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 for each band split. This controls filter roll-off or steepness
527 of filter transfer function.
528 Available values are:
553 Default is @var{4th}.
556 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
559 Set output gain for each band. Default value is 1 for all bands.
566 Split input audio stream into two bands (low and high) with split frequency of 1500 Hz,
567 each band will be in separate stream:
569 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
573 Same as above, but with higher filter order:
575 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500:order=8th[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
579 Same as above, but also with additional middle band (frequencies between 1500 and 8000):
581 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500 8000:order=8th[LOW][MID][HIGH]' -map '[LOW]' low.wav -map '[MID]' mid.wav -map '[HIGH]' high.wav
587 Reduce audio bit resolution.
589 This filter is bit crusher with enhanced functionality. A bit crusher
590 is used to audibly reduce number of bits an audio signal is sampled
591 with. This doesn't change the bit depth at all, it just produces the
592 effect. Material reduced in bit depth sounds more harsh and "digital".
593 This filter is able to even round to continuous values instead of discrete
595 Additionally it has a D/C offset which results in different crushing of
596 the lower and the upper half of the signal.
597 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
599 Another feature of this filter is the logarithmic mode.
600 This setting switches from linear distances between bits to logarithmic ones.
601 The result is a much more "natural" sounding crusher which doesn't gate low
602 signals for example. The human ear has a logarithmic perception,
603 so this kind of crushing is much more pleasant.
604 Logarithmic crushing is also able to get anti-aliased.
606 The filter accepts the following options:
622 Can be linear: @code{lin} or logarithmic: @code{log}.
631 Set sample reduction.
634 Enable LFO. By default disabled.
645 This filter supports the all above options as @ref{commands}.
649 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
653 Remove impulsive noise from input audio.
655 Samples detected as impulsive noise are replaced by interpolated samples using
656 autoregressive modelling.
660 Set window size, in milliseconds. Allowed range is from @code{10} to
661 @code{100}. Default value is @code{55} milliseconds.
662 This sets size of window which will be processed at once.
665 Set window overlap, in percentage of window size. Allowed range is from
666 @code{50} to @code{95}. Default value is @code{75} percent.
667 Setting this to a very high value increases impulsive noise removal but makes
668 whole process much slower.
671 Set autoregression order, in percentage of window size. Allowed range is from
672 @code{0} to @code{25}. Default value is @code{2} percent. This option also
673 controls quality of interpolated samples using neighbour good samples.
676 Set threshold value. Allowed range is from @code{1} to @code{100}.
677 Default value is @code{2}.
678 This controls the strength of impulsive noise which is going to be removed.
679 The lower value, the more samples will be detected as impulsive noise.
682 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
683 @code{10}. Default value is @code{2}.
684 If any two samples detected as noise are spaced less than this value then any
685 sample between those two samples will be also detected as noise.
690 It accepts the following values:
693 Select overlap-add method. Even not interpolated samples are slightly
694 changed with this method.
697 Select overlap-save method. Not interpolated samples remain unchanged.
700 Default value is @code{a}.
704 Remove clipped samples from input audio.
706 Samples detected as clipped are replaced by interpolated samples using
707 autoregressive modelling.
711 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
712 Default value is @code{55} milliseconds.
713 This sets size of window which will be processed at once.
716 Set window overlap, in percentage of window size. Allowed range is from @code{50}
717 to @code{95}. Default value is @code{75} percent.
720 Set autoregression order, in percentage of window size. Allowed range is from
721 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
722 quality of interpolated samples using neighbour good samples.
725 Set threshold value. Allowed range is from @code{1} to @code{100}.
726 Default value is @code{10}. Higher values make clip detection less aggressive.
729 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
730 Default value is @code{1000}. Higher values make clip detection less aggressive.
735 It accepts the following values:
738 Select overlap-add method. Even not interpolated samples are slightly changed
742 Select overlap-save method. Not interpolated samples remain unchanged.
745 Default value is @code{a}.
750 Delay one or more audio channels.
752 Samples in delayed channel are filled with silence.
754 The filter accepts the following option:
758 Set list of delays in milliseconds for each channel separated by '|'.
759 Unused delays will be silently ignored. If number of given delays is
760 smaller than number of channels all remaining channels will not be delayed.
761 If you want to delay exact number of samples, append 'S' to number.
762 If you want instead to delay in seconds, append 's' to number.
765 Use last set delay for all remaining channels. By default is disabled.
766 This option if enabled changes how option @code{delays} is interpreted.
773 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
774 the second channel (and any other channels that may be present) unchanged.
780 Delay second channel by 500 samples, the third channel by 700 samples and leave
781 the first channel (and any other channels that may be present) unchanged.
787 Delay all channels by same number of samples:
789 adelay=delays=64S:all=1
794 Remedy denormals in audio by adding extremely low-level noise.
796 This filter shall be placed before any filter that can produce denormals.
798 A description of the accepted parameters follows.
802 Set level of added noise in dB. Default is @code{-351}.
803 Allowed range is from -451 to -90.
806 Set type of added noise.
819 Default is @code{dc}.
824 This filter supports the all above options as @ref{commands}.
826 @section aderivative, aintegral
828 Compute derivative/integral of audio stream.
830 Applying both filters one after another produces original audio.
834 Apply echoing to the input audio.
836 Echoes are reflected sound and can occur naturally amongst mountains
837 (and sometimes large buildings) when talking or shouting; digital echo
838 effects emulate this behaviour and are often used to help fill out the
839 sound of a single instrument or vocal. The time difference between the
840 original signal and the reflection is the @code{delay}, and the
841 loudness of the reflected signal is the @code{decay}.
842 Multiple echoes can have different delays and decays.
844 A description of the accepted parameters follows.
848 Set input gain of reflected signal. Default is @code{0.6}.
851 Set output gain of reflected signal. Default is @code{0.3}.
854 Set list of time intervals in milliseconds between original signal and reflections
855 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
856 Default is @code{1000}.
859 Set list of loudness of reflected signals separated by '|'.
860 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
861 Default is @code{0.5}.
868 Make it sound as if there are twice as many instruments as are actually playing:
870 aecho=0.8:0.88:60:0.4
874 If delay is very short, then it sounds like a (metallic) robot playing music:
880 A longer delay will sound like an open air concert in the mountains:
882 aecho=0.8:0.9:1000:0.3
886 Same as above but with one more mountain:
888 aecho=0.8:0.9:1000|1800:0.3|0.25
893 Audio emphasis filter creates or restores material directly taken from LPs or
894 emphased CDs with different filter curves. E.g. to store music on vinyl the
895 signal has to be altered by a filter first to even out the disadvantages of
896 this recording medium.
897 Once the material is played back the inverse filter has to be applied to
898 restore the distortion of the frequency response.
900 The filter accepts the following options:
910 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
911 use @code{production} mode. Default is @code{reproduction} mode.
914 Set filter type. Selects medium. Can be one of the following:
926 select Compact Disc (CD).
932 select 50µs (FM-KF).
934 select 75µs (FM-KF).
940 This filter supports the all above options as @ref{commands}.
944 Modify an audio signal according to the specified expressions.
946 This filter accepts one or more expressions (one for each channel),
947 which are evaluated and used to modify a corresponding audio signal.
949 It accepts the following parameters:
953 Set the '|'-separated expressions list for each separate channel. If
954 the number of input channels is greater than the number of
955 expressions, the last specified expression is used for the remaining
958 @item channel_layout, c
959 Set output channel layout. If not specified, the channel layout is
960 specified by the number of expressions. If set to @samp{same}, it will
961 use by default the same input channel layout.
964 Each expression in @var{exprs} can contain the following constants and functions:
968 channel number of the current expression
971 number of the evaluated sample, starting from 0
977 time of the evaluated sample expressed in seconds
980 @item nb_out_channels
981 input and output number of channels
984 the value of input channel with number @var{CH}
987 Note: this filter is slow. For faster processing you should use a
996 aeval=val(ch)/2:c=same
1000 Invert phase of the second channel:
1002 aeval=val(0)|-val(1)
1009 Apply fade-in/out effect to input audio.
1011 A description of the accepted parameters follows.
1015 Specify the effect type, can be either @code{in} for fade-in, or
1016 @code{out} for a fade-out effect. Default is @code{in}.
1018 @item start_sample, ss
1019 Specify the number of the start sample for starting to apply the fade
1020 effect. Default is 0.
1022 @item nb_samples, ns
1023 Specify the number of samples for which the fade effect has to last. At
1024 the end of the fade-in effect the output audio will have the same
1025 volume as the input audio, at the end of the fade-out transition
1026 the output audio will be silence. Default is 44100.
1028 @item start_time, st
1029 Specify the start time of the fade effect. Default is 0.
1030 The value must be specified as a time duration; see
1031 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1032 for the accepted syntax.
1033 If set this option is used instead of @var{start_sample}.
1036 Specify the duration of the fade effect. See
1037 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1038 for the accepted syntax.
1039 At the end of the fade-in effect the output audio will have the same
1040 volume as the input audio, at the end of the fade-out transition
1041 the output audio will be silence.
1042 By default the duration is determined by @var{nb_samples}.
1043 If set this option is used instead of @var{nb_samples}.
1046 Set curve for fade transition.
1048 It accepts the following values:
1051 select triangular, linear slope (default)
1053 select quarter of sine wave
1055 select half of sine wave
1057 select exponential sine wave
1061 select inverted parabola
1075 select inverted quarter of sine wave
1077 select inverted half of sine wave
1079 select double-exponential seat
1081 select double-exponential sigmoid
1083 select logistic sigmoid
1085 select sine cardinal function
1087 select inverted sine cardinal function
1093 @subsection Commands
1095 This filter supports the all above options as @ref{commands}.
1097 @subsection Examples
1101 Fade in first 15 seconds of audio:
1103 afade=t=in:ss=0:d=15
1107 Fade out last 25 seconds of a 900 seconds audio:
1109 afade=t=out:st=875:d=25
1114 Denoise audio samples with FFT.
1116 A description of the accepted parameters follows.
1120 Set the noise reduction in dB, allowed range is 0.01 to 97.
1121 Default value is 12 dB.
1124 Set the noise floor in dB, allowed range is -80 to -20.
1125 Default value is -50 dB.
1130 It accepts the following values:
1139 Select shellac noise.
1142 Select custom noise, defined in @code{bn} option.
1144 Default value is white noise.
1148 Set custom band noise for every one of 15 bands.
1149 Bands are separated by ' ' or '|'.
1152 Set the residual floor in dB, allowed range is -80 to -20.
1153 Default value is -38 dB.
1156 Enable noise tracking. By default is disabled.
1157 With this enabled, noise floor is automatically adjusted.
1160 Enable residual tracking. By default is disabled.
1163 Set the output mode.
1165 It accepts the following values:
1168 Pass input unchanged.
1171 Pass noise filtered out.
1176 Default value is @var{o}.
1180 @subsection Commands
1182 This filter supports the following commands:
1184 @item sample_noise, sn
1185 Start or stop measuring noise profile.
1186 Syntax for the command is : "start" or "stop" string.
1187 After measuring noise profile is stopped it will be
1188 automatically applied in filtering.
1190 @item noise_reduction, nr
1191 Change noise reduction. Argument is single float number.
1192 Syntax for the command is : "@var{noise_reduction}"
1194 @item noise_floor, nf
1195 Change noise floor. Argument is single float number.
1196 Syntax for the command is : "@var{noise_floor}"
1198 @item output_mode, om
1199 Change output mode operation.
1200 Syntax for the command is : "i", "o" or "n" string.
1204 Apply arbitrary expressions to samples in frequency domain.
1208 Set frequency domain real expression for each separate channel separated
1209 by '|'. Default is "re".
1210 If the number of input channels is greater than the number of
1211 expressions, the last specified expression is used for the remaining
1215 Set frequency domain imaginary expression for each separate channel
1216 separated by '|'. Default is "im".
1218 Each expression in @var{real} and @var{imag} can contain the following
1219 constants and functions:
1226 current frequency bin number
1229 number of available bins
1232 channel number of the current expression
1241 current real part of frequency bin of current channel
1244 current imaginary part of frequency bin of current channel
1247 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1250 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1254 Set window size. Allowed range is from 16 to 131072.
1255 Default is @code{4096}
1258 Set window function. Default is @code{hann}.
1261 Set window overlap. If set to 1, the recommended overlap for selected
1262 window function will be picked. Default is @code{0.75}.
1265 @subsection Examples
1269 Leave almost only low frequencies in audio:
1271 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1275 Apply robotize effect:
1277 afftfilt="real='hypot(re,im)*sin(0)':imag='hypot(re,im)*cos(0)':win_size=512:overlap=0.75"
1281 Apply whisper effect:
1283 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"
1290 Apply an arbitrary Finite Impulse Response filter.
1292 This filter is designed for applying long FIR filters,
1293 up to 60 seconds long.
1295 It can be used as component for digital crossover filters,
1296 room equalization, cross talk cancellation, wavefield synthesis,
1297 auralization, ambiophonics, ambisonics and spatialization.
1299 This filter uses the streams higher than first one as FIR coefficients.
1300 If the non-first stream holds a single channel, it will be used
1301 for all input channels in the first stream, otherwise
1302 the number of channels in the non-first stream must be same as
1303 the number of channels in the first stream.
1305 It accepts the following parameters:
1309 Set dry gain. This sets input gain.
1312 Set wet gain. This sets final output gain.
1315 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1318 Enable applying gain measured from power of IR.
1320 Set which approach to use for auto gain measurement.
1324 Do not apply any gain.
1327 select peak gain, very conservative approach. This is default value.
1330 select DC gain, limited application.
1333 select gain to noise approach, this is most popular one.
1337 Set gain to be applied to IR coefficients before filtering.
1338 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1341 Set format of IR stream. Can be @code{mono} or @code{input}.
1342 Default is @code{input}.
1345 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1346 Allowed range is 0.1 to 60 seconds.
1349 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1350 By default it is disabled.
1353 Set for which IR channel to display frequency response. By default is first channel
1354 displayed. This option is used only when @var{response} is enabled.
1357 Set video stream size. This option is used only when @var{response} is enabled.
1360 Set video stream frame rate. This option is used only when @var{response} is enabled.
1363 Set minimal partition size used for convolution. Default is @var{8192}.
1364 Allowed range is from @var{1} to @var{32768}.
1365 Lower values decreases latency at cost of higher CPU usage.
1368 Set maximal partition size used for convolution. Default is @var{8192}.
1369 Allowed range is from @var{8} to @var{32768}.
1370 Lower values may increase CPU usage.
1373 Set number of input impulse responses streams which will be switchable at runtime.
1374 Allowed range is from @var{1} to @var{32}. Default is @var{1}.
1377 Set IR stream which will be used for convolution, starting from @var{0}, should always be
1378 lower than supplied value by @code{nbirs} option. Default is @var{0}.
1379 This option can be changed at runtime via @ref{commands}.
1382 @subsection Examples
1386 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1388 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1395 Set output format constraints for the input audio. The framework will
1396 negotiate the most appropriate format to minimize conversions.
1398 It accepts the following parameters:
1401 @item sample_fmts, f
1402 A '|'-separated list of requested sample formats.
1404 @item sample_rates, r
1405 A '|'-separated list of requested sample rates.
1407 @item channel_layouts, cl
1408 A '|'-separated list of requested channel layouts.
1410 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1411 for the required syntax.
1414 If a parameter is omitted, all values are allowed.
1416 Force the output to either unsigned 8-bit or signed 16-bit stereo
1418 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1422 Apply frequency shift to input audio samples.
1424 The filter accepts the following options:
1428 Specify frequency shift. Allowed range is -INT_MAX to INT_MAX.
1429 Default value is 0.0.
1432 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
1433 Default value is 1.0.
1436 @subsection Commands
1438 This filter supports the all above options as @ref{commands}.
1442 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1443 processing reduces disturbing noise between useful signals.
1445 Gating is done by detecting the volume below a chosen level @var{threshold}
1446 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1447 floor is set via @var{range}. Because an exact manipulation of the signal
1448 would cause distortion of the waveform the reduction can be levelled over
1449 time. This is done by setting @var{attack} and @var{release}.
1451 @var{attack} determines how long the signal has to fall below the threshold
1452 before any reduction will occur and @var{release} sets the time the signal
1453 has to rise above the threshold to reduce the reduction again.
1454 Shorter signals than the chosen attack time will be left untouched.
1458 Set input level before filtering.
1459 Default is 1. Allowed range is from 0.015625 to 64.
1462 Set the mode of operation. Can be @code{upward} or @code{downward}.
1463 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1464 will be amplified, expanding dynamic range in upward direction.
1465 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1468 Set the level of gain reduction when the signal is below the threshold.
1469 Default is 0.06125. Allowed range is from 0 to 1.
1470 Setting this to 0 disables reduction and then filter behaves like expander.
1473 If a signal rises above this level the gain reduction is released.
1474 Default is 0.125. Allowed range is from 0 to 1.
1477 Set a ratio by which the signal is reduced.
1478 Default is 2. Allowed range is from 1 to 9000.
1481 Amount of milliseconds the signal has to rise above the threshold before gain
1483 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1486 Amount of milliseconds the signal has to fall below the threshold before the
1487 reduction is increased again. Default is 250 milliseconds.
1488 Allowed range is from 0.01 to 9000.
1491 Set amount of amplification of signal after processing.
1492 Default is 1. Allowed range is from 1 to 64.
1495 Curve the sharp knee around the threshold to enter gain reduction more softly.
1496 Default is 2.828427125. Allowed range is from 1 to 8.
1499 Choose if exact signal should be taken for detection or an RMS like one.
1500 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1503 Choose if the average level between all channels or the louder channel affects
1505 Default is @code{average}. Can be @code{average} or @code{maximum}.
1508 @subsection Commands
1510 This filter supports the all above options as @ref{commands}.
1514 Apply an arbitrary Infinite Impulse Response filter.
1516 It accepts the following parameters:
1520 Set B/numerator/zeros/reflection coefficients.
1523 Set A/denominator/poles/ladder coefficients.
1535 Set coefficients format.
1539 lattice-ladder function
1541 analog transfer function
1543 digital transfer function
1545 Z-plane zeros/poles, cartesian (default)
1547 Z-plane zeros/poles, polar radians
1549 Z-plane zeros/poles, polar degrees
1555 Set type of processing.
1567 Set filtering precision.
1571 double-precision floating-point (default)
1573 single-precision floating-point
1581 Normalize filter coefficients, by default is enabled.
1582 Enabling it will normalize magnitude response at DC to 0dB.
1585 How much to use filtered signal in output. Default is 1.
1586 Range is between 0 and 1.
1589 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1590 By default it is disabled.
1593 Set for which IR channel to display frequency response. By default is first channel
1594 displayed. This option is used only when @var{response} is enabled.
1597 Set video stream size. This option is used only when @var{response} is enabled.
1600 Coefficients in @code{tf} and @code{sf} format are separated by spaces and are in ascending
1603 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1604 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1607 Different coefficients and gains can be provided for every channel, in such case
1608 use '|' to separate coefficients or gains. Last provided coefficients will be
1609 used for all remaining channels.
1611 @subsection Examples
1615 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1617 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
1621 Same as above but in @code{zp} format:
1623 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
1627 Apply 3-rd order analog normalized Butterworth low-pass filter, using analog transfer function format:
1629 aiir=z=1.3057 0 0 0:p=1.3057 2.3892 2.1860 1:f=sf:r=d
1635 The limiter prevents an input signal from rising over a desired threshold.
1636 This limiter uses lookahead technology to prevent your signal from distorting.
1637 It means that there is a small delay after the signal is processed. Keep in mind
1638 that the delay it produces is the attack time you set.
1640 The filter accepts the following options:
1644 Set input gain. Default is 1.
1647 Set output gain. Default is 1.
1650 Don't let signals above this level pass the limiter. Default is 1.
1653 The limiter will reach its attenuation level in this amount of time in
1654 milliseconds. Default is 5 milliseconds.
1657 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1658 Default is 50 milliseconds.
1661 When gain reduction is always needed ASC takes care of releasing to an
1662 average reduction level rather than reaching a reduction of 0 in the release
1666 Select how much the release time is affected by ASC, 0 means nearly no changes
1667 in release time while 1 produces higher release times.
1670 Auto level output signal. Default is enabled.
1671 This normalizes audio back to 0dB if enabled.
1674 Depending on picked setting it is recommended to upsample input 2x or 4x times
1675 with @ref{aresample} before applying this filter.
1679 Apply a two-pole all-pass filter with central frequency (in Hz)
1680 @var{frequency}, and filter-width @var{width}.
1681 An all-pass filter changes the audio's frequency to phase relationship
1682 without changing its frequency to amplitude relationship.
1684 The filter accepts the following options:
1688 Set frequency in Hz.
1691 Set method to specify band-width of filter.
1706 Specify the band-width of a filter in width_type units.
1709 How much to use filtered signal in output. Default is 1.
1710 Range is between 0 and 1.
1713 Specify which channels to filter, by default all available are filtered.
1716 Normalize biquad coefficients, by default is disabled.
1717 Enabling it will normalize magnitude response at DC to 0dB.
1720 Set the filter order, can be 1 or 2. Default is 2.
1723 Set transform type of IIR filter.
1732 Set precison of filtering.
1735 Pick automatic sample format depending on surround filters.
1737 Always use signed 16-bit.
1739 Always use signed 32-bit.
1741 Always use float 32-bit.
1743 Always use float 64-bit.
1747 @subsection Commands
1749 This filter supports the following commands:
1752 Change allpass frequency.
1753 Syntax for the command is : "@var{frequency}"
1756 Change allpass width_type.
1757 Syntax for the command is : "@var{width_type}"
1760 Change allpass width.
1761 Syntax for the command is : "@var{width}"
1765 Syntax for the command is : "@var{mix}"
1772 The filter accepts the following options:
1776 Set the number of loops. Setting this value to -1 will result in infinite loops.
1780 Set maximal number of samples. Default is 0.
1783 Set first sample of loop. Default is 0.
1789 Merge two or more audio streams into a single multi-channel stream.
1791 The filter accepts the following options:
1796 Set the number of inputs. Default is 2.
1800 If the channel layouts of the inputs are disjoint, and therefore compatible,
1801 the channel layout of the output will be set accordingly and the channels
1802 will be reordered as necessary. If the channel layouts of the inputs are not
1803 disjoint, the output will have all the channels of the first input then all
1804 the channels of the second input, in that order, and the channel layout of
1805 the output will be the default value corresponding to the total number of
1808 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1809 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1810 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1811 first input, b1 is the first channel of the second input).
1813 On the other hand, if both input are in stereo, the output channels will be
1814 in the default order: a1, a2, b1, b2, and the channel layout will be
1815 arbitrarily set to 4.0, which may or may not be the expected value.
1817 All inputs must have the same sample rate, and format.
1819 If inputs do not have the same duration, the output will stop with the
1822 @subsection Examples
1826 Merge two mono files into a stereo stream:
1828 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1832 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1834 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
1840 Mixes multiple audio inputs into a single output.
1842 Note that this filter only supports float samples (the @var{amerge}
1843 and @var{pan} audio filters support many formats). If the @var{amix}
1844 input has integer samples then @ref{aresample} will be automatically
1845 inserted to perform the conversion to float samples.
1849 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1851 will mix 3 input audio streams to a single output with the same duration as the
1852 first input and a dropout transition time of 3 seconds.
1854 It accepts the following parameters:
1858 The number of inputs. If unspecified, it defaults to 2.
1861 How to determine the end-of-stream.
1865 The duration of the longest input. (default)
1868 The duration of the shortest input.
1871 The duration of the first input.
1875 @item dropout_transition
1876 The transition time, in seconds, for volume renormalization when an input
1877 stream ends. The default value is 2 seconds.
1880 Specify weight of each input audio stream as sequence.
1881 Each weight is separated by space. By default all inputs have same weight.
1884 Do not scale inputs but instead do only summation of samples.
1885 Beware of heavy clipping if inputs are not normalized prior of filtering
1886 or output from @var{amix} normalized after filtering. By default is disabled.
1889 @subsection Commands
1891 This filter supports the following commands:
1895 Syntax is same as option with same name.
1900 Multiply first audio stream with second audio stream and store result
1901 in output audio stream. Multiplication is done by multiplying each
1902 sample from first stream with sample at same position from second stream.
1904 With this element-wise multiplication one can create amplitude fades and
1905 amplitude modulations.
1907 @section anequalizer
1909 High-order parametric multiband equalizer for each channel.
1911 It accepts the following parameters:
1915 This option string is in format:
1916 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1917 Each equalizer band is separated by '|'.
1921 Set channel number to which equalization will be applied.
1922 If input doesn't have that channel the entry is ignored.
1925 Set central frequency for band.
1926 If input doesn't have that frequency the entry is ignored.
1929 Set band width in Hertz.
1932 Set band gain in dB.
1935 Set filter type for band, optional, can be:
1939 Butterworth, this is default.
1950 With this option activated frequency response of anequalizer is displayed
1954 Set video stream size. Only useful if curves option is activated.
1957 Set max gain that will be displayed. Only useful if curves option is activated.
1958 Setting this to a reasonable value makes it possible to display gain which is derived from
1959 neighbour bands which are too close to each other and thus produce higher gain
1960 when both are activated.
1963 Set frequency scale used to draw frequency response in video output.
1964 Can be linear or logarithmic. Default is logarithmic.
1967 Set color for each channel curve which is going to be displayed in video stream.
1968 This is list of color names separated by space or by '|'.
1969 Unrecognised or missing colors will be replaced by white color.
1972 @subsection Examples
1976 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1977 for first 2 channels using Chebyshev type 1 filter:
1979 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1983 @subsection Commands
1985 This filter supports the following commands:
1988 Alter existing filter parameters.
1989 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1991 @var{fN} is existing filter number, starting from 0, if no such filter is available
1993 @var{freq} set new frequency parameter.
1994 @var{width} set new width parameter in Hertz.
1995 @var{gain} set new gain parameter in dB.
1997 Full filter invocation with asendcmd may look like this:
1998 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
2003 Reduce broadband noise in audio samples using Non-Local Means algorithm.
2005 Each sample is adjusted by looking for other samples with similar contexts. This
2006 context similarity is defined by comparing their surrounding patches of size
2007 @option{p}. Patches are searched in an area of @option{r} around the sample.
2009 The filter accepts the following options:
2013 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
2016 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
2017 Default value is 2 milliseconds.
2020 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
2021 Default value is 6 milliseconds.
2024 Set the output mode.
2026 It accepts the following values:
2029 Pass input unchanged.
2032 Pass noise filtered out.
2037 Default value is @var{o}.
2041 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
2044 @subsection Commands
2046 This filter supports the all above options as @ref{commands}.
2049 Apply Normalized Least-Mean-Squares algorithm to the first audio stream using the second audio stream.
2051 This adaptive filter is used to mimic a desired filter by finding the filter coefficients that
2052 relate to producing the least mean square of the error signal (difference between the desired,
2053 2nd input audio stream and the actual signal, the 1st input audio stream).
2055 A description of the accepted options follows.
2068 Set the filter leakage.
2071 It accepts the following values:
2080 Pass filtered samples.
2083 Pass difference between desired and filtered samples.
2085 Default value is @var{o}.
2089 @subsection Examples
2093 One of many usages of this filter is noise reduction, input audio is filtered
2094 with same samples that are delayed by fixed amount, one such example for stereo audio is:
2096 asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
2100 @subsection Commands
2102 This filter supports the same commands as options, excluding option @code{order}.
2106 Pass the audio source unchanged to the output.
2110 Pad the end of an audio stream with silence.
2112 This can be used together with @command{ffmpeg} @option{-shortest} to
2113 extend audio streams to the same length as the video stream.
2115 A description of the accepted options follows.
2119 Set silence packet size. Default value is 4096.
2122 Set the number of samples of silence to add to the end. After the
2123 value is reached, the stream is terminated. This option is mutually
2124 exclusive with @option{whole_len}.
2127 Set the minimum total number of samples in the output audio stream. If
2128 the value is longer than the input audio length, silence is added to
2129 the end, until the value is reached. This option is mutually exclusive
2130 with @option{pad_len}.
2133 Specify the duration of samples of silence to add. See
2134 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2135 for the accepted syntax. Used only if set to non-zero value.
2138 Specify the minimum total duration in the output audio stream. See
2139 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2140 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
2141 the input audio length, silence is added to the end, until the value is reached.
2142 This option is mutually exclusive with @option{pad_dur}
2145 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
2146 nor @option{whole_dur} option is set, the filter will add silence to the end of
2147 the input stream indefinitely.
2149 @subsection Examples
2153 Add 1024 samples of silence to the end of the input:
2159 Make sure the audio output will contain at least 10000 samples, pad
2160 the input with silence if required:
2162 apad=whole_len=10000
2166 Use @command{ffmpeg} to pad the audio input with silence, so that the
2167 video stream will always result the shortest and will be converted
2168 until the end in the output file when using the @option{shortest}
2171 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
2176 Add a phasing effect to the input audio.
2178 A phaser filter creates series of peaks and troughs in the frequency spectrum.
2179 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
2181 A description of the accepted parameters follows.
2185 Set input gain. Default is 0.4.
2188 Set output gain. Default is 0.74
2191 Set delay in milliseconds. Default is 3.0.
2194 Set decay. Default is 0.4.
2197 Set modulation speed in Hz. Default is 0.5.
2200 Set modulation type. Default is triangular.
2202 It accepts the following values:
2209 @section aphaseshift
2210 Apply phase shift to input audio samples.
2212 The filter accepts the following options:
2216 Specify phase shift. Allowed range is from -1.0 to 1.0.
2217 Default value is 0.0.
2220 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
2221 Default value is 1.0.
2224 @subsection Commands
2226 This filter supports the all above options as @ref{commands}.
2230 Audio pulsator is something between an autopanner and a tremolo.
2231 But it can produce funny stereo effects as well. Pulsator changes the volume
2232 of the left and right channel based on a LFO (low frequency oscillator) with
2233 different waveforms and shifted phases.
2234 This filter have the ability to define an offset between left and right
2235 channel. An offset of 0 means that both LFO shapes match each other.
2236 The left and right channel are altered equally - a conventional tremolo.
2237 An offset of 50% means that the shape of the right channel is exactly shifted
2238 in phase (or moved backwards about half of the frequency) - pulsator acts as
2239 an autopanner. At 1 both curves match again. Every setting in between moves the
2240 phase shift gapless between all stages and produces some "bypassing" sounds with
2241 sine and triangle waveforms. The more you set the offset near 1 (starting from
2242 the 0.5) the faster the signal passes from the left to the right speaker.
2244 The filter accepts the following options:
2248 Set input gain. By default it is 1. Range is [0.015625 - 64].
2251 Set output gain. By default it is 1. Range is [0.015625 - 64].
2254 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
2255 sawup or sawdown. Default is sine.
2258 Set modulation. Define how much of original signal is affected by the LFO.
2261 Set left channel offset. Default is 0. Allowed range is [0 - 1].
2264 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
2267 Set pulse width. Default is 1. Allowed range is [0 - 2].
2270 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
2273 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
2277 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
2281 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
2282 if timing is set to hz.
2288 Resample the input audio to the specified parameters, using the
2289 libswresample library. If none are specified then the filter will
2290 automatically convert between its input and output.
2292 This filter is also able to stretch/squeeze the audio data to make it match
2293 the timestamps or to inject silence / cut out audio to make it match the
2294 timestamps, do a combination of both or do neither.
2296 The filter accepts the syntax
2297 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
2298 expresses a sample rate and @var{resampler_options} is a list of
2299 @var{key}=@var{value} pairs, separated by ":". See the
2300 @ref{Resampler Options,,"Resampler Options" section in the
2301 ffmpeg-resampler(1) manual,ffmpeg-resampler}
2302 for the complete list of supported options.
2304 @subsection Examples
2308 Resample the input audio to 44100Hz:
2314 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2315 samples per second compensation:
2317 aresample=async=1000
2323 Reverse an audio clip.
2325 Warning: This filter requires memory to buffer the entire clip, so trimming
2328 @subsection Examples
2332 Take the first 5 seconds of a clip, and reverse it.
2334 atrim=end=5,areverse
2340 Reduce noise from speech using Recurrent Neural Networks.
2342 This filter accepts the following options:
2346 Set train model file to load. This option is always required.
2349 Set how much to mix filtered samples into final output.
2350 Allowed range is from -1 to 1. Default value is 1.
2351 Negative values are special, they set how much to keep filtered noise
2352 in the final filter output. Set this option to -1 to hear actual
2353 noise removed from input signal.
2356 @subsection Commands
2358 This filter supports the all above options as @ref{commands}.
2360 @section asetnsamples
2362 Set the number of samples per each output audio frame.
2364 The last output packet may contain a different number of samples, as
2365 the filter will flush all the remaining samples when the input audio
2368 The filter accepts the following options:
2372 @item nb_out_samples, n
2373 Set the number of frames per each output audio frame. The number is
2374 intended as the number of samples @emph{per each channel}.
2375 Default value is 1024.
2378 If set to 1, the filter will pad the last audio frame with zeroes, so
2379 that the last frame will contain the same number of samples as the
2380 previous ones. Default value is 1.
2383 For example, to set the number of per-frame samples to 1234 and
2384 disable padding for the last frame, use:
2386 asetnsamples=n=1234:p=0
2391 Set the sample rate without altering the PCM data.
2392 This will result in a change of speed and pitch.
2394 The filter accepts the following options:
2397 @item sample_rate, r
2398 Set the output sample rate. Default is 44100 Hz.
2403 Show a line containing various information for each input audio frame.
2404 The input audio is not modified.
2406 The shown line contains a sequence of key/value pairs of the form
2407 @var{key}:@var{value}.
2409 The following values are shown in the output:
2413 The (sequential) number of the input frame, starting from 0.
2416 The presentation timestamp of the input frame, in time base units; the time base
2417 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2420 The presentation timestamp of the input frame in seconds.
2423 position of the frame in the input stream, -1 if this information in
2424 unavailable and/or meaningless (for example in case of synthetic audio)
2433 The sample rate for the audio frame.
2436 The number of samples (per channel) in the frame.
2439 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2440 audio, the data is treated as if all the planes were concatenated.
2442 @item plane_checksums
2443 A list of Adler-32 checksums for each data plane.
2447 Apply audio soft clipping.
2449 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2450 along a smooth curve, rather than the abrupt shape of hard-clipping.
2452 This filter accepts the following options:
2456 Set type of soft-clipping.
2458 It accepts the following values:
2472 Set threshold from where to start clipping. Default value is 0dB or 1.
2475 Set gain applied to output. Default value is 0dB or 1.
2478 Set additional parameter which controls sigmoid function.
2481 Set oversampling factor.
2484 @subsection Commands
2486 This filter supports the all above options as @ref{commands}.
2489 Automatic Speech Recognition
2491 This filter uses PocketSphinx for speech recognition. To enable
2492 compilation of this filter, you need to configure FFmpeg with
2493 @code{--enable-pocketsphinx}.
2495 It accepts the following options:
2499 Set sampling rate of input audio. Defaults is @code{16000}.
2500 This need to match speech models, otherwise one will get poor results.
2503 Set dictionary containing acoustic model files.
2506 Set pronunciation dictionary.
2509 Set language model file.
2512 Set language model set.
2515 Set which language model to use.
2518 Set output for log messages.
2521 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2526 Display time domain statistical information about the audio channels.
2527 Statistics are calculated and displayed for each audio channel and,
2528 where applicable, an overall figure is also given.
2530 It accepts the following option:
2533 Short window length in seconds, used for peak and trough RMS measurement.
2534 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2538 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2539 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2542 Available keys for each channel are:
2588 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2589 this @code{lavfi.astats.Overall.Peak_count}.
2591 For description what each key means read below.
2594 Set number of frame after which stats are going to be recalculated.
2595 Default is disabled.
2597 @item measure_perchannel
2598 Select the entries which need to be measured per channel. The metadata keys can
2599 be used as flags, default is @option{all} which measures everything.
2600 @option{none} disables all per channel measurement.
2602 @item measure_overall
2603 Select the entries which need to be measured overall. The metadata keys can
2604 be used as flags, default is @option{all} which measures everything.
2605 @option{none} disables all overall measurement.
2609 A description of each shown parameter follows:
2613 Mean amplitude displacement from zero.
2616 Minimal sample level.
2619 Maximal sample level.
2621 @item Min difference
2622 Minimal difference between two consecutive samples.
2624 @item Max difference
2625 Maximal difference between two consecutive samples.
2627 @item Mean difference
2628 Mean difference between two consecutive samples.
2629 The average of each difference between two consecutive samples.
2631 @item RMS difference
2632 Root Mean Square difference between two consecutive samples.
2636 Standard peak and RMS level measured in dBFS.
2640 Peak and trough values for RMS level measured over a short window.
2643 Standard ratio of peak to RMS level (note: not in dB).
2646 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2647 (i.e. either @var{Min level} or @var{Max level}).
2650 Number of occasions (not the number of samples) that the signal attained either
2651 @var{Min level} or @var{Max level}.
2653 @item Noise floor dB
2654 Minimum local peak measured in dBFS over a short window.
2656 @item Noise floor count
2657 Number of occasions (not the number of samples) that the signal attained
2661 Overall bit depth of audio. Number of bits used for each sample.
2664 Measured dynamic range of audio in dB.
2666 @item Zero crossings
2667 Number of points where the waveform crosses the zero level axis.
2669 @item Zero crossings rate
2670 Rate of Zero crossings and number of audio samples.
2674 Boost subwoofer frequencies.
2676 The filter accepts the following options:
2680 Set dry gain, how much of original signal is kept. Allowed range is from 0 to 1.
2681 Default value is 0.7.
2684 Set wet gain, how much of filtered signal is kept. Allowed range is from 0 to 1.
2685 Default value is 0.7.
2688 Set delay line decay gain value. Allowed range is from 0 to 1.
2689 Default value is 0.7.
2692 Set delay line feedback gain value. Allowed range is from 0 to 1.
2693 Default value is 0.9.
2696 Set cutoff frequency in Hertz. Allowed range is 50 to 900.
2697 Default value is 100.
2700 Set slope amount for cutoff frequency. Allowed range is 0.0001 to 1.
2701 Default value is 0.5.
2704 Set delay. Allowed range is from 1 to 100.
2705 Default value is 20.
2708 @subsection Commands
2710 This filter supports the all above options as @ref{commands}.
2713 Cut subwoofer frequencies.
2715 This filter allows to set custom, steeper
2716 roll off than highpass filter, and thus is able to more attenuate
2717 frequency content in stop-band.
2719 The filter accepts the following options:
2723 Set cutoff frequency in Hertz. Allowed range is 2 to 200.
2724 Default value is 20.
2727 Set filter order. Available values are from 3 to 20.
2728 Default value is 10.
2731 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2734 @subsection Commands
2736 This filter supports the all above options as @ref{commands}.
2739 Cut super frequencies.
2741 The filter accepts the following options:
2745 Set cutoff frequency in Hertz. Allowed range is 20000 to 192000.
2746 Default value is 20000.
2749 Set filter order. Available values are from 3 to 20.
2750 Default value is 10.
2753 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2756 @subsection Commands
2758 This filter supports the all above options as @ref{commands}.
2761 Apply high order Butterworth band-pass filter.
2763 The filter accepts the following options:
2767 Set center frequency in Hertz. Allowed range is 2 to 999999.
2768 Default value is 1000.
2771 Set filter order. Available values are from 4 to 20.
2775 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2778 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2781 @subsection Commands
2783 This filter supports the all above options as @ref{commands}.
2786 Apply high order Butterworth band-stop filter.
2788 The filter accepts the following options:
2792 Set center frequency in Hertz. Allowed range is 2 to 999999.
2793 Default value is 1000.
2796 Set filter order. Available values are from 4 to 20.
2800 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2803 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2806 @subsection Commands
2808 This filter supports the all above options as @ref{commands}.
2814 The filter accepts exactly one parameter, the audio tempo. If not
2815 specified then the filter will assume nominal 1.0 tempo. Tempo must
2816 be in the [0.5, 100.0] range.
2818 Note that tempo greater than 2 will skip some samples rather than
2819 blend them in. If for any reason this is a concern it is always
2820 possible to daisy-chain several instances of atempo to achieve the
2821 desired product tempo.
2823 @subsection Examples
2827 Slow down audio to 80% tempo:
2833 To speed up audio to 300% tempo:
2839 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2841 atempo=sqrt(3),atempo=sqrt(3)
2845 @subsection Commands
2847 This filter supports the following commands:
2850 Change filter tempo scale factor.
2851 Syntax for the command is : "@var{tempo}"
2856 Trim the input so that the output contains one continuous subpart of the input.
2858 It accepts the following parameters:
2861 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2862 sample with the timestamp @var{start} will be the first sample in the output.
2865 Specify time of the first audio sample that will be dropped, i.e. the
2866 audio sample immediately preceding the one with the timestamp @var{end} will be
2867 the last sample in the output.
2870 Same as @var{start}, except this option sets the start timestamp in samples
2874 Same as @var{end}, except this option sets the end timestamp in samples instead
2878 The maximum duration of the output in seconds.
2881 The number of the first sample that should be output.
2884 The number of the first sample that should be dropped.
2887 @option{start}, @option{end}, and @option{duration} are expressed as time
2888 duration specifications; see
2889 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2891 Note that the first two sets of the start/end options and the @option{duration}
2892 option look at the frame timestamp, while the _sample options simply count the
2893 samples that pass through the filter. So start/end_pts and start/end_sample will
2894 give different results when the timestamps are wrong, inexact or do not start at
2895 zero. Also note that this filter does not modify the timestamps. If you wish
2896 to have the output timestamps start at zero, insert the asetpts filter after the
2899 If multiple start or end options are set, this filter tries to be greedy and
2900 keep all samples that match at least one of the specified constraints. To keep
2901 only the part that matches all the constraints at once, chain multiple atrim
2904 The defaults are such that all the input is kept. So it is possible to set e.g.
2905 just the end values to keep everything before the specified time.
2910 Drop everything except the second minute of input:
2912 ffmpeg -i INPUT -af atrim=60:120
2916 Keep only the first 1000 samples:
2918 ffmpeg -i INPUT -af atrim=end_sample=1000
2923 @section axcorrelate
2924 Calculate normalized cross-correlation between two input audio streams.
2926 Resulted samples are always between -1 and 1 inclusive.
2927 If result is 1 it means two input samples are highly correlated in that selected segment.
2928 Result 0 means they are not correlated at all.
2929 If result is -1 it means two input samples are out of phase, which means they cancel each
2932 The filter accepts the following options:
2936 Set size of segment over which cross-correlation is calculated.
2937 Default is 256. Allowed range is from 2 to 131072.
2940 Set algorithm for cross-correlation. Can be @code{slow} or @code{fast}.
2941 Default is @code{slow}. Fast algorithm assumes mean values over any given segment
2942 are always zero and thus need much less calculations to make.
2943 This is generally not true, but is valid for typical audio streams.
2946 @subsection Examples
2950 Calculate correlation between channels in stereo audio stream:
2952 ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
2958 Apply a two-pole Butterworth band-pass filter with central
2959 frequency @var{frequency}, and (3dB-point) band-width width.
2960 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2961 instead of the default: constant 0dB peak gain.
2962 The filter roll off at 6dB per octave (20dB per decade).
2964 The filter accepts the following options:
2968 Set the filter's central frequency. Default is @code{3000}.
2971 Constant skirt gain if set to 1. Defaults to 0.
2974 Set method to specify band-width of filter.
2989 Specify the band-width of a filter in width_type units.
2992 How much to use filtered signal in output. Default is 1.
2993 Range is between 0 and 1.
2996 Specify which channels to filter, by default all available are filtered.
2999 Normalize biquad coefficients, by default is disabled.
3000 Enabling it will normalize magnitude response at DC to 0dB.
3003 Set transform type of IIR filter.
3012 Set precison of filtering.
3015 Pick automatic sample format depending on surround filters.
3017 Always use signed 16-bit.
3019 Always use signed 32-bit.
3021 Always use float 32-bit.
3023 Always use float 64-bit.
3027 @subsection Commands
3029 This filter supports the following commands:
3032 Change bandpass frequency.
3033 Syntax for the command is : "@var{frequency}"
3036 Change bandpass width_type.
3037 Syntax for the command is : "@var{width_type}"
3040 Change bandpass width.
3041 Syntax for the command is : "@var{width}"
3044 Change bandpass mix.
3045 Syntax for the command is : "@var{mix}"
3050 Apply a two-pole Butterworth band-reject filter with central
3051 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
3052 The filter roll off at 6dB per octave (20dB per decade).
3054 The filter accepts the following options:
3058 Set the filter's central frequency. Default is @code{3000}.
3061 Set method to specify band-width of filter.
3076 Specify the band-width of a filter in width_type units.
3079 How much to use filtered signal in output. Default is 1.
3080 Range is between 0 and 1.
3083 Specify which channels to filter, by default all available are filtered.
3086 Normalize biquad coefficients, by default is disabled.
3087 Enabling it will normalize magnitude response at DC to 0dB.
3090 Set transform type of IIR filter.
3099 Set precison of filtering.
3102 Pick automatic sample format depending on surround filters.
3104 Always use signed 16-bit.
3106 Always use signed 32-bit.
3108 Always use float 32-bit.
3110 Always use float 64-bit.
3114 @subsection Commands
3116 This filter supports the following commands:
3119 Change bandreject frequency.
3120 Syntax for the command is : "@var{frequency}"
3123 Change bandreject width_type.
3124 Syntax for the command is : "@var{width_type}"
3127 Change bandreject width.
3128 Syntax for the command is : "@var{width}"
3131 Change bandreject mix.
3132 Syntax for the command is : "@var{mix}"
3135 @section bass, lowshelf
3137 Boost or cut the bass (lower) frequencies of the audio using a two-pole
3138 shelving filter with a response similar to that of a standard
3139 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3141 The filter accepts the following options:
3145 Give the gain at 0 Hz. Its useful range is about -20
3146 (for a large cut) to +20 (for a large boost).
3147 Beware of clipping when using a positive gain.
3150 Set the filter's central frequency and so can be used
3151 to extend or reduce the frequency range to be boosted or cut.
3152 The default value is @code{100} Hz.
3155 Set method to specify band-width of filter.
3170 Determine how steep is the filter's shelf transition.
3173 Set number of poles. Default is 2.
3176 How much to use filtered signal in output. Default is 1.
3177 Range is between 0 and 1.
3180 Specify which channels to filter, by default all available are filtered.
3183 Normalize biquad coefficients, by default is disabled.
3184 Enabling it will normalize magnitude response at DC to 0dB.
3187 Set transform type of IIR filter.
3196 Set precison of filtering.
3199 Pick automatic sample format depending on surround filters.
3201 Always use signed 16-bit.
3203 Always use signed 32-bit.
3205 Always use float 32-bit.
3207 Always use float 64-bit.
3211 @subsection Commands
3213 This filter supports the following commands:
3216 Change bass frequency.
3217 Syntax for the command is : "@var{frequency}"
3220 Change bass width_type.
3221 Syntax for the command is : "@var{width_type}"
3225 Syntax for the command is : "@var{width}"
3229 Syntax for the command is : "@var{gain}"
3233 Syntax for the command is : "@var{mix}"
3238 Apply a biquad IIR filter with the given coefficients.
3239 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
3240 are the numerator and denominator coefficients respectively.
3241 and @var{channels}, @var{c} specify which channels to filter, by default all
3242 available are filtered.
3244 @subsection Commands
3246 This filter supports the following commands:
3254 Change biquad parameter.
3255 Syntax for the command is : "@var{value}"
3258 How much to use filtered signal in output. Default is 1.
3259 Range is between 0 and 1.
3262 Specify which channels to filter, by default all available are filtered.
3265 Normalize biquad coefficients, by default is disabled.
3266 Enabling it will normalize magnitude response at DC to 0dB.
3269 Set transform type of IIR filter.
3278 Set precison of filtering.
3281 Pick automatic sample format depending on surround filters.
3283 Always use signed 16-bit.
3285 Always use signed 32-bit.
3287 Always use float 32-bit.
3289 Always use float 64-bit.
3294 Bauer stereo to binaural transformation, which improves headphone listening of
3295 stereo audio records.
3297 To enable compilation of this filter you need to configure FFmpeg with
3298 @code{--enable-libbs2b}.
3300 It accepts the following parameters:
3304 Pre-defined crossfeed level.
3308 Default level (fcut=700, feed=50).
3311 Chu Moy circuit (fcut=700, feed=60).
3314 Jan Meier circuit (fcut=650, feed=95).
3319 Cut frequency (in Hz).
3328 Remap input channels to new locations.
3330 It accepts the following parameters:
3333 Map channels from input to output. The argument is a '|'-separated list of
3334 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
3335 @var{in_channel} form. @var{in_channel} can be either the name of the input
3336 channel (e.g. FL for front left) or its index in the input channel layout.
3337 @var{out_channel} is the name of the output channel or its index in the output
3338 channel layout. If @var{out_channel} is not given then it is implicitly an
3339 index, starting with zero and increasing by one for each mapping.
3341 @item channel_layout
3342 The channel layout of the output stream.
3345 If no mapping is present, the filter will implicitly map input channels to
3346 output channels, preserving indices.
3348 @subsection Examples
3352 For example, assuming a 5.1+downmix input MOV file,
3354 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
3356 will create an output WAV file tagged as stereo from the downmix channels of
3360 To fix a 5.1 WAV improperly encoded in AAC's native channel order
3362 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
3366 @section channelsplit
3368 Split each channel from an input audio stream into a separate output stream.
3370 It accepts the following parameters:
3372 @item channel_layout
3373 The channel layout of the input stream. The default is "stereo".
3375 A channel layout describing the channels to be extracted as separate output streams
3376 or "all" to extract each input channel as a separate stream. The default is "all".
3378 Choosing channels not present in channel layout in the input will result in an error.
3381 @subsection Examples
3385 For example, assuming a stereo input MP3 file,
3387 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
3389 will create an output Matroska file with two audio streams, one containing only
3390 the left channel and the other the right channel.
3393 Split a 5.1 WAV file into per-channel files:
3395 ffmpeg -i in.wav -filter_complex
3396 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
3397 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
3398 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
3403 Extract only LFE from a 5.1 WAV file:
3405 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
3406 -map '[LFE]' lfe.wav
3411 Add a chorus effect to the audio.
3413 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
3415 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
3416 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
3417 The modulation depth defines the range the modulated delay is played before or after
3418 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
3419 sound tuned around the original one, like in a chorus where some vocals are slightly
3422 It accepts the following parameters:
3425 Set input gain. Default is 0.4.
3428 Set output gain. Default is 0.4.
3431 Set delays. A typical delay is around 40ms to 60ms.
3443 @subsection Examples
3449 chorus=0.7:0.9:55:0.4:0.25:2
3455 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
3459 Fuller sounding chorus with three delays:
3461 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
3466 Compress or expand the audio's dynamic range.
3468 It accepts the following parameters:
3474 A list of times in seconds for each channel over which the instantaneous level
3475 of the input signal is averaged to determine its volume. @var{attacks} refers to
3476 increase of volume and @var{decays} refers to decrease of volume. For most
3477 situations, the attack time (response to the audio getting louder) should be
3478 shorter than the decay time, because the human ear is more sensitive to sudden
3479 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
3480 a typical value for decay is 0.8 seconds.
3481 If specified number of attacks & decays is lower than number of channels, the last
3482 set attack/decay will be used for all remaining channels.
3485 A list of points for the transfer function, specified in dB relative to the
3486 maximum possible signal amplitude. Each key points list must be defined using
3487 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
3488 @code{x0/y0 x1/y1 x2/y2 ....}
3490 The input values must be in strictly increasing order but the transfer function
3491 does not have to be monotonically rising. The point @code{0/0} is assumed but
3492 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
3493 function are @code{-70/-70|-60/-20|1/0}.
3496 Set the curve radius in dB for all joints. It defaults to 0.01.
3499 Set the additional gain in dB to be applied at all points on the transfer
3500 function. This allows for easy adjustment of the overall gain.
3504 Set an initial volume, in dB, to be assumed for each channel when filtering
3505 starts. This permits the user to supply a nominal level initially, so that, for
3506 example, a very large gain is not applied to initial signal levels before the
3507 companding has begun to operate. A typical value for audio which is initially
3508 quiet is -90 dB. It defaults to 0.
3511 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
3512 delayed before being fed to the volume adjuster. Specifying a delay
3513 approximately equal to the attack/decay times allows the filter to effectively
3514 operate in predictive rather than reactive mode. It defaults to 0.
3518 @subsection Examples
3522 Make music with both quiet and loud passages suitable for listening to in a
3525 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
3528 Another example for audio with whisper and explosion parts:
3530 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
3534 A noise gate for when the noise is at a lower level than the signal:
3536 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
3540 Here is another noise gate, this time for when the noise is at a higher level
3541 than the signal (making it, in some ways, similar to squelch):
3543 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
3547 2:1 compression starting at -6dB:
3549 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
3553 2:1 compression starting at -9dB:
3555 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
3559 2:1 compression starting at -12dB:
3561 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
3565 2:1 compression starting at -18dB:
3567 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
3571 3:1 compression starting at -15dB:
3573 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
3579 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
3585 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
3589 Hard limiter at -6dB:
3591 compand=attacks=0:points=-80/-80|-6/-6|20/-6
3595 Hard limiter at -12dB:
3597 compand=attacks=0:points=-80/-80|-12/-12|20/-12
3601 Hard noise gate at -35 dB:
3603 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
3609 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
3613 @section compensationdelay
3615 Compensation Delay Line is a metric based delay to compensate differing
3616 positions of microphones or speakers.
3618 For example, you have recorded guitar with two microphones placed in
3619 different locations. Because the front of sound wave has fixed speed in
3620 normal conditions, the phasing of microphones can vary and depends on
3621 their location and interposition. The best sound mix can be achieved when
3622 these microphones are in phase (synchronized). Note that a distance of
3623 ~30 cm between microphones makes one microphone capture the signal in
3624 antiphase to the other microphone. That makes the final mix sound moody.
3625 This filter helps to solve phasing problems by adding different delays
3626 to each microphone track and make them synchronized.
3628 The best result can be reached when you take one track as base and
3629 synchronize other tracks one by one with it.
3630 Remember that synchronization/delay tolerance depends on sample rate, too.
3631 Higher sample rates will give more tolerance.
3633 The filter accepts the following parameters:
3637 Set millimeters distance. This is compensation distance for fine tuning.
3641 Set cm distance. This is compensation distance for tightening distance setup.
3645 Set meters distance. This is compensation distance for hard distance setup.
3649 Set dry amount. Amount of unprocessed (dry) signal.
3653 Set wet amount. Amount of processed (wet) signal.
3657 Set temperature in degrees Celsius. This is the temperature of the environment.
3662 Apply headphone crossfeed filter.
3664 Crossfeed is the process of blending the left and right channels of stereo
3666 It is mainly used to reduce extreme stereo separation of low frequencies.
3668 The intent is to produce more speaker like sound to the listener.
3670 The filter accepts the following options:
3674 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3675 This sets gain of low shelf filter for side part of stereo image.
3676 Default is -6dB. Max allowed is -30db when strength is set to 1.
3679 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3680 This sets cut off frequency of low shelf filter. Default is cut off near
3681 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3684 Set curve slope of low shelf filter. Default is 0.5.
3685 Allowed range is from 0.01 to 1.
3688 Set input gain. Default is 0.9.
3691 Set output gain. Default is 1.
3694 @subsection Commands
3696 This filter supports the all above options as @ref{commands}.
3698 @section crystalizer
3699 Simple algorithm for audio noise sharpening.
3701 This filter linearly increases differences betweeen each audio sample.
3703 The filter accepts the following options:
3707 Sets the intensity of effect (default: 2.0). Must be in range between -10.0 to 0
3708 (unchanged sound) to 10.0 (maximum effect).
3709 To inverse filtering use negative value.
3712 Enable clipping. By default is enabled.
3715 @subsection Commands
3717 This filter supports the all above options as @ref{commands}.
3720 Apply a DC shift to the audio.
3722 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3723 in the recording chain) from the audio. The effect of a DC offset is reduced
3724 headroom and hence volume. The @ref{astats} filter can be used to determine if
3725 a signal has a DC offset.
3729 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3733 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3734 used to prevent clipping.
3739 Apply de-essing to the audio samples.
3743 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3747 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3751 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3755 Set the output mode.
3757 It accepts the following values:
3760 Pass input unchanged.
3763 Pass ess filtered out.
3768 Default value is @var{o}.
3774 Measure audio dynamic range.
3776 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3777 is found in transition material. And anything less that 8 have very poor dynamics
3778 and is very compressed.
3780 The filter accepts the following options:
3784 Set window length in seconds used to split audio into segments of equal length.
3785 Default is 3 seconds.
3789 Dynamic Audio Normalizer.
3791 This filter applies a certain amount of gain to the input audio in order
3792 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3793 contrast to more "simple" normalization algorithms, the Dynamic Audio
3794 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3795 This allows for applying extra gain to the "quiet" sections of the audio
3796 while avoiding distortions or clipping the "loud" sections. In other words:
3797 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3798 sections, in the sense that the volume of each section is brought to the
3799 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3800 this goal *without* applying "dynamic range compressing". It will retain 100%
3801 of the dynamic range *within* each section of the audio file.
3805 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3806 Default is 500 milliseconds.
3807 The Dynamic Audio Normalizer processes the input audio in small chunks,
3808 referred to as frames. This is required, because a peak magnitude has no
3809 meaning for just a single sample value. Instead, we need to determine the
3810 peak magnitude for a contiguous sequence of sample values. While a "standard"
3811 normalizer would simply use the peak magnitude of the complete file, the
3812 Dynamic Audio Normalizer determines the peak magnitude individually for each
3813 frame. The length of a frame is specified in milliseconds. By default, the
3814 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3815 been found to give good results with most files.
3816 Note that the exact frame length, in number of samples, will be determined
3817 automatically, based on the sampling rate of the individual input audio file.
3820 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3821 number. Default is 31.
3822 Probably the most important parameter of the Dynamic Audio Normalizer is the
3823 @code{window size} of the Gaussian smoothing filter. The filter's window size
3824 is specified in frames, centered around the current frame. For the sake of
3825 simplicity, this must be an odd number. Consequently, the default value of 31
3826 takes into account the current frame, as well as the 15 preceding frames and
3827 the 15 subsequent frames. Using a larger window results in a stronger
3828 smoothing effect and thus in less gain variation, i.e. slower gain
3829 adaptation. Conversely, using a smaller window results in a weaker smoothing
3830 effect and thus in more gain variation, i.e. faster gain adaptation.
3831 In other words, the more you increase this value, the more the Dynamic Audio
3832 Normalizer will behave like a "traditional" normalization filter. On the
3833 contrary, the more you decrease this value, the more the Dynamic Audio
3834 Normalizer will behave like a dynamic range compressor.
3837 Set the target peak value. This specifies the highest permissible magnitude
3838 level for the normalized audio input. This filter will try to approach the
3839 target peak magnitude as closely as possible, but at the same time it also
3840 makes sure that the normalized signal will never exceed the peak magnitude.
3841 A frame's maximum local gain factor is imposed directly by the target peak
3842 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3843 It is not recommended to go above this value.
3846 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3847 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3848 factor for each input frame, i.e. the maximum gain factor that does not
3849 result in clipping or distortion. The maximum gain factor is determined by
3850 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3851 additionally bounds the frame's maximum gain factor by a predetermined
3852 (global) maximum gain factor. This is done in order to avoid excessive gain
3853 factors in "silent" or almost silent frames. By default, the maximum gain
3854 factor is 10.0, For most inputs the default value should be sufficient and
3855 it usually is not recommended to increase this value. Though, for input
3856 with an extremely low overall volume level, it may be necessary to allow even
3857 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3858 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3859 Instead, a "sigmoid" threshold function will be applied. This way, the
3860 gain factors will smoothly approach the threshold value, but never exceed that
3864 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3865 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3866 This means that the maximum local gain factor for each frame is defined
3867 (only) by the frame's highest magnitude sample. This way, the samples can
3868 be amplified as much as possible without exceeding the maximum signal
3869 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3870 Normalizer can also take into account the frame's root mean square,
3871 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3872 determine the power of a time-varying signal. It is therefore considered
3873 that the RMS is a better approximation of the "perceived loudness" than
3874 just looking at the signal's peak magnitude. Consequently, by adjusting all
3875 frames to a constant RMS value, a uniform "perceived loudness" can be
3876 established. If a target RMS value has been specified, a frame's local gain
3877 factor is defined as the factor that would result in exactly that RMS value.
3878 Note, however, that the maximum local gain factor is still restricted by the
3879 frame's highest magnitude sample, in order to prevent clipping.
3882 Enable channels coupling. By default is enabled.
3883 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3884 amount. This means the same gain factor will be applied to all channels, i.e.
3885 the maximum possible gain factor is determined by the "loudest" channel.
3886 However, in some recordings, it may happen that the volume of the different
3887 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3888 In this case, this option can be used to disable the channel coupling. This way,
3889 the gain factor will be determined independently for each channel, depending
3890 only on the individual channel's highest magnitude sample. This allows for
3891 harmonizing the volume of the different channels.
3894 Enable DC bias correction. By default is disabled.
3895 An audio signal (in the time domain) is a sequence of sample values.
3896 In the Dynamic Audio Normalizer these sample values are represented in the
3897 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3898 audio signal, or "waveform", should be centered around the zero point.
3899 That means if we calculate the mean value of all samples in a file, or in a
3900 single frame, then the result should be 0.0 or at least very close to that
3901 value. If, however, there is a significant deviation of the mean value from
3902 0.0, in either positive or negative direction, this is referred to as a
3903 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3904 Audio Normalizer provides optional DC bias correction.
3905 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3906 the mean value, or "DC correction" offset, of each input frame and subtract
3907 that value from all of the frame's sample values which ensures those samples
3908 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3909 boundaries, the DC correction offset values will be interpolated smoothly
3910 between neighbouring frames.
3912 @item altboundary, b
3913 Enable alternative boundary mode. By default is disabled.
3914 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3915 around each frame. This includes the preceding frames as well as the
3916 subsequent frames. However, for the "boundary" frames, located at the very
3917 beginning and at the very end of the audio file, not all neighbouring
3918 frames are available. In particular, for the first few frames in the audio
3919 file, the preceding frames are not known. And, similarly, for the last few
3920 frames in the audio file, the subsequent frames are not known. Thus, the
3921 question arises which gain factors should be assumed for the missing frames
3922 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3923 to deal with this situation. The default boundary mode assumes a gain factor
3924 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3925 "fade out" at the beginning and at the end of the input, respectively.
3928 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3929 By default, the Dynamic Audio Normalizer does not apply "traditional"
3930 compression. This means that signal peaks will not be pruned and thus the
3931 full dynamic range will be retained within each local neighbourhood. However,
3932 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3933 normalization algorithm with a more "traditional" compression.
3934 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3935 (thresholding) function. If (and only if) the compression feature is enabled,
3936 all input frames will be processed by a soft knee thresholding function prior
3937 to the actual normalization process. Put simply, the thresholding function is
3938 going to prune all samples whose magnitude exceeds a certain threshold value.
3939 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3940 value. Instead, the threshold value will be adjusted for each individual
3942 In general, smaller parameters result in stronger compression, and vice versa.
3943 Values below 3.0 are not recommended, because audible distortion may appear.
3946 Set the target threshold value. This specifies the lowest permissible
3947 magnitude level for the audio input which will be normalized.
3948 If input frame volume is above this value frame will be normalized.
3949 Otherwise frame may not be normalized at all. The default value is set
3950 to 0, which means all input frames will be normalized.
3951 This option is mostly useful if digital noise is not wanted to be amplified.
3954 @subsection Commands
3956 This filter supports the all above options as @ref{commands}.
3960 Make audio easier to listen to on headphones.
3962 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3963 so that when listened to on headphones the stereo image is moved from
3964 inside your head (standard for headphones) to outside and in front of
3965 the listener (standard for speakers).
3971 Apply a two-pole peaking equalisation (EQ) filter. With this
3972 filter, the signal-level at and around a selected frequency can
3973 be increased or decreased, whilst (unlike bandpass and bandreject
3974 filters) that at all other frequencies is unchanged.
3976 In order to produce complex equalisation curves, this filter can
3977 be given several times, each with a different central frequency.
3979 The filter accepts the following options:
3983 Set the filter's central frequency in Hz.
3986 Set method to specify band-width of filter.
4001 Specify the band-width of a filter in width_type units.
4004 Set the required gain or attenuation in dB.
4005 Beware of clipping when using a positive gain.
4008 How much to use filtered signal in output. Default is 1.
4009 Range is between 0 and 1.
4012 Specify which channels to filter, by default all available are filtered.
4015 Normalize biquad coefficients, by default is disabled.
4016 Enabling it will normalize magnitude response at DC to 0dB.
4019 Set transform type of IIR filter.
4028 Set precison of filtering.
4031 Pick automatic sample format depending on surround filters.
4033 Always use signed 16-bit.
4035 Always use signed 32-bit.
4037 Always use float 32-bit.
4039 Always use float 64-bit.
4043 @subsection Examples
4046 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
4048 equalizer=f=1000:t=h:width=200:g=-10
4052 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
4054 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
4058 @subsection Commands
4060 This filter supports the following commands:
4063 Change equalizer frequency.
4064 Syntax for the command is : "@var{frequency}"
4067 Change equalizer width_type.
4068 Syntax for the command is : "@var{width_type}"
4071 Change equalizer width.
4072 Syntax for the command is : "@var{width}"
4075 Change equalizer gain.
4076 Syntax for the command is : "@var{gain}"
4079 Change equalizer mix.
4080 Syntax for the command is : "@var{mix}"
4083 @section extrastereo
4085 Linearly increases the difference between left and right channels which
4086 adds some sort of "live" effect to playback.
4088 The filter accepts the following options:
4092 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
4093 (average of both channels), with 1.0 sound will be unchanged, with
4094 -1.0 left and right channels will be swapped.
4097 Enable clipping. By default is enabled.
4100 @subsection Commands
4102 This filter supports the all above options as @ref{commands}.
4104 @section firequalizer
4105 Apply FIR Equalization using arbitrary frequency response.
4107 The filter accepts the following option:
4111 Set gain curve equation (in dB). The expression can contain variables:
4114 the evaluated frequency
4118 channel number, set to 0 when multichannels evaluation is disabled
4120 channel id, see libavutil/channel_layout.h, set to the first channel id when
4121 multichannels evaluation is disabled
4125 channel_layout, see libavutil/channel_layout.h
4130 @item gain_interpolate(f)
4131 interpolate gain on frequency f based on gain_entry
4132 @item cubic_interpolate(f)
4133 same as gain_interpolate, but smoother
4135 This option is also available as command. Default is @code{gain_interpolate(f)}.
4138 Set gain entry for gain_interpolate function. The expression can
4142 store gain entry at frequency f with value g
4144 This option is also available as command.
4147 Set filter delay in seconds. Higher value means more accurate.
4148 Default is @code{0.01}.
4151 Set filter accuracy in Hz. Lower value means more accurate.
4152 Default is @code{5}.
4155 Set window function. Acceptable values are:
4158 rectangular window, useful when gain curve is already smooth
4160 hann window (default)
4166 3-terms continuous 1st derivative nuttall window
4168 minimum 3-terms discontinuous nuttall window
4170 4-terms continuous 1st derivative nuttall window
4172 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
4174 blackman-harris window
4180 If enabled, use fixed number of audio samples. This improves speed when
4181 filtering with large delay. Default is disabled.
4184 Enable multichannels evaluation on gain. Default is disabled.
4187 Enable zero phase mode by subtracting timestamp to compensate delay.
4188 Default is disabled.
4191 Set scale used by gain. Acceptable values are:
4194 linear frequency, linear gain
4196 linear frequency, logarithmic (in dB) gain (default)
4198 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
4200 logarithmic frequency, logarithmic gain
4204 Set file for dumping, suitable for gnuplot.
4207 Set scale for dumpfile. Acceptable values are same with scale option.
4211 Enable 2-channel convolution using complex FFT. This improves speed significantly.
4212 Default is disabled.
4215 Enable minimum phase impulse response. Default is disabled.
4218 @subsection Examples
4223 firequalizer=gain='if(lt(f,1000), 0, -INF)'
4226 lowpass at 1000 Hz with gain_entry:
4228 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
4231 custom equalization:
4233 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
4236 higher delay with zero phase to compensate delay:
4238 firequalizer=delay=0.1:fixed=on:zero_phase=on
4241 lowpass on left channel, highpass on right channel:
4243 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
4244 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
4249 Apply a flanging effect to the audio.
4251 The filter accepts the following options:
4255 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
4258 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
4261 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
4265 Set percentage of delayed signal mixed with original. Range from 0 to 100.
4266 Default value is 71.
4269 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
4272 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
4273 Default value is @var{sinusoidal}.
4276 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
4277 Default value is 25.
4280 Set delay-line interpolation, @var{linear} or @var{quadratic}.
4281 Default is @var{linear}.
4285 Apply Haas effect to audio.
4287 Note that this makes most sense to apply on mono signals.
4288 With this filter applied to mono signals it give some directionality and
4289 stretches its stereo image.
4291 The filter accepts the following options:
4295 Set input level. By default is @var{1}, or 0dB
4298 Set output level. By default is @var{1}, or 0dB.
4301 Set gain applied to side part of signal. By default is @var{1}.
4304 Set kind of middle source. Can be one of the following:
4314 Pick middle part signal of stereo image.
4317 Pick side part signal of stereo image.
4321 Change middle phase. By default is disabled.
4324 Set left channel delay. By default is @var{2.05} milliseconds.
4327 Set left channel balance. By default is @var{-1}.
4330 Set left channel gain. By default is @var{1}.
4333 Change left phase. By default is disabled.
4336 Set right channel delay. By defaults is @var{2.12} milliseconds.
4339 Set right channel balance. By default is @var{1}.
4342 Set right channel gain. By default is @var{1}.
4345 Change right phase. By default is enabled.
4350 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
4351 embedded HDCD codes is expanded into a 20-bit PCM stream.
4353 The filter supports the Peak Extend and Low-level Gain Adjustment features
4354 of HDCD, and detects the Transient Filter flag.
4357 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
4360 When using the filter with wav, note the default encoding for wav is 16-bit,
4361 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
4362 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
4364 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
4365 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
4368 The filter accepts the following options:
4371 @item disable_autoconvert
4372 Disable any automatic format conversion or resampling in the filter graph.
4374 @item process_stereo
4375 Process the stereo channels together. If target_gain does not match between
4376 channels, consider it invalid and use the last valid target_gain.
4379 Set the code detect timer period in ms.
4382 Always extend peaks above -3dBFS even if PE isn't signaled.
4385 Replace audio with a solid tone and adjust the amplitude to signal some
4386 specific aspect of the decoding process. The output file can be loaded in
4387 an audio editor alongside the original to aid analysis.
4389 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
4396 Gain adjustment level at each sample
4398 Samples where peak extend occurs
4400 Samples where the code detect timer is active
4402 Samples where the target gain does not match between channels
4408 Apply head-related transfer functions (HRTFs) to create virtual
4409 loudspeakers around the user for binaural listening via headphones.
4410 The HRIRs are provided via additional streams, for each channel
4411 one stereo input stream is needed.
4413 The filter accepts the following options:
4417 Set mapping of input streams for convolution.
4418 The argument is a '|'-separated list of channel names in order as they
4419 are given as additional stream inputs for filter.
4420 This also specify number of input streams. Number of input streams
4421 must be not less than number of channels in first stream plus one.
4424 Set gain applied to audio. Value is in dB. Default is 0.
4427 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4428 processing audio in time domain which is slow.
4429 @var{freq} is processing audio in frequency domain which is fast.
4430 Default is @var{freq}.
4433 Set custom gain for LFE channels. Value is in dB. Default is 0.
4436 Set size of frame in number of samples which will be processed at once.
4437 Default value is @var{1024}. Allowed range is from 1024 to 96000.
4440 Set format of hrir stream.
4441 Default value is @var{stereo}. Alternative value is @var{multich}.
4442 If value is set to @var{stereo}, number of additional streams should
4443 be greater or equal to number of input channels in first input stream.
4444 Also each additional stream should have stereo number of channels.
4445 If value is set to @var{multich}, number of additional streams should
4446 be exactly one. Also number of input channels of additional stream
4447 should be equal or greater than twice number of channels of first input
4451 @subsection Examples
4455 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4456 each amovie filter use stereo file with IR coefficients as input.
4457 The files give coefficients for each position of virtual loudspeaker:
4460 -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"
4465 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4466 but now in @var{multich} @var{hrir} format.
4468 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"
4475 Apply a high-pass filter with 3dB point frequency.
4476 The filter can be either single-pole, or double-pole (the default).
4477 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4479 The filter accepts the following options:
4483 Set frequency in Hz. Default is 3000.
4486 Set number of poles. Default is 2.
4489 Set method to specify band-width of filter.
4504 Specify the band-width of a filter in width_type units.
4505 Applies only to double-pole filter.
4506 The default is 0.707q and gives a Butterworth response.
4509 How much to use filtered signal in output. Default is 1.
4510 Range is between 0 and 1.
4513 Specify which channels to filter, by default all available are filtered.
4516 Normalize biquad coefficients, by default is disabled.
4517 Enabling it will normalize magnitude response at DC to 0dB.
4520 Set transform type of IIR filter.
4529 Set precison of filtering.
4532 Pick automatic sample format depending on surround filters.
4534 Always use signed 16-bit.
4536 Always use signed 32-bit.
4538 Always use float 32-bit.
4540 Always use float 64-bit.
4544 @subsection Commands
4546 This filter supports the following commands:
4549 Change highpass frequency.
4550 Syntax for the command is : "@var{frequency}"
4553 Change highpass width_type.
4554 Syntax for the command is : "@var{width_type}"
4557 Change highpass width.
4558 Syntax for the command is : "@var{width}"
4561 Change highpass mix.
4562 Syntax for the command is : "@var{mix}"
4567 Join multiple input streams into one multi-channel stream.
4569 It accepts the following parameters:
4573 The number of input streams. It defaults to 2.
4575 @item channel_layout
4576 The desired output channel layout. It defaults to stereo.
4579 Map channels from inputs to output. The argument is a '|'-separated list of
4580 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
4581 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
4582 can be either the name of the input channel (e.g. FL for front left) or its
4583 index in the specified input stream. @var{out_channel} is the name of the output
4587 The filter will attempt to guess the mappings when they are not specified
4588 explicitly. It does so by first trying to find an unused matching input channel
4589 and if that fails it picks the first unused input channel.
4591 Join 3 inputs (with properly set channel layouts):
4593 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
4596 Build a 5.1 output from 6 single-channel streams:
4598 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
4599 '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'
4605 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
4607 To enable compilation of this filter you need to configure FFmpeg with
4608 @code{--enable-ladspa}.
4612 Specifies the name of LADSPA plugin library to load. If the environment
4613 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
4614 each one of the directories specified by the colon separated list in
4615 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
4616 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
4617 @file{/usr/lib/ladspa/}.
4620 Specifies the plugin within the library. Some libraries contain only
4621 one plugin, but others contain many of them. If this is not set filter
4622 will list all available plugins within the specified library.
4625 Set the '|' separated list of controls which are zero or more floating point
4626 values that determine the behavior of the loaded plugin (for example delay,
4628 Controls need to be defined using the following syntax:
4629 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
4630 @var{valuei} is the value set on the @var{i}-th control.
4631 Alternatively they can be also defined using the following syntax:
4632 @var{value0}|@var{value1}|@var{value2}|..., where
4633 @var{valuei} is the value set on the @var{i}-th control.
4634 If @option{controls} is set to @code{help}, all available controls and
4635 their valid ranges are printed.
4637 @item sample_rate, s
4638 Specify the sample rate, default to 44100. Only used if plugin have
4642 Set the number of samples per channel per each output frame, default
4643 is 1024. Only used if plugin have zero inputs.
4646 Set the minimum duration of the sourced audio. See
4647 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4648 for the accepted syntax.
4649 Note that the resulting duration may be greater than the specified duration,
4650 as the generated audio is always cut at the end of a complete frame.
4651 If not specified, or the expressed duration is negative, the audio is
4652 supposed to be generated forever.
4653 Only used if plugin have zero inputs.
4656 Enable latency compensation, by default is disabled.
4657 Only used if plugin have inputs.
4660 @subsection Examples
4664 List all available plugins within amp (LADSPA example plugin) library:
4670 List all available controls and their valid ranges for @code{vcf_notch}
4671 plugin from @code{VCF} library:
4673 ladspa=f=vcf:p=vcf_notch:c=help
4677 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
4680 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
4684 Add reverberation to the audio using TAP-plugins
4685 (Tom's Audio Processing plugins):
4687 ladspa=file=tap_reverb:tap_reverb
4691 Generate white noise, with 0.2 amplitude:
4693 ladspa=file=cmt:noise_source_white:c=c0=.2
4697 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
4698 @code{C* Audio Plugin Suite} (CAPS) library:
4700 ladspa=file=caps:Click:c=c1=20'
4704 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
4706 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
4710 Increase volume by 20dB using fast lookahead limiter from Steve Harris
4711 @code{SWH Plugins} collection:
4713 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
4717 Attenuate low frequencies using Multiband EQ from Steve Harris
4718 @code{SWH Plugins} collection:
4720 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
4724 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
4727 ladspa=caps:Narrower
4731 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
4733 ladspa=caps:White:.2
4737 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
4739 ladspa=caps:Fractal:c=c1=1
4743 Dynamic volume normalization using @code{VLevel} plugin:
4745 ladspa=vlevel-ladspa:vlevel_mono
4749 @subsection Commands
4751 This filter supports the following commands:
4754 Modify the @var{N}-th control value.
4756 If the specified value is not valid, it is ignored and prior one is kept.
4761 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4762 Support for both single pass (livestreams, files) and double pass (files) modes.
4763 This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately
4764 detect true peaks, the audio stream will be upsampled to 192 kHz.
4765 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4767 The filter accepts the following options:
4771 Set integrated loudness target.
4772 Range is -70.0 - -5.0. Default value is -24.0.
4775 Set loudness range target.
4776 Range is 1.0 - 20.0. Default value is 7.0.
4779 Set maximum true peak.
4780 Range is -9.0 - +0.0. Default value is -2.0.
4782 @item measured_I, measured_i
4783 Measured IL of input file.
4784 Range is -99.0 - +0.0.
4786 @item measured_LRA, measured_lra
4787 Measured LRA of input file.
4788 Range is 0.0 - 99.0.
4790 @item measured_TP, measured_tp
4791 Measured true peak of input file.
4792 Range is -99.0 - +99.0.
4794 @item measured_thresh
4795 Measured threshold of input file.
4796 Range is -99.0 - +0.0.
4799 Set offset gain. Gain is applied before the true-peak limiter.
4800 Range is -99.0 - +99.0. Default is +0.0.
4803 Normalize by linearly scaling the source audio.
4804 @code{measured_I}, @code{measured_LRA}, @code{measured_TP},
4805 and @code{measured_thresh} must all be specified. Target LRA shouldn't
4806 be lower than source LRA and the change in integrated loudness shouldn't
4807 result in a true peak which exceeds the target TP. If any of these
4808 conditions aren't met, normalization mode will revert to @var{dynamic}.
4809 Options are @code{true} or @code{false}. Default is @code{true}.
4812 Treat mono input files as "dual-mono". If a mono file is intended for playback
4813 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4814 If set to @code{true}, this option will compensate for this effect.
4815 Multi-channel input files are not affected by this option.
4816 Options are true or false. Default is false.
4819 Set print format for stats. Options are summary, json, or none.
4820 Default value is none.
4825 Apply a low-pass filter with 3dB point frequency.
4826 The filter can be either single-pole or double-pole (the default).
4827 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4829 The filter accepts the following options:
4833 Set frequency in Hz. Default is 500.
4836 Set number of poles. Default is 2.
4839 Set method to specify band-width of filter.
4854 Specify the band-width of a filter in width_type units.
4855 Applies only to double-pole filter.
4856 The default is 0.707q and gives a Butterworth response.
4859 How much to use filtered signal in output. Default is 1.
4860 Range is between 0 and 1.
4863 Specify which channels to filter, by default all available are filtered.
4866 Normalize biquad coefficients, by default is disabled.
4867 Enabling it will normalize magnitude response at DC to 0dB.
4870 Set transform type of IIR filter.
4879 Set precison of filtering.
4882 Pick automatic sample format depending on surround filters.
4884 Always use signed 16-bit.
4886 Always use signed 32-bit.
4888 Always use float 32-bit.
4890 Always use float 64-bit.
4894 @subsection Examples
4897 Lowpass only LFE channel, it LFE is not present it does nothing:
4903 @subsection Commands
4905 This filter supports the following commands:
4908 Change lowpass frequency.
4909 Syntax for the command is : "@var{frequency}"
4912 Change lowpass width_type.
4913 Syntax for the command is : "@var{width_type}"
4916 Change lowpass width.
4917 Syntax for the command is : "@var{width}"
4921 Syntax for the command is : "@var{mix}"
4926 Load a LV2 (LADSPA Version 2) plugin.
4928 To enable compilation of this filter you need to configure FFmpeg with
4929 @code{--enable-lv2}.
4933 Specifies the plugin URI. You may need to escape ':'.
4936 Set the '|' separated list of controls which are zero or more floating point
4937 values that determine the behavior of the loaded plugin (for example delay,
4939 If @option{controls} is set to @code{help}, all available controls and
4940 their valid ranges are printed.
4942 @item sample_rate, s
4943 Specify the sample rate, default to 44100. Only used if plugin have
4947 Set the number of samples per channel per each output frame, default
4948 is 1024. Only used if plugin have zero inputs.
4951 Set the minimum duration of the sourced audio. See
4952 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4953 for the accepted syntax.
4954 Note that the resulting duration may be greater than the specified duration,
4955 as the generated audio is always cut at the end of a complete frame.
4956 If not specified, or the expressed duration is negative, the audio is
4957 supposed to be generated forever.
4958 Only used if plugin have zero inputs.
4961 @subsection Examples
4965 Apply bass enhancer plugin from Calf:
4967 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4971 Apply vinyl plugin from Calf:
4973 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4977 Apply bit crusher plugin from ArtyFX:
4979 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4984 Multiband Compress or expand the audio's dynamic range.
4986 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4987 This is akin to the crossover of a loudspeaker, and results in flat frequency
4988 response when absent compander action.
4990 It accepts the following parameters:
4994 This option syntax is:
4995 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4996 For explanation of each item refer to compand filter documentation.
5002 Mix channels with specific gain levels. The filter accepts the output
5003 channel layout followed by a set of channels definitions.
5005 This filter is also designed to efficiently remap the channels of an audio
5008 The filter accepts parameters of the form:
5009 "@var{l}|@var{outdef}|@var{outdef}|..."
5013 output channel layout or number of channels
5016 output channel specification, of the form:
5017 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
5020 output channel to define, either a channel name (FL, FR, etc.) or a channel
5021 number (c0, c1, etc.)
5024 multiplicative coefficient for the channel, 1 leaving the volume unchanged
5027 input channel to use, see out_name for details; it is not possible to mix
5028 named and numbered input channels
5031 If the `=' in a channel specification is replaced by `<', then the gains for
5032 that specification will be renormalized so that the total is 1, thus
5033 avoiding clipping noise.
5035 @subsection Mixing examples
5037 For example, if you want to down-mix from stereo to mono, but with a bigger
5038 factor for the left channel:
5040 pan=1c|c0=0.9*c0+0.1*c1
5043 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
5044 7-channels surround:
5046 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
5049 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
5050 that should be preferred (see "-ac" option) unless you have very specific
5053 @subsection Remapping examples
5055 The channel remapping will be effective if, and only if:
5058 @item gain coefficients are zeroes or ones,
5059 @item only one input per channel output,
5062 If all these conditions are satisfied, the filter will notify the user ("Pure
5063 channel mapping detected"), and use an optimized and lossless method to do the
5066 For example, if you have a 5.1 source and want a stereo audio stream by
5067 dropping the extra channels:
5069 pan="stereo| c0=FL | c1=FR"
5072 Given the same source, you can also switch front left and front right channels
5073 and keep the input channel layout:
5075 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
5078 If the input is a stereo audio stream, you can mute the front left channel (and
5079 still keep the stereo channel layout) with:
5084 Still with a stereo audio stream input, you can copy the right channel in both
5085 front left and right:
5087 pan="stereo| c0=FR | c1=FR"
5092 ReplayGain scanner filter. This filter takes an audio stream as an input and
5093 outputs it unchanged.
5094 At end of filtering it displays @code{track_gain} and @code{track_peak}.
5098 Convert the audio sample format, sample rate and channel layout. It is
5099 not meant to be used directly.
5102 Apply time-stretching and pitch-shifting with librubberband.
5104 To enable compilation of this filter, you need to configure FFmpeg with
5105 @code{--enable-librubberband}.
5107 The filter accepts the following options:
5111 Set tempo scale factor.
5114 Set pitch scale factor.
5117 Set transients detector.
5118 Possible values are:
5127 Possible values are:
5136 Possible values are:
5143 Set processing window size.
5144 Possible values are:
5153 Possible values are:
5160 Enable formant preservation when shift pitching.
5161 Possible values are:
5169 Possible values are:
5178 Possible values are:
5185 @subsection Commands
5187 This filter supports the following commands:
5190 Change filter tempo scale factor.
5191 Syntax for the command is : "@var{tempo}"
5194 Change filter pitch scale factor.
5195 Syntax for the command is : "@var{pitch}"
5198 @section sidechaincompress
5200 This filter acts like normal compressor but has the ability to compress
5201 detected signal using second input signal.
5202 It needs two input streams and returns one output stream.
5203 First input stream will be processed depending on second stream signal.
5204 The filtered signal then can be filtered with other filters in later stages of
5205 processing. See @ref{pan} and @ref{amerge} filter.
5207 The filter accepts the following options:
5211 Set input gain. Default is 1. Range is between 0.015625 and 64.
5214 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
5215 Default is @code{downward}.
5218 If a signal of second stream raises above this level it will affect the gain
5219 reduction of first stream.
5220 By default is 0.125. Range is between 0.00097563 and 1.
5223 Set a ratio about which the signal is reduced. 1:2 means that if the level
5224 raised 4dB above the threshold, it will be only 2dB above after the reduction.
5225 Default is 2. Range is between 1 and 20.
5228 Amount of milliseconds the signal has to rise above the threshold before gain
5229 reduction starts. Default is 20. Range is between 0.01 and 2000.
5232 Amount of milliseconds the signal has to fall below the threshold before
5233 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
5236 Set the amount by how much signal will be amplified after processing.
5237 Default is 1. Range is from 1 to 64.
5240 Curve the sharp knee around the threshold to enter gain reduction more softly.
5241 Default is 2.82843. Range is between 1 and 8.
5244 Choose if the @code{average} level between all channels of side-chain stream
5245 or the louder(@code{maximum}) channel of side-chain stream affects the
5246 reduction. Default is @code{average}.
5249 Should the exact signal be taken in case of @code{peak} or an RMS one in case
5250 of @code{rms}. Default is @code{rms} which is mainly smoother.
5253 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
5256 How much to use compressed signal in output. Default is 1.
5257 Range is between 0 and 1.
5260 @subsection Commands
5262 This filter supports the all above options as @ref{commands}.
5264 @subsection Examples
5268 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
5269 depending on the signal of 2nd input and later compressed signal to be
5270 merged with 2nd input:
5272 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
5276 @section sidechaingate
5278 A sidechain gate acts like a normal (wideband) gate but has the ability to
5279 filter the detected signal before sending it to the gain reduction stage.
5280 Normally a gate uses the full range signal to detect a level above the
5282 For example: If you cut all lower frequencies from your sidechain signal
5283 the gate will decrease the volume of your track only if not enough highs
5284 appear. With this technique you are able to reduce the resonation of a
5285 natural drum or remove "rumbling" of muted strokes from a heavily distorted
5287 It needs two input streams and returns one output stream.
5288 First input stream will be processed depending on second stream signal.
5290 The filter accepts the following options:
5294 Set input level before filtering.
5295 Default is 1. Allowed range is from 0.015625 to 64.
5298 Set the mode of operation. Can be @code{upward} or @code{downward}.
5299 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
5300 will be amplified, expanding dynamic range in upward direction.
5301 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
5304 Set the level of gain reduction when the signal is below the threshold.
5305 Default is 0.06125. Allowed range is from 0 to 1.
5306 Setting this to 0 disables reduction and then filter behaves like expander.
5309 If a signal rises above this level the gain reduction is released.
5310 Default is 0.125. Allowed range is from 0 to 1.
5313 Set a ratio about which the signal is reduced.
5314 Default is 2. Allowed range is from 1 to 9000.
5317 Amount of milliseconds the signal has to rise above the threshold before gain
5319 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
5322 Amount of milliseconds the signal has to fall below the threshold before the
5323 reduction is increased again. Default is 250 milliseconds.
5324 Allowed range is from 0.01 to 9000.
5327 Set amount of amplification of signal after processing.
5328 Default is 1. Allowed range is from 1 to 64.
5331 Curve the sharp knee around the threshold to enter gain reduction more softly.
5332 Default is 2.828427125. Allowed range is from 1 to 8.
5335 Choose if exact signal should be taken for detection or an RMS like one.
5336 Default is rms. Can be peak or rms.
5339 Choose if the average level between all channels or the louder channel affects
5341 Default is average. Can be average or maximum.
5344 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
5347 @subsection Commands
5349 This filter supports the all above options as @ref{commands}.
5351 @section silencedetect
5353 Detect silence in an audio stream.
5355 This filter logs a message when it detects that the input audio volume is less
5356 or equal to a noise tolerance value for a duration greater or equal to the
5357 minimum detected noise duration.
5359 The printed times and duration are expressed in seconds. The
5360 @code{lavfi.silence_start} or @code{lavfi.silence_start.X} metadata key
5361 is set on the first frame whose timestamp equals or exceeds the detection
5362 duration and it contains the timestamp of the first frame of the silence.
5364 The @code{lavfi.silence_duration} or @code{lavfi.silence_duration.X}
5365 and @code{lavfi.silence_end} or @code{lavfi.silence_end.X} metadata
5366 keys are set on the first frame after the silence. If @option{mono} is
5367 enabled, and each channel is evaluated separately, the @code{.X}
5368 suffixed keys are used, and @code{X} corresponds to the channel number.
5370 The filter accepts the following options:
5374 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
5375 specified value) or amplitude ratio. Default is -60dB, or 0.001.
5378 Set silence duration until notification (default is 2 seconds). See
5379 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5380 for the accepted syntax.
5383 Process each channel separately, instead of combined. By default is disabled.
5386 @subsection Examples
5390 Detect 5 seconds of silence with -50dB noise tolerance:
5392 silencedetect=n=-50dB:d=5
5396 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
5397 tolerance in @file{silence.mp3}:
5399 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
5403 @section silenceremove
5405 Remove silence from the beginning, middle or end of the audio.
5407 The filter accepts the following options:
5411 This value is used to indicate if audio should be trimmed at beginning of
5412 the audio. A value of zero indicates no silence should be trimmed from the
5413 beginning. When specifying a non-zero value, it trims audio up until it
5414 finds non-silence. Normally, when trimming silence from beginning of audio
5415 the @var{start_periods} will be @code{1} but it can be increased to higher
5416 values to trim all audio up to specific count of non-silence periods.
5417 Default value is @code{0}.
5419 @item start_duration
5420 Specify the amount of time that non-silence must be detected before it stops
5421 trimming audio. By increasing the duration, bursts of noises can be treated
5422 as silence and trimmed off. Default value is @code{0}.
5424 @item start_threshold
5425 This indicates what sample value should be treated as silence. For digital
5426 audio, a value of @code{0} may be fine but for audio recorded from analog,
5427 you may wish to increase the value to account for background noise.
5428 Can be specified in dB (in case "dB" is appended to the specified value)
5429 or amplitude ratio. Default value is @code{0}.
5432 Specify max duration of silence at beginning that will be kept after
5433 trimming. Default is 0, which is equal to trimming all samples detected
5437 Specify mode of detection of silence end in start of multi-channel audio.
5438 Can be @var{any} or @var{all}. Default is @var{any}.
5439 With @var{any}, any sample that is detected as non-silence will cause
5440 stopped trimming of silence.
5441 With @var{all}, only if all channels are detected as non-silence will cause
5442 stopped trimming of silence.
5445 Set the count for trimming silence from the end of audio.
5446 To remove silence from the middle of a file, specify a @var{stop_periods}
5447 that is negative. This value is then treated as a positive value and is
5448 used to indicate the effect should restart processing as specified by
5449 @var{start_periods}, making it suitable for removing periods of silence
5450 in the middle of the audio.
5451 Default value is @code{0}.
5454 Specify a duration of silence that must exist before audio is not copied any
5455 more. By specifying a higher duration, silence that is wanted can be left in
5457 Default value is @code{0}.
5459 @item stop_threshold
5460 This is the same as @option{start_threshold} but for trimming silence from
5462 Can be specified in dB (in case "dB" is appended to the specified value)
5463 or amplitude ratio. Default value is @code{0}.
5466 Specify max duration of silence at end that will be kept after
5467 trimming. Default is 0, which is equal to trimming all samples detected
5471 Specify mode of detection of silence start in end of multi-channel audio.
5472 Can be @var{any} or @var{all}. Default is @var{any}.
5473 With @var{any}, any sample that is detected as non-silence will cause
5474 stopped trimming of silence.
5475 With @var{all}, only if all channels are detected as non-silence will cause
5476 stopped trimming of silence.
5479 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
5480 and works better with digital silence which is exactly 0.
5481 Default value is @code{rms}.
5484 Set duration in number of seconds used to calculate size of window in number
5485 of samples for detecting silence.
5486 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
5489 @subsection Examples
5493 The following example shows how this filter can be used to start a recording
5494 that does not contain the delay at the start which usually occurs between
5495 pressing the record button and the start of the performance:
5497 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
5501 Trim all silence encountered from beginning to end where there is more than 1
5502 second of silence in audio:
5504 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
5508 Trim all digital silence samples, using peak detection, from beginning to end
5509 where there is more than 0 samples of digital silence in audio and digital
5510 silence is detected in all channels at same positions in stream:
5512 silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
5518 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
5519 loudspeakers around the user for binaural listening via headphones (audio
5520 formats up to 9 channels supported).
5521 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
5522 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
5523 Austrian Academy of Sciences.
5525 To enable compilation of this filter you need to configure FFmpeg with
5526 @code{--enable-libmysofa}.
5528 The filter accepts the following options:
5532 Set the SOFA file used for rendering.
5535 Set gain applied to audio. Value is in dB. Default is 0.
5538 Set rotation of virtual loudspeakers in deg. Default is 0.
5541 Set elevation of virtual speakers in deg. Default is 0.
5544 Set distance in meters between loudspeakers and the listener with near-field
5545 HRTFs. Default is 1.
5548 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
5549 processing audio in time domain which is slow.
5550 @var{freq} is processing audio in frequency domain which is fast.
5551 Default is @var{freq}.
5554 Set custom positions of virtual loudspeakers. Syntax for this option is:
5555 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
5556 Each virtual loudspeaker is described with short channel name following with
5557 azimuth and elevation in degrees.
5558 Each virtual loudspeaker description is separated by '|'.
5559 For example to override front left and front right channel positions use:
5560 'speakers=FL 45 15|FR 345 15'.
5561 Descriptions with unrecognised channel names are ignored.
5564 Set custom gain for LFE channels. Value is in dB. Default is 0.
5567 Set custom frame size in number of samples. Default is 1024.
5568 Allowed range is from 1024 to 96000. Only used if option @samp{type}
5569 is set to @var{freq}.
5572 Should all IRs be normalized upon importing SOFA file.
5573 By default is enabled.
5576 Should nearest IRs be interpolated with neighbor IRs if exact position
5577 does not match. By default is disabled.
5580 Minphase all IRs upon loading of SOFA file. By default is disabled.
5583 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
5586 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
5589 @subsection Examples
5593 Using ClubFritz6 sofa file:
5595 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
5599 Using ClubFritz12 sofa file and bigger radius with small rotation:
5601 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
5605 Similar as above but with custom speaker positions for front left, front right, back left and back right
5606 and also with custom gain:
5608 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
5615 This filter expands or compresses each half-cycle of audio samples
5616 (local set of samples all above or all below zero and between two nearest zero crossings) depending
5617 on threshold value, so audio reaches target peak value under conditions controlled by below options.
5619 The filter accepts the following options:
5623 Set the expansion target peak value. This specifies the highest allowed absolute amplitude
5624 level for the normalized audio input. Default value is 0.95. Allowed range is from 0.0 to 1.0.
5627 Set the maximum expansion factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5628 This option controls maximum local half-cycle of samples expansion. The maximum expansion
5629 would be such that local peak value reaches target peak value but never to surpass it and that
5630 ratio between new and previous peak value does not surpass this option value.
5632 @item compression, c
5633 Set the maximum compression factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5634 This option controls maximum local half-cycle of samples compression. This option is used
5635 only if @option{threshold} option is set to value greater than 0.0, then in such cases
5636 when local peak is lower or same as value set by @option{threshold} all samples belonging to
5637 that peak's half-cycle will be compressed by current compression factor.
5640 Set the threshold value. Default value is 0.0. Allowed range is from 0.0 to 1.0.
5641 This option specifies which half-cycles of samples will be compressed and which will be expanded.
5642 Any half-cycle samples with their local peak value below or same as this option value will be
5643 compressed by current compression factor, otherwise, if greater than threshold value they will be
5644 expanded with expansion factor so that it could reach peak target value but never surpass it.
5647 Set the expansion raising amount per each half-cycle of samples. Default value is 0.001.
5648 Allowed range is from 0.0 to 1.0. This controls how fast expansion factor is raised per
5649 each new half-cycle until it reaches @option{expansion} value.
5650 Setting this options too high may lead to distortions.
5653 Set the compression raising amount per each half-cycle of samples. Default value is 0.001.
5654 Allowed range is from 0.0 to 1.0. This controls how fast compression factor is raised per
5655 each new half-cycle until it reaches @option{compression} value.
5658 Specify which channels to filter, by default all available channels are filtered.
5661 Enable inverted filtering, by default is disabled. This inverts interpretation of @option{threshold}
5662 option. When enabled any half-cycle of samples with their local peak value below or same as
5663 @option{threshold} option will be expanded otherwise it will be compressed.
5666 Link channels when calculating gain applied to each filtered channel sample, by default is disabled.
5667 When disabled each filtered channel gain calculation is independent, otherwise when this option
5668 is enabled the minimum of all possible gains for each filtered channel is used.
5671 @subsection Commands
5673 This filter supports the all above options as @ref{commands}.
5675 @section stereotools
5677 This filter has some handy utilities to manage stereo signals, for converting
5678 M/S stereo recordings to L/R signal while having control over the parameters
5679 or spreading the stereo image of master track.
5681 The filter accepts the following options:
5685 Set input level before filtering for both channels. Defaults is 1.
5686 Allowed range is from 0.015625 to 64.
5689 Set output level after filtering for both channels. Defaults is 1.
5690 Allowed range is from 0.015625 to 64.
5693 Set input balance between both channels. Default is 0.
5694 Allowed range is from -1 to 1.
5697 Set output balance between both channels. Default is 0.
5698 Allowed range is from -1 to 1.
5701 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
5702 clipping. Disabled by default.
5705 Mute the left channel. Disabled by default.
5708 Mute the right channel. Disabled by default.
5711 Change the phase of the left channel. Disabled by default.
5714 Change the phase of the right channel. Disabled by default.
5717 Set stereo mode. Available values are:
5721 Left/Right to Left/Right, this is default.
5724 Left/Right to Mid/Side.
5727 Mid/Side to Left/Right.
5730 Left/Right to Left/Left.
5733 Left/Right to Right/Right.
5736 Left/Right to Left + Right.
5739 Left/Right to Right/Left.
5742 Mid/Side to Left/Left.
5745 Mid/Side to Right/Right.
5748 Mid/Side to Right/Left.
5751 Left/Right to Left - Right.
5755 Set level of side signal. Default is 1.
5756 Allowed range is from 0.015625 to 64.
5759 Set balance of side signal. Default is 0.
5760 Allowed range is from -1 to 1.
5763 Set level of the middle signal. Default is 1.
5764 Allowed range is from 0.015625 to 64.
5767 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
5770 Set stereo base between mono and inversed channels. Default is 0.
5771 Allowed range is from -1 to 1.
5774 Set delay in milliseconds how much to delay left from right channel and
5775 vice versa. Default is 0. Allowed range is from -20 to 20.
5778 Set S/C level. Default is 1. Allowed range is from 1 to 100.
5781 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
5783 @item bmode_in, bmode_out
5784 Set balance mode for balance_in/balance_out option.
5786 Can be one of the following:
5790 Classic balance mode. Attenuate one channel at time.
5791 Gain is raised up to 1.
5794 Similar as classic mode above but gain is raised up to 2.
5797 Equal power distribution, from -6dB to +6dB range.
5801 @subsection Commands
5803 This filter supports the all above options as @ref{commands}.
5805 @subsection Examples
5809 Apply karaoke like effect:
5811 stereotools=mlev=0.015625
5815 Convert M/S signal to L/R:
5817 "stereotools=mode=ms>lr"
5821 @section stereowiden
5823 This filter enhance the stereo effect by suppressing signal common to both
5824 channels and by delaying the signal of left into right and vice versa,
5825 thereby widening the stereo effect.
5827 The filter accepts the following options:
5831 Time in milliseconds of the delay of left signal into right and vice versa.
5832 Default is 20 milliseconds.
5835 Amount of gain in delayed signal into right and vice versa. Gives a delay
5836 effect of left signal in right output and vice versa which gives widening
5837 effect. Default is 0.3.
5840 Cross feed of left into right with inverted phase. This helps in suppressing
5841 the mono. If the value is 1 it will cancel all the signal common to both
5842 channels. Default is 0.3.
5845 Set level of input signal of original channel. Default is 0.8.
5848 @subsection Commands
5850 This filter supports the all above options except @code{delay} as @ref{commands}.
5852 @section superequalizer
5853 Apply 18 band equalizer.
5855 The filter accepts the following options:
5862 Set 131Hz band gain.
5864 Set 185Hz band gain.
5866 Set 262Hz band gain.
5868 Set 370Hz band gain.
5870 Set 523Hz band gain.
5872 Set 740Hz band gain.
5874 Set 1047Hz band gain.
5876 Set 1480Hz band gain.
5878 Set 2093Hz band gain.
5880 Set 2960Hz band gain.
5882 Set 4186Hz band gain.
5884 Set 5920Hz band gain.
5886 Set 8372Hz band gain.
5888 Set 11840Hz band gain.
5890 Set 16744Hz band gain.
5892 Set 20000Hz band gain.
5896 Apply audio surround upmix filter.
5898 This filter allows to produce multichannel output from audio stream.
5900 The filter accepts the following options:
5904 Set output channel layout. By default, this is @var{5.1}.
5906 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5907 for the required syntax.
5910 Set input channel layout. By default, this is @var{stereo}.
5912 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5913 for the required syntax.
5916 Set input volume level. By default, this is @var{1}.
5919 Set output volume level. By default, this is @var{1}.
5922 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5925 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5928 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5931 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5932 In @var{add} mode, LFE channel is created from input audio and added to output.
5933 In @var{sub} mode, LFE channel is created from input audio and added to output but
5934 also all non-LFE output channels are subtracted with output LFE channel.
5937 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5938 Default is @var{90}.
5941 Set front center input volume. By default, this is @var{1}.
5944 Set front center output volume. By default, this is @var{1}.
5947 Set front left input volume. By default, this is @var{1}.
5950 Set front left output volume. By default, this is @var{1}.
5953 Set front right input volume. By default, this is @var{1}.
5956 Set front right output volume. By default, this is @var{1}.
5959 Set side left input volume. By default, this is @var{1}.
5962 Set side left output volume. By default, this is @var{1}.
5965 Set side right input volume. By default, this is @var{1}.
5968 Set side right output volume. By default, this is @var{1}.
5971 Set back left input volume. By default, this is @var{1}.
5974 Set back left output volume. By default, this is @var{1}.
5977 Set back right input volume. By default, this is @var{1}.
5980 Set back right output volume. By default, this is @var{1}.
5983 Set back center input volume. By default, this is @var{1}.
5986 Set back center output volume. By default, this is @var{1}.
5989 Set LFE input volume. By default, this is @var{1}.
5992 Set LFE output volume. By default, this is @var{1}.
5995 Set spread usage of stereo image across X axis for all channels.
5998 Set spread usage of stereo image across Y axis for all channels.
6000 @item fcx, flx, frx, blx, brx, slx, srx, bcx
6001 Set spread usage of stereo image across X axis for each channel.
6003 @item fcy, fly, fry, bly, bry, sly, sry, bcy
6004 Set spread usage of stereo image across Y axis for each channel.
6007 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
6010 Set window function.
6012 It accepts the following values:
6035 Default is @code{hann}.
6038 Set window overlap. If set to 1, the recommended overlap for selected
6039 window function will be picked. Default is @code{0.5}.
6042 @section treble, highshelf
6044 Boost or cut treble (upper) frequencies of the audio using a two-pole
6045 shelving filter with a response similar to that of a standard
6046 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
6048 The filter accepts the following options:
6052 Give the gain at whichever is the lower of ~22 kHz and the
6053 Nyquist frequency. Its useful range is about -20 (for a large cut)
6054 to +20 (for a large boost). Beware of clipping when using a positive gain.
6057 Set the filter's central frequency and so can be used
6058 to extend or reduce the frequency range to be boosted or cut.
6059 The default value is @code{3000} Hz.
6062 Set method to specify band-width of filter.
6077 Determine how steep is the filter's shelf transition.
6080 Set number of poles. Default is 2.
6083 How much to use filtered signal in output. Default is 1.
6084 Range is between 0 and 1.
6087 Specify which channels to filter, by default all available are filtered.
6090 Normalize biquad coefficients, by default is disabled.
6091 Enabling it will normalize magnitude response at DC to 0dB.
6094 Set transform type of IIR filter.
6103 Set precison of filtering.
6106 Pick automatic sample format depending on surround filters.
6108 Always use signed 16-bit.
6110 Always use signed 32-bit.
6112 Always use float 32-bit.
6114 Always use float 64-bit.
6118 @subsection Commands
6120 This filter supports the following commands:
6123 Change treble frequency.
6124 Syntax for the command is : "@var{frequency}"
6127 Change treble width_type.
6128 Syntax for the command is : "@var{width_type}"
6131 Change treble width.
6132 Syntax for the command is : "@var{width}"
6136 Syntax for the command is : "@var{gain}"
6140 Syntax for the command is : "@var{mix}"
6145 Sinusoidal amplitude modulation.
6147 The filter accepts the following options:
6151 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
6152 (20 Hz or lower) will result in a tremolo effect.
6153 This filter may also be used as a ring modulator by specifying
6154 a modulation frequency higher than 20 Hz.
6155 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6158 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6159 Default value is 0.5.
6164 Sinusoidal phase modulation.
6166 The filter accepts the following options:
6170 Modulation frequency in Hertz.
6171 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6174 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6175 Default value is 0.5.
6180 Adjust the input audio volume.
6182 It accepts the following parameters:
6186 Set audio volume expression.
6188 Output values are clipped to the maximum value.
6190 The output audio volume is given by the relation:
6192 @var{output_volume} = @var{volume} * @var{input_volume}
6195 The default value for @var{volume} is "1.0".
6198 This parameter represents the mathematical precision.
6200 It determines which input sample formats will be allowed, which affects the
6201 precision of the volume scaling.
6205 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
6207 32-bit floating-point; this limits input sample format to FLT. (default)
6209 64-bit floating-point; this limits input sample format to DBL.
6213 Choose the behaviour on encountering ReplayGain side data in input frames.
6217 Remove ReplayGain side data, ignoring its contents (the default).
6220 Ignore ReplayGain side data, but leave it in the frame.
6223 Prefer the track gain, if present.
6226 Prefer the album gain, if present.
6229 @item replaygain_preamp
6230 Pre-amplification gain in dB to apply to the selected replaygain gain.
6232 Default value for @var{replaygain_preamp} is 0.0.
6234 @item replaygain_noclip
6235 Prevent clipping by limiting the gain applied.
6237 Default value for @var{replaygain_noclip} is 1.
6240 Set when the volume expression is evaluated.
6242 It accepts the following values:
6245 only evaluate expression once during the filter initialization, or
6246 when the @samp{volume} command is sent
6249 evaluate expression for each incoming frame
6252 Default value is @samp{once}.
6255 The volume expression can contain the following parameters.
6259 frame number (starting at zero)
6262 @item nb_consumed_samples
6263 number of samples consumed by the filter
6265 number of samples in the current frame
6267 original frame position in the file
6273 PTS at start of stream
6275 time at start of stream
6281 last set volume value
6284 Note that when @option{eval} is set to @samp{once} only the
6285 @var{sample_rate} and @var{tb} variables are available, all other
6286 variables will evaluate to NAN.
6288 @subsection Commands
6290 This filter supports the following commands:
6293 Modify the volume expression.
6294 The command accepts the same syntax of the corresponding option.
6296 If the specified expression is not valid, it is kept at its current
6300 @subsection Examples
6304 Halve the input audio volume:
6308 volume=volume=-6.0206dB
6311 In all the above example the named key for @option{volume} can be
6312 omitted, for example like in:
6318 Increase input audio power by 6 decibels using fixed-point precision:
6320 volume=volume=6dB:precision=fixed
6324 Fade volume after time 10 with an annihilation period of 5 seconds:
6326 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
6330 @section volumedetect
6332 Detect the volume of the input video.
6334 The filter has no parameters. The input is not modified. Statistics about
6335 the volume will be printed in the log when the input stream end is reached.
6337 In particular it will show the mean volume (root mean square), maximum
6338 volume (on a per-sample basis), and the beginning of a histogram of the
6339 registered volume values (from the maximum value to a cumulated 1/1000 of
6342 All volumes are in decibels relative to the maximum PCM value.
6344 @subsection Examples
6346 Here is an excerpt of the output:
6348 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
6349 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
6350 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
6351 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
6352 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
6353 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
6354 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
6355 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
6356 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
6362 The mean square energy is approximately -27 dB, or 10^-2.7.
6364 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
6366 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
6369 In other words, raising the volume by +4 dB does not cause any clipping,
6370 raising it by +5 dB causes clipping for 6 samples, etc.
6372 @c man end AUDIO FILTERS
6374 @chapter Audio Sources
6375 @c man begin AUDIO SOURCES
6377 Below is a description of the currently available audio sources.
6381 Buffer audio frames, and make them available to the filter chain.
6383 This source is mainly intended for a programmatic use, in particular
6384 through the interface defined in @file{libavfilter/buffersrc.h}.
6386 It accepts the following parameters:
6390 The timebase which will be used for timestamps of submitted frames. It must be
6391 either a floating-point number or in @var{numerator}/@var{denominator} form.
6394 The sample rate of the incoming audio buffers.
6397 The sample format of the incoming audio buffers.
6398 Either a sample format name or its corresponding integer representation from
6399 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
6401 @item channel_layout
6402 The channel layout of the incoming audio buffers.
6403 Either a channel layout name from channel_layout_map in
6404 @file{libavutil/channel_layout.c} or its corresponding integer representation
6405 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
6408 The number of channels of the incoming audio buffers.
6409 If both @var{channels} and @var{channel_layout} are specified, then they
6414 @subsection Examples
6417 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
6420 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
6421 Since the sample format with name "s16p" corresponds to the number
6422 6 and the "stereo" channel layout corresponds to the value 0x3, this is
6425 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
6430 Generate an audio signal specified by an expression.
6432 This source accepts in input one or more expressions (one for each
6433 channel), which are evaluated and used to generate a corresponding
6436 This source accepts the following options:
6440 Set the '|'-separated expressions list for each separate channel. In case the
6441 @option{channel_layout} option is not specified, the selected channel layout
6442 depends on the number of provided expressions. Otherwise the last
6443 specified expression is applied to the remaining output channels.
6445 @item channel_layout, c
6446 Set the channel layout. The number of channels in the specified layout
6447 must be equal to the number of specified expressions.
6450 Set the minimum duration of the sourced audio. See
6451 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6452 for the accepted syntax.
6453 Note that the resulting duration may be greater than the specified
6454 duration, as the generated audio is always cut at the end of a
6457 If not specified, or the expressed duration is negative, the audio is
6458 supposed to be generated forever.
6461 Set the number of samples per channel per each output frame,
6464 @item sample_rate, s
6465 Specify the sample rate, default to 44100.
6468 Each expression in @var{exprs} can contain the following constants:
6472 number of the evaluated sample, starting from 0
6475 time of the evaluated sample expressed in seconds, starting from 0
6482 @subsection Examples
6492 Generate a sin signal with frequency of 440 Hz, set sample rate to
6495 aevalsrc="sin(440*2*PI*t):s=8000"
6499 Generate a two channels signal, specify the channel layout (Front
6500 Center + Back Center) explicitly:
6502 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
6506 Generate white noise:
6508 aevalsrc="-2+random(0)"
6512 Generate an amplitude modulated signal:
6514 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
6518 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
6520 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
6527 Generate a FIR coefficients using frequency sampling method.
6529 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6531 The filter accepts the following options:
6535 Set number of filter coefficents in output audio stream.
6536 Default value is 1025.
6539 Set frequency points from where magnitude and phase are set.
6540 This must be in non decreasing order, and first element must be 0, while last element
6541 must be 1. Elements are separated by white spaces.
6544 Set magnitude value for every frequency point set by @option{frequency}.
6545 Number of values must be same as number of frequency points.
6546 Values are separated by white spaces.
6549 Set phase value for every frequency point set by @option{frequency}.
6550 Number of values must be same as number of frequency points.
6551 Values are separated by white spaces.
6553 @item sample_rate, r
6554 Set sample rate, default is 44100.
6557 Set number of samples per each frame. Default is 1024.
6560 Set window function. Default is blackman.
6565 The null audio source, return unprocessed audio frames. It is mainly useful
6566 as a template and to be employed in analysis / debugging tools, or as
6567 the source for filters which ignore the input data (for example the sox
6570 This source accepts the following options:
6574 @item channel_layout, cl
6576 Specifies the channel layout, and can be either an integer or a string
6577 representing a channel layout. The default value of @var{channel_layout}
6580 Check the channel_layout_map definition in
6581 @file{libavutil/channel_layout.c} for the mapping between strings and
6582 channel layout values.
6584 @item sample_rate, r
6585 Specifies the sample rate, and defaults to 44100.
6588 Set the number of samples per requested frames.
6591 Set the duration of the sourced audio. See
6592 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6593 for the accepted syntax.
6595 If not specified, or the expressed duration is negative, the audio is
6596 supposed to be generated forever.
6599 @subsection Examples
6603 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
6605 anullsrc=r=48000:cl=4
6609 Do the same operation with a more obvious syntax:
6611 anullsrc=r=48000:cl=mono
6615 All the parameters need to be explicitly defined.
6619 Synthesize a voice utterance using the libflite library.
6621 To enable compilation of this filter you need to configure FFmpeg with
6622 @code{--enable-libflite}.
6624 Note that versions of the flite library prior to 2.0 are not thread-safe.
6626 The filter accepts the following options:
6631 If set to 1, list the names of the available voices and exit
6632 immediately. Default value is 0.
6635 Set the maximum number of samples per frame. Default value is 512.
6638 Set the filename containing the text to speak.
6641 Set the text to speak.
6644 Set the voice to use for the speech synthesis. Default value is
6645 @code{kal}. See also the @var{list_voices} option.
6648 @subsection Examples
6652 Read from file @file{speech.txt}, and synthesize the text using the
6653 standard flite voice:
6655 flite=textfile=speech.txt
6659 Read the specified text selecting the @code{slt} voice:
6661 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6665 Input text to ffmpeg:
6667 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6671 Make @file{ffplay} speak the specified text, using @code{flite} and
6672 the @code{lavfi} device:
6674 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
6678 For more information about libflite, check:
6679 @url{http://www.festvox.org/flite/}
6683 Generate a noise audio signal.
6685 The filter accepts the following options:
6688 @item sample_rate, r
6689 Specify the sample rate. Default value is 48000 Hz.
6692 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
6696 Specify the duration of the generated audio stream. Not specifying this option
6697 results in noise with an infinite length.
6699 @item color, colour, c
6700 Specify the color of noise. Available noise colors are white, pink, brown,
6701 blue, violet and velvet. Default color is white.
6704 Specify a value used to seed the PRNG.
6707 Set the number of samples per each output frame, default is 1024.
6710 @subsection Examples
6715 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
6717 anoisesrc=d=60:c=pink:r=44100:a=0.5
6723 Generate odd-tap Hilbert transform FIR coefficients.
6725 The resulting stream can be used with @ref{afir} filter for phase-shifting
6726 the signal by 90 degrees.
6728 This is used in many matrix coding schemes and for analytic signal generation.
6729 The process is often written as a multiplication by i (or j), the imaginary unit.
6731 The filter accepts the following options:
6735 @item sample_rate, s
6736 Set sample rate, default is 44100.
6739 Set length of FIR filter, default is 22051.
6742 Set number of samples per each frame.
6745 Set window function to be used when generating FIR coefficients.
6750 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
6752 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6754 The filter accepts the following options:
6757 @item sample_rate, r
6758 Set sample rate, default is 44100.
6761 Set number of samples per each frame. Default is 1024.
6764 Set high-pass frequency. Default is 0.
6767 Set low-pass frequency. Default is 0.
6768 If high-pass frequency is lower than low-pass frequency and low-pass frequency
6769 is higher than 0 then filter will create band-pass filter coefficients,
6770 otherwise band-reject filter coefficients.
6773 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
6776 Set Kaiser window beta.
6779 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
6782 Enable rounding, by default is disabled.
6785 Set number of taps for high-pass filter.
6788 Set number of taps for low-pass filter.
6793 Generate an audio signal made of a sine wave with amplitude 1/8.
6795 The audio signal is bit-exact.
6797 The filter accepts the following options:
6802 Set the carrier frequency. Default is 440 Hz.
6804 @item beep_factor, b
6805 Enable a periodic beep every second with frequency @var{beep_factor} times
6806 the carrier frequency. Default is 0, meaning the beep is disabled.
6808 @item sample_rate, r
6809 Specify the sample rate, default is 44100.
6812 Specify the duration of the generated audio stream.
6814 @item samples_per_frame
6815 Set the number of samples per output frame.
6817 The expression can contain the following constants:
6821 The (sequential) number of the output audio frame, starting from 0.
6824 The PTS (Presentation TimeStamp) of the output audio frame,
6825 expressed in @var{TB} units.
6828 The PTS of the output audio frame, expressed in seconds.
6831 The timebase of the output audio frames.
6834 Default is @code{1024}.
6837 @subsection Examples
6842 Generate a simple 440 Hz sine wave:
6848 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
6852 sine=frequency=220:beep_factor=4:duration=5
6856 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
6859 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
6863 @c man end AUDIO SOURCES
6865 @chapter Audio Sinks
6866 @c man begin AUDIO SINKS
6868 Below is a description of the currently available audio sinks.
6870 @section abuffersink
6872 Buffer audio frames, and make them available to the end of filter chain.
6874 This sink is mainly intended for programmatic use, in particular
6875 through the interface defined in @file{libavfilter/buffersink.h}
6876 or the options system.
6878 It accepts a pointer to an AVABufferSinkContext structure, which
6879 defines the incoming buffers' formats, to be passed as the opaque
6880 parameter to @code{avfilter_init_filter} for initialization.
6883 Null audio sink; do absolutely nothing with the input audio. It is
6884 mainly useful as a template and for use in analysis / debugging
6887 @c man end AUDIO SINKS
6889 @chapter Video Filters
6890 @c man begin VIDEO FILTERS
6892 When you configure your FFmpeg build, you can disable any of the
6893 existing filters using @code{--disable-filters}.
6894 The configure output will show the video filters included in your
6897 Below is a description of the currently available video filters.
6901 Mark a region of interest in a video frame.
6903 The frame data is passed through unchanged, but metadata is attached
6904 to the frame indicating regions of interest which can affect the
6905 behaviour of later encoding. Multiple regions can be marked by
6906 applying the filter multiple times.
6910 Region distance in pixels from the left edge of the frame.
6912 Region distance in pixels from the top edge of the frame.
6914 Region width in pixels.
6916 Region height in pixels.
6918 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
6919 and may contain the following variables:
6922 Width of the input frame.
6924 Height of the input frame.
6928 Quantisation offset to apply within the region.
6930 This must be a real value in the range -1 to +1. A value of zero
6931 indicates no quality change. A negative value asks for better quality
6932 (less quantisation), while a positive value asks for worse quality
6933 (greater quantisation).
6935 The range is calibrated so that the extreme values indicate the
6936 largest possible offset - if the rest of the frame is encoded with the
6937 worst possible quality, an offset of -1 indicates that this region
6938 should be encoded with the best possible quality anyway. Intermediate
6939 values are then interpolated in some codec-dependent way.
6941 For example, in 10-bit H.264 the quantisation parameter varies between
6942 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
6943 this region should be encoded with a QP around one-tenth of the full
6944 range better than the rest of the frame. So, if most of the frame
6945 were to be encoded with a QP of around 30, this region would get a QP
6946 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
6947 An extreme value of -1 would indicate that this region should be
6948 encoded with the best possible quality regardless of the treatment of
6949 the rest of the frame - that is, should be encoded at a QP of -12.
6951 If set to true, remove any existing regions of interest marked on the
6952 frame before adding the new one.
6955 @subsection Examples
6959 Mark the centre quarter of the frame as interesting.
6961 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
6964 Mark the 100-pixel-wide region on the left edge of the frame as very
6965 uninteresting (to be encoded at much lower quality than the rest of
6968 addroi=0:0:100:ih:+1/5
6972 @section alphaextract
6974 Extract the alpha component from the input as a grayscale video. This
6975 is especially useful with the @var{alphamerge} filter.
6979 Add or replace the alpha component of the primary input with the
6980 grayscale value of a second input. This is intended for use with
6981 @var{alphaextract} to allow the transmission or storage of frame
6982 sequences that have alpha in a format that doesn't support an alpha
6985 For example, to reconstruct full frames from a normal YUV-encoded video
6986 and a separate video created with @var{alphaextract}, you might use:
6988 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
6993 Amplify differences between current pixel and pixels of adjacent frames in
6994 same pixel location.
6996 This filter accepts the following options:
7000 Set frame radius. Default is 2. Allowed range is from 1 to 63.
7001 For example radius of 3 will instruct filter to calculate average of 7 frames.
7004 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
7007 Set threshold for difference amplification. Any difference greater or equal to
7008 this value will not alter source pixel. Default is 10.
7009 Allowed range is from 0 to 65535.
7012 Set tolerance for difference amplification. Any difference lower to
7013 this value will not alter source pixel. Default is 0.
7014 Allowed range is from 0 to 65535.
7017 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7018 This option controls maximum possible value that will decrease source pixel value.
7021 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7022 This option controls maximum possible value that will increase source pixel value.
7025 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
7028 @subsection Commands
7030 This filter supports the following @ref{commands} that corresponds to option of same name:
7042 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
7043 and libavformat to work. On the other hand, it is limited to ASS (Advanced
7044 Substation Alpha) subtitles files.
7046 This filter accepts the following option in addition to the common options from
7047 the @ref{subtitles} filter:
7051 Set the shaping engine
7053 Available values are:
7056 The default libass shaping engine, which is the best available.
7058 Fast, font-agnostic shaper that can do only substitutions
7060 Slower shaper using OpenType for substitutions and positioning
7063 The default is @code{auto}.
7067 Apply an Adaptive Temporal Averaging Denoiser to the video input.
7069 The filter accepts the following options:
7073 Set threshold A for 1st plane. Default is 0.02.
7074 Valid range is 0 to 0.3.
7077 Set threshold B for 1st plane. Default is 0.04.
7078 Valid range is 0 to 5.
7081 Set threshold A for 2nd plane. Default is 0.02.
7082 Valid range is 0 to 0.3.
7085 Set threshold B for 2nd plane. Default is 0.04.
7086 Valid range is 0 to 5.
7089 Set threshold A for 3rd plane. Default is 0.02.
7090 Valid range is 0 to 0.3.
7093 Set threshold B for 3rd plane. Default is 0.04.
7094 Valid range is 0 to 5.
7096 Threshold A is designed to react on abrupt changes in the input signal and
7097 threshold B is designed to react on continuous changes in the input signal.
7100 Set number of frames filter will use for averaging. Default is 9. Must be odd
7101 number in range [5, 129].
7104 Set what planes of frame filter will use for averaging. Default is all.
7107 Set what variant of algorithm filter will use for averaging. Default is @code{p} parallel.
7108 Alternatively can be set to @code{s} serial.
7110 Parallel can be faster then serial, while other way around is never true.
7111 Parallel will abort early on first change being greater then thresholds, while serial
7112 will continue processing other side of frames if they are equal or below thresholds.
7117 Set sigma for 1st plane, 2nd plane or 3rd plane. Default is 32767.
7118 Valid range is from 0 to 32767.
7119 This options controls weight for each pixel in radius defined by size.
7120 Default value means every pixel have same weight.
7121 Setting this option to 0 effectively disables filtering.
7124 @subsection Commands
7125 This filter supports same @ref{commands} as options except option @code{s}.
7126 The command accepts the same syntax of the corresponding option.
7130 Apply average blur filter.
7132 The filter accepts the following options:
7136 Set horizontal radius size.
7139 Set which planes to filter. By default all planes are filtered.
7142 Set vertical radius size, if zero it will be same as @code{sizeX}.
7143 Default is @code{0}.
7146 @subsection Commands
7147 This filter supports same commands as options.
7148 The command accepts the same syntax of the corresponding option.
7150 If the specified expression is not valid, it is kept at its current
7155 Compute the bounding box for the non-black pixels in the input frame
7158 This filter computes the bounding box containing all the pixels with a
7159 luminance value greater than the minimum allowed value.
7160 The parameters describing the bounding box are printed on the filter
7163 The filter accepts the following option:
7167 Set the minimal luminance value. Default is @code{16}.
7170 @subsection Commands
7172 This filter supports the all above options as @ref{commands}.
7175 Apply bilateral filter, spatial smoothing while preserving edges.
7177 The filter accepts the following options:
7180 Set sigma of gaussian function to calculate spatial weight.
7181 Allowed range is 0 to 512. Default is 0.1.
7184 Set sigma of gaussian function to calculate range weight.
7185 Allowed range is 0 to 1. Default is 0.1.
7188 Set planes to filter. Default is first only.
7191 @subsection Commands
7193 This filter supports the all above options as @ref{commands}.
7195 @section bitplanenoise
7197 Show and measure bit plane noise.
7199 The filter accepts the following options:
7203 Set which plane to analyze. Default is @code{1}.
7206 Filter out noisy pixels from @code{bitplane} set above.
7207 Default is disabled.
7210 @section blackdetect
7212 Detect video intervals that are (almost) completely black. Can be
7213 useful to detect chapter transitions, commercials, or invalid
7216 The filter outputs its detection analysis to both the log as well as
7217 frame metadata. If a black segment of at least the specified minimum
7218 duration is found, a line with the start and end timestamps as well
7219 as duration is printed to the log with level @code{info}. In addition,
7220 a log line with level @code{debug} is printed per frame showing the
7221 black amount detected for that frame.
7223 The filter also attaches metadata to the first frame of a black
7224 segment with key @code{lavfi.black_start} and to the first frame
7225 after the black segment ends with key @code{lavfi.black_end}. The
7226 value is the frame's timestamp. This metadata is added regardless
7227 of the minimum duration specified.
7229 The filter accepts the following options:
7232 @item black_min_duration, d
7233 Set the minimum detected black duration expressed in seconds. It must
7234 be a non-negative floating point number.
7236 Default value is 2.0.
7238 @item picture_black_ratio_th, pic_th
7239 Set the threshold for considering a picture "black".
7240 Express the minimum value for the ratio:
7242 @var{nb_black_pixels} / @var{nb_pixels}
7245 for which a picture is considered black.
7246 Default value is 0.98.
7248 @item pixel_black_th, pix_th
7249 Set the threshold for considering a pixel "black".
7251 The threshold expresses the maximum pixel luminance value for which a
7252 pixel is considered "black". The provided value is scaled according to
7253 the following equation:
7255 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
7258 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
7259 the input video format, the range is [0-255] for YUV full-range
7260 formats and [16-235] for YUV non full-range formats.
7262 Default value is 0.10.
7265 The following example sets the maximum pixel threshold to the minimum
7266 value, and detects only black intervals of 2 or more seconds:
7268 blackdetect=d=2:pix_th=0.00
7273 Detect frames that are (almost) completely black. Can be useful to
7274 detect chapter transitions or commercials. Output lines consist of
7275 the frame number of the detected frame, the percentage of blackness,
7276 the position in the file if known or -1 and the timestamp in seconds.
7278 In order to display the output lines, you need to set the loglevel at
7279 least to the AV_LOG_INFO value.
7281 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
7282 The value represents the percentage of pixels in the picture that
7283 are below the threshold value.
7285 It accepts the following parameters:
7290 The percentage of the pixels that have to be below the threshold; it defaults to
7293 @item threshold, thresh
7294 The threshold below which a pixel value is considered black; it defaults to
7302 Blend two video frames into each other.
7304 The @code{blend} filter takes two input streams and outputs one
7305 stream, the first input is the "top" layer and second input is
7306 "bottom" layer. By default, the output terminates when the longest input terminates.
7308 The @code{tblend} (time blend) filter takes two consecutive frames
7309 from one single stream, and outputs the result obtained by blending
7310 the new frame on top of the old frame.
7312 A description of the accepted options follows.
7320 Set blend mode for specific pixel component or all pixel components in case
7321 of @var{all_mode}. Default value is @code{normal}.
7323 Available values for component modes are:
7365 Set blend opacity for specific pixel component or all pixel components in case
7366 of @var{all_opacity}. Only used in combination with pixel component blend modes.
7373 Set blend expression for specific pixel component or all pixel components in case
7374 of @var{all_expr}. Note that related mode options will be ignored if those are set.
7376 The expressions can use the following variables:
7380 The sequential number of the filtered frame, starting from @code{0}.
7384 the coordinates of the current sample
7388 the width and height of currently filtered plane
7392 Width and height scale for the plane being filtered. It is the
7393 ratio between the dimensions of the current plane to the luma plane,
7394 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
7395 the luma plane and @code{0.5,0.5} for the chroma planes.
7398 Time of the current frame, expressed in seconds.
7401 Value of pixel component at current location for first video frame (top layer).
7404 Value of pixel component at current location for second video frame (bottom layer).
7408 The @code{blend} filter also supports the @ref{framesync} options.
7410 @subsection Examples
7414 Apply transition from bottom layer to top layer in first 10 seconds:
7416 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
7420 Apply linear horizontal transition from top layer to bottom layer:
7422 blend=all_expr='A*(X/W)+B*(1-X/W)'
7426 Apply 1x1 checkerboard effect:
7428 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
7432 Apply uncover left effect:
7434 blend=all_expr='if(gte(N*SW+X,W),A,B)'
7438 Apply uncover down effect:
7440 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
7444 Apply uncover up-left effect:
7446 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
7450 Split diagonally video and shows top and bottom layer on each side:
7452 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
7456 Display differences between the current and the previous frame:
7458 tblend=all_mode=grainextract
7462 @subsection Commands
7463 This filter supports same @ref{commands} as options.
7467 Denoise frames using Block-Matching 3D algorithm.
7469 The filter accepts the following options.
7473 Set denoising strength. Default value is 1.
7474 Allowed range is from 0 to 999.9.
7475 The denoising algorithm is very sensitive to sigma, so adjust it
7476 according to the source.
7479 Set local patch size. This sets dimensions in 2D.
7482 Set sliding step for processing blocks. Default value is 4.
7483 Allowed range is from 1 to 64.
7484 Smaller values allows processing more reference blocks and is slower.
7487 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
7488 When set to 1, no block matching is done. Larger values allows more blocks
7490 Allowed range is from 1 to 256.
7493 Set radius for search block matching. Default is 9.
7494 Allowed range is from 1 to INT32_MAX.
7497 Set step between two search locations for block matching. Default is 1.
7498 Allowed range is from 1 to 64. Smaller is slower.
7501 Set threshold of mean square error for block matching. Valid range is 0 to
7505 Set thresholding parameter for hard thresholding in 3D transformed domain.
7506 Larger values results in stronger hard-thresholding filtering in frequency
7510 Set filtering estimation mode. Can be @code{basic} or @code{final}.
7511 Default is @code{basic}.
7514 If enabled, filter will use 2nd stream for block matching.
7515 Default is disabled for @code{basic} value of @var{estim} option,
7516 and always enabled if value of @var{estim} is @code{final}.
7519 Set planes to filter. Default is all available except alpha.
7522 @subsection Examples
7526 Basic filtering with bm3d:
7528 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
7532 Same as above, but filtering only luma:
7534 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
7538 Same as above, but with both estimation modes:
7540 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
7544 Same as above, but prefilter with @ref{nlmeans} filter instead:
7546 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
7552 Apply a boxblur algorithm to the input video.
7554 It accepts the following parameters:
7558 @item luma_radius, lr
7559 @item luma_power, lp
7560 @item chroma_radius, cr
7561 @item chroma_power, cp
7562 @item alpha_radius, ar
7563 @item alpha_power, ap
7567 A description of the accepted options follows.
7570 @item luma_radius, lr
7571 @item chroma_radius, cr
7572 @item alpha_radius, ar
7573 Set an expression for the box radius in pixels used for blurring the
7574 corresponding input plane.
7576 The radius value must be a non-negative number, and must not be
7577 greater than the value of the expression @code{min(w,h)/2} for the
7578 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
7581 Default value for @option{luma_radius} is "2". If not specified,
7582 @option{chroma_radius} and @option{alpha_radius} default to the
7583 corresponding value set for @option{luma_radius}.
7585 The expressions can contain the following constants:
7589 The input width and height in pixels.
7593 The input chroma image width and height in pixels.
7597 The horizontal and vertical chroma subsample values. For example, for the
7598 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
7601 @item luma_power, lp
7602 @item chroma_power, cp
7603 @item alpha_power, ap
7604 Specify how many times the boxblur filter is applied to the
7605 corresponding plane.
7607 Default value for @option{luma_power} is 2. If not specified,
7608 @option{chroma_power} and @option{alpha_power} default to the
7609 corresponding value set for @option{luma_power}.
7611 A value of 0 will disable the effect.
7614 @subsection Examples
7618 Apply a boxblur filter with the luma, chroma, and alpha radii
7621 boxblur=luma_radius=2:luma_power=1
7626 Set the luma radius to 2, and alpha and chroma radius to 0:
7628 boxblur=2:1:cr=0:ar=0
7632 Set the luma and chroma radii to a fraction of the video dimension:
7634 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
7640 Deinterlace the input video ("bwdif" stands for "Bob Weaver
7641 Deinterlacing Filter").
7643 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
7644 interpolation algorithms.
7645 It accepts the following parameters:
7649 The interlacing mode to adopt. It accepts one of the following values:
7653 Output one frame for each frame.
7655 Output one frame for each field.
7658 The default value is @code{send_field}.
7661 The picture field parity assumed for the input interlaced video. It accepts one
7662 of the following values:
7666 Assume the top field is first.
7668 Assume the bottom field is first.
7670 Enable automatic detection of field parity.
7673 The default value is @code{auto}.
7674 If the interlacing is unknown or the decoder does not export this information,
7675 top field first will be assumed.
7678 Specify which frames to deinterlace. Accepts one of the following
7683 Deinterlace all frames.
7685 Only deinterlace frames marked as interlaced.
7688 The default value is @code{all}.
7693 Apply Contrast Adaptive Sharpen filter to video stream.
7695 The filter accepts the following options:
7699 Set the sharpening strength. Default value is 0.
7702 Set planes to filter. Default value is to filter all
7703 planes except alpha plane.
7706 @subsection Commands
7707 This filter supports same @ref{commands} as options.
7710 Remove all color information for all colors except for certain one.
7712 The filter accepts the following options:
7716 The color which will not be replaced with neutral chroma.
7719 Similarity percentage with the above color.
7720 0.01 matches only the exact key color, while 1.0 matches everything.
7724 0.0 makes pixels either fully gray, or not gray at all.
7725 Higher values result in more preserved color.
7728 Signals that the color passed is already in YUV instead of RGB.
7730 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7731 This can be used to pass exact YUV values as hexadecimal numbers.
7734 @subsection Commands
7735 This filter supports same @ref{commands} as options.
7736 The command accepts the same syntax of the corresponding option.
7738 If the specified expression is not valid, it is kept at its current
7742 YUV colorspace color/chroma keying.
7744 The filter accepts the following options:
7748 The color which will be replaced with transparency.
7751 Similarity percentage with the key color.
7753 0.01 matches only the exact key color, while 1.0 matches everything.
7758 0.0 makes pixels either fully transparent, or not transparent at all.
7760 Higher values result in semi-transparent pixels, with a higher transparency
7761 the more similar the pixels color is to the key color.
7764 Signals that the color passed is already in YUV instead of RGB.
7766 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7767 This can be used to pass exact YUV values as hexadecimal numbers.
7770 @subsection Commands
7771 This filter supports same @ref{commands} as options.
7772 The command accepts the same syntax of the corresponding option.
7774 If the specified expression is not valid, it is kept at its current
7777 @subsection Examples
7781 Make every green pixel in the input image transparent:
7783 ffmpeg -i input.png -vf chromakey=green out.png
7787 Overlay a greenscreen-video on top of a static black background.
7789 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
7794 Reduce chrominance noise.
7796 The filter accepts the following options:
7800 Set threshold for averaging chrominance values.
7801 Sum of absolute difference of Y, U and V pixel components of current
7802 pixel and neighbour pixels lower than this threshold will be used in
7803 averaging. Luma component is left unchanged and is copied to output.
7804 Default value is 30. Allowed range is from 1 to 200.
7807 Set horizontal radius of rectangle used for averaging.
7808 Allowed range is from 1 to 100. Default value is 5.
7811 Set vertical radius of rectangle used for averaging.
7812 Allowed range is from 1 to 100. Default value is 5.
7815 Set horizontal step when averaging. Default value is 1.
7816 Allowed range is from 1 to 50.
7817 Mostly useful to speed-up filtering.
7820 Set vertical step when averaging. Default value is 1.
7821 Allowed range is from 1 to 50.
7822 Mostly useful to speed-up filtering.
7825 Set Y threshold for averaging chrominance values.
7826 Set finer control for max allowed difference between Y components
7827 of current pixel and neigbour pixels.
7828 Default value is 200. Allowed range is from 1 to 200.
7831 Set U threshold for averaging chrominance values.
7832 Set finer control for max allowed difference between U components
7833 of current pixel and neigbour pixels.
7834 Default value is 200. Allowed range is from 1 to 200.
7837 Set V threshold for averaging chrominance values.
7838 Set finer control for max allowed difference between V components
7839 of current pixel and neigbour pixels.
7840 Default value is 200. Allowed range is from 1 to 200.
7843 @subsection Commands
7844 This filter supports same @ref{commands} as options.
7845 The command accepts the same syntax of the corresponding option.
7847 @section chromashift
7848 Shift chroma pixels horizontally and/or vertically.
7850 The filter accepts the following options:
7853 Set amount to shift chroma-blue horizontally.
7855 Set amount to shift chroma-blue vertically.
7857 Set amount to shift chroma-red horizontally.
7859 Set amount to shift chroma-red vertically.
7861 Set edge mode, can be @var{smear}, default, or @var{warp}.
7864 @subsection Commands
7866 This filter supports the all above options as @ref{commands}.
7870 Display CIE color diagram with pixels overlaid onto it.
7872 The filter accepts the following options:
7887 @item uhdtv, rec2020
7901 Set what gamuts to draw.
7903 See @code{system} option for available values.
7906 Set ciescope size, by default set to 512.
7909 Set intensity used to map input pixel values to CIE diagram.
7912 Set contrast used to draw tongue colors that are out of active color system gamut.
7915 Correct gamma displayed on scope, by default enabled.
7918 Show white point on CIE diagram, by default disabled.
7921 Set input gamma. Used only with XYZ input color space.
7926 Visualize information exported by some codecs.
7928 Some codecs can export information through frames using side-data or other
7929 means. For example, some MPEG based codecs export motion vectors through the
7930 @var{export_mvs} flag in the codec @option{flags2} option.
7932 The filter accepts the following option:
7936 Set motion vectors to visualize.
7938 Available flags for @var{mv} are:
7942 forward predicted MVs of P-frames
7944 forward predicted MVs of B-frames
7946 backward predicted MVs of B-frames
7950 Display quantization parameters using the chroma planes.
7953 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
7955 Available flags for @var{mv_type} are:
7959 forward predicted MVs
7961 backward predicted MVs
7964 @item frame_type, ft
7965 Set frame type to visualize motion vectors of.
7967 Available flags for @var{frame_type} are:
7971 intra-coded frames (I-frames)
7973 predicted frames (P-frames)
7975 bi-directionally predicted frames (B-frames)
7979 @subsection Examples
7983 Visualize forward predicted MVs of all frames using @command{ffplay}:
7985 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
7989 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
7991 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
7995 @section colorbalance
7996 Modify intensity of primary colors (red, green and blue) of input frames.
7998 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
7999 regions for the red-cyan, green-magenta or blue-yellow balance.
8001 A positive adjustment value shifts the balance towards the primary color, a negative
8002 value towards the complementary color.
8004 The filter accepts the following options:
8010 Adjust red, green and blue shadows (darkest pixels).
8015 Adjust red, green and blue midtones (medium pixels).
8020 Adjust red, green and blue highlights (brightest pixels).
8022 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8025 Preserve lightness when changing color balance. Default is disabled.
8028 @subsection Examples
8032 Add red color cast to shadows:
8038 @subsection Commands
8040 This filter supports the all above options as @ref{commands}.
8042 @section colorcontrast
8044 Adjust color contrast between RGB components.
8046 The filter accepts the following options:
8050 Set the red-cyan contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8053 Set the green-magenta contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8056 Set the blue-yellow contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8061 Set the weight of each @code{rc}, @code{gm}, @code{by} option value. Default value is 0.0.
8062 Allowed range is from 0.0 to 1.0. If all weights are 0.0 filtering is disabled.
8065 Set the amount of preserving lightness. Default value is 0.0. Allowed range is from 0.0 to 1.0.
8068 @subsection Commands
8070 This filter supports the all above options as @ref{commands}.
8072 @section colorcorrect
8074 Adjust color white balance selectively for blacks and whites.
8075 This filter operates in YUV colorspace.
8077 The filter accepts the following options:
8081 Set the red shadow spot. Allowed range is from -1.0 to 1.0.
8085 Set the blue shadow spot. Allowed range is from -1.0 to 1.0.
8089 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8093 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8097 Set the amount of saturation. Allowed range is from -3.0 to 3.0.
8101 @subsection Commands
8103 This filter supports the all above options as @ref{commands}.
8105 @section colorchannelmixer
8107 Adjust video input frames by re-mixing color channels.
8109 This filter modifies a color channel by adding the values associated to
8110 the other channels of the same pixels. For example if the value to
8111 modify is red, the output value will be:
8113 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
8116 The filter accepts the following options:
8123 Adjust contribution of input red, green, blue and alpha channels for output red channel.
8124 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
8130 Adjust contribution of input red, green, blue and alpha channels for output green channel.
8131 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
8137 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
8138 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
8144 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
8145 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
8147 Allowed ranges for options are @code{[-2.0, 2.0]}.
8150 Preserve lightness when changing colors. Allowed range is from @code{[0.0, 1.0]}.
8151 Default is @code{0.0}, thus disabled.
8154 @subsection Examples
8158 Convert source to grayscale:
8160 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
8163 Simulate sepia tones:
8165 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
8169 @subsection Commands
8171 This filter supports the all above options as @ref{commands}.
8174 Overlay a solid color on the video stream.
8176 The filter accepts the following options:
8180 Set the color hue. Allowed range is from 0 to 360.
8184 Set the color saturation. Allowed range is from 0 to 1.
8185 Default value is 0.5.
8188 Set the color lightness. Allowed range is from 0 to 1.
8189 Default value is 0.5.
8192 Set the mix of source lightness. By default is set to 1.0.
8193 Allowed range is from 0.0 to 1.0.
8196 @subsection Commands
8198 This filter supports the all above options as @ref{commands}.
8201 RGB colorspace color keying.
8203 The filter accepts the following options:
8207 The color which will be replaced with transparency.
8210 Similarity percentage with the key color.
8212 0.01 matches only the exact key color, while 1.0 matches everything.
8217 0.0 makes pixels either fully transparent, or not transparent at all.
8219 Higher values result in semi-transparent pixels, with a higher transparency
8220 the more similar the pixels color is to the key color.
8223 @subsection Examples
8227 Make every green pixel in the input image transparent:
8229 ffmpeg -i input.png -vf colorkey=green out.png
8233 Overlay a greenscreen-video on top of a static background image.
8235 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
8239 @subsection Commands
8240 This filter supports same @ref{commands} as options.
8241 The command accepts the same syntax of the corresponding option.
8243 If the specified expression is not valid, it is kept at its current
8247 Remove all color information for all RGB colors except for certain one.
8249 The filter accepts the following options:
8253 The color which will not be replaced with neutral gray.
8256 Similarity percentage with the above color.
8257 0.01 matches only the exact key color, while 1.0 matches everything.
8260 Blend percentage. 0.0 makes pixels fully gray.
8261 Higher values result in more preserved color.
8264 @subsection Commands
8265 This filter supports same @ref{commands} as options.
8266 The command accepts the same syntax of the corresponding option.
8268 If the specified expression is not valid, it is kept at its current
8271 @section colorlevels
8273 Adjust video input frames using levels.
8275 The filter accepts the following options:
8282 Adjust red, green, blue and alpha input black point.
8283 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8289 Adjust red, green, blue and alpha input white point.
8290 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
8292 Input levels are used to lighten highlights (bright tones), darken shadows
8293 (dark tones), change the balance of bright and dark tones.
8299 Adjust red, green, blue and alpha output black point.
8300 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
8306 Adjust red, green, blue and alpha output white point.
8307 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
8309 Output levels allows manual selection of a constrained output level range.
8312 @subsection Examples
8316 Make video output darker:
8318 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
8324 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
8328 Make video output lighter:
8330 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
8334 Increase brightness:
8336 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
8340 @subsection Commands
8342 This filter supports the all above options as @ref{commands}.
8344 @section colormatrix
8346 Convert color matrix.
8348 The filter accepts the following options:
8353 Specify the source and destination color matrix. Both values must be
8356 The accepted values are:
8384 For example to convert from BT.601 to SMPTE-240M, use the command:
8386 colormatrix=bt601:smpte240m
8391 Convert colorspace, transfer characteristics or color primaries.
8392 Input video needs to have an even size.
8394 The filter accepts the following options:
8399 Specify all color properties at once.
8401 The accepted values are:
8431 Specify output colorspace.
8433 The accepted values are:
8442 BT.470BG or BT.601-6 625
8445 SMPTE-170M or BT.601-6 525
8454 BT.2020 with non-constant luminance
8460 Specify output transfer characteristics.
8462 The accepted values are:
8474 Constant gamma of 2.2
8477 Constant gamma of 2.8
8480 SMPTE-170M, BT.601-6 625 or BT.601-6 525
8498 BT.2020 for 10-bits content
8501 BT.2020 for 12-bits content
8507 Specify output color primaries.
8509 The accepted values are:
8518 BT.470BG or BT.601-6 625
8521 SMPTE-170M or BT.601-6 525
8545 Specify output color range.
8547 The accepted values are:
8550 TV (restricted) range
8553 MPEG (restricted) range
8564 Specify output color format.
8566 The accepted values are:
8569 YUV 4:2:0 planar 8-bits
8572 YUV 4:2:0 planar 10-bits
8575 YUV 4:2:0 planar 12-bits
8578 YUV 4:2:2 planar 8-bits
8581 YUV 4:2:2 planar 10-bits
8584 YUV 4:2:2 planar 12-bits
8587 YUV 4:4:4 planar 8-bits
8590 YUV 4:4:4 planar 10-bits
8593 YUV 4:4:4 planar 12-bits
8598 Do a fast conversion, which skips gamma/primary correction. This will take
8599 significantly less CPU, but will be mathematically incorrect. To get output
8600 compatible with that produced by the colormatrix filter, use fast=1.
8603 Specify dithering mode.
8605 The accepted values are:
8611 Floyd-Steinberg dithering
8615 Whitepoint adaptation mode.
8617 The accepted values are:
8620 Bradford whitepoint adaptation
8623 von Kries whitepoint adaptation
8626 identity whitepoint adaptation (i.e. no whitepoint adaptation)
8630 Override all input properties at once. Same accepted values as @ref{all}.
8633 Override input colorspace. Same accepted values as @ref{space}.
8636 Override input color primaries. Same accepted values as @ref{primaries}.
8639 Override input transfer characteristics. Same accepted values as @ref{trc}.
8642 Override input color range. Same accepted values as @ref{range}.
8646 The filter converts the transfer characteristics, color space and color
8647 primaries to the specified user values. The output value, if not specified,
8648 is set to a default value based on the "all" property. If that property is
8649 also not specified, the filter will log an error. The output color range and
8650 format default to the same value as the input color range and format. The
8651 input transfer characteristics, color space, color primaries and color range
8652 should be set on the input data. If any of these are missing, the filter will
8653 log an error and no conversion will take place.
8655 For example to convert the input to SMPTE-240M, use the command:
8657 colorspace=smpte240m
8660 @section colortemperature
8661 Adjust color temperature in video to simulate variations in ambient color temperature.
8663 The filter accepts the following options:
8667 Set the temperature in Kelvin. Allowed range is from 1000 to 40000.
8668 Default value is 6500 K.
8671 Set mixing with filtered output. Allowed range is from 0 to 1.
8675 Set the amount of preserving lightness. Allowed range is from 0 to 1.
8679 @subsection Commands
8680 This filter supports same @ref{commands} as options.
8682 @section convolution
8684 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
8686 The filter accepts the following options:
8693 Set matrix for each plane.
8694 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
8695 and from 1 to 49 odd number of signed integers in @var{row} mode.
8701 Set multiplier for calculated value for each plane.
8702 If unset or 0, it will be sum of all matrix elements.
8708 Set bias for each plane. This value is added to the result of the multiplication.
8709 Useful for making the overall image brighter or darker. Default is 0.0.
8715 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
8716 Default is @var{square}.
8719 @subsection Commands
8721 This filter supports the all above options as @ref{commands}.
8723 @subsection Examples
8729 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"
8735 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"
8741 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"
8747 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"
8751 Apply laplacian edge detector which includes diagonals:
8753 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"
8759 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"
8765 Apply 2D convolution of video stream in frequency domain using second stream
8768 The filter accepts the following options:
8772 Set which planes to process.
8775 Set which impulse video frames will be processed, can be @var{first}
8776 or @var{all}. Default is @var{all}.
8779 The @code{convolve} filter also supports the @ref{framesync} options.
8783 Copy the input video source unchanged to the output. This is mainly useful for
8788 Video filtering on GPU using Apple's CoreImage API on OSX.
8790 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8791 processed by video hardware. However, software-based OpenGL implementations
8792 exist which means there is no guarantee for hardware processing. It depends on
8795 There are many filters and image generators provided by Apple that come with a
8796 large variety of options. The filter has to be referenced by its name along
8799 The coreimage filter accepts the following options:
8802 List all available filters and generators along with all their respective
8803 options as well as possible minimum and maximum values along with the default
8810 Specify all filters by their respective name and options.
8811 Use @var{list_filters} to determine all valid filter names and options.
8812 Numerical options are specified by a float value and are automatically clamped
8813 to their respective value range. Vector and color options have to be specified
8814 by a list of space separated float values. Character escaping has to be done.
8815 A special option name @code{default} is available to use default options for a
8818 It is required to specify either @code{default} or at least one of the filter options.
8819 All omitted options are used with their default values.
8820 The syntax of the filter string is as follows:
8822 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8826 Specify a rectangle where the output of the filter chain is copied into the
8827 input image. It is given by a list of space separated float values:
8829 output_rect=x\ y\ width\ height
8831 If not given, the output rectangle equals the dimensions of the input image.
8832 The output rectangle is automatically cropped at the borders of the input
8833 image. Negative values are valid for each component.
8835 output_rect=25\ 25\ 100\ 100
8839 Several filters can be chained for successive processing without GPU-HOST
8840 transfers allowing for fast processing of complex filter chains.
8841 Currently, only filters with zero (generators) or exactly one (filters) input
8842 image and one output image are supported. Also, transition filters are not yet
8845 Some filters generate output images with additional padding depending on the
8846 respective filter kernel. The padding is automatically removed to ensure the
8847 filter output has the same size as the input image.
8849 For image generators, the size of the output image is determined by the
8850 previous output image of the filter chain or the input image of the whole
8851 filterchain, respectively. The generators do not use the pixel information of
8852 this image to generate their output. However, the generated output is
8853 blended onto this image, resulting in partial or complete coverage of the
8856 The @ref{coreimagesrc} video source can be used for generating input images
8857 which are directly fed into the filter chain. By using it, providing input
8858 images by another video source or an input video is not required.
8860 @subsection Examples
8865 List all filters available:
8867 coreimage=list_filters=true
8871 Use the CIBoxBlur filter with default options to blur an image:
8873 coreimage=filter=CIBoxBlur@@default
8877 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8878 its center at 100x100 and a radius of 50 pixels:
8880 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8884 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8885 given as complete and escaped command-line for Apple's standard bash shell:
8887 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8893 Cover a rectangular object
8895 It accepts the following options:
8899 Filepath of the optional cover image, needs to be in yuv420.
8904 It accepts the following values:
8907 cover it by the supplied image
8909 cover it by interpolating the surrounding pixels
8912 Default value is @var{blur}.
8915 @subsection Examples
8919 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8921 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8927 Crop the input video to given dimensions.
8929 It accepts the following parameters:
8933 The width of the output video. It defaults to @code{iw}.
8934 This expression is evaluated only once during the filter
8935 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8938 The height of the output video. It defaults to @code{ih}.
8939 This expression is evaluated only once during the filter
8940 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8943 The horizontal position, in the input video, of the left edge of the output
8944 video. It defaults to @code{(in_w-out_w)/2}.
8945 This expression is evaluated per-frame.
8948 The vertical position, in the input video, of the top edge of the output video.
8949 It defaults to @code{(in_h-out_h)/2}.
8950 This expression is evaluated per-frame.
8953 If set to 1 will force the output display aspect ratio
8954 to be the same of the input, by changing the output sample aspect
8955 ratio. It defaults to 0.
8958 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8959 width/height/x/y as specified and will not be rounded to nearest smaller value.
8963 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8964 expressions containing the following constants:
8969 The computed values for @var{x} and @var{y}. They are evaluated for
8974 The input width and height.
8978 These are the same as @var{in_w} and @var{in_h}.
8982 The output (cropped) width and height.
8986 These are the same as @var{out_w} and @var{out_h}.
8989 same as @var{iw} / @var{ih}
8992 input sample aspect ratio
8995 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8999 horizontal and vertical chroma subsample values. For example for the
9000 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9003 The number of the input frame, starting from 0.
9006 the position in the file of the input frame, NAN if unknown
9009 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
9013 The expression for @var{out_w} may depend on the value of @var{out_h},
9014 and the expression for @var{out_h} may depend on @var{out_w}, but they
9015 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
9016 evaluated after @var{out_w} and @var{out_h}.
9018 The @var{x} and @var{y} parameters specify the expressions for the
9019 position of the top-left corner of the output (non-cropped) area. They
9020 are evaluated for each frame. If the evaluated value is not valid, it
9021 is approximated to the nearest valid value.
9023 The expression for @var{x} may depend on @var{y}, and the expression
9024 for @var{y} may depend on @var{x}.
9026 @subsection Examples
9030 Crop area with size 100x100 at position (12,34).
9035 Using named options, the example above becomes:
9037 crop=w=100:h=100:x=12:y=34
9041 Crop the central input area with size 100x100:
9047 Crop the central input area with size 2/3 of the input video:
9049 crop=2/3*in_w:2/3*in_h
9053 Crop the input video central square:
9060 Delimit the rectangle with the top-left corner placed at position
9061 100:100 and the right-bottom corner corresponding to the right-bottom
9062 corner of the input image.
9064 crop=in_w-100:in_h-100:100:100
9068 Crop 10 pixels from the left and right borders, and 20 pixels from
9069 the top and bottom borders
9071 crop=in_w-2*10:in_h-2*20
9075 Keep only the bottom right quarter of the input image:
9077 crop=in_w/2:in_h/2:in_w/2:in_h/2
9081 Crop height for getting Greek harmony:
9083 crop=in_w:1/PHI*in_w
9087 Apply trembling effect:
9089 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)
9093 Apply erratic camera effect depending on timestamp:
9095 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)"
9099 Set x depending on the value of y:
9101 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
9105 @subsection Commands
9107 This filter supports the following commands:
9113 Set width/height of the output video and the horizontal/vertical position
9115 The command accepts the same syntax of the corresponding option.
9117 If the specified expression is not valid, it is kept at its current
9123 Auto-detect the crop size.
9125 It calculates the necessary cropping parameters and prints the
9126 recommended parameters via the logging system. The detected dimensions
9127 correspond to the non-black area of the input video.
9129 It accepts the following parameters:
9134 Set higher black value threshold, which can be optionally specified
9135 from nothing (0) to everything (255 for 8-bit based formats). An intensity
9136 value greater to the set value is considered non-black. It defaults to 24.
9137 You can also specify a value between 0.0 and 1.0 which will be scaled depending
9138 on the bitdepth of the pixel format.
9141 The value which the width/height should be divisible by. It defaults to
9142 16. The offset is automatically adjusted to center the video. Use 2 to
9143 get only even dimensions (needed for 4:2:2 video). 16 is best when
9144 encoding to most video codecs.
9147 Set the number of initial frames for which evaluation is skipped.
9148 Default is 2. Range is 0 to INT_MAX.
9150 @item reset_count, reset
9151 Set the counter that determines after how many frames cropdetect will
9152 reset the previously detected largest video area and start over to
9153 detect the current optimal crop area. Default value is 0.
9155 This can be useful when channel logos distort the video area. 0
9156 indicates 'never reset', and returns the largest area encountered during
9163 Delay video filtering until a given wallclock timestamp. The filter first
9164 passes on @option{preroll} amount of frames, then it buffers at most
9165 @option{buffer} amount of frames and waits for the cue. After reaching the cue
9166 it forwards the buffered frames and also any subsequent frames coming in its
9169 The filter can be used synchronize the output of multiple ffmpeg processes for
9170 realtime output devices like decklink. By putting the delay in the filtering
9171 chain and pre-buffering frames the process can pass on data to output almost
9172 immediately after the target wallclock timestamp is reached.
9174 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
9180 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
9183 The duration of content to pass on as preroll expressed in seconds. Default is 0.
9186 The maximum duration of content to buffer before waiting for the cue expressed
9187 in seconds. Default is 0.
9194 Apply color adjustments using curves.
9196 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
9197 component (red, green and blue) has its values defined by @var{N} key points
9198 tied from each other using a smooth curve. The x-axis represents the pixel
9199 values from the input frame, and the y-axis the new pixel values to be set for
9202 By default, a component curve is defined by the two points @var{(0;0)} and
9203 @var{(1;1)}. This creates a straight line where each original pixel value is
9204 "adjusted" to its own value, which means no change to the image.
9206 The filter allows you to redefine these two points and add some more. A new
9207 curve (using a natural cubic spline interpolation) will be define to pass
9208 smoothly through all these new coordinates. The new defined points needs to be
9209 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
9210 be in the @var{[0;1]} interval. If the computed curves happened to go outside
9211 the vector spaces, the values will be clipped accordingly.
9213 The filter accepts the following options:
9217 Select one of the available color presets. This option can be used in addition
9218 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
9219 options takes priority on the preset values.
9220 Available presets are:
9223 @item color_negative
9226 @item increase_contrast
9228 @item linear_contrast
9229 @item medium_contrast
9231 @item strong_contrast
9234 Default is @code{none}.
9236 Set the master key points. These points will define a second pass mapping. It
9237 is sometimes called a "luminance" or "value" mapping. It can be used with
9238 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
9239 post-processing LUT.
9241 Set the key points for the red component.
9243 Set the key points for the green component.
9245 Set the key points for the blue component.
9247 Set the key points for all components (not including master).
9248 Can be used in addition to the other key points component
9249 options. In this case, the unset component(s) will fallback on this
9250 @option{all} setting.
9252 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
9254 Save Gnuplot script of the curves in specified file.
9257 To avoid some filtergraph syntax conflicts, each key points list need to be
9258 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
9260 @subsection Examples
9264 Increase slightly the middle level of blue:
9266 curves=blue='0/0 0.5/0.58 1/1'
9272 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'
9274 Here we obtain the following coordinates for each components:
9277 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9279 @code{(0;0) (0.50;0.48) (1;1)}
9281 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9285 The previous example can also be achieved with the associated built-in preset:
9287 curves=preset=vintage
9297 Use a Photoshop preset and redefine the points of the green component:
9299 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9303 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9304 and @command{gnuplot}:
9306 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9307 gnuplot -p /tmp/curves.plt
9313 Video data analysis filter.
9315 This filter shows hexadecimal pixel values of part of video.
9317 The filter accepts the following options:
9321 Set output video size.
9324 Set x offset from where to pick pixels.
9327 Set y offset from where to pick pixels.
9330 Set scope mode, can be one of the following:
9333 Draw hexadecimal pixel values with white color on black background.
9336 Draw hexadecimal pixel values with input video pixel color on black
9340 Draw hexadecimal pixel values on color background picked from input video,
9341 the text color is picked in such way so its always visible.
9345 Draw rows and columns numbers on left and top of video.
9348 Set background opacity.
9351 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9354 Set pixel components to display. By default all pixel components are displayed.
9358 Apply Directional blur filter.
9360 The filter accepts the following options:
9364 Set angle of directional blur. Default is @code{45}.
9367 Set radius of directional blur. Default is @code{5}.
9370 Set which planes to filter. By default all planes are filtered.
9373 @subsection Commands
9374 This filter supports same @ref{commands} as options.
9375 The command accepts the same syntax of the corresponding option.
9377 If the specified expression is not valid, it is kept at its current
9382 Denoise frames using 2D DCT (frequency domain filtering).
9384 This filter is not designed for real time.
9386 The filter accepts the following options:
9390 Set the noise sigma constant.
9392 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9393 coefficient (absolute value) below this threshold with be dropped.
9395 If you need a more advanced filtering, see @option{expr}.
9397 Default is @code{0}.
9400 Set number overlapping pixels for each block. Since the filter can be slow, you
9401 may want to reduce this value, at the cost of a less effective filter and the
9402 risk of various artefacts.
9404 If the overlapping value doesn't permit processing the whole input width or
9405 height, a warning will be displayed and according borders won't be denoised.
9407 Default value is @var{blocksize}-1, which is the best possible setting.
9410 Set the coefficient factor expression.
9412 For each coefficient of a DCT block, this expression will be evaluated as a
9413 multiplier value for the coefficient.
9415 If this is option is set, the @option{sigma} option will be ignored.
9417 The absolute value of the coefficient can be accessed through the @var{c}
9421 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9422 @var{blocksize}, which is the width and height of the processed blocks.
9424 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9425 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9426 on the speed processing. Also, a larger block size does not necessarily means a
9430 @subsection Examples
9432 Apply a denoise with a @option{sigma} of @code{4.5}:
9437 The same operation can be achieved using the expression system:
9439 dctdnoiz=e='gte(c, 4.5*3)'
9442 Violent denoise using a block size of @code{16x16}:
9449 Remove banding artifacts from input video.
9450 It works by replacing banded pixels with average value of referenced pixels.
9452 The filter accepts the following options:
9459 Set banding detection threshold for each plane. Default is 0.02.
9460 Valid range is 0.00003 to 0.5.
9461 If difference between current pixel and reference pixel is less than threshold,
9462 it will be considered as banded.
9465 Banding detection range in pixels. Default is 16. If positive, random number
9466 in range 0 to set value will be used. If negative, exact absolute value
9468 The range defines square of four pixels around current pixel.
9471 Set direction in radians from which four pixel will be compared. If positive,
9472 random direction from 0 to set direction will be picked. If negative, exact of
9473 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9474 will pick only pixels on same row and -PI/2 will pick only pixels on same
9478 If enabled, current pixel is compared with average value of all four
9479 surrounding pixels. The default is enabled. If disabled current pixel is
9480 compared with all four surrounding pixels. The pixel is considered banded
9481 if only all four differences with surrounding pixels are less than threshold.
9484 If enabled, current pixel is changed if and only if all pixel components are banded,
9485 e.g. banding detection threshold is triggered for all color components.
9486 The default is disabled.
9491 Remove blocking artifacts from input video.
9493 The filter accepts the following options:
9497 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9498 This controls what kind of deblocking is applied.
9501 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9507 Set blocking detection thresholds. Allowed range is 0 to 1.
9508 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9509 Using higher threshold gives more deblocking strength.
9510 Setting @var{alpha} controls threshold detection at exact edge of block.
9511 Remaining options controls threshold detection near the edge. Each one for
9512 below/above or left/right. Setting any of those to @var{0} disables
9516 Set planes to filter. Default is to filter all available planes.
9519 @subsection Examples
9523 Deblock using weak filter and block size of 4 pixels.
9525 deblock=filter=weak:block=4
9529 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9530 deblocking more edges.
9532 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9536 Similar as above, but filter only first plane.
9538 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9542 Similar as above, but filter only second and third plane.
9544 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9548 @subsection Commands
9550 This filter supports the all above options as @ref{commands}.
9555 Drop duplicated frames at regular intervals.
9557 The filter accepts the following options:
9561 Set the number of frames from which one will be dropped. Setting this to
9562 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9563 Default is @code{5}.
9566 Set the threshold for duplicate detection. If the difference metric for a frame
9567 is less than or equal to this value, then it is declared as duplicate. Default
9571 Set scene change threshold. Default is @code{15}.
9575 Set the size of the x and y-axis blocks used during metric calculations.
9576 Larger blocks give better noise suppression, but also give worse detection of
9577 small movements. Must be a power of two. Default is @code{32}.
9580 Mark main input as a pre-processed input and activate clean source input
9581 stream. This allows the input to be pre-processed with various filters to help
9582 the metrics calculation while keeping the frame selection lossless. When set to
9583 @code{1}, the first stream is for the pre-processed input, and the second
9584 stream is the clean source from where the kept frames are chosen. Default is
9588 Set whether or not chroma is considered in the metric calculations. Default is
9594 Apply 2D deconvolution of video stream in frequency domain using second stream
9597 The filter accepts the following options:
9601 Set which planes to process.
9604 Set which impulse video frames will be processed, can be @var{first}
9605 or @var{all}. Default is @var{all}.
9608 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9609 and height are not same and not power of 2 or if stream prior to convolving
9613 The @code{deconvolve} filter also supports the @ref{framesync} options.
9617 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9619 It accepts the following options:
9623 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9624 @var{rainbows} for cross-color reduction.
9627 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9630 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9633 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9636 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9641 Apply deflate effect to the video.
9643 This filter replaces the pixel by the local(3x3) average by taking into account
9644 only values lower than the pixel.
9646 It accepts the following options:
9653 Limit the maximum change for each plane, default is 65535.
9654 If 0, plane will remain unchanged.
9657 @subsection Commands
9659 This filter supports the all above options as @ref{commands}.
9663 Remove temporal frame luminance variations.
9665 It accepts the following options:
9669 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9672 Set averaging mode to smooth temporal luminance variations.
9674 Available values are:
9699 Do not actually modify frame. Useful when one only wants metadata.
9704 Remove judder produced by partially interlaced telecined content.
9706 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9707 source was partially telecined content then the output of @code{pullup,dejudder}
9708 will have a variable frame rate. May change the recorded frame rate of the
9709 container. Aside from that change, this filter will not affect constant frame
9712 The option available in this filter is:
9716 Specify the length of the window over which the judder repeats.
9718 Accepts any integer greater than 1. Useful values are:
9722 If the original was telecined from 24 to 30 fps (Film to NTSC).
9725 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9728 If a mixture of the two.
9731 The default is @samp{4}.
9736 Suppress a TV station logo by a simple interpolation of the surrounding
9737 pixels. Just set a rectangle covering the logo and watch it disappear
9738 (and sometimes something even uglier appear - your mileage may vary).
9740 It accepts the following parameters:
9745 Specify the top left corner coordinates of the logo. They must be
9750 Specify the width and height of the logo to clear. They must be
9754 Specify the thickness of the fuzzy edge of the rectangle (added to
9755 @var{w} and @var{h}). The default value is 1. This option is
9756 deprecated, setting higher values should no longer be necessary and
9760 When set to 1, a green rectangle is drawn on the screen to simplify
9761 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9762 The default value is 0.
9764 The rectangle is drawn on the outermost pixels which will be (partly)
9765 replaced with interpolated values. The values of the next pixels
9766 immediately outside this rectangle in each direction will be used to
9767 compute the interpolated pixel values inside the rectangle.
9771 @subsection Examples
9775 Set a rectangle covering the area with top left corner coordinates 0,0
9776 and size 100x77, and a band of size 10:
9778 delogo=x=0:y=0:w=100:h=77:band=10
9786 Remove the rain in the input image/video by applying the derain methods based on
9787 convolutional neural networks. Supported models:
9791 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9792 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9795 Training as well as model generation scripts are provided in
9796 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9798 Native model files (.model) can be generated from TensorFlow model
9799 files (.pb) by using tools/python/convert.py
9801 The filter accepts the following options:
9805 Specify which filter to use. This option accepts the following values:
9809 Derain filter. To conduct derain filter, you need to use a derain model.
9812 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9814 Default value is @samp{derain}.
9817 Specify which DNN backend to use for model loading and execution. This option accepts
9818 the following values:
9822 Native implementation of DNN loading and execution.
9825 TensorFlow backend. To enable this backend you
9826 need to install the TensorFlow for C library (see
9827 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9828 @code{--enable-libtensorflow}
9830 Default value is @samp{native}.
9833 Set path to model file specifying network architecture and its parameters.
9834 Note that different backends use different file formats. TensorFlow and native
9835 backend can load files for only its format.
9838 It can also be finished with @ref{dnn_processing} filter.
9842 Attempt to fix small changes in horizontal and/or vertical shift. This
9843 filter helps remove camera shake from hand-holding a camera, bumping a
9844 tripod, moving on a vehicle, etc.
9846 The filter accepts the following options:
9854 Specify a rectangular area where to limit the search for motion
9856 If desired the search for motion vectors can be limited to a
9857 rectangular area of the frame defined by its top left corner, width
9858 and height. These parameters have the same meaning as the drawbox
9859 filter which can be used to visualise the position of the bounding
9862 This is useful when simultaneous movement of subjects within the frame
9863 might be confused for camera motion by the motion vector search.
9865 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9866 then the full frame is used. This allows later options to be set
9867 without specifying the bounding box for the motion vector search.
9869 Default - search the whole frame.
9873 Specify the maximum extent of movement in x and y directions in the
9874 range 0-64 pixels. Default 16.
9877 Specify how to generate pixels to fill blanks at the edge of the
9878 frame. Available values are:
9881 Fill zeroes at blank locations
9883 Original image at blank locations
9885 Extruded edge value at blank locations
9887 Mirrored edge at blank locations
9889 Default value is @samp{mirror}.
9892 Specify the blocksize to use for motion search. Range 4-128 pixels,
9896 Specify the contrast threshold for blocks. Only blocks with more than
9897 the specified contrast (difference between darkest and lightest
9898 pixels) will be considered. Range 1-255, default 125.
9901 Specify the search strategy. Available values are:
9904 Set exhaustive search
9906 Set less exhaustive search.
9908 Default value is @samp{exhaustive}.
9911 If set then a detailed log of the motion search is written to the
9918 Remove unwanted contamination of foreground colors, caused by reflected color of
9919 greenscreen or bluescreen.
9921 This filter accepts the following options:
9925 Set what type of despill to use.
9928 Set how spillmap will be generated.
9931 Set how much to get rid of still remaining spill.
9934 Controls amount of red in spill area.
9937 Controls amount of green in spill area.
9938 Should be -1 for greenscreen.
9941 Controls amount of blue in spill area.
9942 Should be -1 for bluescreen.
9945 Controls brightness of spill area, preserving colors.
9948 Modify alpha from generated spillmap.
9951 @subsection Commands
9953 This filter supports the all above options as @ref{commands}.
9957 Apply an exact inverse of the telecine operation. It requires a predefined
9958 pattern specified using the pattern option which must be the same as that passed
9959 to the telecine filter.
9961 This filter accepts the following options:
9970 The default value is @code{top}.
9974 A string of numbers representing the pulldown pattern you wish to apply.
9975 The default value is @code{23}.
9978 A number representing position of the first frame with respect to the telecine
9979 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9984 Apply dilation effect to the video.
9986 This filter replaces the pixel by the local(3x3) maximum.
9988 It accepts the following options:
9995 Limit the maximum change for each plane, default is 65535.
9996 If 0, plane will remain unchanged.
9999 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10002 Flags to local 3x3 coordinates maps like this:
10009 @subsection Commands
10011 This filter supports the all above options as @ref{commands}.
10015 Displace pixels as indicated by second and third input stream.
10017 It takes three input streams and outputs one stream, the first input is the
10018 source, and second and third input are displacement maps.
10020 The second input specifies how much to displace pixels along the
10021 x-axis, while the third input specifies how much to displace pixels
10023 If one of displacement map streams terminates, last frame from that
10024 displacement map will be used.
10026 Note that once generated, displacements maps can be reused over and over again.
10028 A description of the accepted options follows.
10032 Set displace behavior for pixels that are out of range.
10034 Available values are:
10037 Missing pixels are replaced by black pixels.
10040 Adjacent pixels will spread out to replace missing pixels.
10043 Out of range pixels are wrapped so they point to pixels of other side.
10046 Out of range pixels will be replaced with mirrored pixels.
10048 Default is @samp{smear}.
10052 @subsection Examples
10056 Add ripple effect to rgb input of video size hd720:
10058 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
10062 Add wave effect to rgb input of video size hd720:
10064 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
10068 @anchor{dnn_processing}
10069 @section dnn_processing
10071 Do image processing with deep neural networks. It works together with another filter
10072 which converts the pixel format of the Frame to what the dnn network requires.
10074 The filter accepts the following options:
10078 Specify which DNN backend to use for model loading and execution. This option accepts
10079 the following values:
10083 Native implementation of DNN loading and execution.
10086 TensorFlow backend. To enable this backend you
10087 need to install the TensorFlow for C library (see
10088 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10089 @code{--enable-libtensorflow}
10092 OpenVINO backend. To enable this backend you
10093 need to build and install the OpenVINO for C library (see
10094 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10095 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10096 be needed if the header files and libraries are not installed into system path)
10100 Default value is @samp{native}.
10103 Set path to model file specifying network architecture and its parameters.
10104 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10105 backend can load files for only its format.
10107 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10110 Set the input name of the dnn network.
10113 Set the output name of the dnn network.
10116 use DNN async execution if set (default: set),
10117 roll back to sync execution if the backend does not support async.
10121 @subsection Examples
10125 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10127 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10131 Halve the pixel value of the frame with format gray32f:
10133 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
10137 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10139 ./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
10143 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10145 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10152 Draw a colored box on the input image.
10154 It accepts the following parameters:
10159 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10163 The expressions which specify the width and height of the box; if 0 they are interpreted as
10164 the input width and height. It defaults to 0.
10167 Specify the color of the box to write. For the general syntax of this option,
10168 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10169 value @code{invert} is used, the box edge color is the same as the
10170 video with inverted luma.
10173 The expression which sets the thickness of the box edge.
10174 A value of @code{fill} will create a filled box. Default value is @code{3}.
10176 See below for the list of accepted constants.
10179 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10180 will overwrite the video's color and alpha pixels.
10181 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10184 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10185 following constants:
10189 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10193 horizontal and vertical chroma subsample values. For example for the
10194 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10198 The input width and height.
10201 The input sample aspect ratio.
10205 The x and y offset coordinates where the box is drawn.
10209 The width and height of the drawn box.
10212 The thickness of the drawn box.
10214 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10215 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10219 @subsection Examples
10223 Draw a black box around the edge of the input image:
10229 Draw a box with color red and an opacity of 50%:
10231 drawbox=10:20:200:60:red@@0.5
10234 The previous example can be specified as:
10236 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10240 Fill the box with pink color:
10242 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10246 Draw a 2-pixel red 2.40:1 mask:
10248 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
10252 @subsection Commands
10253 This filter supports same commands as options.
10254 The command accepts the same syntax of the corresponding option.
10256 If the specified expression is not valid, it is kept at its current
10261 Draw a graph using input video metadata.
10263 It accepts the following parameters:
10267 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10270 Set 1st foreground color expression.
10273 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10276 Set 2nd foreground color expression.
10279 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10282 Set 3rd foreground color expression.
10285 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10288 Set 4th foreground color expression.
10291 Set minimal value of metadata value.
10294 Set maximal value of metadata value.
10297 Set graph background color. Default is white.
10302 Available values for mode is:
10309 Default is @code{line}.
10314 Available values for slide is:
10317 Draw new frame when right border is reached.
10320 Replace old columns with new ones.
10323 Scroll from right to left.
10326 Scroll from left to right.
10329 Draw single picture.
10332 Default is @code{frame}.
10335 Set size of graph video. For the syntax of this option, check the
10336 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10337 The default value is @code{900x256}.
10340 Set the output frame rate. Default value is @code{25}.
10342 The foreground color expressions can use the following variables:
10345 Minimal value of metadata value.
10348 Maximal value of metadata value.
10351 Current metadata key value.
10354 The color is defined as 0xAABBGGRR.
10357 Example using metadata from @ref{signalstats} filter:
10359 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10362 Example using metadata from @ref{ebur128} filter:
10364 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10369 Draw a grid on the input image.
10371 It accepts the following parameters:
10376 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10380 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10381 input width and height, respectively, minus @code{thickness}, so image gets
10382 framed. Default to 0.
10385 Specify the color of the grid. For the general syntax of this option,
10386 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10387 value @code{invert} is used, the grid color is the same as the
10388 video with inverted luma.
10391 The expression which sets the thickness of the grid line. Default value is @code{1}.
10393 See below for the list of accepted constants.
10396 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10397 will overwrite the video's color and alpha pixels.
10398 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10401 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10402 following constants:
10406 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10410 horizontal and vertical chroma subsample values. For example for the
10411 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10415 The input grid cell width and height.
10418 The input sample aspect ratio.
10422 The x and y coordinates of some point of grid intersection (meant to configure offset).
10426 The width and height of the drawn cell.
10429 The thickness of the drawn cell.
10431 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10432 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10436 @subsection Examples
10440 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10442 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10446 Draw a white 3x3 grid with an opacity of 50%:
10448 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10452 @subsection Commands
10453 This filter supports same commands as options.
10454 The command accepts the same syntax of the corresponding option.
10456 If the specified expression is not valid, it is kept at its current
10462 Draw a text string or text from a specified file on top of a video, using the
10463 libfreetype library.
10465 To enable compilation of this filter, you need to configure FFmpeg with
10466 @code{--enable-libfreetype}.
10467 To enable default font fallback and the @var{font} option you need to
10468 configure FFmpeg with @code{--enable-libfontconfig}.
10469 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10470 @code{--enable-libfribidi}.
10474 It accepts the following parameters:
10479 Used to draw a box around text using the background color.
10480 The value must be either 1 (enable) or 0 (disable).
10481 The default value of @var{box} is 0.
10484 Set the width of the border to be drawn around the box using @var{boxcolor}.
10485 The default value of @var{boxborderw} is 0.
10488 The color to be used for drawing box around text. For the syntax of this
10489 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10491 The default value of @var{boxcolor} is "white".
10494 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10495 The default value of @var{line_spacing} is 0.
10498 Set the width of the border to be drawn around the text using @var{bordercolor}.
10499 The default value of @var{borderw} is 0.
10502 Set the color to be used for drawing border around text. For the syntax of this
10503 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10505 The default value of @var{bordercolor} is "black".
10508 Select how the @var{text} is expanded. Can be either @code{none},
10509 @code{strftime} (deprecated) or
10510 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10514 Set a start time for the count. Value is in microseconds. Only applied
10515 in the deprecated strftime expansion mode. To emulate in normal expansion
10516 mode use the @code{pts} function, supplying the start time (in seconds)
10517 as the second argument.
10520 If true, check and fix text coords to avoid clipping.
10523 The color to be used for drawing fonts. For the syntax of this option, check
10524 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10526 The default value of @var{fontcolor} is "black".
10528 @item fontcolor_expr
10529 String which is expanded the same way as @var{text} to obtain dynamic
10530 @var{fontcolor} value. By default this option has empty value and is not
10531 processed. When this option is set, it overrides @var{fontcolor} option.
10534 The font family to be used for drawing text. By default Sans.
10537 The font file to be used for drawing text. The path must be included.
10538 This parameter is mandatory if the fontconfig support is disabled.
10541 Draw the text applying alpha blending. The value can
10542 be a number between 0.0 and 1.0.
10543 The expression accepts the same variables @var{x, y} as well.
10544 The default value is 1.
10545 Please see @var{fontcolor_expr}.
10548 The font size to be used for drawing text.
10549 The default value of @var{fontsize} is 16.
10552 If set to 1, attempt to shape the text (for example, reverse the order of
10553 right-to-left text and join Arabic characters) before drawing it.
10554 Otherwise, just draw the text exactly as given.
10555 By default 1 (if supported).
10557 @item ft_load_flags
10558 The flags to be used for loading the fonts.
10560 The flags map the corresponding flags supported by libfreetype, and are
10561 a combination of the following values:
10568 @item vertical_layout
10569 @item force_autohint
10572 @item ignore_global_advance_width
10574 @item ignore_transform
10576 @item linear_design
10580 Default value is "default".
10582 For more information consult the documentation for the FT_LOAD_*
10586 The color to be used for drawing a shadow behind the drawn text. For the
10587 syntax of this option, check the @ref{color syntax,,"Color" section in the
10588 ffmpeg-utils manual,ffmpeg-utils}.
10590 The default value of @var{shadowcolor} is "black".
10594 The x and y offsets for the text shadow position with respect to the
10595 position of the text. They can be either positive or negative
10596 values. The default value for both is "0".
10599 The starting frame number for the n/frame_num variable. The default value
10603 The size in number of spaces to use for rendering the tab.
10604 Default value is 4.
10607 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10608 format. It can be used with or without text parameter. @var{timecode_rate}
10609 option must be specified.
10611 @item timecode_rate, rate, r
10612 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10613 integer. Minimum value is "1".
10614 Drop-frame timecode is supported for frame rates 30 & 60.
10617 If set to 1, the output of the timecode option will wrap around at 24 hours.
10618 Default is 0 (disabled).
10621 The text string to be drawn. The text must be a sequence of UTF-8
10622 encoded characters.
10623 This parameter is mandatory if no file is specified with the parameter
10627 A text file containing text to be drawn. The text must be a sequence
10628 of UTF-8 encoded characters.
10630 This parameter is mandatory if no text string is specified with the
10631 parameter @var{text}.
10633 If both @var{text} and @var{textfile} are specified, an error is thrown.
10636 If set to 1, the @var{textfile} will be reloaded before each frame.
10637 Be sure to update it atomically, or it may be read partially, or even fail.
10641 The expressions which specify the offsets where text will be drawn
10642 within the video frame. They are relative to the top/left border of the
10645 The default value of @var{x} and @var{y} is "0".
10647 See below for the list of accepted constants and functions.
10650 The parameters for @var{x} and @var{y} are expressions containing the
10651 following constants and functions:
10655 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10659 horizontal and vertical chroma subsample values. For example for the
10660 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10663 the height of each text line
10671 @item max_glyph_a, ascent
10672 the maximum distance from the baseline to the highest/upper grid
10673 coordinate used to place a glyph outline point, for all the rendered
10675 It is a positive value, due to the grid's orientation with the Y axis
10678 @item max_glyph_d, descent
10679 the maximum distance from the baseline to the lowest grid coordinate
10680 used to place a glyph outline point, for all the rendered glyphs.
10681 This is a negative value, due to the grid's orientation, with the Y axis
10685 maximum glyph height, that is the maximum height for all the glyphs
10686 contained in the rendered text, it is equivalent to @var{ascent} -
10690 maximum glyph width, that is the maximum width for all the glyphs
10691 contained in the rendered text
10694 the number of input frame, starting from 0
10696 @item rand(min, max)
10697 return a random number included between @var{min} and @var{max}
10700 The input sample aspect ratio.
10703 timestamp expressed in seconds, NAN if the input timestamp is unknown
10706 the height of the rendered text
10709 the width of the rendered text
10713 the x and y offset coordinates where the text is drawn.
10715 These parameters allow the @var{x} and @var{y} expressions to refer
10716 to each other, so you can for example specify @code{y=x/dar}.
10719 A one character description of the current frame's picture type.
10722 The current packet's position in the input file or stream
10723 (in bytes, from the start of the input). A value of -1 indicates
10724 this info is not available.
10727 The current packet's duration, in seconds.
10730 The current packet's size (in bytes).
10733 @anchor{drawtext_expansion}
10734 @subsection Text expansion
10736 If @option{expansion} is set to @code{strftime},
10737 the filter recognizes strftime() sequences in the provided text and
10738 expands them accordingly. Check the documentation of strftime(). This
10739 feature is deprecated.
10741 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10743 If @option{expansion} is set to @code{normal} (which is the default),
10744 the following expansion mechanism is used.
10746 The backslash character @samp{\}, followed by any character, always expands to
10747 the second character.
10749 Sequences of the form @code{%@{...@}} are expanded. The text between the
10750 braces is a function name, possibly followed by arguments separated by ':'.
10751 If the arguments contain special characters or delimiters (':' or '@}'),
10752 they should be escaped.
10754 Note that they probably must also be escaped as the value for the
10755 @option{text} option in the filter argument string and as the filter
10756 argument in the filtergraph description, and possibly also for the shell,
10757 that makes up to four levels of escaping; using a text file avoids these
10760 The following functions are available:
10765 The expression evaluation result.
10767 It must take one argument specifying the expression to be evaluated,
10768 which accepts the same constants and functions as the @var{x} and
10769 @var{y} values. Note that not all constants should be used, for
10770 example the text size is not known when evaluating the expression, so
10771 the constants @var{text_w} and @var{text_h} will have an undefined
10774 @item expr_int_format, eif
10775 Evaluate the expression's value and output as formatted integer.
10777 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10778 The second argument specifies the output format. Allowed values are @samp{x},
10779 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10780 @code{printf} function.
10781 The third parameter is optional and sets the number of positions taken by the output.
10782 It can be used to add padding with zeros from the left.
10785 The time at which the filter is running, expressed in UTC.
10786 It can accept an argument: a strftime() format string.
10789 The time at which the filter is running, expressed in the local time zone.
10790 It can accept an argument: a strftime() format string.
10793 Frame metadata. Takes one or two arguments.
10795 The first argument is mandatory and specifies the metadata key.
10797 The second argument is optional and specifies a default value, used when the
10798 metadata key is not found or empty.
10800 Available metadata can be identified by inspecting entries
10801 starting with TAG included within each frame section
10802 printed by running @code{ffprobe -show_frames}.
10804 String metadata generated in filters leading to
10805 the drawtext filter are also available.
10808 The frame number, starting from 0.
10811 A one character description of the current picture type.
10814 The timestamp of the current frame.
10815 It can take up to three arguments.
10817 The first argument is the format of the timestamp; it defaults to @code{flt}
10818 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10819 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10820 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10821 @code{localtime} stands for the timestamp of the frame formatted as
10822 local time zone time.
10824 The second argument is an offset added to the timestamp.
10826 If the format is set to @code{hms}, a third argument @code{24HH} may be
10827 supplied to present the hour part of the formatted timestamp in 24h format
10830 If the format is set to @code{localtime} or @code{gmtime},
10831 a third argument may be supplied: a strftime() format string.
10832 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10835 @subsection Commands
10837 This filter supports altering parameters via commands:
10840 Alter existing filter parameters.
10842 Syntax for the argument is the same as for filter invocation, e.g.
10845 fontsize=56:fontcolor=green:text='Hello World'
10848 Full filter invocation with sendcmd would look like this:
10851 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10855 If the entire argument can't be parsed or applied as valid values then the filter will
10856 continue with its existing parameters.
10858 @subsection Examples
10862 Draw "Test Text" with font FreeSerif, using the default values for the
10863 optional parameters.
10866 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10870 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10871 and y=50 (counting from the top-left corner of the screen), text is
10872 yellow with a red box around it. Both the text and the box have an
10876 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10877 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10880 Note that the double quotes are not necessary if spaces are not used
10881 within the parameter list.
10884 Show the text at the center of the video frame:
10886 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10890 Show the text at a random position, switching to a new position every 30 seconds:
10892 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)"
10896 Show a text line sliding from right to left in the last row of the video
10897 frame. The file @file{LONG_LINE} is assumed to contain a single line
10900 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10904 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10906 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10910 Draw a single green letter "g", at the center of the input video.
10911 The glyph baseline is placed at half screen height.
10913 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10917 Show text for 1 second every 3 seconds:
10919 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10923 Use fontconfig to set the font. Note that the colons need to be escaped.
10925 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10929 Draw "Test Text" with font size dependent on height of the video.
10931 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
10935 Print the date of a real-time encoding (see strftime(3)):
10937 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10941 Show text fading in and out (appearing/disappearing):
10944 DS=1.0 # display start
10945 DE=10.0 # display end
10946 FID=1.5 # fade in duration
10947 FOD=5 # fade out duration
10948 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 @}"
10952 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10953 and the @option{fontsize} value are included in the @option{y} offset.
10955 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10956 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10960 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10961 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10962 must have option @option{-export_path_metadata 1} for the special metadata fields
10963 to be available for filters.
10965 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10970 For more information about libfreetype, check:
10971 @url{http://www.freetype.org/}.
10973 For more information about fontconfig, check:
10974 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10976 For more information about libfribidi, check:
10977 @url{http://fribidi.org/}.
10979 @section edgedetect
10981 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10983 The filter accepts the following options:
10988 Set low and high threshold values used by the Canny thresholding
10991 The high threshold selects the "strong" edge pixels, which are then
10992 connected through 8-connectivity with the "weak" edge pixels selected
10993 by the low threshold.
10995 @var{low} and @var{high} threshold values must be chosen in the range
10996 [0,1], and @var{low} should be lesser or equal to @var{high}.
10998 Default value for @var{low} is @code{20/255}, and default value for @var{high}
11002 Define the drawing mode.
11006 Draw white/gray wires on black background.
11009 Mix the colors to create a paint/cartoon effect.
11012 Apply Canny edge detector on all selected planes.
11014 Default value is @var{wires}.
11017 Select planes for filtering. By default all available planes are filtered.
11020 @subsection Examples
11024 Standard edge detection with custom values for the hysteresis thresholding:
11026 edgedetect=low=0.1:high=0.4
11030 Painting effect without thresholding:
11032 edgedetect=mode=colormix:high=0
11038 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
11040 For each input image, the filter will compute the optimal mapping from
11041 the input to the output given the codebook length, that is the number
11042 of distinct output colors.
11044 This filter accepts the following options.
11047 @item codebook_length, l
11048 Set codebook length. The value must be a positive integer, and
11049 represents the number of distinct output colors. Default value is 256.
11052 Set the maximum number of iterations to apply for computing the optimal
11053 mapping. The higher the value the better the result and the higher the
11054 computation time. Default value is 1.
11057 Set a random seed, must be an integer included between 0 and
11058 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
11059 will try to use a good random seed on a best effort basis.
11062 Set pal8 output pixel format. This option does not work with codebook
11063 length greater than 256.
11068 Measure graylevel entropy in histogram of color channels of video frames.
11070 It accepts the following parameters:
11074 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
11076 @var{diff} mode measures entropy of histogram delta values, absolute differences
11077 between neighbour histogram values.
11081 Apply the EPX magnification filter which is designed for pixel art.
11083 It accepts the following option:
11087 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11089 Default is @code{3}.
11093 Set brightness, contrast, saturation and approximate gamma adjustment.
11095 The filter accepts the following options:
11099 Set the contrast expression. The value must be a float value in range
11100 @code{-1000.0} to @code{1000.0}. The default value is "1".
11103 Set the brightness expression. The value must be a float value in
11104 range @code{-1.0} to @code{1.0}. The default value is "0".
11107 Set the saturation expression. The value must be a float in
11108 range @code{0.0} to @code{3.0}. The default value is "1".
11111 Set the gamma expression. The value must be a float in range
11112 @code{0.1} to @code{10.0}. The default value is "1".
11115 Set the gamma expression for red. The value must be a float in
11116 range @code{0.1} to @code{10.0}. The default value is "1".
11119 Set the gamma expression for green. The value must be a float in range
11120 @code{0.1} to @code{10.0}. The default value is "1".
11123 Set the gamma expression for blue. The value must be a float in range
11124 @code{0.1} to @code{10.0}. The default value is "1".
11127 Set the gamma weight expression. It can be used to reduce the effect
11128 of a high gamma value on bright image areas, e.g. keep them from
11129 getting overamplified and just plain white. The value must be a float
11130 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11131 gamma correction all the way down while @code{1.0} leaves it at its
11132 full strength. Default is "1".
11135 Set when the expressions for brightness, contrast, saturation and
11136 gamma expressions are evaluated.
11138 It accepts the following values:
11141 only evaluate expressions once during the filter initialization or
11142 when a command is processed
11145 evaluate expressions for each incoming frame
11148 Default value is @samp{init}.
11151 The expressions accept the following parameters:
11154 frame count of the input frame starting from 0
11157 byte position of the corresponding packet in the input file, NAN if
11161 frame rate of the input video, NAN if the input frame rate is unknown
11164 timestamp expressed in seconds, NAN if the input timestamp is unknown
11167 @subsection Commands
11168 The filter supports the following commands:
11172 Set the contrast expression.
11175 Set the brightness expression.
11178 Set the saturation expression.
11181 Set the gamma expression.
11184 Set the gamma_r expression.
11187 Set gamma_g expression.
11190 Set gamma_b expression.
11193 Set gamma_weight expression.
11195 The command accepts the same syntax of the corresponding option.
11197 If the specified expression is not valid, it is kept at its current
11204 Apply erosion effect to the video.
11206 This filter replaces the pixel by the local(3x3) minimum.
11208 It accepts the following options:
11215 Limit the maximum change for each plane, default is 65535.
11216 If 0, plane will remain unchanged.
11219 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11222 Flags to local 3x3 coordinates maps like this:
11229 @subsection Commands
11231 This filter supports the all above options as @ref{commands}.
11235 Deinterlace the input video ("estdif" stands for "Edge Slope
11236 Tracing Deinterlacing Filter").
11238 Spatial only filter that uses edge slope tracing algorithm
11239 to interpolate missing lines.
11240 It accepts the following parameters:
11244 The interlacing mode to adopt. It accepts one of the following values:
11248 Output one frame for each frame.
11250 Output one frame for each field.
11253 The default value is @code{field}.
11256 The picture field parity assumed for the input interlaced video. It accepts one
11257 of the following values:
11261 Assume the top field is first.
11263 Assume the bottom field is first.
11265 Enable automatic detection of field parity.
11268 The default value is @code{auto}.
11269 If the interlacing is unknown or the decoder does not export this information,
11270 top field first will be assumed.
11273 Specify which frames to deinterlace. Accepts one of the following
11278 Deinterlace all frames.
11280 Only deinterlace frames marked as interlaced.
11283 The default value is @code{all}.
11286 Specify the search radius for edge slope tracing. Default value is 1.
11287 Allowed range is from 1 to 15.
11290 Specify the search radius for best edge matching. Default value is 2.
11291 Allowed range is from 0 to 15.
11294 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11295 of the following values:
11299 Two-point interpolation.
11301 Four-point interpolation.
11303 Six-point interpolation.
11307 @subsection Commands
11308 This filter supports same @ref{commands} as options.
11310 @section extractplanes
11312 Extract color channel components from input video stream into
11313 separate grayscale video streams.
11315 The filter accepts the following option:
11319 Set plane(s) to extract.
11321 Available values for planes are:
11332 Choosing planes not available in the input will result in an error.
11333 That means you cannot select @code{r}, @code{g}, @code{b} planes
11334 with @code{y}, @code{u}, @code{v} planes at same time.
11337 @subsection Examples
11341 Extract luma, u and v color channel component from input video frame
11342 into 3 grayscale outputs:
11344 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
11350 Apply a fade-in/out effect to the input video.
11352 It accepts the following parameters:
11356 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11358 Default is @code{in}.
11360 @item start_frame, s
11361 Specify the number of the frame to start applying the fade
11362 effect at. Default is 0.
11365 The number of frames that the fade effect lasts. At the end of the
11366 fade-in effect, the output video will have the same intensity as the input video.
11367 At the end of the fade-out transition, the output video will be filled with the
11368 selected @option{color}.
11372 If set to 1, fade only alpha channel, if one exists on the input.
11373 Default value is 0.
11375 @item start_time, st
11376 Specify the timestamp (in seconds) of the frame to start to apply the fade
11377 effect. If both start_frame and start_time are specified, the fade will start at
11378 whichever comes last. Default is 0.
11381 The number of seconds for which the fade effect has to last. At the end of the
11382 fade-in effect the output video will have the same intensity as the input video,
11383 at the end of the fade-out transition the output video will be filled with the
11384 selected @option{color}.
11385 If both duration and nb_frames are specified, duration is used. Default is 0
11386 (nb_frames is used by default).
11389 Specify the color of the fade. Default is "black".
11392 @subsection Examples
11396 Fade in the first 30 frames of video:
11401 The command above is equivalent to:
11407 Fade out the last 45 frames of a 200-frame video:
11410 fade=type=out:start_frame=155:nb_frames=45
11414 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11416 fade=in:0:25, fade=out:975:25
11420 Make the first 5 frames yellow, then fade in from frame 5-24:
11422 fade=in:5:20:color=yellow
11426 Fade in alpha over first 25 frames of video:
11428 fade=in:0:25:alpha=1
11432 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11434 fade=t=in:st=5.5:d=0.5
11440 Denoise frames using 3D FFT (frequency domain filtering).
11442 The filter accepts the following options:
11446 Set the noise sigma constant. This sets denoising strength.
11447 Default value is 1. Allowed range is from 0 to 30.
11448 Using very high sigma with low overlap may give blocking artifacts.
11451 Set amount of denoising. By default all detected noise is reduced.
11452 Default value is 1. Allowed range is from 0 to 1.
11455 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11456 Actual size of block in pixels is 2 to power of @var{block}, so by default
11457 block size in pixels is 2^4 which is 16.
11460 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11463 Set number of previous frames to use for denoising. By default is set to 0.
11466 Set number of next frames to to use for denoising. By default is set to 0.
11469 Set planes which will be filtered, by default are all available filtered
11474 Apply arbitrary expressions to samples in frequency domain
11478 Adjust the dc value (gain) of the luma plane of the image. The filter
11479 accepts an integer value in range @code{0} to @code{1000}. The default
11480 value is set to @code{0}.
11483 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11484 filter accepts an integer value in range @code{0} to @code{1000}. The
11485 default value is set to @code{0}.
11488 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11489 filter accepts an integer value in range @code{0} to @code{1000}. The
11490 default value is set to @code{0}.
11493 Set the frequency domain weight expression for the luma plane.
11496 Set the frequency domain weight expression for the 1st chroma plane.
11499 Set the frequency domain weight expression for the 2nd chroma plane.
11502 Set when the expressions are evaluated.
11504 It accepts the following values:
11507 Only evaluate expressions once during the filter initialization.
11510 Evaluate expressions for each incoming frame.
11513 Default value is @samp{init}.
11515 The filter accepts the following variables:
11518 The coordinates of the current sample.
11522 The width and height of the image.
11525 The number of input frame, starting from 0.
11528 @subsection Examples
11534 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11540 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11546 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11552 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11559 Extract a single field from an interlaced image using stride
11560 arithmetic to avoid wasting CPU time. The output frames are marked as
11563 The filter accepts the following options:
11567 Specify whether to extract the top (if the value is @code{0} or
11568 @code{top}) or the bottom field (if the value is @code{1} or
11574 Create new frames by copying the top and bottom fields from surrounding frames
11575 supplied as numbers by the hint file.
11579 Set file containing hints: absolute/relative frame numbers.
11581 There must be one line for each frame in a clip. Each line must contain two
11582 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11583 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11584 is current frame number for @code{absolute} mode or out of [-1, 1] range
11585 for @code{relative} mode. First number tells from which frame to pick up top
11586 field and second number tells from which frame to pick up bottom field.
11588 If optionally followed by @code{+} output frame will be marked as interlaced,
11589 else if followed by @code{-} output frame will be marked as progressive, else
11590 it will be marked same as input frame.
11591 If optionally followed by @code{t} output frame will use only top field, or in
11592 case of @code{b} it will use only bottom field.
11593 If line starts with @code{#} or @code{;} that line is skipped.
11596 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11599 Example of first several lines of @code{hint} file for @code{relative} mode:
11601 0,0 - # first frame
11602 1,0 - # second frame, use third's frame top field and second's frame bottom field
11603 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11618 @section fieldmatch
11620 Field matching filter for inverse telecine. It is meant to reconstruct the
11621 progressive frames from a telecined stream. The filter does not drop duplicated
11622 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11623 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11625 The separation of the field matching and the decimation is notably motivated by
11626 the possibility of inserting a de-interlacing filter fallback between the two.
11627 If the source has mixed telecined and real interlaced content,
11628 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11629 But these remaining combed frames will be marked as interlaced, and thus can be
11630 de-interlaced by a later filter such as @ref{yadif} before decimation.
11632 In addition to the various configuration options, @code{fieldmatch} can take an
11633 optional second stream, activated through the @option{ppsrc} option. If
11634 enabled, the frames reconstruction will be based on the fields and frames from
11635 this second stream. This allows the first input to be pre-processed in order to
11636 help the various algorithms of the filter, while keeping the output lossless
11637 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11638 or brightness/contrast adjustments can help.
11640 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11641 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11642 which @code{fieldmatch} is based on. While the semantic and usage are very
11643 close, some behaviour and options names can differ.
11645 The @ref{decimate} filter currently only works for constant frame rate input.
11646 If your input has mixed telecined (30fps) and progressive content with a lower
11647 framerate like 24fps use the following filterchain to produce the necessary cfr
11648 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11650 The filter accepts the following options:
11654 Specify the assumed field order of the input stream. Available values are:
11658 Auto detect parity (use FFmpeg's internal parity value).
11660 Assume bottom field first.
11662 Assume top field first.
11665 Note that it is sometimes recommended not to trust the parity announced by the
11668 Default value is @var{auto}.
11671 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11672 sense that it won't risk creating jerkiness due to duplicate frames when
11673 possible, but if there are bad edits or blended fields it will end up
11674 outputting combed frames when a good match might actually exist. On the other
11675 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11676 but will almost always find a good frame if there is one. The other values are
11677 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11678 jerkiness and creating duplicate frames versus finding good matches in sections
11679 with bad edits, orphaned fields, blended fields, etc.
11681 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11683 Available values are:
11687 2-way matching (p/c)
11689 2-way matching, and trying 3rd match if still combed (p/c + n)
11691 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11693 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11694 still combed (p/c + n + u/b)
11696 3-way matching (p/c/n)
11698 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11699 detected as combed (p/c/n + u/b)
11702 The parenthesis at the end indicate the matches that would be used for that
11703 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11706 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11709 Default value is @var{pc_n}.
11712 Mark the main input stream as a pre-processed input, and enable the secondary
11713 input stream as the clean source to pick the fields from. See the filter
11714 introduction for more details. It is similar to the @option{clip2} feature from
11717 Default value is @code{0} (disabled).
11720 Set the field to match from. It is recommended to set this to the same value as
11721 @option{order} unless you experience matching failures with that setting. In
11722 certain circumstances changing the field that is used to match from can have a
11723 large impact on matching performance. Available values are:
11727 Automatic (same value as @option{order}).
11729 Match from the bottom field.
11731 Match from the top field.
11734 Default value is @var{auto}.
11737 Set whether or not chroma is included during the match comparisons. In most
11738 cases it is recommended to leave this enabled. You should set this to @code{0}
11739 only if your clip has bad chroma problems such as heavy rainbowing or other
11740 artifacts. Setting this to @code{0} could also be used to speed things up at
11741 the cost of some accuracy.
11743 Default value is @code{1}.
11747 These define an exclusion band which excludes the lines between @option{y0} and
11748 @option{y1} from being included in the field matching decision. An exclusion
11749 band can be used to ignore subtitles, a logo, or other things that may
11750 interfere with the matching. @option{y0} sets the starting scan line and
11751 @option{y1} sets the ending line; all lines in between @option{y0} and
11752 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11753 @option{y0} and @option{y1} to the same value will disable the feature.
11754 @option{y0} and @option{y1} defaults to @code{0}.
11757 Set the scene change detection threshold as a percentage of maximum change on
11758 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11759 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11760 @option{scthresh} is @code{[0.0, 100.0]}.
11762 Default value is @code{12.0}.
11765 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11766 account the combed scores of matches when deciding what match to use as the
11767 final match. Available values are:
11771 No final matching based on combed scores.
11773 Combed scores are only used when a scene change is detected.
11775 Use combed scores all the time.
11778 Default is @var{sc}.
11781 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11782 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11783 Available values are:
11787 No forced calculation.
11789 Force p/c/n calculations.
11791 Force p/c/n/u/b calculations.
11794 Default value is @var{none}.
11797 This is the area combing threshold used for combed frame detection. This
11798 essentially controls how "strong" or "visible" combing must be to be detected.
11799 Larger values mean combing must be more visible and smaller values mean combing
11800 can be less visible or strong and still be detected. Valid settings are from
11801 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11802 be detected as combed). This is basically a pixel difference value. A good
11803 range is @code{[8, 12]}.
11805 Default value is @code{9}.
11808 Sets whether or not chroma is considered in the combed frame decision. Only
11809 disable this if your source has chroma problems (rainbowing, etc.) that are
11810 causing problems for the combed frame detection with chroma enabled. Actually,
11811 using @option{chroma}=@var{0} is usually more reliable, except for the case
11812 where there is chroma only combing in the source.
11814 Default value is @code{0}.
11818 Respectively set the x-axis and y-axis size of the window used during combed
11819 frame detection. This has to do with the size of the area in which
11820 @option{combpel} pixels are required to be detected as combed for a frame to be
11821 declared combed. See the @option{combpel} parameter description for more info.
11822 Possible values are any number that is a power of 2 starting at 4 and going up
11825 Default value is @code{16}.
11828 The number of combed pixels inside any of the @option{blocky} by
11829 @option{blockx} size blocks on the frame for the frame to be detected as
11830 combed. While @option{cthresh} controls how "visible" the combing must be, this
11831 setting controls "how much" combing there must be in any localized area (a
11832 window defined by the @option{blockx} and @option{blocky} settings) on the
11833 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11834 which point no frames will ever be detected as combed). This setting is known
11835 as @option{MI} in TFM/VFM vocabulary.
11837 Default value is @code{80}.
11840 @anchor{p/c/n/u/b meaning}
11841 @subsection p/c/n/u/b meaning
11843 @subsubsection p/c/n
11845 We assume the following telecined stream:
11848 Top fields: 1 2 2 3 4
11849 Bottom fields: 1 2 3 4 4
11852 The numbers correspond to the progressive frame the fields relate to. Here, the
11853 first two frames are progressive, the 3rd and 4th are combed, and so on.
11855 When @code{fieldmatch} is configured to run a matching from bottom
11856 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11861 B 1 2 3 4 4 <-- matching reference
11870 As a result of the field matching, we can see that some frames get duplicated.
11871 To perform a complete inverse telecine, you need to rely on a decimation filter
11872 after this operation. See for instance the @ref{decimate} filter.
11874 The same operation now matching from top fields (@option{field}=@var{top})
11879 T 1 2 2 3 4 <-- matching reference
11889 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11890 basically, they refer to the frame and field of the opposite parity:
11893 @item @var{p} matches the field of the opposite parity in the previous frame
11894 @item @var{c} matches the field of the opposite parity in the current frame
11895 @item @var{n} matches the field of the opposite parity in the next frame
11900 The @var{u} and @var{b} matching are a bit special in the sense that they match
11901 from the opposite parity flag. In the following examples, we assume that we are
11902 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11903 'x' is placed above and below each matched fields.
11905 With bottom matching (@option{field}=@var{bottom}):
11910 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11911 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11919 With top matching (@option{field}=@var{top}):
11924 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11925 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11933 @subsection Examples
11935 Simple IVTC of a top field first telecined stream:
11937 fieldmatch=order=tff:combmatch=none, decimate
11940 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11942 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11945 @section fieldorder
11947 Transform the field order of the input video.
11949 It accepts the following parameters:
11954 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11955 for bottom field first.
11958 The default value is @samp{tff}.
11960 The transformation is done by shifting the picture content up or down
11961 by one line, and filling the remaining line with appropriate picture content.
11962 This method is consistent with most broadcast field order converters.
11964 If the input video is not flagged as being interlaced, or it is already
11965 flagged as being of the required output field order, then this filter does
11966 not alter the incoming video.
11968 It is very useful when converting to or from PAL DV material,
11969 which is bottom field first.
11973 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11976 @section fifo, afifo
11978 Buffer input images and send them when they are requested.
11980 It is mainly useful when auto-inserted by the libavfilter
11983 It does not take parameters.
11985 @section fillborders
11987 Fill borders of the input video, without changing video stream dimensions.
11988 Sometimes video can have garbage at the four edges and you may not want to
11989 crop video input to keep size multiple of some number.
11991 This filter accepts the following options:
11995 Number of pixels to fill from left border.
11998 Number of pixels to fill from right border.
12001 Number of pixels to fill from top border.
12004 Number of pixels to fill from bottom border.
12009 It accepts the following values:
12012 fill pixels using outermost pixels
12015 fill pixels using mirroring (half sample symmetric)
12018 fill pixels with constant value
12021 fill pixels using reflecting (whole sample symmetric)
12024 fill pixels using wrapping
12027 fade pixels to constant value
12030 Default is @var{smear}.
12033 Set color for pixels in fixed or fade mode. Default is @var{black}.
12036 @subsection Commands
12037 This filter supports same @ref{commands} as options.
12038 The command accepts the same syntax of the corresponding option.
12040 If the specified expression is not valid, it is kept at its current
12045 Find a rectangular object
12047 It accepts the following options:
12051 Filepath of the object image, needs to be in gray8.
12054 Detection threshold, default is 0.5.
12057 Number of mipmaps, default is 3.
12059 @item xmin, ymin, xmax, ymax
12060 Specifies the rectangle in which to search.
12063 @subsection Examples
12067 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
12069 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
12075 Flood area with values of same pixel components with another values.
12077 It accepts the following options:
12080 Set pixel x coordinate.
12083 Set pixel y coordinate.
12086 Set source #0 component value.
12089 Set source #1 component value.
12092 Set source #2 component value.
12095 Set source #3 component value.
12098 Set destination #0 component value.
12101 Set destination #1 component value.
12104 Set destination #2 component value.
12107 Set destination #3 component value.
12113 Convert the input video to one of the specified pixel formats.
12114 Libavfilter will try to pick one that is suitable as input to
12117 It accepts the following parameters:
12121 A '|'-separated list of pixel format names, such as
12122 "pix_fmts=yuv420p|monow|rgb24".
12126 @subsection Examples
12130 Convert the input video to the @var{yuv420p} format
12132 format=pix_fmts=yuv420p
12135 Convert the input video to any of the formats in the list
12137 format=pix_fmts=yuv420p|yuv444p|yuv410p
12144 Convert the video to specified constant frame rate by duplicating or dropping
12145 frames as necessary.
12147 It accepts the following parameters:
12151 The desired output frame rate. The default is @code{25}.
12154 Assume the first PTS should be the given value, in seconds. This allows for
12155 padding/trimming at the start of stream. By default, no assumption is made
12156 about the first frame's expected PTS, so no padding or trimming is done.
12157 For example, this could be set to 0 to pad the beginning with duplicates of
12158 the first frame if a video stream starts after the audio stream or to trim any
12159 frames with a negative PTS.
12162 Timestamp (PTS) rounding method.
12164 Possible values are:
12171 round towards -infinity
12173 round towards +infinity
12177 The default is @code{near}.
12180 Action performed when reading the last frame.
12182 Possible values are:
12185 Use same timestamp rounding method as used for other frames.
12187 Pass through last frame if input duration has not been reached yet.
12189 The default is @code{round}.
12193 Alternatively, the options can be specified as a flat string:
12194 @var{fps}[:@var{start_time}[:@var{round}]].
12196 See also the @ref{setpts} filter.
12198 @subsection Examples
12202 A typical usage in order to set the fps to 25:
12208 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12210 fps=fps=film:round=near
12216 Pack two different video streams into a stereoscopic video, setting proper
12217 metadata on supported codecs. The two views should have the same size and
12218 framerate and processing will stop when the shorter video ends. Please note
12219 that you may conveniently adjust view properties with the @ref{scale} and
12222 It accepts the following parameters:
12226 The desired packing format. Supported values are:
12231 The views are next to each other (default).
12234 The views are on top of each other.
12237 The views are packed by line.
12240 The views are packed by column.
12243 The views are temporally interleaved.
12252 # Convert left and right views into a frame-sequential video
12253 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12255 # Convert views into a side-by-side video with the same output resolution as the input
12256 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
12261 Change the frame rate by interpolating new video output frames from the source
12264 This filter is not designed to function correctly with interlaced media. If
12265 you wish to change the frame rate of interlaced media then you are required
12266 to deinterlace before this filter and re-interlace after this filter.
12268 A description of the accepted options follows.
12272 Specify the output frames per second. This option can also be specified
12273 as a value alone. The default is @code{50}.
12276 Specify the start of a range where the output frame will be created as a
12277 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12278 the default is @code{15}.
12281 Specify the end of a range where the output frame will be created as a
12282 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12283 the default is @code{240}.
12286 Specify the level at which a scene change is detected as a value between
12287 0 and 100 to indicate a new scene; a low value reflects a low
12288 probability for the current frame to introduce a new scene, while a higher
12289 value means the current frame is more likely to be one.
12290 The default is @code{8.2}.
12293 Specify flags influencing the filter process.
12295 Available value for @var{flags} is:
12298 @item scene_change_detect, scd
12299 Enable scene change detection using the value of the option @var{scene}.
12300 This flag is enabled by default.
12306 Select one frame every N-th frame.
12308 This filter accepts the following option:
12311 Select frame after every @code{step} frames.
12312 Allowed values are positive integers higher than 0. Default value is @code{1}.
12315 @section freezedetect
12317 Detect frozen video.
12319 This filter logs a message and sets frame metadata when it detects that the
12320 input video has no significant change in content during a specified duration.
12321 Video freeze detection calculates the mean average absolute difference of all
12322 the components of video frames and compares it to a noise floor.
12324 The printed times and duration are expressed in seconds. The
12325 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12326 whose timestamp equals or exceeds the detection duration and it contains the
12327 timestamp of the first frame of the freeze. The
12328 @code{lavfi.freezedetect.freeze_duration} and
12329 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12332 The filter accepts the following options:
12336 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12337 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12341 Set freeze duration until notification (default is 2 seconds).
12344 @section freezeframes
12346 Freeze video frames.
12348 This filter freezes video frames using frame from 2nd input.
12350 The filter accepts the following options:
12354 Set number of first frame from which to start freeze.
12357 Set number of last frame from which to end freeze.
12360 Set number of frame from 2nd input which will be used instead of replaced frames.
12366 Apply a frei0r effect to the input video.
12368 To enable the compilation of this filter, you need to install the frei0r
12369 header and configure FFmpeg with @code{--enable-frei0r}.
12371 It accepts the following parameters:
12376 The name of the frei0r effect to load. If the environment variable
12377 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12378 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12379 Otherwise, the standard frei0r paths are searched, in this order:
12380 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12381 @file{/usr/lib/frei0r-1/}.
12383 @item filter_params
12384 A '|'-separated list of parameters to pass to the frei0r effect.
12388 A frei0r effect parameter can be a boolean (its value is either
12389 "y" or "n"), a double, a color (specified as
12390 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12391 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12392 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12393 a position (specified as @var{X}/@var{Y}, where
12394 @var{X} and @var{Y} are floating point numbers) and/or a string.
12396 The number and types of parameters depend on the loaded effect. If an
12397 effect parameter is not specified, the default value is set.
12399 @subsection Examples
12403 Apply the distort0r effect, setting the first two double parameters:
12405 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12409 Apply the colordistance effect, taking a color as the first parameter:
12411 frei0r=colordistance:0.2/0.3/0.4
12412 frei0r=colordistance:violet
12413 frei0r=colordistance:0x112233
12417 Apply the perspective effect, specifying the top left and top right image
12420 frei0r=perspective:0.2/0.2|0.8/0.2
12424 For more information, see
12425 @url{http://frei0r.dyne.org}
12427 @subsection Commands
12429 This filter supports the @option{filter_params} option as @ref{commands}.
12433 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12435 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12436 processing filter, one of them is performed once per block, not per pixel.
12437 This allows for much higher speed.
12439 The filter accepts the following options:
12443 Set quality. This option defines the number of levels for averaging. It accepts
12444 an integer in the range 4-5. Default value is @code{4}.
12447 Force a constant quantization parameter. It accepts an integer in range 0-63.
12448 If not set, the filter will use the QP from the video stream (if available).
12451 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12452 more details but also more artifacts, while higher values make the image smoother
12453 but also blurrier. Default value is @code{0} − PSNR optimal.
12455 @item use_bframe_qp
12456 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12457 option may cause flicker since the B-Frames have often larger QP. Default is
12458 @code{0} (not enabled).
12464 Apply Gaussian blur filter.
12466 The filter accepts the following options:
12470 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12473 Set number of steps for Gaussian approximation. Default is @code{1}.
12476 Set which planes to filter. By default all planes are filtered.
12479 Set vertical sigma, if negative it will be same as @code{sigma}.
12480 Default is @code{-1}.
12483 @subsection Commands
12484 This filter supports same commands as options.
12485 The command accepts the same syntax of the corresponding option.
12487 If the specified expression is not valid, it is kept at its current
12492 Apply generic equation to each pixel.
12494 The filter accepts the following options:
12497 @item lum_expr, lum
12498 Set the luminance expression.
12500 Set the chrominance blue expression.
12502 Set the chrominance red expression.
12503 @item alpha_expr, a
12504 Set the alpha expression.
12506 Set the red expression.
12507 @item green_expr, g
12508 Set the green expression.
12510 Set the blue expression.
12513 The colorspace is selected according to the specified options. If one
12514 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12515 options is specified, the filter will automatically select a YCbCr
12516 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12517 @option{blue_expr} options is specified, it will select an RGB
12520 If one of the chrominance expression is not defined, it falls back on the other
12521 one. If no alpha expression is specified it will evaluate to opaque value.
12522 If none of chrominance expressions are specified, they will evaluate
12523 to the luminance expression.
12525 The expressions can use the following variables and functions:
12529 The sequential number of the filtered frame, starting from @code{0}.
12533 The coordinates of the current sample.
12537 The width and height of the image.
12541 Width and height scale depending on the currently filtered plane. It is the
12542 ratio between the corresponding luma plane number of pixels and the current
12543 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12544 @code{0.5,0.5} for chroma planes.
12547 Time of the current frame, expressed in seconds.
12550 Return the value of the pixel at location (@var{x},@var{y}) of the current
12554 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12558 Return the value of the pixel at location (@var{x},@var{y}) of the
12559 blue-difference chroma plane. Return 0 if there is no such plane.
12562 Return the value of the pixel at location (@var{x},@var{y}) of the
12563 red-difference chroma plane. Return 0 if there is no such plane.
12568 Return the value of the pixel at location (@var{x},@var{y}) of the
12569 red/green/blue component. Return 0 if there is no such component.
12572 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12573 plane. Return 0 if there is no such plane.
12575 @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)
12576 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12577 sums of samples within a rectangle. See the functions without the sum postfix.
12579 @item interpolation
12580 Set one of interpolation methods:
12585 Default is bilinear.
12588 For functions, if @var{x} and @var{y} are outside the area, the value will be
12589 automatically clipped to the closer edge.
12591 Please note that this filter can use multiple threads in which case each slice
12592 will have its own expression state. If you want to use only a single expression
12593 state because your expressions depend on previous state then you should limit
12594 the number of filter threads to 1.
12596 @subsection Examples
12600 Flip the image horizontally:
12606 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12607 wavelength of 100 pixels:
12609 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12613 Generate a fancy enigmatic moving light:
12615 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
12619 Generate a quick emboss effect:
12621 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12625 Modify RGB components depending on pixel position:
12627 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12631 Create a radial gradient that is the same size as the input (also see
12632 the @ref{vignette} filter):
12634 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12640 Fix the banding artifacts that are sometimes introduced into nearly flat
12641 regions by truncation to 8-bit color depth.
12642 Interpolate the gradients that should go where the bands are, and
12645 It is designed for playback only. Do not use it prior to
12646 lossy compression, because compression tends to lose the dither and
12647 bring back the bands.
12649 It accepts the following parameters:
12654 The maximum amount by which the filter will change any one pixel. This is also
12655 the threshold for detecting nearly flat regions. Acceptable values range from
12656 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12660 The neighborhood to fit the gradient to. A larger radius makes for smoother
12661 gradients, but also prevents the filter from modifying the pixels near detailed
12662 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12663 values will be clipped to the valid range.
12667 Alternatively, the options can be specified as a flat string:
12668 @var{strength}[:@var{radius}]
12670 @subsection Examples
12674 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12680 Specify radius, omitting the strength (which will fall-back to the default
12688 @anchor{graphmonitor}
12689 @section graphmonitor
12690 Show various filtergraph stats.
12692 With this filter one can debug complete filtergraph.
12693 Especially issues with links filling with queued frames.
12695 The filter accepts the following options:
12699 Set video output size. Default is @var{hd720}.
12702 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12705 Set output mode, can be @var{fulll} or @var{compact}.
12706 In @var{compact} mode only filters with some queued frames have displayed stats.
12709 Set flags which enable which stats are shown in video.
12711 Available values for flags are:
12714 Display number of queued frames in each link.
12716 @item frame_count_in
12717 Display number of frames taken from filter.
12719 @item frame_count_out
12720 Display number of frames given out from filter.
12723 Display current filtered frame pts.
12726 Display current filtered frame time.
12729 Display time base for filter link.
12732 Display used format for filter link.
12735 Display video size or number of audio channels in case of audio used by filter link.
12738 Display video frame rate or sample rate in case of audio used by filter link.
12741 Display link output status.
12745 Set upper limit for video rate of output stream, Default value is @var{25}.
12746 This guarantee that output video frame rate will not be higher than this value.
12750 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12751 and corrects the scene colors accordingly.
12753 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12755 The filter accepts the following options:
12759 The order of differentiation to be applied on the scene. Must be chosen in the range
12760 [0,2] and default value is 1.
12763 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12764 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12765 max value instead of calculating Minkowski distance.
12768 The standard deviation of Gaussian blur to be applied on the scene. Must be
12769 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12770 can't be equal to 0 if @var{difford} is greater than 0.
12773 @subsection Examples
12779 greyedge=difford=1:minknorm=5:sigma=2
12785 greyedge=difford=1:minknorm=0:sigma=2
12793 Apply a Hald CLUT to a video stream.
12795 First input is the video stream to process, and second one is the Hald CLUT.
12796 The Hald CLUT input can be a simple picture or a complete video stream.
12798 The filter accepts the following options:
12802 Force termination when the shortest input terminates. Default is @code{0}.
12804 Continue applying the last CLUT after the end of the stream. A value of
12805 @code{0} disable the filter after the last frame of the CLUT is reached.
12806 Default is @code{1}.
12809 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12810 filters share the same internals).
12812 This filter also supports the @ref{framesync} options.
12814 More information about the Hald CLUT can be found on Eskil Steenberg's website
12815 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12817 @subsection Workflow examples
12819 @subsubsection Hald CLUT video stream
12821 Generate an identity Hald CLUT stream altered with various effects:
12823 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
12826 Note: make sure you use a lossless codec.
12828 Then use it with @code{haldclut} to apply it on some random stream:
12830 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12833 The Hald CLUT will be applied to the 10 first seconds (duration of
12834 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12835 to the remaining frames of the @code{mandelbrot} stream.
12837 @subsubsection Hald CLUT with preview
12839 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12840 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12841 biggest possible square starting at the top left of the picture. The remaining
12842 padding pixels (bottom or right) will be ignored. This area can be used to add
12843 a preview of the Hald CLUT.
12845 Typically, the following generated Hald CLUT will be supported by the
12846 @code{haldclut} filter:
12849 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12850 pad=iw+320 [padded_clut];
12851 smptebars=s=320x256, split [a][b];
12852 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12853 [main][b] overlay=W-320" -frames:v 1 clut.png
12856 It contains the original and a preview of the effect of the CLUT: SMPTE color
12857 bars are displayed on the right-top, and below the same color bars processed by
12860 Then, the effect of this Hald CLUT can be visualized with:
12862 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12867 Flip the input video horizontally.
12869 For example, to horizontally flip the input video with @command{ffmpeg}:
12871 ffmpeg -i in.avi -vf "hflip" out.avi
12875 This filter applies a global color histogram equalization on a
12878 It can be used to correct video that has a compressed range of pixel
12879 intensities. The filter redistributes the pixel intensities to
12880 equalize their distribution across the intensity range. It may be
12881 viewed as an "automatically adjusting contrast filter". This filter is
12882 useful only for correcting degraded or poorly captured source
12885 The filter accepts the following options:
12889 Determine the amount of equalization to be applied. As the strength
12890 is reduced, the distribution of pixel intensities more-and-more
12891 approaches that of the input frame. The value must be a float number
12892 in the range [0,1] and defaults to 0.200.
12895 Set the maximum intensity that can generated and scale the output
12896 values appropriately. The strength should be set as desired and then
12897 the intensity can be limited if needed to avoid washing-out. The value
12898 must be a float number in the range [0,1] and defaults to 0.210.
12901 Set the antibanding level. If enabled the filter will randomly vary
12902 the luminance of output pixels by a small amount to avoid banding of
12903 the histogram. Possible values are @code{none}, @code{weak} or
12904 @code{strong}. It defaults to @code{none}.
12910 Compute and draw a color distribution histogram for the input video.
12912 The computed histogram is a representation of the color component
12913 distribution in an image.
12915 Standard histogram displays the color components distribution in an image.
12916 Displays color graph for each color component. Shows distribution of
12917 the Y, U, V, A or R, G, B components, depending on input format, in the
12918 current frame. Below each graph a color component scale meter is shown.
12920 The filter accepts the following options:
12924 Set height of level. Default value is @code{200}.
12925 Allowed range is [50, 2048].
12928 Set height of color scale. Default value is @code{12}.
12929 Allowed range is [0, 40].
12933 It accepts the following values:
12936 Per color component graphs are placed below each other.
12939 Per color component graphs are placed side by side.
12942 Presents information identical to that in the @code{parade}, except
12943 that the graphs representing color components are superimposed directly
12946 Default is @code{stack}.
12949 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12950 Default is @code{linear}.
12953 Set what color components to display.
12954 Default is @code{7}.
12957 Set foreground opacity. Default is @code{0.7}.
12960 Set background opacity. Default is @code{0.5}.
12963 @subsection Examples
12968 Calculate and draw histogram:
12970 ffplay -i input -vf histogram
12978 This is a high precision/quality 3d denoise filter. It aims to reduce
12979 image noise, producing smooth images and making still images really
12980 still. It should enhance compressibility.
12982 It accepts the following optional parameters:
12986 A non-negative floating point number which specifies spatial luma strength.
12987 It defaults to 4.0.
12989 @item chroma_spatial
12990 A non-negative floating point number which specifies spatial chroma strength.
12991 It defaults to 3.0*@var{luma_spatial}/4.0.
12994 A floating point number which specifies luma temporal strength. It defaults to
12995 6.0*@var{luma_spatial}/4.0.
12998 A floating point number which specifies chroma temporal strength. It defaults to
12999 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
13002 @subsection Commands
13003 This filter supports same @ref{commands} as options.
13004 The command accepts the same syntax of the corresponding option.
13006 If the specified expression is not valid, it is kept at its current
13009 @anchor{hwdownload}
13010 @section hwdownload
13012 Download hardware frames to system memory.
13014 The input must be in hardware frames, and the output a non-hardware format.
13015 Not all formats will be supported on the output - it may be necessary to insert
13016 an additional @option{format} filter immediately following in the graph to get
13017 the output in a supported format.
13021 Map hardware frames to system memory or to another device.
13023 This filter has several different modes of operation; which one is used depends
13024 on the input and output formats:
13027 Hardware frame input, normal frame output
13029 Map the input frames to system memory and pass them to the output. If the
13030 original hardware frame is later required (for example, after overlaying
13031 something else on part of it), the @option{hwmap} filter can be used again
13032 in the next mode to retrieve it.
13034 Normal frame input, hardware frame output
13036 If the input is actually a software-mapped hardware frame, then unmap it -
13037 that is, return the original hardware frame.
13039 Otherwise, a device must be provided. Create new hardware surfaces on that
13040 device for the output, then map them back to the software format at the input
13041 and give those frames to the preceding filter. This will then act like the
13042 @option{hwupload} filter, but may be able to avoid an additional copy when
13043 the input is already in a compatible format.
13045 Hardware frame input and output
13047 A device must be supplied for the output, either directly or with the
13048 @option{derive_device} option. The input and output devices must be of
13049 different types and compatible - the exact meaning of this is
13050 system-dependent, but typically it means that they must refer to the same
13051 underlying hardware context (for example, refer to the same graphics card).
13053 If the input frames were originally created on the output device, then unmap
13054 to retrieve the original frames.
13056 Otherwise, map the frames to the output device - create new hardware frames
13057 on the output corresponding to the frames on the input.
13060 The following additional parameters are accepted:
13064 Set the frame mapping mode. Some combination of:
13067 The mapped frame should be readable.
13069 The mapped frame should be writeable.
13071 The mapping will always overwrite the entire frame.
13073 This may improve performance in some cases, as the original contents of the
13074 frame need not be loaded.
13076 The mapping must not involve any copying.
13078 Indirect mappings to copies of frames are created in some cases where either
13079 direct mapping is not possible or it would have unexpected properties.
13080 Setting this flag ensures that the mapping is direct and will fail if that is
13083 Defaults to @var{read+write} if not specified.
13085 @item derive_device @var{type}
13086 Rather than using the device supplied at initialisation, instead derive a new
13087 device of type @var{type} from the device the input frames exist on.
13090 In a hardware to hardware mapping, map in reverse - create frames in the sink
13091 and map them back to the source. This may be necessary in some cases where
13092 a mapping in one direction is required but only the opposite direction is
13093 supported by the devices being used.
13095 This option is dangerous - it may break the preceding filter in undefined
13096 ways if there are any additional constraints on that filter's output.
13097 Do not use it without fully understanding the implications of its use.
13103 Upload system memory frames to hardware surfaces.
13105 The device to upload to must be supplied when the filter is initialised. If
13106 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13107 option or with the @option{derive_device} option. The input and output devices
13108 must be of different types and compatible - the exact meaning of this is
13109 system-dependent, but typically it means that they must refer to the same
13110 underlying hardware context (for example, refer to the same graphics card).
13112 The following additional parameters are accepted:
13115 @item derive_device @var{type}
13116 Rather than using the device supplied at initialisation, instead derive a new
13117 device of type @var{type} from the device the input frames exist on.
13120 @anchor{hwupload_cuda}
13121 @section hwupload_cuda
13123 Upload system memory frames to a CUDA device.
13125 It accepts the following optional parameters:
13129 The number of the CUDA device to use
13134 Apply a high-quality magnification filter designed for pixel art. This filter
13135 was originally created by Maxim Stepin.
13137 It accepts the following option:
13141 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13142 @code{hq3x} and @code{4} for @code{hq4x}.
13143 Default is @code{3}.
13147 Stack input videos horizontally.
13149 All streams must be of same pixel format and of same height.
13151 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13152 to create same output.
13154 The filter accepts the following option:
13158 Set number of input streams. Default is 2.
13161 If set to 1, force the output to terminate when the shortest input
13162 terminates. Default value is 0.
13167 Modify the hue and/or the saturation of the input.
13169 It accepts the following parameters:
13173 Specify the hue angle as a number of degrees. It accepts an expression,
13174 and defaults to "0".
13177 Specify the saturation in the [-10,10] range. It accepts an expression and
13181 Specify the hue angle as a number of radians. It accepts an
13182 expression, and defaults to "0".
13185 Specify the brightness in the [-10,10] range. It accepts an expression and
13189 @option{h} and @option{H} are mutually exclusive, and can't be
13190 specified at the same time.
13192 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13193 expressions containing the following constants:
13197 frame count of the input frame starting from 0
13200 presentation timestamp of the input frame expressed in time base units
13203 frame rate of the input video, NAN if the input frame rate is unknown
13206 timestamp expressed in seconds, NAN if the input timestamp is unknown
13209 time base of the input video
13212 @subsection Examples
13216 Set the hue to 90 degrees and the saturation to 1.0:
13222 Same command but expressing the hue in radians:
13228 Rotate hue and make the saturation swing between 0
13229 and 2 over a period of 1 second:
13231 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13235 Apply a 3 seconds saturation fade-in effect starting at 0:
13237 hue="s=min(t/3\,1)"
13240 The general fade-in expression can be written as:
13242 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13246 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13248 hue="s=max(0\, min(1\, (8-t)/3))"
13251 The general fade-out expression can be written as:
13253 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13258 @subsection Commands
13260 This filter supports the following commands:
13266 Modify the hue and/or the saturation and/or brightness of the input video.
13267 The command accepts the same syntax of the corresponding option.
13269 If the specified expression is not valid, it is kept at its current
13273 @section hysteresis
13275 Grow first stream into second stream by connecting components.
13276 This makes it possible to build more robust edge masks.
13278 This filter accepts the following options:
13282 Set which planes will be processed as bitmap, unprocessed planes will be
13283 copied from first stream.
13284 By default value 0xf, all planes will be processed.
13287 Set threshold which is used in filtering. If pixel component value is higher than
13288 this value filter algorithm for connecting components is activated.
13289 By default value is 0.
13292 The @code{hysteresis} filter also supports the @ref{framesync} options.
13296 Detect video interlacing type.
13298 This filter tries to detect if the input frames are interlaced, progressive,
13299 top or bottom field first. It will also try to detect fields that are
13300 repeated between adjacent frames (a sign of telecine).
13302 Single frame detection considers only immediately adjacent frames when classifying each frame.
13303 Multiple frame detection incorporates the classification history of previous frames.
13305 The filter will log these metadata values:
13308 @item single.current_frame
13309 Detected type of current frame using single-frame detection. One of:
13310 ``tff'' (top field first), ``bff'' (bottom field first),
13311 ``progressive'', or ``undetermined''
13314 Cumulative number of frames detected as top field first using single-frame detection.
13317 Cumulative number of frames detected as top field first using multiple-frame detection.
13320 Cumulative number of frames detected as bottom field first using single-frame detection.
13322 @item multiple.current_frame
13323 Detected type of current frame using multiple-frame detection. One of:
13324 ``tff'' (top field first), ``bff'' (bottom field first),
13325 ``progressive'', or ``undetermined''
13328 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13330 @item single.progressive
13331 Cumulative number of frames detected as progressive using single-frame detection.
13333 @item multiple.progressive
13334 Cumulative number of frames detected as progressive using multiple-frame detection.
13336 @item single.undetermined
13337 Cumulative number of frames that could not be classified using single-frame detection.
13339 @item multiple.undetermined
13340 Cumulative number of frames that could not be classified using multiple-frame detection.
13342 @item repeated.current_frame
13343 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13345 @item repeated.neither
13346 Cumulative number of frames with no repeated field.
13349 Cumulative number of frames with the top field repeated from the previous frame's top field.
13351 @item repeated.bottom
13352 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13355 The filter accepts the following options:
13359 Set interlacing threshold.
13361 Set progressive threshold.
13363 Threshold for repeated field detection.
13365 Number of frames after which a given frame's contribution to the
13366 statistics is halved (i.e., it contributes only 0.5 to its
13367 classification). The default of 0 means that all frames seen are given
13368 full weight of 1.0 forever.
13369 @item analyze_interlaced_flag
13370 When this is not 0 then idet will use the specified number of frames to determine
13371 if the interlaced flag is accurate, it will not count undetermined frames.
13372 If the flag is found to be accurate it will be used without any further
13373 computations, if it is found to be inaccurate it will be cleared without any
13374 further computations. This allows inserting the idet filter as a low computational
13375 method to clean up the interlaced flag
13380 Deinterleave or interleave fields.
13382 This filter allows one to process interlaced images fields without
13383 deinterlacing them. Deinterleaving splits the input frame into 2
13384 fields (so called half pictures). Odd lines are moved to the top
13385 half of the output image, even lines to the bottom half.
13386 You can process (filter) them independently and then re-interleave them.
13388 The filter accepts the following options:
13392 @item chroma_mode, c
13393 @item alpha_mode, a
13394 Available values for @var{luma_mode}, @var{chroma_mode} and
13395 @var{alpha_mode} are:
13401 @item deinterleave, d
13402 Deinterleave fields, placing one above the other.
13404 @item interleave, i
13405 Interleave fields. Reverse the effect of deinterleaving.
13407 Default value is @code{none}.
13409 @item luma_swap, ls
13410 @item chroma_swap, cs
13411 @item alpha_swap, as
13412 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13415 @subsection Commands
13417 This filter supports the all above options as @ref{commands}.
13421 Apply inflate effect to the video.
13423 This filter replaces the pixel by the local(3x3) average by taking into account
13424 only values higher than the pixel.
13426 It accepts the following options:
13433 Limit the maximum change for each plane, default is 65535.
13434 If 0, plane will remain unchanged.
13437 @subsection Commands
13439 This filter supports the all above options as @ref{commands}.
13443 Simple interlacing filter from progressive contents. This interleaves upper (or
13444 lower) lines from odd frames with lower (or upper) lines from even frames,
13445 halving the frame rate and preserving image height.
13448 Original Original New Frame
13449 Frame 'j' Frame 'j+1' (tff)
13450 ========== =========== ==================
13451 Line 0 --------------------> Frame 'j' Line 0
13452 Line 1 Line 1 ----> Frame 'j+1' Line 1
13453 Line 2 ---------------------> Frame 'j' Line 2
13454 Line 3 Line 3 ----> Frame 'j+1' Line 3
13456 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13459 It accepts the following optional parameters:
13463 This determines whether the interlaced frame is taken from the even
13464 (tff - default) or odd (bff) lines of the progressive frame.
13467 Vertical lowpass filter to avoid twitter interlacing and
13468 reduce moire patterns.
13472 Disable vertical lowpass filter
13475 Enable linear filter (default)
13478 Enable complex filter. This will slightly less reduce twitter and moire
13479 but better retain detail and subjective sharpness impression.
13486 Deinterlace input video by applying Donald Graft's adaptive kernel
13487 deinterling. Work on interlaced parts of a video to produce
13488 progressive frames.
13490 The description of the accepted parameters follows.
13494 Set the threshold which affects the filter's tolerance when
13495 determining if a pixel line must be processed. It must be an integer
13496 in the range [0,255] and defaults to 10. A value of 0 will result in
13497 applying the process on every pixels.
13500 Paint pixels exceeding the threshold value to white if set to 1.
13504 Set the fields order. Swap fields if set to 1, leave fields alone if
13508 Enable additional sharpening if set to 1. Default is 0.
13511 Enable twoway sharpening if set to 1. Default is 0.
13514 @subsection Examples
13518 Apply default values:
13520 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13524 Enable additional sharpening:
13530 Paint processed pixels in white:
13537 Apply kirsch operator to input video stream.
13539 The filter accepts the following option:
13543 Set which planes will be processed, unprocessed planes will be copied.
13544 By default value 0xf, all planes will be processed.
13547 Set value which will be multiplied with filtered result.
13550 Set value which will be added to filtered result.
13553 @subsection Commands
13555 This filter supports the all above options as @ref{commands}.
13559 Slowly update darker pixels.
13561 This filter makes short flashes of light appear longer.
13562 This filter accepts the following options:
13566 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13569 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13572 @subsection Commands
13574 This filter supports the all above options as @ref{commands}.
13576 @section lenscorrection
13578 Correct radial lens distortion
13580 This filter can be used to correct for radial distortion as can result from the use
13581 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13582 one can use tools available for example as part of opencv or simply trial-and-error.
13583 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13584 and extract the k1 and k2 coefficients from the resulting matrix.
13586 Note that effectively the same filter is available in the open-source tools Krita and
13587 Digikam from the KDE project.
13589 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13590 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13591 brightness distribution, so you may want to use both filters together in certain
13592 cases, though you will have to take care of ordering, i.e. whether vignetting should
13593 be applied before or after lens correction.
13595 @subsection Options
13597 The filter accepts the following options:
13601 Relative x-coordinate of the focal point of the image, and thereby the center of the
13602 distortion. This value has a range [0,1] and is expressed as fractions of the image
13603 width. Default is 0.5.
13605 Relative y-coordinate of the focal point of the image, and thereby the center of the
13606 distortion. This value has a range [0,1] and is expressed as fractions of the image
13607 height. Default is 0.5.
13609 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13610 no correction. Default is 0.
13612 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13613 0 means no correction. Default is 0.
13615 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13616 Default is @code{nearest}.
13618 Specify the color of the unmapped pixels. For the syntax of this option,
13619 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13620 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13623 The formula that generates the correction is:
13625 @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)
13627 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13628 distances from the focal point in the source and target images, respectively.
13630 @subsection Commands
13632 This filter supports the all above options as @ref{commands}.
13636 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13638 The @code{lensfun} filter requires the camera make, camera model, and lens model
13639 to apply the lens correction. The filter will load the lensfun database and
13640 query it to find the corresponding camera and lens entries in the database. As
13641 long as these entries can be found with the given options, the filter can
13642 perform corrections on frames. Note that incomplete strings will result in the
13643 filter choosing the best match with the given options, and the filter will
13644 output the chosen camera and lens models (logged with level "info"). You must
13645 provide the make, camera model, and lens model as they are required.
13647 The filter accepts the following options:
13651 The make of the camera (for example, "Canon"). This option is required.
13654 The model of the camera (for example, "Canon EOS 100D"). This option is
13658 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13659 option is required.
13662 The type of correction to apply. The following values are valid options:
13666 Enables fixing lens vignetting.
13669 Enables fixing lens geometry. This is the default.
13672 Enables fixing chromatic aberrations.
13675 Enables fixing lens vignetting and lens geometry.
13678 Enables fixing lens vignetting and chromatic aberrations.
13681 Enables fixing both lens geometry and chromatic aberrations.
13684 Enables all possible corrections.
13688 The focal length of the image/video (zoom; expected constant for video). For
13689 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13690 range should be chosen when using that lens. Default 18.
13693 The aperture of the image/video (expected constant for video). Note that
13694 aperture is only used for vignetting correction. Default 3.5.
13696 @item focus_distance
13697 The focus distance of the image/video (expected constant for video). Note that
13698 focus distance is only used for vignetting and only slightly affects the
13699 vignetting correction process. If unknown, leave it at the default value (which
13703 The scale factor which is applied after transformation. After correction the
13704 video is no longer necessarily rectangular. This parameter controls how much of
13705 the resulting image is visible. The value 0 means that a value will be chosen
13706 automatically such that there is little or no unmapped area in the output
13707 image. 1.0 means that no additional scaling is done. Lower values may result
13708 in more of the corrected image being visible, while higher values may avoid
13709 unmapped areas in the output.
13711 @item target_geometry
13712 The target geometry of the output image/video. The following values are valid
13716 @item rectilinear (default)
13719 @item equirectangular
13720 @item fisheye_orthographic
13721 @item fisheye_stereographic
13722 @item fisheye_equisolid
13723 @item fisheye_thoby
13726 Apply the reverse of image correction (instead of correcting distortion, apply
13729 @item interpolation
13730 The type of interpolation used when correcting distortion. The following values
13735 @item linear (default)
13740 @subsection Examples
13744 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13745 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13749 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
13753 Apply the same as before, but only for the first 5 seconds of video.
13756 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
13763 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13764 score between two input videos.
13766 The obtained VMAF score is printed through the logging system.
13768 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13769 After installing the library it can be enabled using:
13770 @code{./configure --enable-libvmaf}.
13771 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13773 The filter has following options:
13777 Set the model path which is to be used for SVM.
13778 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13781 Set the file path to be used to store logs.
13784 Set the format of the log file (csv, json or xml).
13786 @item enable_transform
13787 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13788 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13789 Default value: @code{false}
13792 Invokes the phone model which will generate VMAF scores higher than in the
13793 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13794 Default value: @code{false}
13797 Enables computing psnr along with vmaf.
13798 Default value: @code{false}
13801 Enables computing ssim along with vmaf.
13802 Default value: @code{false}
13805 Enables computing ms_ssim along with vmaf.
13806 Default value: @code{false}
13809 Set the pool method to be used for computing vmaf.
13810 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13813 Set number of threads to be used when computing vmaf.
13814 Default value: @code{0}, which makes use of all available logical processors.
13817 Set interval for frame subsampling used when computing vmaf.
13818 Default value: @code{1}
13820 @item enable_conf_interval
13821 Enables confidence interval.
13822 Default value: @code{false}
13825 This filter also supports the @ref{framesync} options.
13827 @subsection Examples
13830 On the below examples the input file @file{main.mpg} being processed is
13831 compared with the reference file @file{ref.mpg}.
13834 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13838 Example with options:
13840 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13844 Example with options and different containers:
13846 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 -
13852 Limits the pixel components values to the specified range [min, max].
13854 The filter accepts the following options:
13858 Lower bound. Defaults to the lowest allowed value for the input.
13861 Upper bound. Defaults to the highest allowed value for the input.
13864 Specify which planes will be processed. Defaults to all available.
13867 @subsection Commands
13869 This filter supports the all above options as @ref{commands}.
13875 The filter accepts the following options:
13879 Set the number of loops. Setting this value to -1 will result in infinite loops.
13883 Set maximal size in number of frames. Default is 0.
13886 Set first frame of loop. Default is 0.
13889 @subsection Examples
13893 Loop single first frame infinitely:
13895 loop=loop=-1:size=1:start=0
13899 Loop single first frame 10 times:
13901 loop=loop=10:size=1:start=0
13905 Loop 10 first frames 5 times:
13907 loop=loop=5:size=10:start=0
13913 Apply a 1D LUT to an input video.
13915 The filter accepts the following options:
13919 Set the 1D LUT file name.
13921 Currently supported formats:
13930 Select interpolation mode.
13932 Available values are:
13936 Use values from the nearest defined point.
13938 Interpolate values using the linear interpolation.
13940 Interpolate values using the cosine interpolation.
13942 Interpolate values using the cubic interpolation.
13944 Interpolate values using the spline interpolation.
13948 @subsection Commands
13950 This filter supports the all above options as @ref{commands}.
13955 Apply a 3D LUT to an input video.
13957 The filter accepts the following options:
13961 Set the 3D LUT file name.
13963 Currently supported formats:
13977 Select interpolation mode.
13979 Available values are:
13983 Use values from the nearest defined point.
13985 Interpolate values using the 8 points defining a cube.
13987 Interpolate values using a tetrahedron.
13989 Interpolate values using a pyramid.
13991 Interpolate values using a prism.
13997 Turn certain luma values into transparency.
13999 The filter accepts the following options:
14003 Set the luma which will be used as base for transparency.
14004 Default value is @code{0}.
14007 Set the range of luma values to be keyed out.
14008 Default value is @code{0.01}.
14011 Set the range of softness. Default value is @code{0}.
14012 Use this to control gradual transition from zero to full transparency.
14015 @subsection Commands
14016 This filter supports same @ref{commands} as options.
14017 The command accepts the same syntax of the corresponding option.
14019 If the specified expression is not valid, it is kept at its current
14022 @section lut, lutrgb, lutyuv
14024 Compute a look-up table for binding each pixel component input value
14025 to an output value, and apply it to the input video.
14027 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
14028 to an RGB input video.
14030 These filters accept the following parameters:
14033 set first pixel component expression
14035 set second pixel component expression
14037 set third pixel component expression
14039 set fourth pixel component expression, corresponds to the alpha component
14042 set red component expression
14044 set green component expression
14046 set blue component expression
14048 alpha component expression
14051 set Y/luminance component expression
14053 set U/Cb component expression
14055 set V/Cr component expression
14058 Each of them specifies the expression to use for computing the lookup table for
14059 the corresponding pixel component values.
14061 The exact component associated to each of the @var{c*} options depends on the
14064 The @var{lut} filter requires either YUV or RGB pixel formats in input,
14065 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
14067 The expressions can contain the following constants and functions:
14072 The input width and height.
14075 The input value for the pixel component.
14078 The input value, clipped to the @var{minval}-@var{maxval} range.
14081 The maximum value for the pixel component.
14084 The minimum value for the pixel component.
14087 The negated value for the pixel component value, clipped to the
14088 @var{minval}-@var{maxval} range; it corresponds to the expression
14089 "maxval-clipval+minval".
14092 The computed value in @var{val}, clipped to the
14093 @var{minval}-@var{maxval} range.
14095 @item gammaval(gamma)
14096 The computed gamma correction value of the pixel component value,
14097 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14099 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14103 All expressions default to "val".
14105 @subsection Examples
14109 Negate input video:
14111 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14112 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14115 The above is the same as:
14117 lutrgb="r=negval:g=negval:b=negval"
14118 lutyuv="y=negval:u=negval:v=negval"
14128 Remove chroma components, turning the video into a graytone image:
14130 lutyuv="u=128:v=128"
14134 Apply a luma burning effect:
14140 Remove green and blue components:
14146 Set a constant alpha channel value on input:
14148 format=rgba,lutrgb=a="maxval-minval/2"
14152 Correct luminance gamma by a factor of 0.5:
14154 lutyuv=y=gammaval(0.5)
14158 Discard least significant bits of luma:
14160 lutyuv=y='bitand(val, 128+64+32)'
14164 Technicolor like effect:
14166 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14170 @section lut2, tlut2
14172 The @code{lut2} filter takes two input streams and outputs one
14175 The @code{tlut2} (time lut2) filter takes two consecutive frames
14176 from one single stream.
14178 This filter accepts the following parameters:
14181 set first pixel component expression
14183 set second pixel component expression
14185 set third pixel component expression
14187 set fourth pixel component expression, corresponds to the alpha component
14190 set output bit depth, only available for @code{lut2} filter. By default is 0,
14191 which means bit depth is automatically picked from first input format.
14194 The @code{lut2} filter also supports the @ref{framesync} options.
14196 Each of them specifies the expression to use for computing the lookup table for
14197 the corresponding pixel component values.
14199 The exact component associated to each of the @var{c*} options depends on the
14202 The expressions can contain the following constants:
14207 The input width and height.
14210 The first input value for the pixel component.
14213 The second input value for the pixel component.
14216 The first input video bit depth.
14219 The second input video bit depth.
14222 All expressions default to "x".
14224 @subsection Examples
14228 Highlight differences between two RGB video streams:
14230 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)'
14234 Highlight differences between two YUV video streams:
14236 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)'
14240 Show max difference between two video streams:
14242 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)))'
14246 @section maskedclamp
14248 Clamp the first input stream with the second input and third input stream.
14250 Returns the value of first stream to be between second input
14251 stream - @code{undershoot} and third input stream + @code{overshoot}.
14253 This filter accepts the following options:
14256 Default value is @code{0}.
14259 Default value is @code{0}.
14262 Set which planes will be processed as bitmap, unprocessed planes will be
14263 copied from first stream.
14264 By default value 0xf, all planes will be processed.
14267 @subsection Commands
14269 This filter supports the all above options as @ref{commands}.
14273 Merge the second and third input stream into output stream using absolute differences
14274 between second input stream and first input stream and absolute difference between
14275 third input stream and first input stream. The picked value will be from second input
14276 stream if second absolute difference is greater than first one or from third input stream
14279 This filter accepts the following options:
14282 Set which planes will be processed as bitmap, unprocessed planes will be
14283 copied from first stream.
14284 By default value 0xf, all planes will be processed.
14287 @subsection Commands
14289 This filter supports the all above options as @ref{commands}.
14291 @section maskedmerge
14293 Merge the first input stream with the second input stream using per pixel
14294 weights in the third input stream.
14296 A value of 0 in the third stream pixel component means that pixel component
14297 from first stream is returned unchanged, while maximum value (eg. 255 for
14298 8-bit videos) means that pixel component from second stream is returned
14299 unchanged. Intermediate values define the amount of merging between both
14300 input stream's pixel components.
14302 This filter accepts the following options:
14305 Set which planes will be processed as bitmap, unprocessed planes will be
14306 copied from first stream.
14307 By default value 0xf, all planes will be processed.
14310 @subsection Commands
14312 This filter supports the all above options as @ref{commands}.
14316 Merge the second and third input stream into output stream using absolute differences
14317 between second input stream and first input stream and absolute difference between
14318 third input stream and first input stream. The picked value will be from second input
14319 stream if second absolute difference is less than first one or from third input stream
14322 This filter accepts the following options:
14325 Set which planes will be processed as bitmap, unprocessed planes will be
14326 copied from first stream.
14327 By default value 0xf, all planes will be processed.
14330 @subsection Commands
14332 This filter supports the all above options as @ref{commands}.
14334 @section maskedthreshold
14335 Pick pixels comparing absolute difference of two video streams with fixed
14338 If absolute difference between pixel component of first and second video
14339 stream is equal or lower than user supplied threshold than pixel component
14340 from first video stream is picked, otherwise pixel component from second
14341 video stream is picked.
14343 This filter accepts the following options:
14346 Set threshold used when picking pixels from absolute difference from two input
14350 Set which planes will be processed as bitmap, unprocessed planes will be
14351 copied from second stream.
14352 By default value 0xf, all planes will be processed.
14355 @subsection Commands
14357 This filter supports the all above options as @ref{commands}.
14360 Create mask from input video.
14362 For example it is useful to create motion masks after @code{tblend} filter.
14364 This filter accepts the following options:
14368 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14371 Set high threshold. Any pixel component higher than this value will be set to max value
14372 allowed for current pixel format.
14375 Set planes to filter, by default all available planes are filtered.
14378 Fill all frame pixels with this value.
14381 Set max average pixel value for frame. If sum of all pixel components is higher that this
14382 average, output frame will be completely filled with value set by @var{fill} option.
14383 Typically useful for scene changes when used in combination with @code{tblend} filter.
14388 Apply motion-compensation deinterlacing.
14390 It needs one field per frame as input and must thus be used together
14391 with yadif=1/3 or equivalent.
14393 This filter accepts the following options:
14396 Set the deinterlacing mode.
14398 It accepts one of the following values:
14403 use iterative motion estimation
14405 like @samp{slow}, but use multiple reference frames.
14407 Default value is @samp{fast}.
14410 Set the picture field parity assumed for the input video. It must be
14411 one of the following values:
14415 assume top field first
14417 assume bottom field first
14420 Default value is @samp{bff}.
14423 Set per-block quantization parameter (QP) used by the internal
14426 Higher values should result in a smoother motion vector field but less
14427 optimal individual vectors. Default value is 1.
14432 Pick median pixel from certain rectangle defined by radius.
14434 This filter accepts the following options:
14438 Set horizontal radius size. Default value is @code{1}.
14439 Allowed range is integer from 1 to 127.
14442 Set which planes to process. Default is @code{15}, which is all available planes.
14445 Set vertical radius size. Default value is @code{0}.
14446 Allowed range is integer from 0 to 127.
14447 If it is 0, value will be picked from horizontal @code{radius} option.
14450 Set median percentile. Default value is @code{0.5}.
14451 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14452 minimum values, and @code{1} maximum values.
14455 @subsection Commands
14456 This filter supports same @ref{commands} as options.
14457 The command accepts the same syntax of the corresponding option.
14459 If the specified expression is not valid, it is kept at its current
14462 @section mergeplanes
14464 Merge color channel components from several video streams.
14466 The filter accepts up to 4 input streams, and merge selected input
14467 planes to the output video.
14469 This filter accepts the following options:
14472 Set input to output plane mapping. Default is @code{0}.
14474 The mappings is specified as a bitmap. It should be specified as a
14475 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14476 mapping for the first plane of the output stream. 'A' sets the number of
14477 the input stream to use (from 0 to 3), and 'a' the plane number of the
14478 corresponding input to use (from 0 to 3). The rest of the mappings is
14479 similar, 'Bb' describes the mapping for the output stream second
14480 plane, 'Cc' describes the mapping for the output stream third plane and
14481 'Dd' describes the mapping for the output stream fourth plane.
14484 Set output pixel format. Default is @code{yuva444p}.
14487 @subsection Examples
14491 Merge three gray video streams of same width and height into single video stream:
14493 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14497 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14499 [a0][a1]mergeplanes=0x00010210:yuva444p
14503 Swap Y and A plane in yuva444p stream:
14505 format=yuva444p,mergeplanes=0x03010200:yuva444p
14509 Swap U and V plane in yuv420p stream:
14511 format=yuv420p,mergeplanes=0x000201:yuv420p
14515 Cast a rgb24 clip to yuv444p:
14517 format=rgb24,mergeplanes=0x000102:yuv444p
14523 Estimate and export motion vectors using block matching algorithms.
14524 Motion vectors are stored in frame side data to be used by other filters.
14526 This filter accepts the following options:
14529 Specify the motion estimation method. Accepts one of the following values:
14533 Exhaustive search algorithm.
14535 Three step search algorithm.
14537 Two dimensional logarithmic search algorithm.
14539 New three step search algorithm.
14541 Four step search algorithm.
14543 Diamond search algorithm.
14545 Hexagon-based search algorithm.
14547 Enhanced predictive zonal search algorithm.
14549 Uneven multi-hexagon search algorithm.
14551 Default value is @samp{esa}.
14554 Macroblock size. Default @code{16}.
14557 Search parameter. Default @code{7}.
14560 @section midequalizer
14562 Apply Midway Image Equalization effect using two video streams.
14564 Midway Image Equalization adjusts a pair of images to have the same
14565 histogram, while maintaining their dynamics as much as possible. It's
14566 useful for e.g. matching exposures from a pair of stereo cameras.
14568 This filter has two inputs and one output, which must be of same pixel format, but
14569 may be of different sizes. The output of filter is first input adjusted with
14570 midway histogram of both inputs.
14572 This filter accepts the following option:
14576 Set which planes to process. Default is @code{15}, which is all available planes.
14579 @section minterpolate
14581 Convert the video to specified frame rate using motion interpolation.
14583 This filter accepts the following options:
14586 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}.
14589 Motion interpolation mode. Following values are accepted:
14592 Duplicate previous or next frame for interpolating new ones.
14594 Blend source frames. Interpolated frame is mean of previous and next frames.
14596 Motion compensated interpolation. Following options are effective when this mode is selected:
14600 Motion compensation mode. Following values are accepted:
14603 Overlapped block motion compensation.
14605 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14607 Default mode is @samp{obmc}.
14610 Motion estimation mode. Following values are accepted:
14613 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14615 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14617 Default mode is @samp{bilat}.
14620 The algorithm to be used for motion estimation. Following values are accepted:
14623 Exhaustive search algorithm.
14625 Three step search algorithm.
14627 Two dimensional logarithmic search algorithm.
14629 New three step search algorithm.
14631 Four step search algorithm.
14633 Diamond search algorithm.
14635 Hexagon-based search algorithm.
14637 Enhanced predictive zonal search algorithm.
14639 Uneven multi-hexagon search algorithm.
14641 Default algorithm is @samp{epzs}.
14644 Macroblock size. Default @code{16}.
14647 Motion estimation search parameter. Default @code{32}.
14650 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).
14655 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:
14658 Disable scene change detection.
14660 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14662 Default method is @samp{fdiff}.
14664 @item scd_threshold
14665 Scene change detection threshold. Default is @code{10.}.
14670 Mix several video input streams into one video stream.
14672 A description of the accepted options follows.
14676 The number of inputs. If unspecified, it defaults to 2.
14679 Specify weight of each input video stream as sequence.
14680 Each weight is separated by space. If number of weights
14681 is smaller than number of @var{frames} last specified
14682 weight will be used for all remaining unset weights.
14685 Specify scale, if it is set it will be multiplied with sum
14686 of each weight multiplied with pixel values to give final destination
14687 pixel value. By default @var{scale} is auto scaled to sum of weights.
14690 Specify how end of stream is determined.
14693 The duration of the longest input. (default)
14696 The duration of the shortest input.
14699 The duration of the first input.
14703 @subsection Commands
14705 This filter supports the following commands:
14709 Syntax is same as option with same name.
14712 @section mpdecimate
14714 Drop frames that do not differ greatly from the previous frame in
14715 order to reduce frame rate.
14717 The main use of this filter is for very-low-bitrate encoding
14718 (e.g. streaming over dialup modem), but it could in theory be used for
14719 fixing movies that were inverse-telecined incorrectly.
14721 A description of the accepted options follows.
14725 Set the maximum number of consecutive frames which can be dropped (if
14726 positive), or the minimum interval between dropped frames (if
14727 negative). If the value is 0, the frame is dropped disregarding the
14728 number of previous sequentially dropped frames.
14730 Default value is 0.
14735 Set the dropping threshold values.
14737 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14738 represent actual pixel value differences, so a threshold of 64
14739 corresponds to 1 unit of difference for each pixel, or the same spread
14740 out differently over the block.
14742 A frame is a candidate for dropping if no 8x8 blocks differ by more
14743 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14744 meaning the whole image) differ by more than a threshold of @option{lo}.
14746 Default value for @option{hi} is 64*12, default value for @option{lo} is
14747 64*5, and default value for @option{frac} is 0.33.
14753 Negate (invert) the input video.
14755 It accepts the following option:
14760 With value 1, it negates the alpha component, if present. Default value is 0.
14766 Denoise frames using Non-Local Means algorithm.
14768 Each pixel is adjusted by looking for other pixels with similar contexts. This
14769 context similarity is defined by comparing their surrounding patches of size
14770 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14773 Note that the research area defines centers for patches, which means some
14774 patches will be made of pixels outside that research area.
14776 The filter accepts the following options.
14780 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14783 Set patch size. Default is 7. Must be odd number in range [0, 99].
14786 Same as @option{p} but for chroma planes.
14788 The default value is @var{0} and means automatic.
14791 Set research size. Default is 15. Must be odd number in range [0, 99].
14794 Same as @option{r} but for chroma planes.
14796 The default value is @var{0} and means automatic.
14801 Deinterlace video using neural network edge directed interpolation.
14803 This filter accepts the following options:
14807 Mandatory option, without binary file filter can not work.
14808 Currently file can be found here:
14809 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14812 Set which frames to deinterlace, by default it is @code{all}.
14813 Can be @code{all} or @code{interlaced}.
14816 Set mode of operation.
14818 Can be one of the following:
14822 Use frame flags, both fields.
14824 Use frame flags, single field.
14826 Use top field only.
14828 Use bottom field only.
14830 Use both fields, top first.
14832 Use both fields, bottom first.
14836 Set which planes to process, by default filter process all frames.
14839 Set size of local neighborhood around each pixel, used by the predictor neural
14842 Can be one of the following:
14855 Set the number of neurons in predictor neural network.
14856 Can be one of the following:
14867 Controls the number of different neural network predictions that are blended
14868 together to compute the final output value. Can be @code{fast}, default or
14872 Set which set of weights to use in the predictor.
14873 Can be one of the following:
14877 weights trained to minimize absolute error
14879 weights trained to minimize squared error
14883 Controls whether or not the prescreener neural network is used to decide
14884 which pixels should be processed by the predictor neural network and which
14885 can be handled by simple cubic interpolation.
14886 The prescreener is trained to know whether cubic interpolation will be
14887 sufficient for a pixel or whether it should be predicted by the predictor nn.
14888 The computational complexity of the prescreener nn is much less than that of
14889 the predictor nn. Since most pixels can be handled by cubic interpolation,
14890 using the prescreener generally results in much faster processing.
14891 The prescreener is pretty accurate, so the difference between using it and not
14892 using it is almost always unnoticeable.
14894 Can be one of the following:
14904 Default is @code{new}.
14907 @subsection Commands
14908 This filter supports same @ref{commands} as options, excluding @var{weights} option.
14912 Force libavfilter not to use any of the specified pixel formats for the
14913 input to the next filter.
14915 It accepts the following parameters:
14919 A '|'-separated list of pixel format names, such as
14920 pix_fmts=yuv420p|monow|rgb24".
14924 @subsection Examples
14928 Force libavfilter to use a format different from @var{yuv420p} for the
14929 input to the vflip filter:
14931 noformat=pix_fmts=yuv420p,vflip
14935 Convert the input video to any of the formats not contained in the list:
14937 noformat=yuv420p|yuv444p|yuv410p
14943 Add noise on video input frame.
14945 The filter accepts the following options:
14953 Set noise seed for specific pixel component or all pixel components in case
14954 of @var{all_seed}. Default value is @code{123457}.
14956 @item all_strength, alls
14957 @item c0_strength, c0s
14958 @item c1_strength, c1s
14959 @item c2_strength, c2s
14960 @item c3_strength, c3s
14961 Set noise strength for specific pixel component or all pixel components in case
14962 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14964 @item all_flags, allf
14965 @item c0_flags, c0f
14966 @item c1_flags, c1f
14967 @item c2_flags, c2f
14968 @item c3_flags, c3f
14969 Set pixel component flags or set flags for all components if @var{all_flags}.
14970 Available values for component flags are:
14973 averaged temporal noise (smoother)
14975 mix random noise with a (semi)regular pattern
14977 temporal noise (noise pattern changes between frames)
14979 uniform noise (gaussian otherwise)
14983 @subsection Examples
14985 Add temporal and uniform noise to input video:
14987 noise=alls=20:allf=t+u
14992 Normalize RGB video (aka histogram stretching, contrast stretching).
14993 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
14995 For each channel of each frame, the filter computes the input range and maps
14996 it linearly to the user-specified output range. The output range defaults
14997 to the full dynamic range from pure black to pure white.
14999 Temporal smoothing can be used on the input range to reduce flickering (rapid
15000 changes in brightness) caused when small dark or bright objects enter or leave
15001 the scene. This is similar to the auto-exposure (automatic gain control) on a
15002 video camera, and, like a video camera, it may cause a period of over- or
15003 under-exposure of the video.
15005 The R,G,B channels can be normalized independently, which may cause some
15006 color shifting, or linked together as a single channel, which prevents
15007 color shifting. Linked normalization preserves hue. Independent normalization
15008 does not, so it can be used to remove some color casts. Independent and linked
15009 normalization can be combined in any ratio.
15011 The normalize filter accepts the following options:
15016 Colors which define the output range. The minimum input value is mapped to
15017 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
15018 The defaults are black and white respectively. Specifying white for
15019 @var{blackpt} and black for @var{whitept} will give color-inverted,
15020 normalized video. Shades of grey can be used to reduce the dynamic range
15021 (contrast). Specifying saturated colors here can create some interesting
15025 The number of previous frames to use for temporal smoothing. The input range
15026 of each channel is smoothed using a rolling average over the current frame
15027 and the @var{smoothing} previous frames. The default is 0 (no temporal
15031 Controls the ratio of independent (color shifting) channel normalization to
15032 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
15033 independent. Defaults to 1.0 (fully independent).
15036 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
15037 expensive no-op. Defaults to 1.0 (full strength).
15041 @subsection Commands
15042 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
15043 The command accepts the same syntax of the corresponding option.
15045 If the specified expression is not valid, it is kept at its current
15048 @subsection Examples
15050 Stretch video contrast to use the full dynamic range, with no temporal
15051 smoothing; may flicker depending on the source content:
15053 normalize=blackpt=black:whitept=white:smoothing=0
15056 As above, but with 50 frames of temporal smoothing; flicker should be
15057 reduced, depending on the source content:
15059 normalize=blackpt=black:whitept=white:smoothing=50
15062 As above, but with hue-preserving linked channel normalization:
15064 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
15067 As above, but with half strength:
15069 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
15072 Map the darkest input color to red, the brightest input color to cyan:
15074 normalize=blackpt=red:whitept=cyan
15079 Pass the video source unchanged to the output.
15082 Optical Character Recognition
15084 This filter uses Tesseract for optical character recognition. To enable
15085 compilation of this filter, you need to configure FFmpeg with
15086 @code{--enable-libtesseract}.
15088 It accepts the following options:
15092 Set datapath to tesseract data. Default is to use whatever was
15093 set at installation.
15096 Set language, default is "eng".
15099 Set character whitelist.
15102 Set character blacklist.
15105 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15106 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15110 Apply a video transform using libopencv.
15112 To enable this filter, install the libopencv library and headers and
15113 configure FFmpeg with @code{--enable-libopencv}.
15115 It accepts the following parameters:
15120 The name of the libopencv filter to apply.
15122 @item filter_params
15123 The parameters to pass to the libopencv filter. If not specified, the default
15124 values are assumed.
15128 Refer to the official libopencv documentation for more precise
15130 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15132 Several libopencv filters are supported; see the following subsections.
15137 Dilate an image by using a specific structuring element.
15138 It corresponds to the libopencv function @code{cvDilate}.
15140 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15142 @var{struct_el} represents a structuring element, and has the syntax:
15143 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15145 @var{cols} and @var{rows} represent the number of columns and rows of
15146 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15147 point, and @var{shape} the shape for the structuring element. @var{shape}
15148 must be "rect", "cross", "ellipse", or "custom".
15150 If the value for @var{shape} is "custom", it must be followed by a
15151 string of the form "=@var{filename}". The file with name
15152 @var{filename} is assumed to represent a binary image, with each
15153 printable character corresponding to a bright pixel. When a custom
15154 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15155 or columns and rows of the read file are assumed instead.
15157 The default value for @var{struct_el} is "3x3+0x0/rect".
15159 @var{nb_iterations} specifies the number of times the transform is
15160 applied to the image, and defaults to 1.
15164 # Use the default values
15167 # Dilate using a structuring element with a 5x5 cross, iterating two times
15168 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15170 # Read the shape from the file diamond.shape, iterating two times.
15171 # The file diamond.shape may contain a pattern of characters like this
15177 # The specified columns and rows are ignored
15178 # but the anchor point coordinates are not
15179 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15184 Erode an image by using a specific structuring element.
15185 It corresponds to the libopencv function @code{cvErode}.
15187 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15188 with the same syntax and semantics as the @ref{dilate} filter.
15192 Smooth the input video.
15194 The filter takes the following parameters:
15195 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15197 @var{type} is the type of smooth filter to apply, and must be one of
15198 the following values: "blur", "blur_no_scale", "median", "gaussian",
15199 or "bilateral". The default value is "gaussian".
15201 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15202 depends on the smooth type. @var{param1} and
15203 @var{param2} accept integer positive values or 0. @var{param3} and
15204 @var{param4} accept floating point values.
15206 The default value for @var{param1} is 3. The default value for the
15207 other parameters is 0.
15209 These parameters correspond to the parameters assigned to the
15210 libopencv function @code{cvSmooth}.
15212 @section oscilloscope
15214 2D Video Oscilloscope.
15216 Useful to measure spatial impulse, step responses, chroma delays, etc.
15218 It accepts the following parameters:
15222 Set scope center x position.
15225 Set scope center y position.
15228 Set scope size, relative to frame diagonal.
15231 Set scope tilt/rotation.
15237 Set trace center x position.
15240 Set trace center y position.
15243 Set trace width, relative to width of frame.
15246 Set trace height, relative to height of frame.
15249 Set which components to trace. By default it traces first three components.
15252 Draw trace grid. By default is enabled.
15255 Draw some statistics. By default is enabled.
15258 Draw scope. By default is enabled.
15261 @subsection Commands
15262 This filter supports same @ref{commands} as options.
15263 The command accepts the same syntax of the corresponding option.
15265 If the specified expression is not valid, it is kept at its current
15268 @subsection Examples
15272 Inspect full first row of video frame.
15274 oscilloscope=x=0.5:y=0:s=1
15278 Inspect full last row of video frame.
15280 oscilloscope=x=0.5:y=1:s=1
15284 Inspect full 5th line of video frame of height 1080.
15286 oscilloscope=x=0.5:y=5/1080:s=1
15290 Inspect full last column of video frame.
15292 oscilloscope=x=1:y=0.5:s=1:t=1
15300 Overlay one video on top of another.
15302 It takes two inputs and has one output. The first input is the "main"
15303 video on which the second input is overlaid.
15305 It accepts the following parameters:
15307 A description of the accepted options follows.
15312 Set the expression for the x and y coordinates of the overlaid video
15313 on the main video. Default value is "0" for both expressions. In case
15314 the expression is invalid, it is set to a huge value (meaning that the
15315 overlay will not be displayed within the output visible area).
15318 See @ref{framesync}.
15321 Set when the expressions for @option{x}, and @option{y} are evaluated.
15323 It accepts the following values:
15326 only evaluate expressions once during the filter initialization or
15327 when a command is processed
15330 evaluate expressions for each incoming frame
15333 Default value is @samp{frame}.
15336 See @ref{framesync}.
15339 Set the format for the output video.
15341 It accepts the following values:
15344 force YUV420 output
15347 force YUV420p10 output
15350 force YUV422 output
15353 force YUV422p10 output
15356 force YUV444 output
15359 force packed RGB output
15362 force planar RGB output
15365 automatically pick format
15368 Default value is @samp{yuv420}.
15371 See @ref{framesync}.
15374 Set format of alpha of the overlaid video, it can be @var{straight} or
15375 @var{premultiplied}. Default is @var{straight}.
15378 The @option{x}, and @option{y} expressions can contain the following
15384 The main input width and height.
15388 The overlay input width and height.
15392 The computed values for @var{x} and @var{y}. They are evaluated for
15397 horizontal and vertical chroma subsample values of the output
15398 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15402 the number of input frame, starting from 0
15405 the position in the file of the input frame, NAN if unknown
15408 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15412 This filter also supports the @ref{framesync} options.
15414 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15415 when evaluation is done @emph{per frame}, and will evaluate to NAN
15416 when @option{eval} is set to @samp{init}.
15418 Be aware that frames are taken from each input video in timestamp
15419 order, hence, if their initial timestamps differ, it is a good idea
15420 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15421 have them begin in the same zero timestamp, as the example for
15422 the @var{movie} filter does.
15424 You can chain together more overlays but you should test the
15425 efficiency of such approach.
15427 @subsection Commands
15429 This filter supports the following commands:
15433 Modify the x and y of the overlay input.
15434 The command accepts the same syntax of the corresponding option.
15436 If the specified expression is not valid, it is kept at its current
15440 @subsection Examples
15444 Draw the overlay at 10 pixels from the bottom right corner of the main
15447 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15450 Using named options the example above becomes:
15452 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15456 Insert a transparent PNG logo in the bottom left corner of the input,
15457 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15459 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15463 Insert 2 different transparent PNG logos (second logo on bottom
15464 right corner) using the @command{ffmpeg} tool:
15466 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
15470 Add a transparent color layer on top of the main video; @code{WxH}
15471 must specify the size of the main input to the overlay filter:
15473 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15477 Play an original video and a filtered version (here with the deshake
15478 filter) side by side using the @command{ffplay} tool:
15480 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15483 The above command is the same as:
15485 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15489 Make a sliding overlay appearing from the left to the right top part of the
15490 screen starting since time 2:
15492 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15496 Compose output by putting two input videos side to side:
15498 ffmpeg -i left.avi -i right.avi -filter_complex "
15499 nullsrc=size=200x100 [background];
15500 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15501 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15502 [background][left] overlay=shortest=1 [background+left];
15503 [background+left][right] overlay=shortest=1:x=100 [left+right]
15508 Mask 10-20 seconds of a video by applying the delogo filter to a section
15510 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15511 -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]'
15516 Chain several overlays in cascade:
15518 nullsrc=s=200x200 [bg];
15519 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15520 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15521 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15522 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15523 [in3] null, [mid2] overlay=100:100 [out0]
15528 @anchor{overlay_cuda}
15529 @section overlay_cuda
15531 Overlay one video on top of another.
15533 This is the CUDA variant of the @ref{overlay} filter.
15534 It only accepts CUDA frames. The underlying input pixel formats have to match.
15536 It takes two inputs and has one output. The first input is the "main"
15537 video on which the second input is overlaid.
15539 It accepts the following parameters:
15544 Set the x and y coordinates of the overlaid video on the main video.
15545 Default value is "0" for both expressions.
15548 See @ref{framesync}.
15551 See @ref{framesync}.
15554 See @ref{framesync}.
15558 This filter also supports the @ref{framesync} options.
15562 Apply Overcomplete Wavelet denoiser.
15564 The filter accepts the following options:
15570 Larger depth values will denoise lower frequency components more, but
15571 slow down filtering.
15573 Must be an int in the range 8-16, default is @code{8}.
15575 @item luma_strength, ls
15578 Must be a double value in the range 0-1000, default is @code{1.0}.
15580 @item chroma_strength, cs
15581 Set chroma strength.
15583 Must be a double value in the range 0-1000, default is @code{1.0}.
15589 Add paddings to the input image, and place the original input at the
15590 provided @var{x}, @var{y} coordinates.
15592 It accepts the following parameters:
15597 Specify an expression for the size of the output image with the
15598 paddings added. If the value for @var{width} or @var{height} is 0, the
15599 corresponding input size is used for the output.
15601 The @var{width} expression can reference the value set by the
15602 @var{height} expression, and vice versa.
15604 The default value of @var{width} and @var{height} is 0.
15608 Specify the offsets to place the input image at within the padded area,
15609 with respect to the top/left border of the output image.
15611 The @var{x} expression can reference the value set by the @var{y}
15612 expression, and vice versa.
15614 The default value of @var{x} and @var{y} is 0.
15616 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15617 so the input image is centered on the padded area.
15620 Specify the color of the padded area. For the syntax of this option,
15621 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15622 manual,ffmpeg-utils}.
15624 The default value of @var{color} is "black".
15627 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15629 It accepts the following values:
15633 Only evaluate expressions once during the filter initialization or when
15634 a command is processed.
15637 Evaluate expressions for each incoming frame.
15641 Default value is @samp{init}.
15644 Pad to aspect instead to a resolution.
15648 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15649 options are expressions containing the following constants:
15654 The input video width and height.
15658 These are the same as @var{in_w} and @var{in_h}.
15662 The output width and height (the size of the padded area), as
15663 specified by the @var{width} and @var{height} expressions.
15667 These are the same as @var{out_w} and @var{out_h}.
15671 The x and y offsets as specified by the @var{x} and @var{y}
15672 expressions, or NAN if not yet specified.
15675 same as @var{iw} / @var{ih}
15678 input sample aspect ratio
15681 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15685 The horizontal and vertical chroma subsample values. For example for the
15686 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15689 @subsection Examples
15693 Add paddings with the color "violet" to the input video. The output video
15694 size is 640x480, and the top-left corner of the input video is placed at
15697 pad=640:480:0:40:violet
15700 The example above is equivalent to the following command:
15702 pad=width=640:height=480:x=0:y=40:color=violet
15706 Pad the input to get an output with dimensions increased by 3/2,
15707 and put the input video at the center of the padded area:
15709 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15713 Pad the input to get a squared output with size equal to the maximum
15714 value between the input width and height, and put the input video at
15715 the center of the padded area:
15717 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15721 Pad the input to get a final w/h ratio of 16:9:
15723 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15727 In case of anamorphic video, in order to set the output display aspect
15728 correctly, it is necessary to use @var{sar} in the expression,
15729 according to the relation:
15731 (ih * X / ih) * sar = output_dar
15732 X = output_dar / sar
15735 Thus the previous example needs to be modified to:
15737 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15741 Double the output size and put the input video in the bottom-right
15742 corner of the output padded area:
15744 pad="2*iw:2*ih:ow-iw:oh-ih"
15748 @anchor{palettegen}
15749 @section palettegen
15751 Generate one palette for a whole video stream.
15753 It accepts the following options:
15757 Set the maximum number of colors to quantize in the palette.
15758 Note: the palette will still contain 256 colors; the unused palette entries
15761 @item reserve_transparent
15762 Create a palette of 255 colors maximum and reserve the last one for
15763 transparency. Reserving the transparency color is useful for GIF optimization.
15764 If not set, the maximum of colors in the palette will be 256. You probably want
15765 to disable this option for a standalone image.
15768 @item transparency_color
15769 Set the color that will be used as background for transparency.
15772 Set statistics mode.
15774 It accepts the following values:
15777 Compute full frame histograms.
15779 Compute histograms only for the part that differs from previous frame. This
15780 might be relevant to give more importance to the moving part of your input if
15781 the background is static.
15783 Compute new histogram for each frame.
15786 Default value is @var{full}.
15789 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15790 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15791 color quantization of the palette. This information is also visible at
15792 @var{info} logging level.
15794 @subsection Examples
15798 Generate a representative palette of a given video using @command{ffmpeg}:
15800 ffmpeg -i input.mkv -vf palettegen palette.png
15804 @section paletteuse
15806 Use a palette to downsample an input video stream.
15808 The filter takes two inputs: one video stream and a palette. The palette must
15809 be a 256 pixels image.
15811 It accepts the following options:
15815 Select dithering mode. Available algorithms are:
15818 Ordered 8x8 bayer dithering (deterministic)
15820 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
15821 Note: this dithering is sometimes considered "wrong" and is included as a
15823 @item floyd_steinberg
15824 Floyd and Steingberg dithering (error diffusion)
15826 Frankie Sierra dithering v2 (error diffusion)
15828 Frankie Sierra dithering v2 "Lite" (error diffusion)
15831 Default is @var{sierra2_4a}.
15834 When @var{bayer} dithering is selected, this option defines the scale of the
15835 pattern (how much the crosshatch pattern is visible). A low value means more
15836 visible pattern for less banding, and higher value means less visible pattern
15837 at the cost of more banding.
15839 The option must be an integer value in the range [0,5]. Default is @var{2}.
15842 If set, define the zone to process
15846 Only the changing rectangle will be reprocessed. This is similar to GIF
15847 cropping/offsetting compression mechanism. This option can be useful for speed
15848 if only a part of the image is changing, and has use cases such as limiting the
15849 scope of the error diffusal @option{dither} to the rectangle that bounds the
15850 moving scene (it leads to more deterministic output if the scene doesn't change
15851 much, and as a result less moving noise and better GIF compression).
15854 Default is @var{none}.
15857 Take new palette for each output frame.
15859 @item alpha_threshold
15860 Sets the alpha threshold for transparency. Alpha values above this threshold
15861 will be treated as completely opaque, and values below this threshold will be
15862 treated as completely transparent.
15864 The option must be an integer value in the range [0,255]. Default is @var{128}.
15867 @subsection Examples
15871 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
15872 using @command{ffmpeg}:
15874 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
15878 @section perspective
15880 Correct perspective of video not recorded perpendicular to the screen.
15882 A description of the accepted parameters follows.
15893 Set coordinates expression for top left, top right, bottom left and bottom right corners.
15894 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
15895 If the @code{sense} option is set to @code{source}, then the specified points will be sent
15896 to the corners of the destination. If the @code{sense} option is set to @code{destination},
15897 then the corners of the source will be sent to the specified coordinates.
15899 The expressions can use the following variables:
15904 the width and height of video frame.
15908 Output frame count.
15911 @item interpolation
15912 Set interpolation for perspective correction.
15914 It accepts the following values:
15920 Default value is @samp{linear}.
15923 Set interpretation of coordinate options.
15925 It accepts the following values:
15929 Send point in the source specified by the given coordinates to
15930 the corners of the destination.
15932 @item 1, destination
15934 Send the corners of the source to the point in the destination specified
15935 by the given coordinates.
15937 Default value is @samp{source}.
15941 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
15943 It accepts the following values:
15946 only evaluate expressions once during the filter initialization or
15947 when a command is processed
15950 evaluate expressions for each incoming frame
15953 Default value is @samp{init}.
15958 Delay interlaced video by one field time so that the field order changes.
15960 The intended use is to fix PAL movies that have been captured with the
15961 opposite field order to the film-to-video transfer.
15963 A description of the accepted parameters follows.
15969 It accepts the following values:
15972 Capture field order top-first, transfer bottom-first.
15973 Filter will delay the bottom field.
15976 Capture field order bottom-first, transfer top-first.
15977 Filter will delay the top field.
15980 Capture and transfer with the same field order. This mode only exists
15981 for the documentation of the other options to refer to, but if you
15982 actually select it, the filter will faithfully do nothing.
15985 Capture field order determined automatically by field flags, transfer
15987 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
15988 basis using field flags. If no field information is available,
15989 then this works just like @samp{u}.
15992 Capture unknown or varying, transfer opposite.
15993 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
15994 analyzing the images and selecting the alternative that produces best
15995 match between the fields.
15998 Capture top-first, transfer unknown or varying.
15999 Filter selects among @samp{t} and @samp{p} using image analysis.
16002 Capture bottom-first, transfer unknown or varying.
16003 Filter selects among @samp{b} and @samp{p} using image analysis.
16006 Capture determined by field flags, transfer unknown or varying.
16007 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
16008 image analysis. If no field information is available, then this works just
16009 like @samp{U}. This is the default mode.
16012 Both capture and transfer unknown or varying.
16013 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
16017 @subsection Commands
16019 This filter supports the all above options as @ref{commands}.
16021 @section photosensitivity
16022 Reduce various flashes in video, so to help users with epilepsy.
16024 It accepts the following options:
16027 Set how many frames to use when filtering. Default is 30.
16030 Set detection threshold factor. Default is 1.
16034 Set how many pixels to skip when sampling frames. Default is 1.
16035 Allowed range is from 1 to 1024.
16038 Leave frames unchanged. Default is disabled.
16041 @section pixdesctest
16043 Pixel format descriptor test filter, mainly useful for internal
16044 testing. The output video should be equal to the input video.
16048 format=monow, pixdesctest
16051 can be used to test the monowhite pixel format descriptor definition.
16055 Display sample values of color channels. Mainly useful for checking color
16056 and levels. Minimum supported resolution is 640x480.
16058 The filters accept the following options:
16062 Set scope X position, relative offset on X axis.
16065 Set scope Y position, relative offset on Y axis.
16074 Set window opacity. This window also holds statistics about pixel area.
16077 Set window X position, relative offset on X axis.
16080 Set window Y position, relative offset on Y axis.
16085 Enable the specified chain of postprocessing subfilters using libpostproc. This
16086 library should be automatically selected with a GPL build (@code{--enable-gpl}).
16087 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
16088 Each subfilter and some options have a short and a long name that can be used
16089 interchangeably, i.e. dr/dering are the same.
16091 The filters accept the following options:
16095 Set postprocessing subfilters string.
16098 All subfilters share common options to determine their scope:
16102 Honor the quality commands for this subfilter.
16105 Do chrominance filtering, too (default).
16108 Do luminance filtering only (no chrominance).
16111 Do chrominance filtering only (no luminance).
16114 These options can be appended after the subfilter name, separated by a '|'.
16116 Available subfilters are:
16119 @item hb/hdeblock[|difference[|flatness]]
16120 Horizontal deblocking filter
16123 Difference factor where higher values mean more deblocking (default: @code{32}).
16125 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16128 @item vb/vdeblock[|difference[|flatness]]
16129 Vertical deblocking filter
16132 Difference factor where higher values mean more deblocking (default: @code{32}).
16134 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16137 @item ha/hadeblock[|difference[|flatness]]
16138 Accurate horizontal deblocking filter
16141 Difference factor where higher values mean more deblocking (default: @code{32}).
16143 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16146 @item va/vadeblock[|difference[|flatness]]
16147 Accurate vertical deblocking filter
16150 Difference factor where higher values mean more deblocking (default: @code{32}).
16152 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16156 The horizontal and vertical deblocking filters share the difference and
16157 flatness values so you cannot set different horizontal and vertical
16161 @item h1/x1hdeblock
16162 Experimental horizontal deblocking filter
16164 @item v1/x1vdeblock
16165 Experimental vertical deblocking filter
16170 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16173 larger -> stronger filtering
16175 larger -> stronger filtering
16177 larger -> stronger filtering
16180 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16183 Stretch luminance to @code{0-255}.
16186 @item lb/linblenddeint
16187 Linear blend deinterlacing filter that deinterlaces the given block by
16188 filtering all lines with a @code{(1 2 1)} filter.
16190 @item li/linipoldeint
16191 Linear interpolating deinterlacing filter that deinterlaces the given block by
16192 linearly interpolating every second line.
16194 @item ci/cubicipoldeint
16195 Cubic interpolating deinterlacing filter deinterlaces the given block by
16196 cubically interpolating every second line.
16198 @item md/mediandeint
16199 Median deinterlacing filter that deinterlaces the given block by applying a
16200 median filter to every second line.
16202 @item fd/ffmpegdeint
16203 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16204 second line with a @code{(-1 4 2 4 -1)} filter.
16207 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16208 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16210 @item fq/forceQuant[|quantizer]
16211 Overrides the quantizer table from the input with the constant quantizer you
16219 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16222 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16225 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16228 @subsection Examples
16232 Apply horizontal and vertical deblocking, deringing and automatic
16233 brightness/contrast:
16239 Apply default filters without brightness/contrast correction:
16245 Apply default filters and temporal denoiser:
16247 pp=default/tmpnoise|1|2|3
16251 Apply deblocking on luminance only, and switch vertical deblocking on or off
16252 automatically depending on available CPU time:
16259 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16260 similar to spp = 6 with 7 point DCT, where only the center sample is
16263 The filter accepts the following options:
16267 Force a constant quantization parameter. It accepts an integer in range
16268 0 to 63. If not set, the filter will use the QP from the video stream
16272 Set thresholding mode. Available modes are:
16276 Set hard thresholding.
16278 Set soft thresholding (better de-ringing effect, but likely blurrier).
16280 Set medium thresholding (good results, default).
16284 @section premultiply
16285 Apply alpha premultiply effect to input video stream using first plane
16286 of second stream as alpha.
16288 Both streams must have same dimensions and same pixel format.
16290 The filter accepts the following option:
16294 Set which planes will be processed, unprocessed planes will be copied.
16295 By default value 0xf, all planes will be processed.
16298 Do not require 2nd input for processing, instead use alpha plane from input stream.
16302 Apply prewitt operator to input video stream.
16304 The filter accepts the following option:
16308 Set which planes will be processed, unprocessed planes will be copied.
16309 By default value 0xf, all planes will be processed.
16312 Set value which will be multiplied with filtered result.
16315 Set value which will be added to filtered result.
16318 @subsection Commands
16320 This filter supports the all above options as @ref{commands}.
16322 @section pseudocolor
16324 Alter frame colors in video with pseudocolors.
16326 This filter accepts the following options:
16330 set pixel first component expression
16333 set pixel second component expression
16336 set pixel third component expression
16339 set pixel fourth component expression, corresponds to the alpha component
16342 set component to use as base for altering colors
16345 Pick one of built-in LUTs. By default is set to none.
16361 Each of them specifies the expression to use for computing the lookup table for
16362 the corresponding pixel component values.
16364 The expressions can contain the following constants and functions:
16369 The input width and height.
16372 The input value for the pixel component.
16374 @item ymin, umin, vmin, amin
16375 The minimum allowed component value.
16377 @item ymax, umax, vmax, amax
16378 The maximum allowed component value.
16381 All expressions default to "val".
16383 @subsection Commands
16385 This filter supports the all above options as @ref{commands}.
16387 @subsection Examples
16391 Change too high luma values to gradient:
16393 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'"
16399 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16400 Ratio) between two input videos.
16402 This filter takes in input two input videos, the first input is
16403 considered the "main" source and is passed unchanged to the
16404 output. The second input is used as a "reference" video for computing
16407 Both video inputs must have the same resolution and pixel format for
16408 this filter to work correctly. Also it assumes that both inputs
16409 have the same number of frames, which are compared one by one.
16411 The obtained average PSNR is printed through the logging system.
16413 The filter stores the accumulated MSE (mean squared error) of each
16414 frame, and at the end of the processing it is averaged across all frames
16415 equally, and the following formula is applied to obtain the PSNR:
16418 PSNR = 10*log10(MAX^2/MSE)
16421 Where MAX is the average of the maximum values of each component of the
16424 The description of the accepted parameters follows.
16427 @item stats_file, f
16428 If specified the filter will use the named file to save the PSNR of
16429 each individual frame. When filename equals "-" the data is sent to
16432 @item stats_version
16433 Specifies which version of the stats file format to use. Details of
16434 each format are written below.
16435 Default value is 1.
16437 @item stats_add_max
16438 Determines whether the max value is output to the stats log.
16439 Default value is 0.
16440 Requires stats_version >= 2. If this is set and stats_version < 2,
16441 the filter will return an error.
16444 This filter also supports the @ref{framesync} options.
16446 The file printed if @var{stats_file} is selected, contains a sequence of
16447 key/value pairs of the form @var{key}:@var{value} for each compared
16450 If a @var{stats_version} greater than 1 is specified, a header line precedes
16451 the list of per-frame-pair stats, with key value pairs following the frame
16452 format with the following parameters:
16455 @item psnr_log_version
16456 The version of the log file format. Will match @var{stats_version}.
16459 A comma separated list of the per-frame-pair parameters included in
16463 A description of each shown per-frame-pair parameter follows:
16467 sequential number of the input frame, starting from 1
16470 Mean Square Error pixel-by-pixel average difference of the compared
16471 frames, averaged over all the image components.
16473 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16474 Mean Square Error pixel-by-pixel average difference of the compared
16475 frames for the component specified by the suffix.
16477 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16478 Peak Signal to Noise ratio of the compared frames for the component
16479 specified by the suffix.
16481 @item max_avg, max_y, max_u, max_v
16482 Maximum allowed value for each channel, and average over all
16486 @subsection Examples
16491 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16492 [main][ref] psnr="stats_file=stats.log" [out]
16495 On this example the input file being processed is compared with the
16496 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16497 is stored in @file{stats.log}.
16500 Another example with different containers:
16502 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 -
16509 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16510 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16513 The pullup filter is designed to take advantage of future context in making
16514 its decisions. This filter is stateless in the sense that it does not lock
16515 onto a pattern to follow, but it instead looks forward to the following
16516 fields in order to identify matches and rebuild progressive frames.
16518 To produce content with an even framerate, insert the fps filter after
16519 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16520 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16522 The filter accepts the following options:
16529 These options set the amount of "junk" to ignore at the left, right, top, and
16530 bottom of the image, respectively. Left and right are in units of 8 pixels,
16531 while top and bottom are in units of 2 lines.
16532 The default is 8 pixels on each side.
16535 Set the strict breaks. Setting this option to 1 will reduce the chances of
16536 filter generating an occasional mismatched frame, but it may also cause an
16537 excessive number of frames to be dropped during high motion sequences.
16538 Conversely, setting it to -1 will make filter match fields more easily.
16539 This may help processing of video where there is slight blurring between
16540 the fields, but may also cause there to be interlaced frames in the output.
16541 Default value is @code{0}.
16544 Set the metric plane to use. It accepts the following values:
16550 Use chroma blue plane.
16553 Use chroma red plane.
16556 This option may be set to use chroma plane instead of the default luma plane
16557 for doing filter's computations. This may improve accuracy on very clean
16558 source material, but more likely will decrease accuracy, especially if there
16559 is chroma noise (rainbow effect) or any grayscale video.
16560 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16561 load and make pullup usable in realtime on slow machines.
16564 For best results (without duplicated frames in the output file) it is
16565 necessary to change the output frame rate. For example, to inverse
16566 telecine NTSC input:
16568 ffmpeg -i input -vf pullup -r 24000/1001 ...
16573 Change video quantization parameters (QP).
16575 The filter accepts the following option:
16579 Set expression for quantization parameter.
16582 The expression is evaluated through the eval API and can contain, among others,
16583 the following constants:
16587 1 if index is not 129, 0 otherwise.
16590 Sequential index starting from -129 to 128.
16593 @subsection Examples
16597 Some equation like:
16605 Flush video frames from internal cache of frames into a random order.
16606 No frame is discarded.
16607 Inspired by @ref{frei0r} nervous filter.
16611 Set size in number of frames of internal cache, in range from @code{2} to
16612 @code{512}. Default is @code{30}.
16615 Set seed for random number generator, must be an integer included between
16616 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16617 less than @code{0}, the filter will try to use a good random seed on a
16621 @section readeia608
16623 Read closed captioning (EIA-608) information from the top lines of a video frame.
16625 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16626 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16627 with EIA-608 data (starting from 0). A description of each metadata value follows:
16630 @item lavfi.readeia608.X.cc
16631 The two bytes stored as EIA-608 data (printed in hexadecimal).
16633 @item lavfi.readeia608.X.line
16634 The number of the line on which the EIA-608 data was identified and read.
16637 This filter accepts the following options:
16641 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16644 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16647 Set the ratio of width reserved for sync code detection.
16648 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16651 Enable checking the parity bit. In the event of a parity error, the filter will output
16652 @code{0x00} for that character. Default is false.
16655 Lowpass lines prior to further processing. Default is enabled.
16658 @subsection Commands
16660 This filter supports the all above options as @ref{commands}.
16662 @subsection Examples
16666 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16668 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
16674 Read vertical interval timecode (VITC) information from the top lines of a
16677 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16678 timecode value, if a valid timecode has been detected. Further metadata key
16679 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16680 timecode data has been found or not.
16682 This filter accepts the following options:
16686 Set the maximum number of lines to scan for VITC data. If the value is set to
16687 @code{-1} the full video frame is scanned. Default is @code{45}.
16690 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16691 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16694 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16695 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16698 @subsection Examples
16702 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16703 draw @code{--:--:--:--} as a placeholder:
16705 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16711 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16713 Destination pixel at position (X, Y) will be picked from source (x, y) position
16714 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16715 value for pixel will be used for destination pixel.
16717 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16718 will have Xmap/Ymap video stream dimensions.
16719 Xmap and Ymap input video streams are 16bit depth, single channel.
16723 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16724 Default is @code{color}.
16727 Specify the color of the unmapped pixels. For the syntax of this option,
16728 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16729 manual,ffmpeg-utils}. Default color is @code{black}.
16732 @section removegrain
16734 The removegrain filter is a spatial denoiser for progressive video.
16738 Set mode for the first plane.
16741 Set mode for the second plane.
16744 Set mode for the third plane.
16747 Set mode for the fourth plane.
16750 Range of mode is from 0 to 24. Description of each mode follows:
16754 Leave input plane unchanged. Default.
16757 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16760 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16763 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16766 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16767 This is equivalent to a median filter.
16770 Line-sensitive clipping giving the minimal change.
16773 Line-sensitive clipping, intermediate.
16776 Line-sensitive clipping, intermediate.
16779 Line-sensitive clipping, intermediate.
16782 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16785 Replaces the target pixel with the closest neighbour.
16788 [1 2 1] horizontal and vertical kernel blur.
16794 Bob mode, interpolates top field from the line where the neighbours
16795 pixels are the closest.
16798 Bob mode, interpolates bottom field from the line where the neighbours
16799 pixels are the closest.
16802 Bob mode, interpolates top field. Same as 13 but with a more complicated
16803 interpolation formula.
16806 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16807 interpolation formula.
16810 Clips the pixel with the minimum and maximum of respectively the maximum and
16811 minimum of each pair of opposite neighbour pixels.
16814 Line-sensitive clipping using opposite neighbours whose greatest distance from
16815 the current pixel is minimal.
16818 Replaces the pixel with the average of its 8 neighbours.
16821 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
16824 Clips pixels using the averages of opposite neighbour.
16827 Same as mode 21 but simpler and faster.
16830 Small edge and halo removal, but reputed useless.
16836 @section removelogo
16838 Suppress a TV station logo, using an image file to determine which
16839 pixels comprise the logo. It works by filling in the pixels that
16840 comprise the logo with neighboring pixels.
16842 The filter accepts the following options:
16846 Set the filter bitmap file, which can be any image format supported by
16847 libavformat. The width and height of the image file must match those of the
16848 video stream being processed.
16851 Pixels in the provided bitmap image with a value of zero are not
16852 considered part of the logo, non-zero pixels are considered part of
16853 the logo. If you use white (255) for the logo and black (0) for the
16854 rest, you will be safe. For making the filter bitmap, it is
16855 recommended to take a screen capture of a black frame with the logo
16856 visible, and then using a threshold filter followed by the erode
16857 filter once or twice.
16859 If needed, little splotches can be fixed manually. Remember that if
16860 logo pixels are not covered, the filter quality will be much
16861 reduced. Marking too many pixels as part of the logo does not hurt as
16862 much, but it will increase the amount of blurring needed to cover over
16863 the image and will destroy more information than necessary, and extra
16864 pixels will slow things down on a large logo.
16866 @section repeatfields
16868 This filter uses the repeat_field flag from the Video ES headers and hard repeats
16869 fields based on its value.
16873 Reverse a video clip.
16875 Warning: This filter requires memory to buffer the entire clip, so trimming
16878 @subsection Examples
16882 Take the first 5 seconds of a clip, and reverse it.
16889 Shift R/G/B/A pixels horizontally and/or vertically.
16891 The filter accepts the following options:
16894 Set amount to shift red horizontally.
16896 Set amount to shift red vertically.
16898 Set amount to shift green horizontally.
16900 Set amount to shift green vertically.
16902 Set amount to shift blue horizontally.
16904 Set amount to shift blue vertically.
16906 Set amount to shift alpha horizontally.
16908 Set amount to shift alpha vertically.
16910 Set edge mode, can be @var{smear}, default, or @var{warp}.
16913 @subsection Commands
16915 This filter supports the all above options as @ref{commands}.
16918 Apply roberts cross operator to input video stream.
16920 The filter accepts the following option:
16924 Set which planes will be processed, unprocessed planes will be copied.
16925 By default value 0xf, all planes will be processed.
16928 Set value which will be multiplied with filtered result.
16931 Set value which will be added to filtered result.
16934 @subsection Commands
16936 This filter supports the all above options as @ref{commands}.
16940 Rotate video by an arbitrary angle expressed in radians.
16942 The filter accepts the following options:
16944 A description of the optional parameters follows.
16947 Set an expression for the angle by which to rotate the input video
16948 clockwise, expressed as a number of radians. A negative value will
16949 result in a counter-clockwise rotation. By default it is set to "0".
16951 This expression is evaluated for each frame.
16954 Set the output width expression, default value is "iw".
16955 This expression is evaluated just once during configuration.
16958 Set the output height expression, default value is "ih".
16959 This expression is evaluated just once during configuration.
16962 Enable bilinear interpolation if set to 1, a value of 0 disables
16963 it. Default value is 1.
16966 Set the color used to fill the output area not covered by the rotated
16967 image. For the general syntax of this option, check the
16968 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16969 If the special value "none" is selected then no
16970 background is printed (useful for example if the background is never shown).
16972 Default value is "black".
16975 The expressions for the angle and the output size can contain the
16976 following constants and functions:
16980 sequential number of the input frame, starting from 0. It is always NAN
16981 before the first frame is filtered.
16984 time in seconds of the input frame, it is set to 0 when the filter is
16985 configured. It is always NAN before the first frame is filtered.
16989 horizontal and vertical chroma subsample values. For example for the
16990 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16994 the input video width and height
16998 the output width and height, that is the size of the padded area as
16999 specified by the @var{width} and @var{height} expressions
17003 the minimal width/height required for completely containing the input
17004 video rotated by @var{a} radians.
17006 These are only available when computing the @option{out_w} and
17007 @option{out_h} expressions.
17010 @subsection Examples
17014 Rotate the input by PI/6 radians clockwise:
17020 Rotate the input by PI/6 radians counter-clockwise:
17026 Rotate the input by 45 degrees clockwise:
17032 Apply a constant rotation with period T, starting from an angle of PI/3:
17034 rotate=PI/3+2*PI*t/T
17038 Make the input video rotation oscillating with a period of T
17039 seconds and an amplitude of A radians:
17041 rotate=A*sin(2*PI/T*t)
17045 Rotate the video, output size is chosen so that the whole rotating
17046 input video is always completely contained in the output:
17048 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
17052 Rotate the video, reduce the output size so that no background is ever
17055 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
17059 @subsection Commands
17061 The filter supports the following commands:
17065 Set the angle expression.
17066 The command accepts the same syntax of the corresponding option.
17068 If the specified expression is not valid, it is kept at its current
17074 Apply Shape Adaptive Blur.
17076 The filter accepts the following options:
17079 @item luma_radius, lr
17080 Set luma blur filter strength, must be a value in range 0.1-4.0, default
17081 value is 1.0. A greater value will result in a more blurred image, and
17082 in slower processing.
17084 @item luma_pre_filter_radius, lpfr
17085 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
17088 @item luma_strength, ls
17089 Set luma maximum difference between pixels to still be considered, must
17090 be a value in the 0.1-100.0 range, default value is 1.0.
17092 @item chroma_radius, cr
17093 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17094 greater value will result in a more blurred image, and in slower
17097 @item chroma_pre_filter_radius, cpfr
17098 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17100 @item chroma_strength, cs
17101 Set chroma maximum difference between pixels to still be considered,
17102 must be a value in the -0.9-100.0 range.
17105 Each chroma option value, if not explicitly specified, is set to the
17106 corresponding luma option value.
17111 Scale (resize) the input video, using the libswscale library.
17113 The scale filter forces the output display aspect ratio to be the same
17114 of the input, by changing the output sample aspect ratio.
17116 If the input image format is different from the format requested by
17117 the next filter, the scale filter will convert the input to the
17120 @subsection Options
17121 The filter accepts the following options, or any of the options
17122 supported by the libswscale scaler.
17124 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17125 the complete list of scaler options.
17130 Set the output video dimension expression. Default value is the input
17133 If the @var{width} or @var{w} value is 0, the input width is used for
17134 the output. If the @var{height} or @var{h} value is 0, the input height
17135 is used for the output.
17137 If one and only one of the values is -n with n >= 1, the scale filter
17138 will use a value that maintains the aspect ratio of the input image,
17139 calculated from the other specified dimension. After that it will,
17140 however, make sure that the calculated dimension is divisible by n and
17141 adjust the value if necessary.
17143 If both values are -n with n >= 1, the behavior will be identical to
17144 both values being set to 0 as previously detailed.
17146 See below for the list of accepted constants for use in the dimension
17150 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17154 Only evaluate expressions once during the filter initialization or when a command is processed.
17157 Evaluate expressions for each incoming frame.
17161 Default value is @samp{init}.
17165 Set the interlacing mode. It accepts the following values:
17169 Force interlaced aware scaling.
17172 Do not apply interlaced scaling.
17175 Select interlaced aware scaling depending on whether the source frames
17176 are flagged as interlaced or not.
17179 Default value is @samp{0}.
17182 Set libswscale scaling flags. See
17183 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17184 complete list of values. If not explicitly specified the filter applies
17188 @item param0, param1
17189 Set libswscale input parameters for scaling algorithms that need them. See
17190 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17191 complete documentation. If not explicitly specified the filter applies
17197 Set the video size. For the syntax of this option, check the
17198 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17200 @item in_color_matrix
17201 @item out_color_matrix
17202 Set in/output YCbCr color space type.
17204 This allows the autodetected value to be overridden as well as allows forcing
17205 a specific value used for the output and encoder.
17207 If not specified, the color space type depends on the pixel format.
17213 Choose automatically.
17216 Format conforming to International Telecommunication Union (ITU)
17217 Recommendation BT.709.
17220 Set color space conforming to the United States Federal Communications
17221 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17226 Set color space conforming to:
17230 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17233 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17236 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17241 Set color space conforming to SMPTE ST 240:1999.
17244 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17249 Set in/output YCbCr sample range.
17251 This allows the autodetected value to be overridden as well as allows forcing
17252 a specific value used for the output and encoder. If not specified, the
17253 range depends on the pixel format. Possible values:
17257 Choose automatically.
17260 Set full range (0-255 in case of 8-bit luma).
17262 @item mpeg/limited/tv
17263 Set "MPEG" range (16-235 in case of 8-bit luma).
17266 @item force_original_aspect_ratio
17267 Enable decreasing or increasing output video width or height if necessary to
17268 keep the original aspect ratio. Possible values:
17272 Scale the video as specified and disable this feature.
17275 The output video dimensions will automatically be decreased if needed.
17278 The output video dimensions will automatically be increased if needed.
17282 One useful instance of this option is that when you know a specific device's
17283 maximum allowed resolution, you can use this to limit the output video to
17284 that, while retaining the aspect ratio. For example, device A allows
17285 1280x720 playback, and your video is 1920x800. Using this option (set it to
17286 decrease) and specifying 1280x720 to the command line makes the output
17289 Please note that this is a different thing than specifying -1 for @option{w}
17290 or @option{h}, you still need to specify the output resolution for this option
17293 @item force_divisible_by
17294 Ensures that both the output dimensions, width and height, are divisible by the
17295 given integer when used together with @option{force_original_aspect_ratio}. This
17296 works similar to using @code{-n} in the @option{w} and @option{h} options.
17298 This option respects the value set for @option{force_original_aspect_ratio},
17299 increasing or decreasing the resolution accordingly. The video's aspect ratio
17300 may be slightly modified.
17302 This option can be handy if you need to have a video fit within or exceed
17303 a defined resolution using @option{force_original_aspect_ratio} but also have
17304 encoder restrictions on width or height divisibility.
17308 The values of the @option{w} and @option{h} options are expressions
17309 containing the following constants:
17314 The input width and height
17318 These are the same as @var{in_w} and @var{in_h}.
17322 The output (scaled) width and height
17326 These are the same as @var{out_w} and @var{out_h}
17329 The same as @var{iw} / @var{ih}
17332 input sample aspect ratio
17335 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17339 horizontal and vertical input chroma subsample values. For example for the
17340 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17344 horizontal and vertical output chroma subsample values. For example for the
17345 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17348 The (sequential) number of the input frame, starting from 0.
17349 Only available with @code{eval=frame}.
17352 The presentation timestamp of the input frame, expressed as a number of
17353 seconds. Only available with @code{eval=frame}.
17356 The position (byte offset) of the frame in the input stream, or NaN if
17357 this information is unavailable and/or meaningless (for example in case of synthetic video).
17358 Only available with @code{eval=frame}.
17361 @subsection Examples
17365 Scale the input video to a size of 200x100
17370 This is equivalent to:
17381 Specify a size abbreviation for the output size:
17386 which can also be written as:
17392 Scale the input to 2x:
17394 scale=w=2*iw:h=2*ih
17398 The above is the same as:
17400 scale=2*in_w:2*in_h
17404 Scale the input to 2x with forced interlaced scaling:
17406 scale=2*iw:2*ih:interl=1
17410 Scale the input to half size:
17412 scale=w=iw/2:h=ih/2
17416 Increase the width, and set the height to the same size:
17422 Seek Greek harmony:
17429 Increase the height, and set the width to 3/2 of the height:
17431 scale=w=3/2*oh:h=3/5*ih
17435 Increase the size, making the size a multiple of the chroma
17438 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17442 Increase the width to a maximum of 500 pixels,
17443 keeping the same aspect ratio as the input:
17445 scale=w='min(500\, iw*3/2):h=-1'
17449 Make pixels square by combining scale and setsar:
17451 scale='trunc(ih*dar):ih',setsar=1/1
17455 Make pixels square by combining scale and setsar,
17456 making sure the resulting resolution is even (required by some codecs):
17458 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17462 @subsection Commands
17464 This filter supports the following commands:
17468 Set the output video dimension expression.
17469 The command accepts the same syntax of the corresponding option.
17471 If the specified expression is not valid, it is kept at its current
17477 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17478 format conversion on CUDA video frames. Setting the output width and height
17479 works in the same way as for the @var{scale} filter.
17481 The following additional options are accepted:
17484 The pixel format of the output CUDA frames. If set to the string "same" (the
17485 default), the input format will be kept. Note that automatic format negotiation
17486 and conversion is not yet supported for hardware frames
17489 The interpolation algorithm used for resizing. One of the following:
17496 @item cubic2p_bspline
17497 2-parameter cubic (B=1, C=0)
17499 @item cubic2p_catmullrom
17500 2-parameter cubic (B=0, C=1/2)
17502 @item cubic2p_b05c03
17503 2-parameter cubic (B=1/2, C=3/10)
17511 @item force_original_aspect_ratio
17512 Enable decreasing or increasing output video width or height if necessary to
17513 keep the original aspect ratio. Possible values:
17517 Scale the video as specified and disable this feature.
17520 The output video dimensions will automatically be decreased if needed.
17523 The output video dimensions will automatically be increased if needed.
17527 One useful instance of this option is that when you know a specific device's
17528 maximum allowed resolution, you can use this to limit the output video to
17529 that, while retaining the aspect ratio. For example, device A allows
17530 1280x720 playback, and your video is 1920x800. Using this option (set it to
17531 decrease) and specifying 1280x720 to the command line makes the output
17534 Please note that this is a different thing than specifying -1 for @option{w}
17535 or @option{h}, you still need to specify the output resolution for this option
17538 @item force_divisible_by
17539 Ensures that both the output dimensions, width and height, are divisible by the
17540 given integer when used together with @option{force_original_aspect_ratio}. This
17541 works similar to using @code{-n} in the @option{w} and @option{h} options.
17543 This option respects the value set for @option{force_original_aspect_ratio},
17544 increasing or decreasing the resolution accordingly. The video's aspect ratio
17545 may be slightly modified.
17547 This option can be handy if you need to have a video fit within or exceed
17548 a defined resolution using @option{force_original_aspect_ratio} but also have
17549 encoder restrictions on width or height divisibility.
17555 Scale (resize) the input video, based on a reference video.
17557 See the scale filter for available options, scale2ref supports the same but
17558 uses the reference video instead of the main input as basis. scale2ref also
17559 supports the following additional constants for the @option{w} and
17560 @option{h} options:
17565 The main input video's width and height
17568 The same as @var{main_w} / @var{main_h}
17571 The main input video's sample aspect ratio
17573 @item main_dar, mdar
17574 The main input video's display aspect ratio. Calculated from
17575 @code{(main_w / main_h) * main_sar}.
17579 The main input video's horizontal and vertical chroma subsample values.
17580 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17584 The (sequential) number of the main input frame, starting from 0.
17585 Only available with @code{eval=frame}.
17588 The presentation timestamp of the main input frame, expressed as a number of
17589 seconds. Only available with @code{eval=frame}.
17592 The position (byte offset) of the frame in the main input stream, or NaN if
17593 this information is unavailable and/or meaningless (for example in case of synthetic video).
17594 Only available with @code{eval=frame}.
17597 @subsection Examples
17601 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17603 'scale2ref[b][a];[a][b]overlay'
17607 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17609 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17613 @subsection Commands
17615 This filter supports the following commands:
17619 Set the output video dimension expression.
17620 The command accepts the same syntax of the corresponding option.
17622 If the specified expression is not valid, it is kept at its current
17627 Scroll input video horizontally and/or vertically by constant speed.
17629 The filter accepts the following options:
17631 @item horizontal, h
17632 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17633 Negative values changes scrolling direction.
17636 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17637 Negative values changes scrolling direction.
17640 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17643 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17646 @subsection Commands
17648 This filter supports the following @ref{commands}:
17650 @item horizontal, h
17651 Set the horizontal scrolling speed.
17653 Set the vertical scrolling speed.
17659 Detect video scene change.
17661 This filter sets frame metadata with mafd between frame, the scene score, and
17662 forward the frame to the next filter, so they can use these metadata to detect
17663 scene change or others.
17665 In addition, this filter logs a message and sets frame metadata when it detects
17666 a scene change by @option{threshold}.
17668 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17670 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17671 to detect scene change.
17673 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17674 detect scene change with @option{threshold}.
17676 The filter accepts the following options:
17680 Set the scene change detection threshold as a percentage of maximum change. Good
17681 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17684 Default value is @code{10.}.
17687 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17688 You can enable it if you want to get snapshot of scene change frames only.
17691 @anchor{selectivecolor}
17692 @section selectivecolor
17694 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17695 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17696 by the "purity" of the color (that is, how saturated it already is).
17698 This filter is similar to the Adobe Photoshop Selective Color tool.
17700 The filter accepts the following options:
17703 @item correction_method
17704 Select color correction method.
17706 Available values are:
17709 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17712 Specified adjustments are relative to the original component value.
17714 Default is @code{absolute}.
17716 Adjustments for red pixels (pixels where the red component is the maximum)
17718 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17720 Adjustments for green pixels (pixels where the green component is the maximum)
17722 Adjustments for cyan pixels (pixels where the red component is the minimum)
17724 Adjustments for blue pixels (pixels where the blue component is the maximum)
17726 Adjustments for magenta pixels (pixels where the green component is the minimum)
17728 Adjustments for white pixels (pixels where all components are greater than 128)
17730 Adjustments for all pixels except pure black and pure white
17732 Adjustments for black pixels (pixels where all components are lesser than 128)
17734 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17737 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17738 4 space separated floating point adjustment values in the [-1,1] range,
17739 respectively to adjust the amount of cyan, magenta, yellow and black for the
17740 pixels of its range.
17742 @subsection Examples
17746 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17747 increase magenta by 27% in blue areas:
17749 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17753 Use a Photoshop selective color preset:
17755 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17759 @anchor{separatefields}
17760 @section separatefields
17762 The @code{separatefields} takes a frame-based video input and splits
17763 each frame into its components fields, producing a new half height clip
17764 with twice the frame rate and twice the frame count.
17766 This filter use field-dominance information in frame to decide which
17767 of each pair of fields to place first in the output.
17768 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17770 @section setdar, setsar
17772 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17775 This is done by changing the specified Sample (aka Pixel) Aspect
17776 Ratio, according to the following equation:
17778 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17781 Keep in mind that the @code{setdar} filter does not modify the pixel
17782 dimensions of the video frame. Also, the display aspect ratio set by
17783 this filter may be changed by later filters in the filterchain,
17784 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17787 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17788 the filter output video.
17790 Note that as a consequence of the application of this filter, the
17791 output display aspect ratio will change according to the equation
17794 Keep in mind that the sample aspect ratio set by the @code{setsar}
17795 filter may be changed by later filters in the filterchain, e.g. if
17796 another "setsar" or a "setdar" filter is applied.
17798 It accepts the following parameters:
17801 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17802 Set the aspect ratio used by the filter.
17804 The parameter can be a floating point number string, an expression, or
17805 a string of the form @var{num}:@var{den}, where @var{num} and
17806 @var{den} are the numerator and denominator of the aspect ratio. If
17807 the parameter is not specified, it is assumed the value "0".
17808 In case the form "@var{num}:@var{den}" is used, the @code{:} character
17812 Set the maximum integer value to use for expressing numerator and
17813 denominator when reducing the expressed aspect ratio to a rational.
17814 Default value is @code{100}.
17818 The parameter @var{sar} is an expression containing
17819 the following constants:
17823 These are approximated values for the mathematical constants e
17824 (Euler's number), pi (Greek pi), and phi (the golden ratio).
17827 The input width and height.
17830 These are the same as @var{w} / @var{h}.
17833 The input sample aspect ratio.
17836 The input display aspect ratio. It is the same as
17837 (@var{w} / @var{h}) * @var{sar}.
17840 Horizontal and vertical chroma subsample values. For example, for the
17841 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17844 @subsection Examples
17849 To change the display aspect ratio to 16:9, specify one of the following:
17856 To change the sample aspect ratio to 10:11, specify:
17862 To set a display aspect ratio of 16:9, and specify a maximum integer value of
17863 1000 in the aspect ratio reduction, use the command:
17865 setdar=ratio=16/9:max=1000
17873 Force field for the output video frame.
17875 The @code{setfield} filter marks the interlace type field for the
17876 output frames. It does not change the input frame, but only sets the
17877 corresponding property, which affects how the frame is treated by
17878 following filters (e.g. @code{fieldorder} or @code{yadif}).
17880 The filter accepts the following options:
17885 Available values are:
17889 Keep the same field property.
17892 Mark the frame as bottom-field-first.
17895 Mark the frame as top-field-first.
17898 Mark the frame as progressive.
17905 Force frame parameter for the output video frame.
17907 The @code{setparams} filter marks interlace and color range for the
17908 output frames. It does not change the input frame, but only sets the
17909 corresponding property, which affects how the frame is treated by
17914 Available values are:
17918 Keep the same field property (default).
17921 Mark the frame as bottom-field-first.
17924 Mark the frame as top-field-first.
17927 Mark the frame as progressive.
17931 Available values are:
17935 Keep the same color range property (default).
17937 @item unspecified, unknown
17938 Mark the frame as unspecified color range.
17940 @item limited, tv, mpeg
17941 Mark the frame as limited range.
17943 @item full, pc, jpeg
17944 Mark the frame as full range.
17947 @item color_primaries
17948 Set the color primaries.
17949 Available values are:
17953 Keep the same color primaries property (default).
17970 Set the color transfer.
17971 Available values are:
17975 Keep the same color trc property (default).
17997 Set the colorspace.
17998 Available values are:
18002 Keep the same colorspace property (default).
18015 @item chroma-derived-nc
18016 @item chroma-derived-c
18022 Apply shear transform to input video.
18024 This filter supports the following options:
18028 Shear factor in X-direction. Default value is 0.
18029 Allowed range is from -2 to 2.
18032 Shear factor in Y-direction. Default value is 0.
18033 Allowed range is from -2 to 2.
18036 Set the color used to fill the output area not covered by the transformed
18037 video. For the general syntax of this option, check the
18038 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18039 If the special value "none" is selected then no
18040 background is printed (useful for example if the background is never shown).
18042 Default value is "black".
18045 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
18048 @subsection Commands
18050 This filter supports the all above options as @ref{commands}.
18054 Show a line containing various information for each input video frame.
18055 The input video is not modified.
18057 This filter supports the following options:
18061 Calculate checksums of each plane. By default enabled.
18064 The shown line contains a sequence of key/value pairs of the form
18065 @var{key}:@var{value}.
18067 The following values are shown in the output:
18071 The (sequential) number of the input frame, starting from 0.
18074 The Presentation TimeStamp of the input frame, expressed as a number of
18075 time base units. The time base unit depends on the filter input pad.
18078 The Presentation TimeStamp of the input frame, expressed as a number of
18082 The position of the frame in the input stream, or -1 if this information is
18083 unavailable and/or meaningless (for example in case of synthetic video).
18086 The pixel format name.
18089 The sample aspect ratio of the input frame, expressed in the form
18090 @var{num}/@var{den}.
18093 The size of the input frame. For the syntax of this option, check the
18094 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18097 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18098 for bottom field first).
18101 This is 1 if the frame is a key frame, 0 otherwise.
18104 The picture type of the input frame ("I" for an I-frame, "P" for a
18105 P-frame, "B" for a B-frame, or "?" for an unknown type).
18106 Also refer to the documentation of the @code{AVPictureType} enum and of
18107 the @code{av_get_picture_type_char} function defined in
18108 @file{libavutil/avutil.h}.
18111 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18113 @item plane_checksum
18114 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18115 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18118 The mean value of pixels in each plane of the input frame, expressed in the form
18119 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18122 The standard deviation of pixel values in each plane of the input frame, expressed
18123 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18127 @section showpalette
18129 Displays the 256 colors palette of each frame. This filter is only relevant for
18130 @var{pal8} pixel format frames.
18132 It accepts the following option:
18136 Set the size of the box used to represent one palette color entry. Default is
18137 @code{30} (for a @code{30x30} pixel box).
18140 @section shuffleframes
18142 Reorder and/or duplicate and/or drop video frames.
18144 It accepts the following parameters:
18148 Set the destination indexes of input frames.
18149 This is space or '|' separated list of indexes that maps input frames to output
18150 frames. Number of indexes also sets maximal value that each index may have.
18151 '-1' index have special meaning and that is to drop frame.
18154 The first frame has the index 0. The default is to keep the input unchanged.
18156 @subsection Examples
18160 Swap second and third frame of every three frames of the input:
18162 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18166 Swap 10th and 1st frame of every ten frames of the input:
18168 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18172 @section shufflepixels
18174 Reorder pixels in video frames.
18176 This filter accepts the following options:
18180 Set shuffle direction. Can be forward or inverse direction.
18181 Default direction is forward.
18184 Set shuffle mode. Can be horizontal, vertical or block mode.
18188 Set shuffle block_size. In case of horizontal shuffle mode only width
18189 part of size is used, and in case of vertical shuffle mode only height
18190 part of size is used.
18193 Set random seed used with shuffling pixels. Mainly useful to set to be able
18194 to reverse filtering process to get original input.
18195 For example, to reverse forward shuffle you need to use same parameters
18196 and exact same seed and to set direction to inverse.
18199 @section shuffleplanes
18201 Reorder and/or duplicate video planes.
18203 It accepts the following parameters:
18208 The index of the input plane to be used as the first output plane.
18211 The index of the input plane to be used as the second output plane.
18214 The index of the input plane to be used as the third output plane.
18217 The index of the input plane to be used as the fourth output plane.
18221 The first plane has the index 0. The default is to keep the input unchanged.
18223 @subsection Examples
18227 Swap the second and third planes of the input:
18229 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18233 @anchor{signalstats}
18234 @section signalstats
18235 Evaluate various visual metrics that assist in determining issues associated
18236 with the digitization of analog video media.
18238 By default the filter will log these metadata values:
18242 Display the minimal Y value contained within the input frame. Expressed in
18246 Display the Y value at the 10% percentile within the input frame. Expressed in
18250 Display the average Y value within the input frame. Expressed in range of
18254 Display the Y value at the 90% percentile within the input frame. Expressed in
18258 Display the maximum Y value contained within the input frame. Expressed in
18262 Display the minimal U value contained within the input frame. Expressed in
18266 Display the U value at the 10% percentile within the input frame. Expressed in
18270 Display the average U value within the input frame. Expressed in range of
18274 Display the U value at the 90% percentile within the input frame. Expressed in
18278 Display the maximum U value contained within the input frame. Expressed in
18282 Display the minimal V value contained within the input frame. Expressed in
18286 Display the V value at the 10% percentile within the input frame. Expressed in
18290 Display the average V value within the input frame. Expressed in range of
18294 Display the V value at the 90% percentile within the input frame. Expressed in
18298 Display the maximum V value contained within the input frame. Expressed in
18302 Display the minimal saturation value contained within the input frame.
18303 Expressed in range of [0-~181.02].
18306 Display the saturation value at the 10% percentile within the input frame.
18307 Expressed in range of [0-~181.02].
18310 Display the average saturation value within the input frame. Expressed in range
18314 Display the saturation value at the 90% percentile within the input frame.
18315 Expressed in range of [0-~181.02].
18318 Display the maximum saturation value contained within the input frame.
18319 Expressed in range of [0-~181.02].
18322 Display the median value for hue within the input frame. Expressed in range of
18326 Display the average value for hue within the input frame. Expressed in range of
18330 Display the average of sample value difference between all values of the Y
18331 plane in the current frame and corresponding values of the previous input frame.
18332 Expressed in range of [0-255].
18335 Display the average of sample value difference between all values of the U
18336 plane in the current frame and corresponding values of the previous input frame.
18337 Expressed in range of [0-255].
18340 Display the average of sample value difference between all values of the V
18341 plane in the current frame and corresponding values of the previous input frame.
18342 Expressed in range of [0-255].
18345 Display bit depth of Y plane in current frame.
18346 Expressed in range of [0-16].
18349 Display bit depth of U plane in current frame.
18350 Expressed in range of [0-16].
18353 Display bit depth of V plane in current frame.
18354 Expressed in range of [0-16].
18357 The filter accepts the following options:
18363 @option{stat} specify an additional form of image analysis.
18364 @option{out} output video with the specified type of pixel highlighted.
18366 Both options accept the following values:
18370 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18371 unlike the neighboring pixels of the same field. Examples of temporal outliers
18372 include the results of video dropouts, head clogs, or tape tracking issues.
18375 Identify @var{vertical line repetition}. Vertical line repetition includes
18376 similar rows of pixels within a frame. In born-digital video vertical line
18377 repetition is common, but this pattern is uncommon in video digitized from an
18378 analog source. When it occurs in video that results from the digitization of an
18379 analog source it can indicate concealment from a dropout compensator.
18382 Identify pixels that fall outside of legal broadcast range.
18386 Set the highlight color for the @option{out} option. The default color is
18390 @subsection Examples
18394 Output data of various video metrics:
18396 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18400 Output specific data about the minimum and maximum values of the Y plane per frame:
18402 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18406 Playback video while highlighting pixels that are outside of broadcast range in red.
18408 ffplay example.mov -vf signalstats="out=brng:color=red"
18412 Playback video with signalstats metadata drawn over the frame.
18414 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18417 The contents of signalstat_drawtext.txt used in the command are:
18420 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18421 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18422 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18423 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18431 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18432 input. In this case the matching between the inputs can be calculated additionally.
18433 The filter always passes through the first input. The signature of each stream can
18434 be written into a file.
18436 It accepts the following options:
18440 Enable or disable the matching process.
18442 Available values are:
18446 Disable the calculation of a matching (default).
18448 Calculate the matching for the whole video and output whether the whole video
18449 matches or only parts.
18451 Calculate only until a matching is found or the video ends. Should be faster in
18456 Set the number of inputs. The option value must be a non negative integer.
18457 Default value is 1.
18460 Set the path to which the output is written. If there is more than one input,
18461 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18462 integer), that will be replaced with the input number. If no filename is
18463 specified, no output will be written. This is the default.
18466 Choose the output format.
18468 Available values are:
18472 Use the specified binary representation (default).
18474 Use the specified xml representation.
18478 Set threshold to detect one word as similar. The option value must be an integer
18479 greater than zero. The default value is 9000.
18482 Set threshold to detect all words as similar. The option value must be an integer
18483 greater than zero. The default value is 60000.
18486 Set threshold to detect frames as similar. The option value must be an integer
18487 greater than zero. The default value is 116.
18490 Set the minimum length of a sequence in frames to recognize it as matching
18491 sequence. The option value must be a non negative integer value.
18492 The default value is 0.
18495 Set the minimum relation, that matching frames to all frames must have.
18496 The option value must be a double value between 0 and 1. The default value is 0.5.
18499 @subsection Examples
18503 To calculate the signature of an input video and store it in signature.bin:
18505 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18509 To detect whether two videos match and store the signatures in XML format in
18510 signature0.xml and signature1.xml:
18512 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 -
18520 Blur the input video without impacting the outlines.
18522 It accepts the following options:
18525 @item luma_radius, lr
18526 Set the luma radius. The option value must be a float number in
18527 the range [0.1,5.0] that specifies the variance of the gaussian filter
18528 used to blur the image (slower if larger). Default value is 1.0.
18530 @item luma_strength, ls
18531 Set the luma strength. The option value must be a float number
18532 in the range [-1.0,1.0] that configures the blurring. A value included
18533 in [0.0,1.0] will blur the image whereas a value included in
18534 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18536 @item luma_threshold, lt
18537 Set the luma threshold used as a coefficient to determine
18538 whether a pixel should be blurred or not. The option value must be an
18539 integer in the range [-30,30]. A value of 0 will filter all the image,
18540 a value included in [0,30] will filter flat areas and a value included
18541 in [-30,0] will filter edges. Default value is 0.
18543 @item chroma_radius, cr
18544 Set the chroma radius. The option value must be a float number in
18545 the range [0.1,5.0] that specifies the variance of the gaussian filter
18546 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18548 @item chroma_strength, cs
18549 Set the chroma strength. The option value must be a float number
18550 in the range [-1.0,1.0] that configures the blurring. A value included
18551 in [0.0,1.0] will blur the image whereas a value included in
18552 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18554 @item chroma_threshold, ct
18555 Set the chroma threshold used as a coefficient to determine
18556 whether a pixel should be blurred or not. The option value must be an
18557 integer in the range [-30,30]. A value of 0 will filter all the image,
18558 a value included in [0,30] will filter flat areas and a value included
18559 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18562 If a chroma option is not explicitly set, the corresponding luma value
18566 Apply sobel operator to input video stream.
18568 The filter accepts the following option:
18572 Set which planes will be processed, unprocessed planes will be copied.
18573 By default value 0xf, all planes will be processed.
18576 Set value which will be multiplied with filtered result.
18579 Set value which will be added to filtered result.
18582 @subsection Commands
18584 This filter supports the all above options as @ref{commands}.
18589 Apply a simple postprocessing filter that compresses and decompresses the image
18590 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18591 and average the results.
18593 The filter accepts the following options:
18597 Set quality. This option defines the number of levels for averaging. It accepts
18598 an integer in the range 0-6. If set to @code{0}, the filter will have no
18599 effect. A value of @code{6} means the higher quality. For each increment of
18600 that value the speed drops by a factor of approximately 2. Default value is
18604 Force a constant quantization parameter. If not set, the filter will use the QP
18605 from the video stream (if available).
18608 Set thresholding mode. Available modes are:
18612 Set hard thresholding (default).
18614 Set soft thresholding (better de-ringing effect, but likely blurrier).
18617 @item use_bframe_qp
18618 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18619 option may cause flicker since the B-Frames have often larger QP. Default is
18620 @code{0} (not enabled).
18623 @subsection Commands
18625 This filter supports the following commands:
18627 @item quality, level
18628 Set quality level. The value @code{max} can be used to set the maximum level,
18629 currently @code{6}.
18635 Scale the input by applying one of the super-resolution methods based on
18636 convolutional neural networks. Supported models:
18640 Super-Resolution Convolutional Neural Network model (SRCNN).
18641 See @url{https://arxiv.org/abs/1501.00092}.
18644 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18645 See @url{https://arxiv.org/abs/1609.05158}.
18648 Training scripts as well as scripts for model file (.pb) saving can be found at
18649 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18650 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18652 Native model files (.model) can be generated from TensorFlow model
18653 files (.pb) by using tools/python/convert.py
18655 The filter accepts the following options:
18659 Specify which DNN backend to use for model loading and execution. This option accepts
18660 the following values:
18664 Native implementation of DNN loading and execution.
18667 TensorFlow backend. To enable this backend you
18668 need to install the TensorFlow for C library (see
18669 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18670 @code{--enable-libtensorflow}
18673 Default value is @samp{native}.
18676 Set path to model file specifying network architecture and its parameters.
18677 Note that different backends use different file formats. TensorFlow backend
18678 can load files for both formats, while native backend can load files for only
18682 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18683 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18684 input upscaled using bicubic upscaling with proper scale factor.
18687 This feature can also be finished with @ref{dnn_processing} filter.
18691 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18693 This filter takes in input two input videos, the first input is
18694 considered the "main" source and is passed unchanged to the
18695 output. The second input is used as a "reference" video for computing
18698 Both video inputs must have the same resolution and pixel format for
18699 this filter to work correctly. Also it assumes that both inputs
18700 have the same number of frames, which are compared one by one.
18702 The filter stores the calculated SSIM of each frame.
18704 The description of the accepted parameters follows.
18707 @item stats_file, f
18708 If specified the filter will use the named file to save the SSIM of
18709 each individual frame. When filename equals "-" the data is sent to
18713 The file printed if @var{stats_file} is selected, contains a sequence of
18714 key/value pairs of the form @var{key}:@var{value} for each compared
18717 A description of each shown parameter follows:
18721 sequential number of the input frame, starting from 1
18723 @item Y, U, V, R, G, B
18724 SSIM of the compared frames for the component specified by the suffix.
18727 SSIM of the compared frames for the whole frame.
18730 Same as above but in dB representation.
18733 This filter also supports the @ref{framesync} options.
18735 @subsection Examples
18740 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18741 [main][ref] ssim="stats_file=stats.log" [out]
18744 On this example the input file being processed is compared with the
18745 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18746 is stored in @file{stats.log}.
18749 Another example with both psnr and ssim at same time:
18751 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18755 Another example with different containers:
18757 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 -
18763 Convert between different stereoscopic image formats.
18765 The filters accept the following options:
18769 Set stereoscopic image format of input.
18771 Available values for input image formats are:
18774 side by side parallel (left eye left, right eye right)
18777 side by side crosseye (right eye left, left eye right)
18780 side by side parallel with half width resolution
18781 (left eye left, right eye right)
18784 side by side crosseye with half width resolution
18785 (right eye left, left eye right)
18789 above-below (left eye above, right eye below)
18793 above-below (right eye above, left eye below)
18797 above-below with half height resolution
18798 (left eye above, right eye below)
18802 above-below with half height resolution
18803 (right eye above, left eye below)
18806 alternating frames (left eye first, right eye second)
18809 alternating frames (right eye first, left eye second)
18812 interleaved rows (left eye has top row, right eye starts on next row)
18815 interleaved rows (right eye has top row, left eye starts on next row)
18818 interleaved columns, left eye first
18821 interleaved columns, right eye first
18823 Default value is @samp{sbsl}.
18827 Set stereoscopic image format of output.
18831 side by side parallel (left eye left, right eye right)
18834 side by side crosseye (right eye left, left eye right)
18837 side by side parallel with half width resolution
18838 (left eye left, right eye right)
18841 side by side crosseye with half width resolution
18842 (right eye left, left eye right)
18846 above-below (left eye above, right eye below)
18850 above-below (right eye above, left eye below)
18854 above-below with half height resolution
18855 (left eye above, right eye below)
18859 above-below with half height resolution
18860 (right eye above, left eye below)
18863 alternating frames (left eye first, right eye second)
18866 alternating frames (right eye first, left eye second)
18869 interleaved rows (left eye has top row, right eye starts on next row)
18872 interleaved rows (right eye has top row, left eye starts on next row)
18875 anaglyph red/blue gray
18876 (red filter on left eye, blue filter on right eye)
18879 anaglyph red/green gray
18880 (red filter on left eye, green filter on right eye)
18883 anaglyph red/cyan gray
18884 (red filter on left eye, cyan filter on right eye)
18887 anaglyph red/cyan half colored
18888 (red filter on left eye, cyan filter on right eye)
18891 anaglyph red/cyan color
18892 (red filter on left eye, cyan filter on right eye)
18895 anaglyph red/cyan color optimized with the least squares projection of dubois
18896 (red filter on left eye, cyan filter on right eye)
18899 anaglyph green/magenta gray
18900 (green filter on left eye, magenta filter on right eye)
18903 anaglyph green/magenta half colored
18904 (green filter on left eye, magenta filter on right eye)
18907 anaglyph green/magenta colored
18908 (green filter on left eye, magenta filter on right eye)
18911 anaglyph green/magenta color optimized with the least squares projection of dubois
18912 (green filter on left eye, magenta filter on right eye)
18915 anaglyph yellow/blue gray
18916 (yellow filter on left eye, blue filter on right eye)
18919 anaglyph yellow/blue half colored
18920 (yellow filter on left eye, blue filter on right eye)
18923 anaglyph yellow/blue colored
18924 (yellow filter on left eye, blue filter on right eye)
18927 anaglyph yellow/blue color optimized with the least squares projection of dubois
18928 (yellow filter on left eye, blue filter on right eye)
18931 mono output (left eye only)
18934 mono output (right eye only)
18937 checkerboard, left eye first
18940 checkerboard, right eye first
18943 interleaved columns, left eye first
18946 interleaved columns, right eye first
18952 Default value is @samp{arcd}.
18955 @subsection Examples
18959 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
18965 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
18971 @section streamselect, astreamselect
18972 Select video or audio streams.
18974 The filter accepts the following options:
18978 Set number of inputs. Default is 2.
18981 Set input indexes to remap to outputs.
18984 @subsection Commands
18986 The @code{streamselect} and @code{astreamselect} filter supports the following
18991 Set input indexes to remap to outputs.
18994 @subsection Examples
18998 Select first 5 seconds 1st stream and rest of time 2nd stream:
19000 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
19004 Same as above, but for audio:
19006 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
19013 Draw subtitles on top of input video using the libass library.
19015 To enable compilation of this filter you need to configure FFmpeg with
19016 @code{--enable-libass}. This filter also requires a build with libavcodec and
19017 libavformat to convert the passed subtitles file to ASS (Advanced Substation
19018 Alpha) subtitles format.
19020 The filter accepts the following options:
19024 Set the filename of the subtitle file to read. It must be specified.
19026 @item original_size
19027 Specify the size of the original video, the video for which the ASS file
19028 was composed. For the syntax of this option, check the
19029 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19030 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
19031 correctly scale the fonts if the aspect ratio has been changed.
19034 Set a directory path containing fonts that can be used by the filter.
19035 These fonts will be used in addition to whatever the font provider uses.
19038 Process alpha channel, by default alpha channel is untouched.
19041 Set subtitles input character encoding. @code{subtitles} filter only. Only
19042 useful if not UTF-8.
19044 @item stream_index, si
19045 Set subtitles stream index. @code{subtitles} filter only.
19048 Override default style or script info parameters of the subtitles. It accepts a
19049 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
19052 If the first key is not specified, it is assumed that the first value
19053 specifies the @option{filename}.
19055 For example, to render the file @file{sub.srt} on top of the input
19056 video, use the command:
19061 which is equivalent to:
19063 subtitles=filename=sub.srt
19066 To render the default subtitles stream from file @file{video.mkv}, use:
19068 subtitles=video.mkv
19071 To render the second subtitles stream from that file, use:
19073 subtitles=video.mkv:si=1
19076 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
19077 @code{DejaVu Serif}, use:
19079 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
19082 @section super2xsai
19084 Scale the input by 2x and smooth using the Super2xSaI (Scale and
19085 Interpolate) pixel art scaling algorithm.
19087 Useful for enlarging pixel art images without reducing sharpness.
19091 Swap two rectangular objects in video.
19093 This filter accepts the following options:
19103 Set 1st rect x coordinate.
19106 Set 1st rect y coordinate.
19109 Set 2nd rect x coordinate.
19112 Set 2nd rect y coordinate.
19114 All expressions are evaluated once for each frame.
19117 The all options are expressions containing the following constants:
19122 The input width and height.
19125 same as @var{w} / @var{h}
19128 input sample aspect ratio
19131 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19134 The number of the input frame, starting from 0.
19137 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19140 the position in the file of the input frame, NAN if unknown
19147 Blend successive video frames.
19153 Apply telecine process to the video.
19155 This filter accepts the following options:
19164 The default value is @code{top}.
19168 A string of numbers representing the pulldown pattern you wish to apply.
19169 The default value is @code{23}.
19173 Some typical patterns:
19178 24p: 2332 (preferred)
19185 24p: 222222222223 ("Euro pulldown")
19190 @section thistogram
19192 Compute and draw a color distribution histogram for the input video across time.
19194 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19195 at certain time, this filter shows also past histograms of number of frames defined
19196 by @code{width} option.
19198 The computed histogram is a representation of the color component
19199 distribution in an image.
19201 The filter accepts the following options:
19205 Set width of single color component output. Default value is @code{0}.
19206 Value of @code{0} means width will be picked from input video.
19207 This also set number of passed histograms to keep.
19208 Allowed range is [0, 8192].
19210 @item display_mode, d
19212 It accepts the following values:
19215 Per color component graphs are placed below each other.
19218 Per color component graphs are placed side by side.
19221 Presents information identical to that in the @code{parade}, except
19222 that the graphs representing color components are superimposed directly
19225 Default is @code{stack}.
19227 @item levels_mode, m
19228 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19229 Default is @code{linear}.
19231 @item components, c
19232 Set what color components to display.
19233 Default is @code{7}.
19236 Set background opacity. Default is @code{0.9}.
19239 Show envelope. Default is disabled.
19242 Set envelope color. Default is @code{gold}.
19247 Available values for slide is:
19250 Draw new frame when right border is reached.
19253 Replace old columns with new ones.
19256 Scroll from right to left.
19259 Scroll from left to right.
19262 Draw single picture.
19265 Default is @code{replace}.
19270 Apply threshold effect to video stream.
19272 This filter needs four video streams to perform thresholding.
19273 First stream is stream we are filtering.
19274 Second stream is holding threshold values, third stream is holding min values,
19275 and last, fourth stream is holding max values.
19277 The filter accepts the following option:
19281 Set which planes will be processed, unprocessed planes will be copied.
19282 By default value 0xf, all planes will be processed.
19285 For example if first stream pixel's component value is less then threshold value
19286 of pixel component from 2nd threshold stream, third stream value will picked,
19287 otherwise fourth stream pixel component value will be picked.
19289 Using color source filter one can perform various types of thresholding:
19291 @subsection Examples
19295 Binary threshold, using gray color as threshold:
19297 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19301 Inverted binary threshold, using gray color as threshold:
19303 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19307 Truncate binary threshold, using gray color as threshold:
19309 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19313 Threshold to zero, using gray color as threshold:
19315 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19319 Inverted threshold to zero, using gray color as threshold:
19321 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19326 Select the most representative frame in a given sequence of consecutive frames.
19328 The filter accepts the following options:
19332 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19333 will pick one of them, and then handle the next batch of @var{n} frames until
19334 the end. Default is @code{100}.
19337 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19338 value will result in a higher memory usage, so a high value is not recommended.
19340 @subsection Examples
19344 Extract one picture each 50 frames:
19350 Complete example of a thumbnail creation with @command{ffmpeg}:
19352 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19359 Tile several successive frames together.
19361 The @ref{untile} filter can do the reverse.
19363 The filter accepts the following options:
19368 Set the grid size (i.e. the number of lines and columns). For the syntax of
19369 this option, check the
19370 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19373 Set the maximum number of frames to render in the given area. It must be less
19374 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19375 the area will be used.
19378 Set the outer border margin in pixels.
19381 Set the inner border thickness (i.e. the number of pixels between frames). For
19382 more advanced padding options (such as having different values for the edges),
19383 refer to the pad video filter.
19386 Specify the color of the unused area. For the syntax of this option, check the
19387 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19388 The default value of @var{color} is "black".
19391 Set the number of frames to overlap when tiling several successive frames together.
19392 The value must be between @code{0} and @var{nb_frames - 1}.
19395 Set the number of frames to initially be empty before displaying first output frame.
19396 This controls how soon will one get first output frame.
19397 The value must be between @code{0} and @var{nb_frames - 1}.
19400 @subsection Examples
19404 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19406 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19408 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19409 duplicating each output frame to accommodate the originally detected frame
19413 Display @code{5} pictures in an area of @code{3x2} frames,
19414 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19415 mixed flat and named options:
19417 tile=3x2:nb_frames=5:padding=7:margin=2
19421 @section tinterlace
19423 Perform various types of temporal field interlacing.
19425 Frames are counted starting from 1, so the first input frame is
19428 The filter accepts the following options:
19433 Specify the mode of the interlacing. This option can also be specified
19434 as a value alone. See below for a list of values for this option.
19436 Available values are:
19440 Move odd frames into the upper field, even into the lower field,
19441 generating a double height frame at half frame rate.
19445 Frame 1 Frame 2 Frame 3 Frame 4
19447 11111 22222 33333 44444
19448 11111 22222 33333 44444
19449 11111 22222 33333 44444
19450 11111 22222 33333 44444
19464 Only output odd frames, even frames are dropped, generating a frame with
19465 unchanged height at half frame rate.
19470 Frame 1 Frame 2 Frame 3 Frame 4
19472 11111 22222 33333 44444
19473 11111 22222 33333 44444
19474 11111 22222 33333 44444
19475 11111 22222 33333 44444
19485 Only output even frames, odd frames are dropped, generating a frame with
19486 unchanged height at half frame rate.
19491 Frame 1 Frame 2 Frame 3 Frame 4
19493 11111 22222 33333 44444
19494 11111 22222 33333 44444
19495 11111 22222 33333 44444
19496 11111 22222 33333 44444
19506 Expand each frame to full height, but pad alternate lines with black,
19507 generating a frame with double height at the same input frame rate.
19512 Frame 1 Frame 2 Frame 3 Frame 4
19514 11111 22222 33333 44444
19515 11111 22222 33333 44444
19516 11111 22222 33333 44444
19517 11111 22222 33333 44444
19520 11111 ..... 33333 .....
19521 ..... 22222 ..... 44444
19522 11111 ..... 33333 .....
19523 ..... 22222 ..... 44444
19524 11111 ..... 33333 .....
19525 ..... 22222 ..... 44444
19526 11111 ..... 33333 .....
19527 ..... 22222 ..... 44444
19531 @item interleave_top, 4
19532 Interleave the upper field from odd frames with the lower field from
19533 even frames, generating a frame with unchanged height at half frame rate.
19538 Frame 1 Frame 2 Frame 3 Frame 4
19540 11111<- 22222 33333<- 44444
19541 11111 22222<- 33333 44444<-
19542 11111<- 22222 33333<- 44444
19543 11111 22222<- 33333 44444<-
19553 @item interleave_bottom, 5
19554 Interleave the lower field from odd frames with the upper field from
19555 even frames, generating a frame with unchanged height at half frame rate.
19560 Frame 1 Frame 2 Frame 3 Frame 4
19562 11111 22222<- 33333 44444<-
19563 11111<- 22222 33333<- 44444
19564 11111 22222<- 33333 44444<-
19565 11111<- 22222 33333<- 44444
19575 @item interlacex2, 6
19576 Double frame rate with unchanged height. Frames are inserted each
19577 containing the second temporal field from the previous input frame and
19578 the first temporal field from the next input frame. This mode relies on
19579 the top_field_first flag. Useful for interlaced video displays with no
19580 field synchronisation.
19585 Frame 1 Frame 2 Frame 3 Frame 4
19587 11111 22222 33333 44444
19588 11111 22222 33333 44444
19589 11111 22222 33333 44444
19590 11111 22222 33333 44444
19593 11111 22222 22222 33333 33333 44444 44444
19594 11111 11111 22222 22222 33333 33333 44444
19595 11111 22222 22222 33333 33333 44444 44444
19596 11111 11111 22222 22222 33333 33333 44444
19601 Move odd frames into the upper field, even into the lower field,
19602 generating a double height frame at same frame rate.
19607 Frame 1 Frame 2 Frame 3 Frame 4
19609 11111 22222 33333 44444
19610 11111 22222 33333 44444
19611 11111 22222 33333 44444
19612 11111 22222 33333 44444
19615 11111 33333 33333 55555
19616 22222 22222 44444 44444
19617 11111 33333 33333 55555
19618 22222 22222 44444 44444
19619 11111 33333 33333 55555
19620 22222 22222 44444 44444
19621 11111 33333 33333 55555
19622 22222 22222 44444 44444
19627 Numeric values are deprecated but are accepted for backward
19628 compatibility reasons.
19630 Default mode is @code{merge}.
19633 Specify flags influencing the filter process.
19635 Available value for @var{flags} is:
19638 @item low_pass_filter, vlpf
19639 Enable linear vertical low-pass filtering in the filter.
19640 Vertical low-pass filtering is required when creating an interlaced
19641 destination from a progressive source which contains high-frequency
19642 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19645 @item complex_filter, cvlpf
19646 Enable complex vertical low-pass filtering.
19647 This will slightly less reduce interlace 'twitter' and Moire
19648 patterning but better retain detail and subjective sharpness impression.
19651 Bypass already interlaced frames, only adjust the frame rate.
19654 Vertical low-pass filtering and bypassing already interlaced frames can only be
19655 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19660 Pick median pixels from several successive input video frames.
19662 The filter accepts the following options:
19666 Set radius of median filter.
19667 Default is 1. Allowed range is from 1 to 127.
19670 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19673 Set median percentile. Default value is @code{0.5}.
19674 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19675 minimum values, and @code{1} maximum values.
19678 @subsection Commands
19680 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19682 @section tmidequalizer
19684 Apply Temporal Midway Video Equalization effect.
19686 Midway Video Equalization adjusts a sequence of video frames to have the same
19687 histograms, while maintaining their dynamics as much as possible. It's
19688 useful for e.g. matching exposures from a video frames sequence.
19690 This filter accepts the following option:
19694 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19697 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19698 Setting this option to 0 effectively does nothing.
19701 Set which planes to process. Default is @code{15}, which is all available planes.
19706 Mix successive video frames.
19708 A description of the accepted options follows.
19712 The number of successive frames to mix. If unspecified, it defaults to 3.
19715 Specify weight of each input video frame.
19716 Each weight is separated by space. If number of weights is smaller than
19717 number of @var{frames} last specified weight will be used for all remaining
19721 Specify scale, if it is set it will be multiplied with sum
19722 of each weight multiplied with pixel values to give final destination
19723 pixel value. By default @var{scale} is auto scaled to sum of weights.
19726 @subsection Examples
19730 Average 7 successive frames:
19732 tmix=frames=7:weights="1 1 1 1 1 1 1"
19736 Apply simple temporal convolution:
19738 tmix=frames=3:weights="-1 3 -1"
19742 Similar as above but only showing temporal differences:
19744 tmix=frames=3:weights="-1 2 -1":scale=1
19748 @subsection Commands
19750 This filter supports the following commands:
19754 Syntax is same as option with same name.
19759 Tone map colors from different dynamic ranges.
19761 This filter expects data in single precision floating point, as it needs to
19762 operate on (and can output) out-of-range values. Another filter, such as
19763 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19765 The tonemapping algorithms implemented only work on linear light, so input
19766 data should be linearized beforehand (and possibly correctly tagged).
19769 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19772 @subsection Options
19773 The filter accepts the following options.
19777 Set the tone map algorithm to use.
19779 Possible values are:
19782 Do not apply any tone map, only desaturate overbright pixels.
19785 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19786 in-range values, while distorting out-of-range values.
19789 Stretch the entire reference gamut to a linear multiple of the display.
19792 Fit a logarithmic transfer between the tone curves.
19795 Preserve overall image brightness with a simple curve, using nonlinear
19796 contrast, which results in flattening details and degrading color accuracy.
19799 Preserve both dark and bright details better than @var{reinhard}, at the cost
19800 of slightly darkening everything. Use it when detail preservation is more
19801 important than color and brightness accuracy.
19804 Smoothly map out-of-range values, while retaining contrast and colors for
19805 in-range material as much as possible. Use it when color accuracy is more
19806 important than detail preservation.
19812 Tune the tone mapping algorithm.
19814 This affects the following algorithms:
19820 Specifies the scale factor to use while stretching.
19824 Specifies the exponent of the function.
19828 Specify an extra linear coefficient to multiply into the signal before clipping.
19832 Specify the local contrast coefficient at the display peak.
19833 Default to 0.5, which means that in-gamut values will be about half as bright
19840 Specify the transition point from linear to mobius transform. Every value
19841 below this point is guaranteed to be mapped 1:1. The higher the value, the
19842 more accurate the result will be, at the cost of losing bright details.
19843 Default to 0.3, which due to the steep initial slope still preserves in-range
19844 colors fairly accurately.
19848 Apply desaturation for highlights that exceed this level of brightness. The
19849 higher the parameter, the more color information will be preserved. This
19850 setting helps prevent unnaturally blown-out colors for super-highlights, by
19851 (smoothly) turning into white instead. This makes images feel more natural,
19852 at the cost of reducing information about out-of-range colors.
19854 The default of 2.0 is somewhat conservative and will mostly just apply to
19855 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
19857 This option works only if the input frame has a supported color tag.
19860 Override signal/nominal/reference peak with this value. Useful when the
19861 embedded peak information in display metadata is not reliable or when tone
19862 mapping from a lower range to a higher range.
19867 Temporarily pad video frames.
19869 The filter accepts the following options:
19873 Specify number of delay frames before input video stream. Default is 0.
19876 Specify number of padding frames after input video stream.
19877 Set to -1 to pad indefinitely. Default is 0.
19880 Set kind of frames added to beginning of stream.
19881 Can be either @var{add} or @var{clone}.
19882 With @var{add} frames of solid-color are added.
19883 With @var{clone} frames are clones of first frame.
19884 Default is @var{add}.
19887 Set kind of frames added to end of stream.
19888 Can be either @var{add} or @var{clone}.
19889 With @var{add} frames of solid-color are added.
19890 With @var{clone} frames are clones of last frame.
19891 Default is @var{add}.
19893 @item start_duration, stop_duration
19894 Specify the duration of the start/stop delay. See
19895 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19896 for the accepted syntax.
19897 These options override @var{start} and @var{stop}. Default is 0.
19900 Specify the color of the padded area. For the syntax of this option,
19901 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
19902 manual,ffmpeg-utils}.
19904 The default value of @var{color} is "black".
19910 Transpose rows with columns in the input video and optionally flip it.
19912 It accepts the following parameters:
19917 Specify the transposition direction.
19919 Can assume the following values:
19921 @item 0, 4, cclock_flip
19922 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
19930 Rotate by 90 degrees clockwise, that is:
19938 Rotate by 90 degrees counterclockwise, that is:
19945 @item 3, 7, clock_flip
19946 Rotate by 90 degrees clockwise and vertically flip, that is:
19954 For values between 4-7, the transposition is only done if the input
19955 video geometry is portrait and not landscape. These values are
19956 deprecated, the @code{passthrough} option should be used instead.
19958 Numerical values are deprecated, and should be dropped in favor of
19959 symbolic constants.
19962 Do not apply the transposition if the input geometry matches the one
19963 specified by the specified value. It accepts the following values:
19966 Always apply transposition.
19968 Preserve portrait geometry (when @var{height} >= @var{width}).
19970 Preserve landscape geometry (when @var{width} >= @var{height}).
19973 Default value is @code{none}.
19976 For example to rotate by 90 degrees clockwise and preserve portrait
19979 transpose=dir=1:passthrough=portrait
19982 The command above can also be specified as:
19984 transpose=1:portrait
19987 @section transpose_npp
19989 Transpose rows with columns in the input video and optionally flip it.
19990 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
19992 It accepts the following parameters:
19997 Specify the transposition direction.
19999 Can assume the following values:
20002 Rotate by 90 degrees counterclockwise and vertically flip. (default)
20005 Rotate by 90 degrees clockwise.
20008 Rotate by 90 degrees counterclockwise.
20011 Rotate by 90 degrees clockwise and vertically flip.
20015 Do not apply the transposition if the input geometry matches the one
20016 specified by the specified value. It accepts the following values:
20019 Always apply transposition. (default)
20021 Preserve portrait geometry (when @var{height} >= @var{width}).
20023 Preserve landscape geometry (when @var{width} >= @var{height}).
20029 Trim the input so that the output contains one continuous subpart of the input.
20031 It accepts the following parameters:
20034 Specify the time of the start of the kept section, i.e. the frame with the
20035 timestamp @var{start} will be the first frame in the output.
20038 Specify the time of the first frame that will be dropped, i.e. the frame
20039 immediately preceding the one with the timestamp @var{end} will be the last
20040 frame in the output.
20043 This is the same as @var{start}, except this option sets the start timestamp
20044 in timebase units instead of seconds.
20047 This is the same as @var{end}, except this option sets the end timestamp
20048 in timebase units instead of seconds.
20051 The maximum duration of the output in seconds.
20054 The number of the first frame that should be passed to the output.
20057 The number of the first frame that should be dropped.
20060 @option{start}, @option{end}, and @option{duration} are expressed as time
20061 duration specifications; see
20062 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20063 for the accepted syntax.
20065 Note that the first two sets of the start/end options and the @option{duration}
20066 option look at the frame timestamp, while the _frame variants simply count the
20067 frames that pass through the filter. Also note that this filter does not modify
20068 the timestamps. If you wish for the output timestamps to start at zero, insert a
20069 setpts filter after the trim filter.
20071 If multiple start or end options are set, this filter tries to be greedy and
20072 keep all the frames that match at least one of the specified constraints. To keep
20073 only the part that matches all the constraints at once, chain multiple trim
20076 The defaults are such that all the input is kept. So it is possible to set e.g.
20077 just the end values to keep everything before the specified time.
20082 Drop everything except the second minute of input:
20084 ffmpeg -i INPUT -vf trim=60:120
20088 Keep only the first second:
20090 ffmpeg -i INPUT -vf trim=duration=1
20095 @section unpremultiply
20096 Apply alpha unpremultiply effect to input video stream using first plane
20097 of second stream as alpha.
20099 Both streams must have same dimensions and same pixel format.
20101 The filter accepts the following option:
20105 Set which planes will be processed, unprocessed planes will be copied.
20106 By default value 0xf, all planes will be processed.
20108 If the format has 1 or 2 components, then luma is bit 0.
20109 If the format has 3 or 4 components:
20110 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20111 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20112 If present, the alpha channel is always the last bit.
20115 Do not require 2nd input for processing, instead use alpha plane from input stream.
20121 Sharpen or blur the input video.
20123 It accepts the following parameters:
20126 @item luma_msize_x, lx
20127 Set the luma matrix horizontal size. It must be an odd integer between
20128 3 and 23. The default value is 5.
20130 @item luma_msize_y, ly
20131 Set the luma matrix vertical size. It must be an odd integer between 3
20132 and 23. The default value is 5.
20134 @item luma_amount, la
20135 Set the luma effect strength. It must be a floating point number, reasonable
20136 values lay between -1.5 and 1.5.
20138 Negative values will blur the input video, while positive values will
20139 sharpen it, a value of zero will disable the effect.
20141 Default value is 1.0.
20143 @item chroma_msize_x, cx
20144 Set the chroma matrix horizontal size. It must be an odd integer
20145 between 3 and 23. The default value is 5.
20147 @item chroma_msize_y, cy
20148 Set the chroma matrix vertical size. It must be an odd integer
20149 between 3 and 23. The default value is 5.
20151 @item chroma_amount, ca
20152 Set the chroma effect strength. It must be a floating point number, reasonable
20153 values lay between -1.5 and 1.5.
20155 Negative values will blur the input video, while positive values will
20156 sharpen it, a value of zero will disable the effect.
20158 Default value is 0.0.
20162 All parameters are optional and default to the equivalent of the
20163 string '5:5:1.0:5:5:0.0'.
20165 @subsection Examples
20169 Apply strong luma sharpen effect:
20171 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20175 Apply a strong blur of both luma and chroma parameters:
20177 unsharp=7:7:-2:7:7:-2
20184 Decompose a video made of tiled images into the individual images.
20186 The frame rate of the output video is the frame rate of the input video
20187 multiplied by the number of tiles.
20189 This filter does the reverse of @ref{tile}.
20191 The filter accepts the following options:
20196 Set the grid size (i.e. the number of lines and columns). For the syntax of
20197 this option, check the
20198 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20201 @subsection Examples
20205 Produce a 1-second video from a still image file made of 25 frames stacked
20206 vertically, like an analogic film reel:
20208 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20214 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20215 the image at several (or - in the case of @option{quality} level @code{8} - all)
20216 shifts and average the results.
20218 The way this differs from the behavior of spp is that uspp actually encodes &
20219 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20220 DCT similar to MJPEG.
20222 The filter accepts the following options:
20226 Set quality. This option defines the number of levels for averaging. It accepts
20227 an integer in the range 0-8. If set to @code{0}, the filter will have no
20228 effect. A value of @code{8} means the higher quality. For each increment of
20229 that value the speed drops by a factor of approximately 2. Default value is
20233 Force a constant quantization parameter. If not set, the filter will use the QP
20234 from the video stream (if available).
20239 Convert 360 videos between various formats.
20241 The filter accepts the following options:
20247 Set format of the input/output video.
20255 Equirectangular projection.
20260 Cubemap with 3x2/6x1/1x6 layout.
20262 Format specific options:
20267 Set padding proportion for the input/output cubemap. Values in decimals.
20274 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)
20277 Default value is @b{@samp{0}}.
20278 Maximum value is @b{@samp{0.1}}.
20282 Set fixed padding for the input/output cubemap. Values in pixels.
20284 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20288 Set order of faces for the input/output cubemap. Choose one direction for each position.
20290 Designation of directions:
20306 Default value is @b{@samp{rludfb}}.
20310 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20312 Designation of angles:
20315 0 degrees clockwise
20317 90 degrees clockwise
20319 180 degrees clockwise
20321 270 degrees clockwise
20324 Default value is @b{@samp{000000}}.
20328 Equi-Angular Cubemap.
20335 Format specific options:
20340 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20342 If diagonal field of view is set it overrides horizontal and vertical field of view.
20347 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20349 If diagonal field of view is set it overrides horizontal and vertical field of view.
20355 Format specific options:
20360 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20362 If diagonal field of view is set it overrides horizontal and vertical field of view.
20367 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20369 If diagonal field of view is set it overrides horizontal and vertical field of view.
20375 Facebook's 360 formats.
20378 Stereographic format.
20380 Format specific options:
20385 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20387 If diagonal field of view is set it overrides horizontal and vertical field of view.
20392 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20394 If diagonal field of view is set it overrides horizontal and vertical field of view.
20401 Ball format, gives significant distortion toward the back.
20404 Hammer-Aitoff map projection format.
20407 Sinusoidal map projection format.
20410 Fisheye projection.
20412 Format specific options:
20417 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20419 If diagonal field of view is set it overrides horizontal and vertical field of view.
20424 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20426 If diagonal field of view is set it overrides horizontal and vertical field of view.
20430 Pannini projection.
20432 Format specific options:
20435 Set output pannini parameter.
20438 Set input pannini parameter.
20442 Cylindrical projection.
20444 Format specific options:
20449 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20451 If diagonal field of view is set it overrides horizontal and vertical field of view.
20456 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20458 If diagonal field of view is set it overrides horizontal and vertical field of view.
20462 Perspective projection. @i{(output only)}
20464 Format specific options:
20467 Set perspective parameter.
20471 Tetrahedron projection.
20474 Truncated square pyramid projection.
20478 Half equirectangular projection.
20483 Format specific options:
20488 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20490 If diagonal field of view is set it overrides horizontal and vertical field of view.
20495 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20497 If diagonal field of view is set it overrides horizontal and vertical field of view.
20501 Orthographic format.
20503 Format specific options:
20508 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20510 If diagonal field of view is set it overrides horizontal and vertical field of view.
20515 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20517 If diagonal field of view is set it overrides horizontal and vertical field of view.
20521 Octahedron projection.
20525 Set interpolation method.@*
20526 @i{Note: more complex interpolation methods require much more memory to run.}
20536 Bilinear interpolation.
20538 Lagrange9 interpolation.
20541 Bicubic interpolation.
20544 Lanczos interpolation.
20547 Spline16 interpolation.
20550 Gaussian interpolation.
20552 Mitchell interpolation.
20555 Default value is @b{@samp{line}}.
20559 Set the output video resolution.
20561 Default resolution depends on formats.
20565 Set the input/output stereo format.
20576 Default value is @b{@samp{2d}} for input and output format.
20581 Set rotation for the output video. Values in degrees.
20584 Set rotation order for the output video. Choose one item for each position.
20595 Default value is @b{@samp{ypr}}.
20600 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20604 Set if input video is flipped horizontally/vertically. Boolean values.
20607 Set if input video is transposed. Boolean value, by default disabled.
20610 Set if output video needs to be transposed. Boolean value, by default disabled.
20613 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20616 @subsection Examples
20620 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20622 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20625 Extract back view of Equi-Angular Cubemap:
20627 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20630 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20632 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20636 @subsection Commands
20638 This filter supports subset of above options as @ref{commands}.
20640 @section vaguedenoiser
20642 Apply a wavelet based denoiser.
20644 It transforms each frame from the video input into the wavelet domain,
20645 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20646 the obtained coefficients. It does an inverse wavelet transform after.
20647 Due to wavelet properties, it should give a nice smoothed result, and
20648 reduced noise, without blurring picture features.
20650 This filter accepts the following options:
20654 The filtering strength. The higher, the more filtered the video will be.
20655 Hard thresholding can use a higher threshold than soft thresholding
20656 before the video looks overfiltered. Default value is 2.
20659 The filtering method the filter will use.
20661 It accepts the following values:
20664 All values under the threshold will be zeroed.
20667 All values under the threshold will be zeroed. All values above will be
20668 reduced by the threshold.
20671 Scales or nullifies coefficients - intermediary between (more) soft and
20672 (less) hard thresholding.
20675 Default is garrote.
20678 Number of times, the wavelet will decompose the picture. Picture can't
20679 be decomposed beyond a particular point (typically, 8 for a 640x480
20680 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20683 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20686 A list of the planes to process. By default all planes are processed.
20689 The threshold type the filter will use.
20691 It accepts the following values:
20694 Threshold used is same for all decompositions.
20697 Threshold used depends also on each decomposition coefficients.
20700 Default is universal.
20703 @section vectorscope
20705 Display 2 color component values in the two dimensional graph (which is called
20708 This filter accepts the following options:
20712 Set vectorscope mode.
20714 It accepts the following values:
20718 Gray values are displayed on graph, higher brightness means more pixels have
20719 same component color value on location in graph. This is the default mode.
20722 Gray values are displayed on graph. Surrounding pixels values which are not
20723 present in video frame are drawn in gradient of 2 color components which are
20724 set by option @code{x} and @code{y}. The 3rd color component is static.
20727 Actual color components values present in video frame are displayed on graph.
20730 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20731 on graph increases value of another color component, which is luminance by
20732 default values of @code{x} and @code{y}.
20735 Actual colors present in video frame are displayed on graph. If two different
20736 colors map to same position on graph then color with higher value of component
20737 not present in graph is picked.
20740 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20741 component picked from radial gradient.
20745 Set which color component will be represented on X-axis. Default is @code{1}.
20748 Set which color component will be represented on Y-axis. Default is @code{2}.
20751 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20752 of color component which represents frequency of (X, Y) location in graph.
20757 No envelope, this is default.
20760 Instant envelope, even darkest single pixel will be clearly highlighted.
20763 Hold maximum and minimum values presented in graph over time. This way you
20764 can still spot out of range values without constantly looking at vectorscope.
20767 Peak and instant envelope combined together.
20771 Set what kind of graticule to draw.
20780 Set graticule opacity.
20783 Set graticule flags.
20787 Draw graticule for white point.
20790 Draw graticule for black point.
20793 Draw color points short names.
20797 Set background opacity.
20799 @item lthreshold, l
20800 Set low threshold for color component not represented on X or Y axis.
20801 Values lower than this value will be ignored. Default is 0.
20802 Note this value is multiplied with actual max possible value one pixel component
20803 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
20806 @item hthreshold, h
20807 Set high threshold for color component not represented on X or Y axis.
20808 Values higher than this value will be ignored. Default is 1.
20809 Note this value is multiplied with actual max possible value one pixel component
20810 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
20811 is 0.9 * 255 = 230.
20813 @item colorspace, c
20814 Set what kind of colorspace to use when drawing graticule.
20824 Set color tint for gray/tint vectorscope mode. By default both options are zero.
20825 This means no tint, and output will remain gray.
20828 @anchor{vidstabdetect}
20829 @section vidstabdetect
20831 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
20832 @ref{vidstabtransform} for pass 2.
20834 This filter generates a file with relative translation and rotation
20835 transform information about subsequent frames, which is then used by
20836 the @ref{vidstabtransform} filter.
20838 To enable compilation of this filter you need to configure FFmpeg with
20839 @code{--enable-libvidstab}.
20841 This filter accepts the following options:
20845 Set the path to the file used to write the transforms information.
20846 Default value is @file{transforms.trf}.
20849 Set how shaky the video is and how quick the camera is. It accepts an
20850 integer in the range 1-10, a value of 1 means little shakiness, a
20851 value of 10 means strong shakiness. Default value is 5.
20854 Set the accuracy of the detection process. It must be a value in the
20855 range 1-15. A value of 1 means low accuracy, a value of 15 means high
20856 accuracy. Default value is 15.
20859 Set stepsize of the search process. The region around minimum is
20860 scanned with 1 pixel resolution. Default value is 6.
20863 Set minimum contrast. Below this value a local measurement field is
20864 discarded. Must be a floating point value in the range 0-1. Default
20868 Set reference frame number for tripod mode.
20870 If enabled, the motion of the frames is compared to a reference frame
20871 in the filtered stream, identified by the specified number. The idea
20872 is to compensate all movements in a more-or-less static scene and keep
20873 the camera view absolutely still.
20875 If set to 0, it is disabled. The frames are counted starting from 1.
20878 Show fields and transforms in the resulting frames. It accepts an
20879 integer in the range 0-2. Default value is 0, which disables any
20883 @subsection Examples
20887 Use default values:
20893 Analyze strongly shaky movie and put the results in file
20894 @file{mytransforms.trf}:
20896 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
20900 Visualize the result of internal transformations in the resulting
20903 vidstabdetect=show=1
20907 Analyze a video with medium shakiness using @command{ffmpeg}:
20909 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
20913 @anchor{vidstabtransform}
20914 @section vidstabtransform
20916 Video stabilization/deshaking: pass 2 of 2,
20917 see @ref{vidstabdetect} for pass 1.
20919 Read a file with transform information for each frame and
20920 apply/compensate them. Together with the @ref{vidstabdetect}
20921 filter this can be used to deshake videos. See also
20922 @url{http://public.hronopik.de/vid.stab}. It is important to also use
20923 the @ref{unsharp} filter, see below.
20925 To enable compilation of this filter you need to configure FFmpeg with
20926 @code{--enable-libvidstab}.
20928 @subsection Options
20932 Set path to the file used to read the transforms. Default value is
20933 @file{transforms.trf}.
20936 Set the number of frames (value*2 + 1) used for lowpass filtering the
20937 camera movements. Default value is 10.
20939 For example a number of 10 means that 21 frames are used (10 in the
20940 past and 10 in the future) to smoothen the motion in the video. A
20941 larger value leads to a smoother video, but limits the acceleration of
20942 the camera (pan/tilt movements). 0 is a special case where a static
20943 camera is simulated.
20946 Set the camera path optimization algorithm.
20948 Accepted values are:
20951 gaussian kernel low-pass filter on camera motion (default)
20953 averaging on transformations
20957 Set maximal number of pixels to translate frames. Default value is -1,
20961 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
20962 value is -1, meaning no limit.
20965 Specify how to deal with borders that may be visible due to movement
20968 Available values are:
20971 keep image information from previous frame (default)
20973 fill the border black
20977 Invert transforms if set to 1. Default value is 0.
20980 Consider transforms as relative to previous frame if set to 1,
20981 absolute if set to 0. Default value is 0.
20984 Set percentage to zoom. A positive value will result in a zoom-in
20985 effect, a negative value in a zoom-out effect. Default value is 0 (no
20989 Set optimal zooming to avoid borders.
20991 Accepted values are:
20996 optimal static zoom value is determined (only very strong movements
20997 will lead to visible borders) (default)
20999 optimal adaptive zoom value is determined (no borders will be
21000 visible), see @option{zoomspeed}
21003 Note that the value given at zoom is added to the one calculated here.
21006 Set percent to zoom maximally each frame (enabled when
21007 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
21011 Specify type of interpolation.
21013 Available values are:
21018 linear only horizontal
21020 linear in both directions (default)
21022 cubic in both directions (slow)
21026 Enable virtual tripod mode if set to 1, which is equivalent to
21027 @code{relative=0:smoothing=0}. Default value is 0.
21029 Use also @code{tripod} option of @ref{vidstabdetect}.
21032 Increase log verbosity if set to 1. Also the detected global motions
21033 are written to the temporary file @file{global_motions.trf}. Default
21037 @subsection Examples
21041 Use @command{ffmpeg} for a typical stabilization with default values:
21043 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
21046 Note the use of the @ref{unsharp} filter which is always recommended.
21049 Zoom in a bit more and load transform data from a given file:
21051 vidstabtransform=zoom=5:input="mytransforms.trf"
21055 Smoothen the video even more:
21057 vidstabtransform=smoothing=30
21063 Flip the input video vertically.
21065 For example, to vertically flip a video with @command{ffmpeg}:
21067 ffmpeg -i in.avi -vf "vflip" out.avi
21072 Detect variable frame rate video.
21074 This filter tries to detect if the input is variable or constant frame rate.
21076 At end it will output number of frames detected as having variable delta pts,
21077 and ones with constant delta pts.
21078 If there was frames with variable delta, than it will also show min, max and
21079 average delta encountered.
21083 Boost or alter saturation.
21085 The filter accepts the following options:
21088 Set strength of boost if positive value or strength of alter if negative value.
21089 Default is 0. Allowed range is from -2 to 2.
21092 Set the red balance. Default is 1. Allowed range is from -10 to 10.
21095 Set the green balance. Default is 1. Allowed range is from -10 to 10.
21098 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21101 Set the red luma coefficient.
21104 Set the green luma coefficient.
21107 Set the blue luma coefficient.
21110 If @code{intensity} is negative and this is set to 1, colors will change,
21111 otherwise colors will be less saturated, more towards gray.
21114 @subsection Commands
21116 This filter supports the all above options as @ref{commands}.
21121 Make or reverse a natural vignetting effect.
21123 The filter accepts the following options:
21127 Set lens angle expression as a number of radians.
21129 The value is clipped in the @code{[0,PI/2]} range.
21131 Default value: @code{"PI/5"}
21135 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21139 Set forward/backward mode.
21141 Available modes are:
21144 The larger the distance from the central point, the darker the image becomes.
21147 The larger the distance from the central point, the brighter the image becomes.
21148 This can be used to reverse a vignette effect, though there is no automatic
21149 detection to extract the lens @option{angle} and other settings (yet). It can
21150 also be used to create a burning effect.
21153 Default value is @samp{forward}.
21156 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21158 It accepts the following values:
21161 Evaluate expressions only once during the filter initialization.
21164 Evaluate expressions for each incoming frame. This is way slower than the
21165 @samp{init} mode since it requires all the scalers to be re-computed, but it
21166 allows advanced dynamic expressions.
21169 Default value is @samp{init}.
21172 Set dithering to reduce the circular banding effects. Default is @code{1}
21176 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21177 Setting this value to the SAR of the input will make a rectangular vignetting
21178 following the dimensions of the video.
21180 Default is @code{1/1}.
21183 @subsection Expressions
21185 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21186 following parameters.
21191 input width and height
21194 the number of input frame, starting from 0
21197 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21198 @var{TB} units, NAN if undefined
21201 frame rate of the input video, NAN if the input frame rate is unknown
21204 the PTS (Presentation TimeStamp) of the filtered video frame,
21205 expressed in seconds, NAN if undefined
21208 time base of the input video
21212 @subsection Examples
21216 Apply simple strong vignetting effect:
21222 Make a flickering vignetting:
21224 vignette='PI/4+random(1)*PI/50':eval=frame
21229 @section vmafmotion
21231 Obtain the average VMAF motion score of a video.
21232 It is one of the component metrics of VMAF.
21234 The obtained average motion score is printed through the logging system.
21236 The filter accepts the following options:
21240 If specified, the filter will use the named file to save the motion score of
21241 each frame with respect to the previous frame.
21242 When filename equals "-" the data is sent to standard output.
21247 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21251 Stack input videos vertically.
21253 All streams must be of same pixel format and of same width.
21255 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21256 to create same output.
21258 The filter accepts the following options:
21262 Set number of input streams. Default is 2.
21265 If set to 1, force the output to terminate when the shortest input
21266 terminates. Default value is 0.
21271 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21272 Deinterlacing Filter").
21274 Based on the process described by Martin Weston for BBC R&D, and
21275 implemented based on the de-interlace algorithm written by Jim
21276 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21277 uses filter coefficients calculated by BBC R&D.
21279 This filter uses field-dominance information in frame to decide which
21280 of each pair of fields to place first in the output.
21281 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21283 There are two sets of filter coefficients, so called "simple"
21284 and "complex". Which set of filter coefficients is used can
21285 be set by passing an optional parameter:
21289 Set the interlacing filter coefficients. Accepts one of the following values:
21293 Simple filter coefficient set.
21295 More-complex filter coefficient set.
21297 Default value is @samp{complex}.
21300 The interlacing mode to adopt. It accepts one of the following values:
21304 Output one frame for each frame.
21306 Output one frame for each field.
21309 The default value is @code{field}.
21312 The picture field parity assumed for the input interlaced video. It accepts one
21313 of the following values:
21317 Assume the top field is first.
21319 Assume the bottom field is first.
21321 Enable automatic detection of field parity.
21324 The default value is @code{auto}.
21325 If the interlacing is unknown or the decoder does not export this information,
21326 top field first will be assumed.
21329 Specify which frames to deinterlace. Accepts one of the following values:
21333 Deinterlace all frames,
21335 Only deinterlace frames marked as interlaced.
21338 Default value is @samp{all}.
21341 @subsection Commands
21342 This filter supports same @ref{commands} as options.
21345 Video waveform monitor.
21347 The waveform monitor plots color component intensity. By default luminance
21348 only. Each column of the waveform corresponds to a column of pixels in the
21351 It accepts the following options:
21355 Can be either @code{row}, or @code{column}. Default is @code{column}.
21356 In row mode, the graph on the left side represents color component value 0 and
21357 the right side represents value = 255. In column mode, the top side represents
21358 color component value = 0 and bottom side represents value = 255.
21361 Set intensity. Smaller values are useful to find out how many values of the same
21362 luminance are distributed across input rows/columns.
21363 Default value is @code{0.04}. Allowed range is [0, 1].
21366 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21367 In mirrored mode, higher values will be represented on the left
21368 side for @code{row} mode and at the top for @code{column} mode. Default is
21369 @code{1} (mirrored).
21373 It accepts the following values:
21376 Presents information identical to that in the @code{parade}, except
21377 that the graphs representing color components are superimposed directly
21380 This display mode makes it easier to spot relative differences or similarities
21381 in overlapping areas of the color components that are supposed to be identical,
21382 such as neutral whites, grays, or blacks.
21385 Display separate graph for the color components side by side in
21386 @code{row} mode or one below the other in @code{column} mode.
21389 Display separate graph for the color components side by side in
21390 @code{column} mode or one below the other in @code{row} mode.
21392 Using this display mode makes it easy to spot color casts in the highlights
21393 and shadows of an image, by comparing the contours of the top and the bottom
21394 graphs of each waveform. Since whites, grays, and blacks are characterized
21395 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21396 should display three waveforms of roughly equal width/height. If not, the
21397 correction is easy to perform by making level adjustments the three waveforms.
21399 Default is @code{stack}.
21401 @item components, c
21402 Set which color components to display. Default is 1, which means only luminance
21403 or red color component if input is in RGB colorspace. If is set for example to
21404 7 it will display all 3 (if) available color components.
21409 No envelope, this is default.
21412 Instant envelope, minimum and maximum values presented in graph will be easily
21413 visible even with small @code{step} value.
21416 Hold minimum and maximum values presented in graph across time. This way you
21417 can still spot out of range values without constantly looking at waveforms.
21420 Peak and instant envelope combined together.
21426 No filtering, this is default.
21429 Luma and chroma combined together.
21432 Similar as above, but shows difference between blue and red chroma.
21435 Similar as above, but use different colors.
21438 Similar as above, but again with different colors.
21441 Displays only chroma.
21444 Displays actual color value on waveform.
21447 Similar as above, but with luma showing frequency of chroma values.
21451 Set which graticule to display.
21455 Do not display graticule.
21458 Display green graticule showing legal broadcast ranges.
21461 Display orange graticule showing legal broadcast ranges.
21464 Display invert graticule showing legal broadcast ranges.
21468 Set graticule opacity.
21471 Set graticule flags.
21475 Draw numbers above lines. By default enabled.
21478 Draw dots instead of lines.
21482 Set scale used for displaying graticule.
21489 Default is digital.
21492 Set background opacity.
21496 Set tint for output.
21497 Only used with lowpass filter and when display is not overlay and input
21498 pixel formats are not RGB.
21501 @section weave, doubleweave
21503 The @code{weave} takes a field-based video input and join
21504 each two sequential fields into single frame, producing a new double
21505 height clip with half the frame rate and half the frame count.
21507 The @code{doubleweave} works same as @code{weave} but without
21508 halving frame rate and frame count.
21510 It accepts the following option:
21514 Set first field. Available values are:
21518 Set the frame as top-field-first.
21521 Set the frame as bottom-field-first.
21525 @subsection Examples
21529 Interlace video using @ref{select} and @ref{separatefields} filter:
21531 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21536 Apply the xBR high-quality magnification filter which is designed for pixel
21537 art. It follows a set of edge-detection rules, see
21538 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21540 It accepts the following option:
21544 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21545 @code{3xBR} and @code{4} for @code{4xBR}.
21546 Default is @code{3}.
21551 Apply cross fade from one input video stream to another input video stream.
21552 The cross fade is applied for specified duration.
21554 The filter accepts the following options:
21558 Set one of available transition effects:
21606 Default transition effect is fade.
21609 Set cross fade duration in seconds.
21610 Default duration is 1 second.
21613 Set cross fade start relative to first input stream in seconds.
21614 Default offset is 0.
21617 Set expression for custom transition effect.
21619 The expressions can use the following variables and functions:
21624 The coordinates of the current sample.
21628 The width and height of the image.
21631 Progress of transition effect.
21634 Currently processed plane.
21637 Return value of first input at current location and plane.
21640 Return value of second input at current location and plane.
21646 Return the value of the pixel at location (@var{x},@var{y}) of the
21647 first/second/third/fourth component of first input.
21653 Return the value of the pixel at location (@var{x},@var{y}) of the
21654 first/second/third/fourth component of second input.
21658 @subsection Examples
21662 Cross fade from one input video to another input video, with fade transition and duration of transition
21663 of 2 seconds starting at offset of 5 seconds:
21665 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21670 Pick median pixels from several input videos.
21672 The filter accepts the following options:
21676 Set number of inputs.
21677 Default is 3. Allowed range is from 3 to 255.
21678 If number of inputs is even number, than result will be mean value between two median values.
21681 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21684 Set median percentile. Default value is @code{0.5}.
21685 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21686 minimum values, and @code{1} maximum values.
21689 @subsection Commands
21691 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21694 Stack video inputs into custom layout.
21696 All streams must be of same pixel format.
21698 The filter accepts the following options:
21702 Set number of input streams. Default is 2.
21705 Specify layout of inputs.
21706 This option requires the desired layout configuration to be explicitly set by the user.
21707 This sets position of each video input in output. Each input
21708 is separated by '|'.
21709 The first number represents the column, and the second number represents the row.
21710 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21711 where X is video input from which to take width or height.
21712 Multiple values can be used when separated by '+'. In such
21713 case values are summed together.
21715 Note that if inputs are of different sizes gaps may appear, as not all of
21716 the output video frame will be filled. Similarly, videos can overlap each
21717 other if their position doesn't leave enough space for the full frame of
21720 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21721 a layout must be set by the user.
21724 If set to 1, force the output to terminate when the shortest input
21725 terminates. Default value is 0.
21728 If set to valid color, all unused pixels will be filled with that color.
21729 By default fill is set to none, so it is disabled.
21732 @subsection Examples
21736 Display 4 inputs into 2x2 grid.
21740 input1(0, 0) | input3(w0, 0)
21741 input2(0, h0) | input4(w0, h0)
21745 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21748 Note that if inputs are of different sizes, gaps or overlaps may occur.
21751 Display 4 inputs into 1x4 grid.
21758 input4(0, h0+h1+h2)
21762 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21765 Note that if inputs are of different widths, unused space will appear.
21768 Display 9 inputs into 3x3 grid.
21772 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21773 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21774 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21778 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
21781 Note that if inputs are of different sizes, gaps or overlaps may occur.
21784 Display 16 inputs into 4x4 grid.
21788 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
21789 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
21790 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
21791 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
21795 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|
21796 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
21799 Note that if inputs are of different sizes, gaps or overlaps may occur.
21806 Deinterlace the input video ("yadif" means "yet another deinterlacing
21809 It accepts the following parameters:
21815 The interlacing mode to adopt. It accepts one of the following values:
21818 @item 0, send_frame
21819 Output one frame for each frame.
21820 @item 1, send_field
21821 Output one frame for each field.
21822 @item 2, send_frame_nospatial
21823 Like @code{send_frame}, but it skips the spatial interlacing check.
21824 @item 3, send_field_nospatial
21825 Like @code{send_field}, but it skips the spatial interlacing check.
21828 The default value is @code{send_frame}.
21831 The picture field parity assumed for the input interlaced video. It accepts one
21832 of the following values:
21836 Assume the top field is first.
21838 Assume the bottom field is first.
21840 Enable automatic detection of field parity.
21843 The default value is @code{auto}.
21844 If the interlacing is unknown or the decoder does not export this information,
21845 top field first will be assumed.
21848 Specify which frames to deinterlace. Accepts one of the following
21853 Deinterlace all frames.
21854 @item 1, interlaced
21855 Only deinterlace frames marked as interlaced.
21858 The default value is @code{all}.
21861 @section yadif_cuda
21863 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
21864 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
21867 It accepts the following parameters:
21873 The interlacing mode to adopt. It accepts one of the following values:
21876 @item 0, send_frame
21877 Output one frame for each frame.
21878 @item 1, send_field
21879 Output one frame for each field.
21880 @item 2, send_frame_nospatial
21881 Like @code{send_frame}, but it skips the spatial interlacing check.
21882 @item 3, send_field_nospatial
21883 Like @code{send_field}, but it skips the spatial interlacing check.
21886 The default value is @code{send_frame}.
21889 The picture field parity assumed for the input interlaced video. It accepts one
21890 of the following values:
21894 Assume the top field is first.
21896 Assume the bottom field is first.
21898 Enable automatic detection of field parity.
21901 The default value is @code{auto}.
21902 If the interlacing is unknown or the decoder does not export this information,
21903 top field first will be assumed.
21906 Specify which frames to deinterlace. Accepts one of the following
21911 Deinterlace all frames.
21912 @item 1, interlaced
21913 Only deinterlace frames marked as interlaced.
21916 The default value is @code{all}.
21921 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
21922 The algorithm is described in
21923 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
21925 It accepts the following parameters:
21929 Set the window radius. Default value is 3.
21932 Set which planes to filter. Default is only the first plane.
21935 Set blur strength. Default value is 128.
21938 @subsection Commands
21939 This filter supports same @ref{commands} as options.
21943 Apply Zoom & Pan effect.
21945 This filter accepts the following options:
21949 Set the zoom expression. Range is 1-10. Default is 1.
21953 Set the x and y expression. Default is 0.
21956 Set the duration expression in number of frames.
21957 This sets for how many number of frames effect will last for
21958 single input image.
21961 Set the output image size, default is 'hd720'.
21964 Set the output frame rate, default is '25'.
21967 Each expression can contain the following constants:
21986 Output frame count.
21989 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
21991 @item out_time, time, ot
21992 The output timestamp expressed in seconds.
21996 Last calculated 'x' and 'y' position from 'x' and 'y' expression
21997 for current input frame.
22001 'x' and 'y' of last output frame of previous input frame or 0 when there was
22002 not yet such frame (first input frame).
22005 Last calculated zoom from 'z' expression for current input frame.
22008 Last calculated zoom of last output frame of previous input frame.
22011 Number of output frames for current input frame. Calculated from 'd' expression
22012 for each input frame.
22015 number of output frames created for previous input frame
22018 Rational number: input width / input height
22021 sample aspect ratio
22024 display aspect ratio
22028 @subsection Examples
22032 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
22034 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
22038 Zoom in up to 1.5x and pan always at center of picture:
22040 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22044 Same as above but without pausing:
22046 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22050 Zoom in 2x into center of picture only for the first second of the input video:
22052 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22059 Scale (resize) the input video, using the z.lib library:
22060 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
22061 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
22063 The zscale filter forces the output display aspect ratio to be the same
22064 as the input, by changing the output sample aspect ratio.
22066 If the input image format is different from the format requested by
22067 the next filter, the zscale filter will convert the input to the
22070 @subsection Options
22071 The filter accepts the following options.
22076 Set the output video dimension expression. Default value is the input
22079 If the @var{width} or @var{w} value is 0, the input width is used for
22080 the output. If the @var{height} or @var{h} value is 0, the input height
22081 is used for the output.
22083 If one and only one of the values is -n with n >= 1, the zscale filter
22084 will use a value that maintains the aspect ratio of the input image,
22085 calculated from the other specified dimension. After that it will,
22086 however, make sure that the calculated dimension is divisible by n and
22087 adjust the value if necessary.
22089 If both values are -n with n >= 1, the behavior will be identical to
22090 both values being set to 0 as previously detailed.
22092 See below for the list of accepted constants for use in the dimension
22096 Set the video size. For the syntax of this option, check the
22097 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22100 Set the dither type.
22102 Possible values are:
22107 @item error_diffusion
22113 Set the resize filter type.
22115 Possible values are:
22125 Default is bilinear.
22128 Set the color range.
22130 Possible values are:
22137 Default is same as input.
22140 Set the color primaries.
22142 Possible values are:
22152 Default is same as input.
22155 Set the transfer characteristics.
22157 Possible values are:
22171 Default is same as input.
22174 Set the colorspace matrix.
22176 Possible value are:
22187 Default is same as input.
22190 Set the input color range.
22192 Possible values are:
22199 Default is same as input.
22201 @item primariesin, pin
22202 Set the input color primaries.
22204 Possible values are:
22214 Default is same as input.
22216 @item transferin, tin
22217 Set the input transfer characteristics.
22219 Possible values are:
22230 Default is same as input.
22232 @item matrixin, min
22233 Set the input colorspace matrix.
22235 Possible value are:
22247 Set the output chroma location.
22249 Possible values are:
22260 @item chromalin, cin
22261 Set the input chroma location.
22263 Possible values are:
22275 Set the nominal peak luminance.
22278 The values of the @option{w} and @option{h} options are expressions
22279 containing the following constants:
22284 The input width and height
22288 These are the same as @var{in_w} and @var{in_h}.
22292 The output (scaled) width and height
22296 These are the same as @var{out_w} and @var{out_h}
22299 The same as @var{iw} / @var{ih}
22302 input sample aspect ratio
22305 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22309 horizontal and vertical input chroma subsample values. For example for the
22310 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22314 horizontal and vertical output chroma subsample values. For example for the
22315 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22318 @subsection Commands
22320 This filter supports the following commands:
22324 Set the output video dimension expression.
22325 The command accepts the same syntax of the corresponding option.
22327 If the specified expression is not valid, it is kept at its current
22331 @c man end VIDEO FILTERS
22333 @chapter OpenCL Video Filters
22334 @c man begin OPENCL VIDEO FILTERS
22336 Below is a description of the currently available OpenCL video filters.
22338 To enable compilation of these filters you need to configure FFmpeg with
22339 @code{--enable-opencl}.
22341 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22344 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22345 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22346 given device parameters.
22348 @item -filter_hw_device @var{name}
22349 Pass the hardware device called @var{name} to all filters in any filter graph.
22353 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22357 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22359 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22363 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.
22365 @section avgblur_opencl
22367 Apply average blur filter.
22369 The filter accepts the following options:
22373 Set horizontal radius size.
22374 Range is @code{[1, 1024]} and default value is @code{1}.
22377 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22380 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22383 @subsection Example
22387 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.
22389 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22393 @section boxblur_opencl
22395 Apply a boxblur algorithm to the input video.
22397 It accepts the following parameters:
22401 @item luma_radius, lr
22402 @item luma_power, lp
22403 @item chroma_radius, cr
22404 @item chroma_power, cp
22405 @item alpha_radius, ar
22406 @item alpha_power, ap
22410 A description of the accepted options follows.
22413 @item luma_radius, lr
22414 @item chroma_radius, cr
22415 @item alpha_radius, ar
22416 Set an expression for the box radius in pixels used for blurring the
22417 corresponding input plane.
22419 The radius value must be a non-negative number, and must not be
22420 greater than the value of the expression @code{min(w,h)/2} for the
22421 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22424 Default value for @option{luma_radius} is "2". If not specified,
22425 @option{chroma_radius} and @option{alpha_radius} default to the
22426 corresponding value set for @option{luma_radius}.
22428 The expressions can contain the following constants:
22432 The input width and height in pixels.
22436 The input chroma image width and height in pixels.
22440 The horizontal and vertical chroma subsample values. For example, for the
22441 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22444 @item luma_power, lp
22445 @item chroma_power, cp
22446 @item alpha_power, ap
22447 Specify how many times the boxblur filter is applied to the
22448 corresponding plane.
22450 Default value for @option{luma_power} is 2. If not specified,
22451 @option{chroma_power} and @option{alpha_power} default to the
22452 corresponding value set for @option{luma_power}.
22454 A value of 0 will disable the effect.
22457 @subsection Examples
22459 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.
22463 Apply a boxblur filter with the luma, chroma, and alpha radius
22464 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.
22466 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22467 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22471 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.
22473 For the luma plane, a 2x2 box radius will be run once.
22475 For the chroma plane, a 4x4 box radius will be run 5 times.
22477 For the alpha plane, a 3x3 box radius will be run 7 times.
22479 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22483 @section colorkey_opencl
22484 RGB colorspace color keying.
22486 The filter accepts the following options:
22490 The color which will be replaced with transparency.
22493 Similarity percentage with the key color.
22495 0.01 matches only the exact key color, while 1.0 matches everything.
22500 0.0 makes pixels either fully transparent, or not transparent at all.
22502 Higher values result in semi-transparent pixels, with a higher transparency
22503 the more similar the pixels color is to the key color.
22506 @subsection Examples
22510 Make every semi-green pixel in the input transparent with some slight blending:
22512 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22516 @section convolution_opencl
22518 Apply convolution of 3x3, 5x5, 7x7 matrix.
22520 The filter accepts the following options:
22527 Set matrix for each plane.
22528 Matrix is sequence of 9, 25 or 49 signed numbers.
22529 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22535 Set multiplier for calculated value for each plane.
22536 If unset or 0, it will be sum of all matrix elements.
22537 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22543 Set bias for each plane. This value is added to the result of the multiplication.
22544 Useful for making the overall image brighter or darker.
22545 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22549 @subsection Examples
22555 -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
22561 -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
22565 Apply edge enhance:
22567 -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
22573 -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
22577 Apply laplacian edge detector which includes diagonals:
22579 -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
22585 -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
22589 @section erosion_opencl
22591 Apply erosion effect to the video.
22593 This filter replaces the pixel by the local(3x3) minimum.
22595 It accepts the following options:
22602 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22603 If @code{0}, plane will remain unchanged.
22606 Flag which specifies the pixel to refer to.
22607 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22609 Flags to local 3x3 coordinates region centered on @code{x}:
22618 @subsection Example
22622 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.
22624 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22628 @section deshake_opencl
22629 Feature-point based video stabilization filter.
22631 The filter accepts the following options:
22635 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22638 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22640 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22642 Viewing point matches in the output video is only supported for RGB input.
22644 Defaults to @code{0}.
22646 @item adaptive_crop
22647 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22649 Defaults to @code{1}.
22651 @item refine_features
22652 Whether or not feature points should be refined at a sub-pixel level.
22654 This can be turned off for a slight performance gain at the cost of precision.
22656 Defaults to @code{1}.
22658 @item smooth_strength
22659 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22661 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22663 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22665 Defaults to @code{0.0}.
22667 @item smooth_window_multiplier
22668 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22670 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22672 Acceptable values range from @code{0.1} to @code{10.0}.
22674 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22675 potentially improving smoothness, but also increase latency and memory usage.
22677 Defaults to @code{2.0}.
22681 @subsection Examples
22685 Stabilize a video with a fixed, medium smoothing strength:
22687 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22691 Stabilize a video with debugging (both in console and in rendered video):
22693 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22697 @section dilation_opencl
22699 Apply dilation effect to the video.
22701 This filter replaces the pixel by the local(3x3) maximum.
22703 It accepts the following options:
22710 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22711 If @code{0}, plane will remain unchanged.
22714 Flag which specifies the pixel to refer to.
22715 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22717 Flags to local 3x3 coordinates region centered on @code{x}:
22726 @subsection Example
22730 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.
22732 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22736 @section nlmeans_opencl
22738 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22740 @section overlay_opencl
22742 Overlay one video on top of another.
22744 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22745 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22747 The filter accepts the following options:
22752 Set the x coordinate of the overlaid video on the main video.
22753 Default value is @code{0}.
22756 Set the y coordinate of the overlaid video on the main video.
22757 Default value is @code{0}.
22761 @subsection Examples
22765 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22767 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22770 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22772 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22777 @section pad_opencl
22779 Add paddings to the input image, and place the original input at the
22780 provided @var{x}, @var{y} coordinates.
22782 It accepts the following options:
22787 Specify an expression for the size of the output image with the
22788 paddings added. If the value for @var{width} or @var{height} is 0, the
22789 corresponding input size is used for the output.
22791 The @var{width} expression can reference the value set by the
22792 @var{height} expression, and vice versa.
22794 The default value of @var{width} and @var{height} is 0.
22798 Specify the offsets to place the input image at within the padded area,
22799 with respect to the top/left border of the output image.
22801 The @var{x} expression can reference the value set by the @var{y}
22802 expression, and vice versa.
22804 The default value of @var{x} and @var{y} is 0.
22806 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
22807 so the input image is centered on the padded area.
22810 Specify the color of the padded area. For the syntax of this option,
22811 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
22812 manual,ffmpeg-utils}.
22815 Pad to an aspect instead to a resolution.
22818 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
22819 options are expressions containing the following constants:
22824 The input video width and height.
22828 These are the same as @var{in_w} and @var{in_h}.
22832 The output width and height (the size of the padded area), as
22833 specified by the @var{width} and @var{height} expressions.
22837 These are the same as @var{out_w} and @var{out_h}.
22841 The x and y offsets as specified by the @var{x} and @var{y}
22842 expressions, or NAN if not yet specified.
22845 same as @var{iw} / @var{ih}
22848 input sample aspect ratio
22851 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
22854 @section prewitt_opencl
22856 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
22858 The filter accepts the following option:
22862 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22865 Set value which will be multiplied with filtered result.
22866 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22869 Set value which will be added to filtered result.
22870 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22873 @subsection Example
22877 Apply the Prewitt operator with scale set to 2 and delta set to 10.
22879 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
22883 @anchor{program_opencl}
22884 @section program_opencl
22886 Filter video using an OpenCL program.
22891 OpenCL program source file.
22894 Kernel name in program.
22897 Number of inputs to the filter. Defaults to 1.
22900 Size of output frames. Defaults to the same as the first input.
22904 The @code{program_opencl} filter also supports the @ref{framesync} options.
22906 The program source file must contain a kernel function with the given name,
22907 which will be run once for each plane of the output. Each run on a plane
22908 gets enqueued as a separate 2D global NDRange with one work-item for each
22909 pixel to be generated. The global ID offset for each work-item is therefore
22910 the coordinates of a pixel in the destination image.
22912 The kernel function needs to take the following arguments:
22915 Destination image, @var{__write_only image2d_t}.
22917 This image will become the output; the kernel should write all of it.
22919 Frame index, @var{unsigned int}.
22921 This is a counter starting from zero and increasing by one for each frame.
22923 Source images, @var{__read_only image2d_t}.
22925 These are the most recent images on each input. The kernel may read from
22926 them to generate the output, but they can't be written to.
22933 Copy the input to the output (output must be the same size as the input).
22935 __kernel void copy(__write_only image2d_t destination,
22936 unsigned int index,
22937 __read_only image2d_t source)
22939 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
22941 int2 location = (int2)(get_global_id(0), get_global_id(1));
22943 float4 value = read_imagef(source, sampler, location);
22945 write_imagef(destination, location, value);
22950 Apply a simple transformation, rotating the input by an amount increasing
22951 with the index counter. Pixel values are linearly interpolated by the
22952 sampler, and the output need not have the same dimensions as the input.
22954 __kernel void rotate_image(__write_only image2d_t dst,
22955 unsigned int index,
22956 __read_only image2d_t src)
22958 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22959 CLK_FILTER_LINEAR);
22961 float angle = (float)index / 100.0f;
22963 float2 dst_dim = convert_float2(get_image_dim(dst));
22964 float2 src_dim = convert_float2(get_image_dim(src));
22966 float2 dst_cen = dst_dim / 2.0f;
22967 float2 src_cen = src_dim / 2.0f;
22969 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
22971 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
22973 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
22974 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
22976 src_pos = src_pos * src_dim / dst_dim;
22978 float2 src_loc = src_pos + src_cen;
22980 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
22981 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
22982 write_imagef(dst, dst_loc, 0.5f);
22984 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
22989 Blend two inputs together, with the amount of each input used varying
22990 with the index counter.
22992 __kernel void blend_images(__write_only image2d_t dst,
22993 unsigned int index,
22994 __read_only image2d_t src1,
22995 __read_only image2d_t src2)
22997 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22998 CLK_FILTER_LINEAR);
23000 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
23002 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23003 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
23004 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
23006 float4 val1 = read_imagef(src1, sampler, src1_loc);
23007 float4 val2 = read_imagef(src2, sampler, src2_loc);
23009 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
23015 @section roberts_opencl
23016 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
23018 The filter accepts the following option:
23022 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23025 Set value which will be multiplied with filtered result.
23026 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23029 Set value which will be added to filtered result.
23030 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23033 @subsection Example
23037 Apply the Roberts cross operator with scale set to 2 and delta set to 10
23039 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
23043 @section sobel_opencl
23045 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
23047 The filter accepts the following option:
23051 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23054 Set value which will be multiplied with filtered result.
23055 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23058 Set value which will be added to filtered result.
23059 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23062 @subsection Example
23066 Apply sobel operator with scale set to 2 and delta set to 10
23068 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
23072 @section tonemap_opencl
23074 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
23076 It accepts the following parameters:
23080 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
23083 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
23086 Apply desaturation for highlights that exceed this level of brightness. The
23087 higher the parameter, the more color information will be preserved. This
23088 setting helps prevent unnaturally blown-out colors for super-highlights, by
23089 (smoothly) turning into white instead. This makes images feel more natural,
23090 at the cost of reducing information about out-of-range colors.
23092 The default value is 0.5, and the algorithm here is a little different from
23093 the cpu version tonemap currently. A setting of 0.0 disables this option.
23096 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
23097 is used to detect whether the scene has changed or not. If the distance between
23098 the current frame average brightness and the current running average exceeds
23099 a threshold value, we would re-calculate scene average and peak brightness.
23100 The default value is 0.2.
23103 Specify the output pixel format.
23105 Currently supported formats are:
23112 Set the output color range.
23114 Possible values are:
23120 Default is same as input.
23123 Set the output color primaries.
23125 Possible values are:
23131 Default is same as input.
23134 Set the output transfer characteristics.
23136 Possible values are:
23145 Set the output colorspace matrix.
23147 Possible value are:
23153 Default is same as input.
23157 @subsection Example
23161 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23163 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23167 @section unsharp_opencl
23169 Sharpen or blur the input video.
23171 It accepts the following parameters:
23174 @item luma_msize_x, lx
23175 Set the luma matrix horizontal size.
23176 Range is @code{[1, 23]} and default value is @code{5}.
23178 @item luma_msize_y, ly
23179 Set the luma matrix vertical size.
23180 Range is @code{[1, 23]} and default value is @code{5}.
23182 @item luma_amount, la
23183 Set the luma effect strength.
23184 Range is @code{[-10, 10]} and default value is @code{1.0}.
23186 Negative values will blur the input video, while positive values will
23187 sharpen it, a value of zero will disable the effect.
23189 @item chroma_msize_x, cx
23190 Set the chroma matrix horizontal size.
23191 Range is @code{[1, 23]} and default value is @code{5}.
23193 @item chroma_msize_y, cy
23194 Set the chroma matrix vertical size.
23195 Range is @code{[1, 23]} and default value is @code{5}.
23197 @item chroma_amount, ca
23198 Set the chroma effect strength.
23199 Range is @code{[-10, 10]} and default value is @code{0.0}.
23201 Negative values will blur the input video, while positive values will
23202 sharpen it, a value of zero will disable the effect.
23206 All parameters are optional and default to the equivalent of the
23207 string '5:5:1.0:5:5:0.0'.
23209 @subsection Examples
23213 Apply strong luma sharpen effect:
23215 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23219 Apply a strong blur of both luma and chroma parameters:
23221 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23225 @section xfade_opencl
23227 Cross fade two videos with custom transition effect by using OpenCL.
23229 It accepts the following options:
23233 Set one of possible transition effects.
23237 Select custom transition effect, the actual transition description
23238 will be picked from source and kernel options.
23250 Default transition is fade.
23254 OpenCL program source file for custom transition.
23257 Set name of kernel to use for custom transition from program source file.
23260 Set duration of video transition.
23263 Set time of start of transition relative to first video.
23266 The program source file must contain a kernel function with the given name,
23267 which will be run once for each plane of the output. Each run on a plane
23268 gets enqueued as a separate 2D global NDRange with one work-item for each
23269 pixel to be generated. The global ID offset for each work-item is therefore
23270 the coordinates of a pixel in the destination image.
23272 The kernel function needs to take the following arguments:
23275 Destination image, @var{__write_only image2d_t}.
23277 This image will become the output; the kernel should write all of it.
23280 First Source image, @var{__read_only image2d_t}.
23281 Second Source image, @var{__read_only image2d_t}.
23283 These are the most recent images on each input. The kernel may read from
23284 them to generate the output, but they can't be written to.
23287 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23294 Apply dots curtain transition effect:
23296 __kernel void blend_images(__write_only image2d_t dst,
23297 __read_only image2d_t src1,
23298 __read_only image2d_t src2,
23301 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23302 CLK_FILTER_LINEAR);
23303 int2 p = (int2)(get_global_id(0), get_global_id(1));
23304 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23305 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23308 float2 dots = (float2)(20.0, 20.0);
23309 float2 center = (float2)(0,0);
23312 float4 val1 = read_imagef(src1, sampler, p);
23313 float4 val2 = read_imagef(src2, sampler, p);
23314 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23316 write_imagef(dst, p, next ? val1 : val2);
23322 @c man end OPENCL VIDEO FILTERS
23324 @chapter VAAPI Video Filters
23325 @c man begin VAAPI VIDEO FILTERS
23327 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23329 To enable compilation of these filters you need to configure FFmpeg with
23330 @code{--enable-vaapi}.
23332 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}
23334 @section tonemap_vaapi
23336 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23337 It maps the dynamic range of HDR10 content to the SDR content.
23338 It currently only accepts HDR10 as input.
23340 It accepts the following parameters:
23344 Specify the output pixel format.
23346 Currently supported formats are:
23355 Set the output color primaries.
23357 Default is same as input.
23360 Set the output transfer characteristics.
23365 Set the output colorspace matrix.
23367 Default is same as input.
23371 @subsection Example
23375 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23377 tonemap_vaapi=format=p010:t=bt2020-10
23381 @c man end VAAPI VIDEO FILTERS
23383 @chapter Video Sources
23384 @c man begin VIDEO SOURCES
23386 Below is a description of the currently available video sources.
23390 Buffer video frames, and make them available to the filter chain.
23392 This source is mainly intended for a programmatic use, in particular
23393 through the interface defined in @file{libavfilter/buffersrc.h}.
23395 It accepts the following parameters:
23400 Specify the size (width and height) of the buffered video frames. For the
23401 syntax of this option, check the
23402 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23405 The input video width.
23408 The input video height.
23411 A string representing the pixel format of the buffered video frames.
23412 It may be a number corresponding to a pixel format, or a pixel format
23416 Specify the timebase assumed by the timestamps of the buffered frames.
23419 Specify the frame rate expected for the video stream.
23421 @item pixel_aspect, sar
23422 The sample (pixel) aspect ratio of the input video.
23425 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23426 to the filtergraph description to specify swscale flags for automatically
23427 inserted scalers. See @ref{Filtergraph syntax}.
23429 @item hw_frames_ctx
23430 When using a hardware pixel format, this should be a reference to an
23431 AVHWFramesContext describing input frames.
23436 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23439 will instruct the source to accept video frames with size 320x240 and
23440 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23441 square pixels (1:1 sample aspect ratio).
23442 Since the pixel format with name "yuv410p" corresponds to the number 6
23443 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23444 this example corresponds to:
23446 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23449 Alternatively, the options can be specified as a flat string, but this
23450 syntax is deprecated:
23452 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23456 Create a pattern generated by an elementary cellular automaton.
23458 The initial state of the cellular automaton can be defined through the
23459 @option{filename} and @option{pattern} options. If such options are
23460 not specified an initial state is created randomly.
23462 At each new frame a new row in the video is filled with the result of
23463 the cellular automaton next generation. The behavior when the whole
23464 frame is filled is defined by the @option{scroll} option.
23466 This source accepts the following options:
23470 Read the initial cellular automaton state, i.e. the starting row, from
23471 the specified file.
23472 In the file, each non-whitespace character is considered an alive
23473 cell, a newline will terminate the row, and further characters in the
23474 file will be ignored.
23477 Read the initial cellular automaton state, i.e. the starting row, from
23478 the specified string.
23480 Each non-whitespace character in the string is considered an alive
23481 cell, a newline will terminate the row, and further characters in the
23482 string will be ignored.
23485 Set the video rate, that is the number of frames generated per second.
23488 @item random_fill_ratio, ratio
23489 Set the random fill ratio for the initial cellular automaton row. It
23490 is a floating point number value ranging from 0 to 1, defaults to
23493 This option is ignored when a file or a pattern is specified.
23495 @item random_seed, seed
23496 Set the seed for filling randomly the initial row, must be an integer
23497 included between 0 and UINT32_MAX. If not specified, or if explicitly
23498 set to -1, the filter will try to use a good random seed on a best
23502 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23503 Default value is 110.
23506 Set the size of the output video. For the syntax of this option, check the
23507 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23509 If @option{filename} or @option{pattern} is specified, the size is set
23510 by default to the width of the specified initial state row, and the
23511 height is set to @var{width} * PHI.
23513 If @option{size} is set, it must contain the width of the specified
23514 pattern string, and the specified pattern will be centered in the
23517 If a filename or a pattern string is not specified, the size value
23518 defaults to "320x518" (used for a randomly generated initial state).
23521 If set to 1, scroll the output upward when all the rows in the output
23522 have been already filled. If set to 0, the new generated row will be
23523 written over the top row just after the bottom row is filled.
23526 @item start_full, full
23527 If set to 1, completely fill the output with generated rows before
23528 outputting the first frame.
23529 This is the default behavior, for disabling set the value to 0.
23532 If set to 1, stitch the left and right row edges together.
23533 This is the default behavior, for disabling set the value to 0.
23536 @subsection Examples
23540 Read the initial state from @file{pattern}, and specify an output of
23543 cellauto=f=pattern:s=200x400
23547 Generate a random initial row with a width of 200 cells, with a fill
23550 cellauto=ratio=2/3:s=200x200
23554 Create a pattern generated by rule 18 starting by a single alive cell
23555 centered on an initial row with width 100:
23557 cellauto=p=@@:s=100x400:full=0:rule=18
23561 Specify a more elaborated initial pattern:
23563 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23568 @anchor{coreimagesrc}
23569 @section coreimagesrc
23570 Video source generated on GPU using Apple's CoreImage API on OSX.
23572 This video source is a specialized version of the @ref{coreimage} video filter.
23573 Use a core image generator at the beginning of the applied filterchain to
23574 generate the content.
23576 The coreimagesrc video source accepts the following options:
23578 @item list_generators
23579 List all available generators along with all their respective options as well as
23580 possible minimum and maximum values along with the default values.
23582 list_generators=true
23586 Specify the size of the sourced video. For the syntax of this option, check the
23587 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23588 The default value is @code{320x240}.
23591 Specify the frame rate of the sourced video, as the number of frames
23592 generated per second. It has to be a string in the format
23593 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23594 number or a valid video frame rate abbreviation. The default value is
23598 Set the sample aspect ratio of the sourced video.
23601 Set the duration of the sourced video. See
23602 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23603 for the accepted syntax.
23605 If not specified, or the expressed duration is negative, the video is
23606 supposed to be generated forever.
23609 Additionally, all options of the @ref{coreimage} video filter are accepted.
23610 A complete filterchain can be used for further processing of the
23611 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23612 and examples for details.
23614 @subsection Examples
23619 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23620 given as complete and escaped command-line for Apple's standard bash shell:
23622 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23624 This example is equivalent to the QRCode example of @ref{coreimage} without the
23625 need for a nullsrc video source.
23630 Generate several gradients.
23634 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23635 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23638 Set frame rate, expressed as number of frames per second. Default
23641 @item c0, c1, c2, c3, c4, c5, c6, c7
23642 Set 8 colors. Default values for colors is to pick random one.
23644 @item x0, y0, y0, y1
23645 Set gradient line source and destination points. If negative or out of range, random ones
23649 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23652 Set seed for picking gradient line points.
23655 Set the duration of the sourced video. See
23656 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23657 for the accepted syntax.
23659 If not specified, or the expressed duration is negative, the video is
23660 supposed to be generated forever.
23663 Set speed of gradients rotation.
23667 @section mandelbrot
23669 Generate a Mandelbrot set fractal, and progressively zoom towards the
23670 point specified with @var{start_x} and @var{start_y}.
23672 This source accepts the following options:
23677 Set the terminal pts value. Default value is 400.
23680 Set the terminal scale value.
23681 Must be a floating point value. Default value is 0.3.
23684 Set the inner coloring mode, that is the algorithm used to draw the
23685 Mandelbrot fractal internal region.
23687 It shall assume one of the following values:
23692 Show time until convergence.
23694 Set color based on point closest to the origin of the iterations.
23699 Default value is @var{mincol}.
23702 Set the bailout value. Default value is 10.0.
23705 Set the maximum of iterations performed by the rendering
23706 algorithm. Default value is 7189.
23709 Set outer coloring mode.
23710 It shall assume one of following values:
23712 @item iteration_count
23713 Set iteration count mode.
23714 @item normalized_iteration_count
23715 set normalized iteration count mode.
23717 Default value is @var{normalized_iteration_count}.
23720 Set frame rate, expressed as number of frames per second. Default
23724 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23725 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23728 Set the initial scale value. Default value is 3.0.
23731 Set the initial x position. Must be a floating point value between
23732 -100 and 100. Default value is -0.743643887037158704752191506114774.
23735 Set the initial y position. Must be a floating point value between
23736 -100 and 100. Default value is -0.131825904205311970493132056385139.
23741 Generate various test patterns, as generated by the MPlayer test filter.
23743 The size of the generated video is fixed, and is 256x256.
23744 This source is useful in particular for testing encoding features.
23746 This source accepts the following options:
23751 Specify the frame rate of the sourced video, as the number of frames
23752 generated per second. It has to be a string in the format
23753 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23754 number or a valid video frame rate abbreviation. The default value is
23758 Set the duration of the sourced video. See
23759 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23760 for the accepted syntax.
23762 If not specified, or the expressed duration is negative, the video is
23763 supposed to be generated forever.
23767 Set the number or the name of the test to perform. Supported tests are:
23781 @item max_frames, m
23782 Set the maximum number of frames generated for each test, default value is 30.
23786 Default value is "all", which will cycle through the list of all tests.
23791 mptestsrc=t=dc_luma
23794 will generate a "dc_luma" test pattern.
23796 @section frei0r_src
23798 Provide a frei0r source.
23800 To enable compilation of this filter you need to install the frei0r
23801 header and configure FFmpeg with @code{--enable-frei0r}.
23803 This source accepts the following parameters:
23808 The size of the video to generate. For the syntax of this option, check the
23809 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23812 The framerate of the generated video. It may be a string of the form
23813 @var{num}/@var{den} or a frame rate abbreviation.
23816 The name to the frei0r source to load. For more information regarding frei0r and
23817 how to set the parameters, read the @ref{frei0r} section in the video filters
23820 @item filter_params
23821 A '|'-separated list of parameters to pass to the frei0r source.
23825 For example, to generate a frei0r partik0l source with size 200x200
23826 and frame rate 10 which is overlaid on the overlay filter main input:
23828 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
23833 Generate a life pattern.
23835 This source is based on a generalization of John Conway's life game.
23837 The sourced input represents a life grid, each pixel represents a cell
23838 which can be in one of two possible states, alive or dead. Every cell
23839 interacts with its eight neighbours, which are the cells that are
23840 horizontally, vertically, or diagonally adjacent.
23842 At each interaction the grid evolves according to the adopted rule,
23843 which specifies the number of neighbor alive cells which will make a
23844 cell stay alive or born. The @option{rule} option allows one to specify
23847 This source accepts the following options:
23851 Set the file from which to read the initial grid state. In the file,
23852 each non-whitespace character is considered an alive cell, and newline
23853 is used to delimit the end of each row.
23855 If this option is not specified, the initial grid is generated
23859 Set the video rate, that is the number of frames generated per second.
23862 @item random_fill_ratio, ratio
23863 Set the random fill ratio for the initial random grid. It is a
23864 floating point number value ranging from 0 to 1, defaults to 1/PHI.
23865 It is ignored when a file is specified.
23867 @item random_seed, seed
23868 Set the seed for filling the initial random grid, must be an integer
23869 included between 0 and UINT32_MAX. If not specified, or if explicitly
23870 set to -1, the filter will try to use a good random seed on a best
23876 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
23877 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
23878 @var{NS} specifies the number of alive neighbor cells which make a
23879 live cell stay alive, and @var{NB} the number of alive neighbor cells
23880 which make a dead cell to become alive (i.e. to "born").
23881 "s" and "b" can be used in place of "S" and "B", respectively.
23883 Alternatively a rule can be specified by an 18-bits integer. The 9
23884 high order bits are used to encode the next cell state if it is alive
23885 for each number of neighbor alive cells, the low order bits specify
23886 the rule for "borning" new cells. Higher order bits encode for an
23887 higher number of neighbor cells.
23888 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
23889 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
23891 Default value is "S23/B3", which is the original Conway's game of life
23892 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
23893 cells, and will born a new cell if there are three alive cells around
23897 Set the size of the output video. For the syntax of this option, check the
23898 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23900 If @option{filename} is specified, the size is set by default to the
23901 same size of the input file. If @option{size} is set, it must contain
23902 the size specified in the input file, and the initial grid defined in
23903 that file is centered in the larger resulting area.
23905 If a filename is not specified, the size value defaults to "320x240"
23906 (used for a randomly generated initial grid).
23909 If set to 1, stitch the left and right grid edges together, and the
23910 top and bottom edges also. Defaults to 1.
23913 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
23914 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
23915 value from 0 to 255.
23918 Set the color of living (or new born) cells.
23921 Set the color of dead cells. If @option{mold} is set, this is the first color
23922 used to represent a dead cell.
23925 Set mold color, for definitely dead and moldy cells.
23927 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
23928 ffmpeg-utils manual,ffmpeg-utils}.
23931 @subsection Examples
23935 Read a grid from @file{pattern}, and center it on a grid of size
23938 life=f=pattern:s=300x300
23942 Generate a random grid of size 200x200, with a fill ratio of 2/3:
23944 life=ratio=2/3:s=200x200
23948 Specify a custom rule for evolving a randomly generated grid:
23954 Full example with slow death effect (mold) using @command{ffplay}:
23956 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
23963 @anchor{haldclutsrc}
23966 @anchor{pal100bars}
23967 @anchor{rgbtestsrc}
23969 @anchor{smptehdbars}
23972 @anchor{yuvtestsrc}
23973 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
23975 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
23977 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
23979 The @code{color} source provides an uniformly colored input.
23981 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
23982 @ref{haldclut} filter.
23984 The @code{nullsrc} source returns unprocessed video frames. It is
23985 mainly useful to be employed in analysis / debugging tools, or as the
23986 source for filters which ignore the input data.
23988 The @code{pal75bars} source generates a color bars pattern, based on
23989 EBU PAL recommendations with 75% color levels.
23991 The @code{pal100bars} source generates a color bars pattern, based on
23992 EBU PAL recommendations with 100% color levels.
23994 The @code{rgbtestsrc} source generates an RGB test pattern useful for
23995 detecting RGB vs BGR issues. You should see a red, green and blue
23996 stripe from top to bottom.
23998 The @code{smptebars} source generates a color bars pattern, based on
23999 the SMPTE Engineering Guideline EG 1-1990.
24001 The @code{smptehdbars} source generates a color bars pattern, based on
24002 the SMPTE RP 219-2002.
24004 The @code{testsrc} source generates a test video pattern, showing a
24005 color pattern, a scrolling gradient and a timestamp. This is mainly
24006 intended for testing purposes.
24008 The @code{testsrc2} source is similar to testsrc, but supports more
24009 pixel formats instead of just @code{rgb24}. This allows using it as an
24010 input for other tests without requiring a format conversion.
24012 The @code{yuvtestsrc} source generates an YUV test pattern. You should
24013 see a y, cb and cr stripe from top to bottom.
24015 The sources accept the following parameters:
24020 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
24021 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
24022 pixels to be used as identity matrix for 3D lookup tables. Each component is
24023 coded on a @code{1/(N*N)} scale.
24026 Specify the color of the source, only available in the @code{color}
24027 source. For the syntax of this option, check the
24028 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
24031 Specify the size of the sourced video. For the syntax of this option, check the
24032 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24033 The default value is @code{320x240}.
24035 This option is not available with the @code{allrgb}, @code{allyuv}, and
24036 @code{haldclutsrc} filters.
24039 Specify the frame rate of the sourced video, as the number of frames
24040 generated per second. It has to be a string in the format
24041 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
24042 number or a valid video frame rate abbreviation. The default value is
24046 Set the duration of the sourced video. See
24047 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
24048 for the accepted syntax.
24050 If not specified, or the expressed duration is negative, the video is
24051 supposed to be generated forever.
24053 Since the frame rate is used as time base, all frames including the last one
24054 will have their full duration. If the specified duration is not a multiple
24055 of the frame duration, it will be rounded up.
24058 Set the sample aspect ratio of the sourced video.
24061 Specify the alpha (opacity) of the background, only available in the
24062 @code{testsrc2} source. The value must be between 0 (fully transparent) and
24063 255 (fully opaque, the default).
24066 Set the number of decimals to show in the timestamp, only available in the
24067 @code{testsrc} source.
24069 The displayed timestamp value will correspond to the original
24070 timestamp value multiplied by the power of 10 of the specified
24071 value. Default value is 0.
24074 @subsection Examples
24078 Generate a video with a duration of 5.3 seconds, with size
24079 176x144 and a frame rate of 10 frames per second:
24081 testsrc=duration=5.3:size=qcif:rate=10
24085 The following graph description will generate a red source
24086 with an opacity of 0.2, with size "qcif" and a frame rate of 10
24089 color=c=red@@0.2:s=qcif:r=10
24093 If the input content is to be ignored, @code{nullsrc} can be used. The
24094 following command generates noise in the luminance plane by employing
24095 the @code{geq} filter:
24097 nullsrc=s=256x256, geq=random(1)*255:128:128
24101 @subsection Commands
24103 The @code{color} source supports the following commands:
24107 Set the color of the created image. Accepts the same syntax of the
24108 corresponding @option{color} option.
24113 Generate video using an OpenCL program.
24118 OpenCL program source file.
24121 Kernel name in program.
24124 Size of frames to generate. This must be set.
24127 Pixel format to use for the generated frames. This must be set.
24130 Number of frames generated every second. Default value is '25'.
24134 For details of how the program loading works, see the @ref{program_opencl}
24141 Generate a colour ramp by setting pixel values from the position of the pixel
24142 in the output image. (Note that this will work with all pixel formats, but
24143 the generated output will not be the same.)
24145 __kernel void ramp(__write_only image2d_t dst,
24146 unsigned int index)
24148 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24151 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24153 write_imagef(dst, loc, val);
24158 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24160 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24161 unsigned int index)
24163 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24165 float4 value = 0.0f;
24166 int x = loc.x + index;
24167 int y = loc.y + index;
24168 while (x > 0 || y > 0) {
24169 if (x % 3 == 1 && y % 3 == 1) {
24177 write_imagef(dst, loc, value);
24183 @section sierpinski
24185 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24187 This source accepts the following options:
24191 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24192 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24195 Set frame rate, expressed as number of frames per second. Default
24199 Set seed which is used for random panning.
24202 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24205 Set fractal type, can be default @code{carpet} or @code{triangle}.
24208 @c man end VIDEO SOURCES
24210 @chapter Video Sinks
24211 @c man begin VIDEO SINKS
24213 Below is a description of the currently available video sinks.
24215 @section buffersink
24217 Buffer video frames, and make them available to the end of the filter
24220 This sink is mainly intended for programmatic use, in particular
24221 through the interface defined in @file{libavfilter/buffersink.h}
24222 or the options system.
24224 It accepts a pointer to an AVBufferSinkContext structure, which
24225 defines the incoming buffers' formats, to be passed as the opaque
24226 parameter to @code{avfilter_init_filter} for initialization.
24230 Null video sink: do absolutely nothing with the input video. It is
24231 mainly useful as a template and for use in analysis / debugging
24234 @c man end VIDEO SINKS
24236 @chapter Multimedia Filters
24237 @c man begin MULTIMEDIA FILTERS
24239 Below is a description of the currently available multimedia filters.
24243 Convert input audio to a video output, displaying the audio bit scope.
24245 The filter accepts the following options:
24249 Set frame rate, expressed as number of frames per second. Default
24253 Specify the video size for the output. For the syntax of this option, check the
24254 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24255 Default value is @code{1024x256}.
24258 Specify list of colors separated by space or by '|' which will be used to
24259 draw channels. Unrecognized or missing colors will be replaced
24263 @section adrawgraph
24264 Draw a graph using input audio metadata.
24266 See @ref{drawgraph}
24268 @section agraphmonitor
24270 See @ref{graphmonitor}.
24272 @section ahistogram
24274 Convert input audio to a video output, displaying the volume histogram.
24276 The filter accepts the following options:
24280 Specify how histogram is calculated.
24282 It accepts the following values:
24285 Use single histogram for all channels.
24287 Use separate histogram for each channel.
24289 Default is @code{single}.
24292 Set frame rate, expressed as number of frames per second. Default
24296 Specify the video size for the output. For the syntax of this option, check the
24297 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24298 Default value is @code{hd720}.
24303 It accepts the following values:
24314 reverse logarithmic
24316 Default is @code{log}.
24319 Set amplitude scale.
24321 It accepts the following values:
24328 Default is @code{log}.
24331 Set how much frames to accumulate in histogram.
24332 Default is 1. Setting this to -1 accumulates all frames.
24335 Set histogram ratio of window height.
24338 Set sonogram sliding.
24340 It accepts the following values:
24343 replace old rows with new ones.
24345 scroll from top to bottom.
24347 Default is @code{replace}.
24350 @section aphasemeter
24352 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24353 representing mean phase of current audio frame. A video output can also be produced and is
24354 enabled by default. The audio is passed through as first output.
24356 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24357 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24358 and @code{1} means channels are in phase.
24360 The filter accepts the following options, all related to its video output:
24364 Set the output frame rate. Default value is @code{25}.
24367 Set the video size for the output. For the syntax of this option, check the
24368 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24369 Default value is @code{800x400}.
24374 Specify the red, green, blue contrast. Default values are @code{2},
24375 @code{7} and @code{1}.
24376 Allowed range is @code{[0, 255]}.
24379 Set color which will be used for drawing median phase. If color is
24380 @code{none} which is default, no median phase value will be drawn.
24383 Enable video output. Default is enabled.
24386 @subsection phasing detection
24388 The filter also detects out of phase and mono sequences in stereo streams.
24389 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24391 The filter accepts the following options for this detection:
24395 Enable mono and out of phase detection. Default is disabled.
24398 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24399 Allowed range is @code{[0, 1]}.
24402 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24403 Allowed range is @code{[90, 180]}.
24406 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24409 @subsection Examples
24413 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24415 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24419 @section avectorscope
24421 Convert input audio to a video output, representing the audio vector
24424 The filter is used to measure the difference between channels of stereo
24425 audio stream. A monaural signal, consisting of identical left and right
24426 signal, results in straight vertical line. Any stereo separation is visible
24427 as a deviation from this line, creating a Lissajous figure.
24428 If the straight (or deviation from it) but horizontal line appears this
24429 indicates that the left and right channels are out of phase.
24431 The filter accepts the following options:
24435 Set the vectorscope mode.
24437 Available values are:
24440 Lissajous rotated by 45 degrees.
24443 Same as above but not rotated.
24446 Shape resembling half of circle.
24449 Default value is @samp{lissajous}.
24452 Set the video size for the output. For the syntax of this option, check the
24453 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24454 Default value is @code{400x400}.
24457 Set the output frame rate. Default value is @code{25}.
24463 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24464 @code{160}, @code{80} and @code{255}.
24465 Allowed range is @code{[0, 255]}.
24471 Specify the red, green, blue and alpha fade. Default values are @code{15},
24472 @code{10}, @code{5} and @code{5}.
24473 Allowed range is @code{[0, 255]}.
24476 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24477 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24480 Set the vectorscope drawing mode.
24482 Available values are:
24485 Draw dot for each sample.
24488 Draw line between previous and current sample.
24491 Default value is @samp{dot}.
24494 Specify amplitude scale of audio samples.
24496 Available values are:
24512 Swap left channel axis with right channel axis.
24522 Mirror only x axis.
24525 Mirror only y axis.
24533 @subsection Examples
24537 Complete example using @command{ffplay}:
24539 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24540 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24544 @section bench, abench
24546 Benchmark part of a filtergraph.
24548 The filter accepts the following options:
24552 Start or stop a timer.
24554 Available values are:
24557 Get the current time, set it as frame metadata (using the key
24558 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24561 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24562 the input frame metadata to get the time difference. Time difference, average,
24563 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24564 @code{min}) are then printed. The timestamps are expressed in seconds.
24568 @subsection Examples
24572 Benchmark @ref{selectivecolor} filter:
24574 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24580 Concatenate audio and video streams, joining them together one after the
24583 The filter works on segments of synchronized video and audio streams. All
24584 segments must have the same number of streams of each type, and that will
24585 also be the number of streams at output.
24587 The filter accepts the following options:
24592 Set the number of segments. Default is 2.
24595 Set the number of output video streams, that is also the number of video
24596 streams in each segment. Default is 1.
24599 Set the number of output audio streams, that is also the number of audio
24600 streams in each segment. Default is 0.
24603 Activate unsafe mode: do not fail if segments have a different format.
24607 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24608 @var{a} audio outputs.
24610 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24611 segment, in the same order as the outputs, then the inputs for the second
24614 Related streams do not always have exactly the same duration, for various
24615 reasons including codec frame size or sloppy authoring. For that reason,
24616 related synchronized streams (e.g. a video and its audio track) should be
24617 concatenated at once. The concat filter will use the duration of the longest
24618 stream in each segment (except the last one), and if necessary pad shorter
24619 audio streams with silence.
24621 For this filter to work correctly, all segments must start at timestamp 0.
24623 All corresponding streams must have the same parameters in all segments; the
24624 filtering system will automatically select a common pixel format for video
24625 streams, and a common sample format, sample rate and channel layout for
24626 audio streams, but other settings, such as resolution, must be converted
24627 explicitly by the user.
24629 Different frame rates are acceptable but will result in variable frame rate
24630 at output; be sure to configure the output file to handle it.
24632 @subsection Examples
24636 Concatenate an opening, an episode and an ending, all in bilingual version
24637 (video in stream 0, audio in streams 1 and 2):
24639 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24640 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24641 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24642 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24646 Concatenate two parts, handling audio and video separately, using the
24647 (a)movie sources, and adjusting the resolution:
24649 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24650 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24651 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24653 Note that a desync will happen at the stitch if the audio and video streams
24654 do not have exactly the same duration in the first file.
24658 @subsection Commands
24660 This filter supports the following commands:
24663 Close the current segment and step to the next one
24669 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24670 level. By default, it logs a message at a frequency of 10Hz with the
24671 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24672 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24674 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24675 sample format is double-precision floating point. The input stream will be converted to
24676 this specification, if needed. Users may need to insert aformat and/or aresample filters
24677 after this filter to obtain the original parameters.
24679 The filter also has a video output (see the @var{video} option) with a real
24680 time graph to observe the loudness evolution. The graphic contains the logged
24681 message mentioned above, so it is not printed anymore when this option is set,
24682 unless the verbose logging is set. The main graphing area contains the
24683 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24684 the momentary loudness (400 milliseconds), but can optionally be configured
24685 to instead display short-term loudness (see @var{gauge}).
24687 The green area marks a +/- 1LU target range around the target loudness
24688 (-23LUFS by default, unless modified through @var{target}).
24690 More information about the Loudness Recommendation EBU R128 on
24691 @url{http://tech.ebu.ch/loudness}.
24693 The filter accepts the following options:
24698 Activate the video output. The audio stream is passed unchanged whether this
24699 option is set or no. The video stream will be the first output stream if
24700 activated. Default is @code{0}.
24703 Set the video size. This option is for video only. For the syntax of this
24705 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24706 Default and minimum resolution is @code{640x480}.
24709 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24710 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24711 other integer value between this range is allowed.
24714 Set metadata injection. If set to @code{1}, the audio input will be segmented
24715 into 100ms output frames, each of them containing various loudness information
24716 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24718 Default is @code{0}.
24721 Force the frame logging level.
24723 Available values are:
24726 information logging level
24728 verbose logging level
24731 By default, the logging level is set to @var{info}. If the @option{video} or
24732 the @option{metadata} options are set, it switches to @var{verbose}.
24737 Available modes can be cumulated (the option is a @code{flag} type). Possible
24741 Disable any peak mode (default).
24743 Enable sample-peak mode.
24745 Simple peak mode looking for the higher sample value. It logs a message
24746 for sample-peak (identified by @code{SPK}).
24748 Enable true-peak mode.
24750 If enabled, the peak lookup is done on an over-sampled version of the input
24751 stream for better peak accuracy. It logs a message for true-peak.
24752 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24753 This mode requires a build with @code{libswresample}.
24757 Treat mono input files as "dual mono". If a mono file is intended for playback
24758 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24759 If set to @code{true}, this option will compensate for this effect.
24760 Multi-channel input files are not affected by this option.
24763 Set a specific pan law to be used for the measurement of dual mono files.
24764 This parameter is optional, and has a default value of -3.01dB.
24767 Set a specific target level (in LUFS) used as relative zero in the visualization.
24768 This parameter is optional and has a default value of -23LUFS as specified
24769 by EBU R128. However, material published online may prefer a level of -16LUFS
24770 (e.g. for use with podcasts or video platforms).
24773 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24774 @code{shortterm}. By default the momentary value will be used, but in certain
24775 scenarios it may be more useful to observe the short term value instead (e.g.
24779 Sets the display scale for the loudness. Valid parameters are @code{absolute}
24780 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
24781 video output, not the summary or continuous log output.
24784 @subsection Examples
24788 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
24790 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
24794 Run an analysis with @command{ffmpeg}:
24796 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
24800 @section interleave, ainterleave
24802 Temporally interleave frames from several inputs.
24804 @code{interleave} works with video inputs, @code{ainterleave} with audio.
24806 These filters read frames from several inputs and send the oldest
24807 queued frame to the output.
24809 Input streams must have well defined, monotonically increasing frame
24812 In order to submit one frame to output, these filters need to enqueue
24813 at least one frame for each input, so they cannot work in case one
24814 input is not yet terminated and will not receive incoming frames.
24816 For example consider the case when one input is a @code{select} filter
24817 which always drops input frames. The @code{interleave} filter will keep
24818 reading from that input, but it will never be able to send new frames
24819 to output until the input sends an end-of-stream signal.
24821 Also, depending on inputs synchronization, the filters will drop
24822 frames in case one input receives more frames than the other ones, and
24823 the queue is already filled.
24825 These filters accept the following options:
24829 Set the number of different inputs, it is 2 by default.
24832 How to determine the end-of-stream.
24836 The duration of the longest input. (default)
24839 The duration of the shortest input.
24842 The duration of the first input.
24847 @subsection Examples
24851 Interleave frames belonging to different streams using @command{ffmpeg}:
24853 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
24857 Add flickering blur effect:
24859 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
24863 @section metadata, ametadata
24865 Manipulate frame metadata.
24867 This filter accepts the following options:
24871 Set mode of operation of the filter.
24873 Can be one of the following:
24877 If both @code{value} and @code{key} is set, select frames
24878 which have such metadata. If only @code{key} is set, select
24879 every frame that has such key in metadata.
24882 Add new metadata @code{key} and @code{value}. If key is already available
24886 Modify value of already present key.
24889 If @code{value} is set, delete only keys that have such value.
24890 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
24894 Print key and its value if metadata was found. If @code{key} is not set print all
24895 metadata values available in frame.
24899 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
24902 Set metadata value which will be used. This option is mandatory for
24903 @code{modify} and @code{add} mode.
24906 Which function to use when comparing metadata value and @code{value}.
24908 Can be one of following:
24912 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
24915 Values are interpreted as strings, returns true if metadata value starts with
24916 the @code{value} option string.
24919 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
24922 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
24925 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
24928 Values are interpreted as floats, returns true if expression from option @code{expr}
24932 Values are interpreted as strings, returns true if metadata value ends with
24933 the @code{value} option string.
24937 Set expression which is used when @code{function} is set to @code{expr}.
24938 The expression is evaluated through the eval API and can contain the following
24943 Float representation of @code{value} from metadata key.
24946 Float representation of @code{value} as supplied by user in @code{value} option.
24950 If specified in @code{print} mode, output is written to the named file. Instead of
24951 plain filename any writable url can be specified. Filename ``-'' is a shorthand
24952 for standard output. If @code{file} option is not set, output is written to the log
24953 with AV_LOG_INFO loglevel.
24956 Reduces buffering in print mode when output is written to a URL set using @var{file}.
24960 @subsection Examples
24964 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
24967 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
24970 Print silencedetect output to file @file{metadata.txt}.
24972 silencedetect,ametadata=mode=print:file=metadata.txt
24975 Direct all metadata to a pipe with file descriptor 4.
24977 metadata=mode=print:file='pipe\:4'
24981 @section perms, aperms
24983 Set read/write permissions for the output frames.
24985 These filters are mainly aimed at developers to test direct path in the
24986 following filter in the filtergraph.
24988 The filters accept the following options:
24992 Select the permissions mode.
24994 It accepts the following values:
24997 Do nothing. This is the default.
24999 Set all the output frames read-only.
25001 Set all the output frames directly writable.
25003 Make the frame read-only if writable, and writable if read-only.
25005 Set each output frame read-only or writable randomly.
25009 Set the seed for the @var{random} mode, must be an integer included between
25010 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
25011 @code{-1}, the filter will try to use a good random seed on a best effort
25015 Note: in case of auto-inserted filter between the permission filter and the
25016 following one, the permission might not be received as expected in that
25017 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
25018 perms/aperms filter can avoid this problem.
25020 @section realtime, arealtime
25022 Slow down filtering to match real time approximately.
25024 These filters will pause the filtering for a variable amount of time to
25025 match the output rate with the input timestamps.
25026 They are similar to the @option{re} option to @code{ffmpeg}.
25028 They accept the following options:
25032 Time limit for the pauses. Any pause longer than that will be considered
25033 a timestamp discontinuity and reset the timer. Default is 2 seconds.
25035 Speed factor for processing. The value must be a float larger than zero.
25036 Values larger than 1.0 will result in faster than realtime processing,
25037 smaller will slow processing down. The @var{limit} is automatically adapted
25038 accordingly. Default is 1.0.
25040 A processing speed faster than what is possible without these filters cannot
25045 @section select, aselect
25047 Select frames to pass in output.
25049 This filter accepts the following options:
25054 Set expression, which is evaluated for each input frame.
25056 If the expression is evaluated to zero, the frame is discarded.
25058 If the evaluation result is negative or NaN, the frame is sent to the
25059 first output; otherwise it is sent to the output with index
25060 @code{ceil(val)-1}, assuming that the input index starts from 0.
25062 For example a value of @code{1.2} corresponds to the output with index
25063 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
25066 Set the number of outputs. The output to which to send the selected
25067 frame is based on the result of the evaluation. Default value is 1.
25070 The expression can contain the following constants:
25074 The (sequential) number of the filtered frame, starting from 0.
25077 The (sequential) number of the selected frame, starting from 0.
25079 @item prev_selected_n
25080 The sequential number of the last selected frame. It's NAN if undefined.
25083 The timebase of the input timestamps.
25086 The PTS (Presentation TimeStamp) of the filtered video frame,
25087 expressed in @var{TB} units. It's NAN if undefined.
25090 The PTS of the filtered video frame,
25091 expressed in seconds. It's NAN if undefined.
25094 The PTS of the previously filtered video frame. It's NAN if undefined.
25096 @item prev_selected_pts
25097 The PTS of the last previously filtered video frame. It's NAN if undefined.
25099 @item prev_selected_t
25100 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25103 The PTS of the first video frame in the video. It's NAN if undefined.
25106 The time of the first video frame in the video. It's NAN if undefined.
25108 @item pict_type @emph{(video only)}
25109 The type of the filtered frame. It can assume one of the following
25121 @item interlace_type @emph{(video only)}
25122 The frame interlace type. It can assume one of the following values:
25125 The frame is progressive (not interlaced).
25127 The frame is top-field-first.
25129 The frame is bottom-field-first.
25132 @item consumed_sample_n @emph{(audio only)}
25133 the number of selected samples before the current frame
25135 @item samples_n @emph{(audio only)}
25136 the number of samples in the current frame
25138 @item sample_rate @emph{(audio only)}
25139 the input sample rate
25142 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25145 the position in the file of the filtered frame, -1 if the information
25146 is not available (e.g. for synthetic video)
25148 @item scene @emph{(video only)}
25149 value between 0 and 1 to indicate a new scene; a low value reflects a low
25150 probability for the current frame to introduce a new scene, while a higher
25151 value means the current frame is more likely to be one (see the example below)
25153 @item concatdec_select
25154 The concat demuxer can select only part of a concat input file by setting an
25155 inpoint and an outpoint, but the output packets may not be entirely contained
25156 in the selected interval. By using this variable, it is possible to skip frames
25157 generated by the concat demuxer which are not exactly contained in the selected
25160 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25161 and the @var{lavf.concat.duration} packet metadata values which are also
25162 present in the decoded frames.
25164 The @var{concatdec_select} variable is -1 if the frame pts is at least
25165 start_time and either the duration metadata is missing or the frame pts is less
25166 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25169 That basically means that an input frame is selected if its pts is within the
25170 interval set by the concat demuxer.
25174 The default value of the select expression is "1".
25176 @subsection Examples
25180 Select all frames in input:
25185 The example above is the same as:
25197 Select only I-frames:
25199 select='eq(pict_type\,I)'
25203 Select one frame every 100:
25205 select='not(mod(n\,100))'
25209 Select only frames contained in the 10-20 time interval:
25211 select=between(t\,10\,20)
25215 Select only I-frames contained in the 10-20 time interval:
25217 select=between(t\,10\,20)*eq(pict_type\,I)
25221 Select frames with a minimum distance of 10 seconds:
25223 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25227 Use aselect to select only audio frames with samples number > 100:
25229 aselect='gt(samples_n\,100)'
25233 Create a mosaic of the first scenes:
25235 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25238 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25242 Send even and odd frames to separate outputs, and compose them:
25244 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25248 Select useful frames from an ffconcat file which is using inpoints and
25249 outpoints but where the source files are not intra frame only.
25251 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25255 @section sendcmd, asendcmd
25257 Send commands to filters in the filtergraph.
25259 These filters read commands to be sent to other filters in the
25262 @code{sendcmd} must be inserted between two video filters,
25263 @code{asendcmd} must be inserted between two audio filters, but apart
25264 from that they act the same way.
25266 The specification of commands can be provided in the filter arguments
25267 with the @var{commands} option, or in a file specified by the
25268 @var{filename} option.
25270 These filters accept the following options:
25273 Set the commands to be read and sent to the other filters.
25275 Set the filename of the commands to be read and sent to the other
25279 @subsection Commands syntax
25281 A commands description consists of a sequence of interval
25282 specifications, comprising a list of commands to be executed when a
25283 particular event related to that interval occurs. The occurring event
25284 is typically the current frame time entering or leaving a given time
25287 An interval is specified by the following syntax:
25289 @var{START}[-@var{END}] @var{COMMANDS};
25292 The time interval is specified by the @var{START} and @var{END} times.
25293 @var{END} is optional and defaults to the maximum time.
25295 The current frame time is considered within the specified interval if
25296 it is included in the interval [@var{START}, @var{END}), that is when
25297 the time is greater or equal to @var{START} and is lesser than
25300 @var{COMMANDS} consists of a sequence of one or more command
25301 specifications, separated by ",", relating to that interval. The
25302 syntax of a command specification is given by:
25304 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25307 @var{FLAGS} is optional and specifies the type of events relating to
25308 the time interval which enable sending the specified command, and must
25309 be a non-null sequence of identifier flags separated by "+" or "|" and
25310 enclosed between "[" and "]".
25312 The following flags are recognized:
25315 The command is sent when the current frame timestamp enters the
25316 specified interval. In other words, the command is sent when the
25317 previous frame timestamp was not in the given interval, and the
25321 The command is sent when the current frame timestamp leaves the
25322 specified interval. In other words, the command is sent when the
25323 previous frame timestamp was in the given interval, and the
25327 The command @var{ARG} is interpreted as expression and result of
25328 expression is passed as @var{ARG}.
25330 The expression is evaluated through the eval API and can contain the following
25335 Original position in the file of the frame, or undefined if undefined
25336 for the current frame.
25339 The presentation timestamp in input.
25342 The count of the input frame for video or audio, starting from 0.
25345 The time in seconds of the current frame.
25348 The start time in seconds of the current command interval.
25351 The end time in seconds of the current command interval.
25354 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25359 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25362 @var{TARGET} specifies the target of the command, usually the name of
25363 the filter class or a specific filter instance name.
25365 @var{COMMAND} specifies the name of the command for the target filter.
25367 @var{ARG} is optional and specifies the optional list of argument for
25368 the given @var{COMMAND}.
25370 Between one interval specification and another, whitespaces, or
25371 sequences of characters starting with @code{#} until the end of line,
25372 are ignored and can be used to annotate comments.
25374 A simplified BNF description of the commands specification syntax
25377 @var{COMMAND_FLAG} ::= "enter" | "leave"
25378 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25379 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25380 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25381 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25382 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25385 @subsection Examples
25389 Specify audio tempo change at second 4:
25391 asendcmd=c='4.0 atempo tempo 1.5',atempo
25395 Target a specific filter instance:
25397 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25401 Specify a list of drawtext and hue commands in a file.
25403 # show text in the interval 5-10
25404 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25405 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25407 # desaturate the image in the interval 15-20
25408 15.0-20.0 [enter] hue s 0,
25409 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25411 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25413 # apply an exponential saturation fade-out effect, starting from time 25
25414 25 [enter] hue s exp(25-t)
25417 A filtergraph allowing to read and process the above command list
25418 stored in a file @file{test.cmd}, can be specified with:
25420 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25425 @section setpts, asetpts
25427 Change the PTS (presentation timestamp) of the input frames.
25429 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25431 This filter accepts the following options:
25436 The expression which is evaluated for each frame to construct its timestamp.
25440 The expression is evaluated through the eval API and can contain the following
25444 @item FRAME_RATE, FR
25445 frame rate, only defined for constant frame-rate video
25448 The presentation timestamp in input
25451 The count of the input frame for video or the number of consumed samples,
25452 not including the current frame for audio, starting from 0.
25454 @item NB_CONSUMED_SAMPLES
25455 The number of consumed samples, not including the current frame (only
25458 @item NB_SAMPLES, S
25459 The number of samples in the current frame (only audio)
25461 @item SAMPLE_RATE, SR
25462 The audio sample rate.
25465 The PTS of the first frame.
25468 the time in seconds of the first frame
25471 State whether the current frame is interlaced.
25474 the time in seconds of the current frame
25477 original position in the file of the frame, or undefined if undefined
25478 for the current frame
25481 The previous input PTS.
25484 previous input time in seconds
25487 The previous output PTS.
25490 previous output time in seconds
25493 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25497 The wallclock (RTC) time at the start of the movie in microseconds.
25500 The timebase of the input timestamps.
25504 @subsection Examples
25508 Start counting PTS from zero
25510 setpts=PTS-STARTPTS
25514 Apply fast motion effect:
25520 Apply slow motion effect:
25526 Set fixed rate of 25 frames per second:
25532 Set fixed rate 25 fps with some jitter:
25534 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25538 Apply an offset of 10 seconds to the input PTS:
25544 Generate timestamps from a "live source" and rebase onto the current timebase:
25546 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25550 Generate timestamps by counting samples:
25559 Force color range for the output video frame.
25561 The @code{setrange} filter marks the color range property for the
25562 output frames. It does not change the input frame, but only sets the
25563 corresponding property, which affects how the frame is treated by
25566 The filter accepts the following options:
25571 Available values are:
25575 Keep the same color range property.
25577 @item unspecified, unknown
25578 Set the color range as unspecified.
25580 @item limited, tv, mpeg
25581 Set the color range as limited.
25583 @item full, pc, jpeg
25584 Set the color range as full.
25588 @section settb, asettb
25590 Set the timebase to use for the output frames timestamps.
25591 It is mainly useful for testing timebase configuration.
25593 It accepts the following parameters:
25598 The expression which is evaluated into the output timebase.
25602 The value for @option{tb} is an arithmetic expression representing a
25603 rational. The expression can contain the constants "AVTB" (the default
25604 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25605 audio only). Default value is "intb".
25607 @subsection Examples
25611 Set the timebase to 1/25:
25617 Set the timebase to 1/10:
25623 Set the timebase to 1001/1000:
25629 Set the timebase to 2*intb:
25635 Set the default timebase value:
25642 Convert input audio to a video output representing frequency spectrum
25643 logarithmically using Brown-Puckette constant Q transform algorithm with
25644 direct frequency domain coefficient calculation (but the transform itself
25645 is not really constant Q, instead the Q factor is actually variable/clamped),
25646 with musical tone scale, from E0 to D#10.
25648 The filter accepts the following options:
25652 Specify the video size for the output. It must be even. For the syntax of this option,
25653 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25654 Default value is @code{1920x1080}.
25657 Set the output frame rate. Default value is @code{25}.
25660 Set the bargraph height. It must be even. Default value is @code{-1} which
25661 computes the bargraph height automatically.
25664 Set the axis height. It must be even. Default value is @code{-1} which computes
25665 the axis height automatically.
25668 Set the sonogram height. It must be even. Default value is @code{-1} which
25669 computes the sonogram height automatically.
25672 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25673 instead. Default value is @code{1}.
25675 @item sono_v, volume
25676 Specify the sonogram volume expression. It can contain variables:
25679 the @var{bar_v} evaluated expression
25680 @item frequency, freq, f
25681 the frequency where it is evaluated
25682 @item timeclamp, tc
25683 the value of @var{timeclamp} option
25687 @item a_weighting(f)
25688 A-weighting of equal loudness
25689 @item b_weighting(f)
25690 B-weighting of equal loudness
25691 @item c_weighting(f)
25692 C-weighting of equal loudness.
25694 Default value is @code{16}.
25696 @item bar_v, volume2
25697 Specify the bargraph volume expression. It can contain variables:
25700 the @var{sono_v} evaluated expression
25701 @item frequency, freq, f
25702 the frequency where it is evaluated
25703 @item timeclamp, tc
25704 the value of @var{timeclamp} option
25708 @item a_weighting(f)
25709 A-weighting of equal loudness
25710 @item b_weighting(f)
25711 B-weighting of equal loudness
25712 @item c_weighting(f)
25713 C-weighting of equal loudness.
25715 Default value is @code{sono_v}.
25717 @item sono_g, gamma
25718 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25719 higher gamma makes the spectrum having more range. Default value is @code{3}.
25720 Acceptable range is @code{[1, 7]}.
25722 @item bar_g, gamma2
25723 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25727 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25728 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25730 @item timeclamp, tc
25731 Specify the transform timeclamp. At low frequency, there is trade-off between
25732 accuracy in time domain and frequency domain. If timeclamp is lower,
25733 event in time domain is represented more accurately (such as fast bass drum),
25734 otherwise event in frequency domain is represented more accurately
25735 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25738 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25739 limits future samples by applying asymmetric windowing in time domain, useful
25740 when low latency is required. Accepted range is @code{[0, 1]}.
25743 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25744 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25747 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25748 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25751 This option is deprecated and ignored.
25754 Specify the transform length in time domain. Use this option to control accuracy
25755 trade-off between time domain and frequency domain at every frequency sample.
25756 It can contain variables:
25758 @item frequency, freq, f
25759 the frequency where it is evaluated
25760 @item timeclamp, tc
25761 the value of @var{timeclamp} option.
25763 Default value is @code{384*tc/(384+tc*f)}.
25766 Specify the transform count for every video frame. Default value is @code{6}.
25767 Acceptable range is @code{[1, 30]}.
25770 Specify the transform count for every single pixel. Default value is @code{0},
25771 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25774 Specify font file for use with freetype to draw the axis. If not specified,
25775 use embedded font. Note that drawing with font file or embedded font is not
25776 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
25780 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
25781 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
25785 Specify font color expression. This is arithmetic expression that should return
25786 integer value 0xRRGGBB. It can contain variables:
25788 @item frequency, freq, f
25789 the frequency where it is evaluated
25790 @item timeclamp, tc
25791 the value of @var{timeclamp} option
25796 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
25797 @item r(x), g(x), b(x)
25798 red, green, and blue value of intensity x.
25800 Default value is @code{st(0, (midi(f)-59.5)/12);
25801 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
25802 r(1-ld(1)) + b(ld(1))}.
25805 Specify image file to draw the axis. This option override @var{fontfile} and
25806 @var{fontcolor} option.
25809 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
25810 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
25811 Default value is @code{1}.
25814 Set colorspace. The accepted values are:
25817 Unspecified (default)
25826 BT.470BG or BT.601-6 625
25829 SMPTE-170M or BT.601-6 525
25835 BT.2020 with non-constant luminance
25840 Set spectrogram color scheme. This is list of floating point values with format
25841 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
25842 The default is @code{1|0.5|0|0|0.5|1}.
25846 @subsection Examples
25850 Playing audio while showing the spectrum:
25852 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
25856 Same as above, but with frame rate 30 fps:
25858 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
25862 Playing at 1280x720:
25864 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
25868 Disable sonogram display:
25874 A1 and its harmonics: A1, A2, (near)E3, A3:
25876 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),
25877 asplit[a][out1]; [a] showcqt [out0]'
25881 Same as above, but with more accuracy in frequency domain:
25883 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),
25884 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
25890 bar_v=10:sono_v=bar_v*a_weighting(f)
25894 Custom gamma, now spectrum is linear to the amplitude.
25900 Custom tlength equation:
25902 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)))'
25906 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
25908 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
25912 Custom font using fontconfig:
25914 font='Courier New,Monospace,mono|bold'
25918 Custom frequency range with custom axis using image file:
25920 axisfile=myaxis.png:basefreq=40:endfreq=10000
25926 Convert input audio to video output representing the audio power spectrum.
25927 Audio amplitude is on Y-axis while frequency is on X-axis.
25929 The filter accepts the following options:
25933 Specify size of video. For the syntax of this option, check the
25934 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25935 Default is @code{1024x512}.
25939 This set how each frequency bin will be represented.
25941 It accepts the following values:
25947 Default is @code{bar}.
25950 Set amplitude scale.
25952 It accepts the following values:
25966 Default is @code{log}.
25969 Set frequency scale.
25971 It accepts the following values:
25980 Reverse logarithmic scale.
25982 Default is @code{lin}.
25985 Set window size. Allowed range is from 16 to 65536.
25987 Default is @code{2048}
25990 Set windowing function.
25992 It accepts the following values:
26015 Default is @code{hanning}.
26018 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26019 which means optimal overlap for selected window function will be picked.
26022 Set time averaging. Setting this to 0 will display current maximal peaks.
26023 Default is @code{1}, which means time averaging is disabled.
26026 Specify list of colors separated by space or by '|' which will be used to
26027 draw channel frequencies. Unrecognized or missing colors will be replaced
26031 Set channel display mode.
26033 It accepts the following values:
26038 Default is @code{combined}.
26041 Set minimum amplitude used in @code{log} amplitude scaler.
26044 Set data display mode.
26046 It accepts the following values:
26052 Default is @code{magnitude}.
26055 @section showspatial
26057 Convert stereo input audio to a video output, representing the spatial relationship
26058 between two channels.
26060 The filter accepts the following options:
26064 Specify the video size for the output. For the syntax of this option, check the
26065 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26066 Default value is @code{512x512}.
26069 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
26072 Set window function.
26074 It accepts the following values:
26099 Default value is @code{hann}.
26102 Set ratio of overlap window. Default value is @code{0.5}.
26103 When value is @code{1} overlap is set to recommended size for specific
26104 window function currently used.
26107 @anchor{showspectrum}
26108 @section showspectrum
26110 Convert input audio to a video output, representing the audio frequency
26113 The filter accepts the following options:
26117 Specify the video size for the output. For the syntax of this option, check the
26118 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26119 Default value is @code{640x512}.
26122 Specify how the spectrum should slide along the window.
26124 It accepts the following values:
26127 the samples start again on the left when they reach the right
26129 the samples scroll from right to left
26131 frames are only produced when the samples reach the right
26133 the samples scroll from left to right
26136 Default value is @code{replace}.
26139 Specify display mode.
26141 It accepts the following values:
26144 all channels are displayed in the same row
26146 all channels are displayed in separate rows
26149 Default value is @samp{combined}.
26152 Specify display color mode.
26154 It accepts the following values:
26157 each channel is displayed in a separate color
26159 each channel is displayed using the same color scheme
26161 each channel is displayed using the rainbow color scheme
26163 each channel is displayed using the moreland color scheme
26165 each channel is displayed using the nebulae color scheme
26167 each channel is displayed using the fire color scheme
26169 each channel is displayed using the fiery color scheme
26171 each channel is displayed using the fruit color scheme
26173 each channel is displayed using the cool color scheme
26175 each channel is displayed using the magma color scheme
26177 each channel is displayed using the green color scheme
26179 each channel is displayed using the viridis color scheme
26181 each channel is displayed using the plasma color scheme
26183 each channel is displayed using the cividis color scheme
26185 each channel is displayed using the terrain color scheme
26188 Default value is @samp{channel}.
26191 Specify scale used for calculating intensity color values.
26193 It accepts the following values:
26198 square root, default
26209 Default value is @samp{sqrt}.
26212 Specify frequency scale.
26214 It accepts the following values:
26222 Default value is @samp{lin}.
26225 Set saturation modifier for displayed colors. Negative values provide
26226 alternative color scheme. @code{0} is no saturation at all.
26227 Saturation must be in [-10.0, 10.0] range.
26228 Default value is @code{1}.
26231 Set window function.
26233 It accepts the following values:
26258 Default value is @code{hann}.
26261 Set orientation of time vs frequency axis. Can be @code{vertical} or
26262 @code{horizontal}. Default is @code{vertical}.
26265 Set ratio of overlap window. Default value is @code{0}.
26266 When value is @code{1} overlap is set to recommended size for specific
26267 window function currently used.
26270 Set scale gain for calculating intensity color values.
26271 Default value is @code{1}.
26274 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26277 Set color rotation, must be in [-1.0, 1.0] range.
26278 Default value is @code{0}.
26281 Set start frequency from which to display spectrogram. Default is @code{0}.
26284 Set stop frequency to which to display spectrogram. Default is @code{0}.
26287 Set upper frame rate limit. Default is @code{auto}, unlimited.
26290 Draw time and frequency axes and legends. Default is disabled.
26293 The usage is very similar to the showwaves filter; see the examples in that
26296 @subsection Examples
26300 Large window with logarithmic color scaling:
26302 showspectrum=s=1280x480:scale=log
26306 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26308 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26309 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26313 @section showspectrumpic
26315 Convert input audio to a single video frame, representing the audio frequency
26318 The filter accepts the following options:
26322 Specify the video size for the output. For the syntax of this option, check the
26323 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26324 Default value is @code{4096x2048}.
26327 Specify display mode.
26329 It accepts the following values:
26332 all channels are displayed in the same row
26334 all channels are displayed in separate rows
26336 Default value is @samp{combined}.
26339 Specify display color mode.
26341 It accepts the following values:
26344 each channel is displayed in a separate color
26346 each channel is displayed using the same color scheme
26348 each channel is displayed using the rainbow color scheme
26350 each channel is displayed using the moreland color scheme
26352 each channel is displayed using the nebulae color scheme
26354 each channel is displayed using the fire color scheme
26356 each channel is displayed using the fiery color scheme
26358 each channel is displayed using the fruit color scheme
26360 each channel is displayed using the cool color scheme
26362 each channel is displayed using the magma color scheme
26364 each channel is displayed using the green color scheme
26366 each channel is displayed using the viridis color scheme
26368 each channel is displayed using the plasma color scheme
26370 each channel is displayed using the cividis color scheme
26372 each channel is displayed using the terrain color scheme
26374 Default value is @samp{intensity}.
26377 Specify scale used for calculating intensity color values.
26379 It accepts the following values:
26384 square root, default
26394 Default value is @samp{log}.
26397 Specify frequency scale.
26399 It accepts the following values:
26407 Default value is @samp{lin}.
26410 Set saturation modifier for displayed colors. Negative values provide
26411 alternative color scheme. @code{0} is no saturation at all.
26412 Saturation must be in [-10.0, 10.0] range.
26413 Default value is @code{1}.
26416 Set window function.
26418 It accepts the following values:
26442 Default value is @code{hann}.
26445 Set orientation of time vs frequency axis. Can be @code{vertical} or
26446 @code{horizontal}. Default is @code{vertical}.
26449 Set scale gain for calculating intensity color values.
26450 Default value is @code{1}.
26453 Draw time and frequency axes and legends. Default is enabled.
26456 Set color rotation, must be in [-1.0, 1.0] range.
26457 Default value is @code{0}.
26460 Set start frequency from which to display spectrogram. Default is @code{0}.
26463 Set stop frequency to which to display spectrogram. Default is @code{0}.
26466 @subsection Examples
26470 Extract an audio spectrogram of a whole audio track
26471 in a 1024x1024 picture using @command{ffmpeg}:
26473 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26477 @section showvolume
26479 Convert input audio volume to a video output.
26481 The filter accepts the following options:
26488 Set border width, allowed range is [0, 5]. Default is 1.
26491 Set channel width, allowed range is [80, 8192]. Default is 400.
26494 Set channel height, allowed range is [1, 900]. Default is 20.
26497 Set fade, allowed range is [0, 1]. Default is 0.95.
26500 Set volume color expression.
26502 The expression can use the following variables:
26506 Current max volume of channel in dB.
26512 Current channel number, starting from 0.
26516 If set, displays channel names. Default is enabled.
26519 If set, displays volume values. Default is enabled.
26522 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26523 default is @code{h}.
26526 Set step size, allowed range is [0, 5]. Default is 0, which means
26530 Set background opacity, allowed range is [0, 1]. Default is 0.
26533 Set metering mode, can be peak: @code{p} or rms: @code{r},
26534 default is @code{p}.
26537 Set display scale, can be linear: @code{lin} or log: @code{log},
26538 default is @code{lin}.
26542 If set to > 0., display a line for the max level
26543 in the previous seconds.
26544 default is disabled: @code{0.}
26547 The color of the max line. Use when @code{dm} option is set to > 0.
26548 default is: @code{orange}
26553 Convert input audio to a video output, representing the samples waves.
26555 The filter accepts the following options:
26559 Specify the video size for the output. For the syntax of this option, check the
26560 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26561 Default value is @code{600x240}.
26566 Available values are:
26569 Draw a point for each sample.
26572 Draw a vertical line for each sample.
26575 Draw a point for each sample and a line between them.
26578 Draw a centered vertical line for each sample.
26581 Default value is @code{point}.
26584 Set the number of samples which are printed on the same column. A
26585 larger value will decrease the frame rate. Must be a positive
26586 integer. This option can be set only if the value for @var{rate}
26587 is not explicitly specified.
26590 Set the (approximate) output frame rate. This is done by setting the
26591 option @var{n}. Default value is "25".
26593 @item split_channels
26594 Set if channels should be drawn separately or overlap. Default value is 0.
26597 Set colors separated by '|' which are going to be used for drawing of each channel.
26600 Set amplitude scale.
26602 Available values are:
26620 Set the draw mode. This is mostly useful to set for high @var{n}.
26622 Available values are:
26625 Scale pixel values for each drawn sample.
26628 Draw every sample directly.
26631 Default value is @code{scale}.
26634 @subsection Examples
26638 Output the input file audio and the corresponding video representation
26641 amovie=a.mp3,asplit[out0],showwaves[out1]
26645 Create a synthetic signal and show it with showwaves, forcing a
26646 frame rate of 30 frames per second:
26648 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26652 @section showwavespic
26654 Convert input audio to a single video frame, representing the samples waves.
26656 The filter accepts the following options:
26660 Specify the video size for the output. For the syntax of this option, check the
26661 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26662 Default value is @code{600x240}.
26664 @item split_channels
26665 Set if channels should be drawn separately or overlap. Default value is 0.
26668 Set colors separated by '|' which are going to be used for drawing of each channel.
26671 Set amplitude scale.
26673 Available values are:
26693 Available values are:
26696 Scale pixel values for each drawn sample.
26699 Draw every sample directly.
26702 Default value is @code{scale}.
26705 Set the filter mode.
26707 Available values are:
26710 Use average samples values for each drawn sample.
26713 Use peak samples values for each drawn sample.
26716 Default value is @code{average}.
26719 @subsection Examples
26723 Extract a channel split representation of the wave form of a whole audio track
26724 in a 1024x800 picture using @command{ffmpeg}:
26726 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26730 @section sidedata, asidedata
26732 Delete frame side data, or select frames based on it.
26734 This filter accepts the following options:
26738 Set mode of operation of the filter.
26740 Can be one of the following:
26744 Select every frame with side data of @code{type}.
26747 Delete side data of @code{type}. If @code{type} is not set, delete all side
26753 Set side data type used with all modes. Must be set for @code{select} mode. For
26754 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26755 in @file{libavutil/frame.h}. For example, to choose
26756 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26760 @section spectrumsynth
26762 Synthesize audio from 2 input video spectrums, first input stream represents
26763 magnitude across time and second represents phase across time.
26764 The filter will transform from frequency domain as displayed in videos back
26765 to time domain as presented in audio output.
26767 This filter is primarily created for reversing processed @ref{showspectrum}
26768 filter outputs, but can synthesize sound from other spectrograms too.
26769 But in such case results are going to be poor if the phase data is not
26770 available, because in such cases phase data need to be recreated, usually
26771 it's just recreated from random noise.
26772 For best results use gray only output (@code{channel} color mode in
26773 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26774 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26775 @code{data} option. Inputs videos should generally use @code{fullframe}
26776 slide mode as that saves resources needed for decoding video.
26778 The filter accepts the following options:
26782 Specify sample rate of output audio, the sample rate of audio from which
26783 spectrum was generated may differ.
26786 Set number of channels represented in input video spectrums.
26789 Set scale which was used when generating magnitude input spectrum.
26790 Can be @code{lin} or @code{log}. Default is @code{log}.
26793 Set slide which was used when generating inputs spectrums.
26794 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
26795 Default is @code{fullframe}.
26798 Set window function used for resynthesis.
26801 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26802 which means optimal overlap for selected window function will be picked.
26805 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
26806 Default is @code{vertical}.
26809 @subsection Examples
26813 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
26814 then resynthesize videos back to audio with spectrumsynth:
26816 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
26817 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
26818 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
26822 @section split, asplit
26824 Split input into several identical outputs.
26826 @code{asplit} works with audio input, @code{split} with video.
26828 The filter accepts a single parameter which specifies the number of outputs. If
26829 unspecified, it defaults to 2.
26831 @subsection Examples
26835 Create two separate outputs from the same input:
26837 [in] split [out0][out1]
26841 To create 3 or more outputs, you need to specify the number of
26844 [in] asplit=3 [out0][out1][out2]
26848 Create two separate outputs from the same input, one cropped and
26851 [in] split [splitout1][splitout2];
26852 [splitout1] crop=100:100:0:0 [cropout];
26853 [splitout2] pad=200:200:100:100 [padout];
26857 Create 5 copies of the input audio with @command{ffmpeg}:
26859 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
26865 Receive commands sent through a libzmq client, and forward them to
26866 filters in the filtergraph.
26868 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
26869 must be inserted between two video filters, @code{azmq} between two
26870 audio filters. Both are capable to send messages to any filter type.
26872 To enable these filters you need to install the libzmq library and
26873 headers and configure FFmpeg with @code{--enable-libzmq}.
26875 For more information about libzmq see:
26876 @url{http://www.zeromq.org/}
26878 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
26879 receives messages sent through a network interface defined by the
26880 @option{bind_address} (or the abbreviation "@option{b}") option.
26881 Default value of this option is @file{tcp://localhost:5555}. You may
26882 want to alter this value to your needs, but do not forget to escape any
26883 ':' signs (see @ref{filtergraph escaping}).
26885 The received message must be in the form:
26887 @var{TARGET} @var{COMMAND} [@var{ARG}]
26890 @var{TARGET} specifies the target of the command, usually the name of
26891 the filter class or a specific filter instance name. The default
26892 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
26893 but you can override this by using the @samp{filter_name@@id} syntax
26894 (see @ref{Filtergraph syntax}).
26896 @var{COMMAND} specifies the name of the command for the target filter.
26898 @var{ARG} is optional and specifies the optional argument list for the
26899 given @var{COMMAND}.
26901 Upon reception, the message is processed and the corresponding command
26902 is injected into the filtergraph. Depending on the result, the filter
26903 will send a reply to the client, adopting the format:
26905 @var{ERROR_CODE} @var{ERROR_REASON}
26909 @var{MESSAGE} is optional.
26911 @subsection Examples
26913 Look at @file{tools/zmqsend} for an example of a zmq client which can
26914 be used to send commands processed by these filters.
26916 Consider the following filtergraph generated by @command{ffplay}.
26917 In this example the last overlay filter has an instance name. All other
26918 filters will have default instance names.
26921 ffplay -dumpgraph 1 -f lavfi "
26922 color=s=100x100:c=red [l];
26923 color=s=100x100:c=blue [r];
26924 nullsrc=s=200x100, zmq [bg];
26925 [bg][l] overlay [bg+l];
26926 [bg+l][r] overlay@@my=x=100 "
26929 To change the color of the left side of the video, the following
26930 command can be used:
26932 echo Parsed_color_0 c yellow | tools/zmqsend
26935 To change the right side:
26937 echo Parsed_color_1 c pink | tools/zmqsend
26940 To change the position of the right side:
26942 echo overlay@@my x 150 | tools/zmqsend
26946 @c man end MULTIMEDIA FILTERS
26948 @chapter Multimedia Sources
26949 @c man begin MULTIMEDIA SOURCES
26951 Below is a description of the currently available multimedia sources.
26955 This is the same as @ref{movie} source, except it selects an audio
26961 Read audio and/or video stream(s) from a movie container.
26963 It accepts the following parameters:
26967 The name of the resource to read (not necessarily a file; it can also be a
26968 device or a stream accessed through some protocol).
26970 @item format_name, f
26971 Specifies the format assumed for the movie to read, and can be either
26972 the name of a container or an input device. If not specified, the
26973 format is guessed from @var{movie_name} or by probing.
26975 @item seek_point, sp
26976 Specifies the seek point in seconds. The frames will be output
26977 starting from this seek point. The parameter is evaluated with
26978 @code{av_strtod}, so the numerical value may be suffixed by an IS
26979 postfix. The default value is "0".
26982 Specifies the streams to read. Several streams can be specified,
26983 separated by "+". The source will then have as many outputs, in the
26984 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
26985 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
26986 respectively the default (best suited) video and audio stream. Default
26987 is "dv", or "da" if the filter is called as "amovie".
26989 @item stream_index, si
26990 Specifies the index of the video stream to read. If the value is -1,
26991 the most suitable video stream will be automatically selected. The default
26992 value is "-1". Deprecated. If the filter is called "amovie", it will select
26993 audio instead of video.
26996 Specifies how many times to read the stream in sequence.
26997 If the value is 0, the stream will be looped infinitely.
26998 Default value is "1".
27000 Note that when the movie is looped the source timestamps are not
27001 changed, so it will generate non monotonically increasing timestamps.
27003 @item discontinuity
27004 Specifies the time difference between frames above which the point is
27005 considered a timestamp discontinuity which is removed by adjusting the later
27009 It allows overlaying a second video on top of the main input of
27010 a filtergraph, as shown in this graph:
27012 input -----------> deltapts0 --> overlay --> output
27015 movie --> scale--> deltapts1 -------+
27017 @subsection Examples
27021 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
27022 on top of the input labelled "in":
27024 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
27025 [in] setpts=PTS-STARTPTS [main];
27026 [main][over] overlay=16:16 [out]
27030 Read from a video4linux2 device, and overlay it on top of the input
27033 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
27034 [in] setpts=PTS-STARTPTS [main];
27035 [main][over] overlay=16:16 [out]
27039 Read the first video stream and the audio stream with id 0x81 from
27040 dvd.vob; the video is connected to the pad named "video" and the audio is
27041 connected to the pad named "audio":
27043 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
27047 @subsection Commands
27049 Both movie and amovie support the following commands:
27052 Perform seek using "av_seek_frame".
27053 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
27056 @var{stream_index}: If stream_index is -1, a default
27057 stream is selected, and @var{timestamp} is automatically converted
27058 from AV_TIME_BASE units to the stream specific time_base.
27060 @var{timestamp}: Timestamp in AVStream.time_base units
27061 or, if no stream is specified, in AV_TIME_BASE units.
27063 @var{flags}: Flags which select direction and seeking mode.
27067 Get movie duration in AV_TIME_BASE units.
27071 @c man end MULTIMEDIA SOURCES