
23109 Set amplitude scale.
23111 It accepts the following values:
23118 Default is @code{log}.
23121 Set how much frames to accumulate in histogram.
23122 Default is 1. Setting this to -1 accumulates all frames.
23125 Set histogram ratio of window height.
23128 Set sonogram sliding.
23130 It accepts the following values:
23133 replace old rows with new ones.
23135 scroll from top to bottom.
23137 Default is @code{replace}.
23140 @section aphasemeter
23142 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
23143 representing mean phase of current audio frame. A video output can also be produced and is
23144 enabled by default. The audio is passed through as first output.
23146 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
23147 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
23148 and @code{1} means channels are in phase.
23150 The filter accepts the following options, all related to its video output:
23154 Set the output frame rate. Default value is @code{25}.
23157 Set the video size for the output. For the syntax of this option, check the
23158 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23159 Default value is @code{800x400}.
23164 Specify the red, green, blue contrast. Default values are @code{2},
23165 @code{7} and @code{1}.
23166 Allowed range is @code{[0, 255]}.
23169 Set color which will be used for drawing median phase. If color is
23170 @code{none} which is default, no median phase value will be drawn.
23173 Enable video output. Default is enabled.
23176 @section avectorscope
23178 Convert input audio to a video output, representing the audio vector
23181 The filter is used to measure the difference between channels of stereo
23182 audio stream. A monaural signal, consisting of identical left and right
23183 signal, results in straight vertical line. Any stereo separation is visible
23184 as a deviation from this line, creating a Lissajous figure.
23185 If the straight (or deviation from it) but horizontal line appears this
23186 indicates that the left and right channels are out of phase.
23188 The filter accepts the following options:
23192 Set the vectorscope mode.
23194 Available values are:
23197 Lissajous rotated by 45 degrees.
23200 Same as above but not rotated.
23203 Shape resembling half of circle.
23206 Default value is @samp{lissajous}.
23209 Set the video size for the output. For the syntax of this option, check the
23210 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23211 Default value is @code{400x400}.
23214 Set the output frame rate. Default value is @code{25}.
23220 Specify the red, green, blue and alpha contrast. Default values are @code{40},
23221 @code{160}, @code{80} and @code{255}.
23222 Allowed range is @code{[0, 255]}.
23228 Specify the red, green, blue and alpha fade. Default values are @code{15},
23229 @code{10}, @code{5} and @code{5}.
23230 Allowed range is @code{[0, 255]}.
23233 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
23234 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
23237 Set the vectorscope drawing mode.
23239 Available values are:
23242 Draw dot for each sample.
23245 Draw line between previous and current sample.
23248 Default value is @samp{dot}.
23251 Specify amplitude scale of audio samples.
23253 Available values are:
23269 Swap left channel axis with right channel axis.
23279 Mirror only x axis.
23282 Mirror only y axis.
23290 @subsection Examples
23294 Complete example using @command{ffplay}:
23296 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
23297 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
23301 @section bench, abench
23303 Benchmark part of a filtergraph.
23305 The filter accepts the following options:
23309 Start or stop a timer.
23311 Available values are:
23314 Get the current time, set it as frame metadata (using the key
23315 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
23318 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
23319 the input frame metadata to get the time difference. Time difference, average,
23320 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
23321 @code{min}) are then printed. The timestamps are expressed in seconds.
23325 @subsection Examples
23329 Benchmark @ref{selectivecolor} filter:
23331 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
23337 Concatenate audio and video streams, joining them together one after the
23340 The filter works on segments of synchronized video and audio streams. All
23341 segments must have the same number of streams of each type, and that will
23342 also be the number of streams at output.
23344 The filter accepts the following options:
23349 Set the number of segments. Default is 2.
23352 Set the number of output video streams, that is also the number of video
23353 streams in each segment. Default is 1.
23356 Set the number of output audio streams, that is also the number of audio
23357 streams in each segment. Default is 0.
23360 Activate unsafe mode: do not fail if segments have a different format.
23364 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
23365 @var{a} audio outputs.
23367 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
23368 segment, in the same order as the outputs, then the inputs for the second
23371 Related streams do not always have exactly the same duration, for various
23372 reasons including codec frame size or sloppy authoring. For that reason,
23373 related synchronized streams (e.g. a video and its audio track) should be
23374 concatenated at once. The concat filter will use the duration of the longest
23375 stream in each segment (except the last one), and if necessary pad shorter
23376 audio streams with silence.
23378 For this filter to work correctly, all segments must start at timestamp 0.
23380 All corresponding streams must have the same parameters in all segments; the
23381 filtering system will automatically select a common pixel format for video
23382 streams, and a common sample format, sample rate and channel layout for
23383 audio streams, but other settings, such as resolution, must be converted
23384 explicitly by the user.
23386 Different frame rates are acceptable but will result in variable frame rate
23387 at output; be sure to configure the output file to handle it.
23389 @subsection Examples
23393 Concatenate an opening, an episode and an ending, all in bilingual version
23394 (video in stream 0, audio in streams 1 and 2):
23396 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
23397 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
23398 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
23399 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
23403 Concatenate two parts, handling audio and video separately, using the
23404 (a)movie sources, and adjusting the resolution:
23406 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
23407 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
23408 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
23410 Note that a desync will happen at the stitch if the audio and video streams
23411 do not have exactly the same duration in the first file.
23415 @subsection Commands
23417 This filter supports the following commands:
23420 Close the current segment and step to the next one
23426 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
23427 level. By default, it logs a message at a frequency of 10Hz with the
23428 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
23429 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
23431 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
23432 sample format is double-precision floating point. The input stream will be converted to
23433 this specification, if needed. Users may need to insert aformat and/or aresample filters
23434 after this filter to obtain the original parameters.
23436 The filter also has a video output (see the @var{video} option) with a real
23437 time graph to observe the loudness evolution. The graphic contains the logged
23438 message mentioned above, so it is not printed anymore when this option is set,
23439 unless the verbose logging is set. The main graphing area contains the
23440 short-term loudness (3 seconds of analysis), and the gauge on the right is for
23441 the momentary loudness (400 milliseconds), but can optionally be configured
23442 to instead display short-term loudness (see @var{gauge}).
23444 The green area marks a +/- 1LU target range around the target loudness
23445 (-23LUFS by default, unless modified through @var{target}).
23447 More information about the Loudness Recommendation EBU R128 on
23448 @url{http://tech.ebu.ch/loudness}.
23450 The filter accepts the following options:
23455 Activate the video output. The audio stream is passed unchanged whether this
23456 option is set or no. The video stream will be the first output stream if
23457 activated. Default is @code{0}.
23460 Set the video size. This option is for video only. For the syntax of this
23462 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23463 Default and minimum resolution is @code{640x480}.
23466 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
23467 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
23468 other integer value between this range is allowed.
23471 Set metadata injection. If set to @code{1}, the audio input will be segmented
23472 into 100ms output frames, each of them containing various loudness information
23473 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
23475 Default is @code{0}.
23478 Force the frame logging level.
23480 Available values are:
23483 information logging level
23485 verbose logging level
23488 By default, the logging level is set to @var{info}. If the @option{video} or
23489 the @option{metadata} options are set, it switches to @var{verbose}.
23494 Available modes can be cumulated (the option is a @code{flag} type). Possible
23498 Disable any peak mode (default).
23500 Enable sample-peak mode.
23502 Simple peak mode looking for the higher sample value. It logs a message
23503 for sample-peak (identified by @code{SPK}).
23505 Enable true-peak mode.
23507 If enabled, the peak lookup is done on an over-sampled version of the input
23508 stream for better peak accuracy. It logs a message for true-peak.
23509 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
23510 This mode requires a build with @code{libswresample}.
23514 Treat mono input files as "dual mono". If a mono file is intended for playback
23515 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
23516 If set to @code{true}, this option will compensate for this effect.
23517 Multi-channel input files are not affected by this option.
23520 Set a specific pan law to be used for the measurement of dual mono files.
23521 This parameter is optional, and has a default value of -3.01dB.
23524 Set a specific target level (in LUFS) used as relative zero in the visualization.
23525 This parameter is optional and has a default value of -23LUFS as specified
23526 by EBU R128. However, material published online may prefer a level of -16LUFS
23527 (e.g. for use with podcasts or video platforms).
23530 Set the value displayed by the gauge. Valid values are @code{momentary} and s
23531 @code{shortterm}. By default the momentary value will be used, but in certain
23532 scenarios it may be more useful to observe the short term value instead (e.g.
23536 Sets the display scale for the loudness. Valid parameters are @code{absolute}
23537 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
23538 video output, not the summary or continuous log output.
23541 @subsection Examples
23545 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
23547 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
23551 Run an analysis with @command{ffmpeg}:
23553 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
23557 @section interleave, ainterleave
23559 Temporally interleave frames from several inputs.
23561 @code{interleave} works with video inputs, @code{ainterleave} with audio.
23563 These filters read frames from several inputs and send the oldest
23564 queued frame to the output.
23566 Input streams must have well defined, monotonically increasing frame
23569 In order to submit one frame to output, these filters need to enqueue
23570 at least one frame for each input, so they cannot work in case one
23571 input is not yet terminated and will not receive incoming frames.
23573 For example consider the case when one input is a @code{select} filter
23574 which always drops input frames. The @code{interleave} filter will keep
23575 reading from that input, but it will never be able to send new frames
23576 to output until the input sends an end-of-stream signal.
23578 Also, depending on inputs synchronization, the filters will drop
23579 frames in case one input receives more frames than the other ones, and
23580 the queue is already filled.
23582 These filters accept the following options:
23586 Set the number of different inputs, it is 2 by default.
23589 How to determine the end-of-stream.
23593 The duration of the longest input. (default)
23596 The duration of the shortest input.
23599 The duration of the first input.
23604 @subsection Examples
23608 Interleave frames belonging to different streams using @command{ffmpeg}:
23610 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
23614 Add flickering blur effect:
23616 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
23620 @section metadata, ametadata
23622 Manipulate frame metadata.
23624 This filter accepts the following options:
23628 Set mode of operation of the filter.
23630 Can be one of the following:
23634 If both @code{value} and @code{key} is set, select frames
23635 which have such metadata. If only @code{key} is set, select
23636 every frame that has such key in metadata.
23639 Add new metadata @code{key} and @code{value}. If key is already available
23643 Modify value of already present key.
23646 If @code{value} is set, delete only keys that have such value.
23647 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
23651 Print key and its value if metadata was found. If @code{key} is not set print all
23652 metadata values available in frame.
23656 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
23659 Set metadata value which will be used. This option is mandatory for
23660 @code{modify} and @code{add} mode.
23663 Which function to use when comparing metadata value and @code{value}.
23665 Can be one of following:
23669 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
23672 Values are interpreted as strings, returns true if metadata value starts with
23673 the @code{value} option string.
23676 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
23679 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
23682 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
23685 Values are interpreted as floats, returns true if expression from option @code{expr}
23689 Values are interpreted as strings, returns true if metadata value ends with
23690 the @code{value} option string.
23694 Set expression which is used when @code{function} is set to @code{expr}.
23695 The expression is evaluated through the eval API and can contain the following
23700 Float representation of @code{value} from metadata key.
23703 Float representation of @code{value} as supplied by user in @code{value} option.
23707 If specified in @code{print} mode, output is written to the named file. Instead of
23708 plain filename any writable url can be specified. Filename ``-'' is a shorthand
23709 for standard output. If @code{file} option is not set, output is written to the log
23710 with AV_LOG_INFO loglevel.
23713 Reduces buffering in print mode when output is written to a URL set using @var{file}.
23717 @subsection Examples
23721 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
23724 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
23727 Print silencedetect output to file @file{metadata.txt}.
23729 silencedetect,ametadata=mode=print:file=metadata.txt
23732 Direct all metadata to a pipe with file descriptor 4.
23734 metadata=mode=print:file='pipe\:4'
23738 @section perms, aperms
23740 Set read/write permissions for the output frames.
23742 These filters are mainly aimed at developers to test direct path in the
23743 following filter in the filtergraph.
23745 The filters accept the following options:
23749 Select the permissions mode.
23751 It accepts the following values:
23754 Do nothing. This is the default.
23756 Set all the output frames read-only.
23758 Set all the output frames directly writable.
23760 Make the frame read-only if writable, and writable if read-only.
23762 Set each output frame read-only or writable randomly.
23766 Set the seed for the @var{random} mode, must be an integer included between
23767 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
23768 @code{-1}, the filter will try to use a good random seed on a best effort
23772 Note: in case of auto-inserted filter between the permission filter and the
23773 following one, the permission might not be received as expected in that
23774 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
23775 perms/aperms filter can avoid this problem.
23777 @section realtime, arealtime
23779 Slow down filtering to match real time approximately.
23781 These filters will pause the filtering for a variable amount of time to
23782 match the output rate with the input timestamps.
23783 They are similar to the @option{re} option to @code{ffmpeg}.
23785 They accept the following options:
23789 Time limit for the pauses. Any pause longer than that will be considered
23790 a timestamp discontinuity and reset the timer. Default is 2 seconds.
23792 Speed factor for processing. The value must be a float larger than zero.
23793 Values larger than 1.0 will result in faster than realtime processing,
23794 smaller will slow processing down. The @var{limit} is automatically adapted
23795 accordingly. Default is 1.0.
23797 A processing speed faster than what is possible without these filters cannot
23802 @section select, aselect
23804 Select frames to pass in output.
23806 This filter accepts the following options:
23811 Set expression, which is evaluated for each input frame.
23813 If the expression is evaluated to zero, the frame is discarded.
23815 If the evaluation result is negative or NaN, the frame is sent to the
23816 first output; otherwise it is sent to the output with index
23817 @code{ceil(val)-1}, assuming that the input index starts from 0.
23819 For example a value of @code{1.2} corresponds to the output with index
23820 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
23823 Set the number of outputs. The output to which to send the selected
23824 frame is based on the result of the evaluation. Default value is 1.
23827 The expression can contain the following constants:
23831 The (sequential) number of the filtered frame, starting from 0.
23834 The (sequential) number of the selected frame, starting from 0.
23836 @item prev_selected_n
23837 The sequential number of the last selected frame. It's NAN if undefined.
23840 The timebase of the input timestamps.
23843 The PTS (Presentation TimeStamp) of the filtered video frame,
23844 expressed in @var{TB} units. It's NAN if undefined.
23847 The PTS of the filtered video frame,
23848 expressed in seconds. It's NAN if undefined.
23851 The PTS of the previously filtered video frame. It's NAN if undefined.
23853 @item prev_selected_pts
23854 The PTS of the last previously filtered video frame. It's NAN if undefined.
23856 @item prev_selected_t
23857 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
23860 The PTS of the first video frame in the video. It's NAN if undefined.
23863 The time of the first video frame in the video. It's NAN if undefined.
23865 @item pict_type @emph{(video only)}
23866 The type of the filtered frame. It can assume one of the following
23878 @item interlace_type @emph{(video only)}
23879 The frame interlace type. It can assume one of the following values:
23882 The frame is progressive (not interlaced).
23884 The frame is top-field-first.
23886 The frame is bottom-field-first.
23889 @item consumed_sample_n @emph{(audio only)}
23890 the number of selected samples before the current frame
23892 @item samples_n @emph{(audio only)}
23893 the number of samples in the current frame
23895 @item sample_rate @emph{(audio only)}
23896 the input sample rate
23899 This is 1 if the filtered frame is a key-frame, 0 otherwise.
23902 the position in the file of the filtered frame, -1 if the information
23903 is not available (e.g. for synthetic video)
23905 @item scene @emph{(video only)}
23906 value between 0 and 1 to indicate a new scene; a low value reflects a low
23907 probability for the current frame to introduce a new scene, while a higher
23908 value means the current frame is more likely to be one (see the example below)
23910 @item concatdec_select
23911 The concat demuxer can select only part of a concat input file by setting an
23912 inpoint and an outpoint, but the output packets may not be entirely contained
23913 in the selected interval. By using this variable, it is possible to skip frames
23914 generated by the concat demuxer which are not exactly contained in the selected
23917 This works by comparing the frame pts against the @var{lavf.concat.start_time}
23918 and the @var{lavf.concat.duration} packet metadata values which are also
23919 present in the decoded frames.
23921 The @var{concatdec_select} variable is -1 if the frame pts is at least
23922 start_time and either the duration metadata is missing or the frame pts is less
23923 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
23926 That basically means that an input frame is selected if its pts is within the
23927 interval set by the concat demuxer.
23931 The default value of the select expression is "1".
23933 @subsection Examples
23937 Select all frames in input:
23942 The example above is the same as:
23954 Select only I-frames:
23956 select='eq(pict_type\,I)'
23960 Select one frame every 100:
23962 select='not(mod(n\,100))'
23966 Select only frames contained in the 10-20 time interval:
23968 select=between(t\,10\,20)
23972 Select only I-frames contained in the 10-20 time interval:
23974 select=between(t\,10\,20)*eq(pict_type\,I)
23978 Select frames with a minimum distance of 10 seconds:
23980 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
23984 Use aselect to select only audio frames with samples number > 100:
23986 aselect='gt(samples_n\,100)'
23990 Create a mosaic of the first scenes:
23992 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
23995 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
23999 Send even and odd frames to separate outputs, and compose them:
24001 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
24005 Select useful frames from an ffconcat file which is using inpoints and
24006 outpoints but where the source files are not intra frame only.
24008 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
24012 @section sendcmd, asendcmd
24014 Send commands to filters in the filtergraph.
24016 These filters read commands to be sent to other filters in the
24019 @code{sendcmd} must be inserted between two video filters,
24020 @code{asendcmd} must be inserted between two audio filters, but apart
24021 from that they act the same way.
24023 The specification of commands can be provided in the filter arguments
24024 with the @var{commands} option, or in a file specified by the
24025 @var{filename} option.
24027 These filters accept the following options:
24030 Set the commands to be read and sent to the other filters.
24032 Set the filename of the commands to be read and sent to the other
24036 @subsection Commands syntax
24038 A commands description consists of a sequence of interval
24039 specifications, comprising a list of commands to be executed when a
24040 particular event related to that interval occurs. The occurring event
24041 is typically the current frame time entering or leaving a given time
24044 An interval is specified by the following syntax:
24046 @var{START}[-@var{END}] @var{COMMANDS};
24049 The time interval is specified by the @var{START} and @var{END} times.
24050 @var{END} is optional and defaults to the maximum time.
24052 The current frame time is considered within the specified interval if
24053 it is included in the interval [@var{START}, @var{END}), that is when
24054 the time is greater or equal to @var{START} and is lesser than
24057 @var{COMMANDS} consists of a sequence of one or more command
24058 specifications, separated by ",", relating to that interval. The
24059 syntax of a command specification is given by:
24061 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
24064 @var{FLAGS} is optional and specifies the type of events relating to
24065 the time interval which enable sending the specified command, and must
24066 be a non-null sequence of identifier flags separated by "+" or "|" and
24067 enclosed between "[" and "]".
24069 The following flags are recognized:
24072 The command is sent when the current frame timestamp enters the
24073 specified interval. In other words, the command is sent when the
24074 previous frame timestamp was not in the given interval, and the
24078 The command is sent when the current frame timestamp leaves the
24079 specified interval. In other words, the command is sent when the
24080 previous frame timestamp was in the given interval, and the
24084 The command @var{ARG} is interpreted as expression and result of
24085 expression is passed as @var{ARG}.
24087 The expression is evaluated through the eval API and can contain the following
24092 Original position in the file of the frame, or undefined if undefined
24093 for the current frame.
24096 The presentation timestamp in input.
24099 The count of the input frame for video or audio, starting from 0.
24102 The time in seconds of the current frame.
24105 The start time in seconds of the current command interval.
24108 The end time in seconds of the current command interval.
24111 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
24116 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
24119 @var{TARGET} specifies the target of the command, usually the name of
24120 the filter class or a specific filter instance name.
24122 @var{COMMAND} specifies the name of the command for the target filter.
24124 @var{ARG} is optional and specifies the optional list of argument for
24125 the given @var{COMMAND}.
24127 Between one interval specification and another, whitespaces, or
24128 sequences of characters starting with @code{#} until the end of line,
24129 are ignored and can be used to annotate comments.
24131 A simplified BNF description of the commands specification syntax
24134 @var{COMMAND_FLAG} ::= "enter" | "leave"
24135 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
24136 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
24137 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
24138 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
24139 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
24142 @subsection Examples
24146 Specify audio tempo change at second 4:
24148 asendcmd=c='4.0 atempo tempo 1.5',atempo
24152 Target a specific filter instance:
24154 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
24158 Specify a list of drawtext and hue commands in a file.
24160 # show text in the interval 5-10
24161 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
24162 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
24164 # desaturate the image in the interval 15-20
24165 15.0-20.0 [enter] hue s 0,
24166 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
24168 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
24170 # apply an exponential saturation fade-out effect, starting from time 25
24171 25 [enter] hue s exp(25-t)
24174 A filtergraph allowing to read and process the above command list
24175 stored in a file @file{test.cmd}, can be specified with:
24177 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
24182 @section setpts, asetpts
24184 Change the PTS (presentation timestamp) of the input frames.
24186 @code{setpts} works on video frames, @code{asetpts} on audio frames.
24188 This filter accepts the following options:
24193 The expression which is evaluated for each frame to construct its timestamp.
24197 The expression is evaluated through the eval API and can contain the following
24201 @item FRAME_RATE, FR
24202 frame rate, only defined for constant frame-rate video
24205 The presentation timestamp in input
24208 The count of the input frame for video or the number of consumed samples,
24209 not including the current frame for audio, starting from 0.
24211 @item NB_CONSUMED_SAMPLES
24212 The number of consumed samples, not including the current frame (only
24215 @item NB_SAMPLES, S
24216 The number of samples in the current frame (only audio)
24218 @item SAMPLE_RATE, SR
24219 The audio sample rate.
24222 The PTS of the first frame.
24225 the time in seconds of the first frame
24228 State whether the current frame is interlaced.
24231 the time in seconds of the current frame
24234 original position in the file of the frame, or undefined if undefined
24235 for the current frame
24238 The previous input PTS.
24241 previous input time in seconds
24244 The previous output PTS.
24247 previous output time in seconds
24250 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
24254 The wallclock (RTC) time at the start of the movie in microseconds.
24257 The timebase of the input timestamps.
24261 @subsection Examples
24265 Start counting PTS from zero
24267 setpts=PTS-STARTPTS
24271 Apply fast motion effect:
24277 Apply slow motion effect:
24283 Set fixed rate of 25 frames per second:
24289 Set fixed rate 25 fps with some jitter:
24291 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
24295 Apply an offset of 10 seconds to the input PTS:
24301 Generate timestamps from a "live source" and rebase onto the current timebase:
24303 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
24307 Generate timestamps by counting samples:
24316 Force color range for the output video frame.
24318 The @code{setrange} filter marks the color range property for the
24319 output frames. It does not change the input frame, but only sets the
24320 corresponding property, which affects how the frame is treated by
24323 The filter accepts the following options:
24328 Available values are:
24332 Keep the same color range property.
24334 @item unspecified, unknown
24335 Set the color range as unspecified.
24337 @item limited, tv, mpeg
24338 Set the color range as limited.
24340 @item full, pc, jpeg
24341 Set the color range as full.
24345 @section settb, asettb
24347 Set the timebase to use for the output frames timestamps.
24348 It is mainly useful for testing timebase configuration.
24350 It accepts the following parameters:
24355 The expression which is evaluated into the output timebase.
24359 The value for @option{tb} is an arithmetic expression representing a
24360 rational. The expression can contain the constants "AVTB" (the default
24361 timebase), "intb" (the input timebase) and "sr" (the sample rate,
24362 audio only). Default value is "intb".
24364 @subsection Examples
24368 Set the timebase to 1/25:
24374 Set the timebase to 1/10:
24380 Set the timebase to 1001/1000:
24386 Set the timebase to 2*intb:
24392 Set the default timebase value:
24399 Convert input audio to a video output representing frequency spectrum
24400 logarithmically using Brown-Puckette constant Q transform algorithm with
24401 direct frequency domain coefficient calculation (but the transform itself
24402 is not really constant Q, instead the Q factor is actually variable/clamped),
24403 with musical tone scale, from E0 to D#10.
24405 The filter accepts the following options:
24409 Specify the video size for the output. It must be even. For the syntax of this option,
24410 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24411 Default value is @code{1920x1080}.
24414 Set the output frame rate. Default value is @code{25}.
24417 Set the bargraph height. It must be even. Default value is @code{-1} which
24418 computes the bargraph height automatically.
24421 Set the axis height. It must be even. Default value is @code{-1} which computes
24422 the axis height automatically.
24425 Set the sonogram height. It must be even. Default value is @code{-1} which
24426 computes the sonogram height automatically.
24429 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
24430 instead. Default value is @code{1}.
24432 @item sono_v, volume
24433 Specify the sonogram volume expression. It can contain variables:
24436 the @var{bar_v} evaluated expression
24437 @item frequency, freq, f
24438 the frequency where it is evaluated
24439 @item timeclamp, tc
24440 the value of @var{timeclamp} option
24444 @item a_weighting(f)
24445 A-weighting of equal loudness
24446 @item b_weighting(f)
24447 B-weighting of equal loudness
24448 @item c_weighting(f)
24449 C-weighting of equal loudness.
24451 Default value is @code{16}.
24453 @item bar_v, volume2
24454 Specify the bargraph volume expression. It can contain variables:
24457 the @var{sono_v} evaluated expression
24458 @item frequency, freq, f
24459 the frequency where it is evaluated
24460 @item timeclamp, tc
24461 the value of @var{timeclamp} option
24465 @item a_weighting(f)
24466 A-weighting of equal loudness
24467 @item b_weighting(f)
24468 B-weighting of equal loudness
24469 @item c_weighting(f)
24470 C-weighting of equal loudness.
24472 Default value is @code{sono_v}.
24474 @item sono_g, gamma
24475 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
24476 higher gamma makes the spectrum having more range. Default value is @code{3}.
24477 Acceptable range is @code{[1, 7]}.
24479 @item bar_g, gamma2
24480 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
24484 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
24485 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
24487 @item timeclamp, tc
24488 Specify the transform timeclamp. At low frequency, there is trade-off between
24489 accuracy in time domain and frequency domain. If timeclamp is lower,
24490 event in time domain is represented more accurately (such as fast bass drum),
24491 otherwise event in frequency domain is represented more accurately
24492 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
24495 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
24496 limits future samples by applying asymmetric windowing in time domain, useful
24497 when low latency is required. Accepted range is @code{[0, 1]}.
24500 Specify the transform base frequency. Default value is @code{20.01523126408007475},
24501 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
24504 Specify the transform end frequency. Default value is @code{20495.59681441799654},
24505 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
24508 This option is deprecated and ignored.
24511 Specify the transform length in time domain. Use this option to control accuracy
24512 trade-off between time domain and frequency domain at every frequency sample.
24513 It can contain variables:
24515 @item frequency, freq, f
24516 the frequency where it is evaluated
24517 @item timeclamp, tc
24518 the value of @var{timeclamp} option.
24520 Default value is @code{384*tc/(384+tc*f)}.
24523 Specify the transform count for every video frame. Default value is @code{6}.
24524 Acceptable range is @code{[1, 30]}.
24527 Specify the transform count for every single pixel. Default value is @code{0},
24528 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
24531 Specify font file for use with freetype to draw the axis. If not specified,
24532 use embedded font. Note that drawing with font file or embedded font is not
24533 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
24537 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
24538 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
24542 Specify font color expression. This is arithmetic expression that should return
24543 integer value 0xRRGGBB. It can contain variables:
24545 @item frequency, freq, f
24546 the frequency where it is evaluated
24547 @item timeclamp, tc
24548 the value of @var{timeclamp} option
24553 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
24554 @item r(x), g(x), b(x)
24555 red, green, and blue value of intensity x.
24557 Default value is @code{st(0, (midi(f)-59.5)/12);
24558 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
24559 r(1-ld(1)) + b(ld(1))}.
24562 Specify image file to draw the axis. This option override @var{fontfile} and
24563 @var{fontcolor} option.
24566 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
24567 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
24568 Default value is @code{1}.
24571 Set colorspace. The accepted values are:
24574 Unspecified (default)
24583 BT.470BG or BT.601-6 625
24586 SMPTE-170M or BT.601-6 525
24592 BT.2020 with non-constant luminance
24597 Set spectrogram color scheme. This is list of floating point values with format
24598 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
24599 The default is @code{1|0.5|0|0|0.5|1}.
24603 @subsection Examples
24607 Playing audio while showing the spectrum:
24609 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
24613 Same as above, but with frame rate 30 fps:
24615 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
24619 Playing at 1280x720:
24621 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
24625 Disable sonogram display:
24631 A1 and its harmonics: A1, A2, (near)E3, A3:
24633 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),
24634 asplit[a][out1]; [a] showcqt [out0]'
24638 Same as above, but with more accuracy in frequency domain:
24640 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),
24641 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
24647 bar_v=10:sono_v=bar_v*a_weighting(f)
24651 Custom gamma, now spectrum is linear to the amplitude.
24657 Custom tlength equation:
24659 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)))'
24663 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
24665 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
24669 Custom font using fontconfig:
24671 font='Courier New,Monospace,mono|bold'
24675 Custom frequency range with custom axis using image file:
24677 axisfile=myaxis.png:basefreq=40:endfreq=10000
24683 Convert input audio to video output representing the audio power spectrum.
24684 Audio amplitude is on Y-axis while frequency is on X-axis.
24686 The filter accepts the following options:
24690 Specify size of video. For the syntax of this option, check the
24691 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24692 Default is @code{1024x512}.
24696 This set how each frequency bin will be represented.
24698 It accepts the following values:
24704 Default is @code{bar}.
24707 Set amplitude scale.
24709 It accepts the following values:
24723 Default is @code{log}.
24726 Set frequency scale.
24728 It accepts the following values:
24737 Reverse logarithmic scale.
24739 Default is @code{lin}.
24742 Set window size. Allowed range is from 16 to 65536.
24744 Default is @code{2048}
24747 Set windowing function.
24749 It accepts the following values:
24772 Default is @code{hanning}.
24775 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
24776 which means optimal overlap for selected window function will be picked.
24779 Set time averaging. Setting this to 0 will display current maximal peaks.
24780 Default is @code{1}, which means time averaging is disabled.
24783 Specify list of colors separated by space or by '|' which will be used to
24784 draw channel frequencies. Unrecognized or missing colors will be replaced
24788 Set channel display mode.
24790 It accepts the following values:
24795 Default is @code{combined}.
24798 Set minimum amplitude used in @code{log} amplitude scaler.
24802 @section showspatial
24804 Convert stereo input audio to a video output, representing the spatial relationship
24805 between two channels.
24807 The filter accepts the following options:
24811 Specify the video size for the output. For the syntax of this option, check the
24812 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24813 Default value is @code{512x512}.
24816 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
24819 Set window function.
24821 It accepts the following values:
24846 Default value is @code{hann}.
24849 Set ratio of overlap window. Default value is @code{0.5}.
24850 When value is @code{1} overlap is set to recommended size for specific
24851 window function currently used.
24854 @anchor{showspectrum}
24855 @section showspectrum
24857 Convert input audio to a video output, representing the audio frequency
24860 The filter accepts the following options:
24864 Specify the video size for the output. For the syntax of this option, check the
24865 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24866 Default value is @code{640x512}.
24869 Specify how the spectrum should slide along the window.
24871 It accepts the following values:
24874 the samples start again on the left when they reach the right
24876 the samples scroll from right to left
24878 frames are only produced when the samples reach the right
24880 the samples scroll from left to right
24883 Default value is @code{replace}.
24886 Specify display mode.
24888 It accepts the following values:
24891 all channels are displayed in the same row
24893 all channels are displayed in separate rows
24896 Default value is @samp{combined}.
24899 Specify display color mode.
24901 It accepts the following values:
24904 each channel is displayed in a separate color
24906 each channel is displayed using the same color scheme
24908 each channel is displayed using the rainbow color scheme
24910 each channel is displayed using the moreland color scheme
24912 each channel is displayed using the nebulae color scheme
24914 each channel is displayed using the fire color scheme
24916 each channel is displayed using the fiery color scheme
24918 each channel is displayed using the fruit color scheme
24920 each channel is displayed using the cool color scheme
24922 each channel is displayed using the magma color scheme
24924 each channel is displayed using the green color scheme
24926 each channel is displayed using the viridis color scheme
24928 each channel is displayed using the plasma color scheme
24930 each channel is displayed using the cividis color scheme
24932 each channel is displayed using the terrain color scheme
24935 Default value is @samp{channel}.
24938 Specify scale used for calculating intensity color values.
24940 It accepts the following values:
24945 square root, default
24956 Default value is @samp{sqrt}.
24959 Specify frequency scale.
24961 It accepts the following values:
24969 Default value is @samp{lin}.
24972 Set saturation modifier for displayed colors. Negative values provide
24973 alternative color scheme. @code{0} is no saturation at all.
24974 Saturation must be in [-10.0, 10.0] range.
24975 Default value is @code{1}.
24978 Set window function.
24980 It accepts the following values:
25005 Default value is @code{hann}.
25008 Set orientation of time vs frequency axis. Can be @code{vertical} or
25009 @code{horizontal}. Default is @code{vertical}.
25012 Set ratio of overlap window. Default value is @code{0}.
25013 When value is @code{1} overlap is set to recommended size for specific
25014 window function currently used.
25017 Set scale gain for calculating intensity color values.
25018 Default value is @code{1}.
25021 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
25024 Set color rotation, must be in [-1.0, 1.0] range.
25025 Default value is @code{0}.
25028 Set start frequency from which to display spectrogram. Default is @code{0}.
25031 Set stop frequency to which to display spectrogram. Default is @code{0}.
25034 Set upper frame rate limit. Default is @code{auto}, unlimited.
25037 Draw time and frequency axes and legends. Default is disabled.
25040 The usage is very similar to the showwaves filter; see the examples in that
25043 @subsection Examples
25047 Large window with logarithmic color scaling:
25049 showspectrum=s=1280x480:scale=log
25053 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
25055 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
25056 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
25060 @section showspectrumpic
25062 Convert input audio to a single video frame, representing the audio frequency
25065 The filter accepts the following options:
25069 Specify the video size for the output. For the syntax of this option, check the
25070 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25071 Default value is @code{4096x2048}.
25074 Specify display mode.
25076 It accepts the following values:
25079 all channels are displayed in the same row
25081 all channels are displayed in separate rows
25083 Default value is @samp{combined}.
25086 Specify display color mode.
25088 It accepts the following values:
25091 each channel is displayed in a separate color
25093 each channel is displayed using the same color scheme
25095 each channel is displayed using the rainbow color scheme
25097 each channel is displayed using the moreland color scheme
25099 each channel is displayed using the nebulae color scheme
25101 each channel is displayed using the fire color scheme
25103 each channel is displayed using the fiery color scheme
25105 each channel is displayed using the fruit color scheme
25107 each channel is displayed using the cool color scheme
25109 each channel is displayed using the magma color scheme
25111 each channel is displayed using the green color scheme
25113 each channel is displayed using the viridis color scheme
25115 each channel is displayed using the plasma color scheme
25117 each channel is displayed using the cividis color scheme
25119 each channel is displayed using the terrain color scheme
25121 Default value is @samp{intensity}.
25124 Specify scale used for calculating intensity color values.
25126 It accepts the following values:
25131 square root, default
25141 Default value is @samp{log}.
25144 Specify frequency scale.
25146 It accepts the following values:
25154 Default value is @samp{lin}.
25157 Set saturation modifier for displayed colors. Negative values provide
25158 alternative color scheme. @code{0} is no saturation at all.
25159 Saturation must be in [-10.0, 10.0] range.
25160 Default value is @code{1}.
25163 Set window function.
25165 It accepts the following values:
25189 Default value is @code{hann}.
25192 Set orientation of time vs frequency axis. Can be @code{vertical} or
25193 @code{horizontal}. Default is @code{vertical}.
25196 Set scale gain for calculating intensity color values.
25197 Default value is @code{1}.
25200 Draw time and frequency axes and legends. Default is enabled.
25203 Set color rotation, must be in [-1.0, 1.0] range.
25204 Default value is @code{0}.
25207 Set start frequency from which to display spectrogram. Default is @code{0}.
25210 Set stop frequency to which to display spectrogram. Default is @code{0}.
25213 @subsection Examples
25217 Extract an audio spectrogram of a whole audio track
25218 in a 1024x1024 picture using @command{ffmpeg}:
25220 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
25224 @section showvolume
25226 Convert input audio volume to a video output.
25228 The filter accepts the following options:
25235 Set border width, allowed range is [0, 5]. Default is 1.
25238 Set channel width, allowed range is [80, 8192]. Default is 400.
25241 Set channel height, allowed range is [1, 900]. Default is 20.
25244 Set fade, allowed range is [0, 1]. Default is 0.95.
25247 Set volume color expression.
25249 The expression can use the following variables:
25253 Current max volume of channel in dB.
25259 Current channel number, starting from 0.
25263 If set, displays channel names. Default is enabled.
25266 If set, displays volume values. Default is enabled.
25269 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
25270 default is @code{h}.
25273 Set step size, allowed range is [0, 5]. Default is 0, which means
25277 Set background opacity, allowed range is [0, 1]. Default is 0.
25280 Set metering mode, can be peak: @code{p} or rms: @code{r},
25281 default is @code{p}.
25284 Set display scale, can be linear: @code{lin} or log: @code{log},
25285 default is @code{lin}.
25289 If set to > 0., display a line for the max level
25290 in the previous seconds.
25291 default is disabled: @code{0.}
25294 The color of the max line. Use when @code{dm} option is set to > 0.
25295 default is: @code{orange}
25300 Convert input audio to a video output, representing the samples waves.
25302 The filter accepts the following options:
25306 Specify the video size for the output. For the syntax of this option, check the
25307 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25308 Default value is @code{600x240}.
25313 Available values are:
25316 Draw a point for each sample.
25319 Draw a vertical line for each sample.
25322 Draw a point for each sample and a line between them.
25325 Draw a centered vertical line for each sample.
25328 Default value is @code{point}.
25331 Set the number of samples which are printed on the same column. A
25332 larger value will decrease the frame rate. Must be a positive
25333 integer. This option can be set only if the value for @var{rate}
25334 is not explicitly specified.
25337 Set the (approximate) output frame rate. This is done by setting the
25338 option @var{n}. Default value is "25".
25340 @item split_channels
25341 Set if channels should be drawn separately or overlap. Default value is 0.
25344 Set colors separated by '|' which are going to be used for drawing of each channel.
25347 Set amplitude scale.
25349 Available values are:
25367 Set the draw mode. This is mostly useful to set for high @var{n}.
25369 Available values are:
25372 Scale pixel values for each drawn sample.
25375 Draw every sample directly.
25378 Default value is @code{scale}.
25381 @subsection Examples
25385 Output the input file audio and the corresponding video representation
25388 amovie=a.mp3,asplit[out0],showwaves[out1]
25392 Create a synthetic signal and show it with showwaves, forcing a
25393 frame rate of 30 frames per second:
25395 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
25399 @section showwavespic
25401 Convert input audio to a single video frame, representing the samples waves.
25403 The filter accepts the following options:
25407 Specify the video size for the output. For the syntax of this option, check the
25408 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25409 Default value is @code{600x240}.
25411 @item split_channels
25412 Set if channels should be drawn separately or overlap. Default value is 0.
25415 Set colors separated by '|' which are going to be used for drawing of each channel.
25418 Set amplitude scale.
25420 Available values are:
25440 Available values are:
25443 Scale pixel values for each drawn sample.
25446 Draw every sample directly.
25449 Default value is @code{scale}.
25452 @subsection Examples
25456 Extract a channel split representation of the wave form of a whole audio track
25457 in a 1024x800 picture using @command{ffmpeg}:
25459 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
25463 @section sidedata, asidedata
25465 Delete frame side data, or select frames based on it.
25467 This filter accepts the following options:
25471 Set mode of operation of the filter.
25473 Can be one of the following:
25477 Select every frame with side data of @code{type}.
25480 Delete side data of @code{type}. If @code{type} is not set, delete all side
25486 Set side data type used with all modes. Must be set for @code{select} mode. For
25487 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
25488 in @file{libavutil/frame.h}. For example, to choose
25489 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
25493 @section spectrumsynth
25495 Synthesize audio from 2 input video spectrums, first input stream represents
25496 magnitude across time and second represents phase across time.
25497 The filter will transform from frequency domain as displayed in videos back
25498 to time domain as presented in audio output.
25500 This filter is primarily created for reversing processed @ref{showspectrum}
25501 filter outputs, but can synthesize sound from other spectrograms too.
25502 But in such case results are going to be poor if the phase data is not
25503 available, because in such cases phase data need to be recreated, usually
25504 it's just recreated from random noise.
25505 For best results use gray only output (@code{channel} color mode in
25506 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
25507 @code{lin} scale for phase video. To produce phase, for 2nd video, use
25508 @code{data} option. Inputs videos should generally use @code{fullframe}
25509 slide mode as that saves resources needed for decoding video.
25511 The filter accepts the following options:
25515 Specify sample rate of output audio, the sample rate of audio from which
25516 spectrum was generated may differ.
25519 Set number of channels represented in input video spectrums.
25522 Set scale which was used when generating magnitude input spectrum.
25523 Can be @code{lin} or @code{log}. Default is @code{log}.
25526 Set slide which was used when generating inputs spectrums.
25527 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
25528 Default is @code{fullframe}.
25531 Set window function used for resynthesis.
25534 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
25535 which means optimal overlap for selected window function will be picked.
25538 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
25539 Default is @code{vertical}.
25542 @subsection Examples
25546 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
25547 then resynthesize videos back to audio with spectrumsynth:
25549 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
25550 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
25551 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
25555 @section split, asplit
25557 Split input into several identical outputs.
25559 @code{asplit} works with audio input, @code{split} with video.
25561 The filter accepts a single parameter which specifies the number of outputs. If
25562 unspecified, it defaults to 2.
25564 @subsection Examples
25568 Create two separate outputs from the same input:
25570 [in] split [out0][out1]
25574 To create 3 or more outputs, you need to specify the number of
25577 [in] asplit=3 [out0][out1][out2]
25581 Create two separate outputs from the same input, one cropped and
25584 [in] split [splitout1][splitout2];
25585 [splitout1] crop=100:100:0:0 [cropout];
25586 [splitout2] pad=200:200:100:100 [padout];
25590 Create 5 copies of the input audio with @command{ffmpeg}:
25592 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
25598 Receive commands sent through a libzmq client, and forward them to
25599 filters in the filtergraph.
25601 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
25602 must be inserted between two video filters, @code{azmq} between two
25603 audio filters. Both are capable to send messages to any filter type.
25605 To enable these filters you need to install the libzmq library and
25606 headers and configure FFmpeg with @code{--enable-libzmq}.
25608 For more information about libzmq see:
25609 @url{http://www.zeromq.org/}
25611 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
25612 receives messages sent through a network interface defined by the
25613 @option{bind_address} (or the abbreviation "@option{b}") option.
25614 Default value of this option is @file{tcp://localhost:5555}. You may
25615 want to alter this value to your needs, but do not forget to escape any
25616 ':' signs (see @ref{filtergraph escaping}).
25618 The received message must be in the form:
25620 @var{TARGET} @var{COMMAND} [@var{ARG}]
25623 @var{TARGET} specifies the target of the command, usually the name of
25624 the filter class or a specific filter instance name. The default
25625 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
25626 but you can override this by using the @samp{filter_name@@id} syntax
25627 (see @ref{Filtergraph syntax}).
25629 @var{COMMAND} specifies the name of the command for the target filter.
25631 @var{ARG} is optional and specifies the optional argument list for the
25632 given @var{COMMAND}.
25634 Upon reception, the message is processed and the corresponding command
25635 is injected into the filtergraph. Depending on the result, the filter
25636 will send a reply to the client, adopting the format:
25638 @var{ERROR_CODE} @var{ERROR_REASON}
25642 @var{MESSAGE} is optional.
25644 @subsection Examples
25646 Look at @file{tools/zmqsend} for an example of a zmq client which can
25647 be used to send commands processed by these filters.
25649 Consider the following filtergraph generated by @command{ffplay}.
25650 In this example the last overlay filter has an instance name. All other
25651 filters will have default instance names.
25654 ffplay -dumpgraph 1 -f lavfi "
25655 color=s=100x100:c=red [l];
25656 color=s=100x100:c=blue [r];
25657 nullsrc=s=200x100, zmq [bg];
25658 [bg][l] overlay [bg+l];
25659 [bg+l][r] overlay@@my=x=100 "
25662 To change the color of the left side of the video, the following
25663 command can be used:
25665 echo Parsed_color_0 c yellow | tools/zmqsend
25668 To change the right side:
25670 echo Parsed_color_1 c pink | tools/zmqsend
25673 To change the position of the right side:
25675 echo overlay@@my x 150 | tools/zmqsend
25679 @c man end MULTIMEDIA FILTERS
25681 @chapter Multimedia Sources
25682 @c man begin MULTIMEDIA SOURCES
25684 Below is a description of the currently available multimedia sources.
25688 This is the same as @ref{movie} source, except it selects an audio
25694 Read audio and/or video stream(s) from a movie container.
25696 It accepts the following parameters:
25700 The name of the resource to read (not necessarily a file; it can also be a
25701 device or a stream accessed through some protocol).
25703 @item format_name, f
25704 Specifies the format assumed for the movie to read, and can be either
25705 the name of a container or an input device. If not specified, the
25706 format is guessed from @var{movie_name} or by probing.
25708 @item seek_point, sp
25709 Specifies the seek point in seconds. The frames will be output
25710 starting from this seek point. The parameter is evaluated with
25711 @code{av_strtod}, so the numerical value may be suffixed by an IS
25712 postfix. The default value is "0".
25715 Specifies the streams to read. Several streams can be specified,
25716 separated by "+". The source will then have as many outputs, in the
25717 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
25718 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
25719 respectively the default (best suited) video and audio stream. Default
25720 is "dv", or "da" if the filter is called as "amovie".
25722 @item stream_index, si
25723 Specifies the index of the video stream to read. If the value is -1,
25724 the most suitable video stream will be automatically selected. The default
25725 value is "-1". Deprecated. If the filter is called "amovie", it will select
25726 audio instead of video.
25729 Specifies how many times to read the stream in sequence.
25730 If the value is 0, the stream will be looped infinitely.
25731 Default value is "1".
25733 Note that when the movie is looped the source timestamps are not
25734 changed, so it will generate non monotonically increasing timestamps.
25736 @item discontinuity
25737 Specifies the time difference between frames above which the point is
25738 considered a timestamp discontinuity which is removed by adjusting the later
25742 It allows overlaying a second video on top of the main input of
25743 a filtergraph, as shown in this graph:
25745 input -----------> deltapts0 --> overlay --> output
25748 movie --> scale--> deltapts1 -------+
25750 @subsection Examples
25754 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
25755 on top of the input labelled "in":
25757 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
25758 [in] setpts=PTS-STARTPTS [main];
25759 [main][over] overlay=16:16 [out]
25763 Read from a video4linux2 device, and overlay it on top of the input
25766 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
25767 [in] setpts=PTS-STARTPTS [main];
25768 [main][over] overlay=16:16 [out]
25772 Read the first video stream and the audio stream with id 0x81 from
25773 dvd.vob; the video is connected to the pad named "video" and the audio is
25774 connected to the pad named "audio":
25776 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
25780 @subsection Commands
25782 Both movie and amovie support the following commands:
25785 Perform seek using "av_seek_frame".
25786 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
25789 @var{stream_index}: If stream_index is -1, a default
25790 stream is selected, and @var{timestamp} is automatically converted
25791 from AV_TIME_BASE units to the stream specific time_base.
25793 @var{timestamp}: Timestamp in AVStream.time_base units
25794 or, if no stream is specified, in AV_TIME_BASE units.
25796 @var{flags}: Flags which select direction and seeking mode.
25800 Get movie duration in AV_TIME_BASE units.
25804 @c man end MULTIMEDIA SOURCES