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 @subsection Commands
1886 This filter supports the following commands:
1889 Syntax is same as option with same name.
1894 Multiply first audio stream with second audio stream and store result
1895 in output audio stream. Multiplication is done by multiplying each
1896 sample from first stream with sample at same position from second stream.
1898 With this element-wise multiplication one can create amplitude fades and
1899 amplitude modulations.
1901 @section anequalizer
1903 High-order parametric multiband equalizer for each channel.
1905 It accepts the following parameters:
1909 This option string is in format:
1910 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1911 Each equalizer band is separated by '|'.
1915 Set channel number to which equalization will be applied.
1916 If input doesn't have that channel the entry is ignored.
1919 Set central frequency for band.
1920 If input doesn't have that frequency the entry is ignored.
1923 Set band width in Hertz.
1926 Set band gain in dB.
1929 Set filter type for band, optional, can be:
1933 Butterworth, this is default.
1944 With this option activated frequency response of anequalizer is displayed
1948 Set video stream size. Only useful if curves option is activated.
1951 Set max gain that will be displayed. Only useful if curves option is activated.
1952 Setting this to a reasonable value makes it possible to display gain which is derived from
1953 neighbour bands which are too close to each other and thus produce higher gain
1954 when both are activated.
1957 Set frequency scale used to draw frequency response in video output.
1958 Can be linear or logarithmic. Default is logarithmic.
1961 Set color for each channel curve which is going to be displayed in video stream.
1962 This is list of color names separated by space or by '|'.
1963 Unrecognised or missing colors will be replaced by white color.
1966 @subsection Examples
1970 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1971 for first 2 channels using Chebyshev type 1 filter:
1973 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1977 @subsection Commands
1979 This filter supports the following commands:
1982 Alter existing filter parameters.
1983 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1985 @var{fN} is existing filter number, starting from 0, if no such filter is available
1987 @var{freq} set new frequency parameter.
1988 @var{width} set new width parameter in Hertz.
1989 @var{gain} set new gain parameter in dB.
1991 Full filter invocation with asendcmd may look like this:
1992 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1997 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1999 Each sample is adjusted by looking for other samples with similar contexts. This
2000 context similarity is defined by comparing their surrounding patches of size
2001 @option{p}. Patches are searched in an area of @option{r} around the sample.
2003 The filter accepts the following options:
2007 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
2010 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
2011 Default value is 2 milliseconds.
2014 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
2015 Default value is 6 milliseconds.
2018 Set the output mode.
2020 It accepts the following values:
2023 Pass input unchanged.
2026 Pass noise filtered out.
2031 Default value is @var{o}.
2035 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
2038 @subsection Commands
2040 This filter supports the all above options as @ref{commands}.
2043 Apply Normalized Least-Mean-Squares algorithm to the first audio stream using the second audio stream.
2045 This adaptive filter is used to mimic a desired filter by finding the filter coefficients that
2046 relate to producing the least mean square of the error signal (difference between the desired,
2047 2nd input audio stream and the actual signal, the 1st input audio stream).
2049 A description of the accepted options follows.
2062 Set the filter leakage.
2065 It accepts the following values:
2074 Pass filtered samples.
2077 Pass difference between desired and filtered samples.
2079 Default value is @var{o}.
2083 @subsection Examples
2087 One of many usages of this filter is noise reduction, input audio is filtered
2088 with same samples that are delayed by fixed amount, one such example for stereo audio is:
2090 asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
2094 @subsection Commands
2096 This filter supports the same commands as options, excluding option @code{order}.
2100 Pass the audio source unchanged to the output.
2104 Pad the end of an audio stream with silence.
2106 This can be used together with @command{ffmpeg} @option{-shortest} to
2107 extend audio streams to the same length as the video stream.
2109 A description of the accepted options follows.
2113 Set silence packet size. Default value is 4096.
2116 Set the number of samples of silence to add to the end. After the
2117 value is reached, the stream is terminated. This option is mutually
2118 exclusive with @option{whole_len}.
2121 Set the minimum total number of samples in the output audio stream. If
2122 the value is longer than the input audio length, silence is added to
2123 the end, until the value is reached. This option is mutually exclusive
2124 with @option{pad_len}.
2127 Specify the duration of samples of silence to add. See
2128 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2129 for the accepted syntax. Used only if set to non-zero value.
2132 Specify the minimum total duration in the output audio stream. See
2133 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2134 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
2135 the input audio length, silence is added to the end, until the value is reached.
2136 This option is mutually exclusive with @option{pad_dur}
2139 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
2140 nor @option{whole_dur} option is set, the filter will add silence to the end of
2141 the input stream indefinitely.
2143 @subsection Examples
2147 Add 1024 samples of silence to the end of the input:
2153 Make sure the audio output will contain at least 10000 samples, pad
2154 the input with silence if required:
2156 apad=whole_len=10000
2160 Use @command{ffmpeg} to pad the audio input with silence, so that the
2161 video stream will always result the shortest and will be converted
2162 until the end in the output file when using the @option{shortest}
2165 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
2170 Add a phasing effect to the input audio.
2172 A phaser filter creates series of peaks and troughs in the frequency spectrum.
2173 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
2175 A description of the accepted parameters follows.
2179 Set input gain. Default is 0.4.
2182 Set output gain. Default is 0.74
2185 Set delay in milliseconds. Default is 3.0.
2188 Set decay. Default is 0.4.
2191 Set modulation speed in Hz. Default is 0.5.
2194 Set modulation type. Default is triangular.
2196 It accepts the following values:
2203 @section aphaseshift
2204 Apply phase shift to input audio samples.
2206 The filter accepts the following options:
2210 Specify phase shift. Allowed range is from -1.0 to 1.0.
2211 Default value is 0.0.
2214 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
2215 Default value is 1.0.
2218 @subsection Commands
2220 This filter supports the all above options as @ref{commands}.
2224 Audio pulsator is something between an autopanner and a tremolo.
2225 But it can produce funny stereo effects as well. Pulsator changes the volume
2226 of the left and right channel based on a LFO (low frequency oscillator) with
2227 different waveforms and shifted phases.
2228 This filter have the ability to define an offset between left and right
2229 channel. An offset of 0 means that both LFO shapes match each other.
2230 The left and right channel are altered equally - a conventional tremolo.
2231 An offset of 50% means that the shape of the right channel is exactly shifted
2232 in phase (or moved backwards about half of the frequency) - pulsator acts as
2233 an autopanner. At 1 both curves match again. Every setting in between moves the
2234 phase shift gapless between all stages and produces some "bypassing" sounds with
2235 sine and triangle waveforms. The more you set the offset near 1 (starting from
2236 the 0.5) the faster the signal passes from the left to the right speaker.
2238 The filter accepts the following options:
2242 Set input gain. By default it is 1. Range is [0.015625 - 64].
2245 Set output gain. By default it is 1. Range is [0.015625 - 64].
2248 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
2249 sawup or sawdown. Default is sine.
2252 Set modulation. Define how much of original signal is affected by the LFO.
2255 Set left channel offset. Default is 0. Allowed range is [0 - 1].
2258 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
2261 Set pulse width. Default is 1. Allowed range is [0 - 2].
2264 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
2267 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
2271 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
2275 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
2276 if timing is set to hz.
2282 Resample the input audio to the specified parameters, using the
2283 libswresample library. If none are specified then the filter will
2284 automatically convert between its input and output.
2286 This filter is also able to stretch/squeeze the audio data to make it match
2287 the timestamps or to inject silence / cut out audio to make it match the
2288 timestamps, do a combination of both or do neither.
2290 The filter accepts the syntax
2291 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
2292 expresses a sample rate and @var{resampler_options} is a list of
2293 @var{key}=@var{value} pairs, separated by ":". See the
2294 @ref{Resampler Options,,"Resampler Options" section in the
2295 ffmpeg-resampler(1) manual,ffmpeg-resampler}
2296 for the complete list of supported options.
2298 @subsection Examples
2302 Resample the input audio to 44100Hz:
2308 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2309 samples per second compensation:
2311 aresample=async=1000
2317 Reverse an audio clip.
2319 Warning: This filter requires memory to buffer the entire clip, so trimming
2322 @subsection Examples
2326 Take the first 5 seconds of a clip, and reverse it.
2328 atrim=end=5,areverse
2334 Reduce noise from speech using Recurrent Neural Networks.
2336 This filter accepts the following options:
2340 Set train model file to load. This option is always required.
2343 Set how much to mix filtered samples into final output.
2344 Allowed range is from -1 to 1. Default value is 1.
2345 Negative values are special, they set how much to keep filtered noise
2346 in the final filter output. Set this option to -1 to hear actual
2347 noise removed from input signal.
2350 @section asetnsamples
2352 Set the number of samples per each output audio frame.
2354 The last output packet may contain a different number of samples, as
2355 the filter will flush all the remaining samples when the input audio
2358 The filter accepts the following options:
2362 @item nb_out_samples, n
2363 Set the number of frames per each output audio frame. The number is
2364 intended as the number of samples @emph{per each channel}.
2365 Default value is 1024.
2368 If set to 1, the filter will pad the last audio frame with zeroes, so
2369 that the last frame will contain the same number of samples as the
2370 previous ones. Default value is 1.
2373 For example, to set the number of per-frame samples to 1234 and
2374 disable padding for the last frame, use:
2376 asetnsamples=n=1234:p=0
2381 Set the sample rate without altering the PCM data.
2382 This will result in a change of speed and pitch.
2384 The filter accepts the following options:
2387 @item sample_rate, r
2388 Set the output sample rate. Default is 44100 Hz.
2393 Show a line containing various information for each input audio frame.
2394 The input audio is not modified.
2396 The shown line contains a sequence of key/value pairs of the form
2397 @var{key}:@var{value}.
2399 The following values are shown in the output:
2403 The (sequential) number of the input frame, starting from 0.
2406 The presentation timestamp of the input frame, in time base units; the time base
2407 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2410 The presentation timestamp of the input frame in seconds.
2413 position of the frame in the input stream, -1 if this information in
2414 unavailable and/or meaningless (for example in case of synthetic audio)
2423 The sample rate for the audio frame.
2426 The number of samples (per channel) in the frame.
2429 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2430 audio, the data is treated as if all the planes were concatenated.
2432 @item plane_checksums
2433 A list of Adler-32 checksums for each data plane.
2437 Apply audio soft clipping.
2439 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2440 along a smooth curve, rather than the abrupt shape of hard-clipping.
2442 This filter accepts the following options:
2446 Set type of soft-clipping.
2448 It accepts the following values:
2462 Set threshold from where to start clipping. Default value is 0dB or 1.
2465 Set gain applied to output. Default value is 0dB or 1.
2468 Set additional parameter which controls sigmoid function.
2471 Set oversampling factor.
2474 @subsection Commands
2476 This filter supports the all above options as @ref{commands}.
2479 Automatic Speech Recognition
2481 This filter uses PocketSphinx for speech recognition. To enable
2482 compilation of this filter, you need to configure FFmpeg with
2483 @code{--enable-pocketsphinx}.
2485 It accepts the following options:
2489 Set sampling rate of input audio. Defaults is @code{16000}.
2490 This need to match speech models, otherwise one will get poor results.
2493 Set dictionary containing acoustic model files.
2496 Set pronunciation dictionary.
2499 Set language model file.
2502 Set language model set.
2505 Set which language model to use.
2508 Set output for log messages.
2511 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2516 Display time domain statistical information about the audio channels.
2517 Statistics are calculated and displayed for each audio channel and,
2518 where applicable, an overall figure is also given.
2520 It accepts the following option:
2523 Short window length in seconds, used for peak and trough RMS measurement.
2524 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2528 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2529 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2532 Available keys for each channel are:
2578 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2579 this @code{lavfi.astats.Overall.Peak_count}.
2581 For description what each key means read below.
2584 Set number of frame after which stats are going to be recalculated.
2585 Default is disabled.
2587 @item measure_perchannel
2588 Select the entries which need to be measured per channel. The metadata keys can
2589 be used as flags, default is @option{all} which measures everything.
2590 @option{none} disables all per channel measurement.
2592 @item measure_overall
2593 Select the entries which need to be measured overall. The metadata keys can
2594 be used as flags, default is @option{all} which measures everything.
2595 @option{none} disables all overall measurement.
2599 A description of each shown parameter follows:
2603 Mean amplitude displacement from zero.
2606 Minimal sample level.
2609 Maximal sample level.
2611 @item Min difference
2612 Minimal difference between two consecutive samples.
2614 @item Max difference
2615 Maximal difference between two consecutive samples.
2617 @item Mean difference
2618 Mean difference between two consecutive samples.
2619 The average of each difference between two consecutive samples.
2621 @item RMS difference
2622 Root Mean Square difference between two consecutive samples.
2626 Standard peak and RMS level measured in dBFS.
2630 Peak and trough values for RMS level measured over a short window.
2633 Standard ratio of peak to RMS level (note: not in dB).
2636 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2637 (i.e. either @var{Min level} or @var{Max level}).
2640 Number of occasions (not the number of samples) that the signal attained either
2641 @var{Min level} or @var{Max level}.
2643 @item Noise floor dB
2644 Minimum local peak measured in dBFS over a short window.
2646 @item Noise floor count
2647 Number of occasions (not the number of samples) that the signal attained
2651 Overall bit depth of audio. Number of bits used for each sample.
2654 Measured dynamic range of audio in dB.
2656 @item Zero crossings
2657 Number of points where the waveform crosses the zero level axis.
2659 @item Zero crossings rate
2660 Rate of Zero crossings and number of audio samples.
2664 Boost subwoofer frequencies.
2666 The filter accepts the following options:
2670 Set dry gain, how much of original signal is kept. Allowed range is from 0 to 1.
2671 Default value is 0.7.
2674 Set wet gain, how much of filtered signal is kept. Allowed range is from 0 to 1.
2675 Default value is 0.7.
2678 Set delay line decay gain value. Allowed range is from 0 to 1.
2679 Default value is 0.7.
2682 Set delay line feedback gain value. Allowed range is from 0 to 1.
2683 Default value is 0.9.
2686 Set cutoff frequency in Hertz. Allowed range is 50 to 900.
2687 Default value is 100.
2690 Set slope amount for cutoff frequency. Allowed range is 0.0001 to 1.
2691 Default value is 0.5.
2694 Set delay. Allowed range is from 1 to 100.
2695 Default value is 20.
2698 @subsection Commands
2700 This filter supports the all above options as @ref{commands}.
2703 Cut subwoofer frequencies.
2705 This filter allows to set custom, steeper
2706 roll off than highpass filter, and thus is able to more attenuate
2707 frequency content in stop-band.
2709 The filter accepts the following options:
2713 Set cutoff frequency in Hertz. Allowed range is 2 to 200.
2714 Default value is 20.
2717 Set filter order. Available values are from 3 to 20.
2718 Default value is 10.
2721 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2724 @subsection Commands
2726 This filter supports the all above options as @ref{commands}.
2729 Cut super frequencies.
2731 The filter accepts the following options:
2735 Set cutoff frequency in Hertz. Allowed range is 20000 to 192000.
2736 Default value is 20000.
2739 Set filter order. Available values are from 3 to 20.
2740 Default value is 10.
2743 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2746 @subsection Commands
2748 This filter supports the all above options as @ref{commands}.
2751 Apply high order Butterworth band-pass filter.
2753 The filter accepts the following options:
2757 Set center frequency in Hertz. Allowed range is 2 to 999999.
2758 Default value is 1000.
2761 Set filter order. Available values are from 4 to 20.
2765 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2768 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2771 @subsection Commands
2773 This filter supports the all above options as @ref{commands}.
2776 Apply high order Butterworth band-stop filter.
2778 The filter accepts the following options:
2782 Set center frequency in Hertz. Allowed range is 2 to 999999.
2783 Default value is 1000.
2786 Set filter order. Available values are from 4 to 20.
2790 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2793 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2796 @subsection Commands
2798 This filter supports the all above options as @ref{commands}.
2804 The filter accepts exactly one parameter, the audio tempo. If not
2805 specified then the filter will assume nominal 1.0 tempo. Tempo must
2806 be in the [0.5, 100.0] range.
2808 Note that tempo greater than 2 will skip some samples rather than
2809 blend them in. If for any reason this is a concern it is always
2810 possible to daisy-chain several instances of atempo to achieve the
2811 desired product tempo.
2813 @subsection Examples
2817 Slow down audio to 80% tempo:
2823 To speed up audio to 300% tempo:
2829 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2831 atempo=sqrt(3),atempo=sqrt(3)
2835 @subsection Commands
2837 This filter supports the following commands:
2840 Change filter tempo scale factor.
2841 Syntax for the command is : "@var{tempo}"
2846 Trim the input so that the output contains one continuous subpart of the input.
2848 It accepts the following parameters:
2851 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2852 sample with the timestamp @var{start} will be the first sample in the output.
2855 Specify time of the first audio sample that will be dropped, i.e. the
2856 audio sample immediately preceding the one with the timestamp @var{end} will be
2857 the last sample in the output.
2860 Same as @var{start}, except this option sets the start timestamp in samples
2864 Same as @var{end}, except this option sets the end timestamp in samples instead
2868 The maximum duration of the output in seconds.
2871 The number of the first sample that should be output.
2874 The number of the first sample that should be dropped.
2877 @option{start}, @option{end}, and @option{duration} are expressed as time
2878 duration specifications; see
2879 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2881 Note that the first two sets of the start/end options and the @option{duration}
2882 option look at the frame timestamp, while the _sample options simply count the
2883 samples that pass through the filter. So start/end_pts and start/end_sample will
2884 give different results when the timestamps are wrong, inexact or do not start at
2885 zero. Also note that this filter does not modify the timestamps. If you wish
2886 to have the output timestamps start at zero, insert the asetpts filter after the
2889 If multiple start or end options are set, this filter tries to be greedy and
2890 keep all samples that match at least one of the specified constraints. To keep
2891 only the part that matches all the constraints at once, chain multiple atrim
2894 The defaults are such that all the input is kept. So it is possible to set e.g.
2895 just the end values to keep everything before the specified time.
2900 Drop everything except the second minute of input:
2902 ffmpeg -i INPUT -af atrim=60:120
2906 Keep only the first 1000 samples:
2908 ffmpeg -i INPUT -af atrim=end_sample=1000
2913 @section axcorrelate
2914 Calculate normalized cross-correlation between two input audio streams.
2916 Resulted samples are always between -1 and 1 inclusive.
2917 If result is 1 it means two input samples are highly correlated in that selected segment.
2918 Result 0 means they are not correlated at all.
2919 If result is -1 it means two input samples are out of phase, which means they cancel each
2922 The filter accepts the following options:
2926 Set size of segment over which cross-correlation is calculated.
2927 Default is 256. Allowed range is from 2 to 131072.
2930 Set algorithm for cross-correlation. Can be @code{slow} or @code{fast}.
2931 Default is @code{slow}. Fast algorithm assumes mean values over any given segment
2932 are always zero and thus need much less calculations to make.
2933 This is generally not true, but is valid for typical audio streams.
2936 @subsection Examples
2940 Calculate correlation between channels in stereo audio stream:
2942 ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
2948 Apply a two-pole Butterworth band-pass filter with central
2949 frequency @var{frequency}, and (3dB-point) band-width width.
2950 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2951 instead of the default: constant 0dB peak gain.
2952 The filter roll off at 6dB per octave (20dB per decade).
2954 The filter accepts the following options:
2958 Set the filter's central frequency. Default is @code{3000}.
2961 Constant skirt gain if set to 1. Defaults to 0.
2964 Set method to specify band-width of filter.
2979 Specify the band-width of a filter in width_type units.
2982 How much to use filtered signal in output. Default is 1.
2983 Range is between 0 and 1.
2986 Specify which channels to filter, by default all available are filtered.
2989 Normalize biquad coefficients, by default is disabled.
2990 Enabling it will normalize magnitude response at DC to 0dB.
2993 Set transform type of IIR filter.
3002 Set precison of filtering.
3005 Pick automatic sample format depending on surround filters.
3007 Always use signed 16-bit.
3009 Always use signed 32-bit.
3011 Always use float 32-bit.
3013 Always use float 64-bit.
3017 @subsection Commands
3019 This filter supports the following commands:
3022 Change bandpass frequency.
3023 Syntax for the command is : "@var{frequency}"
3026 Change bandpass width_type.
3027 Syntax for the command is : "@var{width_type}"
3030 Change bandpass width.
3031 Syntax for the command is : "@var{width}"
3034 Change bandpass mix.
3035 Syntax for the command is : "@var{mix}"
3040 Apply a two-pole Butterworth band-reject filter with central
3041 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
3042 The filter roll off at 6dB per octave (20dB per decade).
3044 The filter accepts the following options:
3048 Set the filter's central frequency. Default is @code{3000}.
3051 Set method to specify band-width of filter.
3066 Specify the band-width of a filter in width_type units.
3069 How much to use filtered signal in output. Default is 1.
3070 Range is between 0 and 1.
3073 Specify which channels to filter, by default all available are filtered.
3076 Normalize biquad coefficients, by default is disabled.
3077 Enabling it will normalize magnitude response at DC to 0dB.
3080 Set transform type of IIR filter.
3089 Set precison of filtering.
3092 Pick automatic sample format depending on surround filters.
3094 Always use signed 16-bit.
3096 Always use signed 32-bit.
3098 Always use float 32-bit.
3100 Always use float 64-bit.
3104 @subsection Commands
3106 This filter supports the following commands:
3109 Change bandreject frequency.
3110 Syntax for the command is : "@var{frequency}"
3113 Change bandreject width_type.
3114 Syntax for the command is : "@var{width_type}"
3117 Change bandreject width.
3118 Syntax for the command is : "@var{width}"
3121 Change bandreject mix.
3122 Syntax for the command is : "@var{mix}"
3125 @section bass, lowshelf
3127 Boost or cut the bass (lower) frequencies of the audio using a two-pole
3128 shelving filter with a response similar to that of a standard
3129 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3131 The filter accepts the following options:
3135 Give the gain at 0 Hz. Its useful range is about -20
3136 (for a large cut) to +20 (for a large boost).
3137 Beware of clipping when using a positive gain.
3140 Set the filter's central frequency and so can be used
3141 to extend or reduce the frequency range to be boosted or cut.
3142 The default value is @code{100} Hz.
3145 Set method to specify band-width of filter.
3160 Determine how steep is the filter's shelf transition.
3163 Set number of poles. Default is 2.
3166 How much to use filtered signal in output. Default is 1.
3167 Range is between 0 and 1.
3170 Specify which channels to filter, by default all available are filtered.
3173 Normalize biquad coefficients, by default is disabled.
3174 Enabling it will normalize magnitude response at DC to 0dB.
3177 Set transform type of IIR filter.
3186 Set precison of filtering.
3189 Pick automatic sample format depending on surround filters.
3191 Always use signed 16-bit.
3193 Always use signed 32-bit.
3195 Always use float 32-bit.
3197 Always use float 64-bit.
3201 @subsection Commands
3203 This filter supports the following commands:
3206 Change bass frequency.
3207 Syntax for the command is : "@var{frequency}"
3210 Change bass width_type.
3211 Syntax for the command is : "@var{width_type}"
3215 Syntax for the command is : "@var{width}"
3219 Syntax for the command is : "@var{gain}"
3223 Syntax for the command is : "@var{mix}"
3228 Apply a biquad IIR filter with the given coefficients.
3229 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
3230 are the numerator and denominator coefficients respectively.
3231 and @var{channels}, @var{c} specify which channels to filter, by default all
3232 available are filtered.
3234 @subsection Commands
3236 This filter supports the following commands:
3244 Change biquad parameter.
3245 Syntax for the command is : "@var{value}"
3248 How much to use filtered signal in output. Default is 1.
3249 Range is between 0 and 1.
3252 Specify which channels to filter, by default all available are filtered.
3255 Normalize biquad coefficients, by default is disabled.
3256 Enabling it will normalize magnitude response at DC to 0dB.
3259 Set transform type of IIR filter.
3268 Set precison of filtering.
3271 Pick automatic sample format depending on surround filters.
3273 Always use signed 16-bit.
3275 Always use signed 32-bit.
3277 Always use float 32-bit.
3279 Always use float 64-bit.
3284 Bauer stereo to binaural transformation, which improves headphone listening of
3285 stereo audio records.
3287 To enable compilation of this filter you need to configure FFmpeg with
3288 @code{--enable-libbs2b}.
3290 It accepts the following parameters:
3294 Pre-defined crossfeed level.
3298 Default level (fcut=700, feed=50).
3301 Chu Moy circuit (fcut=700, feed=60).
3304 Jan Meier circuit (fcut=650, feed=95).
3309 Cut frequency (in Hz).
3318 Remap input channels to new locations.
3320 It accepts the following parameters:
3323 Map channels from input to output. The argument is a '|'-separated list of
3324 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
3325 @var{in_channel} form. @var{in_channel} can be either the name of the input
3326 channel (e.g. FL for front left) or its index in the input channel layout.
3327 @var{out_channel} is the name of the output channel or its index in the output
3328 channel layout. If @var{out_channel} is not given then it is implicitly an
3329 index, starting with zero and increasing by one for each mapping.
3331 @item channel_layout
3332 The channel layout of the output stream.
3335 If no mapping is present, the filter will implicitly map input channels to
3336 output channels, preserving indices.
3338 @subsection Examples
3342 For example, assuming a 5.1+downmix input MOV file,
3344 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
3346 will create an output WAV file tagged as stereo from the downmix channels of
3350 To fix a 5.1 WAV improperly encoded in AAC's native channel order
3352 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
3356 @section channelsplit
3358 Split each channel from an input audio stream into a separate output stream.
3360 It accepts the following parameters:
3362 @item channel_layout
3363 The channel layout of the input stream. The default is "stereo".
3365 A channel layout describing the channels to be extracted as separate output streams
3366 or "all" to extract each input channel as a separate stream. The default is "all".
3368 Choosing channels not present in channel layout in the input will result in an error.
3371 @subsection Examples
3375 For example, assuming a stereo input MP3 file,
3377 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
3379 will create an output Matroska file with two audio streams, one containing only
3380 the left channel and the other the right channel.
3383 Split a 5.1 WAV file into per-channel files:
3385 ffmpeg -i in.wav -filter_complex
3386 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
3387 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
3388 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
3393 Extract only LFE from a 5.1 WAV file:
3395 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
3396 -map '[LFE]' lfe.wav
3401 Add a chorus effect to the audio.
3403 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
3405 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
3406 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
3407 The modulation depth defines the range the modulated delay is played before or after
3408 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
3409 sound tuned around the original one, like in a chorus where some vocals are slightly
3412 It accepts the following parameters:
3415 Set input gain. Default is 0.4.
3418 Set output gain. Default is 0.4.
3421 Set delays. A typical delay is around 40ms to 60ms.
3433 @subsection Examples
3439 chorus=0.7:0.9:55:0.4:0.25:2
3445 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
3449 Fuller sounding chorus with three delays:
3451 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
3456 Compress or expand the audio's dynamic range.
3458 It accepts the following parameters:
3464 A list of times in seconds for each channel over which the instantaneous level
3465 of the input signal is averaged to determine its volume. @var{attacks} refers to
3466 increase of volume and @var{decays} refers to decrease of volume. For most
3467 situations, the attack time (response to the audio getting louder) should be
3468 shorter than the decay time, because the human ear is more sensitive to sudden
3469 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
3470 a typical value for decay is 0.8 seconds.
3471 If specified number of attacks & decays is lower than number of channels, the last
3472 set attack/decay will be used for all remaining channels.
3475 A list of points for the transfer function, specified in dB relative to the
3476 maximum possible signal amplitude. Each key points list must be defined using
3477 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
3478 @code{x0/y0 x1/y1 x2/y2 ....}
3480 The input values must be in strictly increasing order but the transfer function
3481 does not have to be monotonically rising. The point @code{0/0} is assumed but
3482 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
3483 function are @code{-70/-70|-60/-20|1/0}.
3486 Set the curve radius in dB for all joints. It defaults to 0.01.
3489 Set the additional gain in dB to be applied at all points on the transfer
3490 function. This allows for easy adjustment of the overall gain.
3494 Set an initial volume, in dB, to be assumed for each channel when filtering
3495 starts. This permits the user to supply a nominal level initially, so that, for
3496 example, a very large gain is not applied to initial signal levels before the
3497 companding has begun to operate. A typical value for audio which is initially
3498 quiet is -90 dB. It defaults to 0.
3501 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
3502 delayed before being fed to the volume adjuster. Specifying a delay
3503 approximately equal to the attack/decay times allows the filter to effectively
3504 operate in predictive rather than reactive mode. It defaults to 0.
3508 @subsection Examples
3512 Make music with both quiet and loud passages suitable for listening to in a
3515 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
3518 Another example for audio with whisper and explosion parts:
3520 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
3524 A noise gate for when the noise is at a lower level than the signal:
3526 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
3530 Here is another noise gate, this time for when the noise is at a higher level
3531 than the signal (making it, in some ways, similar to squelch):
3533 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
3537 2:1 compression starting at -6dB:
3539 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
3543 2:1 compression starting at -9dB:
3545 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
3549 2:1 compression starting at -12dB:
3551 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
3555 2:1 compression starting at -18dB:
3557 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
3561 3:1 compression starting at -15dB:
3563 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
3569 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
3575 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
3579 Hard limiter at -6dB:
3581 compand=attacks=0:points=-80/-80|-6/-6|20/-6
3585 Hard limiter at -12dB:
3587 compand=attacks=0:points=-80/-80|-12/-12|20/-12
3591 Hard noise gate at -35 dB:
3593 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
3599 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
3603 @section compensationdelay
3605 Compensation Delay Line is a metric based delay to compensate differing
3606 positions of microphones or speakers.
3608 For example, you have recorded guitar with two microphones placed in
3609 different locations. Because the front of sound wave has fixed speed in
3610 normal conditions, the phasing of microphones can vary and depends on
3611 their location and interposition. The best sound mix can be achieved when
3612 these microphones are in phase (synchronized). Note that a distance of
3613 ~30 cm between microphones makes one microphone capture the signal in
3614 antiphase to the other microphone. That makes the final mix sound moody.
3615 This filter helps to solve phasing problems by adding different delays
3616 to each microphone track and make them synchronized.
3618 The best result can be reached when you take one track as base and
3619 synchronize other tracks one by one with it.
3620 Remember that synchronization/delay tolerance depends on sample rate, too.
3621 Higher sample rates will give more tolerance.
3623 The filter accepts the following parameters:
3627 Set millimeters distance. This is compensation distance for fine tuning.
3631 Set cm distance. This is compensation distance for tightening distance setup.
3635 Set meters distance. This is compensation distance for hard distance setup.
3639 Set dry amount. Amount of unprocessed (dry) signal.
3643 Set wet amount. Amount of processed (wet) signal.
3647 Set temperature in degrees Celsius. This is the temperature of the environment.
3652 Apply headphone crossfeed filter.
3654 Crossfeed is the process of blending the left and right channels of stereo
3656 It is mainly used to reduce extreme stereo separation of low frequencies.
3658 The intent is to produce more speaker like sound to the listener.
3660 The filter accepts the following options:
3664 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3665 This sets gain of low shelf filter for side part of stereo image.
3666 Default is -6dB. Max allowed is -30db when strength is set to 1.
3669 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3670 This sets cut off frequency of low shelf filter. Default is cut off near
3671 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3674 Set curve slope of low shelf filter. Default is 0.5.
3675 Allowed range is from 0.01 to 1.
3678 Set input gain. Default is 0.9.
3681 Set output gain. Default is 1.
3684 @subsection Commands
3686 This filter supports the all above options as @ref{commands}.
3688 @section crystalizer
3689 Simple algorithm for audio noise sharpening.
3691 This filter linearly increases differences betweeen each audio sample.
3693 The filter accepts the following options:
3697 Sets the intensity of effect (default: 2.0). Must be in range between -10.0 to 0
3698 (unchanged sound) to 10.0 (maximum effect).
3699 To inverse filtering use negative value.
3702 Enable clipping. By default is enabled.
3705 @subsection Commands
3707 This filter supports the all above options as @ref{commands}.
3710 Apply a DC shift to the audio.
3712 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3713 in the recording chain) from the audio. The effect of a DC offset is reduced
3714 headroom and hence volume. The @ref{astats} filter can be used to determine if
3715 a signal has a DC offset.
3719 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3723 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3724 used to prevent clipping.
3729 Apply de-essing to the audio samples.
3733 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3737 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3741 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3745 Set the output mode.
3747 It accepts the following values:
3750 Pass input unchanged.
3753 Pass ess filtered out.
3758 Default value is @var{o}.
3764 Measure audio dynamic range.
3766 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3767 is found in transition material. And anything less that 8 have very poor dynamics
3768 and is very compressed.
3770 The filter accepts the following options:
3774 Set window length in seconds used to split audio into segments of equal length.
3775 Default is 3 seconds.
3779 Dynamic Audio Normalizer.
3781 This filter applies a certain amount of gain to the input audio in order
3782 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3783 contrast to more "simple" normalization algorithms, the Dynamic Audio
3784 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3785 This allows for applying extra gain to the "quiet" sections of the audio
3786 while avoiding distortions or clipping the "loud" sections. In other words:
3787 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3788 sections, in the sense that the volume of each section is brought to the
3789 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3790 this goal *without* applying "dynamic range compressing". It will retain 100%
3791 of the dynamic range *within* each section of the audio file.
3795 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3796 Default is 500 milliseconds.
3797 The Dynamic Audio Normalizer processes the input audio in small chunks,
3798 referred to as frames. This is required, because a peak magnitude has no
3799 meaning for just a single sample value. Instead, we need to determine the
3800 peak magnitude for a contiguous sequence of sample values. While a "standard"
3801 normalizer would simply use the peak magnitude of the complete file, the
3802 Dynamic Audio Normalizer determines the peak magnitude individually for each
3803 frame. The length of a frame is specified in milliseconds. By default, the
3804 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3805 been found to give good results with most files.
3806 Note that the exact frame length, in number of samples, will be determined
3807 automatically, based on the sampling rate of the individual input audio file.
3810 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3811 number. Default is 31.
3812 Probably the most important parameter of the Dynamic Audio Normalizer is the
3813 @code{window size} of the Gaussian smoothing filter. The filter's window size
3814 is specified in frames, centered around the current frame. For the sake of
3815 simplicity, this must be an odd number. Consequently, the default value of 31
3816 takes into account the current frame, as well as the 15 preceding frames and
3817 the 15 subsequent frames. Using a larger window results in a stronger
3818 smoothing effect and thus in less gain variation, i.e. slower gain
3819 adaptation. Conversely, using a smaller window results in a weaker smoothing
3820 effect and thus in more gain variation, i.e. faster gain adaptation.
3821 In other words, the more you increase this value, the more the Dynamic Audio
3822 Normalizer will behave like a "traditional" normalization filter. On the
3823 contrary, the more you decrease this value, the more the Dynamic Audio
3824 Normalizer will behave like a dynamic range compressor.
3827 Set the target peak value. This specifies the highest permissible magnitude
3828 level for the normalized audio input. This filter will try to approach the
3829 target peak magnitude as closely as possible, but at the same time it also
3830 makes sure that the normalized signal will never exceed the peak magnitude.
3831 A frame's maximum local gain factor is imposed directly by the target peak
3832 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3833 It is not recommended to go above this value.
3836 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3837 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3838 factor for each input frame, i.e. the maximum gain factor that does not
3839 result in clipping or distortion. The maximum gain factor is determined by
3840 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3841 additionally bounds the frame's maximum gain factor by a predetermined
3842 (global) maximum gain factor. This is done in order to avoid excessive gain
3843 factors in "silent" or almost silent frames. By default, the maximum gain
3844 factor is 10.0, For most inputs the default value should be sufficient and
3845 it usually is not recommended to increase this value. Though, for input
3846 with an extremely low overall volume level, it may be necessary to allow even
3847 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3848 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3849 Instead, a "sigmoid" threshold function will be applied. This way, the
3850 gain factors will smoothly approach the threshold value, but never exceed that
3854 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3855 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3856 This means that the maximum local gain factor for each frame is defined
3857 (only) by the frame's highest magnitude sample. This way, the samples can
3858 be amplified as much as possible without exceeding the maximum signal
3859 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3860 Normalizer can also take into account the frame's root mean square,
3861 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3862 determine the power of a time-varying signal. It is therefore considered
3863 that the RMS is a better approximation of the "perceived loudness" than
3864 just looking at the signal's peak magnitude. Consequently, by adjusting all
3865 frames to a constant RMS value, a uniform "perceived loudness" can be
3866 established. If a target RMS value has been specified, a frame's local gain
3867 factor is defined as the factor that would result in exactly that RMS value.
3868 Note, however, that the maximum local gain factor is still restricted by the
3869 frame's highest magnitude sample, in order to prevent clipping.
3872 Enable channels coupling. By default is enabled.
3873 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3874 amount. This means the same gain factor will be applied to all channels, i.e.
3875 the maximum possible gain factor is determined by the "loudest" channel.
3876 However, in some recordings, it may happen that the volume of the different
3877 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3878 In this case, this option can be used to disable the channel coupling. This way,
3879 the gain factor will be determined independently for each channel, depending
3880 only on the individual channel's highest magnitude sample. This allows for
3881 harmonizing the volume of the different channels.
3884 Enable DC bias correction. By default is disabled.
3885 An audio signal (in the time domain) is a sequence of sample values.
3886 In the Dynamic Audio Normalizer these sample values are represented in the
3887 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3888 audio signal, or "waveform", should be centered around the zero point.
3889 That means if we calculate the mean value of all samples in a file, or in a
3890 single frame, then the result should be 0.0 or at least very close to that
3891 value. If, however, there is a significant deviation of the mean value from
3892 0.0, in either positive or negative direction, this is referred to as a
3893 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3894 Audio Normalizer provides optional DC bias correction.
3895 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3896 the mean value, or "DC correction" offset, of each input frame and subtract
3897 that value from all of the frame's sample values which ensures those samples
3898 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3899 boundaries, the DC correction offset values will be interpolated smoothly
3900 between neighbouring frames.
3902 @item altboundary, b
3903 Enable alternative boundary mode. By default is disabled.
3904 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3905 around each frame. This includes the preceding frames as well as the
3906 subsequent frames. However, for the "boundary" frames, located at the very
3907 beginning and at the very end of the audio file, not all neighbouring
3908 frames are available. In particular, for the first few frames in the audio
3909 file, the preceding frames are not known. And, similarly, for the last few
3910 frames in the audio file, the subsequent frames are not known. Thus, the
3911 question arises which gain factors should be assumed for the missing frames
3912 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3913 to deal with this situation. The default boundary mode assumes a gain factor
3914 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3915 "fade out" at the beginning and at the end of the input, respectively.
3918 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3919 By default, the Dynamic Audio Normalizer does not apply "traditional"
3920 compression. This means that signal peaks will not be pruned and thus the
3921 full dynamic range will be retained within each local neighbourhood. However,
3922 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3923 normalization algorithm with a more "traditional" compression.
3924 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3925 (thresholding) function. If (and only if) the compression feature is enabled,
3926 all input frames will be processed by a soft knee thresholding function prior
3927 to the actual normalization process. Put simply, the thresholding function is
3928 going to prune all samples whose magnitude exceeds a certain threshold value.
3929 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3930 value. Instead, the threshold value will be adjusted for each individual
3932 In general, smaller parameters result in stronger compression, and vice versa.
3933 Values below 3.0 are not recommended, because audible distortion may appear.
3936 Set the target threshold value. This specifies the lowest permissible
3937 magnitude level for the audio input which will be normalized.
3938 If input frame volume is above this value frame will be normalized.
3939 Otherwise frame may not be normalized at all. The default value is set
3940 to 0, which means all input frames will be normalized.
3941 This option is mostly useful if digital noise is not wanted to be amplified.
3944 @subsection Commands
3946 This filter supports the all above options as @ref{commands}.
3950 Make audio easier to listen to on headphones.
3952 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3953 so that when listened to on headphones the stereo image is moved from
3954 inside your head (standard for headphones) to outside and in front of
3955 the listener (standard for speakers).
3961 Apply a two-pole peaking equalisation (EQ) filter. With this
3962 filter, the signal-level at and around a selected frequency can
3963 be increased or decreased, whilst (unlike bandpass and bandreject
3964 filters) that at all other frequencies is unchanged.
3966 In order to produce complex equalisation curves, this filter can
3967 be given several times, each with a different central frequency.
3969 The filter accepts the following options:
3973 Set the filter's central frequency in Hz.
3976 Set method to specify band-width of filter.
3991 Specify the band-width of a filter in width_type units.
3994 Set the required gain or attenuation in dB.
3995 Beware of clipping when using a positive gain.
3998 How much to use filtered signal in output. Default is 1.
3999 Range is between 0 and 1.
4002 Specify which channels to filter, by default all available are filtered.
4005 Normalize biquad coefficients, by default is disabled.
4006 Enabling it will normalize magnitude response at DC to 0dB.
4009 Set transform type of IIR filter.
4018 Set precison of filtering.
4021 Pick automatic sample format depending on surround filters.
4023 Always use signed 16-bit.
4025 Always use signed 32-bit.
4027 Always use float 32-bit.
4029 Always use float 64-bit.
4033 @subsection Examples
4036 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
4038 equalizer=f=1000:t=h:width=200:g=-10
4042 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
4044 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
4048 @subsection Commands
4050 This filter supports the following commands:
4053 Change equalizer frequency.
4054 Syntax for the command is : "@var{frequency}"
4057 Change equalizer width_type.
4058 Syntax for the command is : "@var{width_type}"
4061 Change equalizer width.
4062 Syntax for the command is : "@var{width}"
4065 Change equalizer gain.
4066 Syntax for the command is : "@var{gain}"
4069 Change equalizer mix.
4070 Syntax for the command is : "@var{mix}"
4073 @section extrastereo
4075 Linearly increases the difference between left and right channels which
4076 adds some sort of "live" effect to playback.
4078 The filter accepts the following options:
4082 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
4083 (average of both channels), with 1.0 sound will be unchanged, with
4084 -1.0 left and right channels will be swapped.
4087 Enable clipping. By default is enabled.
4090 @subsection Commands
4092 This filter supports the all above options as @ref{commands}.
4094 @section firequalizer
4095 Apply FIR Equalization using arbitrary frequency response.
4097 The filter accepts the following option:
4101 Set gain curve equation (in dB). The expression can contain variables:
4104 the evaluated frequency
4108 channel number, set to 0 when multichannels evaluation is disabled
4110 channel id, see libavutil/channel_layout.h, set to the first channel id when
4111 multichannels evaluation is disabled
4115 channel_layout, see libavutil/channel_layout.h
4120 @item gain_interpolate(f)
4121 interpolate gain on frequency f based on gain_entry
4122 @item cubic_interpolate(f)
4123 same as gain_interpolate, but smoother
4125 This option is also available as command. Default is @code{gain_interpolate(f)}.
4128 Set gain entry for gain_interpolate function. The expression can
4132 store gain entry at frequency f with value g
4134 This option is also available as command.
4137 Set filter delay in seconds. Higher value means more accurate.
4138 Default is @code{0.01}.
4141 Set filter accuracy in Hz. Lower value means more accurate.
4142 Default is @code{5}.
4145 Set window function. Acceptable values are:
4148 rectangular window, useful when gain curve is already smooth
4150 hann window (default)
4156 3-terms continuous 1st derivative nuttall window
4158 minimum 3-terms discontinuous nuttall window
4160 4-terms continuous 1st derivative nuttall window
4162 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
4164 blackman-harris window
4170 If enabled, use fixed number of audio samples. This improves speed when
4171 filtering with large delay. Default is disabled.
4174 Enable multichannels evaluation on gain. Default is disabled.
4177 Enable zero phase mode by subtracting timestamp to compensate delay.
4178 Default is disabled.
4181 Set scale used by gain. Acceptable values are:
4184 linear frequency, linear gain
4186 linear frequency, logarithmic (in dB) gain (default)
4188 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
4190 logarithmic frequency, logarithmic gain
4194 Set file for dumping, suitable for gnuplot.
4197 Set scale for dumpfile. Acceptable values are same with scale option.
4201 Enable 2-channel convolution using complex FFT. This improves speed significantly.
4202 Default is disabled.
4205 Enable minimum phase impulse response. Default is disabled.
4208 @subsection Examples
4213 firequalizer=gain='if(lt(f,1000), 0, -INF)'
4216 lowpass at 1000 Hz with gain_entry:
4218 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
4221 custom equalization:
4223 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
4226 higher delay with zero phase to compensate delay:
4228 firequalizer=delay=0.1:fixed=on:zero_phase=on
4231 lowpass on left channel, highpass on right channel:
4233 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
4234 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
4239 Apply a flanging effect to the audio.
4241 The filter accepts the following options:
4245 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
4248 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
4251 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
4255 Set percentage of delayed signal mixed with original. Range from 0 to 100.
4256 Default value is 71.
4259 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
4262 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
4263 Default value is @var{sinusoidal}.
4266 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
4267 Default value is 25.
4270 Set delay-line interpolation, @var{linear} or @var{quadratic}.
4271 Default is @var{linear}.
4275 Apply Haas effect to audio.
4277 Note that this makes most sense to apply on mono signals.
4278 With this filter applied to mono signals it give some directionality and
4279 stretches its stereo image.
4281 The filter accepts the following options:
4285 Set input level. By default is @var{1}, or 0dB
4288 Set output level. By default is @var{1}, or 0dB.
4291 Set gain applied to side part of signal. By default is @var{1}.
4294 Set kind of middle source. Can be one of the following:
4304 Pick middle part signal of stereo image.
4307 Pick side part signal of stereo image.
4311 Change middle phase. By default is disabled.
4314 Set left channel delay. By default is @var{2.05} milliseconds.
4317 Set left channel balance. By default is @var{-1}.
4320 Set left channel gain. By default is @var{1}.
4323 Change left phase. By default is disabled.
4326 Set right channel delay. By defaults is @var{2.12} milliseconds.
4329 Set right channel balance. By default is @var{1}.
4332 Set right channel gain. By default is @var{1}.
4335 Change right phase. By default is enabled.
4340 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
4341 embedded HDCD codes is expanded into a 20-bit PCM stream.
4343 The filter supports the Peak Extend and Low-level Gain Adjustment features
4344 of HDCD, and detects the Transient Filter flag.
4347 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
4350 When using the filter with wav, note the default encoding for wav is 16-bit,
4351 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
4352 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
4354 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
4355 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
4358 The filter accepts the following options:
4361 @item disable_autoconvert
4362 Disable any automatic format conversion or resampling in the filter graph.
4364 @item process_stereo
4365 Process the stereo channels together. If target_gain does not match between
4366 channels, consider it invalid and use the last valid target_gain.
4369 Set the code detect timer period in ms.
4372 Always extend peaks above -3dBFS even if PE isn't signaled.
4375 Replace audio with a solid tone and adjust the amplitude to signal some
4376 specific aspect of the decoding process. The output file can be loaded in
4377 an audio editor alongside the original to aid analysis.
4379 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
4386 Gain adjustment level at each sample
4388 Samples where peak extend occurs
4390 Samples where the code detect timer is active
4392 Samples where the target gain does not match between channels
4398 Apply head-related transfer functions (HRTFs) to create virtual
4399 loudspeakers around the user for binaural listening via headphones.
4400 The HRIRs are provided via additional streams, for each channel
4401 one stereo input stream is needed.
4403 The filter accepts the following options:
4407 Set mapping of input streams for convolution.
4408 The argument is a '|'-separated list of channel names in order as they
4409 are given as additional stream inputs for filter.
4410 This also specify number of input streams. Number of input streams
4411 must be not less than number of channels in first stream plus one.
4414 Set gain applied to audio. Value is in dB. Default is 0.
4417 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4418 processing audio in time domain which is slow.
4419 @var{freq} is processing audio in frequency domain which is fast.
4420 Default is @var{freq}.
4423 Set custom gain for LFE channels. Value is in dB. Default is 0.
4426 Set size of frame in number of samples which will be processed at once.
4427 Default value is @var{1024}. Allowed range is from 1024 to 96000.
4430 Set format of hrir stream.
4431 Default value is @var{stereo}. Alternative value is @var{multich}.
4432 If value is set to @var{stereo}, number of additional streams should
4433 be greater or equal to number of input channels in first input stream.
4434 Also each additional stream should have stereo number of channels.
4435 If value is set to @var{multich}, number of additional streams should
4436 be exactly one. Also number of input channels of additional stream
4437 should be equal or greater than twice number of channels of first input
4441 @subsection Examples
4445 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4446 each amovie filter use stereo file with IR coefficients as input.
4447 The files give coefficients for each position of virtual loudspeaker:
4450 -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"
4455 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4456 but now in @var{multich} @var{hrir} format.
4458 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"
4465 Apply a high-pass filter with 3dB point frequency.
4466 The filter can be either single-pole, or double-pole (the default).
4467 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4469 The filter accepts the following options:
4473 Set frequency in Hz. Default is 3000.
4476 Set number of poles. Default is 2.
4479 Set method to specify band-width of filter.
4494 Specify the band-width of a filter in width_type units.
4495 Applies only to double-pole filter.
4496 The default is 0.707q and gives a Butterworth response.
4499 How much to use filtered signal in output. Default is 1.
4500 Range is between 0 and 1.
4503 Specify which channels to filter, by default all available are filtered.
4506 Normalize biquad coefficients, by default is disabled.
4507 Enabling it will normalize magnitude response at DC to 0dB.
4510 Set transform type of IIR filter.
4519 Set precison of filtering.
4522 Pick automatic sample format depending on surround filters.
4524 Always use signed 16-bit.
4526 Always use signed 32-bit.
4528 Always use float 32-bit.
4530 Always use float 64-bit.
4534 @subsection Commands
4536 This filter supports the following commands:
4539 Change highpass frequency.
4540 Syntax for the command is : "@var{frequency}"
4543 Change highpass width_type.
4544 Syntax for the command is : "@var{width_type}"
4547 Change highpass width.
4548 Syntax for the command is : "@var{width}"
4551 Change highpass mix.
4552 Syntax for the command is : "@var{mix}"
4557 Join multiple input streams into one multi-channel stream.
4559 It accepts the following parameters:
4563 The number of input streams. It defaults to 2.
4565 @item channel_layout
4566 The desired output channel layout. It defaults to stereo.
4569 Map channels from inputs to output. The argument is a '|'-separated list of
4570 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
4571 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
4572 can be either the name of the input channel (e.g. FL for front left) or its
4573 index in the specified input stream. @var{out_channel} is the name of the output
4577 The filter will attempt to guess the mappings when they are not specified
4578 explicitly. It does so by first trying to find an unused matching input channel
4579 and if that fails it picks the first unused input channel.
4581 Join 3 inputs (with properly set channel layouts):
4583 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
4586 Build a 5.1 output from 6 single-channel streams:
4588 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
4589 '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'
4595 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
4597 To enable compilation of this filter you need to configure FFmpeg with
4598 @code{--enable-ladspa}.
4602 Specifies the name of LADSPA plugin library to load. If the environment
4603 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
4604 each one of the directories specified by the colon separated list in
4605 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
4606 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
4607 @file{/usr/lib/ladspa/}.
4610 Specifies the plugin within the library. Some libraries contain only
4611 one plugin, but others contain many of them. If this is not set filter
4612 will list all available plugins within the specified library.
4615 Set the '|' separated list of controls which are zero or more floating point
4616 values that determine the behavior of the loaded plugin (for example delay,
4618 Controls need to be defined using the following syntax:
4619 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
4620 @var{valuei} is the value set on the @var{i}-th control.
4621 Alternatively they can be also defined using the following syntax:
4622 @var{value0}|@var{value1}|@var{value2}|..., where
4623 @var{valuei} is the value set on the @var{i}-th control.
4624 If @option{controls} is set to @code{help}, all available controls and
4625 their valid ranges are printed.
4627 @item sample_rate, s
4628 Specify the sample rate, default to 44100. Only used if plugin have
4632 Set the number of samples per channel per each output frame, default
4633 is 1024. Only used if plugin have zero inputs.
4636 Set the minimum duration of the sourced audio. See
4637 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4638 for the accepted syntax.
4639 Note that the resulting duration may be greater than the specified duration,
4640 as the generated audio is always cut at the end of a complete frame.
4641 If not specified, or the expressed duration is negative, the audio is
4642 supposed to be generated forever.
4643 Only used if plugin have zero inputs.
4646 Enable latency compensation, by default is disabled.
4647 Only used if plugin have inputs.
4650 @subsection Examples
4654 List all available plugins within amp (LADSPA example plugin) library:
4660 List all available controls and their valid ranges for @code{vcf_notch}
4661 plugin from @code{VCF} library:
4663 ladspa=f=vcf:p=vcf_notch:c=help
4667 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
4670 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
4674 Add reverberation to the audio using TAP-plugins
4675 (Tom's Audio Processing plugins):
4677 ladspa=file=tap_reverb:tap_reverb
4681 Generate white noise, with 0.2 amplitude:
4683 ladspa=file=cmt:noise_source_white:c=c0=.2
4687 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
4688 @code{C* Audio Plugin Suite} (CAPS) library:
4690 ladspa=file=caps:Click:c=c1=20'
4694 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
4696 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
4700 Increase volume by 20dB using fast lookahead limiter from Steve Harris
4701 @code{SWH Plugins} collection:
4703 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
4707 Attenuate low frequencies using Multiband EQ from Steve Harris
4708 @code{SWH Plugins} collection:
4710 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
4714 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
4717 ladspa=caps:Narrower
4721 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
4723 ladspa=caps:White:.2
4727 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
4729 ladspa=caps:Fractal:c=c1=1
4733 Dynamic volume normalization using @code{VLevel} plugin:
4735 ladspa=vlevel-ladspa:vlevel_mono
4739 @subsection Commands
4741 This filter supports the following commands:
4744 Modify the @var{N}-th control value.
4746 If the specified value is not valid, it is ignored and prior one is kept.
4751 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4752 Support for both single pass (livestreams, files) and double pass (files) modes.
4753 This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately
4754 detect true peaks, the audio stream will be upsampled to 192 kHz.
4755 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4757 The filter accepts the following options:
4761 Set integrated loudness target.
4762 Range is -70.0 - -5.0. Default value is -24.0.
4765 Set loudness range target.
4766 Range is 1.0 - 20.0. Default value is 7.0.
4769 Set maximum true peak.
4770 Range is -9.0 - +0.0. Default value is -2.0.
4772 @item measured_I, measured_i
4773 Measured IL of input file.
4774 Range is -99.0 - +0.0.
4776 @item measured_LRA, measured_lra
4777 Measured LRA of input file.
4778 Range is 0.0 - 99.0.
4780 @item measured_TP, measured_tp
4781 Measured true peak of input file.
4782 Range is -99.0 - +99.0.
4784 @item measured_thresh
4785 Measured threshold of input file.
4786 Range is -99.0 - +0.0.
4789 Set offset gain. Gain is applied before the true-peak limiter.
4790 Range is -99.0 - +99.0. Default is +0.0.
4793 Normalize by linearly scaling the source audio.
4794 @code{measured_I}, @code{measured_LRA}, @code{measured_TP},
4795 and @code{measured_thresh} must all be specified. Target LRA shouldn't
4796 be lower than source LRA and the change in integrated loudness shouldn't
4797 result in a true peak which exceeds the target TP. If any of these
4798 conditions aren't met, normalization mode will revert to @var{dynamic}.
4799 Options are @code{true} or @code{false}. Default is @code{true}.
4802 Treat mono input files as "dual-mono". If a mono file is intended for playback
4803 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4804 If set to @code{true}, this option will compensate for this effect.
4805 Multi-channel input files are not affected by this option.
4806 Options are true or false. Default is false.
4809 Set print format for stats. Options are summary, json, or none.
4810 Default value is none.
4815 Apply a low-pass filter with 3dB point frequency.
4816 The filter can be either single-pole or double-pole (the default).
4817 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4819 The filter accepts the following options:
4823 Set frequency in Hz. Default is 500.
4826 Set number of poles. Default is 2.
4829 Set method to specify band-width of filter.
4844 Specify the band-width of a filter in width_type units.
4845 Applies only to double-pole filter.
4846 The default is 0.707q and gives a Butterworth response.
4849 How much to use filtered signal in output. Default is 1.
4850 Range is between 0 and 1.
4853 Specify which channels to filter, by default all available are filtered.
4856 Normalize biquad coefficients, by default is disabled.
4857 Enabling it will normalize magnitude response at DC to 0dB.
4860 Set transform type of IIR filter.
4869 Set precison of filtering.
4872 Pick automatic sample format depending on surround filters.
4874 Always use signed 16-bit.
4876 Always use signed 32-bit.
4878 Always use float 32-bit.
4880 Always use float 64-bit.
4884 @subsection Examples
4887 Lowpass only LFE channel, it LFE is not present it does nothing:
4893 @subsection Commands
4895 This filter supports the following commands:
4898 Change lowpass frequency.
4899 Syntax for the command is : "@var{frequency}"
4902 Change lowpass width_type.
4903 Syntax for the command is : "@var{width_type}"
4906 Change lowpass width.
4907 Syntax for the command is : "@var{width}"
4911 Syntax for the command is : "@var{mix}"
4916 Load a LV2 (LADSPA Version 2) plugin.
4918 To enable compilation of this filter you need to configure FFmpeg with
4919 @code{--enable-lv2}.
4923 Specifies the plugin URI. You may need to escape ':'.
4926 Set the '|' separated list of controls which are zero or more floating point
4927 values that determine the behavior of the loaded plugin (for example delay,
4929 If @option{controls} is set to @code{help}, all available controls and
4930 their valid ranges are printed.
4932 @item sample_rate, s
4933 Specify the sample rate, default to 44100. Only used if plugin have
4937 Set the number of samples per channel per each output frame, default
4938 is 1024. Only used if plugin have zero inputs.
4941 Set the minimum duration of the sourced audio. See
4942 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4943 for the accepted syntax.
4944 Note that the resulting duration may be greater than the specified duration,
4945 as the generated audio is always cut at the end of a complete frame.
4946 If not specified, or the expressed duration is negative, the audio is
4947 supposed to be generated forever.
4948 Only used if plugin have zero inputs.
4951 @subsection Examples
4955 Apply bass enhancer plugin from Calf:
4957 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4961 Apply vinyl plugin from Calf:
4963 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4967 Apply bit crusher plugin from ArtyFX:
4969 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4974 Multiband Compress or expand the audio's dynamic range.
4976 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4977 This is akin to the crossover of a loudspeaker, and results in flat frequency
4978 response when absent compander action.
4980 It accepts the following parameters:
4984 This option syntax is:
4985 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4986 For explanation of each item refer to compand filter documentation.
4992 Mix channels with specific gain levels. The filter accepts the output
4993 channel layout followed by a set of channels definitions.
4995 This filter is also designed to efficiently remap the channels of an audio
4998 The filter accepts parameters of the form:
4999 "@var{l}|@var{outdef}|@var{outdef}|..."
5003 output channel layout or number of channels
5006 output channel specification, of the form:
5007 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
5010 output channel to define, either a channel name (FL, FR, etc.) or a channel
5011 number (c0, c1, etc.)
5014 multiplicative coefficient for the channel, 1 leaving the volume unchanged
5017 input channel to use, see out_name for details; it is not possible to mix
5018 named and numbered input channels
5021 If the `=' in a channel specification is replaced by `<', then the gains for
5022 that specification will be renormalized so that the total is 1, thus
5023 avoiding clipping noise.
5025 @subsection Mixing examples
5027 For example, if you want to down-mix from stereo to mono, but with a bigger
5028 factor for the left channel:
5030 pan=1c|c0=0.9*c0+0.1*c1
5033 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
5034 7-channels surround:
5036 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
5039 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
5040 that should be preferred (see "-ac" option) unless you have very specific
5043 @subsection Remapping examples
5045 The channel remapping will be effective if, and only if:
5048 @item gain coefficients are zeroes or ones,
5049 @item only one input per channel output,
5052 If all these conditions are satisfied, the filter will notify the user ("Pure
5053 channel mapping detected"), and use an optimized and lossless method to do the
5056 For example, if you have a 5.1 source and want a stereo audio stream by
5057 dropping the extra channels:
5059 pan="stereo| c0=FL | c1=FR"
5062 Given the same source, you can also switch front left and front right channels
5063 and keep the input channel layout:
5065 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
5068 If the input is a stereo audio stream, you can mute the front left channel (and
5069 still keep the stereo channel layout) with:
5074 Still with a stereo audio stream input, you can copy the right channel in both
5075 front left and right:
5077 pan="stereo| c0=FR | c1=FR"
5082 ReplayGain scanner filter. This filter takes an audio stream as an input and
5083 outputs it unchanged.
5084 At end of filtering it displays @code{track_gain} and @code{track_peak}.
5088 Convert the audio sample format, sample rate and channel layout. It is
5089 not meant to be used directly.
5092 Apply time-stretching and pitch-shifting with librubberband.
5094 To enable compilation of this filter, you need to configure FFmpeg with
5095 @code{--enable-librubberband}.
5097 The filter accepts the following options:
5101 Set tempo scale factor.
5104 Set pitch scale factor.
5107 Set transients detector.
5108 Possible values are:
5117 Possible values are:
5126 Possible values are:
5133 Set processing window size.
5134 Possible values are:
5143 Possible values are:
5150 Enable formant preservation when shift pitching.
5151 Possible values are:
5159 Possible values are:
5168 Possible values are:
5175 @subsection Commands
5177 This filter supports the following commands:
5180 Change filter tempo scale factor.
5181 Syntax for the command is : "@var{tempo}"
5184 Change filter pitch scale factor.
5185 Syntax for the command is : "@var{pitch}"
5188 @section sidechaincompress
5190 This filter acts like normal compressor but has the ability to compress
5191 detected signal using second input signal.
5192 It needs two input streams and returns one output stream.
5193 First input stream will be processed depending on second stream signal.
5194 The filtered signal then can be filtered with other filters in later stages of
5195 processing. See @ref{pan} and @ref{amerge} filter.
5197 The filter accepts the following options:
5201 Set input gain. Default is 1. Range is between 0.015625 and 64.
5204 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
5205 Default is @code{downward}.
5208 If a signal of second stream raises above this level it will affect the gain
5209 reduction of first stream.
5210 By default is 0.125. Range is between 0.00097563 and 1.
5213 Set a ratio about which the signal is reduced. 1:2 means that if the level
5214 raised 4dB above the threshold, it will be only 2dB above after the reduction.
5215 Default is 2. Range is between 1 and 20.
5218 Amount of milliseconds the signal has to rise above the threshold before gain
5219 reduction starts. Default is 20. Range is between 0.01 and 2000.
5222 Amount of milliseconds the signal has to fall below the threshold before
5223 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
5226 Set the amount by how much signal will be amplified after processing.
5227 Default is 1. Range is from 1 to 64.
5230 Curve the sharp knee around the threshold to enter gain reduction more softly.
5231 Default is 2.82843. Range is between 1 and 8.
5234 Choose if the @code{average} level between all channels of side-chain stream
5235 or the louder(@code{maximum}) channel of side-chain stream affects the
5236 reduction. Default is @code{average}.
5239 Should the exact signal be taken in case of @code{peak} or an RMS one in case
5240 of @code{rms}. Default is @code{rms} which is mainly smoother.
5243 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
5246 How much to use compressed signal in output. Default is 1.
5247 Range is between 0 and 1.
5250 @subsection Commands
5252 This filter supports the all above options as @ref{commands}.
5254 @subsection Examples
5258 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
5259 depending on the signal of 2nd input and later compressed signal to be
5260 merged with 2nd input:
5262 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
5266 @section sidechaingate
5268 A sidechain gate acts like a normal (wideband) gate but has the ability to
5269 filter the detected signal before sending it to the gain reduction stage.
5270 Normally a gate uses the full range signal to detect a level above the
5272 For example: If you cut all lower frequencies from your sidechain signal
5273 the gate will decrease the volume of your track only if not enough highs
5274 appear. With this technique you are able to reduce the resonation of a
5275 natural drum or remove "rumbling" of muted strokes from a heavily distorted
5277 It needs two input streams and returns one output stream.
5278 First input stream will be processed depending on second stream signal.
5280 The filter accepts the following options:
5284 Set input level before filtering.
5285 Default is 1. Allowed range is from 0.015625 to 64.
5288 Set the mode of operation. Can be @code{upward} or @code{downward}.
5289 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
5290 will be amplified, expanding dynamic range in upward direction.
5291 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
5294 Set the level of gain reduction when the signal is below the threshold.
5295 Default is 0.06125. Allowed range is from 0 to 1.
5296 Setting this to 0 disables reduction and then filter behaves like expander.
5299 If a signal rises above this level the gain reduction is released.
5300 Default is 0.125. Allowed range is from 0 to 1.
5303 Set a ratio about which the signal is reduced.
5304 Default is 2. Allowed range is from 1 to 9000.
5307 Amount of milliseconds the signal has to rise above the threshold before gain
5309 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
5312 Amount of milliseconds the signal has to fall below the threshold before the
5313 reduction is increased again. Default is 250 milliseconds.
5314 Allowed range is from 0.01 to 9000.
5317 Set amount of amplification of signal after processing.
5318 Default is 1. Allowed range is from 1 to 64.
5321 Curve the sharp knee around the threshold to enter gain reduction more softly.
5322 Default is 2.828427125. Allowed range is from 1 to 8.
5325 Choose if exact signal should be taken for detection or an RMS like one.
5326 Default is rms. Can be peak or rms.
5329 Choose if the average level between all channels or the louder channel affects
5331 Default is average. Can be average or maximum.
5334 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
5337 @subsection Commands
5339 This filter supports the all above options as @ref{commands}.
5341 @section silencedetect
5343 Detect silence in an audio stream.
5345 This filter logs a message when it detects that the input audio volume is less
5346 or equal to a noise tolerance value for a duration greater or equal to the
5347 minimum detected noise duration.
5349 The printed times and duration are expressed in seconds. The
5350 @code{lavfi.silence_start} or @code{lavfi.silence_start.X} metadata key
5351 is set on the first frame whose timestamp equals or exceeds the detection
5352 duration and it contains the timestamp of the first frame of the silence.
5354 The @code{lavfi.silence_duration} or @code{lavfi.silence_duration.X}
5355 and @code{lavfi.silence_end} or @code{lavfi.silence_end.X} metadata
5356 keys are set on the first frame after the silence. If @option{mono} is
5357 enabled, and each channel is evaluated separately, the @code{.X}
5358 suffixed keys are used, and @code{X} corresponds to the channel number.
5360 The filter accepts the following options:
5364 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
5365 specified value) or amplitude ratio. Default is -60dB, or 0.001.
5368 Set silence duration until notification (default is 2 seconds). See
5369 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5370 for the accepted syntax.
5373 Process each channel separately, instead of combined. By default is disabled.
5376 @subsection Examples
5380 Detect 5 seconds of silence with -50dB noise tolerance:
5382 silencedetect=n=-50dB:d=5
5386 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
5387 tolerance in @file{silence.mp3}:
5389 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
5393 @section silenceremove
5395 Remove silence from the beginning, middle or end of the audio.
5397 The filter accepts the following options:
5401 This value is used to indicate if audio should be trimmed at beginning of
5402 the audio. A value of zero indicates no silence should be trimmed from the
5403 beginning. When specifying a non-zero value, it trims audio up until it
5404 finds non-silence. Normally, when trimming silence from beginning of audio
5405 the @var{start_periods} will be @code{1} but it can be increased to higher
5406 values to trim all audio up to specific count of non-silence periods.
5407 Default value is @code{0}.
5409 @item start_duration
5410 Specify the amount of time that non-silence must be detected before it stops
5411 trimming audio. By increasing the duration, bursts of noises can be treated
5412 as silence and trimmed off. Default value is @code{0}.
5414 @item start_threshold
5415 This indicates what sample value should be treated as silence. For digital
5416 audio, a value of @code{0} may be fine but for audio recorded from analog,
5417 you may wish to increase the value to account for background noise.
5418 Can be specified in dB (in case "dB" is appended to the specified value)
5419 or amplitude ratio. Default value is @code{0}.
5422 Specify max duration of silence at beginning that will be kept after
5423 trimming. Default is 0, which is equal to trimming all samples detected
5427 Specify mode of detection of silence end in start of multi-channel audio.
5428 Can be @var{any} or @var{all}. Default is @var{any}.
5429 With @var{any}, any sample that is detected as non-silence will cause
5430 stopped trimming of silence.
5431 With @var{all}, only if all channels are detected as non-silence will cause
5432 stopped trimming of silence.
5435 Set the count for trimming silence from the end of audio.
5436 To remove silence from the middle of a file, specify a @var{stop_periods}
5437 that is negative. This value is then treated as a positive value and is
5438 used to indicate the effect should restart processing as specified by
5439 @var{start_periods}, making it suitable for removing periods of silence
5440 in the middle of the audio.
5441 Default value is @code{0}.
5444 Specify a duration of silence that must exist before audio is not copied any
5445 more. By specifying a higher duration, silence that is wanted can be left in
5447 Default value is @code{0}.
5449 @item stop_threshold
5450 This is the same as @option{start_threshold} but for trimming silence from
5452 Can be specified in dB (in case "dB" is appended to the specified value)
5453 or amplitude ratio. Default value is @code{0}.
5456 Specify max duration of silence at end that will be kept after
5457 trimming. Default is 0, which is equal to trimming all samples detected
5461 Specify mode of detection of silence start in end of multi-channel audio.
5462 Can be @var{any} or @var{all}. Default is @var{any}.
5463 With @var{any}, any sample that is detected as non-silence will cause
5464 stopped trimming of silence.
5465 With @var{all}, only if all channels are detected as non-silence will cause
5466 stopped trimming of silence.
5469 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
5470 and works better with digital silence which is exactly 0.
5471 Default value is @code{rms}.
5474 Set duration in number of seconds used to calculate size of window in number
5475 of samples for detecting silence.
5476 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
5479 @subsection Examples
5483 The following example shows how this filter can be used to start a recording
5484 that does not contain the delay at the start which usually occurs between
5485 pressing the record button and the start of the performance:
5487 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
5491 Trim all silence encountered from beginning to end where there is more than 1
5492 second of silence in audio:
5494 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
5498 Trim all digital silence samples, using peak detection, from beginning to end
5499 where there is more than 0 samples of digital silence in audio and digital
5500 silence is detected in all channels at same positions in stream:
5502 silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
5508 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
5509 loudspeakers around the user for binaural listening via headphones (audio
5510 formats up to 9 channels supported).
5511 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
5512 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
5513 Austrian Academy of Sciences.
5515 To enable compilation of this filter you need to configure FFmpeg with
5516 @code{--enable-libmysofa}.
5518 The filter accepts the following options:
5522 Set the SOFA file used for rendering.
5525 Set gain applied to audio. Value is in dB. Default is 0.
5528 Set rotation of virtual loudspeakers in deg. Default is 0.
5531 Set elevation of virtual speakers in deg. Default is 0.
5534 Set distance in meters between loudspeakers and the listener with near-field
5535 HRTFs. Default is 1.
5538 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
5539 processing audio in time domain which is slow.
5540 @var{freq} is processing audio in frequency domain which is fast.
5541 Default is @var{freq}.
5544 Set custom positions of virtual loudspeakers. Syntax for this option is:
5545 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
5546 Each virtual loudspeaker is described with short channel name following with
5547 azimuth and elevation in degrees.
5548 Each virtual loudspeaker description is separated by '|'.
5549 For example to override front left and front right channel positions use:
5550 'speakers=FL 45 15|FR 345 15'.
5551 Descriptions with unrecognised channel names are ignored.
5554 Set custom gain for LFE channels. Value is in dB. Default is 0.
5557 Set custom frame size in number of samples. Default is 1024.
5558 Allowed range is from 1024 to 96000. Only used if option @samp{type}
5559 is set to @var{freq}.
5562 Should all IRs be normalized upon importing SOFA file.
5563 By default is enabled.
5566 Should nearest IRs be interpolated with neighbor IRs if exact position
5567 does not match. By default is disabled.
5570 Minphase all IRs upon loading of SOFA file. By default is disabled.
5573 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
5576 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
5579 @subsection Examples
5583 Using ClubFritz6 sofa file:
5585 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
5589 Using ClubFritz12 sofa file and bigger radius with small rotation:
5591 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
5595 Similar as above but with custom speaker positions for front left, front right, back left and back right
5596 and also with custom gain:
5598 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
5605 This filter expands or compresses each half-cycle of audio samples
5606 (local set of samples all above or all below zero and between two nearest zero crossings) depending
5607 on threshold value, so audio reaches target peak value under conditions controlled by below options.
5609 The filter accepts the following options:
5613 Set the expansion target peak value. This specifies the highest allowed absolute amplitude
5614 level for the normalized audio input. Default value is 0.95. Allowed range is from 0.0 to 1.0.
5617 Set the maximum expansion factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5618 This option controls maximum local half-cycle of samples expansion. The maximum expansion
5619 would be such that local peak value reaches target peak value but never to surpass it and that
5620 ratio between new and previous peak value does not surpass this option value.
5622 @item compression, c
5623 Set the maximum compression factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5624 This option controls maximum local half-cycle of samples compression. This option is used
5625 only if @option{threshold} option is set to value greater than 0.0, then in such cases
5626 when local peak is lower or same as value set by @option{threshold} all samples belonging to
5627 that peak's half-cycle will be compressed by current compression factor.
5630 Set the threshold value. Default value is 0.0. Allowed range is from 0.0 to 1.0.
5631 This option specifies which half-cycles of samples will be compressed and which will be expanded.
5632 Any half-cycle samples with their local peak value below or same as this option value will be
5633 compressed by current compression factor, otherwise, if greater than threshold value they will be
5634 expanded with expansion factor so that it could reach peak target value but never surpass it.
5637 Set the expansion raising amount per each half-cycle of samples. Default value is 0.001.
5638 Allowed range is from 0.0 to 1.0. This controls how fast expansion factor is raised per
5639 each new half-cycle until it reaches @option{expansion} value.
5640 Setting this options too high may lead to distortions.
5643 Set the compression raising amount per each half-cycle of samples. Default value is 0.001.
5644 Allowed range is from 0.0 to 1.0. This controls how fast compression factor is raised per
5645 each new half-cycle until it reaches @option{compression} value.
5648 Specify which channels to filter, by default all available channels are filtered.
5651 Enable inverted filtering, by default is disabled. This inverts interpretation of @option{threshold}
5652 option. When enabled any half-cycle of samples with their local peak value below or same as
5653 @option{threshold} option will be expanded otherwise it will be compressed.
5656 Link channels when calculating gain applied to each filtered channel sample, by default is disabled.
5657 When disabled each filtered channel gain calculation is independent, otherwise when this option
5658 is enabled the minimum of all possible gains for each filtered channel is used.
5661 @subsection Commands
5663 This filter supports the all above options as @ref{commands}.
5665 @section stereotools
5667 This filter has some handy utilities to manage stereo signals, for converting
5668 M/S stereo recordings to L/R signal while having control over the parameters
5669 or spreading the stereo image of master track.
5671 The filter accepts the following options:
5675 Set input level before filtering for both channels. Defaults is 1.
5676 Allowed range is from 0.015625 to 64.
5679 Set output level after filtering for both channels. Defaults is 1.
5680 Allowed range is from 0.015625 to 64.
5683 Set input balance between both channels. Default is 0.
5684 Allowed range is from -1 to 1.
5687 Set output balance between both channels. Default is 0.
5688 Allowed range is from -1 to 1.
5691 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
5692 clipping. Disabled by default.
5695 Mute the left channel. Disabled by default.
5698 Mute the right channel. Disabled by default.
5701 Change the phase of the left channel. Disabled by default.
5704 Change the phase of the right channel. Disabled by default.
5707 Set stereo mode. Available values are:
5711 Left/Right to Left/Right, this is default.
5714 Left/Right to Mid/Side.
5717 Mid/Side to Left/Right.
5720 Left/Right to Left/Left.
5723 Left/Right to Right/Right.
5726 Left/Right to Left + Right.
5729 Left/Right to Right/Left.
5732 Mid/Side to Left/Left.
5735 Mid/Side to Right/Right.
5738 Mid/Side to Right/Left.
5741 Left/Right to Left - Right.
5745 Set level of side signal. Default is 1.
5746 Allowed range is from 0.015625 to 64.
5749 Set balance of side signal. Default is 0.
5750 Allowed range is from -1 to 1.
5753 Set level of the middle signal. Default is 1.
5754 Allowed range is from 0.015625 to 64.
5757 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
5760 Set stereo base between mono and inversed channels. Default is 0.
5761 Allowed range is from -1 to 1.
5764 Set delay in milliseconds how much to delay left from right channel and
5765 vice versa. Default is 0. Allowed range is from -20 to 20.
5768 Set S/C level. Default is 1. Allowed range is from 1 to 100.
5771 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
5773 @item bmode_in, bmode_out
5774 Set balance mode for balance_in/balance_out option.
5776 Can be one of the following:
5780 Classic balance mode. Attenuate one channel at time.
5781 Gain is raised up to 1.
5784 Similar as classic mode above but gain is raised up to 2.
5787 Equal power distribution, from -6dB to +6dB range.
5791 @subsection Commands
5793 This filter supports the all above options as @ref{commands}.
5795 @subsection Examples
5799 Apply karaoke like effect:
5801 stereotools=mlev=0.015625
5805 Convert M/S signal to L/R:
5807 "stereotools=mode=ms>lr"
5811 @section stereowiden
5813 This filter enhance the stereo effect by suppressing signal common to both
5814 channels and by delaying the signal of left into right and vice versa,
5815 thereby widening the stereo effect.
5817 The filter accepts the following options:
5821 Time in milliseconds of the delay of left signal into right and vice versa.
5822 Default is 20 milliseconds.
5825 Amount of gain in delayed signal into right and vice versa. Gives a delay
5826 effect of left signal in right output and vice versa which gives widening
5827 effect. Default is 0.3.
5830 Cross feed of left into right with inverted phase. This helps in suppressing
5831 the mono. If the value is 1 it will cancel all the signal common to both
5832 channels. Default is 0.3.
5835 Set level of input signal of original channel. Default is 0.8.
5838 @subsection Commands
5840 This filter supports the all above options except @code{delay} as @ref{commands}.
5842 @section superequalizer
5843 Apply 18 band equalizer.
5845 The filter accepts the following options:
5852 Set 131Hz band gain.
5854 Set 185Hz band gain.
5856 Set 262Hz band gain.
5858 Set 370Hz band gain.
5860 Set 523Hz band gain.
5862 Set 740Hz band gain.
5864 Set 1047Hz band gain.
5866 Set 1480Hz band gain.
5868 Set 2093Hz band gain.
5870 Set 2960Hz band gain.
5872 Set 4186Hz band gain.
5874 Set 5920Hz band gain.
5876 Set 8372Hz band gain.
5878 Set 11840Hz band gain.
5880 Set 16744Hz band gain.
5882 Set 20000Hz band gain.
5886 Apply audio surround upmix filter.
5888 This filter allows to produce multichannel output from audio stream.
5890 The filter accepts the following options:
5894 Set output channel layout. By default, this is @var{5.1}.
5896 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5897 for the required syntax.
5900 Set input channel layout. By default, this is @var{stereo}.
5902 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5903 for the required syntax.
5906 Set input volume level. By default, this is @var{1}.
5909 Set output volume level. By default, this is @var{1}.
5912 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5915 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5918 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5921 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5922 In @var{add} mode, LFE channel is created from input audio and added to output.
5923 In @var{sub} mode, LFE channel is created from input audio and added to output but
5924 also all non-LFE output channels are subtracted with output LFE channel.
5927 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5928 Default is @var{90}.
5931 Set front center input volume. By default, this is @var{1}.
5934 Set front center output volume. By default, this is @var{1}.
5937 Set front left input volume. By default, this is @var{1}.
5940 Set front left output volume. By default, this is @var{1}.
5943 Set front right input volume. By default, this is @var{1}.
5946 Set front right output volume. By default, this is @var{1}.
5949 Set side left input volume. By default, this is @var{1}.
5952 Set side left output volume. By default, this is @var{1}.
5955 Set side right input volume. By default, this is @var{1}.
5958 Set side right output volume. By default, this is @var{1}.
5961 Set back left input volume. By default, this is @var{1}.
5964 Set back left output volume. By default, this is @var{1}.
5967 Set back right input volume. By default, this is @var{1}.
5970 Set back right output volume. By default, this is @var{1}.
5973 Set back center input volume. By default, this is @var{1}.
5976 Set back center output volume. By default, this is @var{1}.
5979 Set LFE input volume. By default, this is @var{1}.
5982 Set LFE output volume. By default, this is @var{1}.
5985 Set spread usage of stereo image across X axis for all channels.
5988 Set spread usage of stereo image across Y axis for all channels.
5990 @item fcx, flx, frx, blx, brx, slx, srx, bcx
5991 Set spread usage of stereo image across X axis for each channel.
5993 @item fcy, fly, fry, bly, bry, sly, sry, bcy
5994 Set spread usage of stereo image across Y axis for each channel.
5997 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
6000 Set window function.
6002 It accepts the following values:
6025 Default is @code{hann}.
6028 Set window overlap. If set to 1, the recommended overlap for selected
6029 window function will be picked. Default is @code{0.5}.
6032 @section treble, highshelf
6034 Boost or cut treble (upper) frequencies of the audio using a two-pole
6035 shelving filter with a response similar to that of a standard
6036 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
6038 The filter accepts the following options:
6042 Give the gain at whichever is the lower of ~22 kHz and the
6043 Nyquist frequency. Its useful range is about -20 (for a large cut)
6044 to +20 (for a large boost). Beware of clipping when using a positive gain.
6047 Set the filter's central frequency and so can be used
6048 to extend or reduce the frequency range to be boosted or cut.
6049 The default value is @code{3000} Hz.
6052 Set method to specify band-width of filter.
6067 Determine how steep is the filter's shelf transition.
6070 Set number of poles. Default is 2.
6073 How much to use filtered signal in output. Default is 1.
6074 Range is between 0 and 1.
6077 Specify which channels to filter, by default all available are filtered.
6080 Normalize biquad coefficients, by default is disabled.
6081 Enabling it will normalize magnitude response at DC to 0dB.
6084 Set transform type of IIR filter.
6093 Set precison of filtering.
6096 Pick automatic sample format depending on surround filters.
6098 Always use signed 16-bit.
6100 Always use signed 32-bit.
6102 Always use float 32-bit.
6104 Always use float 64-bit.
6108 @subsection Commands
6110 This filter supports the following commands:
6113 Change treble frequency.
6114 Syntax for the command is : "@var{frequency}"
6117 Change treble width_type.
6118 Syntax for the command is : "@var{width_type}"
6121 Change treble width.
6122 Syntax for the command is : "@var{width}"
6126 Syntax for the command is : "@var{gain}"
6130 Syntax for the command is : "@var{mix}"
6135 Sinusoidal amplitude modulation.
6137 The filter accepts the following options:
6141 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
6142 (20 Hz or lower) will result in a tremolo effect.
6143 This filter may also be used as a ring modulator by specifying
6144 a modulation frequency higher than 20 Hz.
6145 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6148 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6149 Default value is 0.5.
6154 Sinusoidal phase modulation.
6156 The filter accepts the following options:
6160 Modulation frequency in Hertz.
6161 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6164 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6165 Default value is 0.5.
6170 Adjust the input audio volume.
6172 It accepts the following parameters:
6176 Set audio volume expression.
6178 Output values are clipped to the maximum value.
6180 The output audio volume is given by the relation:
6182 @var{output_volume} = @var{volume} * @var{input_volume}
6185 The default value for @var{volume} is "1.0".
6188 This parameter represents the mathematical precision.
6190 It determines which input sample formats will be allowed, which affects the
6191 precision of the volume scaling.
6195 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
6197 32-bit floating-point; this limits input sample format to FLT. (default)
6199 64-bit floating-point; this limits input sample format to DBL.
6203 Choose the behaviour on encountering ReplayGain side data in input frames.
6207 Remove ReplayGain side data, ignoring its contents (the default).
6210 Ignore ReplayGain side data, but leave it in the frame.
6213 Prefer the track gain, if present.
6216 Prefer the album gain, if present.
6219 @item replaygain_preamp
6220 Pre-amplification gain in dB to apply to the selected replaygain gain.
6222 Default value for @var{replaygain_preamp} is 0.0.
6224 @item replaygain_noclip
6225 Prevent clipping by limiting the gain applied.
6227 Default value for @var{replaygain_noclip} is 1.
6230 Set when the volume expression is evaluated.
6232 It accepts the following values:
6235 only evaluate expression once during the filter initialization, or
6236 when the @samp{volume} command is sent
6239 evaluate expression for each incoming frame
6242 Default value is @samp{once}.
6245 The volume expression can contain the following parameters.
6249 frame number (starting at zero)
6252 @item nb_consumed_samples
6253 number of samples consumed by the filter
6255 number of samples in the current frame
6257 original frame position in the file
6263 PTS at start of stream
6265 time at start of stream
6271 last set volume value
6274 Note that when @option{eval} is set to @samp{once} only the
6275 @var{sample_rate} and @var{tb} variables are available, all other
6276 variables will evaluate to NAN.
6278 @subsection Commands
6280 This filter supports the following commands:
6283 Modify the volume expression.
6284 The command accepts the same syntax of the corresponding option.
6286 If the specified expression is not valid, it is kept at its current
6290 @subsection Examples
6294 Halve the input audio volume:
6298 volume=volume=-6.0206dB
6301 In all the above example the named key for @option{volume} can be
6302 omitted, for example like in:
6308 Increase input audio power by 6 decibels using fixed-point precision:
6310 volume=volume=6dB:precision=fixed
6314 Fade volume after time 10 with an annihilation period of 5 seconds:
6316 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
6320 @section volumedetect
6322 Detect the volume of the input video.
6324 The filter has no parameters. The input is not modified. Statistics about
6325 the volume will be printed in the log when the input stream end is reached.
6327 In particular it will show the mean volume (root mean square), maximum
6328 volume (on a per-sample basis), and the beginning of a histogram of the
6329 registered volume values (from the maximum value to a cumulated 1/1000 of
6332 All volumes are in decibels relative to the maximum PCM value.
6334 @subsection Examples
6336 Here is an excerpt of the output:
6338 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
6339 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
6340 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
6341 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
6342 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
6343 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
6344 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
6345 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
6346 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
6352 The mean square energy is approximately -27 dB, or 10^-2.7.
6354 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
6356 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
6359 In other words, raising the volume by +4 dB does not cause any clipping,
6360 raising it by +5 dB causes clipping for 6 samples, etc.
6362 @c man end AUDIO FILTERS
6364 @chapter Audio Sources
6365 @c man begin AUDIO SOURCES
6367 Below is a description of the currently available audio sources.
6371 Buffer audio frames, and make them available to the filter chain.
6373 This source is mainly intended for a programmatic use, in particular
6374 through the interface defined in @file{libavfilter/buffersrc.h}.
6376 It accepts the following parameters:
6380 The timebase which will be used for timestamps of submitted frames. It must be
6381 either a floating-point number or in @var{numerator}/@var{denominator} form.
6384 The sample rate of the incoming audio buffers.
6387 The sample format of the incoming audio buffers.
6388 Either a sample format name or its corresponding integer representation from
6389 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
6391 @item channel_layout
6392 The channel layout of the incoming audio buffers.
6393 Either a channel layout name from channel_layout_map in
6394 @file{libavutil/channel_layout.c} or its corresponding integer representation
6395 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
6398 The number of channels of the incoming audio buffers.
6399 If both @var{channels} and @var{channel_layout} are specified, then they
6404 @subsection Examples
6407 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
6410 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
6411 Since the sample format with name "s16p" corresponds to the number
6412 6 and the "stereo" channel layout corresponds to the value 0x3, this is
6415 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
6420 Generate an audio signal specified by an expression.
6422 This source accepts in input one or more expressions (one for each
6423 channel), which are evaluated and used to generate a corresponding
6426 This source accepts the following options:
6430 Set the '|'-separated expressions list for each separate channel. In case the
6431 @option{channel_layout} option is not specified, the selected channel layout
6432 depends on the number of provided expressions. Otherwise the last
6433 specified expression is applied to the remaining output channels.
6435 @item channel_layout, c
6436 Set the channel layout. The number of channels in the specified layout
6437 must be equal to the number of specified expressions.
6440 Set the minimum duration of the sourced audio. See
6441 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6442 for the accepted syntax.
6443 Note that the resulting duration may be greater than the specified
6444 duration, as the generated audio is always cut at the end of a
6447 If not specified, or the expressed duration is negative, the audio is
6448 supposed to be generated forever.
6451 Set the number of samples per channel per each output frame,
6454 @item sample_rate, s
6455 Specify the sample rate, default to 44100.
6458 Each expression in @var{exprs} can contain the following constants:
6462 number of the evaluated sample, starting from 0
6465 time of the evaluated sample expressed in seconds, starting from 0
6472 @subsection Examples
6482 Generate a sin signal with frequency of 440 Hz, set sample rate to
6485 aevalsrc="sin(440*2*PI*t):s=8000"
6489 Generate a two channels signal, specify the channel layout (Front
6490 Center + Back Center) explicitly:
6492 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
6496 Generate white noise:
6498 aevalsrc="-2+random(0)"
6502 Generate an amplitude modulated signal:
6504 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
6508 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
6510 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
6517 Generate a FIR coefficients using frequency sampling method.
6519 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6521 The filter accepts the following options:
6525 Set number of filter coefficents in output audio stream.
6526 Default value is 1025.
6529 Set frequency points from where magnitude and phase are set.
6530 This must be in non decreasing order, and first element must be 0, while last element
6531 must be 1. Elements are separated by white spaces.
6534 Set magnitude value for every frequency point set by @option{frequency}.
6535 Number of values must be same as number of frequency points.
6536 Values are separated by white spaces.
6539 Set phase value for every frequency point set by @option{frequency}.
6540 Number of values must be same as number of frequency points.
6541 Values are separated by white spaces.
6543 @item sample_rate, r
6544 Set sample rate, default is 44100.
6547 Set number of samples per each frame. Default is 1024.
6550 Set window function. Default is blackman.
6555 The null audio source, return unprocessed audio frames. It is mainly useful
6556 as a template and to be employed in analysis / debugging tools, or as
6557 the source for filters which ignore the input data (for example the sox
6560 This source accepts the following options:
6564 @item channel_layout, cl
6566 Specifies the channel layout, and can be either an integer or a string
6567 representing a channel layout. The default value of @var{channel_layout}
6570 Check the channel_layout_map definition in
6571 @file{libavutil/channel_layout.c} for the mapping between strings and
6572 channel layout values.
6574 @item sample_rate, r
6575 Specifies the sample rate, and defaults to 44100.
6578 Set the number of samples per requested frames.
6581 Set the duration of the sourced audio. See
6582 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6583 for the accepted syntax.
6585 If not specified, or the expressed duration is negative, the audio is
6586 supposed to be generated forever.
6589 @subsection Examples
6593 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
6595 anullsrc=r=48000:cl=4
6599 Do the same operation with a more obvious syntax:
6601 anullsrc=r=48000:cl=mono
6605 All the parameters need to be explicitly defined.
6609 Synthesize a voice utterance using the libflite library.
6611 To enable compilation of this filter you need to configure FFmpeg with
6612 @code{--enable-libflite}.
6614 Note that versions of the flite library prior to 2.0 are not thread-safe.
6616 The filter accepts the following options:
6621 If set to 1, list the names of the available voices and exit
6622 immediately. Default value is 0.
6625 Set the maximum number of samples per frame. Default value is 512.
6628 Set the filename containing the text to speak.
6631 Set the text to speak.
6634 Set the voice to use for the speech synthesis. Default value is
6635 @code{kal}. See also the @var{list_voices} option.
6638 @subsection Examples
6642 Read from file @file{speech.txt}, and synthesize the text using the
6643 standard flite voice:
6645 flite=textfile=speech.txt
6649 Read the specified text selecting the @code{slt} voice:
6651 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6655 Input text to ffmpeg:
6657 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6661 Make @file{ffplay} speak the specified text, using @code{flite} and
6662 the @code{lavfi} device:
6664 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
6668 For more information about libflite, check:
6669 @url{http://www.festvox.org/flite/}
6673 Generate a noise audio signal.
6675 The filter accepts the following options:
6678 @item sample_rate, r
6679 Specify the sample rate. Default value is 48000 Hz.
6682 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
6686 Specify the duration of the generated audio stream. Not specifying this option
6687 results in noise with an infinite length.
6689 @item color, colour, c
6690 Specify the color of noise. Available noise colors are white, pink, brown,
6691 blue, violet and velvet. Default color is white.
6694 Specify a value used to seed the PRNG.
6697 Set the number of samples per each output frame, default is 1024.
6700 @subsection Examples
6705 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
6707 anoisesrc=d=60:c=pink:r=44100:a=0.5
6713 Generate odd-tap Hilbert transform FIR coefficients.
6715 The resulting stream can be used with @ref{afir} filter for phase-shifting
6716 the signal by 90 degrees.
6718 This is used in many matrix coding schemes and for analytic signal generation.
6719 The process is often written as a multiplication by i (or j), the imaginary unit.
6721 The filter accepts the following options:
6725 @item sample_rate, s
6726 Set sample rate, default is 44100.
6729 Set length of FIR filter, default is 22051.
6732 Set number of samples per each frame.
6735 Set window function to be used when generating FIR coefficients.
6740 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
6742 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6744 The filter accepts the following options:
6747 @item sample_rate, r
6748 Set sample rate, default is 44100.
6751 Set number of samples per each frame. Default is 1024.
6754 Set high-pass frequency. Default is 0.
6757 Set low-pass frequency. Default is 0.
6758 If high-pass frequency is lower than low-pass frequency and low-pass frequency
6759 is higher than 0 then filter will create band-pass filter coefficients,
6760 otherwise band-reject filter coefficients.
6763 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
6766 Set Kaiser window beta.
6769 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
6772 Enable rounding, by default is disabled.
6775 Set number of taps for high-pass filter.
6778 Set number of taps for low-pass filter.
6783 Generate an audio signal made of a sine wave with amplitude 1/8.
6785 The audio signal is bit-exact.
6787 The filter accepts the following options:
6792 Set the carrier frequency. Default is 440 Hz.
6794 @item beep_factor, b
6795 Enable a periodic beep every second with frequency @var{beep_factor} times
6796 the carrier frequency. Default is 0, meaning the beep is disabled.
6798 @item sample_rate, r
6799 Specify the sample rate, default is 44100.
6802 Specify the duration of the generated audio stream.
6804 @item samples_per_frame
6805 Set the number of samples per output frame.
6807 The expression can contain the following constants:
6811 The (sequential) number of the output audio frame, starting from 0.
6814 The PTS (Presentation TimeStamp) of the output audio frame,
6815 expressed in @var{TB} units.
6818 The PTS of the output audio frame, expressed in seconds.
6821 The timebase of the output audio frames.
6824 Default is @code{1024}.
6827 @subsection Examples
6832 Generate a simple 440 Hz sine wave:
6838 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
6842 sine=frequency=220:beep_factor=4:duration=5
6846 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
6849 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
6853 @c man end AUDIO SOURCES
6855 @chapter Audio Sinks
6856 @c man begin AUDIO SINKS
6858 Below is a description of the currently available audio sinks.
6860 @section abuffersink
6862 Buffer audio frames, and make them available to the end of filter chain.
6864 This sink is mainly intended for programmatic use, in particular
6865 through the interface defined in @file{libavfilter/buffersink.h}
6866 or the options system.
6868 It accepts a pointer to an AVABufferSinkContext structure, which
6869 defines the incoming buffers' formats, to be passed as the opaque
6870 parameter to @code{avfilter_init_filter} for initialization.
6873 Null audio sink; do absolutely nothing with the input audio. It is
6874 mainly useful as a template and for use in analysis / debugging
6877 @c man end AUDIO SINKS
6879 @chapter Video Filters
6880 @c man begin VIDEO FILTERS
6882 When you configure your FFmpeg build, you can disable any of the
6883 existing filters using @code{--disable-filters}.
6884 The configure output will show the video filters included in your
6887 Below is a description of the currently available video filters.
6891 Mark a region of interest in a video frame.
6893 The frame data is passed through unchanged, but metadata is attached
6894 to the frame indicating regions of interest which can affect the
6895 behaviour of later encoding. Multiple regions can be marked by
6896 applying the filter multiple times.
6900 Region distance in pixels from the left edge of the frame.
6902 Region distance in pixels from the top edge of the frame.
6904 Region width in pixels.
6906 Region height in pixels.
6908 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
6909 and may contain the following variables:
6912 Width of the input frame.
6914 Height of the input frame.
6918 Quantisation offset to apply within the region.
6920 This must be a real value in the range -1 to +1. A value of zero
6921 indicates no quality change. A negative value asks for better quality
6922 (less quantisation), while a positive value asks for worse quality
6923 (greater quantisation).
6925 The range is calibrated so that the extreme values indicate the
6926 largest possible offset - if the rest of the frame is encoded with the
6927 worst possible quality, an offset of -1 indicates that this region
6928 should be encoded with the best possible quality anyway. Intermediate
6929 values are then interpolated in some codec-dependent way.
6931 For example, in 10-bit H.264 the quantisation parameter varies between
6932 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
6933 this region should be encoded with a QP around one-tenth of the full
6934 range better than the rest of the frame. So, if most of the frame
6935 were to be encoded with a QP of around 30, this region would get a QP
6936 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
6937 An extreme value of -1 would indicate that this region should be
6938 encoded with the best possible quality regardless of the treatment of
6939 the rest of the frame - that is, should be encoded at a QP of -12.
6941 If set to true, remove any existing regions of interest marked on the
6942 frame before adding the new one.
6945 @subsection Examples
6949 Mark the centre quarter of the frame as interesting.
6951 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
6954 Mark the 100-pixel-wide region on the left edge of the frame as very
6955 uninteresting (to be encoded at much lower quality than the rest of
6958 addroi=0:0:100:ih:+1/5
6962 @section alphaextract
6964 Extract the alpha component from the input as a grayscale video. This
6965 is especially useful with the @var{alphamerge} filter.
6969 Add or replace the alpha component of the primary input with the
6970 grayscale value of a second input. This is intended for use with
6971 @var{alphaextract} to allow the transmission or storage of frame
6972 sequences that have alpha in a format that doesn't support an alpha
6975 For example, to reconstruct full frames from a normal YUV-encoded video
6976 and a separate video created with @var{alphaextract}, you might use:
6978 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
6983 Amplify differences between current pixel and pixels of adjacent frames in
6984 same pixel location.
6986 This filter accepts the following options:
6990 Set frame radius. Default is 2. Allowed range is from 1 to 63.
6991 For example radius of 3 will instruct filter to calculate average of 7 frames.
6994 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
6997 Set threshold for difference amplification. Any difference greater or equal to
6998 this value will not alter source pixel. Default is 10.
6999 Allowed range is from 0 to 65535.
7002 Set tolerance for difference amplification. Any difference lower to
7003 this value will not alter source pixel. Default is 0.
7004 Allowed range is from 0 to 65535.
7007 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7008 This option controls maximum possible value that will decrease source pixel value.
7011 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7012 This option controls maximum possible value that will increase source pixel value.
7015 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
7018 @subsection Commands
7020 This filter supports the following @ref{commands} that corresponds to option of same name:
7032 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
7033 and libavformat to work. On the other hand, it is limited to ASS (Advanced
7034 Substation Alpha) subtitles files.
7036 This filter accepts the following option in addition to the common options from
7037 the @ref{subtitles} filter:
7041 Set the shaping engine
7043 Available values are:
7046 The default libass shaping engine, which is the best available.
7048 Fast, font-agnostic shaper that can do only substitutions
7050 Slower shaper using OpenType for substitutions and positioning
7053 The default is @code{auto}.
7057 Apply an Adaptive Temporal Averaging Denoiser to the video input.
7059 The filter accepts the following options:
7063 Set threshold A for 1st plane. Default is 0.02.
7064 Valid range is 0 to 0.3.
7067 Set threshold B for 1st plane. Default is 0.04.
7068 Valid range is 0 to 5.
7071 Set threshold A for 2nd plane. Default is 0.02.
7072 Valid range is 0 to 0.3.
7075 Set threshold B for 2nd plane. Default is 0.04.
7076 Valid range is 0 to 5.
7079 Set threshold A for 3rd plane. Default is 0.02.
7080 Valid range is 0 to 0.3.
7083 Set threshold B for 3rd plane. Default is 0.04.
7084 Valid range is 0 to 5.
7086 Threshold A is designed to react on abrupt changes in the input signal and
7087 threshold B is designed to react on continuous changes in the input signal.
7090 Set number of frames filter will use for averaging. Default is 9. Must be odd
7091 number in range [5, 129].
7094 Set what planes of frame filter will use for averaging. Default is all.
7097 Set what variant of algorithm filter will use for averaging. Default is @code{p} parallel.
7098 Alternatively can be set to @code{s} serial.
7100 Parallel can be faster then serial, while other way around is never true.
7101 Parallel will abort early on first change being greater then thresholds, while serial
7102 will continue processing other side of frames if they are equal or below thresholds.
7107 Set sigma for 1st plane, 2nd plane or 3rd plane. Default is 32767.
7108 Valid range is from 0 to 32767.
7109 This options controls weight for each pixel in radius defined by size.
7110 Default value means every pixel have same weight.
7111 Setting this option to 0 effectively disables filtering.
7114 @subsection Commands
7115 This filter supports same @ref{commands} as options except option @code{s}.
7116 The command accepts the same syntax of the corresponding option.
7120 Apply average blur filter.
7122 The filter accepts the following options:
7126 Set horizontal radius size.
7129 Set which planes to filter. By default all planes are filtered.
7132 Set vertical radius size, if zero it will be same as @code{sizeX}.
7133 Default is @code{0}.
7136 @subsection Commands
7137 This filter supports same commands as options.
7138 The command accepts the same syntax of the corresponding option.
7140 If the specified expression is not valid, it is kept at its current
7145 Compute the bounding box for the non-black pixels in the input frame
7148 This filter computes the bounding box containing all the pixels with a
7149 luminance value greater than the minimum allowed value.
7150 The parameters describing the bounding box are printed on the filter
7153 The filter accepts the following option:
7157 Set the minimal luminance value. Default is @code{16}.
7160 @subsection Commands
7162 This filter supports the all above options as @ref{commands}.
7165 Apply bilateral filter, spatial smoothing while preserving edges.
7167 The filter accepts the following options:
7170 Set sigma of gaussian function to calculate spatial weight.
7171 Allowed range is 0 to 512. Default is 0.1.
7174 Set sigma of gaussian function to calculate range weight.
7175 Allowed range is 0 to 1. Default is 0.1.
7178 Set planes to filter. Default is first only.
7181 @subsection Commands
7183 This filter supports the all above options as @ref{commands}.
7185 @section bitplanenoise
7187 Show and measure bit plane noise.
7189 The filter accepts the following options:
7193 Set which plane to analyze. Default is @code{1}.
7196 Filter out noisy pixels from @code{bitplane} set above.
7197 Default is disabled.
7200 @section blackdetect
7202 Detect video intervals that are (almost) completely black. Can be
7203 useful to detect chapter transitions, commercials, or invalid
7206 The filter outputs its detection analysis to both the log as well as
7207 frame metadata. If a black segment of at least the specified minimum
7208 duration is found, a line with the start and end timestamps as well
7209 as duration is printed to the log with level @code{info}. In addition,
7210 a log line with level @code{debug} is printed per frame showing the
7211 black amount detected for that frame.
7213 The filter also attaches metadata to the first frame of a black
7214 segment with key @code{lavfi.black_start} and to the first frame
7215 after the black segment ends with key @code{lavfi.black_end}. The
7216 value is the frame's timestamp. This metadata is added regardless
7217 of the minimum duration specified.
7219 The filter accepts the following options:
7222 @item black_min_duration, d
7223 Set the minimum detected black duration expressed in seconds. It must
7224 be a non-negative floating point number.
7226 Default value is 2.0.
7228 @item picture_black_ratio_th, pic_th
7229 Set the threshold for considering a picture "black".
7230 Express the minimum value for the ratio:
7232 @var{nb_black_pixels} / @var{nb_pixels}
7235 for which a picture is considered black.
7236 Default value is 0.98.
7238 @item pixel_black_th, pix_th
7239 Set the threshold for considering a pixel "black".
7241 The threshold expresses the maximum pixel luminance value for which a
7242 pixel is considered "black". The provided value is scaled according to
7243 the following equation:
7245 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
7248 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
7249 the input video format, the range is [0-255] for YUV full-range
7250 formats and [16-235] for YUV non full-range formats.
7252 Default value is 0.10.
7255 The following example sets the maximum pixel threshold to the minimum
7256 value, and detects only black intervals of 2 or more seconds:
7258 blackdetect=d=2:pix_th=0.00
7263 Detect frames that are (almost) completely black. Can be useful to
7264 detect chapter transitions or commercials. Output lines consist of
7265 the frame number of the detected frame, the percentage of blackness,
7266 the position in the file if known or -1 and the timestamp in seconds.
7268 In order to display the output lines, you need to set the loglevel at
7269 least to the AV_LOG_INFO value.
7271 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
7272 The value represents the percentage of pixels in the picture that
7273 are below the threshold value.
7275 It accepts the following parameters:
7280 The percentage of the pixels that have to be below the threshold; it defaults to
7283 @item threshold, thresh
7284 The threshold below which a pixel value is considered black; it defaults to
7292 Blend two video frames into each other.
7294 The @code{blend} filter takes two input streams and outputs one
7295 stream, the first input is the "top" layer and second input is
7296 "bottom" layer. By default, the output terminates when the longest input terminates.
7298 The @code{tblend} (time blend) filter takes two consecutive frames
7299 from one single stream, and outputs the result obtained by blending
7300 the new frame on top of the old frame.
7302 A description of the accepted options follows.
7310 Set blend mode for specific pixel component or all pixel components in case
7311 of @var{all_mode}. Default value is @code{normal}.
7313 Available values for component modes are:
7355 Set blend opacity for specific pixel component or all pixel components in case
7356 of @var{all_opacity}. Only used in combination with pixel component blend modes.
7363 Set blend expression for specific pixel component or all pixel components in case
7364 of @var{all_expr}. Note that related mode options will be ignored if those are set.
7366 The expressions can use the following variables:
7370 The sequential number of the filtered frame, starting from @code{0}.
7374 the coordinates of the current sample
7378 the width and height of currently filtered plane
7382 Width and height scale for the plane being filtered. It is the
7383 ratio between the dimensions of the current plane to the luma plane,
7384 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
7385 the luma plane and @code{0.5,0.5} for the chroma planes.
7388 Time of the current frame, expressed in seconds.
7391 Value of pixel component at current location for first video frame (top layer).
7394 Value of pixel component at current location for second video frame (bottom layer).
7398 The @code{blend} filter also supports the @ref{framesync} options.
7400 @subsection Examples
7404 Apply transition from bottom layer to top layer in first 10 seconds:
7406 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
7410 Apply linear horizontal transition from top layer to bottom layer:
7412 blend=all_expr='A*(X/W)+B*(1-X/W)'
7416 Apply 1x1 checkerboard effect:
7418 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
7422 Apply uncover left effect:
7424 blend=all_expr='if(gte(N*SW+X,W),A,B)'
7428 Apply uncover down effect:
7430 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
7434 Apply uncover up-left effect:
7436 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
7440 Split diagonally video and shows top and bottom layer on each side:
7442 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
7446 Display differences between the current and the previous frame:
7448 tblend=all_mode=grainextract
7454 Denoise frames using Block-Matching 3D algorithm.
7456 The filter accepts the following options.
7460 Set denoising strength. Default value is 1.
7461 Allowed range is from 0 to 999.9.
7462 The denoising algorithm is very sensitive to sigma, so adjust it
7463 according to the source.
7466 Set local patch size. This sets dimensions in 2D.
7469 Set sliding step for processing blocks. Default value is 4.
7470 Allowed range is from 1 to 64.
7471 Smaller values allows processing more reference blocks and is slower.
7474 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
7475 When set to 1, no block matching is done. Larger values allows more blocks
7477 Allowed range is from 1 to 256.
7480 Set radius for search block matching. Default is 9.
7481 Allowed range is from 1 to INT32_MAX.
7484 Set step between two search locations for block matching. Default is 1.
7485 Allowed range is from 1 to 64. Smaller is slower.
7488 Set threshold of mean square error for block matching. Valid range is 0 to
7492 Set thresholding parameter for hard thresholding in 3D transformed domain.
7493 Larger values results in stronger hard-thresholding filtering in frequency
7497 Set filtering estimation mode. Can be @code{basic} or @code{final}.
7498 Default is @code{basic}.
7501 If enabled, filter will use 2nd stream for block matching.
7502 Default is disabled for @code{basic} value of @var{estim} option,
7503 and always enabled if value of @var{estim} is @code{final}.
7506 Set planes to filter. Default is all available except alpha.
7509 @subsection Examples
7513 Basic filtering with bm3d:
7515 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
7519 Same as above, but filtering only luma:
7521 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
7525 Same as above, but with both estimation modes:
7527 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
7531 Same as above, but prefilter with @ref{nlmeans} filter instead:
7533 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
7539 Apply a boxblur algorithm to the input video.
7541 It accepts the following parameters:
7545 @item luma_radius, lr
7546 @item luma_power, lp
7547 @item chroma_radius, cr
7548 @item chroma_power, cp
7549 @item alpha_radius, ar
7550 @item alpha_power, ap
7554 A description of the accepted options follows.
7557 @item luma_radius, lr
7558 @item chroma_radius, cr
7559 @item alpha_radius, ar
7560 Set an expression for the box radius in pixels used for blurring the
7561 corresponding input plane.
7563 The radius value must be a non-negative number, and must not be
7564 greater than the value of the expression @code{min(w,h)/2} for the
7565 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
7568 Default value for @option{luma_radius} is "2". If not specified,
7569 @option{chroma_radius} and @option{alpha_radius} default to the
7570 corresponding value set for @option{luma_radius}.
7572 The expressions can contain the following constants:
7576 The input width and height in pixels.
7580 The input chroma image width and height in pixels.
7584 The horizontal and vertical chroma subsample values. For example, for the
7585 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
7588 @item luma_power, lp
7589 @item chroma_power, cp
7590 @item alpha_power, ap
7591 Specify how many times the boxblur filter is applied to the
7592 corresponding plane.
7594 Default value for @option{luma_power} is 2. If not specified,
7595 @option{chroma_power} and @option{alpha_power} default to the
7596 corresponding value set for @option{luma_power}.
7598 A value of 0 will disable the effect.
7601 @subsection Examples
7605 Apply a boxblur filter with the luma, chroma, and alpha radii
7608 boxblur=luma_radius=2:luma_power=1
7613 Set the luma radius to 2, and alpha and chroma radius to 0:
7615 boxblur=2:1:cr=0:ar=0
7619 Set the luma and chroma radii to a fraction of the video dimension:
7621 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
7627 Deinterlace the input video ("bwdif" stands for "Bob Weaver
7628 Deinterlacing Filter").
7630 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
7631 interpolation algorithms.
7632 It accepts the following parameters:
7636 The interlacing mode to adopt. It accepts one of the following values:
7640 Output one frame for each frame.
7642 Output one frame for each field.
7645 The default value is @code{send_field}.
7648 The picture field parity assumed for the input interlaced video. It accepts one
7649 of the following values:
7653 Assume the top field is first.
7655 Assume the bottom field is first.
7657 Enable automatic detection of field parity.
7660 The default value is @code{auto}.
7661 If the interlacing is unknown or the decoder does not export this information,
7662 top field first will be assumed.
7665 Specify which frames to deinterlace. Accepts one of the following
7670 Deinterlace all frames.
7672 Only deinterlace frames marked as interlaced.
7675 The default value is @code{all}.
7680 Apply Contrast Adaptive Sharpen filter to video stream.
7682 The filter accepts the following options:
7686 Set the sharpening strength. Default value is 0.
7689 Set planes to filter. Default value is to filter all
7690 planes except alpha plane.
7693 @subsection Commands
7694 This filter supports same @ref{commands} as options.
7697 Remove all color information for all colors except for certain one.
7699 The filter accepts the following options:
7703 The color which will not be replaced with neutral chroma.
7706 Similarity percentage with the above color.
7707 0.01 matches only the exact key color, while 1.0 matches everything.
7711 0.0 makes pixels either fully gray, or not gray at all.
7712 Higher values result in more preserved color.
7715 Signals that the color passed is already in YUV instead of RGB.
7717 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7718 This can be used to pass exact YUV values as hexadecimal numbers.
7721 @subsection Commands
7722 This filter supports same @ref{commands} as options.
7723 The command accepts the same syntax of the corresponding option.
7725 If the specified expression is not valid, it is kept at its current
7729 YUV colorspace color/chroma keying.
7731 The filter accepts the following options:
7735 The color which will be replaced with transparency.
7738 Similarity percentage with the key color.
7740 0.01 matches only the exact key color, while 1.0 matches everything.
7745 0.0 makes pixels either fully transparent, or not transparent at all.
7747 Higher values result in semi-transparent pixels, with a higher transparency
7748 the more similar the pixels color is to the key color.
7751 Signals that the color passed is already in YUV instead of RGB.
7753 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7754 This can be used to pass exact YUV values as hexadecimal numbers.
7757 @subsection Commands
7758 This filter supports same @ref{commands} as options.
7759 The command accepts the same syntax of the corresponding option.
7761 If the specified expression is not valid, it is kept at its current
7764 @subsection Examples
7768 Make every green pixel in the input image transparent:
7770 ffmpeg -i input.png -vf chromakey=green out.png
7774 Overlay a greenscreen-video on top of a static black background.
7776 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
7781 Reduce chrominance noise.
7783 The filter accepts the following options:
7787 Set threshold for averaging chrominance values.
7788 Sum of absolute difference of Y, U and V pixel components of current
7789 pixel and neighbour pixels lower than this threshold will be used in
7790 averaging. Luma component is left unchanged and is copied to output.
7791 Default value is 30. Allowed range is from 1 to 200.
7794 Set horizontal radius of rectangle used for averaging.
7795 Allowed range is from 1 to 100. Default value is 5.
7798 Set vertical radius of rectangle used for averaging.
7799 Allowed range is from 1 to 100. Default value is 5.
7802 Set horizontal step when averaging. Default value is 1.
7803 Allowed range is from 1 to 50.
7804 Mostly useful to speed-up filtering.
7807 Set vertical step when averaging. Default value is 1.
7808 Allowed range is from 1 to 50.
7809 Mostly useful to speed-up filtering.
7812 Set Y threshold for averaging chrominance values.
7813 Set finer control for max allowed difference between Y components
7814 of current pixel and neigbour pixels.
7815 Default value is 200. Allowed range is from 1 to 200.
7818 Set U threshold for averaging chrominance values.
7819 Set finer control for max allowed difference between U components
7820 of current pixel and neigbour pixels.
7821 Default value is 200. Allowed range is from 1 to 200.
7824 Set V threshold for averaging chrominance values.
7825 Set finer control for max allowed difference between V components
7826 of current pixel and neigbour pixels.
7827 Default value is 200. Allowed range is from 1 to 200.
7830 @subsection Commands
7831 This filter supports same @ref{commands} as options.
7832 The command accepts the same syntax of the corresponding option.
7834 @section chromashift
7835 Shift chroma pixels horizontally and/or vertically.
7837 The filter accepts the following options:
7840 Set amount to shift chroma-blue horizontally.
7842 Set amount to shift chroma-blue vertically.
7844 Set amount to shift chroma-red horizontally.
7846 Set amount to shift chroma-red vertically.
7848 Set edge mode, can be @var{smear}, default, or @var{warp}.
7851 @subsection Commands
7853 This filter supports the all above options as @ref{commands}.
7857 Display CIE color diagram with pixels overlaid onto it.
7859 The filter accepts the following options:
7874 @item uhdtv, rec2020
7888 Set what gamuts to draw.
7890 See @code{system} option for available values.
7893 Set ciescope size, by default set to 512.
7896 Set intensity used to map input pixel values to CIE diagram.
7899 Set contrast used to draw tongue colors that are out of active color system gamut.
7902 Correct gamma displayed on scope, by default enabled.
7905 Show white point on CIE diagram, by default disabled.
7908 Set input gamma. Used only with XYZ input color space.
7913 Visualize information exported by some codecs.
7915 Some codecs can export information through frames using side-data or other
7916 means. For example, some MPEG based codecs export motion vectors through the
7917 @var{export_mvs} flag in the codec @option{flags2} option.
7919 The filter accepts the following option:
7923 Set motion vectors to visualize.
7925 Available flags for @var{mv} are:
7929 forward predicted MVs of P-frames
7931 forward predicted MVs of B-frames
7933 backward predicted MVs of B-frames
7937 Display quantization parameters using the chroma planes.
7940 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
7942 Available flags for @var{mv_type} are:
7946 forward predicted MVs
7948 backward predicted MVs
7951 @item frame_type, ft
7952 Set frame type to visualize motion vectors of.
7954 Available flags for @var{frame_type} are:
7958 intra-coded frames (I-frames)
7960 predicted frames (P-frames)
7962 bi-directionally predicted frames (B-frames)
7966 @subsection Examples
7970 Visualize forward predicted MVs of all frames using @command{ffplay}:
7972 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
7976 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
7978 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
7982 @section colorbalance
7983 Modify intensity of primary colors (red, green and blue) of input frames.
7985 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
7986 regions for the red-cyan, green-magenta or blue-yellow balance.
7988 A positive adjustment value shifts the balance towards the primary color, a negative
7989 value towards the complementary color.
7991 The filter accepts the following options:
7997 Adjust red, green and blue shadows (darkest pixels).
8002 Adjust red, green and blue midtones (medium pixels).
8007 Adjust red, green and blue highlights (brightest pixels).
8009 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8012 Preserve lightness when changing color balance. Default is disabled.
8015 @subsection Examples
8019 Add red color cast to shadows:
8025 @subsection Commands
8027 This filter supports the all above options as @ref{commands}.
8029 @section colorcontrast
8031 Adjust color contrast between RGB components.
8033 The filter accepts the following options:
8037 Set the red-cyan contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8040 Set the green-magenta contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8043 Set the blue-yellow contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8048 Set the weight of each @code{rc}, @code{gm}, @code{by} option value. Default value is 0.0.
8049 Allowed range is from 0.0 to 1.0. If all weights are 0.0 filtering is disabled.
8052 Set the amount of preserving lightness. Default value is 0.0. Allowed range is from 0.0 to 1.0.
8055 @subsection Commands
8057 This filter supports the all above options as @ref{commands}.
8059 @section colorchannelmixer
8061 Adjust video input frames by re-mixing color channels.
8063 This filter modifies a color channel by adding the values associated to
8064 the other channels of the same pixels. For example if the value to
8065 modify is red, the output value will be:
8067 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
8070 The filter accepts the following options:
8077 Adjust contribution of input red, green, blue and alpha channels for output red channel.
8078 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
8084 Adjust contribution of input red, green, blue and alpha channels for output green channel.
8085 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
8091 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
8092 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
8098 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
8099 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
8101 Allowed ranges for options are @code{[-2.0, 2.0]}.
8104 Preserve lightness when changing colors. Allowed range is from @code{[0.0, 1.0]}.
8105 Default is @code{0.0}, thus disabled.
8108 @subsection Examples
8112 Convert source to grayscale:
8114 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
8117 Simulate sepia tones:
8119 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
8123 @subsection Commands
8125 This filter supports the all above options as @ref{commands}.
8128 RGB colorspace color keying.
8130 The filter accepts the following options:
8134 The color which will be replaced with transparency.
8137 Similarity percentage with the key color.
8139 0.01 matches only the exact key color, while 1.0 matches everything.
8144 0.0 makes pixels either fully transparent, or not transparent at all.
8146 Higher values result in semi-transparent pixels, with a higher transparency
8147 the more similar the pixels color is to the key color.
8150 @subsection Examples
8154 Make every green pixel in the input image transparent:
8156 ffmpeg -i input.png -vf colorkey=green out.png
8160 Overlay a greenscreen-video on top of a static background image.
8162 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
8166 @subsection Commands
8167 This filter supports same @ref{commands} as options.
8168 The command accepts the same syntax of the corresponding option.
8170 If the specified expression is not valid, it is kept at its current
8174 Remove all color information for all RGB colors except for certain one.
8176 The filter accepts the following options:
8180 The color which will not be replaced with neutral gray.
8183 Similarity percentage with the above color.
8184 0.01 matches only the exact key color, while 1.0 matches everything.
8187 Blend percentage. 0.0 makes pixels fully gray.
8188 Higher values result in more preserved color.
8191 @subsection Commands
8192 This filter supports same @ref{commands} as options.
8193 The command accepts the same syntax of the corresponding option.
8195 If the specified expression is not valid, it is kept at its current
8198 @section colorlevels
8200 Adjust video input frames using levels.
8202 The filter accepts the following options:
8209 Adjust red, green, blue and alpha input black point.
8210 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8216 Adjust red, green, blue and alpha input white point.
8217 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
8219 Input levels are used to lighten highlights (bright tones), darken shadows
8220 (dark tones), change the balance of bright and dark tones.
8226 Adjust red, green, blue and alpha output black point.
8227 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
8233 Adjust red, green, blue and alpha output white point.
8234 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
8236 Output levels allows manual selection of a constrained output level range.
8239 @subsection Examples
8243 Make video output darker:
8245 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
8251 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
8255 Make video output lighter:
8257 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
8261 Increase brightness:
8263 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
8267 @subsection Commands
8269 This filter supports the all above options as @ref{commands}.
8271 @section colormatrix
8273 Convert color matrix.
8275 The filter accepts the following options:
8280 Specify the source and destination color matrix. Both values must be
8283 The accepted values are:
8311 For example to convert from BT.601 to SMPTE-240M, use the command:
8313 colormatrix=bt601:smpte240m
8318 Convert colorspace, transfer characteristics or color primaries.
8319 Input video needs to have an even size.
8321 The filter accepts the following options:
8326 Specify all color properties at once.
8328 The accepted values are:
8358 Specify output colorspace.
8360 The accepted values are:
8369 BT.470BG or BT.601-6 625
8372 SMPTE-170M or BT.601-6 525
8381 BT.2020 with non-constant luminance
8387 Specify output transfer characteristics.
8389 The accepted values are:
8401 Constant gamma of 2.2
8404 Constant gamma of 2.8
8407 SMPTE-170M, BT.601-6 625 or BT.601-6 525
8425 BT.2020 for 10-bits content
8428 BT.2020 for 12-bits content
8434 Specify output color primaries.
8436 The accepted values are:
8445 BT.470BG or BT.601-6 625
8448 SMPTE-170M or BT.601-6 525
8472 Specify output color range.
8474 The accepted values are:
8477 TV (restricted) range
8480 MPEG (restricted) range
8491 Specify output color format.
8493 The accepted values are:
8496 YUV 4:2:0 planar 8-bits
8499 YUV 4:2:0 planar 10-bits
8502 YUV 4:2:0 planar 12-bits
8505 YUV 4:2:2 planar 8-bits
8508 YUV 4:2:2 planar 10-bits
8511 YUV 4:2:2 planar 12-bits
8514 YUV 4:4:4 planar 8-bits
8517 YUV 4:4:4 planar 10-bits
8520 YUV 4:4:4 planar 12-bits
8525 Do a fast conversion, which skips gamma/primary correction. This will take
8526 significantly less CPU, but will be mathematically incorrect. To get output
8527 compatible with that produced by the colormatrix filter, use fast=1.
8530 Specify dithering mode.
8532 The accepted values are:
8538 Floyd-Steinberg dithering
8542 Whitepoint adaptation mode.
8544 The accepted values are:
8547 Bradford whitepoint adaptation
8550 von Kries whitepoint adaptation
8553 identity whitepoint adaptation (i.e. no whitepoint adaptation)
8557 Override all input properties at once. Same accepted values as @ref{all}.
8560 Override input colorspace. Same accepted values as @ref{space}.
8563 Override input color primaries. Same accepted values as @ref{primaries}.
8566 Override input transfer characteristics. Same accepted values as @ref{trc}.
8569 Override input color range. Same accepted values as @ref{range}.
8573 The filter converts the transfer characteristics, color space and color
8574 primaries to the specified user values. The output value, if not specified,
8575 is set to a default value based on the "all" property. If that property is
8576 also not specified, the filter will log an error. The output color range and
8577 format default to the same value as the input color range and format. The
8578 input transfer characteristics, color space, color primaries and color range
8579 should be set on the input data. If any of these are missing, the filter will
8580 log an error and no conversion will take place.
8582 For example to convert the input to SMPTE-240M, use the command:
8584 colorspace=smpte240m
8587 @section colortemperature
8588 Adjust color temperature in video to simulate variations in ambient color temperature.
8590 The filter accepts the following options:
8594 Set the temperature in Kelvin. Allowed range is from 1000 to 40000.
8595 Default value is 6500 K.
8598 Set mixing with filtered output. Allowed range is from 0 to 1.
8602 Set the amount of preserving lightness. Allowed range is from 0 to 1.
8606 @subsection Commands
8607 This filter supports same @ref{commands} as options.
8609 @section convolution
8611 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
8613 The filter accepts the following options:
8620 Set matrix for each plane.
8621 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
8622 and from 1 to 49 odd number of signed integers in @var{row} mode.
8628 Set multiplier for calculated value for each plane.
8629 If unset or 0, it will be sum of all matrix elements.
8635 Set bias for each plane. This value is added to the result of the multiplication.
8636 Useful for making the overall image brighter or darker. Default is 0.0.
8642 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
8643 Default is @var{square}.
8646 @subsection Commands
8648 This filter supports the all above options as @ref{commands}.
8650 @subsection Examples
8656 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"
8662 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"
8668 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"
8674 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"
8678 Apply laplacian edge detector which includes diagonals:
8680 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"
8686 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"
8692 Apply 2D convolution of video stream in frequency domain using second stream
8695 The filter accepts the following options:
8699 Set which planes to process.
8702 Set which impulse video frames will be processed, can be @var{first}
8703 or @var{all}. Default is @var{all}.
8706 The @code{convolve} filter also supports the @ref{framesync} options.
8710 Copy the input video source unchanged to the output. This is mainly useful for
8715 Video filtering on GPU using Apple's CoreImage API on OSX.
8717 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8718 processed by video hardware. However, software-based OpenGL implementations
8719 exist which means there is no guarantee for hardware processing. It depends on
8722 There are many filters and image generators provided by Apple that come with a
8723 large variety of options. The filter has to be referenced by its name along
8726 The coreimage filter accepts the following options:
8729 List all available filters and generators along with all their respective
8730 options as well as possible minimum and maximum values along with the default
8737 Specify all filters by their respective name and options.
8738 Use @var{list_filters} to determine all valid filter names and options.
8739 Numerical options are specified by a float value and are automatically clamped
8740 to their respective value range. Vector and color options have to be specified
8741 by a list of space separated float values. Character escaping has to be done.
8742 A special option name @code{default} is available to use default options for a
8745 It is required to specify either @code{default} or at least one of the filter options.
8746 All omitted options are used with their default values.
8747 The syntax of the filter string is as follows:
8749 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8753 Specify a rectangle where the output of the filter chain is copied into the
8754 input image. It is given by a list of space separated float values:
8756 output_rect=x\ y\ width\ height
8758 If not given, the output rectangle equals the dimensions of the input image.
8759 The output rectangle is automatically cropped at the borders of the input
8760 image. Negative values are valid for each component.
8762 output_rect=25\ 25\ 100\ 100
8766 Several filters can be chained for successive processing without GPU-HOST
8767 transfers allowing for fast processing of complex filter chains.
8768 Currently, only filters with zero (generators) or exactly one (filters) input
8769 image and one output image are supported. Also, transition filters are not yet
8772 Some filters generate output images with additional padding depending on the
8773 respective filter kernel. The padding is automatically removed to ensure the
8774 filter output has the same size as the input image.
8776 For image generators, the size of the output image is determined by the
8777 previous output image of the filter chain or the input image of the whole
8778 filterchain, respectively. The generators do not use the pixel information of
8779 this image to generate their output. However, the generated output is
8780 blended onto this image, resulting in partial or complete coverage of the
8783 The @ref{coreimagesrc} video source can be used for generating input images
8784 which are directly fed into the filter chain. By using it, providing input
8785 images by another video source or an input video is not required.
8787 @subsection Examples
8792 List all filters available:
8794 coreimage=list_filters=true
8798 Use the CIBoxBlur filter with default options to blur an image:
8800 coreimage=filter=CIBoxBlur@@default
8804 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8805 its center at 100x100 and a radius of 50 pixels:
8807 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8811 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8812 given as complete and escaped command-line for Apple's standard bash shell:
8814 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8820 Cover a rectangular object
8822 It accepts the following options:
8826 Filepath of the optional cover image, needs to be in yuv420.
8831 It accepts the following values:
8834 cover it by the supplied image
8836 cover it by interpolating the surrounding pixels
8839 Default value is @var{blur}.
8842 @subsection Examples
8846 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8848 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8854 Crop the input video to given dimensions.
8856 It accepts the following parameters:
8860 The width of the output video. It defaults to @code{iw}.
8861 This expression is evaluated only once during the filter
8862 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8865 The height of the output video. It defaults to @code{ih}.
8866 This expression is evaluated only once during the filter
8867 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8870 The horizontal position, in the input video, of the left edge of the output
8871 video. It defaults to @code{(in_w-out_w)/2}.
8872 This expression is evaluated per-frame.
8875 The vertical position, in the input video, of the top edge of the output video.
8876 It defaults to @code{(in_h-out_h)/2}.
8877 This expression is evaluated per-frame.
8880 If set to 1 will force the output display aspect ratio
8881 to be the same of the input, by changing the output sample aspect
8882 ratio. It defaults to 0.
8885 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8886 width/height/x/y as specified and will not be rounded to nearest smaller value.
8890 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8891 expressions containing the following constants:
8896 The computed values for @var{x} and @var{y}. They are evaluated for
8901 The input width and height.
8905 These are the same as @var{in_w} and @var{in_h}.
8909 The output (cropped) width and height.
8913 These are the same as @var{out_w} and @var{out_h}.
8916 same as @var{iw} / @var{ih}
8919 input sample aspect ratio
8922 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8926 horizontal and vertical chroma subsample values. For example for the
8927 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8930 The number of the input frame, starting from 0.
8933 the position in the file of the input frame, NAN if unknown
8936 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
8940 The expression for @var{out_w} may depend on the value of @var{out_h},
8941 and the expression for @var{out_h} may depend on @var{out_w}, but they
8942 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
8943 evaluated after @var{out_w} and @var{out_h}.
8945 The @var{x} and @var{y} parameters specify the expressions for the
8946 position of the top-left corner of the output (non-cropped) area. They
8947 are evaluated for each frame. If the evaluated value is not valid, it
8948 is approximated to the nearest valid value.
8950 The expression for @var{x} may depend on @var{y}, and the expression
8951 for @var{y} may depend on @var{x}.
8953 @subsection Examples
8957 Crop area with size 100x100 at position (12,34).
8962 Using named options, the example above becomes:
8964 crop=w=100:h=100:x=12:y=34
8968 Crop the central input area with size 100x100:
8974 Crop the central input area with size 2/3 of the input video:
8976 crop=2/3*in_w:2/3*in_h
8980 Crop the input video central square:
8987 Delimit the rectangle with the top-left corner placed at position
8988 100:100 and the right-bottom corner corresponding to the right-bottom
8989 corner of the input image.
8991 crop=in_w-100:in_h-100:100:100
8995 Crop 10 pixels from the left and right borders, and 20 pixels from
8996 the top and bottom borders
8998 crop=in_w-2*10:in_h-2*20
9002 Keep only the bottom right quarter of the input image:
9004 crop=in_w/2:in_h/2:in_w/2:in_h/2
9008 Crop height for getting Greek harmony:
9010 crop=in_w:1/PHI*in_w
9014 Apply trembling effect:
9016 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)
9020 Apply erratic camera effect depending on timestamp:
9022 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)"
9026 Set x depending on the value of y:
9028 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
9032 @subsection Commands
9034 This filter supports the following commands:
9040 Set width/height of the output video and the horizontal/vertical position
9042 The command accepts the same syntax of the corresponding option.
9044 If the specified expression is not valid, it is kept at its current
9050 Auto-detect the crop size.
9052 It calculates the necessary cropping parameters and prints the
9053 recommended parameters via the logging system. The detected dimensions
9054 correspond to the non-black area of the input video.
9056 It accepts the following parameters:
9061 Set higher black value threshold, which can be optionally specified
9062 from nothing (0) to everything (255 for 8-bit based formats). An intensity
9063 value greater to the set value is considered non-black. It defaults to 24.
9064 You can also specify a value between 0.0 and 1.0 which will be scaled depending
9065 on the bitdepth of the pixel format.
9068 The value which the width/height should be divisible by. It defaults to
9069 16. The offset is automatically adjusted to center the video. Use 2 to
9070 get only even dimensions (needed for 4:2:2 video). 16 is best when
9071 encoding to most video codecs.
9074 Set the number of initial frames for which evaluation is skipped.
9075 Default is 2. Range is 0 to INT_MAX.
9077 @item reset_count, reset
9078 Set the counter that determines after how many frames cropdetect will
9079 reset the previously detected largest video area and start over to
9080 detect the current optimal crop area. Default value is 0.
9082 This can be useful when channel logos distort the video area. 0
9083 indicates 'never reset', and returns the largest area encountered during
9090 Delay video filtering until a given wallclock timestamp. The filter first
9091 passes on @option{preroll} amount of frames, then it buffers at most
9092 @option{buffer} amount of frames and waits for the cue. After reaching the cue
9093 it forwards the buffered frames and also any subsequent frames coming in its
9096 The filter can be used synchronize the output of multiple ffmpeg processes for
9097 realtime output devices like decklink. By putting the delay in the filtering
9098 chain and pre-buffering frames the process can pass on data to output almost
9099 immediately after the target wallclock timestamp is reached.
9101 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
9107 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
9110 The duration of content to pass on as preroll expressed in seconds. Default is 0.
9113 The maximum duration of content to buffer before waiting for the cue expressed
9114 in seconds. Default is 0.
9121 Apply color adjustments using curves.
9123 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
9124 component (red, green and blue) has its values defined by @var{N} key points
9125 tied from each other using a smooth curve. The x-axis represents the pixel
9126 values from the input frame, and the y-axis the new pixel values to be set for
9129 By default, a component curve is defined by the two points @var{(0;0)} and
9130 @var{(1;1)}. This creates a straight line where each original pixel value is
9131 "adjusted" to its own value, which means no change to the image.
9133 The filter allows you to redefine these two points and add some more. A new
9134 curve (using a natural cubic spline interpolation) will be define to pass
9135 smoothly through all these new coordinates. The new defined points needs to be
9136 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
9137 be in the @var{[0;1]} interval. If the computed curves happened to go outside
9138 the vector spaces, the values will be clipped accordingly.
9140 The filter accepts the following options:
9144 Select one of the available color presets. This option can be used in addition
9145 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
9146 options takes priority on the preset values.
9147 Available presets are:
9150 @item color_negative
9153 @item increase_contrast
9155 @item linear_contrast
9156 @item medium_contrast
9158 @item strong_contrast
9161 Default is @code{none}.
9163 Set the master key points. These points will define a second pass mapping. It
9164 is sometimes called a "luminance" or "value" mapping. It can be used with
9165 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
9166 post-processing LUT.
9168 Set the key points for the red component.
9170 Set the key points for the green component.
9172 Set the key points for the blue component.
9174 Set the key points for all components (not including master).
9175 Can be used in addition to the other key points component
9176 options. In this case, the unset component(s) will fallback on this
9177 @option{all} setting.
9179 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
9181 Save Gnuplot script of the curves in specified file.
9184 To avoid some filtergraph syntax conflicts, each key points list need to be
9185 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
9187 @subsection Examples
9191 Increase slightly the middle level of blue:
9193 curves=blue='0/0 0.5/0.58 1/1'
9199 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'
9201 Here we obtain the following coordinates for each components:
9204 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9206 @code{(0;0) (0.50;0.48) (1;1)}
9208 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9212 The previous example can also be achieved with the associated built-in preset:
9214 curves=preset=vintage
9224 Use a Photoshop preset and redefine the points of the green component:
9226 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9230 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9231 and @command{gnuplot}:
9233 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9234 gnuplot -p /tmp/curves.plt
9240 Video data analysis filter.
9242 This filter shows hexadecimal pixel values of part of video.
9244 The filter accepts the following options:
9248 Set output video size.
9251 Set x offset from where to pick pixels.
9254 Set y offset from where to pick pixels.
9257 Set scope mode, can be one of the following:
9260 Draw hexadecimal pixel values with white color on black background.
9263 Draw hexadecimal pixel values with input video pixel color on black
9267 Draw hexadecimal pixel values on color background picked from input video,
9268 the text color is picked in such way so its always visible.
9272 Draw rows and columns numbers on left and top of video.
9275 Set background opacity.
9278 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9281 Set pixel components to display. By default all pixel components are displayed.
9285 Apply Directional blur filter.
9287 The filter accepts the following options:
9291 Set angle of directional blur. Default is @code{45}.
9294 Set radius of directional blur. Default is @code{5}.
9297 Set which planes to filter. By default all planes are filtered.
9300 @subsection Commands
9301 This filter supports same @ref{commands} as options.
9302 The command accepts the same syntax of the corresponding option.
9304 If the specified expression is not valid, it is kept at its current
9309 Denoise frames using 2D DCT (frequency domain filtering).
9311 This filter is not designed for real time.
9313 The filter accepts the following options:
9317 Set the noise sigma constant.
9319 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9320 coefficient (absolute value) below this threshold with be dropped.
9322 If you need a more advanced filtering, see @option{expr}.
9324 Default is @code{0}.
9327 Set number overlapping pixels for each block. Since the filter can be slow, you
9328 may want to reduce this value, at the cost of a less effective filter and the
9329 risk of various artefacts.
9331 If the overlapping value doesn't permit processing the whole input width or
9332 height, a warning will be displayed and according borders won't be denoised.
9334 Default value is @var{blocksize}-1, which is the best possible setting.
9337 Set the coefficient factor expression.
9339 For each coefficient of a DCT block, this expression will be evaluated as a
9340 multiplier value for the coefficient.
9342 If this is option is set, the @option{sigma} option will be ignored.
9344 The absolute value of the coefficient can be accessed through the @var{c}
9348 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9349 @var{blocksize}, which is the width and height of the processed blocks.
9351 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9352 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9353 on the speed processing. Also, a larger block size does not necessarily means a
9357 @subsection Examples
9359 Apply a denoise with a @option{sigma} of @code{4.5}:
9364 The same operation can be achieved using the expression system:
9366 dctdnoiz=e='gte(c, 4.5*3)'
9369 Violent denoise using a block size of @code{16x16}:
9376 Remove banding artifacts from input video.
9377 It works by replacing banded pixels with average value of referenced pixels.
9379 The filter accepts the following options:
9386 Set banding detection threshold for each plane. Default is 0.02.
9387 Valid range is 0.00003 to 0.5.
9388 If difference between current pixel and reference pixel is less than threshold,
9389 it will be considered as banded.
9392 Banding detection range in pixels. Default is 16. If positive, random number
9393 in range 0 to set value will be used. If negative, exact absolute value
9395 The range defines square of four pixels around current pixel.
9398 Set direction in radians from which four pixel will be compared. If positive,
9399 random direction from 0 to set direction will be picked. If negative, exact of
9400 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9401 will pick only pixels on same row and -PI/2 will pick only pixels on same
9405 If enabled, current pixel is compared with average value of all four
9406 surrounding pixels. The default is enabled. If disabled current pixel is
9407 compared with all four surrounding pixels. The pixel is considered banded
9408 if only all four differences with surrounding pixels are less than threshold.
9411 If enabled, current pixel is changed if and only if all pixel components are banded,
9412 e.g. banding detection threshold is triggered for all color components.
9413 The default is disabled.
9418 Remove blocking artifacts from input video.
9420 The filter accepts the following options:
9424 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9425 This controls what kind of deblocking is applied.
9428 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9434 Set blocking detection thresholds. Allowed range is 0 to 1.
9435 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9436 Using higher threshold gives more deblocking strength.
9437 Setting @var{alpha} controls threshold detection at exact edge of block.
9438 Remaining options controls threshold detection near the edge. Each one for
9439 below/above or left/right. Setting any of those to @var{0} disables
9443 Set planes to filter. Default is to filter all available planes.
9446 @subsection Examples
9450 Deblock using weak filter and block size of 4 pixels.
9452 deblock=filter=weak:block=4
9456 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9457 deblocking more edges.
9459 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9463 Similar as above, but filter only first plane.
9465 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9469 Similar as above, but filter only second and third plane.
9471 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9478 Drop duplicated frames at regular intervals.
9480 The filter accepts the following options:
9484 Set the number of frames from which one will be dropped. Setting this to
9485 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9486 Default is @code{5}.
9489 Set the threshold for duplicate detection. If the difference metric for a frame
9490 is less than or equal to this value, then it is declared as duplicate. Default
9494 Set scene change threshold. Default is @code{15}.
9498 Set the size of the x and y-axis blocks used during metric calculations.
9499 Larger blocks give better noise suppression, but also give worse detection of
9500 small movements. Must be a power of two. Default is @code{32}.
9503 Mark main input as a pre-processed input and activate clean source input
9504 stream. This allows the input to be pre-processed with various filters to help
9505 the metrics calculation while keeping the frame selection lossless. When set to
9506 @code{1}, the first stream is for the pre-processed input, and the second
9507 stream is the clean source from where the kept frames are chosen. Default is
9511 Set whether or not chroma is considered in the metric calculations. Default is
9517 Apply 2D deconvolution of video stream in frequency domain using second stream
9520 The filter accepts the following options:
9524 Set which planes to process.
9527 Set which impulse video frames will be processed, can be @var{first}
9528 or @var{all}. Default is @var{all}.
9531 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9532 and height are not same and not power of 2 or if stream prior to convolving
9536 The @code{deconvolve} filter also supports the @ref{framesync} options.
9540 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9542 It accepts the following options:
9546 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9547 @var{rainbows} for cross-color reduction.
9550 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9553 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9556 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9559 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9564 Apply deflate effect to the video.
9566 This filter replaces the pixel by the local(3x3) average by taking into account
9567 only values lower than the pixel.
9569 It accepts the following options:
9576 Limit the maximum change for each plane, default is 65535.
9577 If 0, plane will remain unchanged.
9580 @subsection Commands
9582 This filter supports the all above options as @ref{commands}.
9586 Remove temporal frame luminance variations.
9588 It accepts the following options:
9592 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9595 Set averaging mode to smooth temporal luminance variations.
9597 Available values are:
9622 Do not actually modify frame. Useful when one only wants metadata.
9627 Remove judder produced by partially interlaced telecined content.
9629 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9630 source was partially telecined content then the output of @code{pullup,dejudder}
9631 will have a variable frame rate. May change the recorded frame rate of the
9632 container. Aside from that change, this filter will not affect constant frame
9635 The option available in this filter is:
9639 Specify the length of the window over which the judder repeats.
9641 Accepts any integer greater than 1. Useful values are:
9645 If the original was telecined from 24 to 30 fps (Film to NTSC).
9648 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9651 If a mixture of the two.
9654 The default is @samp{4}.
9659 Suppress a TV station logo by a simple interpolation of the surrounding
9660 pixels. Just set a rectangle covering the logo and watch it disappear
9661 (and sometimes something even uglier appear - your mileage may vary).
9663 It accepts the following parameters:
9668 Specify the top left corner coordinates of the logo. They must be
9673 Specify the width and height of the logo to clear. They must be
9677 Specify the thickness of the fuzzy edge of the rectangle (added to
9678 @var{w} and @var{h}). The default value is 1. This option is
9679 deprecated, setting higher values should no longer be necessary and
9683 When set to 1, a green rectangle is drawn on the screen to simplify
9684 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9685 The default value is 0.
9687 The rectangle is drawn on the outermost pixels which will be (partly)
9688 replaced with interpolated values. The values of the next pixels
9689 immediately outside this rectangle in each direction will be used to
9690 compute the interpolated pixel values inside the rectangle.
9694 @subsection Examples
9698 Set a rectangle covering the area with top left corner coordinates 0,0
9699 and size 100x77, and a band of size 10:
9701 delogo=x=0:y=0:w=100:h=77:band=10
9709 Remove the rain in the input image/video by applying the derain methods based on
9710 convolutional neural networks. Supported models:
9714 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9715 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9718 Training as well as model generation scripts are provided in
9719 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9721 Native model files (.model) can be generated from TensorFlow model
9722 files (.pb) by using tools/python/convert.py
9724 The filter accepts the following options:
9728 Specify which filter to use. This option accepts the following values:
9732 Derain filter. To conduct derain filter, you need to use a derain model.
9735 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9737 Default value is @samp{derain}.
9740 Specify which DNN backend to use for model loading and execution. This option accepts
9741 the following values:
9745 Native implementation of DNN loading and execution.
9748 TensorFlow backend. To enable this backend you
9749 need to install the TensorFlow for C library (see
9750 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9751 @code{--enable-libtensorflow}
9753 Default value is @samp{native}.
9756 Set path to model file specifying network architecture and its parameters.
9757 Note that different backends use different file formats. TensorFlow and native
9758 backend can load files for only its format.
9761 It can also be finished with @ref{dnn_processing} filter.
9765 Attempt to fix small changes in horizontal and/or vertical shift. This
9766 filter helps remove camera shake from hand-holding a camera, bumping a
9767 tripod, moving on a vehicle, etc.
9769 The filter accepts the following options:
9777 Specify a rectangular area where to limit the search for motion
9779 If desired the search for motion vectors can be limited to a
9780 rectangular area of the frame defined by its top left corner, width
9781 and height. These parameters have the same meaning as the drawbox
9782 filter which can be used to visualise the position of the bounding
9785 This is useful when simultaneous movement of subjects within the frame
9786 might be confused for camera motion by the motion vector search.
9788 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9789 then the full frame is used. This allows later options to be set
9790 without specifying the bounding box for the motion vector search.
9792 Default - search the whole frame.
9796 Specify the maximum extent of movement in x and y directions in the
9797 range 0-64 pixels. Default 16.
9800 Specify how to generate pixels to fill blanks at the edge of the
9801 frame. Available values are:
9804 Fill zeroes at blank locations
9806 Original image at blank locations
9808 Extruded edge value at blank locations
9810 Mirrored edge at blank locations
9812 Default value is @samp{mirror}.
9815 Specify the blocksize to use for motion search. Range 4-128 pixels,
9819 Specify the contrast threshold for blocks. Only blocks with more than
9820 the specified contrast (difference between darkest and lightest
9821 pixels) will be considered. Range 1-255, default 125.
9824 Specify the search strategy. Available values are:
9827 Set exhaustive search
9829 Set less exhaustive search.
9831 Default value is @samp{exhaustive}.
9834 If set then a detailed log of the motion search is written to the
9841 Remove unwanted contamination of foreground colors, caused by reflected color of
9842 greenscreen or bluescreen.
9844 This filter accepts the following options:
9848 Set what type of despill to use.
9851 Set how spillmap will be generated.
9854 Set how much to get rid of still remaining spill.
9857 Controls amount of red in spill area.
9860 Controls amount of green in spill area.
9861 Should be -1 for greenscreen.
9864 Controls amount of blue in spill area.
9865 Should be -1 for bluescreen.
9868 Controls brightness of spill area, preserving colors.
9871 Modify alpha from generated spillmap.
9874 @subsection Commands
9876 This filter supports the all above options as @ref{commands}.
9880 Apply an exact inverse of the telecine operation. It requires a predefined
9881 pattern specified using the pattern option which must be the same as that passed
9882 to the telecine filter.
9884 This filter accepts the following options:
9893 The default value is @code{top}.
9897 A string of numbers representing the pulldown pattern you wish to apply.
9898 The default value is @code{23}.
9901 A number representing position of the first frame with respect to the telecine
9902 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9907 Apply dilation effect to the video.
9909 This filter replaces the pixel by the local(3x3) maximum.
9911 It accepts the following options:
9918 Limit the maximum change for each plane, default is 65535.
9919 If 0, plane will remain unchanged.
9922 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9925 Flags to local 3x3 coordinates maps like this:
9932 @subsection Commands
9934 This filter supports the all above options as @ref{commands}.
9938 Displace pixels as indicated by second and third input stream.
9940 It takes three input streams and outputs one stream, the first input is the
9941 source, and second and third input are displacement maps.
9943 The second input specifies how much to displace pixels along the
9944 x-axis, while the third input specifies how much to displace pixels
9946 If one of displacement map streams terminates, last frame from that
9947 displacement map will be used.
9949 Note that once generated, displacements maps can be reused over and over again.
9951 A description of the accepted options follows.
9955 Set displace behavior for pixels that are out of range.
9957 Available values are:
9960 Missing pixels are replaced by black pixels.
9963 Adjacent pixels will spread out to replace missing pixels.
9966 Out of range pixels are wrapped so they point to pixels of other side.
9969 Out of range pixels will be replaced with mirrored pixels.
9971 Default is @samp{smear}.
9975 @subsection Examples
9979 Add ripple effect to rgb input of video size hd720:
9981 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
9985 Add wave effect to rgb input of video size hd720:
9987 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
9991 @anchor{dnn_processing}
9992 @section dnn_processing
9994 Do image processing with deep neural networks. It works together with another filter
9995 which converts the pixel format of the Frame to what the dnn network requires.
9997 The filter accepts the following options:
10001 Specify which DNN backend to use for model loading and execution. This option accepts
10002 the following values:
10006 Native implementation of DNN loading and execution.
10009 TensorFlow backend. To enable this backend you
10010 need to install the TensorFlow for C library (see
10011 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10012 @code{--enable-libtensorflow}
10015 OpenVINO backend. To enable this backend you
10016 need to build and install the OpenVINO for C library (see
10017 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10018 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10019 be needed if the header files and libraries are not installed into system path)
10023 Default value is @samp{native}.
10026 Set path to model file specifying network architecture and its parameters.
10027 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10028 backend can load files for only its format.
10030 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10033 Set the input name of the dnn network.
10036 Set the output name of the dnn network.
10039 use DNN async execution if set (default: set),
10040 roll back to sync execution if the backend does not support async.
10044 @subsection Examples
10048 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10050 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10054 Halve the pixel value of the frame with format gray32f:
10056 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
10060 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10062 ./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
10066 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10068 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10075 Draw a colored box on the input image.
10077 It accepts the following parameters:
10082 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10086 The expressions which specify the width and height of the box; if 0 they are interpreted as
10087 the input width and height. It defaults to 0.
10090 Specify the color of the box to write. For the general syntax of this option,
10091 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10092 value @code{invert} is used, the box edge color is the same as the
10093 video with inverted luma.
10096 The expression which sets the thickness of the box edge.
10097 A value of @code{fill} will create a filled box. Default value is @code{3}.
10099 See below for the list of accepted constants.
10102 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10103 will overwrite the video's color and alpha pixels.
10104 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10107 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10108 following constants:
10112 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10116 horizontal and vertical chroma subsample values. For example for the
10117 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10121 The input width and height.
10124 The input sample aspect ratio.
10128 The x and y offset coordinates where the box is drawn.
10132 The width and height of the drawn box.
10135 The thickness of the drawn box.
10137 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10138 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10142 @subsection Examples
10146 Draw a black box around the edge of the input image:
10152 Draw a box with color red and an opacity of 50%:
10154 drawbox=10:20:200:60:red@@0.5
10157 The previous example can be specified as:
10159 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10163 Fill the box with pink color:
10165 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10169 Draw a 2-pixel red 2.40:1 mask:
10171 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
10175 @subsection Commands
10176 This filter supports same commands as options.
10177 The command accepts the same syntax of the corresponding option.
10179 If the specified expression is not valid, it is kept at its current
10184 Draw a graph using input video metadata.
10186 It accepts the following parameters:
10190 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10193 Set 1st foreground color expression.
10196 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10199 Set 2nd foreground color expression.
10202 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10205 Set 3rd foreground color expression.
10208 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10211 Set 4th foreground color expression.
10214 Set minimal value of metadata value.
10217 Set maximal value of metadata value.
10220 Set graph background color. Default is white.
10225 Available values for mode is:
10232 Default is @code{line}.
10237 Available values for slide is:
10240 Draw new frame when right border is reached.
10243 Replace old columns with new ones.
10246 Scroll from right to left.
10249 Scroll from left to right.
10252 Draw single picture.
10255 Default is @code{frame}.
10258 Set size of graph video. For the syntax of this option, check the
10259 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10260 The default value is @code{900x256}.
10263 Set the output frame rate. Default value is @code{25}.
10265 The foreground color expressions can use the following variables:
10268 Minimal value of metadata value.
10271 Maximal value of metadata value.
10274 Current metadata key value.
10277 The color is defined as 0xAABBGGRR.
10280 Example using metadata from @ref{signalstats} filter:
10282 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10285 Example using metadata from @ref{ebur128} filter:
10287 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10292 Draw a grid on the input image.
10294 It accepts the following parameters:
10299 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10303 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10304 input width and height, respectively, minus @code{thickness}, so image gets
10305 framed. Default to 0.
10308 Specify the color of the grid. For the general syntax of this option,
10309 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10310 value @code{invert} is used, the grid color is the same as the
10311 video with inverted luma.
10314 The expression which sets the thickness of the grid line. Default value is @code{1}.
10316 See below for the list of accepted constants.
10319 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10320 will overwrite the video's color and alpha pixels.
10321 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10324 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10325 following constants:
10329 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10333 horizontal and vertical chroma subsample values. For example for the
10334 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10338 The input grid cell width and height.
10341 The input sample aspect ratio.
10345 The x and y coordinates of some point of grid intersection (meant to configure offset).
10349 The width and height of the drawn cell.
10352 The thickness of the drawn cell.
10354 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10355 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10359 @subsection Examples
10363 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10365 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10369 Draw a white 3x3 grid with an opacity of 50%:
10371 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10375 @subsection Commands
10376 This filter supports same commands as options.
10377 The command accepts the same syntax of the corresponding option.
10379 If the specified expression is not valid, it is kept at its current
10385 Draw a text string or text from a specified file on top of a video, using the
10386 libfreetype library.
10388 To enable compilation of this filter, you need to configure FFmpeg with
10389 @code{--enable-libfreetype}.
10390 To enable default font fallback and the @var{font} option you need to
10391 configure FFmpeg with @code{--enable-libfontconfig}.
10392 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10393 @code{--enable-libfribidi}.
10397 It accepts the following parameters:
10402 Used to draw a box around text using the background color.
10403 The value must be either 1 (enable) or 0 (disable).
10404 The default value of @var{box} is 0.
10407 Set the width of the border to be drawn around the box using @var{boxcolor}.
10408 The default value of @var{boxborderw} is 0.
10411 The color to be used for drawing box around text. For the syntax of this
10412 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10414 The default value of @var{boxcolor} is "white".
10417 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10418 The default value of @var{line_spacing} is 0.
10421 Set the width of the border to be drawn around the text using @var{bordercolor}.
10422 The default value of @var{borderw} is 0.
10425 Set the color to be used for drawing border around text. For the syntax of this
10426 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10428 The default value of @var{bordercolor} is "black".
10431 Select how the @var{text} is expanded. Can be either @code{none},
10432 @code{strftime} (deprecated) or
10433 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10437 Set a start time for the count. Value is in microseconds. Only applied
10438 in the deprecated strftime expansion mode. To emulate in normal expansion
10439 mode use the @code{pts} function, supplying the start time (in seconds)
10440 as the second argument.
10443 If true, check and fix text coords to avoid clipping.
10446 The color to be used for drawing fonts. For the syntax of this option, check
10447 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10449 The default value of @var{fontcolor} is "black".
10451 @item fontcolor_expr
10452 String which is expanded the same way as @var{text} to obtain dynamic
10453 @var{fontcolor} value. By default this option has empty value and is not
10454 processed. When this option is set, it overrides @var{fontcolor} option.
10457 The font family to be used for drawing text. By default Sans.
10460 The font file to be used for drawing text. The path must be included.
10461 This parameter is mandatory if the fontconfig support is disabled.
10464 Draw the text applying alpha blending. The value can
10465 be a number between 0.0 and 1.0.
10466 The expression accepts the same variables @var{x, y} as well.
10467 The default value is 1.
10468 Please see @var{fontcolor_expr}.
10471 The font size to be used for drawing text.
10472 The default value of @var{fontsize} is 16.
10475 If set to 1, attempt to shape the text (for example, reverse the order of
10476 right-to-left text and join Arabic characters) before drawing it.
10477 Otherwise, just draw the text exactly as given.
10478 By default 1 (if supported).
10480 @item ft_load_flags
10481 The flags to be used for loading the fonts.
10483 The flags map the corresponding flags supported by libfreetype, and are
10484 a combination of the following values:
10491 @item vertical_layout
10492 @item force_autohint
10495 @item ignore_global_advance_width
10497 @item ignore_transform
10499 @item linear_design
10503 Default value is "default".
10505 For more information consult the documentation for the FT_LOAD_*
10509 The color to be used for drawing a shadow behind the drawn text. For the
10510 syntax of this option, check the @ref{color syntax,,"Color" section in the
10511 ffmpeg-utils manual,ffmpeg-utils}.
10513 The default value of @var{shadowcolor} is "black".
10517 The x and y offsets for the text shadow position with respect to the
10518 position of the text. They can be either positive or negative
10519 values. The default value for both is "0".
10522 The starting frame number for the n/frame_num variable. The default value
10526 The size in number of spaces to use for rendering the tab.
10527 Default value is 4.
10530 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10531 format. It can be used with or without text parameter. @var{timecode_rate}
10532 option must be specified.
10534 @item timecode_rate, rate, r
10535 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10536 integer. Minimum value is "1".
10537 Drop-frame timecode is supported for frame rates 30 & 60.
10540 If set to 1, the output of the timecode option will wrap around at 24 hours.
10541 Default is 0 (disabled).
10544 The text string to be drawn. The text must be a sequence of UTF-8
10545 encoded characters.
10546 This parameter is mandatory if no file is specified with the parameter
10550 A text file containing text to be drawn. The text must be a sequence
10551 of UTF-8 encoded characters.
10553 This parameter is mandatory if no text string is specified with the
10554 parameter @var{text}.
10556 If both @var{text} and @var{textfile} are specified, an error is thrown.
10559 If set to 1, the @var{textfile} will be reloaded before each frame.
10560 Be sure to update it atomically, or it may be read partially, or even fail.
10564 The expressions which specify the offsets where text will be drawn
10565 within the video frame. They are relative to the top/left border of the
10568 The default value of @var{x} and @var{y} is "0".
10570 See below for the list of accepted constants and functions.
10573 The parameters for @var{x} and @var{y} are expressions containing the
10574 following constants and functions:
10578 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10582 horizontal and vertical chroma subsample values. For example for the
10583 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10586 the height of each text line
10594 @item max_glyph_a, ascent
10595 the maximum distance from the baseline to the highest/upper grid
10596 coordinate used to place a glyph outline point, for all the rendered
10598 It is a positive value, due to the grid's orientation with the Y axis
10601 @item max_glyph_d, descent
10602 the maximum distance from the baseline to the lowest grid coordinate
10603 used to place a glyph outline point, for all the rendered glyphs.
10604 This is a negative value, due to the grid's orientation, with the Y axis
10608 maximum glyph height, that is the maximum height for all the glyphs
10609 contained in the rendered text, it is equivalent to @var{ascent} -
10613 maximum glyph width, that is the maximum width for all the glyphs
10614 contained in the rendered text
10617 the number of input frame, starting from 0
10619 @item rand(min, max)
10620 return a random number included between @var{min} and @var{max}
10623 The input sample aspect ratio.
10626 timestamp expressed in seconds, NAN if the input timestamp is unknown
10629 the height of the rendered text
10632 the width of the rendered text
10636 the x and y offset coordinates where the text is drawn.
10638 These parameters allow the @var{x} and @var{y} expressions to refer
10639 to each other, so you can for example specify @code{y=x/dar}.
10642 A one character description of the current frame's picture type.
10645 The current packet's position in the input file or stream
10646 (in bytes, from the start of the input). A value of -1 indicates
10647 this info is not available.
10650 The current packet's duration, in seconds.
10653 The current packet's size (in bytes).
10656 @anchor{drawtext_expansion}
10657 @subsection Text expansion
10659 If @option{expansion} is set to @code{strftime},
10660 the filter recognizes strftime() sequences in the provided text and
10661 expands them accordingly. Check the documentation of strftime(). This
10662 feature is deprecated.
10664 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10666 If @option{expansion} is set to @code{normal} (which is the default),
10667 the following expansion mechanism is used.
10669 The backslash character @samp{\}, followed by any character, always expands to
10670 the second character.
10672 Sequences of the form @code{%@{...@}} are expanded. The text between the
10673 braces is a function name, possibly followed by arguments separated by ':'.
10674 If the arguments contain special characters or delimiters (':' or '@}'),
10675 they should be escaped.
10677 Note that they probably must also be escaped as the value for the
10678 @option{text} option in the filter argument string and as the filter
10679 argument in the filtergraph description, and possibly also for the shell,
10680 that makes up to four levels of escaping; using a text file avoids these
10683 The following functions are available:
10688 The expression evaluation result.
10690 It must take one argument specifying the expression to be evaluated,
10691 which accepts the same constants and functions as the @var{x} and
10692 @var{y} values. Note that not all constants should be used, for
10693 example the text size is not known when evaluating the expression, so
10694 the constants @var{text_w} and @var{text_h} will have an undefined
10697 @item expr_int_format, eif
10698 Evaluate the expression's value and output as formatted integer.
10700 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10701 The second argument specifies the output format. Allowed values are @samp{x},
10702 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10703 @code{printf} function.
10704 The third parameter is optional and sets the number of positions taken by the output.
10705 It can be used to add padding with zeros from the left.
10708 The time at which the filter is running, expressed in UTC.
10709 It can accept an argument: a strftime() format string.
10712 The time at which the filter is running, expressed in the local time zone.
10713 It can accept an argument: a strftime() format string.
10716 Frame metadata. Takes one or two arguments.
10718 The first argument is mandatory and specifies the metadata key.
10720 The second argument is optional and specifies a default value, used when the
10721 metadata key is not found or empty.
10723 Available metadata can be identified by inspecting entries
10724 starting with TAG included within each frame section
10725 printed by running @code{ffprobe -show_frames}.
10727 String metadata generated in filters leading to
10728 the drawtext filter are also available.
10731 The frame number, starting from 0.
10734 A one character description of the current picture type.
10737 The timestamp of the current frame.
10738 It can take up to three arguments.
10740 The first argument is the format of the timestamp; it defaults to @code{flt}
10741 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10742 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10743 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10744 @code{localtime} stands for the timestamp of the frame formatted as
10745 local time zone time.
10747 The second argument is an offset added to the timestamp.
10749 If the format is set to @code{hms}, a third argument @code{24HH} may be
10750 supplied to present the hour part of the formatted timestamp in 24h format
10753 If the format is set to @code{localtime} or @code{gmtime},
10754 a third argument may be supplied: a strftime() format string.
10755 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10758 @subsection Commands
10760 This filter supports altering parameters via commands:
10763 Alter existing filter parameters.
10765 Syntax for the argument is the same as for filter invocation, e.g.
10768 fontsize=56:fontcolor=green:text='Hello World'
10771 Full filter invocation with sendcmd would look like this:
10774 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10778 If the entire argument can't be parsed or applied as valid values then the filter will
10779 continue with its existing parameters.
10781 @subsection Examples
10785 Draw "Test Text" with font FreeSerif, using the default values for the
10786 optional parameters.
10789 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10793 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10794 and y=50 (counting from the top-left corner of the screen), text is
10795 yellow with a red box around it. Both the text and the box have an
10799 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10800 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10803 Note that the double quotes are not necessary if spaces are not used
10804 within the parameter list.
10807 Show the text at the center of the video frame:
10809 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10813 Show the text at a random position, switching to a new position every 30 seconds:
10815 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)"
10819 Show a text line sliding from right to left in the last row of the video
10820 frame. The file @file{LONG_LINE} is assumed to contain a single line
10823 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10827 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10829 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10833 Draw a single green letter "g", at the center of the input video.
10834 The glyph baseline is placed at half screen height.
10836 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10840 Show text for 1 second every 3 seconds:
10842 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10846 Use fontconfig to set the font. Note that the colons need to be escaped.
10848 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10852 Draw "Test Text" with font size dependent on height of the video.
10854 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
10858 Print the date of a real-time encoding (see strftime(3)):
10860 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10864 Show text fading in and out (appearing/disappearing):
10867 DS=1.0 # display start
10868 DE=10.0 # display end
10869 FID=1.5 # fade in duration
10870 FOD=5 # fade out duration
10871 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 @}"
10875 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10876 and the @option{fontsize} value are included in the @option{y} offset.
10878 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10879 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10883 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10884 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10885 must have option @option{-export_path_metadata 1} for the special metadata fields
10886 to be available for filters.
10888 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10893 For more information about libfreetype, check:
10894 @url{http://www.freetype.org/}.
10896 For more information about fontconfig, check:
10897 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10899 For more information about libfribidi, check:
10900 @url{http://fribidi.org/}.
10902 @section edgedetect
10904 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10906 The filter accepts the following options:
10911 Set low and high threshold values used by the Canny thresholding
10914 The high threshold selects the "strong" edge pixels, which are then
10915 connected through 8-connectivity with the "weak" edge pixels selected
10916 by the low threshold.
10918 @var{low} and @var{high} threshold values must be chosen in the range
10919 [0,1], and @var{low} should be lesser or equal to @var{high}.
10921 Default value for @var{low} is @code{20/255}, and default value for @var{high}
10925 Define the drawing mode.
10929 Draw white/gray wires on black background.
10932 Mix the colors to create a paint/cartoon effect.
10935 Apply Canny edge detector on all selected planes.
10937 Default value is @var{wires}.
10940 Select planes for filtering. By default all available planes are filtered.
10943 @subsection Examples
10947 Standard edge detection with custom values for the hysteresis thresholding:
10949 edgedetect=low=0.1:high=0.4
10953 Painting effect without thresholding:
10955 edgedetect=mode=colormix:high=0
10961 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
10963 For each input image, the filter will compute the optimal mapping from
10964 the input to the output given the codebook length, that is the number
10965 of distinct output colors.
10967 This filter accepts the following options.
10970 @item codebook_length, l
10971 Set codebook length. The value must be a positive integer, and
10972 represents the number of distinct output colors. Default value is 256.
10975 Set the maximum number of iterations to apply for computing the optimal
10976 mapping. The higher the value the better the result and the higher the
10977 computation time. Default value is 1.
10980 Set a random seed, must be an integer included between 0 and
10981 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
10982 will try to use a good random seed on a best effort basis.
10985 Set pal8 output pixel format. This option does not work with codebook
10986 length greater than 256.
10991 Measure graylevel entropy in histogram of color channels of video frames.
10993 It accepts the following parameters:
10997 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
10999 @var{diff} mode measures entropy of histogram delta values, absolute differences
11000 between neighbour histogram values.
11004 Apply the EPX magnification filter which is designed for pixel art.
11006 It accepts the following option:
11010 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11012 Default is @code{3}.
11016 Set brightness, contrast, saturation and approximate gamma adjustment.
11018 The filter accepts the following options:
11022 Set the contrast expression. The value must be a float value in range
11023 @code{-1000.0} to @code{1000.0}. The default value is "1".
11026 Set the brightness expression. The value must be a float value in
11027 range @code{-1.0} to @code{1.0}. The default value is "0".
11030 Set the saturation expression. The value must be a float in
11031 range @code{0.0} to @code{3.0}. The default value is "1".
11034 Set the gamma expression. The value must be a float in range
11035 @code{0.1} to @code{10.0}. The default value is "1".
11038 Set the gamma expression for red. The value must be a float in
11039 range @code{0.1} to @code{10.0}. The default value is "1".
11042 Set the gamma expression for green. The value must be a float in range
11043 @code{0.1} to @code{10.0}. The default value is "1".
11046 Set the gamma expression for blue. The value must be a float in range
11047 @code{0.1} to @code{10.0}. The default value is "1".
11050 Set the gamma weight expression. It can be used to reduce the effect
11051 of a high gamma value on bright image areas, e.g. keep them from
11052 getting overamplified and just plain white. The value must be a float
11053 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11054 gamma correction all the way down while @code{1.0} leaves it at its
11055 full strength. Default is "1".
11058 Set when the expressions for brightness, contrast, saturation and
11059 gamma expressions are evaluated.
11061 It accepts the following values:
11064 only evaluate expressions once during the filter initialization or
11065 when a command is processed
11068 evaluate expressions for each incoming frame
11071 Default value is @samp{init}.
11074 The expressions accept the following parameters:
11077 frame count of the input frame starting from 0
11080 byte position of the corresponding packet in the input file, NAN if
11084 frame rate of the input video, NAN if the input frame rate is unknown
11087 timestamp expressed in seconds, NAN if the input timestamp is unknown
11090 @subsection Commands
11091 The filter supports the following commands:
11095 Set the contrast expression.
11098 Set the brightness expression.
11101 Set the saturation expression.
11104 Set the gamma expression.
11107 Set the gamma_r expression.
11110 Set gamma_g expression.
11113 Set gamma_b expression.
11116 Set gamma_weight expression.
11118 The command accepts the same syntax of the corresponding option.
11120 If the specified expression is not valid, it is kept at its current
11127 Apply erosion effect to the video.
11129 This filter replaces the pixel by the local(3x3) minimum.
11131 It accepts the following options:
11138 Limit the maximum change for each plane, default is 65535.
11139 If 0, plane will remain unchanged.
11142 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11145 Flags to local 3x3 coordinates maps like this:
11152 @subsection Commands
11154 This filter supports the all above options as @ref{commands}.
11158 Deinterlace the input video ("estdif" stands for "Edge Slope
11159 Tracing Deinterlacing Filter").
11161 Spatial only filter that uses edge slope tracing algorithm
11162 to interpolate missing lines.
11163 It accepts the following parameters:
11167 The interlacing mode to adopt. It accepts one of the following values:
11171 Output one frame for each frame.
11173 Output one frame for each field.
11176 The default value is @code{field}.
11179 The picture field parity assumed for the input interlaced video. It accepts one
11180 of the following values:
11184 Assume the top field is first.
11186 Assume the bottom field is first.
11188 Enable automatic detection of field parity.
11191 The default value is @code{auto}.
11192 If the interlacing is unknown or the decoder does not export this information,
11193 top field first will be assumed.
11196 Specify which frames to deinterlace. Accepts one of the following
11201 Deinterlace all frames.
11203 Only deinterlace frames marked as interlaced.
11206 The default value is @code{all}.
11209 Specify the search radius for edge slope tracing. Default value is 1.
11210 Allowed range is from 1 to 15.
11213 Specify the search radius for best edge matching. Default value is 2.
11214 Allowed range is from 0 to 15.
11217 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11218 of the following values:
11222 Two-point interpolation.
11224 Four-point interpolation.
11226 Six-point interpolation.
11230 @subsection Commands
11231 This filter supports same @ref{commands} as options.
11233 @section extractplanes
11235 Extract color channel components from input video stream into
11236 separate grayscale video streams.
11238 The filter accepts the following option:
11242 Set plane(s) to extract.
11244 Available values for planes are:
11255 Choosing planes not available in the input will result in an error.
11256 That means you cannot select @code{r}, @code{g}, @code{b} planes
11257 with @code{y}, @code{u}, @code{v} planes at same time.
11260 @subsection Examples
11264 Extract luma, u and v color channel component from input video frame
11265 into 3 grayscale outputs:
11267 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
11273 Apply a fade-in/out effect to the input video.
11275 It accepts the following parameters:
11279 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11281 Default is @code{in}.
11283 @item start_frame, s
11284 Specify the number of the frame to start applying the fade
11285 effect at. Default is 0.
11288 The number of frames that the fade effect lasts. At the end of the
11289 fade-in effect, the output video will have the same intensity as the input video.
11290 At the end of the fade-out transition, the output video will be filled with the
11291 selected @option{color}.
11295 If set to 1, fade only alpha channel, if one exists on the input.
11296 Default value is 0.
11298 @item start_time, st
11299 Specify the timestamp (in seconds) of the frame to start to apply the fade
11300 effect. If both start_frame and start_time are specified, the fade will start at
11301 whichever comes last. Default is 0.
11304 The number of seconds for which the fade effect has to last. At the end of the
11305 fade-in effect the output video will have the same intensity as the input video,
11306 at the end of the fade-out transition the output video will be filled with the
11307 selected @option{color}.
11308 If both duration and nb_frames are specified, duration is used. Default is 0
11309 (nb_frames is used by default).
11312 Specify the color of the fade. Default is "black".
11315 @subsection Examples
11319 Fade in the first 30 frames of video:
11324 The command above is equivalent to:
11330 Fade out the last 45 frames of a 200-frame video:
11333 fade=type=out:start_frame=155:nb_frames=45
11337 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11339 fade=in:0:25, fade=out:975:25
11343 Make the first 5 frames yellow, then fade in from frame 5-24:
11345 fade=in:5:20:color=yellow
11349 Fade in alpha over first 25 frames of video:
11351 fade=in:0:25:alpha=1
11355 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11357 fade=t=in:st=5.5:d=0.5
11363 Denoise frames using 3D FFT (frequency domain filtering).
11365 The filter accepts the following options:
11369 Set the noise sigma constant. This sets denoising strength.
11370 Default value is 1. Allowed range is from 0 to 30.
11371 Using very high sigma with low overlap may give blocking artifacts.
11374 Set amount of denoising. By default all detected noise is reduced.
11375 Default value is 1. Allowed range is from 0 to 1.
11378 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11379 Actual size of block in pixels is 2 to power of @var{block}, so by default
11380 block size in pixels is 2^4 which is 16.
11383 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11386 Set number of previous frames to use for denoising. By default is set to 0.
11389 Set number of next frames to to use for denoising. By default is set to 0.
11392 Set planes which will be filtered, by default are all available filtered
11397 Apply arbitrary expressions to samples in frequency domain
11401 Adjust the dc value (gain) of the luma plane of the image. The filter
11402 accepts an integer value in range @code{0} to @code{1000}. The default
11403 value is set to @code{0}.
11406 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11407 filter accepts an integer value in range @code{0} to @code{1000}. The
11408 default value is set to @code{0}.
11411 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11412 filter accepts an integer value in range @code{0} to @code{1000}. The
11413 default value is set to @code{0}.
11416 Set the frequency domain weight expression for the luma plane.
11419 Set the frequency domain weight expression for the 1st chroma plane.
11422 Set the frequency domain weight expression for the 2nd chroma plane.
11425 Set when the expressions are evaluated.
11427 It accepts the following values:
11430 Only evaluate expressions once during the filter initialization.
11433 Evaluate expressions for each incoming frame.
11436 Default value is @samp{init}.
11438 The filter accepts the following variables:
11441 The coordinates of the current sample.
11445 The width and height of the image.
11448 The number of input frame, starting from 0.
11451 @subsection Examples
11457 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11463 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11469 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11475 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11482 Extract a single field from an interlaced image using stride
11483 arithmetic to avoid wasting CPU time. The output frames are marked as
11486 The filter accepts the following options:
11490 Specify whether to extract the top (if the value is @code{0} or
11491 @code{top}) or the bottom field (if the value is @code{1} or
11497 Create new frames by copying the top and bottom fields from surrounding frames
11498 supplied as numbers by the hint file.
11502 Set file containing hints: absolute/relative frame numbers.
11504 There must be one line for each frame in a clip. Each line must contain two
11505 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11506 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11507 is current frame number for @code{absolute} mode or out of [-1, 1] range
11508 for @code{relative} mode. First number tells from which frame to pick up top
11509 field and second number tells from which frame to pick up bottom field.
11511 If optionally followed by @code{+} output frame will be marked as interlaced,
11512 else if followed by @code{-} output frame will be marked as progressive, else
11513 it will be marked same as input frame.
11514 If optionally followed by @code{t} output frame will use only top field, or in
11515 case of @code{b} it will use only bottom field.
11516 If line starts with @code{#} or @code{;} that line is skipped.
11519 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11522 Example of first several lines of @code{hint} file for @code{relative} mode:
11524 0,0 - # first frame
11525 1,0 - # second frame, use third's frame top field and second's frame bottom field
11526 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11541 @section fieldmatch
11543 Field matching filter for inverse telecine. It is meant to reconstruct the
11544 progressive frames from a telecined stream. The filter does not drop duplicated
11545 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11546 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11548 The separation of the field matching and the decimation is notably motivated by
11549 the possibility of inserting a de-interlacing filter fallback between the two.
11550 If the source has mixed telecined and real interlaced content,
11551 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11552 But these remaining combed frames will be marked as interlaced, and thus can be
11553 de-interlaced by a later filter such as @ref{yadif} before decimation.
11555 In addition to the various configuration options, @code{fieldmatch} can take an
11556 optional second stream, activated through the @option{ppsrc} option. If
11557 enabled, the frames reconstruction will be based on the fields and frames from
11558 this second stream. This allows the first input to be pre-processed in order to
11559 help the various algorithms of the filter, while keeping the output lossless
11560 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11561 or brightness/contrast adjustments can help.
11563 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11564 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11565 which @code{fieldmatch} is based on. While the semantic and usage are very
11566 close, some behaviour and options names can differ.
11568 The @ref{decimate} filter currently only works for constant frame rate input.
11569 If your input has mixed telecined (30fps) and progressive content with a lower
11570 framerate like 24fps use the following filterchain to produce the necessary cfr
11571 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11573 The filter accepts the following options:
11577 Specify the assumed field order of the input stream. Available values are:
11581 Auto detect parity (use FFmpeg's internal parity value).
11583 Assume bottom field first.
11585 Assume top field first.
11588 Note that it is sometimes recommended not to trust the parity announced by the
11591 Default value is @var{auto}.
11594 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11595 sense that it won't risk creating jerkiness due to duplicate frames when
11596 possible, but if there are bad edits or blended fields it will end up
11597 outputting combed frames when a good match might actually exist. On the other
11598 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11599 but will almost always find a good frame if there is one. The other values are
11600 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11601 jerkiness and creating duplicate frames versus finding good matches in sections
11602 with bad edits, orphaned fields, blended fields, etc.
11604 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11606 Available values are:
11610 2-way matching (p/c)
11612 2-way matching, and trying 3rd match if still combed (p/c + n)
11614 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11616 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11617 still combed (p/c + n + u/b)
11619 3-way matching (p/c/n)
11621 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11622 detected as combed (p/c/n + u/b)
11625 The parenthesis at the end indicate the matches that would be used for that
11626 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11629 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11632 Default value is @var{pc_n}.
11635 Mark the main input stream as a pre-processed input, and enable the secondary
11636 input stream as the clean source to pick the fields from. See the filter
11637 introduction for more details. It is similar to the @option{clip2} feature from
11640 Default value is @code{0} (disabled).
11643 Set the field to match from. It is recommended to set this to the same value as
11644 @option{order} unless you experience matching failures with that setting. In
11645 certain circumstances changing the field that is used to match from can have a
11646 large impact on matching performance. Available values are:
11650 Automatic (same value as @option{order}).
11652 Match from the bottom field.
11654 Match from the top field.
11657 Default value is @var{auto}.
11660 Set whether or not chroma is included during the match comparisons. In most
11661 cases it is recommended to leave this enabled. You should set this to @code{0}
11662 only if your clip has bad chroma problems such as heavy rainbowing or other
11663 artifacts. Setting this to @code{0} could also be used to speed things up at
11664 the cost of some accuracy.
11666 Default value is @code{1}.
11670 These define an exclusion band which excludes the lines between @option{y0} and
11671 @option{y1} from being included in the field matching decision. An exclusion
11672 band can be used to ignore subtitles, a logo, or other things that may
11673 interfere with the matching. @option{y0} sets the starting scan line and
11674 @option{y1} sets the ending line; all lines in between @option{y0} and
11675 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11676 @option{y0} and @option{y1} to the same value will disable the feature.
11677 @option{y0} and @option{y1} defaults to @code{0}.
11680 Set the scene change detection threshold as a percentage of maximum change on
11681 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11682 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11683 @option{scthresh} is @code{[0.0, 100.0]}.
11685 Default value is @code{12.0}.
11688 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11689 account the combed scores of matches when deciding what match to use as the
11690 final match. Available values are:
11694 No final matching based on combed scores.
11696 Combed scores are only used when a scene change is detected.
11698 Use combed scores all the time.
11701 Default is @var{sc}.
11704 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11705 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11706 Available values are:
11710 No forced calculation.
11712 Force p/c/n calculations.
11714 Force p/c/n/u/b calculations.
11717 Default value is @var{none}.
11720 This is the area combing threshold used for combed frame detection. This
11721 essentially controls how "strong" or "visible" combing must be to be detected.
11722 Larger values mean combing must be more visible and smaller values mean combing
11723 can be less visible or strong and still be detected. Valid settings are from
11724 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11725 be detected as combed). This is basically a pixel difference value. A good
11726 range is @code{[8, 12]}.
11728 Default value is @code{9}.
11731 Sets whether or not chroma is considered in the combed frame decision. Only
11732 disable this if your source has chroma problems (rainbowing, etc.) that are
11733 causing problems for the combed frame detection with chroma enabled. Actually,
11734 using @option{chroma}=@var{0} is usually more reliable, except for the case
11735 where there is chroma only combing in the source.
11737 Default value is @code{0}.
11741 Respectively set the x-axis and y-axis size of the window used during combed
11742 frame detection. This has to do with the size of the area in which
11743 @option{combpel} pixels are required to be detected as combed for a frame to be
11744 declared combed. See the @option{combpel} parameter description for more info.
11745 Possible values are any number that is a power of 2 starting at 4 and going up
11748 Default value is @code{16}.
11751 The number of combed pixels inside any of the @option{blocky} by
11752 @option{blockx} size blocks on the frame for the frame to be detected as
11753 combed. While @option{cthresh} controls how "visible" the combing must be, this
11754 setting controls "how much" combing there must be in any localized area (a
11755 window defined by the @option{blockx} and @option{blocky} settings) on the
11756 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11757 which point no frames will ever be detected as combed). This setting is known
11758 as @option{MI} in TFM/VFM vocabulary.
11760 Default value is @code{80}.
11763 @anchor{p/c/n/u/b meaning}
11764 @subsection p/c/n/u/b meaning
11766 @subsubsection p/c/n
11768 We assume the following telecined stream:
11771 Top fields: 1 2 2 3 4
11772 Bottom fields: 1 2 3 4 4
11775 The numbers correspond to the progressive frame the fields relate to. Here, the
11776 first two frames are progressive, the 3rd and 4th are combed, and so on.
11778 When @code{fieldmatch} is configured to run a matching from bottom
11779 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11784 B 1 2 3 4 4 <-- matching reference
11793 As a result of the field matching, we can see that some frames get duplicated.
11794 To perform a complete inverse telecine, you need to rely on a decimation filter
11795 after this operation. See for instance the @ref{decimate} filter.
11797 The same operation now matching from top fields (@option{field}=@var{top})
11802 T 1 2 2 3 4 <-- matching reference
11812 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11813 basically, they refer to the frame and field of the opposite parity:
11816 @item @var{p} matches the field of the opposite parity in the previous frame
11817 @item @var{c} matches the field of the opposite parity in the current frame
11818 @item @var{n} matches the field of the opposite parity in the next frame
11823 The @var{u} and @var{b} matching are a bit special in the sense that they match
11824 from the opposite parity flag. In the following examples, we assume that we are
11825 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11826 'x' is placed above and below each matched fields.
11828 With bottom matching (@option{field}=@var{bottom}):
11833 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11834 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11842 With top matching (@option{field}=@var{top}):
11847 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11848 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11856 @subsection Examples
11858 Simple IVTC of a top field first telecined stream:
11860 fieldmatch=order=tff:combmatch=none, decimate
11863 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11865 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11868 @section fieldorder
11870 Transform the field order of the input video.
11872 It accepts the following parameters:
11877 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11878 for bottom field first.
11881 The default value is @samp{tff}.
11883 The transformation is done by shifting the picture content up or down
11884 by one line, and filling the remaining line with appropriate picture content.
11885 This method is consistent with most broadcast field order converters.
11887 If the input video is not flagged as being interlaced, or it is already
11888 flagged as being of the required output field order, then this filter does
11889 not alter the incoming video.
11891 It is very useful when converting to or from PAL DV material,
11892 which is bottom field first.
11896 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11899 @section fifo, afifo
11901 Buffer input images and send them when they are requested.
11903 It is mainly useful when auto-inserted by the libavfilter
11906 It does not take parameters.
11908 @section fillborders
11910 Fill borders of the input video, without changing video stream dimensions.
11911 Sometimes video can have garbage at the four edges and you may not want to
11912 crop video input to keep size multiple of some number.
11914 This filter accepts the following options:
11918 Number of pixels to fill from left border.
11921 Number of pixels to fill from right border.
11924 Number of pixels to fill from top border.
11927 Number of pixels to fill from bottom border.
11932 It accepts the following values:
11935 fill pixels using outermost pixels
11938 fill pixels using mirroring (half sample symmetric)
11941 fill pixels with constant value
11944 fill pixels using reflecting (whole sample symmetric)
11947 fill pixels using wrapping
11950 fade pixels to constant value
11953 Default is @var{smear}.
11956 Set color for pixels in fixed or fade mode. Default is @var{black}.
11959 @subsection Commands
11960 This filter supports same @ref{commands} as options.
11961 The command accepts the same syntax of the corresponding option.
11963 If the specified expression is not valid, it is kept at its current
11968 Find a rectangular object
11970 It accepts the following options:
11974 Filepath of the object image, needs to be in gray8.
11977 Detection threshold, default is 0.5.
11980 Number of mipmaps, default is 3.
11982 @item xmin, ymin, xmax, ymax
11983 Specifies the rectangle in which to search.
11986 @subsection Examples
11990 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
11992 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
11998 Flood area with values of same pixel components with another values.
12000 It accepts the following options:
12003 Set pixel x coordinate.
12006 Set pixel y coordinate.
12009 Set source #0 component value.
12012 Set source #1 component value.
12015 Set source #2 component value.
12018 Set source #3 component value.
12021 Set destination #0 component value.
12024 Set destination #1 component value.
12027 Set destination #2 component value.
12030 Set destination #3 component value.
12036 Convert the input video to one of the specified pixel formats.
12037 Libavfilter will try to pick one that is suitable as input to
12040 It accepts the following parameters:
12044 A '|'-separated list of pixel format names, such as
12045 "pix_fmts=yuv420p|monow|rgb24".
12049 @subsection Examples
12053 Convert the input video to the @var{yuv420p} format
12055 format=pix_fmts=yuv420p
12058 Convert the input video to any of the formats in the list
12060 format=pix_fmts=yuv420p|yuv444p|yuv410p
12067 Convert the video to specified constant frame rate by duplicating or dropping
12068 frames as necessary.
12070 It accepts the following parameters:
12074 The desired output frame rate. The default is @code{25}.
12077 Assume the first PTS should be the given value, in seconds. This allows for
12078 padding/trimming at the start of stream. By default, no assumption is made
12079 about the first frame's expected PTS, so no padding or trimming is done.
12080 For example, this could be set to 0 to pad the beginning with duplicates of
12081 the first frame if a video stream starts after the audio stream or to trim any
12082 frames with a negative PTS.
12085 Timestamp (PTS) rounding method.
12087 Possible values are:
12094 round towards -infinity
12096 round towards +infinity
12100 The default is @code{near}.
12103 Action performed when reading the last frame.
12105 Possible values are:
12108 Use same timestamp rounding method as used for other frames.
12110 Pass through last frame if input duration has not been reached yet.
12112 The default is @code{round}.
12116 Alternatively, the options can be specified as a flat string:
12117 @var{fps}[:@var{start_time}[:@var{round}]].
12119 See also the @ref{setpts} filter.
12121 @subsection Examples
12125 A typical usage in order to set the fps to 25:
12131 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12133 fps=fps=film:round=near
12139 Pack two different video streams into a stereoscopic video, setting proper
12140 metadata on supported codecs. The two views should have the same size and
12141 framerate and processing will stop when the shorter video ends. Please note
12142 that you may conveniently adjust view properties with the @ref{scale} and
12145 It accepts the following parameters:
12149 The desired packing format. Supported values are:
12154 The views are next to each other (default).
12157 The views are on top of each other.
12160 The views are packed by line.
12163 The views are packed by column.
12166 The views are temporally interleaved.
12175 # Convert left and right views into a frame-sequential video
12176 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12178 # Convert views into a side-by-side video with the same output resolution as the input
12179 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
12184 Change the frame rate by interpolating new video output frames from the source
12187 This filter is not designed to function correctly with interlaced media. If
12188 you wish to change the frame rate of interlaced media then you are required
12189 to deinterlace before this filter and re-interlace after this filter.
12191 A description of the accepted options follows.
12195 Specify the output frames per second. This option can also be specified
12196 as a value alone. The default is @code{50}.
12199 Specify the start of a range where the output frame will be created as a
12200 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12201 the default is @code{15}.
12204 Specify the end of a range where the output frame will be created as a
12205 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12206 the default is @code{240}.
12209 Specify the level at which a scene change is detected as a value between
12210 0 and 100 to indicate a new scene; a low value reflects a low
12211 probability for the current frame to introduce a new scene, while a higher
12212 value means the current frame is more likely to be one.
12213 The default is @code{8.2}.
12216 Specify flags influencing the filter process.
12218 Available value for @var{flags} is:
12221 @item scene_change_detect, scd
12222 Enable scene change detection using the value of the option @var{scene}.
12223 This flag is enabled by default.
12229 Select one frame every N-th frame.
12231 This filter accepts the following option:
12234 Select frame after every @code{step} frames.
12235 Allowed values are positive integers higher than 0. Default value is @code{1}.
12238 @section freezedetect
12240 Detect frozen video.
12242 This filter logs a message and sets frame metadata when it detects that the
12243 input video has no significant change in content during a specified duration.
12244 Video freeze detection calculates the mean average absolute difference of all
12245 the components of video frames and compares it to a noise floor.
12247 The printed times and duration are expressed in seconds. The
12248 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12249 whose timestamp equals or exceeds the detection duration and it contains the
12250 timestamp of the first frame of the freeze. The
12251 @code{lavfi.freezedetect.freeze_duration} and
12252 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12255 The filter accepts the following options:
12259 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12260 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12264 Set freeze duration until notification (default is 2 seconds).
12267 @section freezeframes
12269 Freeze video frames.
12271 This filter freezes video frames using frame from 2nd input.
12273 The filter accepts the following options:
12277 Set number of first frame from which to start freeze.
12280 Set number of last frame from which to end freeze.
12283 Set number of frame from 2nd input which will be used instead of replaced frames.
12289 Apply a frei0r effect to the input video.
12291 To enable the compilation of this filter, you need to install the frei0r
12292 header and configure FFmpeg with @code{--enable-frei0r}.
12294 It accepts the following parameters:
12299 The name of the frei0r effect to load. If the environment variable
12300 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12301 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12302 Otherwise, the standard frei0r paths are searched, in this order:
12303 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12304 @file{/usr/lib/frei0r-1/}.
12306 @item filter_params
12307 A '|'-separated list of parameters to pass to the frei0r effect.
12311 A frei0r effect parameter can be a boolean (its value is either
12312 "y" or "n"), a double, a color (specified as
12313 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12314 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12315 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12316 a position (specified as @var{X}/@var{Y}, where
12317 @var{X} and @var{Y} are floating point numbers) and/or a string.
12319 The number and types of parameters depend on the loaded effect. If an
12320 effect parameter is not specified, the default value is set.
12322 @subsection Examples
12326 Apply the distort0r effect, setting the first two double parameters:
12328 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12332 Apply the colordistance effect, taking a color as the first parameter:
12334 frei0r=colordistance:0.2/0.3/0.4
12335 frei0r=colordistance:violet
12336 frei0r=colordistance:0x112233
12340 Apply the perspective effect, specifying the top left and top right image
12343 frei0r=perspective:0.2/0.2|0.8/0.2
12347 For more information, see
12348 @url{http://frei0r.dyne.org}
12350 @subsection Commands
12352 This filter supports the @option{filter_params} option as @ref{commands}.
12356 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12358 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12359 processing filter, one of them is performed once per block, not per pixel.
12360 This allows for much higher speed.
12362 The filter accepts the following options:
12366 Set quality. This option defines the number of levels for averaging. It accepts
12367 an integer in the range 4-5. Default value is @code{4}.
12370 Force a constant quantization parameter. It accepts an integer in range 0-63.
12371 If not set, the filter will use the QP from the video stream (if available).
12374 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12375 more details but also more artifacts, while higher values make the image smoother
12376 but also blurrier. Default value is @code{0} − PSNR optimal.
12378 @item use_bframe_qp
12379 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12380 option may cause flicker since the B-Frames have often larger QP. Default is
12381 @code{0} (not enabled).
12387 Apply Gaussian blur filter.
12389 The filter accepts the following options:
12393 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12396 Set number of steps for Gaussian approximation. Default is @code{1}.
12399 Set which planes to filter. By default all planes are filtered.
12402 Set vertical sigma, if negative it will be same as @code{sigma}.
12403 Default is @code{-1}.
12406 @subsection Commands
12407 This filter supports same commands as options.
12408 The command accepts the same syntax of the corresponding option.
12410 If the specified expression is not valid, it is kept at its current
12415 Apply generic equation to each pixel.
12417 The filter accepts the following options:
12420 @item lum_expr, lum
12421 Set the luminance expression.
12423 Set the chrominance blue expression.
12425 Set the chrominance red expression.
12426 @item alpha_expr, a
12427 Set the alpha expression.
12429 Set the red expression.
12430 @item green_expr, g
12431 Set the green expression.
12433 Set the blue expression.
12436 The colorspace is selected according to the specified options. If one
12437 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12438 options is specified, the filter will automatically select a YCbCr
12439 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12440 @option{blue_expr} options is specified, it will select an RGB
12443 If one of the chrominance expression is not defined, it falls back on the other
12444 one. If no alpha expression is specified it will evaluate to opaque value.
12445 If none of chrominance expressions are specified, they will evaluate
12446 to the luminance expression.
12448 The expressions can use the following variables and functions:
12452 The sequential number of the filtered frame, starting from @code{0}.
12456 The coordinates of the current sample.
12460 The width and height of the image.
12464 Width and height scale depending on the currently filtered plane. It is the
12465 ratio between the corresponding luma plane number of pixels and the current
12466 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12467 @code{0.5,0.5} for chroma planes.
12470 Time of the current frame, expressed in seconds.
12473 Return the value of the pixel at location (@var{x},@var{y}) of the current
12477 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12481 Return the value of the pixel at location (@var{x},@var{y}) of the
12482 blue-difference chroma plane. Return 0 if there is no such plane.
12485 Return the value of the pixel at location (@var{x},@var{y}) of the
12486 red-difference chroma plane. Return 0 if there is no such plane.
12491 Return the value of the pixel at location (@var{x},@var{y}) of the
12492 red/green/blue component. Return 0 if there is no such component.
12495 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12496 plane. Return 0 if there is no such plane.
12498 @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)
12499 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12500 sums of samples within a rectangle. See the functions without the sum postfix.
12502 @item interpolation
12503 Set one of interpolation methods:
12508 Default is bilinear.
12511 For functions, if @var{x} and @var{y} are outside the area, the value will be
12512 automatically clipped to the closer edge.
12514 Please note that this filter can use multiple threads in which case each slice
12515 will have its own expression state. If you want to use only a single expression
12516 state because your expressions depend on previous state then you should limit
12517 the number of filter threads to 1.
12519 @subsection Examples
12523 Flip the image horizontally:
12529 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12530 wavelength of 100 pixels:
12532 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12536 Generate a fancy enigmatic moving light:
12538 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
12542 Generate a quick emboss effect:
12544 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12548 Modify RGB components depending on pixel position:
12550 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12554 Create a radial gradient that is the same size as the input (also see
12555 the @ref{vignette} filter):
12557 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12563 Fix the banding artifacts that are sometimes introduced into nearly flat
12564 regions by truncation to 8-bit color depth.
12565 Interpolate the gradients that should go where the bands are, and
12568 It is designed for playback only. Do not use it prior to
12569 lossy compression, because compression tends to lose the dither and
12570 bring back the bands.
12572 It accepts the following parameters:
12577 The maximum amount by which the filter will change any one pixel. This is also
12578 the threshold for detecting nearly flat regions. Acceptable values range from
12579 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12583 The neighborhood to fit the gradient to. A larger radius makes for smoother
12584 gradients, but also prevents the filter from modifying the pixels near detailed
12585 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12586 values will be clipped to the valid range.
12590 Alternatively, the options can be specified as a flat string:
12591 @var{strength}[:@var{radius}]
12593 @subsection Examples
12597 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12603 Specify radius, omitting the strength (which will fall-back to the default
12611 @anchor{graphmonitor}
12612 @section graphmonitor
12613 Show various filtergraph stats.
12615 With this filter one can debug complete filtergraph.
12616 Especially issues with links filling with queued frames.
12618 The filter accepts the following options:
12622 Set video output size. Default is @var{hd720}.
12625 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12628 Set output mode, can be @var{fulll} or @var{compact}.
12629 In @var{compact} mode only filters with some queued frames have displayed stats.
12632 Set flags which enable which stats are shown in video.
12634 Available values for flags are:
12637 Display number of queued frames in each link.
12639 @item frame_count_in
12640 Display number of frames taken from filter.
12642 @item frame_count_out
12643 Display number of frames given out from filter.
12646 Display current filtered frame pts.
12649 Display current filtered frame time.
12652 Display time base for filter link.
12655 Display used format for filter link.
12658 Display video size or number of audio channels in case of audio used by filter link.
12661 Display video frame rate or sample rate in case of audio used by filter link.
12664 Display link output status.
12668 Set upper limit for video rate of output stream, Default value is @var{25}.
12669 This guarantee that output video frame rate will not be higher than this value.
12673 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12674 and corrects the scene colors accordingly.
12676 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12678 The filter accepts the following options:
12682 The order of differentiation to be applied on the scene. Must be chosen in the range
12683 [0,2] and default value is 1.
12686 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12687 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12688 max value instead of calculating Minkowski distance.
12691 The standard deviation of Gaussian blur to be applied on the scene. Must be
12692 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12693 can't be equal to 0 if @var{difford} is greater than 0.
12696 @subsection Examples
12702 greyedge=difford=1:minknorm=5:sigma=2
12708 greyedge=difford=1:minknorm=0:sigma=2
12716 Apply a Hald CLUT to a video stream.
12718 First input is the video stream to process, and second one is the Hald CLUT.
12719 The Hald CLUT input can be a simple picture or a complete video stream.
12721 The filter accepts the following options:
12725 Force termination when the shortest input terminates. Default is @code{0}.
12727 Continue applying the last CLUT after the end of the stream. A value of
12728 @code{0} disable the filter after the last frame of the CLUT is reached.
12729 Default is @code{1}.
12732 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12733 filters share the same internals).
12735 This filter also supports the @ref{framesync} options.
12737 More information about the Hald CLUT can be found on Eskil Steenberg's website
12738 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12740 @subsection Workflow examples
12742 @subsubsection Hald CLUT video stream
12744 Generate an identity Hald CLUT stream altered with various effects:
12746 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
12749 Note: make sure you use a lossless codec.
12751 Then use it with @code{haldclut} to apply it on some random stream:
12753 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12756 The Hald CLUT will be applied to the 10 first seconds (duration of
12757 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12758 to the remaining frames of the @code{mandelbrot} stream.
12760 @subsubsection Hald CLUT with preview
12762 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12763 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12764 biggest possible square starting at the top left of the picture. The remaining
12765 padding pixels (bottom or right) will be ignored. This area can be used to add
12766 a preview of the Hald CLUT.
12768 Typically, the following generated Hald CLUT will be supported by the
12769 @code{haldclut} filter:
12772 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12773 pad=iw+320 [padded_clut];
12774 smptebars=s=320x256, split [a][b];
12775 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12776 [main][b] overlay=W-320" -frames:v 1 clut.png
12779 It contains the original and a preview of the effect of the CLUT: SMPTE color
12780 bars are displayed on the right-top, and below the same color bars processed by
12783 Then, the effect of this Hald CLUT can be visualized with:
12785 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12790 Flip the input video horizontally.
12792 For example, to horizontally flip the input video with @command{ffmpeg}:
12794 ffmpeg -i in.avi -vf "hflip" out.avi
12798 This filter applies a global color histogram equalization on a
12801 It can be used to correct video that has a compressed range of pixel
12802 intensities. The filter redistributes the pixel intensities to
12803 equalize their distribution across the intensity range. It may be
12804 viewed as an "automatically adjusting contrast filter". This filter is
12805 useful only for correcting degraded or poorly captured source
12808 The filter accepts the following options:
12812 Determine the amount of equalization to be applied. As the strength
12813 is reduced, the distribution of pixel intensities more-and-more
12814 approaches that of the input frame. The value must be a float number
12815 in the range [0,1] and defaults to 0.200.
12818 Set the maximum intensity that can generated and scale the output
12819 values appropriately. The strength should be set as desired and then
12820 the intensity can be limited if needed to avoid washing-out. The value
12821 must be a float number in the range [0,1] and defaults to 0.210.
12824 Set the antibanding level. If enabled the filter will randomly vary
12825 the luminance of output pixels by a small amount to avoid banding of
12826 the histogram. Possible values are @code{none}, @code{weak} or
12827 @code{strong}. It defaults to @code{none}.
12833 Compute and draw a color distribution histogram for the input video.
12835 The computed histogram is a representation of the color component
12836 distribution in an image.
12838 Standard histogram displays the color components distribution in an image.
12839 Displays color graph for each color component. Shows distribution of
12840 the Y, U, V, A or R, G, B components, depending on input format, in the
12841 current frame. Below each graph a color component scale meter is shown.
12843 The filter accepts the following options:
12847 Set height of level. Default value is @code{200}.
12848 Allowed range is [50, 2048].
12851 Set height of color scale. Default value is @code{12}.
12852 Allowed range is [0, 40].
12856 It accepts the following values:
12859 Per color component graphs are placed below each other.
12862 Per color component graphs are placed side by side.
12865 Presents information identical to that in the @code{parade}, except
12866 that the graphs representing color components are superimposed directly
12869 Default is @code{stack}.
12872 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12873 Default is @code{linear}.
12876 Set what color components to display.
12877 Default is @code{7}.
12880 Set foreground opacity. Default is @code{0.7}.
12883 Set background opacity. Default is @code{0.5}.
12886 @subsection Examples
12891 Calculate and draw histogram:
12893 ffplay -i input -vf histogram
12901 This is a high precision/quality 3d denoise filter. It aims to reduce
12902 image noise, producing smooth images and making still images really
12903 still. It should enhance compressibility.
12905 It accepts the following optional parameters:
12909 A non-negative floating point number which specifies spatial luma strength.
12910 It defaults to 4.0.
12912 @item chroma_spatial
12913 A non-negative floating point number which specifies spatial chroma strength.
12914 It defaults to 3.0*@var{luma_spatial}/4.0.
12917 A floating point number which specifies luma temporal strength. It defaults to
12918 6.0*@var{luma_spatial}/4.0.
12921 A floating point number which specifies chroma temporal strength. It defaults to
12922 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
12925 @subsection Commands
12926 This filter supports same @ref{commands} as options.
12927 The command accepts the same syntax of the corresponding option.
12929 If the specified expression is not valid, it is kept at its current
12932 @anchor{hwdownload}
12933 @section hwdownload
12935 Download hardware frames to system memory.
12937 The input must be in hardware frames, and the output a non-hardware format.
12938 Not all formats will be supported on the output - it may be necessary to insert
12939 an additional @option{format} filter immediately following in the graph to get
12940 the output in a supported format.
12944 Map hardware frames to system memory or to another device.
12946 This filter has several different modes of operation; which one is used depends
12947 on the input and output formats:
12950 Hardware frame input, normal frame output
12952 Map the input frames to system memory and pass them to the output. If the
12953 original hardware frame is later required (for example, after overlaying
12954 something else on part of it), the @option{hwmap} filter can be used again
12955 in the next mode to retrieve it.
12957 Normal frame input, hardware frame output
12959 If the input is actually a software-mapped hardware frame, then unmap it -
12960 that is, return the original hardware frame.
12962 Otherwise, a device must be provided. Create new hardware surfaces on that
12963 device for the output, then map them back to the software format at the input
12964 and give those frames to the preceding filter. This will then act like the
12965 @option{hwupload} filter, but may be able to avoid an additional copy when
12966 the input is already in a compatible format.
12968 Hardware frame input and output
12970 A device must be supplied for the output, either directly or with the
12971 @option{derive_device} option. The input and output devices must be of
12972 different types and compatible - the exact meaning of this is
12973 system-dependent, but typically it means that they must refer to the same
12974 underlying hardware context (for example, refer to the same graphics card).
12976 If the input frames were originally created on the output device, then unmap
12977 to retrieve the original frames.
12979 Otherwise, map the frames to the output device - create new hardware frames
12980 on the output corresponding to the frames on the input.
12983 The following additional parameters are accepted:
12987 Set the frame mapping mode. Some combination of:
12990 The mapped frame should be readable.
12992 The mapped frame should be writeable.
12994 The mapping will always overwrite the entire frame.
12996 This may improve performance in some cases, as the original contents of the
12997 frame need not be loaded.
12999 The mapping must not involve any copying.
13001 Indirect mappings to copies of frames are created in some cases where either
13002 direct mapping is not possible or it would have unexpected properties.
13003 Setting this flag ensures that the mapping is direct and will fail if that is
13006 Defaults to @var{read+write} if not specified.
13008 @item derive_device @var{type}
13009 Rather than using the device supplied at initialisation, instead derive a new
13010 device of type @var{type} from the device the input frames exist on.
13013 In a hardware to hardware mapping, map in reverse - create frames in the sink
13014 and map them back to the source. This may be necessary in some cases where
13015 a mapping in one direction is required but only the opposite direction is
13016 supported by the devices being used.
13018 This option is dangerous - it may break the preceding filter in undefined
13019 ways if there are any additional constraints on that filter's output.
13020 Do not use it without fully understanding the implications of its use.
13026 Upload system memory frames to hardware surfaces.
13028 The device to upload to must be supplied when the filter is initialised. If
13029 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13030 option or with the @option{derive_device} option. The input and output devices
13031 must be of different types and compatible - the exact meaning of this is
13032 system-dependent, but typically it means that they must refer to the same
13033 underlying hardware context (for example, refer to the same graphics card).
13035 The following additional parameters are accepted:
13038 @item derive_device @var{type}
13039 Rather than using the device supplied at initialisation, instead derive a new
13040 device of type @var{type} from the device the input frames exist on.
13043 @anchor{hwupload_cuda}
13044 @section hwupload_cuda
13046 Upload system memory frames to a CUDA device.
13048 It accepts the following optional parameters:
13052 The number of the CUDA device to use
13057 Apply a high-quality magnification filter designed for pixel art. This filter
13058 was originally created by Maxim Stepin.
13060 It accepts the following option:
13064 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13065 @code{hq3x} and @code{4} for @code{hq4x}.
13066 Default is @code{3}.
13070 Stack input videos horizontally.
13072 All streams must be of same pixel format and of same height.
13074 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13075 to create same output.
13077 The filter accepts the following option:
13081 Set number of input streams. Default is 2.
13084 If set to 1, force the output to terminate when the shortest input
13085 terminates. Default value is 0.
13090 Modify the hue and/or the saturation of the input.
13092 It accepts the following parameters:
13096 Specify the hue angle as a number of degrees. It accepts an expression,
13097 and defaults to "0".
13100 Specify the saturation in the [-10,10] range. It accepts an expression and
13104 Specify the hue angle as a number of radians. It accepts an
13105 expression, and defaults to "0".
13108 Specify the brightness in the [-10,10] range. It accepts an expression and
13112 @option{h} and @option{H} are mutually exclusive, and can't be
13113 specified at the same time.
13115 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13116 expressions containing the following constants:
13120 frame count of the input frame starting from 0
13123 presentation timestamp of the input frame expressed in time base units
13126 frame rate of the input video, NAN if the input frame rate is unknown
13129 timestamp expressed in seconds, NAN if the input timestamp is unknown
13132 time base of the input video
13135 @subsection Examples
13139 Set the hue to 90 degrees and the saturation to 1.0:
13145 Same command but expressing the hue in radians:
13151 Rotate hue and make the saturation swing between 0
13152 and 2 over a period of 1 second:
13154 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13158 Apply a 3 seconds saturation fade-in effect starting at 0:
13160 hue="s=min(t/3\,1)"
13163 The general fade-in expression can be written as:
13165 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13169 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13171 hue="s=max(0\, min(1\, (8-t)/3))"
13174 The general fade-out expression can be written as:
13176 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13181 @subsection Commands
13183 This filter supports the following commands:
13189 Modify the hue and/or the saturation and/or brightness of the input video.
13190 The command accepts the same syntax of the corresponding option.
13192 If the specified expression is not valid, it is kept at its current
13196 @section hysteresis
13198 Grow first stream into second stream by connecting components.
13199 This makes it possible to build more robust edge masks.
13201 This filter accepts the following options:
13205 Set which planes will be processed as bitmap, unprocessed planes will be
13206 copied from first stream.
13207 By default value 0xf, all planes will be processed.
13210 Set threshold which is used in filtering. If pixel component value is higher than
13211 this value filter algorithm for connecting components is activated.
13212 By default value is 0.
13215 The @code{hysteresis} filter also supports the @ref{framesync} options.
13219 Detect video interlacing type.
13221 This filter tries to detect if the input frames are interlaced, progressive,
13222 top or bottom field first. It will also try to detect fields that are
13223 repeated between adjacent frames (a sign of telecine).
13225 Single frame detection considers only immediately adjacent frames when classifying each frame.
13226 Multiple frame detection incorporates the classification history of previous frames.
13228 The filter will log these metadata values:
13231 @item single.current_frame
13232 Detected type of current frame using single-frame detection. One of:
13233 ``tff'' (top field first), ``bff'' (bottom field first),
13234 ``progressive'', or ``undetermined''
13237 Cumulative number of frames detected as top field first using single-frame detection.
13240 Cumulative number of frames detected as top field first using multiple-frame detection.
13243 Cumulative number of frames detected as bottom field first using single-frame detection.
13245 @item multiple.current_frame
13246 Detected type of current frame using multiple-frame detection. One of:
13247 ``tff'' (top field first), ``bff'' (bottom field first),
13248 ``progressive'', or ``undetermined''
13251 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13253 @item single.progressive
13254 Cumulative number of frames detected as progressive using single-frame detection.
13256 @item multiple.progressive
13257 Cumulative number of frames detected as progressive using multiple-frame detection.
13259 @item single.undetermined
13260 Cumulative number of frames that could not be classified using single-frame detection.
13262 @item multiple.undetermined
13263 Cumulative number of frames that could not be classified using multiple-frame detection.
13265 @item repeated.current_frame
13266 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13268 @item repeated.neither
13269 Cumulative number of frames with no repeated field.
13272 Cumulative number of frames with the top field repeated from the previous frame's top field.
13274 @item repeated.bottom
13275 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13278 The filter accepts the following options:
13282 Set interlacing threshold.
13284 Set progressive threshold.
13286 Threshold for repeated field detection.
13288 Number of frames after which a given frame's contribution to the
13289 statistics is halved (i.e., it contributes only 0.5 to its
13290 classification). The default of 0 means that all frames seen are given
13291 full weight of 1.0 forever.
13292 @item analyze_interlaced_flag
13293 When this is not 0 then idet will use the specified number of frames to determine
13294 if the interlaced flag is accurate, it will not count undetermined frames.
13295 If the flag is found to be accurate it will be used without any further
13296 computations, if it is found to be inaccurate it will be cleared without any
13297 further computations. This allows inserting the idet filter as a low computational
13298 method to clean up the interlaced flag
13303 Deinterleave or interleave fields.
13305 This filter allows one to process interlaced images fields without
13306 deinterlacing them. Deinterleaving splits the input frame into 2
13307 fields (so called half pictures). Odd lines are moved to the top
13308 half of the output image, even lines to the bottom half.
13309 You can process (filter) them independently and then re-interleave them.
13311 The filter accepts the following options:
13315 @item chroma_mode, c
13316 @item alpha_mode, a
13317 Available values for @var{luma_mode}, @var{chroma_mode} and
13318 @var{alpha_mode} are:
13324 @item deinterleave, d
13325 Deinterleave fields, placing one above the other.
13327 @item interleave, i
13328 Interleave fields. Reverse the effect of deinterleaving.
13330 Default value is @code{none}.
13332 @item luma_swap, ls
13333 @item chroma_swap, cs
13334 @item alpha_swap, as
13335 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13338 @subsection Commands
13340 This filter supports the all above options as @ref{commands}.
13344 Apply inflate effect to the video.
13346 This filter replaces the pixel by the local(3x3) average by taking into account
13347 only values higher than the pixel.
13349 It accepts the following options:
13356 Limit the maximum change for each plane, default is 65535.
13357 If 0, plane will remain unchanged.
13360 @subsection Commands
13362 This filter supports the all above options as @ref{commands}.
13366 Simple interlacing filter from progressive contents. This interleaves upper (or
13367 lower) lines from odd frames with lower (or upper) lines from even frames,
13368 halving the frame rate and preserving image height.
13371 Original Original New Frame
13372 Frame 'j' Frame 'j+1' (tff)
13373 ========== =========== ==================
13374 Line 0 --------------------> Frame 'j' Line 0
13375 Line 1 Line 1 ----> Frame 'j+1' Line 1
13376 Line 2 ---------------------> Frame 'j' Line 2
13377 Line 3 Line 3 ----> Frame 'j+1' Line 3
13379 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13382 It accepts the following optional parameters:
13386 This determines whether the interlaced frame is taken from the even
13387 (tff - default) or odd (bff) lines of the progressive frame.
13390 Vertical lowpass filter to avoid twitter interlacing and
13391 reduce moire patterns.
13395 Disable vertical lowpass filter
13398 Enable linear filter (default)
13401 Enable complex filter. This will slightly less reduce twitter and moire
13402 but better retain detail and subjective sharpness impression.
13409 Deinterlace input video by applying Donald Graft's adaptive kernel
13410 deinterling. Work on interlaced parts of a video to produce
13411 progressive frames.
13413 The description of the accepted parameters follows.
13417 Set the threshold which affects the filter's tolerance when
13418 determining if a pixel line must be processed. It must be an integer
13419 in the range [0,255] and defaults to 10. A value of 0 will result in
13420 applying the process on every pixels.
13423 Paint pixels exceeding the threshold value to white if set to 1.
13427 Set the fields order. Swap fields if set to 1, leave fields alone if
13431 Enable additional sharpening if set to 1. Default is 0.
13434 Enable twoway sharpening if set to 1. Default is 0.
13437 @subsection Examples
13441 Apply default values:
13443 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13447 Enable additional sharpening:
13453 Paint processed pixels in white:
13460 Apply kirsch operator to input video stream.
13462 The filter accepts the following option:
13466 Set which planes will be processed, unprocessed planes will be copied.
13467 By default value 0xf, all planes will be processed.
13470 Set value which will be multiplied with filtered result.
13473 Set value which will be added to filtered result.
13476 @subsection Commands
13478 This filter supports the all above options as @ref{commands}.
13482 Slowly update darker pixels.
13484 This filter makes short flashes of light appear longer.
13485 This filter accepts the following options:
13489 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13492 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13495 @subsection Commands
13497 This filter supports the all above options as @ref{commands}.
13499 @section lenscorrection
13501 Correct radial lens distortion
13503 This filter can be used to correct for radial distortion as can result from the use
13504 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13505 one can use tools available for example as part of opencv or simply trial-and-error.
13506 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13507 and extract the k1 and k2 coefficients from the resulting matrix.
13509 Note that effectively the same filter is available in the open-source tools Krita and
13510 Digikam from the KDE project.
13512 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13513 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13514 brightness distribution, so you may want to use both filters together in certain
13515 cases, though you will have to take care of ordering, i.e. whether vignetting should
13516 be applied before or after lens correction.
13518 @subsection Options
13520 The filter accepts the following options:
13524 Relative x-coordinate of the focal point of the image, and thereby the center of the
13525 distortion. This value has a range [0,1] and is expressed as fractions of the image
13526 width. Default is 0.5.
13528 Relative y-coordinate of the focal point of the image, and thereby the center of the
13529 distortion. This value has a range [0,1] and is expressed as fractions of the image
13530 height. Default is 0.5.
13532 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13533 no correction. Default is 0.
13535 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13536 0 means no correction. Default is 0.
13538 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13539 Default is @code{nearest}.
13541 Specify the color of the unmapped pixels. For the syntax of this option,
13542 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13543 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13546 The formula that generates the correction is:
13548 @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)
13550 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13551 distances from the focal point in the source and target images, respectively.
13553 @subsection Commands
13555 This filter supports the all above options as @ref{commands}.
13559 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13561 The @code{lensfun} filter requires the camera make, camera model, and lens model
13562 to apply the lens correction. The filter will load the lensfun database and
13563 query it to find the corresponding camera and lens entries in the database. As
13564 long as these entries can be found with the given options, the filter can
13565 perform corrections on frames. Note that incomplete strings will result in the
13566 filter choosing the best match with the given options, and the filter will
13567 output the chosen camera and lens models (logged with level "info"). You must
13568 provide the make, camera model, and lens model as they are required.
13570 The filter accepts the following options:
13574 The make of the camera (for example, "Canon"). This option is required.
13577 The model of the camera (for example, "Canon EOS 100D"). This option is
13581 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13582 option is required.
13585 The type of correction to apply. The following values are valid options:
13589 Enables fixing lens vignetting.
13592 Enables fixing lens geometry. This is the default.
13595 Enables fixing chromatic aberrations.
13598 Enables fixing lens vignetting and lens geometry.
13601 Enables fixing lens vignetting and chromatic aberrations.
13604 Enables fixing both lens geometry and chromatic aberrations.
13607 Enables all possible corrections.
13611 The focal length of the image/video (zoom; expected constant for video). For
13612 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13613 range should be chosen when using that lens. Default 18.
13616 The aperture of the image/video (expected constant for video). Note that
13617 aperture is only used for vignetting correction. Default 3.5.
13619 @item focus_distance
13620 The focus distance of the image/video (expected constant for video). Note that
13621 focus distance is only used for vignetting and only slightly affects the
13622 vignetting correction process. If unknown, leave it at the default value (which
13626 The scale factor which is applied after transformation. After correction the
13627 video is no longer necessarily rectangular. This parameter controls how much of
13628 the resulting image is visible. The value 0 means that a value will be chosen
13629 automatically such that there is little or no unmapped area in the output
13630 image. 1.0 means that no additional scaling is done. Lower values may result
13631 in more of the corrected image being visible, while higher values may avoid
13632 unmapped areas in the output.
13634 @item target_geometry
13635 The target geometry of the output image/video. The following values are valid
13639 @item rectilinear (default)
13642 @item equirectangular
13643 @item fisheye_orthographic
13644 @item fisheye_stereographic
13645 @item fisheye_equisolid
13646 @item fisheye_thoby
13649 Apply the reverse of image correction (instead of correcting distortion, apply
13652 @item interpolation
13653 The type of interpolation used when correcting distortion. The following values
13658 @item linear (default)
13663 @subsection Examples
13667 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13668 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13672 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
13676 Apply the same as before, but only for the first 5 seconds of video.
13679 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
13686 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13687 score between two input videos.
13689 The obtained VMAF score is printed through the logging system.
13691 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13692 After installing the library it can be enabled using:
13693 @code{./configure --enable-libvmaf}.
13694 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13696 The filter has following options:
13700 Set the model path which is to be used for SVM.
13701 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13704 Set the file path to be used to store logs.
13707 Set the format of the log file (csv, json or xml).
13709 @item enable_transform
13710 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13711 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13712 Default value: @code{false}
13715 Invokes the phone model which will generate VMAF scores higher than in the
13716 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13717 Default value: @code{false}
13720 Enables computing psnr along with vmaf.
13721 Default value: @code{false}
13724 Enables computing ssim along with vmaf.
13725 Default value: @code{false}
13728 Enables computing ms_ssim along with vmaf.
13729 Default value: @code{false}
13732 Set the pool method to be used for computing vmaf.
13733 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13736 Set number of threads to be used when computing vmaf.
13737 Default value: @code{0}, which makes use of all available logical processors.
13740 Set interval for frame subsampling used when computing vmaf.
13741 Default value: @code{1}
13743 @item enable_conf_interval
13744 Enables confidence interval.
13745 Default value: @code{false}
13748 This filter also supports the @ref{framesync} options.
13750 @subsection Examples
13753 On the below examples the input file @file{main.mpg} being processed is
13754 compared with the reference file @file{ref.mpg}.
13757 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13761 Example with options:
13763 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13767 Example with options and different containers:
13769 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 -
13775 Limits the pixel components values to the specified range [min, max].
13777 The filter accepts the following options:
13781 Lower bound. Defaults to the lowest allowed value for the input.
13784 Upper bound. Defaults to the highest allowed value for the input.
13787 Specify which planes will be processed. Defaults to all available.
13790 @subsection Commands
13792 This filter supports the all above options as @ref{commands}.
13798 The filter accepts the following options:
13802 Set the number of loops. Setting this value to -1 will result in infinite loops.
13806 Set maximal size in number of frames. Default is 0.
13809 Set first frame of loop. Default is 0.
13812 @subsection Examples
13816 Loop single first frame infinitely:
13818 loop=loop=-1:size=1:start=0
13822 Loop single first frame 10 times:
13824 loop=loop=10:size=1:start=0
13828 Loop 10 first frames 5 times:
13830 loop=loop=5:size=10:start=0
13836 Apply a 1D LUT to an input video.
13838 The filter accepts the following options:
13842 Set the 1D LUT file name.
13844 Currently supported formats:
13853 Select interpolation mode.
13855 Available values are:
13859 Use values from the nearest defined point.
13861 Interpolate values using the linear interpolation.
13863 Interpolate values using the cosine interpolation.
13865 Interpolate values using the cubic interpolation.
13867 Interpolate values using the spline interpolation.
13874 Apply a 3D LUT to an input video.
13876 The filter accepts the following options:
13880 Set the 3D LUT file name.
13882 Currently supported formats:
13896 Select interpolation mode.
13898 Available values are:
13902 Use values from the nearest defined point.
13904 Interpolate values using the 8 points defining a cube.
13906 Interpolate values using a tetrahedron.
13908 Interpolate values using a pyramid.
13910 Interpolate values using a prism.
13916 Turn certain luma values into transparency.
13918 The filter accepts the following options:
13922 Set the luma which will be used as base for transparency.
13923 Default value is @code{0}.
13926 Set the range of luma values to be keyed out.
13927 Default value is @code{0.01}.
13930 Set the range of softness. Default value is @code{0}.
13931 Use this to control gradual transition from zero to full transparency.
13934 @subsection Commands
13935 This filter supports same @ref{commands} as options.
13936 The command accepts the same syntax of the corresponding option.
13938 If the specified expression is not valid, it is kept at its current
13941 @section lut, lutrgb, lutyuv
13943 Compute a look-up table for binding each pixel component input value
13944 to an output value, and apply it to the input video.
13946 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
13947 to an RGB input video.
13949 These filters accept the following parameters:
13952 set first pixel component expression
13954 set second pixel component expression
13956 set third pixel component expression
13958 set fourth pixel component expression, corresponds to the alpha component
13961 set red component expression
13963 set green component expression
13965 set blue component expression
13967 alpha component expression
13970 set Y/luminance component expression
13972 set U/Cb component expression
13974 set V/Cr component expression
13977 Each of them specifies the expression to use for computing the lookup table for
13978 the corresponding pixel component values.
13980 The exact component associated to each of the @var{c*} options depends on the
13983 The @var{lut} filter requires either YUV or RGB pixel formats in input,
13984 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
13986 The expressions can contain the following constants and functions:
13991 The input width and height.
13994 The input value for the pixel component.
13997 The input value, clipped to the @var{minval}-@var{maxval} range.
14000 The maximum value for the pixel component.
14003 The minimum value for the pixel component.
14006 The negated value for the pixel component value, clipped to the
14007 @var{minval}-@var{maxval} range; it corresponds to the expression
14008 "maxval-clipval+minval".
14011 The computed value in @var{val}, clipped to the
14012 @var{minval}-@var{maxval} range.
14014 @item gammaval(gamma)
14015 The computed gamma correction value of the pixel component value,
14016 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14018 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14022 All expressions default to "val".
14024 @subsection Examples
14028 Negate input video:
14030 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14031 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14034 The above is the same as:
14036 lutrgb="r=negval:g=negval:b=negval"
14037 lutyuv="y=negval:u=negval:v=negval"
14047 Remove chroma components, turning the video into a graytone image:
14049 lutyuv="u=128:v=128"
14053 Apply a luma burning effect:
14059 Remove green and blue components:
14065 Set a constant alpha channel value on input:
14067 format=rgba,lutrgb=a="maxval-minval/2"
14071 Correct luminance gamma by a factor of 0.5:
14073 lutyuv=y=gammaval(0.5)
14077 Discard least significant bits of luma:
14079 lutyuv=y='bitand(val, 128+64+32)'
14083 Technicolor like effect:
14085 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14089 @section lut2, tlut2
14091 The @code{lut2} filter takes two input streams and outputs one
14094 The @code{tlut2} (time lut2) filter takes two consecutive frames
14095 from one single stream.
14097 This filter accepts the following parameters:
14100 set first pixel component expression
14102 set second pixel component expression
14104 set third pixel component expression
14106 set fourth pixel component expression, corresponds to the alpha component
14109 set output bit depth, only available for @code{lut2} filter. By default is 0,
14110 which means bit depth is automatically picked from first input format.
14113 The @code{lut2} filter also supports the @ref{framesync} options.
14115 Each of them specifies the expression to use for computing the lookup table for
14116 the corresponding pixel component values.
14118 The exact component associated to each of the @var{c*} options depends on the
14121 The expressions can contain the following constants:
14126 The input width and height.
14129 The first input value for the pixel component.
14132 The second input value for the pixel component.
14135 The first input video bit depth.
14138 The second input video bit depth.
14141 All expressions default to "x".
14143 @subsection Examples
14147 Highlight differences between two RGB video streams:
14149 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)'
14153 Highlight differences between two YUV video streams:
14155 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)'
14159 Show max difference between two video streams:
14161 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)))'
14165 @section maskedclamp
14167 Clamp the first input stream with the second input and third input stream.
14169 Returns the value of first stream to be between second input
14170 stream - @code{undershoot} and third input stream + @code{overshoot}.
14172 This filter accepts the following options:
14175 Default value is @code{0}.
14178 Default value is @code{0}.
14181 Set which planes will be processed as bitmap, unprocessed planes will be
14182 copied from first stream.
14183 By default value 0xf, all planes will be processed.
14186 @subsection Commands
14188 This filter supports the all above options as @ref{commands}.
14192 Merge the second and third input stream into output stream using absolute differences
14193 between second input stream and first input stream and absolute difference between
14194 third input stream and first input stream. The picked value will be from second input
14195 stream if second absolute difference is greater than first one or from third input stream
14198 This filter accepts the following options:
14201 Set which planes will be processed as bitmap, unprocessed planes will be
14202 copied from first stream.
14203 By default value 0xf, all planes will be processed.
14206 @subsection Commands
14208 This filter supports the all above options as @ref{commands}.
14210 @section maskedmerge
14212 Merge the first input stream with the second input stream using per pixel
14213 weights in the third input stream.
14215 A value of 0 in the third stream pixel component means that pixel component
14216 from first stream is returned unchanged, while maximum value (eg. 255 for
14217 8-bit videos) means that pixel component from second stream is returned
14218 unchanged. Intermediate values define the amount of merging between both
14219 input stream's pixel components.
14221 This filter accepts the following options:
14224 Set which planes will be processed as bitmap, unprocessed planes will be
14225 copied from first stream.
14226 By default value 0xf, all planes will be processed.
14229 @subsection Commands
14231 This filter supports the all above options as @ref{commands}.
14235 Merge the second and third input stream into output stream using absolute differences
14236 between second input stream and first input stream and absolute difference between
14237 third input stream and first input stream. The picked value will be from second input
14238 stream if second absolute difference is less than first one or from third input stream
14241 This filter accepts the following options:
14244 Set which planes will be processed as bitmap, unprocessed planes will be
14245 copied from first stream.
14246 By default value 0xf, all planes will be processed.
14249 @subsection Commands
14251 This filter supports the all above options as @ref{commands}.
14253 @section maskedthreshold
14254 Pick pixels comparing absolute difference of two video streams with fixed
14257 If absolute difference between pixel component of first and second video
14258 stream is equal or lower than user supplied threshold than pixel component
14259 from first video stream is picked, otherwise pixel component from second
14260 video stream is picked.
14262 This filter accepts the following options:
14265 Set threshold used when picking pixels from absolute difference from two input
14269 Set which planes will be processed as bitmap, unprocessed planes will be
14270 copied from second stream.
14271 By default value 0xf, all planes will be processed.
14274 @subsection Commands
14276 This filter supports the all above options as @ref{commands}.
14279 Create mask from input video.
14281 For example it is useful to create motion masks after @code{tblend} filter.
14283 This filter accepts the following options:
14287 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14290 Set high threshold. Any pixel component higher than this value will be set to max value
14291 allowed for current pixel format.
14294 Set planes to filter, by default all available planes are filtered.
14297 Fill all frame pixels with this value.
14300 Set max average pixel value for frame. If sum of all pixel components is higher that this
14301 average, output frame will be completely filled with value set by @var{fill} option.
14302 Typically useful for scene changes when used in combination with @code{tblend} filter.
14307 Apply motion-compensation deinterlacing.
14309 It needs one field per frame as input and must thus be used together
14310 with yadif=1/3 or equivalent.
14312 This filter accepts the following options:
14315 Set the deinterlacing mode.
14317 It accepts one of the following values:
14322 use iterative motion estimation
14324 like @samp{slow}, but use multiple reference frames.
14326 Default value is @samp{fast}.
14329 Set the picture field parity assumed for the input video. It must be
14330 one of the following values:
14334 assume top field first
14336 assume bottom field first
14339 Default value is @samp{bff}.
14342 Set per-block quantization parameter (QP) used by the internal
14345 Higher values should result in a smoother motion vector field but less
14346 optimal individual vectors. Default value is 1.
14351 Pick median pixel from certain rectangle defined by radius.
14353 This filter accepts the following options:
14357 Set horizontal radius size. Default value is @code{1}.
14358 Allowed range is integer from 1 to 127.
14361 Set which planes to process. Default is @code{15}, which is all available planes.
14364 Set vertical radius size. Default value is @code{0}.
14365 Allowed range is integer from 0 to 127.
14366 If it is 0, value will be picked from horizontal @code{radius} option.
14369 Set median percentile. Default value is @code{0.5}.
14370 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14371 minimum values, and @code{1} maximum values.
14374 @subsection Commands
14375 This filter supports same @ref{commands} as options.
14376 The command accepts the same syntax of the corresponding option.
14378 If the specified expression is not valid, it is kept at its current
14381 @section mergeplanes
14383 Merge color channel components from several video streams.
14385 The filter accepts up to 4 input streams, and merge selected input
14386 planes to the output video.
14388 This filter accepts the following options:
14391 Set input to output plane mapping. Default is @code{0}.
14393 The mappings is specified as a bitmap. It should be specified as a
14394 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14395 mapping for the first plane of the output stream. 'A' sets the number of
14396 the input stream to use (from 0 to 3), and 'a' the plane number of the
14397 corresponding input to use (from 0 to 3). The rest of the mappings is
14398 similar, 'Bb' describes the mapping for the output stream second
14399 plane, 'Cc' describes the mapping for the output stream third plane and
14400 'Dd' describes the mapping for the output stream fourth plane.
14403 Set output pixel format. Default is @code{yuva444p}.
14406 @subsection Examples
14410 Merge three gray video streams of same width and height into single video stream:
14412 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14416 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14418 [a0][a1]mergeplanes=0x00010210:yuva444p
14422 Swap Y and A plane in yuva444p stream:
14424 format=yuva444p,mergeplanes=0x03010200:yuva444p
14428 Swap U and V plane in yuv420p stream:
14430 format=yuv420p,mergeplanes=0x000201:yuv420p
14434 Cast a rgb24 clip to yuv444p:
14436 format=rgb24,mergeplanes=0x000102:yuv444p
14442 Estimate and export motion vectors using block matching algorithms.
14443 Motion vectors are stored in frame side data to be used by other filters.
14445 This filter accepts the following options:
14448 Specify the motion estimation method. Accepts one of the following values:
14452 Exhaustive search algorithm.
14454 Three step search algorithm.
14456 Two dimensional logarithmic search algorithm.
14458 New three step search algorithm.
14460 Four step search algorithm.
14462 Diamond search algorithm.
14464 Hexagon-based search algorithm.
14466 Enhanced predictive zonal search algorithm.
14468 Uneven multi-hexagon search algorithm.
14470 Default value is @samp{esa}.
14473 Macroblock size. Default @code{16}.
14476 Search parameter. Default @code{7}.
14479 @section midequalizer
14481 Apply Midway Image Equalization effect using two video streams.
14483 Midway Image Equalization adjusts a pair of images to have the same
14484 histogram, while maintaining their dynamics as much as possible. It's
14485 useful for e.g. matching exposures from a pair of stereo cameras.
14487 This filter has two inputs and one output, which must be of same pixel format, but
14488 may be of different sizes. The output of filter is first input adjusted with
14489 midway histogram of both inputs.
14491 This filter accepts the following option:
14495 Set which planes to process. Default is @code{15}, which is all available planes.
14498 @section minterpolate
14500 Convert the video to specified frame rate using motion interpolation.
14502 This filter accepts the following options:
14505 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}.
14508 Motion interpolation mode. Following values are accepted:
14511 Duplicate previous or next frame for interpolating new ones.
14513 Blend source frames. Interpolated frame is mean of previous and next frames.
14515 Motion compensated interpolation. Following options are effective when this mode is selected:
14519 Motion compensation mode. Following values are accepted:
14522 Overlapped block motion compensation.
14524 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14526 Default mode is @samp{obmc}.
14529 Motion estimation mode. Following values are accepted:
14532 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14534 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14536 Default mode is @samp{bilat}.
14539 The algorithm to be used for motion estimation. Following values are accepted:
14542 Exhaustive search algorithm.
14544 Three step search algorithm.
14546 Two dimensional logarithmic search algorithm.
14548 New three step search algorithm.
14550 Four step search algorithm.
14552 Diamond search algorithm.
14554 Hexagon-based search algorithm.
14556 Enhanced predictive zonal search algorithm.
14558 Uneven multi-hexagon search algorithm.
14560 Default algorithm is @samp{epzs}.
14563 Macroblock size. Default @code{16}.
14566 Motion estimation search parameter. Default @code{32}.
14569 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).
14574 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:
14577 Disable scene change detection.
14579 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14581 Default method is @samp{fdiff}.
14583 @item scd_threshold
14584 Scene change detection threshold. Default is @code{10.}.
14589 Mix several video input streams into one video stream.
14591 A description of the accepted options follows.
14595 The number of inputs. If unspecified, it defaults to 2.
14598 Specify weight of each input video stream as sequence.
14599 Each weight is separated by space. If number of weights
14600 is smaller than number of @var{frames} last specified
14601 weight will be used for all remaining unset weights.
14604 Specify scale, if it is set it will be multiplied with sum
14605 of each weight multiplied with pixel values to give final destination
14606 pixel value. By default @var{scale} is auto scaled to sum of weights.
14609 Specify how end of stream is determined.
14612 The duration of the longest input. (default)
14615 The duration of the shortest input.
14618 The duration of the first input.
14622 @section mpdecimate
14624 Drop frames that do not differ greatly from the previous frame in
14625 order to reduce frame rate.
14627 The main use of this filter is for very-low-bitrate encoding
14628 (e.g. streaming over dialup modem), but it could in theory be used for
14629 fixing movies that were inverse-telecined incorrectly.
14631 A description of the accepted options follows.
14635 Set the maximum number of consecutive frames which can be dropped (if
14636 positive), or the minimum interval between dropped frames (if
14637 negative). If the value is 0, the frame is dropped disregarding the
14638 number of previous sequentially dropped frames.
14640 Default value is 0.
14645 Set the dropping threshold values.
14647 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14648 represent actual pixel value differences, so a threshold of 64
14649 corresponds to 1 unit of difference for each pixel, or the same spread
14650 out differently over the block.
14652 A frame is a candidate for dropping if no 8x8 blocks differ by more
14653 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14654 meaning the whole image) differ by more than a threshold of @option{lo}.
14656 Default value for @option{hi} is 64*12, default value for @option{lo} is
14657 64*5, and default value for @option{frac} is 0.33.
14663 Negate (invert) the input video.
14665 It accepts the following option:
14670 With value 1, it negates the alpha component, if present. Default value is 0.
14676 Denoise frames using Non-Local Means algorithm.
14678 Each pixel is adjusted by looking for other pixels with similar contexts. This
14679 context similarity is defined by comparing their surrounding patches of size
14680 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14683 Note that the research area defines centers for patches, which means some
14684 patches will be made of pixels outside that research area.
14686 The filter accepts the following options.
14690 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14693 Set patch size. Default is 7. Must be odd number in range [0, 99].
14696 Same as @option{p} but for chroma planes.
14698 The default value is @var{0} and means automatic.
14701 Set research size. Default is 15. Must be odd number in range [0, 99].
14704 Same as @option{r} but for chroma planes.
14706 The default value is @var{0} and means automatic.
14711 Deinterlace video using neural network edge directed interpolation.
14713 This filter accepts the following options:
14717 Mandatory option, without binary file filter can not work.
14718 Currently file can be found here:
14719 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14722 Set which frames to deinterlace, by default it is @code{all}.
14723 Can be @code{all} or @code{interlaced}.
14726 Set mode of operation.
14728 Can be one of the following:
14732 Use frame flags, both fields.
14734 Use frame flags, single field.
14736 Use top field only.
14738 Use bottom field only.
14740 Use both fields, top first.
14742 Use both fields, bottom first.
14746 Set which planes to process, by default filter process all frames.
14749 Set size of local neighborhood around each pixel, used by the predictor neural
14752 Can be one of the following:
14765 Set the number of neurons in predictor neural network.
14766 Can be one of the following:
14777 Controls the number of different neural network predictions that are blended
14778 together to compute the final output value. Can be @code{fast}, default or
14782 Set which set of weights to use in the predictor.
14783 Can be one of the following:
14787 weights trained to minimize absolute error
14789 weights trained to minimize squared error
14793 Controls whether or not the prescreener neural network is used to decide
14794 which pixels should be processed by the predictor neural network and which
14795 can be handled by simple cubic interpolation.
14796 The prescreener is trained to know whether cubic interpolation will be
14797 sufficient for a pixel or whether it should be predicted by the predictor nn.
14798 The computational complexity of the prescreener nn is much less than that of
14799 the predictor nn. Since most pixels can be handled by cubic interpolation,
14800 using the prescreener generally results in much faster processing.
14801 The prescreener is pretty accurate, so the difference between using it and not
14802 using it is almost always unnoticeable.
14804 Can be one of the following:
14814 Default is @code{new}.
14817 @subsection Commands
14818 This filter supports same @ref{commands} as options, excluding @var{weights} option.
14822 Force libavfilter not to use any of the specified pixel formats for the
14823 input to the next filter.
14825 It accepts the following parameters:
14829 A '|'-separated list of pixel format names, such as
14830 pix_fmts=yuv420p|monow|rgb24".
14834 @subsection Examples
14838 Force libavfilter to use a format different from @var{yuv420p} for the
14839 input to the vflip filter:
14841 noformat=pix_fmts=yuv420p,vflip
14845 Convert the input video to any of the formats not contained in the list:
14847 noformat=yuv420p|yuv444p|yuv410p
14853 Add noise on video input frame.
14855 The filter accepts the following options:
14863 Set noise seed for specific pixel component or all pixel components in case
14864 of @var{all_seed}. Default value is @code{123457}.
14866 @item all_strength, alls
14867 @item c0_strength, c0s
14868 @item c1_strength, c1s
14869 @item c2_strength, c2s
14870 @item c3_strength, c3s
14871 Set noise strength for specific pixel component or all pixel components in case
14872 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14874 @item all_flags, allf
14875 @item c0_flags, c0f
14876 @item c1_flags, c1f
14877 @item c2_flags, c2f
14878 @item c3_flags, c3f
14879 Set pixel component flags or set flags for all components if @var{all_flags}.
14880 Available values for component flags are:
14883 averaged temporal noise (smoother)
14885 mix random noise with a (semi)regular pattern
14887 temporal noise (noise pattern changes between frames)
14889 uniform noise (gaussian otherwise)
14893 @subsection Examples
14895 Add temporal and uniform noise to input video:
14897 noise=alls=20:allf=t+u
14902 Normalize RGB video (aka histogram stretching, contrast stretching).
14903 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
14905 For each channel of each frame, the filter computes the input range and maps
14906 it linearly to the user-specified output range. The output range defaults
14907 to the full dynamic range from pure black to pure white.
14909 Temporal smoothing can be used on the input range to reduce flickering (rapid
14910 changes in brightness) caused when small dark or bright objects enter or leave
14911 the scene. This is similar to the auto-exposure (automatic gain control) on a
14912 video camera, and, like a video camera, it may cause a period of over- or
14913 under-exposure of the video.
14915 The R,G,B channels can be normalized independently, which may cause some
14916 color shifting, or linked together as a single channel, which prevents
14917 color shifting. Linked normalization preserves hue. Independent normalization
14918 does not, so it can be used to remove some color casts. Independent and linked
14919 normalization can be combined in any ratio.
14921 The normalize filter accepts the following options:
14926 Colors which define the output range. The minimum input value is mapped to
14927 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
14928 The defaults are black and white respectively. Specifying white for
14929 @var{blackpt} and black for @var{whitept} will give color-inverted,
14930 normalized video. Shades of grey can be used to reduce the dynamic range
14931 (contrast). Specifying saturated colors here can create some interesting
14935 The number of previous frames to use for temporal smoothing. The input range
14936 of each channel is smoothed using a rolling average over the current frame
14937 and the @var{smoothing} previous frames. The default is 0 (no temporal
14941 Controls the ratio of independent (color shifting) channel normalization to
14942 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
14943 independent. Defaults to 1.0 (fully independent).
14946 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
14947 expensive no-op. Defaults to 1.0 (full strength).
14951 @subsection Commands
14952 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
14953 The command accepts the same syntax of the corresponding option.
14955 If the specified expression is not valid, it is kept at its current
14958 @subsection Examples
14960 Stretch video contrast to use the full dynamic range, with no temporal
14961 smoothing; may flicker depending on the source content:
14963 normalize=blackpt=black:whitept=white:smoothing=0
14966 As above, but with 50 frames of temporal smoothing; flicker should be
14967 reduced, depending on the source content:
14969 normalize=blackpt=black:whitept=white:smoothing=50
14972 As above, but with hue-preserving linked channel normalization:
14974 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
14977 As above, but with half strength:
14979 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
14982 Map the darkest input color to red, the brightest input color to cyan:
14984 normalize=blackpt=red:whitept=cyan
14989 Pass the video source unchanged to the output.
14992 Optical Character Recognition
14994 This filter uses Tesseract for optical character recognition. To enable
14995 compilation of this filter, you need to configure FFmpeg with
14996 @code{--enable-libtesseract}.
14998 It accepts the following options:
15002 Set datapath to tesseract data. Default is to use whatever was
15003 set at installation.
15006 Set language, default is "eng".
15009 Set character whitelist.
15012 Set character blacklist.
15015 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15016 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15020 Apply a video transform using libopencv.
15022 To enable this filter, install the libopencv library and headers and
15023 configure FFmpeg with @code{--enable-libopencv}.
15025 It accepts the following parameters:
15030 The name of the libopencv filter to apply.
15032 @item filter_params
15033 The parameters to pass to the libopencv filter. If not specified, the default
15034 values are assumed.
15038 Refer to the official libopencv documentation for more precise
15040 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15042 Several libopencv filters are supported; see the following subsections.
15047 Dilate an image by using a specific structuring element.
15048 It corresponds to the libopencv function @code{cvDilate}.
15050 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15052 @var{struct_el} represents a structuring element, and has the syntax:
15053 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15055 @var{cols} and @var{rows} represent the number of columns and rows of
15056 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15057 point, and @var{shape} the shape for the structuring element. @var{shape}
15058 must be "rect", "cross", "ellipse", or "custom".
15060 If the value for @var{shape} is "custom", it must be followed by a
15061 string of the form "=@var{filename}". The file with name
15062 @var{filename} is assumed to represent a binary image, with each
15063 printable character corresponding to a bright pixel. When a custom
15064 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15065 or columns and rows of the read file are assumed instead.
15067 The default value for @var{struct_el} is "3x3+0x0/rect".
15069 @var{nb_iterations} specifies the number of times the transform is
15070 applied to the image, and defaults to 1.
15074 # Use the default values
15077 # Dilate using a structuring element with a 5x5 cross, iterating two times
15078 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15080 # Read the shape from the file diamond.shape, iterating two times.
15081 # The file diamond.shape may contain a pattern of characters like this
15087 # The specified columns and rows are ignored
15088 # but the anchor point coordinates are not
15089 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15094 Erode an image by using a specific structuring element.
15095 It corresponds to the libopencv function @code{cvErode}.
15097 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15098 with the same syntax and semantics as the @ref{dilate} filter.
15102 Smooth the input video.
15104 The filter takes the following parameters:
15105 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15107 @var{type} is the type of smooth filter to apply, and must be one of
15108 the following values: "blur", "blur_no_scale", "median", "gaussian",
15109 or "bilateral". The default value is "gaussian".
15111 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15112 depends on the smooth type. @var{param1} and
15113 @var{param2} accept integer positive values or 0. @var{param3} and
15114 @var{param4} accept floating point values.
15116 The default value for @var{param1} is 3. The default value for the
15117 other parameters is 0.
15119 These parameters correspond to the parameters assigned to the
15120 libopencv function @code{cvSmooth}.
15122 @section oscilloscope
15124 2D Video Oscilloscope.
15126 Useful to measure spatial impulse, step responses, chroma delays, etc.
15128 It accepts the following parameters:
15132 Set scope center x position.
15135 Set scope center y position.
15138 Set scope size, relative to frame diagonal.
15141 Set scope tilt/rotation.
15147 Set trace center x position.
15150 Set trace center y position.
15153 Set trace width, relative to width of frame.
15156 Set trace height, relative to height of frame.
15159 Set which components to trace. By default it traces first three components.
15162 Draw trace grid. By default is enabled.
15165 Draw some statistics. By default is enabled.
15168 Draw scope. By default is enabled.
15171 @subsection Commands
15172 This filter supports same @ref{commands} as options.
15173 The command accepts the same syntax of the corresponding option.
15175 If the specified expression is not valid, it is kept at its current
15178 @subsection Examples
15182 Inspect full first row of video frame.
15184 oscilloscope=x=0.5:y=0:s=1
15188 Inspect full last row of video frame.
15190 oscilloscope=x=0.5:y=1:s=1
15194 Inspect full 5th line of video frame of height 1080.
15196 oscilloscope=x=0.5:y=5/1080:s=1
15200 Inspect full last column of video frame.
15202 oscilloscope=x=1:y=0.5:s=1:t=1
15210 Overlay one video on top of another.
15212 It takes two inputs and has one output. The first input is the "main"
15213 video on which the second input is overlaid.
15215 It accepts the following parameters:
15217 A description of the accepted options follows.
15222 Set the expression for the x and y coordinates of the overlaid video
15223 on the main video. Default value is "0" for both expressions. In case
15224 the expression is invalid, it is set to a huge value (meaning that the
15225 overlay will not be displayed within the output visible area).
15228 See @ref{framesync}.
15231 Set when the expressions for @option{x}, and @option{y} are evaluated.
15233 It accepts the following values:
15236 only evaluate expressions once during the filter initialization or
15237 when a command is processed
15240 evaluate expressions for each incoming frame
15243 Default value is @samp{frame}.
15246 See @ref{framesync}.
15249 Set the format for the output video.
15251 It accepts the following values:
15254 force YUV420 output
15257 force YUV420p10 output
15260 force YUV422 output
15263 force YUV422p10 output
15266 force YUV444 output
15269 force packed RGB output
15272 force planar RGB output
15275 automatically pick format
15278 Default value is @samp{yuv420}.
15281 See @ref{framesync}.
15284 Set format of alpha of the overlaid video, it can be @var{straight} or
15285 @var{premultiplied}. Default is @var{straight}.
15288 The @option{x}, and @option{y} expressions can contain the following
15294 The main input width and height.
15298 The overlay input width and height.
15302 The computed values for @var{x} and @var{y}. They are evaluated for
15307 horizontal and vertical chroma subsample values of the output
15308 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15312 the number of input frame, starting from 0
15315 the position in the file of the input frame, NAN if unknown
15318 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15322 This filter also supports the @ref{framesync} options.
15324 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15325 when evaluation is done @emph{per frame}, and will evaluate to NAN
15326 when @option{eval} is set to @samp{init}.
15328 Be aware that frames are taken from each input video in timestamp
15329 order, hence, if their initial timestamps differ, it is a good idea
15330 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15331 have them begin in the same zero timestamp, as the example for
15332 the @var{movie} filter does.
15334 You can chain together more overlays but you should test the
15335 efficiency of such approach.
15337 @subsection Commands
15339 This filter supports the following commands:
15343 Modify the x and y of the overlay input.
15344 The command accepts the same syntax of the corresponding option.
15346 If the specified expression is not valid, it is kept at its current
15350 @subsection Examples
15354 Draw the overlay at 10 pixels from the bottom right corner of the main
15357 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15360 Using named options the example above becomes:
15362 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15366 Insert a transparent PNG logo in the bottom left corner of the input,
15367 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15369 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15373 Insert 2 different transparent PNG logos (second logo on bottom
15374 right corner) using the @command{ffmpeg} tool:
15376 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
15380 Add a transparent color layer on top of the main video; @code{WxH}
15381 must specify the size of the main input to the overlay filter:
15383 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15387 Play an original video and a filtered version (here with the deshake
15388 filter) side by side using the @command{ffplay} tool:
15390 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15393 The above command is the same as:
15395 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15399 Make a sliding overlay appearing from the left to the right top part of the
15400 screen starting since time 2:
15402 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15406 Compose output by putting two input videos side to side:
15408 ffmpeg -i left.avi -i right.avi -filter_complex "
15409 nullsrc=size=200x100 [background];
15410 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15411 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15412 [background][left] overlay=shortest=1 [background+left];
15413 [background+left][right] overlay=shortest=1:x=100 [left+right]
15418 Mask 10-20 seconds of a video by applying the delogo filter to a section
15420 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15421 -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]'
15426 Chain several overlays in cascade:
15428 nullsrc=s=200x200 [bg];
15429 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15430 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15431 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15432 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15433 [in3] null, [mid2] overlay=100:100 [out0]
15438 @anchor{overlay_cuda}
15439 @section overlay_cuda
15441 Overlay one video on top of another.
15443 This is the CUDA variant of the @ref{overlay} filter.
15444 It only accepts CUDA frames. The underlying input pixel formats have to match.
15446 It takes two inputs and has one output. The first input is the "main"
15447 video on which the second input is overlaid.
15449 It accepts the following parameters:
15454 Set the x and y coordinates of the overlaid video on the main video.
15455 Default value is "0" for both expressions.
15458 See @ref{framesync}.
15461 See @ref{framesync}.
15464 See @ref{framesync}.
15468 This filter also supports the @ref{framesync} options.
15472 Apply Overcomplete Wavelet denoiser.
15474 The filter accepts the following options:
15480 Larger depth values will denoise lower frequency components more, but
15481 slow down filtering.
15483 Must be an int in the range 8-16, default is @code{8}.
15485 @item luma_strength, ls
15488 Must be a double value in the range 0-1000, default is @code{1.0}.
15490 @item chroma_strength, cs
15491 Set chroma strength.
15493 Must be a double value in the range 0-1000, default is @code{1.0}.
15499 Add paddings to the input image, and place the original input at the
15500 provided @var{x}, @var{y} coordinates.
15502 It accepts the following parameters:
15507 Specify an expression for the size of the output image with the
15508 paddings added. If the value for @var{width} or @var{height} is 0, the
15509 corresponding input size is used for the output.
15511 The @var{width} expression can reference the value set by the
15512 @var{height} expression, and vice versa.
15514 The default value of @var{width} and @var{height} is 0.
15518 Specify the offsets to place the input image at within the padded area,
15519 with respect to the top/left border of the output image.
15521 The @var{x} expression can reference the value set by the @var{y}
15522 expression, and vice versa.
15524 The default value of @var{x} and @var{y} is 0.
15526 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15527 so the input image is centered on the padded area.
15530 Specify the color of the padded area. For the syntax of this option,
15531 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15532 manual,ffmpeg-utils}.
15534 The default value of @var{color} is "black".
15537 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15539 It accepts the following values:
15543 Only evaluate expressions once during the filter initialization or when
15544 a command is processed.
15547 Evaluate expressions for each incoming frame.
15551 Default value is @samp{init}.
15554 Pad to aspect instead to a resolution.
15558 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15559 options are expressions containing the following constants:
15564 The input video width and height.
15568 These are the same as @var{in_w} and @var{in_h}.
15572 The output width and height (the size of the padded area), as
15573 specified by the @var{width} and @var{height} expressions.
15577 These are the same as @var{out_w} and @var{out_h}.
15581 The x and y offsets as specified by the @var{x} and @var{y}
15582 expressions, or NAN if not yet specified.
15585 same as @var{iw} / @var{ih}
15588 input sample aspect ratio
15591 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15595 The horizontal and vertical chroma subsample values. For example for the
15596 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15599 @subsection Examples
15603 Add paddings with the color "violet" to the input video. The output video
15604 size is 640x480, and the top-left corner of the input video is placed at
15607 pad=640:480:0:40:violet
15610 The example above is equivalent to the following command:
15612 pad=width=640:height=480:x=0:y=40:color=violet
15616 Pad the input to get an output with dimensions increased by 3/2,
15617 and put the input video at the center of the padded area:
15619 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15623 Pad the input to get a squared output with size equal to the maximum
15624 value between the input width and height, and put the input video at
15625 the center of the padded area:
15627 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15631 Pad the input to get a final w/h ratio of 16:9:
15633 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15637 In case of anamorphic video, in order to set the output display aspect
15638 correctly, it is necessary to use @var{sar} in the expression,
15639 according to the relation:
15641 (ih * X / ih) * sar = output_dar
15642 X = output_dar / sar
15645 Thus the previous example needs to be modified to:
15647 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15651 Double the output size and put the input video in the bottom-right
15652 corner of the output padded area:
15654 pad="2*iw:2*ih:ow-iw:oh-ih"
15658 @anchor{palettegen}
15659 @section palettegen
15661 Generate one palette for a whole video stream.
15663 It accepts the following options:
15667 Set the maximum number of colors to quantize in the palette.
15668 Note: the palette will still contain 256 colors; the unused palette entries
15671 @item reserve_transparent
15672 Create a palette of 255 colors maximum and reserve the last one for
15673 transparency. Reserving the transparency color is useful for GIF optimization.
15674 If not set, the maximum of colors in the palette will be 256. You probably want
15675 to disable this option for a standalone image.
15678 @item transparency_color
15679 Set the color that will be used as background for transparency.
15682 Set statistics mode.
15684 It accepts the following values:
15687 Compute full frame histograms.
15689 Compute histograms only for the part that differs from previous frame. This
15690 might be relevant to give more importance to the moving part of your input if
15691 the background is static.
15693 Compute new histogram for each frame.
15696 Default value is @var{full}.
15699 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15700 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15701 color quantization of the palette. This information is also visible at
15702 @var{info} logging level.
15704 @subsection Examples
15708 Generate a representative palette of a given video using @command{ffmpeg}:
15710 ffmpeg -i input.mkv -vf palettegen palette.png
15714 @section paletteuse
15716 Use a palette to downsample an input video stream.
15718 The filter takes two inputs: one video stream and a palette. The palette must
15719 be a 256 pixels image.
15721 It accepts the following options:
15725 Select dithering mode. Available algorithms are:
15728 Ordered 8x8 bayer dithering (deterministic)
15730 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
15731 Note: this dithering is sometimes considered "wrong" and is included as a
15733 @item floyd_steinberg
15734 Floyd and Steingberg dithering (error diffusion)
15736 Frankie Sierra dithering v2 (error diffusion)
15738 Frankie Sierra dithering v2 "Lite" (error diffusion)
15741 Default is @var{sierra2_4a}.
15744 When @var{bayer} dithering is selected, this option defines the scale of the
15745 pattern (how much the crosshatch pattern is visible). A low value means more
15746 visible pattern for less banding, and higher value means less visible pattern
15747 at the cost of more banding.
15749 The option must be an integer value in the range [0,5]. Default is @var{2}.
15752 If set, define the zone to process
15756 Only the changing rectangle will be reprocessed. This is similar to GIF
15757 cropping/offsetting compression mechanism. This option can be useful for speed
15758 if only a part of the image is changing, and has use cases such as limiting the
15759 scope of the error diffusal @option{dither} to the rectangle that bounds the
15760 moving scene (it leads to more deterministic output if the scene doesn't change
15761 much, and as a result less moving noise and better GIF compression).
15764 Default is @var{none}.
15767 Take new palette for each output frame.
15769 @item alpha_threshold
15770 Sets the alpha threshold for transparency. Alpha values above this threshold
15771 will be treated as completely opaque, and values below this threshold will be
15772 treated as completely transparent.
15774 The option must be an integer value in the range [0,255]. Default is @var{128}.
15777 @subsection Examples
15781 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
15782 using @command{ffmpeg}:
15784 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
15788 @section perspective
15790 Correct perspective of video not recorded perpendicular to the screen.
15792 A description of the accepted parameters follows.
15803 Set coordinates expression for top left, top right, bottom left and bottom right corners.
15804 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
15805 If the @code{sense} option is set to @code{source}, then the specified points will be sent
15806 to the corners of the destination. If the @code{sense} option is set to @code{destination},
15807 then the corners of the source will be sent to the specified coordinates.
15809 The expressions can use the following variables:
15814 the width and height of video frame.
15818 Output frame count.
15821 @item interpolation
15822 Set interpolation for perspective correction.
15824 It accepts the following values:
15830 Default value is @samp{linear}.
15833 Set interpretation of coordinate options.
15835 It accepts the following values:
15839 Send point in the source specified by the given coordinates to
15840 the corners of the destination.
15842 @item 1, destination
15844 Send the corners of the source to the point in the destination specified
15845 by the given coordinates.
15847 Default value is @samp{source}.
15851 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
15853 It accepts the following values:
15856 only evaluate expressions once during the filter initialization or
15857 when a command is processed
15860 evaluate expressions for each incoming frame
15863 Default value is @samp{init}.
15868 Delay interlaced video by one field time so that the field order changes.
15870 The intended use is to fix PAL movies that have been captured with the
15871 opposite field order to the film-to-video transfer.
15873 A description of the accepted parameters follows.
15879 It accepts the following values:
15882 Capture field order top-first, transfer bottom-first.
15883 Filter will delay the bottom field.
15886 Capture field order bottom-first, transfer top-first.
15887 Filter will delay the top field.
15890 Capture and transfer with the same field order. This mode only exists
15891 for the documentation of the other options to refer to, but if you
15892 actually select it, the filter will faithfully do nothing.
15895 Capture field order determined automatically by field flags, transfer
15897 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
15898 basis using field flags. If no field information is available,
15899 then this works just like @samp{u}.
15902 Capture unknown or varying, transfer opposite.
15903 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
15904 analyzing the images and selecting the alternative that produces best
15905 match between the fields.
15908 Capture top-first, transfer unknown or varying.
15909 Filter selects among @samp{t} and @samp{p} using image analysis.
15912 Capture bottom-first, transfer unknown or varying.
15913 Filter selects among @samp{b} and @samp{p} using image analysis.
15916 Capture determined by field flags, transfer unknown or varying.
15917 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
15918 image analysis. If no field information is available, then this works just
15919 like @samp{U}. This is the default mode.
15922 Both capture and transfer unknown or varying.
15923 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
15927 @subsection Commands
15929 This filter supports the all above options as @ref{commands}.
15931 @section photosensitivity
15932 Reduce various flashes in video, so to help users with epilepsy.
15934 It accepts the following options:
15937 Set how many frames to use when filtering. Default is 30.
15940 Set detection threshold factor. Default is 1.
15944 Set how many pixels to skip when sampling frames. Default is 1.
15945 Allowed range is from 1 to 1024.
15948 Leave frames unchanged. Default is disabled.
15951 @section pixdesctest
15953 Pixel format descriptor test filter, mainly useful for internal
15954 testing. The output video should be equal to the input video.
15958 format=monow, pixdesctest
15961 can be used to test the monowhite pixel format descriptor definition.
15965 Display sample values of color channels. Mainly useful for checking color
15966 and levels. Minimum supported resolution is 640x480.
15968 The filters accept the following options:
15972 Set scope X position, relative offset on X axis.
15975 Set scope Y position, relative offset on Y axis.
15984 Set window opacity. This window also holds statistics about pixel area.
15987 Set window X position, relative offset on X axis.
15990 Set window Y position, relative offset on Y axis.
15995 Enable the specified chain of postprocessing subfilters using libpostproc. This
15996 library should be automatically selected with a GPL build (@code{--enable-gpl}).
15997 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
15998 Each subfilter and some options have a short and a long name that can be used
15999 interchangeably, i.e. dr/dering are the same.
16001 The filters accept the following options:
16005 Set postprocessing subfilters string.
16008 All subfilters share common options to determine their scope:
16012 Honor the quality commands for this subfilter.
16015 Do chrominance filtering, too (default).
16018 Do luminance filtering only (no chrominance).
16021 Do chrominance filtering only (no luminance).
16024 These options can be appended after the subfilter name, separated by a '|'.
16026 Available subfilters are:
16029 @item hb/hdeblock[|difference[|flatness]]
16030 Horizontal deblocking filter
16033 Difference factor where higher values mean more deblocking (default: @code{32}).
16035 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16038 @item vb/vdeblock[|difference[|flatness]]
16039 Vertical deblocking filter
16042 Difference factor where higher values mean more deblocking (default: @code{32}).
16044 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16047 @item ha/hadeblock[|difference[|flatness]]
16048 Accurate horizontal deblocking filter
16051 Difference factor where higher values mean more deblocking (default: @code{32}).
16053 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16056 @item va/vadeblock[|difference[|flatness]]
16057 Accurate vertical deblocking filter
16060 Difference factor where higher values mean more deblocking (default: @code{32}).
16062 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16066 The horizontal and vertical deblocking filters share the difference and
16067 flatness values so you cannot set different horizontal and vertical
16071 @item h1/x1hdeblock
16072 Experimental horizontal deblocking filter
16074 @item v1/x1vdeblock
16075 Experimental vertical deblocking filter
16080 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16083 larger -> stronger filtering
16085 larger -> stronger filtering
16087 larger -> stronger filtering
16090 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16093 Stretch luminance to @code{0-255}.
16096 @item lb/linblenddeint
16097 Linear blend deinterlacing filter that deinterlaces the given block by
16098 filtering all lines with a @code{(1 2 1)} filter.
16100 @item li/linipoldeint
16101 Linear interpolating deinterlacing filter that deinterlaces the given block by
16102 linearly interpolating every second line.
16104 @item ci/cubicipoldeint
16105 Cubic interpolating deinterlacing filter deinterlaces the given block by
16106 cubically interpolating every second line.
16108 @item md/mediandeint
16109 Median deinterlacing filter that deinterlaces the given block by applying a
16110 median filter to every second line.
16112 @item fd/ffmpegdeint
16113 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16114 second line with a @code{(-1 4 2 4 -1)} filter.
16117 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16118 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16120 @item fq/forceQuant[|quantizer]
16121 Overrides the quantizer table from the input with the constant quantizer you
16129 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16132 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16135 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16138 @subsection Examples
16142 Apply horizontal and vertical deblocking, deringing and automatic
16143 brightness/contrast:
16149 Apply default filters without brightness/contrast correction:
16155 Apply default filters and temporal denoiser:
16157 pp=default/tmpnoise|1|2|3
16161 Apply deblocking on luminance only, and switch vertical deblocking on or off
16162 automatically depending on available CPU time:
16169 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16170 similar to spp = 6 with 7 point DCT, where only the center sample is
16173 The filter accepts the following options:
16177 Force a constant quantization parameter. It accepts an integer in range
16178 0 to 63. If not set, the filter will use the QP from the video stream
16182 Set thresholding mode. Available modes are:
16186 Set hard thresholding.
16188 Set soft thresholding (better de-ringing effect, but likely blurrier).
16190 Set medium thresholding (good results, default).
16194 @section premultiply
16195 Apply alpha premultiply effect to input video stream using first plane
16196 of second stream as alpha.
16198 Both streams must have same dimensions and same pixel format.
16200 The filter accepts the following option:
16204 Set which planes will be processed, unprocessed planes will be copied.
16205 By default value 0xf, all planes will be processed.
16208 Do not require 2nd input for processing, instead use alpha plane from input stream.
16212 Apply prewitt operator to input video stream.
16214 The filter accepts the following option:
16218 Set which planes will be processed, unprocessed planes will be copied.
16219 By default value 0xf, all planes will be processed.
16222 Set value which will be multiplied with filtered result.
16225 Set value which will be added to filtered result.
16228 @subsection Commands
16230 This filter supports the all above options as @ref{commands}.
16232 @section pseudocolor
16234 Alter frame colors in video with pseudocolors.
16236 This filter accepts the following options:
16240 set pixel first component expression
16243 set pixel second component expression
16246 set pixel third component expression
16249 set pixel fourth component expression, corresponds to the alpha component
16252 set component to use as base for altering colors
16255 Pick one of built-in LUTs. By default is set to none.
16271 Each of them specifies the expression to use for computing the lookup table for
16272 the corresponding pixel component values.
16274 The expressions can contain the following constants and functions:
16279 The input width and height.
16282 The input value for the pixel component.
16284 @item ymin, umin, vmin, amin
16285 The minimum allowed component value.
16287 @item ymax, umax, vmax, amax
16288 The maximum allowed component value.
16291 All expressions default to "val".
16293 @subsection Commands
16295 This filter supports the all above options as @ref{commands}.
16297 @subsection Examples
16301 Change too high luma values to gradient:
16303 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'"
16309 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16310 Ratio) between two input videos.
16312 This filter takes in input two input videos, the first input is
16313 considered the "main" source and is passed unchanged to the
16314 output. The second input is used as a "reference" video for computing
16317 Both video inputs must have the same resolution and pixel format for
16318 this filter to work correctly. Also it assumes that both inputs
16319 have the same number of frames, which are compared one by one.
16321 The obtained average PSNR is printed through the logging system.
16323 The filter stores the accumulated MSE (mean squared error) of each
16324 frame, and at the end of the processing it is averaged across all frames
16325 equally, and the following formula is applied to obtain the PSNR:
16328 PSNR = 10*log10(MAX^2/MSE)
16331 Where MAX is the average of the maximum values of each component of the
16334 The description of the accepted parameters follows.
16337 @item stats_file, f
16338 If specified the filter will use the named file to save the PSNR of
16339 each individual frame. When filename equals "-" the data is sent to
16342 @item stats_version
16343 Specifies which version of the stats file format to use. Details of
16344 each format are written below.
16345 Default value is 1.
16347 @item stats_add_max
16348 Determines whether the max value is output to the stats log.
16349 Default value is 0.
16350 Requires stats_version >= 2. If this is set and stats_version < 2,
16351 the filter will return an error.
16354 This filter also supports the @ref{framesync} options.
16356 The file printed if @var{stats_file} is selected, contains a sequence of
16357 key/value pairs of the form @var{key}:@var{value} for each compared
16360 If a @var{stats_version} greater than 1 is specified, a header line precedes
16361 the list of per-frame-pair stats, with key value pairs following the frame
16362 format with the following parameters:
16365 @item psnr_log_version
16366 The version of the log file format. Will match @var{stats_version}.
16369 A comma separated list of the per-frame-pair parameters included in
16373 A description of each shown per-frame-pair parameter follows:
16377 sequential number of the input frame, starting from 1
16380 Mean Square Error pixel-by-pixel average difference of the compared
16381 frames, averaged over all the image components.
16383 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16384 Mean Square Error pixel-by-pixel average difference of the compared
16385 frames for the component specified by the suffix.
16387 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16388 Peak Signal to Noise ratio of the compared frames for the component
16389 specified by the suffix.
16391 @item max_avg, max_y, max_u, max_v
16392 Maximum allowed value for each channel, and average over all
16396 @subsection Examples
16401 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16402 [main][ref] psnr="stats_file=stats.log" [out]
16405 On this example the input file being processed is compared with the
16406 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16407 is stored in @file{stats.log}.
16410 Another example with different containers:
16412 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 -
16419 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16420 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16423 The pullup filter is designed to take advantage of future context in making
16424 its decisions. This filter is stateless in the sense that it does not lock
16425 onto a pattern to follow, but it instead looks forward to the following
16426 fields in order to identify matches and rebuild progressive frames.
16428 To produce content with an even framerate, insert the fps filter after
16429 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16430 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16432 The filter accepts the following options:
16439 These options set the amount of "junk" to ignore at the left, right, top, and
16440 bottom of the image, respectively. Left and right are in units of 8 pixels,
16441 while top and bottom are in units of 2 lines.
16442 The default is 8 pixels on each side.
16445 Set the strict breaks. Setting this option to 1 will reduce the chances of
16446 filter generating an occasional mismatched frame, but it may also cause an
16447 excessive number of frames to be dropped during high motion sequences.
16448 Conversely, setting it to -1 will make filter match fields more easily.
16449 This may help processing of video where there is slight blurring between
16450 the fields, but may also cause there to be interlaced frames in the output.
16451 Default value is @code{0}.
16454 Set the metric plane to use. It accepts the following values:
16460 Use chroma blue plane.
16463 Use chroma red plane.
16466 This option may be set to use chroma plane instead of the default luma plane
16467 for doing filter's computations. This may improve accuracy on very clean
16468 source material, but more likely will decrease accuracy, especially if there
16469 is chroma noise (rainbow effect) or any grayscale video.
16470 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16471 load and make pullup usable in realtime on slow machines.
16474 For best results (without duplicated frames in the output file) it is
16475 necessary to change the output frame rate. For example, to inverse
16476 telecine NTSC input:
16478 ffmpeg -i input -vf pullup -r 24000/1001 ...
16483 Change video quantization parameters (QP).
16485 The filter accepts the following option:
16489 Set expression for quantization parameter.
16492 The expression is evaluated through the eval API and can contain, among others,
16493 the following constants:
16497 1 if index is not 129, 0 otherwise.
16500 Sequential index starting from -129 to 128.
16503 @subsection Examples
16507 Some equation like:
16515 Flush video frames from internal cache of frames into a random order.
16516 No frame is discarded.
16517 Inspired by @ref{frei0r} nervous filter.
16521 Set size in number of frames of internal cache, in range from @code{2} to
16522 @code{512}. Default is @code{30}.
16525 Set seed for random number generator, must be an integer included between
16526 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16527 less than @code{0}, the filter will try to use a good random seed on a
16531 @section readeia608
16533 Read closed captioning (EIA-608) information from the top lines of a video frame.
16535 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16536 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16537 with EIA-608 data (starting from 0). A description of each metadata value follows:
16540 @item lavfi.readeia608.X.cc
16541 The two bytes stored as EIA-608 data (printed in hexadecimal).
16543 @item lavfi.readeia608.X.line
16544 The number of the line on which the EIA-608 data was identified and read.
16547 This filter accepts the following options:
16551 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16554 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16557 Set the ratio of width reserved for sync code detection.
16558 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16561 Enable checking the parity bit. In the event of a parity error, the filter will output
16562 @code{0x00} for that character. Default is false.
16565 Lowpass lines prior to further processing. Default is enabled.
16568 @subsection Commands
16570 This filter supports the all above options as @ref{commands}.
16572 @subsection Examples
16576 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16578 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
16584 Read vertical interval timecode (VITC) information from the top lines of a
16587 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16588 timecode value, if a valid timecode has been detected. Further metadata key
16589 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16590 timecode data has been found or not.
16592 This filter accepts the following options:
16596 Set the maximum number of lines to scan for VITC data. If the value is set to
16597 @code{-1} the full video frame is scanned. Default is @code{45}.
16600 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16601 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16604 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16605 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16608 @subsection Examples
16612 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16613 draw @code{--:--:--:--} as a placeholder:
16615 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16621 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16623 Destination pixel at position (X, Y) will be picked from source (x, y) position
16624 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16625 value for pixel will be used for destination pixel.
16627 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16628 will have Xmap/Ymap video stream dimensions.
16629 Xmap and Ymap input video streams are 16bit depth, single channel.
16633 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16634 Default is @code{color}.
16637 Specify the color of the unmapped pixels. For the syntax of this option,
16638 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16639 manual,ffmpeg-utils}. Default color is @code{black}.
16642 @section removegrain
16644 The removegrain filter is a spatial denoiser for progressive video.
16648 Set mode for the first plane.
16651 Set mode for the second plane.
16654 Set mode for the third plane.
16657 Set mode for the fourth plane.
16660 Range of mode is from 0 to 24. Description of each mode follows:
16664 Leave input plane unchanged. Default.
16667 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16670 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16673 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16676 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16677 This is equivalent to a median filter.
16680 Line-sensitive clipping giving the minimal change.
16683 Line-sensitive clipping, intermediate.
16686 Line-sensitive clipping, intermediate.
16689 Line-sensitive clipping, intermediate.
16692 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16695 Replaces the target pixel with the closest neighbour.
16698 [1 2 1] horizontal and vertical kernel blur.
16704 Bob mode, interpolates top field from the line where the neighbours
16705 pixels are the closest.
16708 Bob mode, interpolates bottom field from the line where the neighbours
16709 pixels are the closest.
16712 Bob mode, interpolates top field. Same as 13 but with a more complicated
16713 interpolation formula.
16716 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16717 interpolation formula.
16720 Clips the pixel with the minimum and maximum of respectively the maximum and
16721 minimum of each pair of opposite neighbour pixels.
16724 Line-sensitive clipping using opposite neighbours whose greatest distance from
16725 the current pixel is minimal.
16728 Replaces the pixel with the average of its 8 neighbours.
16731 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
16734 Clips pixels using the averages of opposite neighbour.
16737 Same as mode 21 but simpler and faster.
16740 Small edge and halo removal, but reputed useless.
16746 @section removelogo
16748 Suppress a TV station logo, using an image file to determine which
16749 pixels comprise the logo. It works by filling in the pixels that
16750 comprise the logo with neighboring pixels.
16752 The filter accepts the following options:
16756 Set the filter bitmap file, which can be any image format supported by
16757 libavformat. The width and height of the image file must match those of the
16758 video stream being processed.
16761 Pixels in the provided bitmap image with a value of zero are not
16762 considered part of the logo, non-zero pixels are considered part of
16763 the logo. If you use white (255) for the logo and black (0) for the
16764 rest, you will be safe. For making the filter bitmap, it is
16765 recommended to take a screen capture of a black frame with the logo
16766 visible, and then using a threshold filter followed by the erode
16767 filter once or twice.
16769 If needed, little splotches can be fixed manually. Remember that if
16770 logo pixels are not covered, the filter quality will be much
16771 reduced. Marking too many pixels as part of the logo does not hurt as
16772 much, but it will increase the amount of blurring needed to cover over
16773 the image and will destroy more information than necessary, and extra
16774 pixels will slow things down on a large logo.
16776 @section repeatfields
16778 This filter uses the repeat_field flag from the Video ES headers and hard repeats
16779 fields based on its value.
16783 Reverse a video clip.
16785 Warning: This filter requires memory to buffer the entire clip, so trimming
16788 @subsection Examples
16792 Take the first 5 seconds of a clip, and reverse it.
16799 Shift R/G/B/A pixels horizontally and/or vertically.
16801 The filter accepts the following options:
16804 Set amount to shift red horizontally.
16806 Set amount to shift red vertically.
16808 Set amount to shift green horizontally.
16810 Set amount to shift green vertically.
16812 Set amount to shift blue horizontally.
16814 Set amount to shift blue vertically.
16816 Set amount to shift alpha horizontally.
16818 Set amount to shift alpha vertically.
16820 Set edge mode, can be @var{smear}, default, or @var{warp}.
16823 @subsection Commands
16825 This filter supports the all above options as @ref{commands}.
16828 Apply roberts cross operator to input video stream.
16830 The filter accepts the following option:
16834 Set which planes will be processed, unprocessed planes will be copied.
16835 By default value 0xf, all planes will be processed.
16838 Set value which will be multiplied with filtered result.
16841 Set value which will be added to filtered result.
16844 @subsection Commands
16846 This filter supports the all above options as @ref{commands}.
16850 Rotate video by an arbitrary angle expressed in radians.
16852 The filter accepts the following options:
16854 A description of the optional parameters follows.
16857 Set an expression for the angle by which to rotate the input video
16858 clockwise, expressed as a number of radians. A negative value will
16859 result in a counter-clockwise rotation. By default it is set to "0".
16861 This expression is evaluated for each frame.
16864 Set the output width expression, default value is "iw".
16865 This expression is evaluated just once during configuration.
16868 Set the output height expression, default value is "ih".
16869 This expression is evaluated just once during configuration.
16872 Enable bilinear interpolation if set to 1, a value of 0 disables
16873 it. Default value is 1.
16876 Set the color used to fill the output area not covered by the rotated
16877 image. For the general syntax of this option, check the
16878 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16879 If the special value "none" is selected then no
16880 background is printed (useful for example if the background is never shown).
16882 Default value is "black".
16885 The expressions for the angle and the output size can contain the
16886 following constants and functions:
16890 sequential number of the input frame, starting from 0. It is always NAN
16891 before the first frame is filtered.
16894 time in seconds of the input frame, it is set to 0 when the filter is
16895 configured. It is always NAN before the first frame is filtered.
16899 horizontal and vertical chroma subsample values. For example for the
16900 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16904 the input video width and height
16908 the output width and height, that is the size of the padded area as
16909 specified by the @var{width} and @var{height} expressions
16913 the minimal width/height required for completely containing the input
16914 video rotated by @var{a} radians.
16916 These are only available when computing the @option{out_w} and
16917 @option{out_h} expressions.
16920 @subsection Examples
16924 Rotate the input by PI/6 radians clockwise:
16930 Rotate the input by PI/6 radians counter-clockwise:
16936 Rotate the input by 45 degrees clockwise:
16942 Apply a constant rotation with period T, starting from an angle of PI/3:
16944 rotate=PI/3+2*PI*t/T
16948 Make the input video rotation oscillating with a period of T
16949 seconds and an amplitude of A radians:
16951 rotate=A*sin(2*PI/T*t)
16955 Rotate the video, output size is chosen so that the whole rotating
16956 input video is always completely contained in the output:
16958 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
16962 Rotate the video, reduce the output size so that no background is ever
16965 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
16969 @subsection Commands
16971 The filter supports the following commands:
16975 Set the angle expression.
16976 The command accepts the same syntax of the corresponding option.
16978 If the specified expression is not valid, it is kept at its current
16984 Apply Shape Adaptive Blur.
16986 The filter accepts the following options:
16989 @item luma_radius, lr
16990 Set luma blur filter strength, must be a value in range 0.1-4.0, default
16991 value is 1.0. A greater value will result in a more blurred image, and
16992 in slower processing.
16994 @item luma_pre_filter_radius, lpfr
16995 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
16998 @item luma_strength, ls
16999 Set luma maximum difference between pixels to still be considered, must
17000 be a value in the 0.1-100.0 range, default value is 1.0.
17002 @item chroma_radius, cr
17003 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17004 greater value will result in a more blurred image, and in slower
17007 @item chroma_pre_filter_radius, cpfr
17008 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17010 @item chroma_strength, cs
17011 Set chroma maximum difference between pixels to still be considered,
17012 must be a value in the -0.9-100.0 range.
17015 Each chroma option value, if not explicitly specified, is set to the
17016 corresponding luma option value.
17021 Scale (resize) the input video, using the libswscale library.
17023 The scale filter forces the output display aspect ratio to be the same
17024 of the input, by changing the output sample aspect ratio.
17026 If the input image format is different from the format requested by
17027 the next filter, the scale filter will convert the input to the
17030 @subsection Options
17031 The filter accepts the following options, or any of the options
17032 supported by the libswscale scaler.
17034 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17035 the complete list of scaler options.
17040 Set the output video dimension expression. Default value is the input
17043 If the @var{width} or @var{w} value is 0, the input width is used for
17044 the output. If the @var{height} or @var{h} value is 0, the input height
17045 is used for the output.
17047 If one and only one of the values is -n with n >= 1, the scale filter
17048 will use a value that maintains the aspect ratio of the input image,
17049 calculated from the other specified dimension. After that it will,
17050 however, make sure that the calculated dimension is divisible by n and
17051 adjust the value if necessary.
17053 If both values are -n with n >= 1, the behavior will be identical to
17054 both values being set to 0 as previously detailed.
17056 See below for the list of accepted constants for use in the dimension
17060 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17064 Only evaluate expressions once during the filter initialization or when a command is processed.
17067 Evaluate expressions for each incoming frame.
17071 Default value is @samp{init}.
17075 Set the interlacing mode. It accepts the following values:
17079 Force interlaced aware scaling.
17082 Do not apply interlaced scaling.
17085 Select interlaced aware scaling depending on whether the source frames
17086 are flagged as interlaced or not.
17089 Default value is @samp{0}.
17092 Set libswscale scaling flags. See
17093 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17094 complete list of values. If not explicitly specified the filter applies
17098 @item param0, param1
17099 Set libswscale input parameters for scaling algorithms that need them. See
17100 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17101 complete documentation. If not explicitly specified the filter applies
17107 Set the video size. For the syntax of this option, check the
17108 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17110 @item in_color_matrix
17111 @item out_color_matrix
17112 Set in/output YCbCr color space type.
17114 This allows the autodetected value to be overridden as well as allows forcing
17115 a specific value used for the output and encoder.
17117 If not specified, the color space type depends on the pixel format.
17123 Choose automatically.
17126 Format conforming to International Telecommunication Union (ITU)
17127 Recommendation BT.709.
17130 Set color space conforming to the United States Federal Communications
17131 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17136 Set color space conforming to:
17140 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17143 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17146 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17151 Set color space conforming to SMPTE ST 240:1999.
17154 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17159 Set in/output YCbCr sample range.
17161 This allows the autodetected value to be overridden as well as allows forcing
17162 a specific value used for the output and encoder. If not specified, the
17163 range depends on the pixel format. Possible values:
17167 Choose automatically.
17170 Set full range (0-255 in case of 8-bit luma).
17172 @item mpeg/limited/tv
17173 Set "MPEG" range (16-235 in case of 8-bit luma).
17176 @item force_original_aspect_ratio
17177 Enable decreasing or increasing output video width or height if necessary to
17178 keep the original aspect ratio. Possible values:
17182 Scale the video as specified and disable this feature.
17185 The output video dimensions will automatically be decreased if needed.
17188 The output video dimensions will automatically be increased if needed.
17192 One useful instance of this option is that when you know a specific device's
17193 maximum allowed resolution, you can use this to limit the output video to
17194 that, while retaining the aspect ratio. For example, device A allows
17195 1280x720 playback, and your video is 1920x800. Using this option (set it to
17196 decrease) and specifying 1280x720 to the command line makes the output
17199 Please note that this is a different thing than specifying -1 for @option{w}
17200 or @option{h}, you still need to specify the output resolution for this option
17203 @item force_divisible_by
17204 Ensures that both the output dimensions, width and height, are divisible by the
17205 given integer when used together with @option{force_original_aspect_ratio}. This
17206 works similar to using @code{-n} in the @option{w} and @option{h} options.
17208 This option respects the value set for @option{force_original_aspect_ratio},
17209 increasing or decreasing the resolution accordingly. The video's aspect ratio
17210 may be slightly modified.
17212 This option can be handy if you need to have a video fit within or exceed
17213 a defined resolution using @option{force_original_aspect_ratio} but also have
17214 encoder restrictions on width or height divisibility.
17218 The values of the @option{w} and @option{h} options are expressions
17219 containing the following constants:
17224 The input width and height
17228 These are the same as @var{in_w} and @var{in_h}.
17232 The output (scaled) width and height
17236 These are the same as @var{out_w} and @var{out_h}
17239 The same as @var{iw} / @var{ih}
17242 input sample aspect ratio
17245 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17249 horizontal and vertical input chroma subsample values. For example for the
17250 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17254 horizontal and vertical output chroma subsample values. For example for the
17255 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17258 The (sequential) number of the input frame, starting from 0.
17259 Only available with @code{eval=frame}.
17262 The presentation timestamp of the input frame, expressed as a number of
17263 seconds. Only available with @code{eval=frame}.
17266 The position (byte offset) of the frame in the input stream, or NaN if
17267 this information is unavailable and/or meaningless (for example in case of synthetic video).
17268 Only available with @code{eval=frame}.
17271 @subsection Examples
17275 Scale the input video to a size of 200x100
17280 This is equivalent to:
17291 Specify a size abbreviation for the output size:
17296 which can also be written as:
17302 Scale the input to 2x:
17304 scale=w=2*iw:h=2*ih
17308 The above is the same as:
17310 scale=2*in_w:2*in_h
17314 Scale the input to 2x with forced interlaced scaling:
17316 scale=2*iw:2*ih:interl=1
17320 Scale the input to half size:
17322 scale=w=iw/2:h=ih/2
17326 Increase the width, and set the height to the same size:
17332 Seek Greek harmony:
17339 Increase the height, and set the width to 3/2 of the height:
17341 scale=w=3/2*oh:h=3/5*ih
17345 Increase the size, making the size a multiple of the chroma
17348 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17352 Increase the width to a maximum of 500 pixels,
17353 keeping the same aspect ratio as the input:
17355 scale=w='min(500\, iw*3/2):h=-1'
17359 Make pixels square by combining scale and setsar:
17361 scale='trunc(ih*dar):ih',setsar=1/1
17365 Make pixels square by combining scale and setsar,
17366 making sure the resulting resolution is even (required by some codecs):
17368 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17372 @subsection Commands
17374 This filter supports the following commands:
17378 Set the output video dimension expression.
17379 The command accepts the same syntax of the corresponding option.
17381 If the specified expression is not valid, it is kept at its current
17387 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17388 format conversion on CUDA video frames. Setting the output width and height
17389 works in the same way as for the @var{scale} filter.
17391 The following additional options are accepted:
17394 The pixel format of the output CUDA frames. If set to the string "same" (the
17395 default), the input format will be kept. Note that automatic format negotiation
17396 and conversion is not yet supported for hardware frames
17399 The interpolation algorithm used for resizing. One of the following:
17406 @item cubic2p_bspline
17407 2-parameter cubic (B=1, C=0)
17409 @item cubic2p_catmullrom
17410 2-parameter cubic (B=0, C=1/2)
17412 @item cubic2p_b05c03
17413 2-parameter cubic (B=1/2, C=3/10)
17421 @item force_original_aspect_ratio
17422 Enable decreasing or increasing output video width or height if necessary to
17423 keep the original aspect ratio. Possible values:
17427 Scale the video as specified and disable this feature.
17430 The output video dimensions will automatically be decreased if needed.
17433 The output video dimensions will automatically be increased if needed.
17437 One useful instance of this option is that when you know a specific device's
17438 maximum allowed resolution, you can use this to limit the output video to
17439 that, while retaining the aspect ratio. For example, device A allows
17440 1280x720 playback, and your video is 1920x800. Using this option (set it to
17441 decrease) and specifying 1280x720 to the command line makes the output
17444 Please note that this is a different thing than specifying -1 for @option{w}
17445 or @option{h}, you still need to specify the output resolution for this option
17448 @item force_divisible_by
17449 Ensures that both the output dimensions, width and height, are divisible by the
17450 given integer when used together with @option{force_original_aspect_ratio}. This
17451 works similar to using @code{-n} in the @option{w} and @option{h} options.
17453 This option respects the value set for @option{force_original_aspect_ratio},
17454 increasing or decreasing the resolution accordingly. The video's aspect ratio
17455 may be slightly modified.
17457 This option can be handy if you need to have a video fit within or exceed
17458 a defined resolution using @option{force_original_aspect_ratio} but also have
17459 encoder restrictions on width or height divisibility.
17465 Scale (resize) the input video, based on a reference video.
17467 See the scale filter for available options, scale2ref supports the same but
17468 uses the reference video instead of the main input as basis. scale2ref also
17469 supports the following additional constants for the @option{w} and
17470 @option{h} options:
17475 The main input video's width and height
17478 The same as @var{main_w} / @var{main_h}
17481 The main input video's sample aspect ratio
17483 @item main_dar, mdar
17484 The main input video's display aspect ratio. Calculated from
17485 @code{(main_w / main_h) * main_sar}.
17489 The main input video's horizontal and vertical chroma subsample values.
17490 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17494 The (sequential) number of the main input frame, starting from 0.
17495 Only available with @code{eval=frame}.
17498 The presentation timestamp of the main input frame, expressed as a number of
17499 seconds. Only available with @code{eval=frame}.
17502 The position (byte offset) of the frame in the main input stream, or NaN if
17503 this information is unavailable and/or meaningless (for example in case of synthetic video).
17504 Only available with @code{eval=frame}.
17507 @subsection Examples
17511 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17513 'scale2ref[b][a];[a][b]overlay'
17517 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17519 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17523 @subsection Commands
17525 This filter supports the following commands:
17529 Set the output video dimension expression.
17530 The command accepts the same syntax of the corresponding option.
17532 If the specified expression is not valid, it is kept at its current
17537 Scroll input video horizontally and/or vertically by constant speed.
17539 The filter accepts the following options:
17541 @item horizontal, h
17542 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17543 Negative values changes scrolling direction.
17546 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17547 Negative values changes scrolling direction.
17550 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17553 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17556 @subsection Commands
17558 This filter supports the following @ref{commands}:
17560 @item horizontal, h
17561 Set the horizontal scrolling speed.
17563 Set the vertical scrolling speed.
17569 Detect video scene change.
17571 This filter sets frame metadata with mafd between frame, the scene score, and
17572 forward the frame to the next filter, so they can use these metadata to detect
17573 scene change or others.
17575 In addition, this filter logs a message and sets frame metadata when it detects
17576 a scene change by @option{threshold}.
17578 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17580 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17581 to detect scene change.
17583 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17584 detect scene change with @option{threshold}.
17586 The filter accepts the following options:
17590 Set the scene change detection threshold as a percentage of maximum change. Good
17591 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17594 Default value is @code{10.}.
17597 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17598 You can enable it if you want to get snapshot of scene change frames only.
17601 @anchor{selectivecolor}
17602 @section selectivecolor
17604 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17605 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17606 by the "purity" of the color (that is, how saturated it already is).
17608 This filter is similar to the Adobe Photoshop Selective Color tool.
17610 The filter accepts the following options:
17613 @item correction_method
17614 Select color correction method.
17616 Available values are:
17619 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17622 Specified adjustments are relative to the original component value.
17624 Default is @code{absolute}.
17626 Adjustments for red pixels (pixels where the red component is the maximum)
17628 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17630 Adjustments for green pixels (pixels where the green component is the maximum)
17632 Adjustments for cyan pixels (pixels where the red component is the minimum)
17634 Adjustments for blue pixels (pixels where the blue component is the maximum)
17636 Adjustments for magenta pixels (pixels where the green component is the minimum)
17638 Adjustments for white pixels (pixels where all components are greater than 128)
17640 Adjustments for all pixels except pure black and pure white
17642 Adjustments for black pixels (pixels where all components are lesser than 128)
17644 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17647 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17648 4 space separated floating point adjustment values in the [-1,1] range,
17649 respectively to adjust the amount of cyan, magenta, yellow and black for the
17650 pixels of its range.
17652 @subsection Examples
17656 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17657 increase magenta by 27% in blue areas:
17659 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17663 Use a Photoshop selective color preset:
17665 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17669 @anchor{separatefields}
17670 @section separatefields
17672 The @code{separatefields} takes a frame-based video input and splits
17673 each frame into its components fields, producing a new half height clip
17674 with twice the frame rate and twice the frame count.
17676 This filter use field-dominance information in frame to decide which
17677 of each pair of fields to place first in the output.
17678 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17680 @section setdar, setsar
17682 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17685 This is done by changing the specified Sample (aka Pixel) Aspect
17686 Ratio, according to the following equation:
17688 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17691 Keep in mind that the @code{setdar} filter does not modify the pixel
17692 dimensions of the video frame. Also, the display aspect ratio set by
17693 this filter may be changed by later filters in the filterchain,
17694 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17697 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17698 the filter output video.
17700 Note that as a consequence of the application of this filter, the
17701 output display aspect ratio will change according to the equation
17704 Keep in mind that the sample aspect ratio set by the @code{setsar}
17705 filter may be changed by later filters in the filterchain, e.g. if
17706 another "setsar" or a "setdar" filter is applied.
17708 It accepts the following parameters:
17711 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17712 Set the aspect ratio used by the filter.
17714 The parameter can be a floating point number string, an expression, or
17715 a string of the form @var{num}:@var{den}, where @var{num} and
17716 @var{den} are the numerator and denominator of the aspect ratio. If
17717 the parameter is not specified, it is assumed the value "0".
17718 In case the form "@var{num}:@var{den}" is used, the @code{:} character
17722 Set the maximum integer value to use for expressing numerator and
17723 denominator when reducing the expressed aspect ratio to a rational.
17724 Default value is @code{100}.
17728 The parameter @var{sar} is an expression containing
17729 the following constants:
17733 These are approximated values for the mathematical constants e
17734 (Euler's number), pi (Greek pi), and phi (the golden ratio).
17737 The input width and height.
17740 These are the same as @var{w} / @var{h}.
17743 The input sample aspect ratio.
17746 The input display aspect ratio. It is the same as
17747 (@var{w} / @var{h}) * @var{sar}.
17750 Horizontal and vertical chroma subsample values. For example, for the
17751 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17754 @subsection Examples
17759 To change the display aspect ratio to 16:9, specify one of the following:
17766 To change the sample aspect ratio to 10:11, specify:
17772 To set a display aspect ratio of 16:9, and specify a maximum integer value of
17773 1000 in the aspect ratio reduction, use the command:
17775 setdar=ratio=16/9:max=1000
17783 Force field for the output video frame.
17785 The @code{setfield} filter marks the interlace type field for the
17786 output frames. It does not change the input frame, but only sets the
17787 corresponding property, which affects how the frame is treated by
17788 following filters (e.g. @code{fieldorder} or @code{yadif}).
17790 The filter accepts the following options:
17795 Available values are:
17799 Keep the same field property.
17802 Mark the frame as bottom-field-first.
17805 Mark the frame as top-field-first.
17808 Mark the frame as progressive.
17815 Force frame parameter for the output video frame.
17817 The @code{setparams} filter marks interlace and color range for the
17818 output frames. It does not change the input frame, but only sets the
17819 corresponding property, which affects how the frame is treated by
17824 Available values are:
17828 Keep the same field property (default).
17831 Mark the frame as bottom-field-first.
17834 Mark the frame as top-field-first.
17837 Mark the frame as progressive.
17841 Available values are:
17845 Keep the same color range property (default).
17847 @item unspecified, unknown
17848 Mark the frame as unspecified color range.
17850 @item limited, tv, mpeg
17851 Mark the frame as limited range.
17853 @item full, pc, jpeg
17854 Mark the frame as full range.
17857 @item color_primaries
17858 Set the color primaries.
17859 Available values are:
17863 Keep the same color primaries property (default).
17880 Set the color transfer.
17881 Available values are:
17885 Keep the same color trc property (default).
17907 Set the colorspace.
17908 Available values are:
17912 Keep the same colorspace property (default).
17925 @item chroma-derived-nc
17926 @item chroma-derived-c
17932 Apply shear transform to input video.
17934 This filter supports the following options:
17938 Shear factor in X-direction. Default value is 0.
17939 Allowed range is from -2 to 2.
17942 Shear factor in Y-direction. Default value is 0.
17943 Allowed range is from -2 to 2.
17946 Set the color used to fill the output area not covered by the transformed
17947 video. For the general syntax of this option, check the
17948 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17949 If the special value "none" is selected then no
17950 background is printed (useful for example if the background is never shown).
17952 Default value is "black".
17955 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
17958 @subsection Commands
17960 This filter supports the all above options as @ref{commands}.
17964 Show a line containing various information for each input video frame.
17965 The input video is not modified.
17967 This filter supports the following options:
17971 Calculate checksums of each plane. By default enabled.
17974 The shown line contains a sequence of key/value pairs of the form
17975 @var{key}:@var{value}.
17977 The following values are shown in the output:
17981 The (sequential) number of the input frame, starting from 0.
17984 The Presentation TimeStamp of the input frame, expressed as a number of
17985 time base units. The time base unit depends on the filter input pad.
17988 The Presentation TimeStamp of the input frame, expressed as a number of
17992 The position of the frame in the input stream, or -1 if this information is
17993 unavailable and/or meaningless (for example in case of synthetic video).
17996 The pixel format name.
17999 The sample aspect ratio of the input frame, expressed in the form
18000 @var{num}/@var{den}.
18003 The size of the input frame. For the syntax of this option, check the
18004 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18007 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18008 for bottom field first).
18011 This is 1 if the frame is a key frame, 0 otherwise.
18014 The picture type of the input frame ("I" for an I-frame, "P" for a
18015 P-frame, "B" for a B-frame, or "?" for an unknown type).
18016 Also refer to the documentation of the @code{AVPictureType} enum and of
18017 the @code{av_get_picture_type_char} function defined in
18018 @file{libavutil/avutil.h}.
18021 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18023 @item plane_checksum
18024 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18025 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18028 The mean value of pixels in each plane of the input frame, expressed in the form
18029 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18032 The standard deviation of pixel values in each plane of the input frame, expressed
18033 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18037 @section showpalette
18039 Displays the 256 colors palette of each frame. This filter is only relevant for
18040 @var{pal8} pixel format frames.
18042 It accepts the following option:
18046 Set the size of the box used to represent one palette color entry. Default is
18047 @code{30} (for a @code{30x30} pixel box).
18050 @section shuffleframes
18052 Reorder and/or duplicate and/or drop video frames.
18054 It accepts the following parameters:
18058 Set the destination indexes of input frames.
18059 This is space or '|' separated list of indexes that maps input frames to output
18060 frames. Number of indexes also sets maximal value that each index may have.
18061 '-1' index have special meaning and that is to drop frame.
18064 The first frame has the index 0. The default is to keep the input unchanged.
18066 @subsection Examples
18070 Swap second and third frame of every three frames of the input:
18072 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18076 Swap 10th and 1st frame of every ten frames of the input:
18078 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18082 @section shufflepixels
18084 Reorder pixels in video frames.
18086 This filter accepts the following options:
18090 Set shuffle direction. Can be forward or inverse direction.
18091 Default direction is forward.
18094 Set shuffle mode. Can be horizontal, vertical or block mode.
18098 Set shuffle block_size. In case of horizontal shuffle mode only width
18099 part of size is used, and in case of vertical shuffle mode only height
18100 part of size is used.
18103 Set random seed used with shuffling pixels. Mainly useful to set to be able
18104 to reverse filtering process to get original input.
18105 For example, to reverse forward shuffle you need to use same parameters
18106 and exact same seed and to set direction to inverse.
18109 @section shuffleplanes
18111 Reorder and/or duplicate video planes.
18113 It accepts the following parameters:
18118 The index of the input plane to be used as the first output plane.
18121 The index of the input plane to be used as the second output plane.
18124 The index of the input plane to be used as the third output plane.
18127 The index of the input plane to be used as the fourth output plane.
18131 The first plane has the index 0. The default is to keep the input unchanged.
18133 @subsection Examples
18137 Swap the second and third planes of the input:
18139 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18143 @anchor{signalstats}
18144 @section signalstats
18145 Evaluate various visual metrics that assist in determining issues associated
18146 with the digitization of analog video media.
18148 By default the filter will log these metadata values:
18152 Display the minimal Y value contained within the input frame. Expressed in
18156 Display the Y value at the 10% percentile within the input frame. Expressed in
18160 Display the average Y value within the input frame. Expressed in range of
18164 Display the Y value at the 90% percentile within the input frame. Expressed in
18168 Display the maximum Y value contained within the input frame. Expressed in
18172 Display the minimal U value contained within the input frame. Expressed in
18176 Display the U value at the 10% percentile within the input frame. Expressed in
18180 Display the average U value within the input frame. Expressed in range of
18184 Display the U value at the 90% percentile within the input frame. Expressed in
18188 Display the maximum U value contained within the input frame. Expressed in
18192 Display the minimal V value contained within the input frame. Expressed in
18196 Display the V value at the 10% percentile within the input frame. Expressed in
18200 Display the average V value within the input frame. Expressed in range of
18204 Display the V value at the 90% percentile within the input frame. Expressed in
18208 Display the maximum V value contained within the input frame. Expressed in
18212 Display the minimal saturation value contained within the input frame.
18213 Expressed in range of [0-~181.02].
18216 Display the saturation value at the 10% percentile within the input frame.
18217 Expressed in range of [0-~181.02].
18220 Display the average saturation value within the input frame. Expressed in range
18224 Display the saturation value at the 90% percentile within the input frame.
18225 Expressed in range of [0-~181.02].
18228 Display the maximum saturation value contained within the input frame.
18229 Expressed in range of [0-~181.02].
18232 Display the median value for hue within the input frame. Expressed in range of
18236 Display the average value for hue within the input frame. Expressed in range of
18240 Display the average of sample value difference between all values of the Y
18241 plane in the current frame and corresponding values of the previous input frame.
18242 Expressed in range of [0-255].
18245 Display the average of sample value difference between all values of the U
18246 plane in the current frame and corresponding values of the previous input frame.
18247 Expressed in range of [0-255].
18250 Display the average of sample value difference between all values of the V
18251 plane in the current frame and corresponding values of the previous input frame.
18252 Expressed in range of [0-255].
18255 Display bit depth of Y plane in current frame.
18256 Expressed in range of [0-16].
18259 Display bit depth of U plane in current frame.
18260 Expressed in range of [0-16].
18263 Display bit depth of V plane in current frame.
18264 Expressed in range of [0-16].
18267 The filter accepts the following options:
18273 @option{stat} specify an additional form of image analysis.
18274 @option{out} output video with the specified type of pixel highlighted.
18276 Both options accept the following values:
18280 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18281 unlike the neighboring pixels of the same field. Examples of temporal outliers
18282 include the results of video dropouts, head clogs, or tape tracking issues.
18285 Identify @var{vertical line repetition}. Vertical line repetition includes
18286 similar rows of pixels within a frame. In born-digital video vertical line
18287 repetition is common, but this pattern is uncommon in video digitized from an
18288 analog source. When it occurs in video that results from the digitization of an
18289 analog source it can indicate concealment from a dropout compensator.
18292 Identify pixels that fall outside of legal broadcast range.
18296 Set the highlight color for the @option{out} option. The default color is
18300 @subsection Examples
18304 Output data of various video metrics:
18306 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18310 Output specific data about the minimum and maximum values of the Y plane per frame:
18312 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18316 Playback video while highlighting pixels that are outside of broadcast range in red.
18318 ffplay example.mov -vf signalstats="out=brng:color=red"
18322 Playback video with signalstats metadata drawn over the frame.
18324 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18327 The contents of signalstat_drawtext.txt used in the command are:
18330 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18331 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18332 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18333 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18341 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18342 input. In this case the matching between the inputs can be calculated additionally.
18343 The filter always passes through the first input. The signature of each stream can
18344 be written into a file.
18346 It accepts the following options:
18350 Enable or disable the matching process.
18352 Available values are:
18356 Disable the calculation of a matching (default).
18358 Calculate the matching for the whole video and output whether the whole video
18359 matches or only parts.
18361 Calculate only until a matching is found or the video ends. Should be faster in
18366 Set the number of inputs. The option value must be a non negative integer.
18367 Default value is 1.
18370 Set the path to which the output is written. If there is more than one input,
18371 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18372 integer), that will be replaced with the input number. If no filename is
18373 specified, no output will be written. This is the default.
18376 Choose the output format.
18378 Available values are:
18382 Use the specified binary representation (default).
18384 Use the specified xml representation.
18388 Set threshold to detect one word as similar. The option value must be an integer
18389 greater than zero. The default value is 9000.
18392 Set threshold to detect all words as similar. The option value must be an integer
18393 greater than zero. The default value is 60000.
18396 Set threshold to detect frames as similar. The option value must be an integer
18397 greater than zero. The default value is 116.
18400 Set the minimum length of a sequence in frames to recognize it as matching
18401 sequence. The option value must be a non negative integer value.
18402 The default value is 0.
18405 Set the minimum relation, that matching frames to all frames must have.
18406 The option value must be a double value between 0 and 1. The default value is 0.5.
18409 @subsection Examples
18413 To calculate the signature of an input video and store it in signature.bin:
18415 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18419 To detect whether two videos match and store the signatures in XML format in
18420 signature0.xml and signature1.xml:
18422 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 -
18430 Blur the input video without impacting the outlines.
18432 It accepts the following options:
18435 @item luma_radius, lr
18436 Set the luma radius. The option value must be a float number in
18437 the range [0.1,5.0] that specifies the variance of the gaussian filter
18438 used to blur the image (slower if larger). Default value is 1.0.
18440 @item luma_strength, ls
18441 Set the luma strength. The option value must be a float number
18442 in the range [-1.0,1.0] that configures the blurring. A value included
18443 in [0.0,1.0] will blur the image whereas a value included in
18444 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18446 @item luma_threshold, lt
18447 Set the luma threshold used as a coefficient to determine
18448 whether a pixel should be blurred or not. The option value must be an
18449 integer in the range [-30,30]. A value of 0 will filter all the image,
18450 a value included in [0,30] will filter flat areas and a value included
18451 in [-30,0] will filter edges. Default value is 0.
18453 @item chroma_radius, cr
18454 Set the chroma radius. The option value must be a float number in
18455 the range [0.1,5.0] that specifies the variance of the gaussian filter
18456 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18458 @item chroma_strength, cs
18459 Set the chroma strength. The option value must be a float number
18460 in the range [-1.0,1.0] that configures the blurring. A value included
18461 in [0.0,1.0] will blur the image whereas a value included in
18462 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18464 @item chroma_threshold, ct
18465 Set the chroma threshold used as a coefficient to determine
18466 whether a pixel should be blurred or not. The option value must be an
18467 integer in the range [-30,30]. A value of 0 will filter all the image,
18468 a value included in [0,30] will filter flat areas and a value included
18469 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18472 If a chroma option is not explicitly set, the corresponding luma value
18476 Apply sobel operator to input video stream.
18478 The filter accepts the following option:
18482 Set which planes will be processed, unprocessed planes will be copied.
18483 By default value 0xf, all planes will be processed.
18486 Set value which will be multiplied with filtered result.
18489 Set value which will be added to filtered result.
18492 @subsection Commands
18494 This filter supports the all above options as @ref{commands}.
18499 Apply a simple postprocessing filter that compresses and decompresses the image
18500 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18501 and average the results.
18503 The filter accepts the following options:
18507 Set quality. This option defines the number of levels for averaging. It accepts
18508 an integer in the range 0-6. If set to @code{0}, the filter will have no
18509 effect. A value of @code{6} means the higher quality. For each increment of
18510 that value the speed drops by a factor of approximately 2. Default value is
18514 Force a constant quantization parameter. If not set, the filter will use the QP
18515 from the video stream (if available).
18518 Set thresholding mode. Available modes are:
18522 Set hard thresholding (default).
18524 Set soft thresholding (better de-ringing effect, but likely blurrier).
18527 @item use_bframe_qp
18528 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18529 option may cause flicker since the B-Frames have often larger QP. Default is
18530 @code{0} (not enabled).
18533 @subsection Commands
18535 This filter supports the following commands:
18537 @item quality, level
18538 Set quality level. The value @code{max} can be used to set the maximum level,
18539 currently @code{6}.
18545 Scale the input by applying one of the super-resolution methods based on
18546 convolutional neural networks. Supported models:
18550 Super-Resolution Convolutional Neural Network model (SRCNN).
18551 See @url{https://arxiv.org/abs/1501.00092}.
18554 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18555 See @url{https://arxiv.org/abs/1609.05158}.
18558 Training scripts as well as scripts for model file (.pb) saving can be found at
18559 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18560 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18562 Native model files (.model) can be generated from TensorFlow model
18563 files (.pb) by using tools/python/convert.py
18565 The filter accepts the following options:
18569 Specify which DNN backend to use for model loading and execution. This option accepts
18570 the following values:
18574 Native implementation of DNN loading and execution.
18577 TensorFlow backend. To enable this backend you
18578 need to install the TensorFlow for C library (see
18579 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18580 @code{--enable-libtensorflow}
18583 Default value is @samp{native}.
18586 Set path to model file specifying network architecture and its parameters.
18587 Note that different backends use different file formats. TensorFlow backend
18588 can load files for both formats, while native backend can load files for only
18592 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18593 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18594 input upscaled using bicubic upscaling with proper scale factor.
18597 This feature can also be finished with @ref{dnn_processing} filter.
18601 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18603 This filter takes in input two input videos, the first input is
18604 considered the "main" source and is passed unchanged to the
18605 output. The second input is used as a "reference" video for computing
18608 Both video inputs must have the same resolution and pixel format for
18609 this filter to work correctly. Also it assumes that both inputs
18610 have the same number of frames, which are compared one by one.
18612 The filter stores the calculated SSIM of each frame.
18614 The description of the accepted parameters follows.
18617 @item stats_file, f
18618 If specified the filter will use the named file to save the SSIM of
18619 each individual frame. When filename equals "-" the data is sent to
18623 The file printed if @var{stats_file} is selected, contains a sequence of
18624 key/value pairs of the form @var{key}:@var{value} for each compared
18627 A description of each shown parameter follows:
18631 sequential number of the input frame, starting from 1
18633 @item Y, U, V, R, G, B
18634 SSIM of the compared frames for the component specified by the suffix.
18637 SSIM of the compared frames for the whole frame.
18640 Same as above but in dB representation.
18643 This filter also supports the @ref{framesync} options.
18645 @subsection Examples
18650 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18651 [main][ref] ssim="stats_file=stats.log" [out]
18654 On this example the input file being processed is compared with the
18655 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18656 is stored in @file{stats.log}.
18659 Another example with both psnr and ssim at same time:
18661 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18665 Another example with different containers:
18667 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 -
18673 Convert between different stereoscopic image formats.
18675 The filters accept the following options:
18679 Set stereoscopic image format of input.
18681 Available values for input image formats are:
18684 side by side parallel (left eye left, right eye right)
18687 side by side crosseye (right eye left, left eye right)
18690 side by side parallel with half width resolution
18691 (left eye left, right eye right)
18694 side by side crosseye with half width resolution
18695 (right eye left, left eye right)
18699 above-below (left eye above, right eye below)
18703 above-below (right eye above, left eye below)
18707 above-below with half height resolution
18708 (left eye above, right eye below)
18712 above-below with half height resolution
18713 (right eye above, left eye below)
18716 alternating frames (left eye first, right eye second)
18719 alternating frames (right eye first, left eye second)
18722 interleaved rows (left eye has top row, right eye starts on next row)
18725 interleaved rows (right eye has top row, left eye starts on next row)
18728 interleaved columns, left eye first
18731 interleaved columns, right eye first
18733 Default value is @samp{sbsl}.
18737 Set stereoscopic image format of output.
18741 side by side parallel (left eye left, right eye right)
18744 side by side crosseye (right eye left, left eye right)
18747 side by side parallel with half width resolution
18748 (left eye left, right eye right)
18751 side by side crosseye with half width resolution
18752 (right eye left, left eye right)
18756 above-below (left eye above, right eye below)
18760 above-below (right eye above, left eye below)
18764 above-below with half height resolution
18765 (left eye above, right eye below)
18769 above-below with half height resolution
18770 (right eye above, left eye below)
18773 alternating frames (left eye first, right eye second)
18776 alternating frames (right eye first, left eye second)
18779 interleaved rows (left eye has top row, right eye starts on next row)
18782 interleaved rows (right eye has top row, left eye starts on next row)
18785 anaglyph red/blue gray
18786 (red filter on left eye, blue filter on right eye)
18789 anaglyph red/green gray
18790 (red filter on left eye, green filter on right eye)
18793 anaglyph red/cyan gray
18794 (red filter on left eye, cyan filter on right eye)
18797 anaglyph red/cyan half colored
18798 (red filter on left eye, cyan filter on right eye)
18801 anaglyph red/cyan color
18802 (red filter on left eye, cyan filter on right eye)
18805 anaglyph red/cyan color optimized with the least squares projection of dubois
18806 (red filter on left eye, cyan filter on right eye)
18809 anaglyph green/magenta gray
18810 (green filter on left eye, magenta filter on right eye)
18813 anaglyph green/magenta half colored
18814 (green filter on left eye, magenta filter on right eye)
18817 anaglyph green/magenta colored
18818 (green filter on left eye, magenta filter on right eye)
18821 anaglyph green/magenta color optimized with the least squares projection of dubois
18822 (green filter on left eye, magenta filter on right eye)
18825 anaglyph yellow/blue gray
18826 (yellow filter on left eye, blue filter on right eye)
18829 anaglyph yellow/blue half colored
18830 (yellow filter on left eye, blue filter on right eye)
18833 anaglyph yellow/blue colored
18834 (yellow filter on left eye, blue filter on right eye)
18837 anaglyph yellow/blue color optimized with the least squares projection of dubois
18838 (yellow filter on left eye, blue filter on right eye)
18841 mono output (left eye only)
18844 mono output (right eye only)
18847 checkerboard, left eye first
18850 checkerboard, right eye first
18853 interleaved columns, left eye first
18856 interleaved columns, right eye first
18862 Default value is @samp{arcd}.
18865 @subsection Examples
18869 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
18875 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
18881 @section streamselect, astreamselect
18882 Select video or audio streams.
18884 The filter accepts the following options:
18888 Set number of inputs. Default is 2.
18891 Set input indexes to remap to outputs.
18894 @subsection Commands
18896 The @code{streamselect} and @code{astreamselect} filter supports the following
18901 Set input indexes to remap to outputs.
18904 @subsection Examples
18908 Select first 5 seconds 1st stream and rest of time 2nd stream:
18910 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
18914 Same as above, but for audio:
18916 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
18923 Draw subtitles on top of input video using the libass library.
18925 To enable compilation of this filter you need to configure FFmpeg with
18926 @code{--enable-libass}. This filter also requires a build with libavcodec and
18927 libavformat to convert the passed subtitles file to ASS (Advanced Substation
18928 Alpha) subtitles format.
18930 The filter accepts the following options:
18934 Set the filename of the subtitle file to read. It must be specified.
18936 @item original_size
18937 Specify the size of the original video, the video for which the ASS file
18938 was composed. For the syntax of this option, check the
18939 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18940 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
18941 correctly scale the fonts if the aspect ratio has been changed.
18944 Set a directory path containing fonts that can be used by the filter.
18945 These fonts will be used in addition to whatever the font provider uses.
18948 Process alpha channel, by default alpha channel is untouched.
18951 Set subtitles input character encoding. @code{subtitles} filter only. Only
18952 useful if not UTF-8.
18954 @item stream_index, si
18955 Set subtitles stream index. @code{subtitles} filter only.
18958 Override default style or script info parameters of the subtitles. It accepts a
18959 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
18962 If the first key is not specified, it is assumed that the first value
18963 specifies the @option{filename}.
18965 For example, to render the file @file{sub.srt} on top of the input
18966 video, use the command:
18971 which is equivalent to:
18973 subtitles=filename=sub.srt
18976 To render the default subtitles stream from file @file{video.mkv}, use:
18978 subtitles=video.mkv
18981 To render the second subtitles stream from that file, use:
18983 subtitles=video.mkv:si=1
18986 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
18987 @code{DejaVu Serif}, use:
18989 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
18992 @section super2xsai
18994 Scale the input by 2x and smooth using the Super2xSaI (Scale and
18995 Interpolate) pixel art scaling algorithm.
18997 Useful for enlarging pixel art images without reducing sharpness.
19001 Swap two rectangular objects in video.
19003 This filter accepts the following options:
19013 Set 1st rect x coordinate.
19016 Set 1st rect y coordinate.
19019 Set 2nd rect x coordinate.
19022 Set 2nd rect y coordinate.
19024 All expressions are evaluated once for each frame.
19027 The all options are expressions containing the following constants:
19032 The input width and height.
19035 same as @var{w} / @var{h}
19038 input sample aspect ratio
19041 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19044 The number of the input frame, starting from 0.
19047 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19050 the position in the file of the input frame, NAN if unknown
19057 Blend successive video frames.
19063 Apply telecine process to the video.
19065 This filter accepts the following options:
19074 The default value is @code{top}.
19078 A string of numbers representing the pulldown pattern you wish to apply.
19079 The default value is @code{23}.
19083 Some typical patterns:
19088 24p: 2332 (preferred)
19095 24p: 222222222223 ("Euro pulldown")
19100 @section thistogram
19102 Compute and draw a color distribution histogram for the input video across time.
19104 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19105 at certain time, this filter shows also past histograms of number of frames defined
19106 by @code{width} option.
19108 The computed histogram is a representation of the color component
19109 distribution in an image.
19111 The filter accepts the following options:
19115 Set width of single color component output. Default value is @code{0}.
19116 Value of @code{0} means width will be picked from input video.
19117 This also set number of passed histograms to keep.
19118 Allowed range is [0, 8192].
19120 @item display_mode, d
19122 It accepts the following values:
19125 Per color component graphs are placed below each other.
19128 Per color component graphs are placed side by side.
19131 Presents information identical to that in the @code{parade}, except
19132 that the graphs representing color components are superimposed directly
19135 Default is @code{stack}.
19137 @item levels_mode, m
19138 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19139 Default is @code{linear}.
19141 @item components, c
19142 Set what color components to display.
19143 Default is @code{7}.
19146 Set background opacity. Default is @code{0.9}.
19149 Show envelope. Default is disabled.
19152 Set envelope color. Default is @code{gold}.
19157 Available values for slide is:
19160 Draw new frame when right border is reached.
19163 Replace old columns with new ones.
19166 Scroll from right to left.
19169 Scroll from left to right.
19172 Draw single picture.
19175 Default is @code{replace}.
19180 Apply threshold effect to video stream.
19182 This filter needs four video streams to perform thresholding.
19183 First stream is stream we are filtering.
19184 Second stream is holding threshold values, third stream is holding min values,
19185 and last, fourth stream is holding max values.
19187 The filter accepts the following option:
19191 Set which planes will be processed, unprocessed planes will be copied.
19192 By default value 0xf, all planes will be processed.
19195 For example if first stream pixel's component value is less then threshold value
19196 of pixel component from 2nd threshold stream, third stream value will picked,
19197 otherwise fourth stream pixel component value will be picked.
19199 Using color source filter one can perform various types of thresholding:
19201 @subsection Examples
19205 Binary threshold, using gray color as threshold:
19207 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19211 Inverted binary threshold, using gray color as threshold:
19213 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19217 Truncate binary threshold, using gray color as threshold:
19219 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19223 Threshold to zero, using gray color as threshold:
19225 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19229 Inverted threshold to zero, using gray color as threshold:
19231 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19236 Select the most representative frame in a given sequence of consecutive frames.
19238 The filter accepts the following options:
19242 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19243 will pick one of them, and then handle the next batch of @var{n} frames until
19244 the end. Default is @code{100}.
19247 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19248 value will result in a higher memory usage, so a high value is not recommended.
19250 @subsection Examples
19254 Extract one picture each 50 frames:
19260 Complete example of a thumbnail creation with @command{ffmpeg}:
19262 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19269 Tile several successive frames together.
19271 The @ref{untile} filter can do the reverse.
19273 The filter accepts the following options:
19278 Set the grid size (i.e. the number of lines and columns). For the syntax of
19279 this option, check the
19280 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19283 Set the maximum number of frames to render in the given area. It must be less
19284 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19285 the area will be used.
19288 Set the outer border margin in pixels.
19291 Set the inner border thickness (i.e. the number of pixels between frames). For
19292 more advanced padding options (such as having different values for the edges),
19293 refer to the pad video filter.
19296 Specify the color of the unused area. For the syntax of this option, check the
19297 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19298 The default value of @var{color} is "black".
19301 Set the number of frames to overlap when tiling several successive frames together.
19302 The value must be between @code{0} and @var{nb_frames - 1}.
19305 Set the number of frames to initially be empty before displaying first output frame.
19306 This controls how soon will one get first output frame.
19307 The value must be between @code{0} and @var{nb_frames - 1}.
19310 @subsection Examples
19314 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19316 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19318 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19319 duplicating each output frame to accommodate the originally detected frame
19323 Display @code{5} pictures in an area of @code{3x2} frames,
19324 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19325 mixed flat and named options:
19327 tile=3x2:nb_frames=5:padding=7:margin=2
19331 @section tinterlace
19333 Perform various types of temporal field interlacing.
19335 Frames are counted starting from 1, so the first input frame is
19338 The filter accepts the following options:
19343 Specify the mode of the interlacing. This option can also be specified
19344 as a value alone. See below for a list of values for this option.
19346 Available values are:
19350 Move odd frames into the upper field, even into the lower field,
19351 generating a double height frame at half frame rate.
19355 Frame 1 Frame 2 Frame 3 Frame 4
19357 11111 22222 33333 44444
19358 11111 22222 33333 44444
19359 11111 22222 33333 44444
19360 11111 22222 33333 44444
19374 Only output odd frames, even frames are dropped, generating a frame with
19375 unchanged height at half frame rate.
19380 Frame 1 Frame 2 Frame 3 Frame 4
19382 11111 22222 33333 44444
19383 11111 22222 33333 44444
19384 11111 22222 33333 44444
19385 11111 22222 33333 44444
19395 Only output even frames, odd frames are dropped, generating a frame with
19396 unchanged height at half frame rate.
19401 Frame 1 Frame 2 Frame 3 Frame 4
19403 11111 22222 33333 44444
19404 11111 22222 33333 44444
19405 11111 22222 33333 44444
19406 11111 22222 33333 44444
19416 Expand each frame to full height, but pad alternate lines with black,
19417 generating a frame with double height at the same input frame rate.
19422 Frame 1 Frame 2 Frame 3 Frame 4
19424 11111 22222 33333 44444
19425 11111 22222 33333 44444
19426 11111 22222 33333 44444
19427 11111 22222 33333 44444
19430 11111 ..... 33333 .....
19431 ..... 22222 ..... 44444
19432 11111 ..... 33333 .....
19433 ..... 22222 ..... 44444
19434 11111 ..... 33333 .....
19435 ..... 22222 ..... 44444
19436 11111 ..... 33333 .....
19437 ..... 22222 ..... 44444
19441 @item interleave_top, 4
19442 Interleave the upper field from odd frames with the lower field from
19443 even frames, generating a frame with unchanged height at half frame rate.
19448 Frame 1 Frame 2 Frame 3 Frame 4
19450 11111<- 22222 33333<- 44444
19451 11111 22222<- 33333 44444<-
19452 11111<- 22222 33333<- 44444
19453 11111 22222<- 33333 44444<-
19463 @item interleave_bottom, 5
19464 Interleave the lower field from odd frames with the upper field from
19465 even frames, generating a frame with 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 @item interlacex2, 6
19486 Double frame rate with unchanged height. Frames are inserted each
19487 containing the second temporal field from the previous input frame and
19488 the first temporal field from the next input frame. This mode relies on
19489 the top_field_first flag. Useful for interlaced video displays with no
19490 field synchronisation.
19495 Frame 1 Frame 2 Frame 3 Frame 4
19497 11111 22222 33333 44444
19498 11111 22222 33333 44444
19499 11111 22222 33333 44444
19500 11111 22222 33333 44444
19503 11111 22222 22222 33333 33333 44444 44444
19504 11111 11111 22222 22222 33333 33333 44444
19505 11111 22222 22222 33333 33333 44444 44444
19506 11111 11111 22222 22222 33333 33333 44444
19511 Move odd frames into the upper field, even into the lower field,
19512 generating a double height frame at same frame rate.
19517 Frame 1 Frame 2 Frame 3 Frame 4
19519 11111 22222 33333 44444
19520 11111 22222 33333 44444
19521 11111 22222 33333 44444
19522 11111 22222 33333 44444
19525 11111 33333 33333 55555
19526 22222 22222 44444 44444
19527 11111 33333 33333 55555
19528 22222 22222 44444 44444
19529 11111 33333 33333 55555
19530 22222 22222 44444 44444
19531 11111 33333 33333 55555
19532 22222 22222 44444 44444
19537 Numeric values are deprecated but are accepted for backward
19538 compatibility reasons.
19540 Default mode is @code{merge}.
19543 Specify flags influencing the filter process.
19545 Available value for @var{flags} is:
19548 @item low_pass_filter, vlpf
19549 Enable linear vertical low-pass filtering in the filter.
19550 Vertical low-pass filtering is required when creating an interlaced
19551 destination from a progressive source which contains high-frequency
19552 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19555 @item complex_filter, cvlpf
19556 Enable complex vertical low-pass filtering.
19557 This will slightly less reduce interlace 'twitter' and Moire
19558 patterning but better retain detail and subjective sharpness impression.
19561 Bypass already interlaced frames, only adjust the frame rate.
19564 Vertical low-pass filtering and bypassing already interlaced frames can only be
19565 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19570 Pick median pixels from several successive input video frames.
19572 The filter accepts the following options:
19576 Set radius of median filter.
19577 Default is 1. Allowed range is from 1 to 127.
19580 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19583 Set median percentile. Default value is @code{0.5}.
19584 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19585 minimum values, and @code{1} maximum values.
19588 @subsection Commands
19590 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19592 @section tmidequalizer
19594 Apply Temporal Midway Video Equalization effect.
19596 Midway Video Equalization adjusts a sequence of video frames to have the same
19597 histograms, while maintaining their dynamics as much as possible. It's
19598 useful for e.g. matching exposures from a video frames sequence.
19600 This filter accepts the following option:
19604 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19607 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19608 Setting this option to 0 effectively does nothing.
19611 Set which planes to process. Default is @code{15}, which is all available planes.
19616 Mix successive video frames.
19618 A description of the accepted options follows.
19622 The number of successive frames to mix. If unspecified, it defaults to 3.
19625 Specify weight of each input video frame.
19626 Each weight is separated by space. If number of weights is smaller than
19627 number of @var{frames} last specified weight will be used for all remaining
19631 Specify scale, if it is set it will be multiplied with sum
19632 of each weight multiplied with pixel values to give final destination
19633 pixel value. By default @var{scale} is auto scaled to sum of weights.
19636 @subsection Examples
19640 Average 7 successive frames:
19642 tmix=frames=7:weights="1 1 1 1 1 1 1"
19646 Apply simple temporal convolution:
19648 tmix=frames=3:weights="-1 3 -1"
19652 Similar as above but only showing temporal differences:
19654 tmix=frames=3:weights="-1 2 -1":scale=1
19660 Tone map colors from different dynamic ranges.
19662 This filter expects data in single precision floating point, as it needs to
19663 operate on (and can output) out-of-range values. Another filter, such as
19664 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19666 The tonemapping algorithms implemented only work on linear light, so input
19667 data should be linearized beforehand (and possibly correctly tagged).
19670 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19673 @subsection Options
19674 The filter accepts the following options.
19678 Set the tone map algorithm to use.
19680 Possible values are:
19683 Do not apply any tone map, only desaturate overbright pixels.
19686 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19687 in-range values, while distorting out-of-range values.
19690 Stretch the entire reference gamut to a linear multiple of the display.
19693 Fit a logarithmic transfer between the tone curves.
19696 Preserve overall image brightness with a simple curve, using nonlinear
19697 contrast, which results in flattening details and degrading color accuracy.
19700 Preserve both dark and bright details better than @var{reinhard}, at the cost
19701 of slightly darkening everything. Use it when detail preservation is more
19702 important than color and brightness accuracy.
19705 Smoothly map out-of-range values, while retaining contrast and colors for
19706 in-range material as much as possible. Use it when color accuracy is more
19707 important than detail preservation.
19713 Tune the tone mapping algorithm.
19715 This affects the following algorithms:
19721 Specifies the scale factor to use while stretching.
19725 Specifies the exponent of the function.
19729 Specify an extra linear coefficient to multiply into the signal before clipping.
19733 Specify the local contrast coefficient at the display peak.
19734 Default to 0.5, which means that in-gamut values will be about half as bright
19741 Specify the transition point from linear to mobius transform. Every value
19742 below this point is guaranteed to be mapped 1:1. The higher the value, the
19743 more accurate the result will be, at the cost of losing bright details.
19744 Default to 0.3, which due to the steep initial slope still preserves in-range
19745 colors fairly accurately.
19749 Apply desaturation for highlights that exceed this level of brightness. The
19750 higher the parameter, the more color information will be preserved. This
19751 setting helps prevent unnaturally blown-out colors for super-highlights, by
19752 (smoothly) turning into white instead. This makes images feel more natural,
19753 at the cost of reducing information about out-of-range colors.
19755 The default of 2.0 is somewhat conservative and will mostly just apply to
19756 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
19758 This option works only if the input frame has a supported color tag.
19761 Override signal/nominal/reference peak with this value. Useful when the
19762 embedded peak information in display metadata is not reliable or when tone
19763 mapping from a lower range to a higher range.
19768 Temporarily pad video frames.
19770 The filter accepts the following options:
19774 Specify number of delay frames before input video stream. Default is 0.
19777 Specify number of padding frames after input video stream.
19778 Set to -1 to pad indefinitely. Default is 0.
19781 Set kind of frames added to beginning of stream.
19782 Can be either @var{add} or @var{clone}.
19783 With @var{add} frames of solid-color are added.
19784 With @var{clone} frames are clones of first frame.
19785 Default is @var{add}.
19788 Set kind of frames added to end of stream.
19789 Can be either @var{add} or @var{clone}.
19790 With @var{add} frames of solid-color are added.
19791 With @var{clone} frames are clones of last frame.
19792 Default is @var{add}.
19794 @item start_duration, stop_duration
19795 Specify the duration of the start/stop delay. See
19796 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19797 for the accepted syntax.
19798 These options override @var{start} and @var{stop}. Default is 0.
19801 Specify the color of the padded area. For the syntax of this option,
19802 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
19803 manual,ffmpeg-utils}.
19805 The default value of @var{color} is "black".
19811 Transpose rows with columns in the input video and optionally flip it.
19813 It accepts the following parameters:
19818 Specify the transposition direction.
19820 Can assume the following values:
19822 @item 0, 4, cclock_flip
19823 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
19831 Rotate by 90 degrees clockwise, that is:
19839 Rotate by 90 degrees counterclockwise, that is:
19846 @item 3, 7, clock_flip
19847 Rotate by 90 degrees clockwise and vertically flip, that is:
19855 For values between 4-7, the transposition is only done if the input
19856 video geometry is portrait and not landscape. These values are
19857 deprecated, the @code{passthrough} option should be used instead.
19859 Numerical values are deprecated, and should be dropped in favor of
19860 symbolic constants.
19863 Do not apply the transposition if the input geometry matches the one
19864 specified by the specified value. It accepts the following values:
19867 Always apply transposition.
19869 Preserve portrait geometry (when @var{height} >= @var{width}).
19871 Preserve landscape geometry (when @var{width} >= @var{height}).
19874 Default value is @code{none}.
19877 For example to rotate by 90 degrees clockwise and preserve portrait
19880 transpose=dir=1:passthrough=portrait
19883 The command above can also be specified as:
19885 transpose=1:portrait
19888 @section transpose_npp
19890 Transpose rows with columns in the input video and optionally flip it.
19891 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
19893 It accepts the following parameters:
19898 Specify the transposition direction.
19900 Can assume the following values:
19903 Rotate by 90 degrees counterclockwise and vertically flip. (default)
19906 Rotate by 90 degrees clockwise.
19909 Rotate by 90 degrees counterclockwise.
19912 Rotate by 90 degrees clockwise and vertically flip.
19916 Do not apply the transposition if the input geometry matches the one
19917 specified by the specified value. It accepts the following values:
19920 Always apply transposition. (default)
19922 Preserve portrait geometry (when @var{height} >= @var{width}).
19924 Preserve landscape geometry (when @var{width} >= @var{height}).
19930 Trim the input so that the output contains one continuous subpart of the input.
19932 It accepts the following parameters:
19935 Specify the time of the start of the kept section, i.e. the frame with the
19936 timestamp @var{start} will be the first frame in the output.
19939 Specify the time of the first frame that will be dropped, i.e. the frame
19940 immediately preceding the one with the timestamp @var{end} will be the last
19941 frame in the output.
19944 This is the same as @var{start}, except this option sets the start timestamp
19945 in timebase units instead of seconds.
19948 This is the same as @var{end}, except this option sets the end timestamp
19949 in timebase units instead of seconds.
19952 The maximum duration of the output in seconds.
19955 The number of the first frame that should be passed to the output.
19958 The number of the first frame that should be dropped.
19961 @option{start}, @option{end}, and @option{duration} are expressed as time
19962 duration specifications; see
19963 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19964 for the accepted syntax.
19966 Note that the first two sets of the start/end options and the @option{duration}
19967 option look at the frame timestamp, while the _frame variants simply count the
19968 frames that pass through the filter. Also note that this filter does not modify
19969 the timestamps. If you wish for the output timestamps to start at zero, insert a
19970 setpts filter after the trim filter.
19972 If multiple start or end options are set, this filter tries to be greedy and
19973 keep all the frames that match at least one of the specified constraints. To keep
19974 only the part that matches all the constraints at once, chain multiple trim
19977 The defaults are such that all the input is kept. So it is possible to set e.g.
19978 just the end values to keep everything before the specified time.
19983 Drop everything except the second minute of input:
19985 ffmpeg -i INPUT -vf trim=60:120
19989 Keep only the first second:
19991 ffmpeg -i INPUT -vf trim=duration=1
19996 @section unpremultiply
19997 Apply alpha unpremultiply effect to input video stream using first plane
19998 of second stream as alpha.
20000 Both streams must have same dimensions and same pixel format.
20002 The filter accepts the following option:
20006 Set which planes will be processed, unprocessed planes will be copied.
20007 By default value 0xf, all planes will be processed.
20009 If the format has 1 or 2 components, then luma is bit 0.
20010 If the format has 3 or 4 components:
20011 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20012 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20013 If present, the alpha channel is always the last bit.
20016 Do not require 2nd input for processing, instead use alpha plane from input stream.
20022 Sharpen or blur the input video.
20024 It accepts the following parameters:
20027 @item luma_msize_x, lx
20028 Set the luma matrix horizontal size. It must be an odd integer between
20029 3 and 23. The default value is 5.
20031 @item luma_msize_y, ly
20032 Set the luma matrix vertical size. It must be an odd integer between 3
20033 and 23. The default value is 5.
20035 @item luma_amount, la
20036 Set the luma effect strength. It must be a floating point number, reasonable
20037 values lay between -1.5 and 1.5.
20039 Negative values will blur the input video, while positive values will
20040 sharpen it, a value of zero will disable the effect.
20042 Default value is 1.0.
20044 @item chroma_msize_x, cx
20045 Set the chroma matrix horizontal size. It must be an odd integer
20046 between 3 and 23. The default value is 5.
20048 @item chroma_msize_y, cy
20049 Set the chroma matrix vertical size. It must be an odd integer
20050 between 3 and 23. The default value is 5.
20052 @item chroma_amount, ca
20053 Set the chroma effect strength. It must be a floating point number, reasonable
20054 values lay between -1.5 and 1.5.
20056 Negative values will blur the input video, while positive values will
20057 sharpen it, a value of zero will disable the effect.
20059 Default value is 0.0.
20063 All parameters are optional and default to the equivalent of the
20064 string '5:5:1.0:5:5:0.0'.
20066 @subsection Examples
20070 Apply strong luma sharpen effect:
20072 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20076 Apply a strong blur of both luma and chroma parameters:
20078 unsharp=7:7:-2:7:7:-2
20085 Decompose a video made of tiled images into the individual images.
20087 The frame rate of the output video is the frame rate of the input video
20088 multiplied by the number of tiles.
20090 This filter does the reverse of @ref{tile}.
20092 The filter accepts the following options:
20097 Set the grid size (i.e. the number of lines and columns). For the syntax of
20098 this option, check the
20099 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20102 @subsection Examples
20106 Produce a 1-second video from a still image file made of 25 frames stacked
20107 vertically, like an analogic film reel:
20109 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20115 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20116 the image at several (or - in the case of @option{quality} level @code{8} - all)
20117 shifts and average the results.
20119 The way this differs from the behavior of spp is that uspp actually encodes &
20120 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20121 DCT similar to MJPEG.
20123 The filter accepts the following options:
20127 Set quality. This option defines the number of levels for averaging. It accepts
20128 an integer in the range 0-8. If set to @code{0}, the filter will have no
20129 effect. A value of @code{8} means the higher quality. For each increment of
20130 that value the speed drops by a factor of approximately 2. Default value is
20134 Force a constant quantization parameter. If not set, the filter will use the QP
20135 from the video stream (if available).
20140 Convert 360 videos between various formats.
20142 The filter accepts the following options:
20148 Set format of the input/output video.
20156 Equirectangular projection.
20161 Cubemap with 3x2/6x1/1x6 layout.
20163 Format specific options:
20168 Set padding proportion for the input/output cubemap. Values in decimals.
20175 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)
20178 Default value is @b{@samp{0}}.
20179 Maximum value is @b{@samp{0.1}}.
20183 Set fixed padding for the input/output cubemap. Values in pixels.
20185 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20189 Set order of faces for the input/output cubemap. Choose one direction for each position.
20191 Designation of directions:
20207 Default value is @b{@samp{rludfb}}.
20211 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20213 Designation of angles:
20216 0 degrees clockwise
20218 90 degrees clockwise
20220 180 degrees clockwise
20222 270 degrees clockwise
20225 Default value is @b{@samp{000000}}.
20229 Equi-Angular Cubemap.
20236 Format specific options:
20241 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20243 If diagonal field of view is set it overrides horizontal and vertical field of view.
20248 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20250 If diagonal field of view is set it overrides horizontal and vertical field of view.
20256 Format specific options:
20261 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20263 If diagonal field of view is set it overrides horizontal and vertical field of view.
20268 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20270 If diagonal field of view is set it overrides horizontal and vertical field of view.
20276 Facebook's 360 formats.
20279 Stereographic format.
20281 Format specific options:
20286 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20288 If diagonal field of view is set it overrides horizontal and vertical field of view.
20293 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20295 If diagonal field of view is set it overrides horizontal and vertical field of view.
20302 Ball format, gives significant distortion toward the back.
20305 Hammer-Aitoff map projection format.
20308 Sinusoidal map projection format.
20311 Fisheye projection.
20313 Format specific options:
20318 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20320 If diagonal field of view is set it overrides horizontal and vertical field of view.
20325 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20327 If diagonal field of view is set it overrides horizontal and vertical field of view.
20331 Pannini projection.
20333 Format specific options:
20336 Set output pannini parameter.
20339 Set input pannini parameter.
20343 Cylindrical projection.
20345 Format specific options:
20350 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20352 If diagonal field of view is set it overrides horizontal and vertical field of view.
20357 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20359 If diagonal field of view is set it overrides horizontal and vertical field of view.
20363 Perspective projection. @i{(output only)}
20365 Format specific options:
20368 Set perspective parameter.
20372 Tetrahedron projection.
20375 Truncated square pyramid projection.
20379 Half equirectangular projection.
20384 Format specific options:
20389 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20391 If diagonal field of view is set it overrides horizontal and vertical field of view.
20396 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20398 If diagonal field of view is set it overrides horizontal and vertical field of view.
20402 Orthographic format.
20404 Format specific options:
20409 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20411 If diagonal field of view is set it overrides horizontal and vertical field of view.
20416 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20418 If diagonal field of view is set it overrides horizontal and vertical field of view.
20422 Octahedron projection.
20426 Set interpolation method.@*
20427 @i{Note: more complex interpolation methods require much more memory to run.}
20437 Bilinear interpolation.
20439 Lagrange9 interpolation.
20442 Bicubic interpolation.
20445 Lanczos interpolation.
20448 Spline16 interpolation.
20451 Gaussian interpolation.
20453 Mitchell interpolation.
20456 Default value is @b{@samp{line}}.
20460 Set the output video resolution.
20462 Default resolution depends on formats.
20466 Set the input/output stereo format.
20477 Default value is @b{@samp{2d}} for input and output format.
20482 Set rotation for the output video. Values in degrees.
20485 Set rotation order for the output video. Choose one item for each position.
20496 Default value is @b{@samp{ypr}}.
20501 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20505 Set if input video is flipped horizontally/vertically. Boolean values.
20508 Set if input video is transposed. Boolean value, by default disabled.
20511 Set if output video needs to be transposed. Boolean value, by default disabled.
20514 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20517 @subsection Examples
20521 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20523 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20526 Extract back view of Equi-Angular Cubemap:
20528 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20531 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20533 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20537 @subsection Commands
20539 This filter supports subset of above options as @ref{commands}.
20541 @section vaguedenoiser
20543 Apply a wavelet based denoiser.
20545 It transforms each frame from the video input into the wavelet domain,
20546 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20547 the obtained coefficients. It does an inverse wavelet transform after.
20548 Due to wavelet properties, it should give a nice smoothed result, and
20549 reduced noise, without blurring picture features.
20551 This filter accepts the following options:
20555 The filtering strength. The higher, the more filtered the video will be.
20556 Hard thresholding can use a higher threshold than soft thresholding
20557 before the video looks overfiltered. Default value is 2.
20560 The filtering method the filter will use.
20562 It accepts the following values:
20565 All values under the threshold will be zeroed.
20568 All values under the threshold will be zeroed. All values above will be
20569 reduced by the threshold.
20572 Scales or nullifies coefficients - intermediary between (more) soft and
20573 (less) hard thresholding.
20576 Default is garrote.
20579 Number of times, the wavelet will decompose the picture. Picture can't
20580 be decomposed beyond a particular point (typically, 8 for a 640x480
20581 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20584 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20587 A list of the planes to process. By default all planes are processed.
20590 The threshold type the filter will use.
20592 It accepts the following values:
20595 Threshold used is same for all decompositions.
20598 Threshold used depends also on each decomposition coefficients.
20601 Default is universal.
20604 @section vectorscope
20606 Display 2 color component values in the two dimensional graph (which is called
20609 This filter accepts the following options:
20613 Set vectorscope mode.
20615 It accepts the following values:
20619 Gray values are displayed on graph, higher brightness means more pixels have
20620 same component color value on location in graph. This is the default mode.
20623 Gray values are displayed on graph. Surrounding pixels values which are not
20624 present in video frame are drawn in gradient of 2 color components which are
20625 set by option @code{x} and @code{y}. The 3rd color component is static.
20628 Actual color components values present in video frame are displayed on graph.
20631 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20632 on graph increases value of another color component, which is luminance by
20633 default values of @code{x} and @code{y}.
20636 Actual colors present in video frame are displayed on graph. If two different
20637 colors map to same position on graph then color with higher value of component
20638 not present in graph is picked.
20641 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20642 component picked from radial gradient.
20646 Set which color component will be represented on X-axis. Default is @code{1}.
20649 Set which color component will be represented on Y-axis. Default is @code{2}.
20652 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20653 of color component which represents frequency of (X, Y) location in graph.
20658 No envelope, this is default.
20661 Instant envelope, even darkest single pixel will be clearly highlighted.
20664 Hold maximum and minimum values presented in graph over time. This way you
20665 can still spot out of range values without constantly looking at vectorscope.
20668 Peak and instant envelope combined together.
20672 Set what kind of graticule to draw.
20681 Set graticule opacity.
20684 Set graticule flags.
20688 Draw graticule for white point.
20691 Draw graticule for black point.
20694 Draw color points short names.
20698 Set background opacity.
20700 @item lthreshold, l
20701 Set low threshold for color component not represented on X or Y axis.
20702 Values lower than this value will be ignored. Default is 0.
20703 Note this value is multiplied with actual max possible value one pixel component
20704 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
20707 @item hthreshold, h
20708 Set high threshold for color component not represented on X or Y axis.
20709 Values higher than this value will be ignored. Default is 1.
20710 Note this value is multiplied with actual max possible value one pixel component
20711 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
20712 is 0.9 * 255 = 230.
20714 @item colorspace, c
20715 Set what kind of colorspace to use when drawing graticule.
20725 Set color tint for gray/tint vectorscope mode. By default both options are zero.
20726 This means no tint, and output will remain gray.
20729 @anchor{vidstabdetect}
20730 @section vidstabdetect
20732 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
20733 @ref{vidstabtransform} for pass 2.
20735 This filter generates a file with relative translation and rotation
20736 transform information about subsequent frames, which is then used by
20737 the @ref{vidstabtransform} filter.
20739 To enable compilation of this filter you need to configure FFmpeg with
20740 @code{--enable-libvidstab}.
20742 This filter accepts the following options:
20746 Set the path to the file used to write the transforms information.
20747 Default value is @file{transforms.trf}.
20750 Set how shaky the video is and how quick the camera is. It accepts an
20751 integer in the range 1-10, a value of 1 means little shakiness, a
20752 value of 10 means strong shakiness. Default value is 5.
20755 Set the accuracy of the detection process. It must be a value in the
20756 range 1-15. A value of 1 means low accuracy, a value of 15 means high
20757 accuracy. Default value is 15.
20760 Set stepsize of the search process. The region around minimum is
20761 scanned with 1 pixel resolution. Default value is 6.
20764 Set minimum contrast. Below this value a local measurement field is
20765 discarded. Must be a floating point value in the range 0-1. Default
20769 Set reference frame number for tripod mode.
20771 If enabled, the motion of the frames is compared to a reference frame
20772 in the filtered stream, identified by the specified number. The idea
20773 is to compensate all movements in a more-or-less static scene and keep
20774 the camera view absolutely still.
20776 If set to 0, it is disabled. The frames are counted starting from 1.
20779 Show fields and transforms in the resulting frames. It accepts an
20780 integer in the range 0-2. Default value is 0, which disables any
20784 @subsection Examples
20788 Use default values:
20794 Analyze strongly shaky movie and put the results in file
20795 @file{mytransforms.trf}:
20797 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
20801 Visualize the result of internal transformations in the resulting
20804 vidstabdetect=show=1
20808 Analyze a video with medium shakiness using @command{ffmpeg}:
20810 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
20814 @anchor{vidstabtransform}
20815 @section vidstabtransform
20817 Video stabilization/deshaking: pass 2 of 2,
20818 see @ref{vidstabdetect} for pass 1.
20820 Read a file with transform information for each frame and
20821 apply/compensate them. Together with the @ref{vidstabdetect}
20822 filter this can be used to deshake videos. See also
20823 @url{http://public.hronopik.de/vid.stab}. It is important to also use
20824 the @ref{unsharp} filter, see below.
20826 To enable compilation of this filter you need to configure FFmpeg with
20827 @code{--enable-libvidstab}.
20829 @subsection Options
20833 Set path to the file used to read the transforms. Default value is
20834 @file{transforms.trf}.
20837 Set the number of frames (value*2 + 1) used for lowpass filtering the
20838 camera movements. Default value is 10.
20840 For example a number of 10 means that 21 frames are used (10 in the
20841 past and 10 in the future) to smoothen the motion in the video. A
20842 larger value leads to a smoother video, but limits the acceleration of
20843 the camera (pan/tilt movements). 0 is a special case where a static
20844 camera is simulated.
20847 Set the camera path optimization algorithm.
20849 Accepted values are:
20852 gaussian kernel low-pass filter on camera motion (default)
20854 averaging on transformations
20858 Set maximal number of pixels to translate frames. Default value is -1,
20862 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
20863 value is -1, meaning no limit.
20866 Specify how to deal with borders that may be visible due to movement
20869 Available values are:
20872 keep image information from previous frame (default)
20874 fill the border black
20878 Invert transforms if set to 1. Default value is 0.
20881 Consider transforms as relative to previous frame if set to 1,
20882 absolute if set to 0. Default value is 0.
20885 Set percentage to zoom. A positive value will result in a zoom-in
20886 effect, a negative value in a zoom-out effect. Default value is 0 (no
20890 Set optimal zooming to avoid borders.
20892 Accepted values are:
20897 optimal static zoom value is determined (only very strong movements
20898 will lead to visible borders) (default)
20900 optimal adaptive zoom value is determined (no borders will be
20901 visible), see @option{zoomspeed}
20904 Note that the value given at zoom is added to the one calculated here.
20907 Set percent to zoom maximally each frame (enabled when
20908 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
20912 Specify type of interpolation.
20914 Available values are:
20919 linear only horizontal
20921 linear in both directions (default)
20923 cubic in both directions (slow)
20927 Enable virtual tripod mode if set to 1, which is equivalent to
20928 @code{relative=0:smoothing=0}. Default value is 0.
20930 Use also @code{tripod} option of @ref{vidstabdetect}.
20933 Increase log verbosity if set to 1. Also the detected global motions
20934 are written to the temporary file @file{global_motions.trf}. Default
20938 @subsection Examples
20942 Use @command{ffmpeg} for a typical stabilization with default values:
20944 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
20947 Note the use of the @ref{unsharp} filter which is always recommended.
20950 Zoom in a bit more and load transform data from a given file:
20952 vidstabtransform=zoom=5:input="mytransforms.trf"
20956 Smoothen the video even more:
20958 vidstabtransform=smoothing=30
20964 Flip the input video vertically.
20966 For example, to vertically flip a video with @command{ffmpeg}:
20968 ffmpeg -i in.avi -vf "vflip" out.avi
20973 Detect variable frame rate video.
20975 This filter tries to detect if the input is variable or constant frame rate.
20977 At end it will output number of frames detected as having variable delta pts,
20978 and ones with constant delta pts.
20979 If there was frames with variable delta, than it will also show min, max and
20980 average delta encountered.
20984 Boost or alter saturation.
20986 The filter accepts the following options:
20989 Set strength of boost if positive value or strength of alter if negative value.
20990 Default is 0. Allowed range is from -2 to 2.
20993 Set the red balance. Default is 1. Allowed range is from -10 to 10.
20996 Set the green balance. Default is 1. Allowed range is from -10 to 10.
20999 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21002 Set the red luma coefficient.
21005 Set the green luma coefficient.
21008 Set the blue luma coefficient.
21011 If @code{intensity} is negative and this is set to 1, colors will change,
21012 otherwise colors will be less saturated, more towards gray.
21015 @subsection Commands
21017 This filter supports the all above options as @ref{commands}.
21022 Make or reverse a natural vignetting effect.
21024 The filter accepts the following options:
21028 Set lens angle expression as a number of radians.
21030 The value is clipped in the @code{[0,PI/2]} range.
21032 Default value: @code{"PI/5"}
21036 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21040 Set forward/backward mode.
21042 Available modes are:
21045 The larger the distance from the central point, the darker the image becomes.
21048 The larger the distance from the central point, the brighter the image becomes.
21049 This can be used to reverse a vignette effect, though there is no automatic
21050 detection to extract the lens @option{angle} and other settings (yet). It can
21051 also be used to create a burning effect.
21054 Default value is @samp{forward}.
21057 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21059 It accepts the following values:
21062 Evaluate expressions only once during the filter initialization.
21065 Evaluate expressions for each incoming frame. This is way slower than the
21066 @samp{init} mode since it requires all the scalers to be re-computed, but it
21067 allows advanced dynamic expressions.
21070 Default value is @samp{init}.
21073 Set dithering to reduce the circular banding effects. Default is @code{1}
21077 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21078 Setting this value to the SAR of the input will make a rectangular vignetting
21079 following the dimensions of the video.
21081 Default is @code{1/1}.
21084 @subsection Expressions
21086 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21087 following parameters.
21092 input width and height
21095 the number of input frame, starting from 0
21098 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21099 @var{TB} units, NAN if undefined
21102 frame rate of the input video, NAN if the input frame rate is unknown
21105 the PTS (Presentation TimeStamp) of the filtered video frame,
21106 expressed in seconds, NAN if undefined
21109 time base of the input video
21113 @subsection Examples
21117 Apply simple strong vignetting effect:
21123 Make a flickering vignetting:
21125 vignette='PI/4+random(1)*PI/50':eval=frame
21130 @section vmafmotion
21132 Obtain the average VMAF motion score of a video.
21133 It is one of the component metrics of VMAF.
21135 The obtained average motion score is printed through the logging system.
21137 The filter accepts the following options:
21141 If specified, the filter will use the named file to save the motion score of
21142 each frame with respect to the previous frame.
21143 When filename equals "-" the data is sent to standard output.
21148 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21152 Stack input videos vertically.
21154 All streams must be of same pixel format and of same width.
21156 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21157 to create same output.
21159 The filter accepts the following options:
21163 Set number of input streams. Default is 2.
21166 If set to 1, force the output to terminate when the shortest input
21167 terminates. Default value is 0.
21172 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21173 Deinterlacing Filter").
21175 Based on the process described by Martin Weston for BBC R&D, and
21176 implemented based on the de-interlace algorithm written by Jim
21177 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21178 uses filter coefficients calculated by BBC R&D.
21180 This filter uses field-dominance information in frame to decide which
21181 of each pair of fields to place first in the output.
21182 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21184 There are two sets of filter coefficients, so called "simple"
21185 and "complex". Which set of filter coefficients is used can
21186 be set by passing an optional parameter:
21190 Set the interlacing filter coefficients. Accepts one of the following values:
21194 Simple filter coefficient set.
21196 More-complex filter coefficient set.
21198 Default value is @samp{complex}.
21201 The interlacing mode to adopt. It accepts one of the following values:
21205 Output one frame for each frame.
21207 Output one frame for each field.
21210 The default value is @code{field}.
21213 The picture field parity assumed for the input interlaced video. It accepts one
21214 of the following values:
21218 Assume the top field is first.
21220 Assume the bottom field is first.
21222 Enable automatic detection of field parity.
21225 The default value is @code{auto}.
21226 If the interlacing is unknown or the decoder does not export this information,
21227 top field first will be assumed.
21230 Specify which frames to deinterlace. Accepts one of the following values:
21234 Deinterlace all frames,
21236 Only deinterlace frames marked as interlaced.
21239 Default value is @samp{all}.
21242 @subsection Commands
21243 This filter supports same @ref{commands} as options.
21246 Video waveform monitor.
21248 The waveform monitor plots color component intensity. By default luminance
21249 only. Each column of the waveform corresponds to a column of pixels in the
21252 It accepts the following options:
21256 Can be either @code{row}, or @code{column}. Default is @code{column}.
21257 In row mode, the graph on the left side represents color component value 0 and
21258 the right side represents value = 255. In column mode, the top side represents
21259 color component value = 0 and bottom side represents value = 255.
21262 Set intensity. Smaller values are useful to find out how many values of the same
21263 luminance are distributed across input rows/columns.
21264 Default value is @code{0.04}. Allowed range is [0, 1].
21267 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21268 In mirrored mode, higher values will be represented on the left
21269 side for @code{row} mode and at the top for @code{column} mode. Default is
21270 @code{1} (mirrored).
21274 It accepts the following values:
21277 Presents information identical to that in the @code{parade}, except
21278 that the graphs representing color components are superimposed directly
21281 This display mode makes it easier to spot relative differences or similarities
21282 in overlapping areas of the color components that are supposed to be identical,
21283 such as neutral whites, grays, or blacks.
21286 Display separate graph for the color components side by side in
21287 @code{row} mode or one below the other in @code{column} mode.
21290 Display separate graph for the color components side by side in
21291 @code{column} mode or one below the other in @code{row} mode.
21293 Using this display mode makes it easy to spot color casts in the highlights
21294 and shadows of an image, by comparing the contours of the top and the bottom
21295 graphs of each waveform. Since whites, grays, and blacks are characterized
21296 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21297 should display three waveforms of roughly equal width/height. If not, the
21298 correction is easy to perform by making level adjustments the three waveforms.
21300 Default is @code{stack}.
21302 @item components, c
21303 Set which color components to display. Default is 1, which means only luminance
21304 or red color component if input is in RGB colorspace. If is set for example to
21305 7 it will display all 3 (if) available color components.
21310 No envelope, this is default.
21313 Instant envelope, minimum and maximum values presented in graph will be easily
21314 visible even with small @code{step} value.
21317 Hold minimum and maximum values presented in graph across time. This way you
21318 can still spot out of range values without constantly looking at waveforms.
21321 Peak and instant envelope combined together.
21327 No filtering, this is default.
21330 Luma and chroma combined together.
21333 Similar as above, but shows difference between blue and red chroma.
21336 Similar as above, but use different colors.
21339 Similar as above, but again with different colors.
21342 Displays only chroma.
21345 Displays actual color value on waveform.
21348 Similar as above, but with luma showing frequency of chroma values.
21352 Set which graticule to display.
21356 Do not display graticule.
21359 Display green graticule showing legal broadcast ranges.
21362 Display orange graticule showing legal broadcast ranges.
21365 Display invert graticule showing legal broadcast ranges.
21369 Set graticule opacity.
21372 Set graticule flags.
21376 Draw numbers above lines. By default enabled.
21379 Draw dots instead of lines.
21383 Set scale used for displaying graticule.
21390 Default is digital.
21393 Set background opacity.
21397 Set tint for output.
21398 Only used with lowpass filter and when display is not overlay and input
21399 pixel formats are not RGB.
21402 @section weave, doubleweave
21404 The @code{weave} takes a field-based video input and join
21405 each two sequential fields into single frame, producing a new double
21406 height clip with half the frame rate and half the frame count.
21408 The @code{doubleweave} works same as @code{weave} but without
21409 halving frame rate and frame count.
21411 It accepts the following option:
21415 Set first field. Available values are:
21419 Set the frame as top-field-first.
21422 Set the frame as bottom-field-first.
21426 @subsection Examples
21430 Interlace video using @ref{select} and @ref{separatefields} filter:
21432 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21437 Apply the xBR high-quality magnification filter which is designed for pixel
21438 art. It follows a set of edge-detection rules, see
21439 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21441 It accepts the following option:
21445 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21446 @code{3xBR} and @code{4} for @code{4xBR}.
21447 Default is @code{3}.
21452 Apply cross fade from one input video stream to another input video stream.
21453 The cross fade is applied for specified duration.
21455 The filter accepts the following options:
21459 Set one of available transition effects:
21507 Default transition effect is fade.
21510 Set cross fade duration in seconds.
21511 Default duration is 1 second.
21514 Set cross fade start relative to first input stream in seconds.
21515 Default offset is 0.
21518 Set expression for custom transition effect.
21520 The expressions can use the following variables and functions:
21525 The coordinates of the current sample.
21529 The width and height of the image.
21532 Progress of transition effect.
21535 Currently processed plane.
21538 Return value of first input at current location and plane.
21541 Return value of second input at current location and plane.
21547 Return the value of the pixel at location (@var{x},@var{y}) of the
21548 first/second/third/fourth component of first input.
21554 Return the value of the pixel at location (@var{x},@var{y}) of the
21555 first/second/third/fourth component of second input.
21559 @subsection Examples
21563 Cross fade from one input video to another input video, with fade transition and duration of transition
21564 of 2 seconds starting at offset of 5 seconds:
21566 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21571 Pick median pixels from several input videos.
21573 The filter accepts the following options:
21577 Set number of inputs.
21578 Default is 3. Allowed range is from 3 to 255.
21579 If number of inputs is even number, than result will be mean value between two median values.
21582 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21585 Set median percentile. Default value is @code{0.5}.
21586 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21587 minimum values, and @code{1} maximum values.
21590 @subsection Commands
21592 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21595 Stack video inputs into custom layout.
21597 All streams must be of same pixel format.
21599 The filter accepts the following options:
21603 Set number of input streams. Default is 2.
21606 Specify layout of inputs.
21607 This option requires the desired layout configuration to be explicitly set by the user.
21608 This sets position of each video input in output. Each input
21609 is separated by '|'.
21610 The first number represents the column, and the second number represents the row.
21611 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21612 where X is video input from which to take width or height.
21613 Multiple values can be used when separated by '+'. In such
21614 case values are summed together.
21616 Note that if inputs are of different sizes gaps may appear, as not all of
21617 the output video frame will be filled. Similarly, videos can overlap each
21618 other if their position doesn't leave enough space for the full frame of
21621 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21622 a layout must be set by the user.
21625 If set to 1, force the output to terminate when the shortest input
21626 terminates. Default value is 0.
21629 If set to valid color, all unused pixels will be filled with that color.
21630 By default fill is set to none, so it is disabled.
21633 @subsection Examples
21637 Display 4 inputs into 2x2 grid.
21641 input1(0, 0) | input3(w0, 0)
21642 input2(0, h0) | input4(w0, h0)
21646 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21649 Note that if inputs are of different sizes, gaps or overlaps may occur.
21652 Display 4 inputs into 1x4 grid.
21659 input4(0, h0+h1+h2)
21663 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21666 Note that if inputs are of different widths, unused space will appear.
21669 Display 9 inputs into 3x3 grid.
21673 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21674 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21675 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21679 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
21682 Note that if inputs are of different sizes, gaps or overlaps may occur.
21685 Display 16 inputs into 4x4 grid.
21689 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
21690 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
21691 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
21692 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
21696 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|
21697 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
21700 Note that if inputs are of different sizes, gaps or overlaps may occur.
21707 Deinterlace the input video ("yadif" means "yet another deinterlacing
21710 It accepts the following parameters:
21716 The interlacing mode to adopt. It accepts one of the following values:
21719 @item 0, send_frame
21720 Output one frame for each frame.
21721 @item 1, send_field
21722 Output one frame for each field.
21723 @item 2, send_frame_nospatial
21724 Like @code{send_frame}, but it skips the spatial interlacing check.
21725 @item 3, send_field_nospatial
21726 Like @code{send_field}, but it skips the spatial interlacing check.
21729 The default value is @code{send_frame}.
21732 The picture field parity assumed for the input interlaced video. It accepts one
21733 of the following values:
21737 Assume the top field is first.
21739 Assume the bottom field is first.
21741 Enable automatic detection of field parity.
21744 The default value is @code{auto}.
21745 If the interlacing is unknown or the decoder does not export this information,
21746 top field first will be assumed.
21749 Specify which frames to deinterlace. Accepts one of the following
21754 Deinterlace all frames.
21755 @item 1, interlaced
21756 Only deinterlace frames marked as interlaced.
21759 The default value is @code{all}.
21762 @section yadif_cuda
21764 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
21765 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
21768 It accepts the following parameters:
21774 The interlacing mode to adopt. It accepts one of the following values:
21777 @item 0, send_frame
21778 Output one frame for each frame.
21779 @item 1, send_field
21780 Output one frame for each field.
21781 @item 2, send_frame_nospatial
21782 Like @code{send_frame}, but it skips the spatial interlacing check.
21783 @item 3, send_field_nospatial
21784 Like @code{send_field}, but it skips the spatial interlacing check.
21787 The default value is @code{send_frame}.
21790 The picture field parity assumed for the input interlaced video. It accepts one
21791 of the following values:
21795 Assume the top field is first.
21797 Assume the bottom field is first.
21799 Enable automatic detection of field parity.
21802 The default value is @code{auto}.
21803 If the interlacing is unknown or the decoder does not export this information,
21804 top field first will be assumed.
21807 Specify which frames to deinterlace. Accepts one of the following
21812 Deinterlace all frames.
21813 @item 1, interlaced
21814 Only deinterlace frames marked as interlaced.
21817 The default value is @code{all}.
21822 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
21823 The algorithm is described in
21824 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
21826 It accepts the following parameters:
21830 Set the window radius. Default value is 3.
21833 Set which planes to filter. Default is only the first plane.
21836 Set blur strength. Default value is 128.
21839 @subsection Commands
21840 This filter supports same @ref{commands} as options.
21844 Apply Zoom & Pan effect.
21846 This filter accepts the following options:
21850 Set the zoom expression. Range is 1-10. Default is 1.
21854 Set the x and y expression. Default is 0.
21857 Set the duration expression in number of frames.
21858 This sets for how many number of frames effect will last for
21859 single input image.
21862 Set the output image size, default is 'hd720'.
21865 Set the output frame rate, default is '25'.
21868 Each expression can contain the following constants:
21887 Output frame count.
21890 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
21892 @item out_time, time, ot
21893 The output timestamp expressed in seconds.
21897 Last calculated 'x' and 'y' position from 'x' and 'y' expression
21898 for current input frame.
21902 'x' and 'y' of last output frame of previous input frame or 0 when there was
21903 not yet such frame (first input frame).
21906 Last calculated zoom from 'z' expression for current input frame.
21909 Last calculated zoom of last output frame of previous input frame.
21912 Number of output frames for current input frame. Calculated from 'd' expression
21913 for each input frame.
21916 number of output frames created for previous input frame
21919 Rational number: input width / input height
21922 sample aspect ratio
21925 display aspect ratio
21929 @subsection Examples
21933 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
21935 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
21939 Zoom in up to 1.5x and pan always at center of picture:
21941 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
21945 Same as above but without pausing:
21947 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
21951 Zoom in 2x into center of picture only for the first second of the input video:
21953 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
21960 Scale (resize) the input video, using the z.lib library:
21961 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
21962 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
21964 The zscale filter forces the output display aspect ratio to be the same
21965 as the input, by changing the output sample aspect ratio.
21967 If the input image format is different from the format requested by
21968 the next filter, the zscale filter will convert the input to the
21971 @subsection Options
21972 The filter accepts the following options.
21977 Set the output video dimension expression. Default value is the input
21980 If the @var{width} or @var{w} value is 0, the input width is used for
21981 the output. If the @var{height} or @var{h} value is 0, the input height
21982 is used for the output.
21984 If one and only one of the values is -n with n >= 1, the zscale filter
21985 will use a value that maintains the aspect ratio of the input image,
21986 calculated from the other specified dimension. After that it will,
21987 however, make sure that the calculated dimension is divisible by n and
21988 adjust the value if necessary.
21990 If both values are -n with n >= 1, the behavior will be identical to
21991 both values being set to 0 as previously detailed.
21993 See below for the list of accepted constants for use in the dimension
21997 Set the video size. For the syntax of this option, check the
21998 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22001 Set the dither type.
22003 Possible values are:
22008 @item error_diffusion
22014 Set the resize filter type.
22016 Possible values are:
22026 Default is bilinear.
22029 Set the color range.
22031 Possible values are:
22038 Default is same as input.
22041 Set the color primaries.
22043 Possible values are:
22053 Default is same as input.
22056 Set the transfer characteristics.
22058 Possible values are:
22072 Default is same as input.
22075 Set the colorspace matrix.
22077 Possible value are:
22088 Default is same as input.
22091 Set the input color range.
22093 Possible values are:
22100 Default is same as input.
22102 @item primariesin, pin
22103 Set the input color primaries.
22105 Possible values are:
22115 Default is same as input.
22117 @item transferin, tin
22118 Set the input transfer characteristics.
22120 Possible values are:
22131 Default is same as input.
22133 @item matrixin, min
22134 Set the input colorspace matrix.
22136 Possible value are:
22148 Set the output chroma location.
22150 Possible values are:
22161 @item chromalin, cin
22162 Set the input chroma location.
22164 Possible values are:
22176 Set the nominal peak luminance.
22179 The values of the @option{w} and @option{h} options are expressions
22180 containing the following constants:
22185 The input width and height
22189 These are the same as @var{in_w} and @var{in_h}.
22193 The output (scaled) width and height
22197 These are the same as @var{out_w} and @var{out_h}
22200 The same as @var{iw} / @var{ih}
22203 input sample aspect ratio
22206 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22210 horizontal and vertical input chroma subsample values. For example for the
22211 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22215 horizontal and vertical output chroma subsample values. For example for the
22216 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22219 @subsection Commands
22221 This filter supports the following commands:
22225 Set the output video dimension expression.
22226 The command accepts the same syntax of the corresponding option.
22228 If the specified expression is not valid, it is kept at its current
22232 @c man end VIDEO FILTERS
22234 @chapter OpenCL Video Filters
22235 @c man begin OPENCL VIDEO FILTERS
22237 Below is a description of the currently available OpenCL video filters.
22239 To enable compilation of these filters you need to configure FFmpeg with
22240 @code{--enable-opencl}.
22242 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22245 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22246 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22247 given device parameters.
22249 @item -filter_hw_device @var{name}
22250 Pass the hardware device called @var{name} to all filters in any filter graph.
22254 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22258 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22260 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22264 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.
22266 @section avgblur_opencl
22268 Apply average blur filter.
22270 The filter accepts the following options:
22274 Set horizontal radius size.
22275 Range is @code{[1, 1024]} and default value is @code{1}.
22278 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22281 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22284 @subsection Example
22288 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.
22290 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22294 @section boxblur_opencl
22296 Apply a boxblur algorithm to the input video.
22298 It accepts the following parameters:
22302 @item luma_radius, lr
22303 @item luma_power, lp
22304 @item chroma_radius, cr
22305 @item chroma_power, cp
22306 @item alpha_radius, ar
22307 @item alpha_power, ap
22311 A description of the accepted options follows.
22314 @item luma_radius, lr
22315 @item chroma_radius, cr
22316 @item alpha_radius, ar
22317 Set an expression for the box radius in pixels used for blurring the
22318 corresponding input plane.
22320 The radius value must be a non-negative number, and must not be
22321 greater than the value of the expression @code{min(w,h)/2} for the
22322 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22325 Default value for @option{luma_radius} is "2". If not specified,
22326 @option{chroma_radius} and @option{alpha_radius} default to the
22327 corresponding value set for @option{luma_radius}.
22329 The expressions can contain the following constants:
22333 The input width and height in pixels.
22337 The input chroma image width and height in pixels.
22341 The horizontal and vertical chroma subsample values. For example, for the
22342 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22345 @item luma_power, lp
22346 @item chroma_power, cp
22347 @item alpha_power, ap
22348 Specify how many times the boxblur filter is applied to the
22349 corresponding plane.
22351 Default value for @option{luma_power} is 2. If not specified,
22352 @option{chroma_power} and @option{alpha_power} default to the
22353 corresponding value set for @option{luma_power}.
22355 A value of 0 will disable the effect.
22358 @subsection Examples
22360 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.
22364 Apply a boxblur filter with the luma, chroma, and alpha radius
22365 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.
22367 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22368 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22372 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.
22374 For the luma plane, a 2x2 box radius will be run once.
22376 For the chroma plane, a 4x4 box radius will be run 5 times.
22378 For the alpha plane, a 3x3 box radius will be run 7 times.
22380 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22384 @section colorkey_opencl
22385 RGB colorspace color keying.
22387 The filter accepts the following options:
22391 The color which will be replaced with transparency.
22394 Similarity percentage with the key color.
22396 0.01 matches only the exact key color, while 1.0 matches everything.
22401 0.0 makes pixels either fully transparent, or not transparent at all.
22403 Higher values result in semi-transparent pixels, with a higher transparency
22404 the more similar the pixels color is to the key color.
22407 @subsection Examples
22411 Make every semi-green pixel in the input transparent with some slight blending:
22413 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22417 @section convolution_opencl
22419 Apply convolution of 3x3, 5x5, 7x7 matrix.
22421 The filter accepts the following options:
22428 Set matrix for each plane.
22429 Matrix is sequence of 9, 25 or 49 signed numbers.
22430 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22436 Set multiplier for calculated value for each plane.
22437 If unset or 0, it will be sum of all matrix elements.
22438 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22444 Set bias for each plane. This value is added to the result of the multiplication.
22445 Useful for making the overall image brighter or darker.
22446 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22450 @subsection Examples
22456 -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
22462 -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
22466 Apply edge enhance:
22468 -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
22474 -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
22478 Apply laplacian edge detector which includes diagonals:
22480 -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
22486 -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
22490 @section erosion_opencl
22492 Apply erosion effect to the video.
22494 This filter replaces the pixel by the local(3x3) minimum.
22496 It accepts the following options:
22503 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22504 If @code{0}, plane will remain unchanged.
22507 Flag which specifies the pixel to refer to.
22508 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22510 Flags to local 3x3 coordinates region centered on @code{x}:
22519 @subsection Example
22523 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.
22525 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22529 @section deshake_opencl
22530 Feature-point based video stabilization filter.
22532 The filter accepts the following options:
22536 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22539 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22541 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22543 Viewing point matches in the output video is only supported for RGB input.
22545 Defaults to @code{0}.
22547 @item adaptive_crop
22548 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22550 Defaults to @code{1}.
22552 @item refine_features
22553 Whether or not feature points should be refined at a sub-pixel level.
22555 This can be turned off for a slight performance gain at the cost of precision.
22557 Defaults to @code{1}.
22559 @item smooth_strength
22560 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22562 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22564 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22566 Defaults to @code{0.0}.
22568 @item smooth_window_multiplier
22569 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22571 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22573 Acceptable values range from @code{0.1} to @code{10.0}.
22575 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22576 potentially improving smoothness, but also increase latency and memory usage.
22578 Defaults to @code{2.0}.
22582 @subsection Examples
22586 Stabilize a video with a fixed, medium smoothing strength:
22588 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22592 Stabilize a video with debugging (both in console and in rendered video):
22594 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22598 @section dilation_opencl
22600 Apply dilation effect to the video.
22602 This filter replaces the pixel by the local(3x3) maximum.
22604 It accepts the following options:
22611 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22612 If @code{0}, plane will remain unchanged.
22615 Flag which specifies the pixel to refer to.
22616 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22618 Flags to local 3x3 coordinates region centered on @code{x}:
22627 @subsection Example
22631 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.
22633 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22637 @section nlmeans_opencl
22639 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22641 @section overlay_opencl
22643 Overlay one video on top of another.
22645 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22646 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22648 The filter accepts the following options:
22653 Set the x coordinate of the overlaid video on the main video.
22654 Default value is @code{0}.
22657 Set the y coordinate of the overlaid video on the main video.
22658 Default value is @code{0}.
22662 @subsection Examples
22666 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22668 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22671 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22673 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22678 @section pad_opencl
22680 Add paddings to the input image, and place the original input at the
22681 provided @var{x}, @var{y} coordinates.
22683 It accepts the following options:
22688 Specify an expression for the size of the output image with the
22689 paddings added. If the value for @var{width} or @var{height} is 0, the
22690 corresponding input size is used for the output.
22692 The @var{width} expression can reference the value set by the
22693 @var{height} expression, and vice versa.
22695 The default value of @var{width} and @var{height} is 0.
22699 Specify the offsets to place the input image at within the padded area,
22700 with respect to the top/left border of the output image.
22702 The @var{x} expression can reference the value set by the @var{y}
22703 expression, and vice versa.
22705 The default value of @var{x} and @var{y} is 0.
22707 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
22708 so the input image is centered on the padded area.
22711 Specify the color of the padded area. For the syntax of this option,
22712 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
22713 manual,ffmpeg-utils}.
22716 Pad to an aspect instead to a resolution.
22719 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
22720 options are expressions containing the following constants:
22725 The input video width and height.
22729 These are the same as @var{in_w} and @var{in_h}.
22733 The output width and height (the size of the padded area), as
22734 specified by the @var{width} and @var{height} expressions.
22738 These are the same as @var{out_w} and @var{out_h}.
22742 The x and y offsets as specified by the @var{x} and @var{y}
22743 expressions, or NAN if not yet specified.
22746 same as @var{iw} / @var{ih}
22749 input sample aspect ratio
22752 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
22755 @section prewitt_opencl
22757 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
22759 The filter accepts the following option:
22763 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22766 Set value which will be multiplied with filtered result.
22767 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22770 Set value which will be added to filtered result.
22771 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22774 @subsection Example
22778 Apply the Prewitt operator with scale set to 2 and delta set to 10.
22780 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
22784 @anchor{program_opencl}
22785 @section program_opencl
22787 Filter video using an OpenCL program.
22792 OpenCL program source file.
22795 Kernel name in program.
22798 Number of inputs to the filter. Defaults to 1.
22801 Size of output frames. Defaults to the same as the first input.
22805 The @code{program_opencl} filter also supports the @ref{framesync} options.
22807 The program source file must contain a kernel function with the given name,
22808 which will be run once for each plane of the output. Each run on a plane
22809 gets enqueued as a separate 2D global NDRange with one work-item for each
22810 pixel to be generated. The global ID offset for each work-item is therefore
22811 the coordinates of a pixel in the destination image.
22813 The kernel function needs to take the following arguments:
22816 Destination image, @var{__write_only image2d_t}.
22818 This image will become the output; the kernel should write all of it.
22820 Frame index, @var{unsigned int}.
22822 This is a counter starting from zero and increasing by one for each frame.
22824 Source images, @var{__read_only image2d_t}.
22826 These are the most recent images on each input. The kernel may read from
22827 them to generate the output, but they can't be written to.
22834 Copy the input to the output (output must be the same size as the input).
22836 __kernel void copy(__write_only image2d_t destination,
22837 unsigned int index,
22838 __read_only image2d_t source)
22840 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
22842 int2 location = (int2)(get_global_id(0), get_global_id(1));
22844 float4 value = read_imagef(source, sampler, location);
22846 write_imagef(destination, location, value);
22851 Apply a simple transformation, rotating the input by an amount increasing
22852 with the index counter. Pixel values are linearly interpolated by the
22853 sampler, and the output need not have the same dimensions as the input.
22855 __kernel void rotate_image(__write_only image2d_t dst,
22856 unsigned int index,
22857 __read_only image2d_t src)
22859 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22860 CLK_FILTER_LINEAR);
22862 float angle = (float)index / 100.0f;
22864 float2 dst_dim = convert_float2(get_image_dim(dst));
22865 float2 src_dim = convert_float2(get_image_dim(src));
22867 float2 dst_cen = dst_dim / 2.0f;
22868 float2 src_cen = src_dim / 2.0f;
22870 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
22872 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
22874 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
22875 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
22877 src_pos = src_pos * src_dim / dst_dim;
22879 float2 src_loc = src_pos + src_cen;
22881 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
22882 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
22883 write_imagef(dst, dst_loc, 0.5f);
22885 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
22890 Blend two inputs together, with the amount of each input used varying
22891 with the index counter.
22893 __kernel void blend_images(__write_only image2d_t dst,
22894 unsigned int index,
22895 __read_only image2d_t src1,
22896 __read_only image2d_t src2)
22898 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22899 CLK_FILTER_LINEAR);
22901 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
22903 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
22904 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
22905 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
22907 float4 val1 = read_imagef(src1, sampler, src1_loc);
22908 float4 val2 = read_imagef(src2, sampler, src2_loc);
22910 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
22916 @section roberts_opencl
22917 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
22919 The filter accepts the following option:
22923 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22926 Set value which will be multiplied with filtered result.
22927 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22930 Set value which will be added to filtered result.
22931 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22934 @subsection Example
22938 Apply the Roberts cross operator with scale set to 2 and delta set to 10
22940 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
22944 @section sobel_opencl
22946 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
22948 The filter accepts the following option:
22952 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22955 Set value which will be multiplied with filtered result.
22956 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22959 Set value which will be added to filtered result.
22960 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22963 @subsection Example
22967 Apply sobel operator with scale set to 2 and delta set to 10
22969 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
22973 @section tonemap_opencl
22975 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
22977 It accepts the following parameters:
22981 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
22984 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
22987 Apply desaturation for highlights that exceed this level of brightness. The
22988 higher the parameter, the more color information will be preserved. This
22989 setting helps prevent unnaturally blown-out colors for super-highlights, by
22990 (smoothly) turning into white instead. This makes images feel more natural,
22991 at the cost of reducing information about out-of-range colors.
22993 The default value is 0.5, and the algorithm here is a little different from
22994 the cpu version tonemap currently. A setting of 0.0 disables this option.
22997 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
22998 is used to detect whether the scene has changed or not. If the distance between
22999 the current frame average brightness and the current running average exceeds
23000 a threshold value, we would re-calculate scene average and peak brightness.
23001 The default value is 0.2.
23004 Specify the output pixel format.
23006 Currently supported formats are:
23013 Set the output color range.
23015 Possible values are:
23021 Default is same as input.
23024 Set the output color primaries.
23026 Possible values are:
23032 Default is same as input.
23035 Set the output transfer characteristics.
23037 Possible values are:
23046 Set the output colorspace matrix.
23048 Possible value are:
23054 Default is same as input.
23058 @subsection Example
23062 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23064 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23068 @section unsharp_opencl
23070 Sharpen or blur the input video.
23072 It accepts the following parameters:
23075 @item luma_msize_x, lx
23076 Set the luma matrix horizontal size.
23077 Range is @code{[1, 23]} and default value is @code{5}.
23079 @item luma_msize_y, ly
23080 Set the luma matrix vertical size.
23081 Range is @code{[1, 23]} and default value is @code{5}.
23083 @item luma_amount, la
23084 Set the luma effect strength.
23085 Range is @code{[-10, 10]} and default value is @code{1.0}.
23087 Negative values will blur the input video, while positive values will
23088 sharpen it, a value of zero will disable the effect.
23090 @item chroma_msize_x, cx
23091 Set the chroma matrix horizontal size.
23092 Range is @code{[1, 23]} and default value is @code{5}.
23094 @item chroma_msize_y, cy
23095 Set the chroma matrix vertical size.
23096 Range is @code{[1, 23]} and default value is @code{5}.
23098 @item chroma_amount, ca
23099 Set the chroma effect strength.
23100 Range is @code{[-10, 10]} and default value is @code{0.0}.
23102 Negative values will blur the input video, while positive values will
23103 sharpen it, a value of zero will disable the effect.
23107 All parameters are optional and default to the equivalent of the
23108 string '5:5:1.0:5:5:0.0'.
23110 @subsection Examples
23114 Apply strong luma sharpen effect:
23116 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23120 Apply a strong blur of both luma and chroma parameters:
23122 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23126 @section xfade_opencl
23128 Cross fade two videos with custom transition effect by using OpenCL.
23130 It accepts the following options:
23134 Set one of possible transition effects.
23138 Select custom transition effect, the actual transition description
23139 will be picked from source and kernel options.
23151 Default transition is fade.
23155 OpenCL program source file for custom transition.
23158 Set name of kernel to use for custom transition from program source file.
23161 Set duration of video transition.
23164 Set time of start of transition relative to first video.
23167 The program source file must contain a kernel function with the given name,
23168 which will be run once for each plane of the output. Each run on a plane
23169 gets enqueued as a separate 2D global NDRange with one work-item for each
23170 pixel to be generated. The global ID offset for each work-item is therefore
23171 the coordinates of a pixel in the destination image.
23173 The kernel function needs to take the following arguments:
23176 Destination image, @var{__write_only image2d_t}.
23178 This image will become the output; the kernel should write all of it.
23181 First Source image, @var{__read_only image2d_t}.
23182 Second Source image, @var{__read_only image2d_t}.
23184 These are the most recent images on each input. The kernel may read from
23185 them to generate the output, but they can't be written to.
23188 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23195 Apply dots curtain transition effect:
23197 __kernel void blend_images(__write_only image2d_t dst,
23198 __read_only image2d_t src1,
23199 __read_only image2d_t src2,
23202 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23203 CLK_FILTER_LINEAR);
23204 int2 p = (int2)(get_global_id(0), get_global_id(1));
23205 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23206 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23209 float2 dots = (float2)(20.0, 20.0);
23210 float2 center = (float2)(0,0);
23213 float4 val1 = read_imagef(src1, sampler, p);
23214 float4 val2 = read_imagef(src2, sampler, p);
23215 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23217 write_imagef(dst, p, next ? val1 : val2);
23223 @c man end OPENCL VIDEO FILTERS
23225 @chapter VAAPI Video Filters
23226 @c man begin VAAPI VIDEO FILTERS
23228 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23230 To enable compilation of these filters you need to configure FFmpeg with
23231 @code{--enable-vaapi}.
23233 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}
23235 @section tonemap_vaapi
23237 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23238 It maps the dynamic range of HDR10 content to the SDR content.
23239 It currently only accepts HDR10 as input.
23241 It accepts the following parameters:
23245 Specify the output pixel format.
23247 Currently supported formats are:
23256 Set the output color primaries.
23258 Default is same as input.
23261 Set the output transfer characteristics.
23266 Set the output colorspace matrix.
23268 Default is same as input.
23272 @subsection Example
23276 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23278 tonemap_vaapi=format=p010:t=bt2020-10
23282 @c man end VAAPI VIDEO FILTERS
23284 @chapter Video Sources
23285 @c man begin VIDEO SOURCES
23287 Below is a description of the currently available video sources.
23291 Buffer video frames, and make them available to the filter chain.
23293 This source is mainly intended for a programmatic use, in particular
23294 through the interface defined in @file{libavfilter/buffersrc.h}.
23296 It accepts the following parameters:
23301 Specify the size (width and height) of the buffered video frames. For the
23302 syntax of this option, check the
23303 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23306 The input video width.
23309 The input video height.
23312 A string representing the pixel format of the buffered video frames.
23313 It may be a number corresponding to a pixel format, or a pixel format
23317 Specify the timebase assumed by the timestamps of the buffered frames.
23320 Specify the frame rate expected for the video stream.
23322 @item pixel_aspect, sar
23323 The sample (pixel) aspect ratio of the input video.
23326 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23327 to the filtergraph description to specify swscale flags for automatically
23328 inserted scalers. See @ref{Filtergraph syntax}.
23330 @item hw_frames_ctx
23331 When using a hardware pixel format, this should be a reference to an
23332 AVHWFramesContext describing input frames.
23337 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23340 will instruct the source to accept video frames with size 320x240 and
23341 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23342 square pixels (1:1 sample aspect ratio).
23343 Since the pixel format with name "yuv410p" corresponds to the number 6
23344 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23345 this example corresponds to:
23347 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23350 Alternatively, the options can be specified as a flat string, but this
23351 syntax is deprecated:
23353 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23357 Create a pattern generated by an elementary cellular automaton.
23359 The initial state of the cellular automaton can be defined through the
23360 @option{filename} and @option{pattern} options. If such options are
23361 not specified an initial state is created randomly.
23363 At each new frame a new row in the video is filled with the result of
23364 the cellular automaton next generation. The behavior when the whole
23365 frame is filled is defined by the @option{scroll} option.
23367 This source accepts the following options:
23371 Read the initial cellular automaton state, i.e. the starting row, from
23372 the specified file.
23373 In the file, each non-whitespace character is considered an alive
23374 cell, a newline will terminate the row, and further characters in the
23375 file will be ignored.
23378 Read the initial cellular automaton state, i.e. the starting row, from
23379 the specified string.
23381 Each non-whitespace character in the string is considered an alive
23382 cell, a newline will terminate the row, and further characters in the
23383 string will be ignored.
23386 Set the video rate, that is the number of frames generated per second.
23389 @item random_fill_ratio, ratio
23390 Set the random fill ratio for the initial cellular automaton row. It
23391 is a floating point number value ranging from 0 to 1, defaults to
23394 This option is ignored when a file or a pattern is specified.
23396 @item random_seed, seed
23397 Set the seed for filling randomly the initial row, must be an integer
23398 included between 0 and UINT32_MAX. If not specified, or if explicitly
23399 set to -1, the filter will try to use a good random seed on a best
23403 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23404 Default value is 110.
23407 Set the size of the output video. For the syntax of this option, check the
23408 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23410 If @option{filename} or @option{pattern} is specified, the size is set
23411 by default to the width of the specified initial state row, and the
23412 height is set to @var{width} * PHI.
23414 If @option{size} is set, it must contain the width of the specified
23415 pattern string, and the specified pattern will be centered in the
23418 If a filename or a pattern string is not specified, the size value
23419 defaults to "320x518" (used for a randomly generated initial state).
23422 If set to 1, scroll the output upward when all the rows in the output
23423 have been already filled. If set to 0, the new generated row will be
23424 written over the top row just after the bottom row is filled.
23427 @item start_full, full
23428 If set to 1, completely fill the output with generated rows before
23429 outputting the first frame.
23430 This is the default behavior, for disabling set the value to 0.
23433 If set to 1, stitch the left and right row edges together.
23434 This is the default behavior, for disabling set the value to 0.
23437 @subsection Examples
23441 Read the initial state from @file{pattern}, and specify an output of
23444 cellauto=f=pattern:s=200x400
23448 Generate a random initial row with a width of 200 cells, with a fill
23451 cellauto=ratio=2/3:s=200x200
23455 Create a pattern generated by rule 18 starting by a single alive cell
23456 centered on an initial row with width 100:
23458 cellauto=p=@@:s=100x400:full=0:rule=18
23462 Specify a more elaborated initial pattern:
23464 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23469 @anchor{coreimagesrc}
23470 @section coreimagesrc
23471 Video source generated on GPU using Apple's CoreImage API on OSX.
23473 This video source is a specialized version of the @ref{coreimage} video filter.
23474 Use a core image generator at the beginning of the applied filterchain to
23475 generate the content.
23477 The coreimagesrc video source accepts the following options:
23479 @item list_generators
23480 List all available generators along with all their respective options as well as
23481 possible minimum and maximum values along with the default values.
23483 list_generators=true
23487 Specify the size of the sourced video. For the syntax of this option, check the
23488 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23489 The default value is @code{320x240}.
23492 Specify the frame rate of the sourced video, as the number of frames
23493 generated per second. It has to be a string in the format
23494 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23495 number or a valid video frame rate abbreviation. The default value is
23499 Set the sample aspect ratio of the sourced video.
23502 Set the duration of the sourced video. See
23503 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23504 for the accepted syntax.
23506 If not specified, or the expressed duration is negative, the video is
23507 supposed to be generated forever.
23510 Additionally, all options of the @ref{coreimage} video filter are accepted.
23511 A complete filterchain can be used for further processing of the
23512 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23513 and examples for details.
23515 @subsection Examples
23520 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23521 given as complete and escaped command-line for Apple's standard bash shell:
23523 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23525 This example is equivalent to the QRCode example of @ref{coreimage} without the
23526 need for a nullsrc video source.
23531 Generate several gradients.
23535 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23536 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23539 Set frame rate, expressed as number of frames per second. Default
23542 @item c0, c1, c2, c3, c4, c5, c6, c7
23543 Set 8 colors. Default values for colors is to pick random one.
23545 @item x0, y0, y0, y1
23546 Set gradient line source and destination points. If negative or out of range, random ones
23550 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23553 Set seed for picking gradient line points.
23556 Set the duration of the sourced video. See
23557 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23558 for the accepted syntax.
23560 If not specified, or the expressed duration is negative, the video is
23561 supposed to be generated forever.
23564 Set speed of gradients rotation.
23568 @section mandelbrot
23570 Generate a Mandelbrot set fractal, and progressively zoom towards the
23571 point specified with @var{start_x} and @var{start_y}.
23573 This source accepts the following options:
23578 Set the terminal pts value. Default value is 400.
23581 Set the terminal scale value.
23582 Must be a floating point value. Default value is 0.3.
23585 Set the inner coloring mode, that is the algorithm used to draw the
23586 Mandelbrot fractal internal region.
23588 It shall assume one of the following values:
23593 Show time until convergence.
23595 Set color based on point closest to the origin of the iterations.
23600 Default value is @var{mincol}.
23603 Set the bailout value. Default value is 10.0.
23606 Set the maximum of iterations performed by the rendering
23607 algorithm. Default value is 7189.
23610 Set outer coloring mode.
23611 It shall assume one of following values:
23613 @item iteration_count
23614 Set iteration count mode.
23615 @item normalized_iteration_count
23616 set normalized iteration count mode.
23618 Default value is @var{normalized_iteration_count}.
23621 Set frame rate, expressed as number of frames per second. Default
23625 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23626 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23629 Set the initial scale value. Default value is 3.0.
23632 Set the initial x position. Must be a floating point value between
23633 -100 and 100. Default value is -0.743643887037158704752191506114774.
23636 Set the initial y position. Must be a floating point value between
23637 -100 and 100. Default value is -0.131825904205311970493132056385139.
23642 Generate various test patterns, as generated by the MPlayer test filter.
23644 The size of the generated video is fixed, and is 256x256.
23645 This source is useful in particular for testing encoding features.
23647 This source accepts the following options:
23652 Specify the frame rate of the sourced video, as the number of frames
23653 generated per second. It has to be a string in the format
23654 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23655 number or a valid video frame rate abbreviation. The default value is
23659 Set the duration of the sourced video. See
23660 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23661 for the accepted syntax.
23663 If not specified, or the expressed duration is negative, the video is
23664 supposed to be generated forever.
23668 Set the number or the name of the test to perform. Supported tests are:
23682 @item max_frames, m
23683 Set the maximum number of frames generated for each test, default value is 30.
23687 Default value is "all", which will cycle through the list of all tests.
23692 mptestsrc=t=dc_luma
23695 will generate a "dc_luma" test pattern.
23697 @section frei0r_src
23699 Provide a frei0r source.
23701 To enable compilation of this filter you need to install the frei0r
23702 header and configure FFmpeg with @code{--enable-frei0r}.
23704 This source accepts the following parameters:
23709 The size of the video to generate. For the syntax of this option, check the
23710 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23713 The framerate of the generated video. It may be a string of the form
23714 @var{num}/@var{den} or a frame rate abbreviation.
23717 The name to the frei0r source to load. For more information regarding frei0r and
23718 how to set the parameters, read the @ref{frei0r} section in the video filters
23721 @item filter_params
23722 A '|'-separated list of parameters to pass to the frei0r source.
23726 For example, to generate a frei0r partik0l source with size 200x200
23727 and frame rate 10 which is overlaid on the overlay filter main input:
23729 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
23734 Generate a life pattern.
23736 This source is based on a generalization of John Conway's life game.
23738 The sourced input represents a life grid, each pixel represents a cell
23739 which can be in one of two possible states, alive or dead. Every cell
23740 interacts with its eight neighbours, which are the cells that are
23741 horizontally, vertically, or diagonally adjacent.
23743 At each interaction the grid evolves according to the adopted rule,
23744 which specifies the number of neighbor alive cells which will make a
23745 cell stay alive or born. The @option{rule} option allows one to specify
23748 This source accepts the following options:
23752 Set the file from which to read the initial grid state. In the file,
23753 each non-whitespace character is considered an alive cell, and newline
23754 is used to delimit the end of each row.
23756 If this option is not specified, the initial grid is generated
23760 Set the video rate, that is the number of frames generated per second.
23763 @item random_fill_ratio, ratio
23764 Set the random fill ratio for the initial random grid. It is a
23765 floating point number value ranging from 0 to 1, defaults to 1/PHI.
23766 It is ignored when a file is specified.
23768 @item random_seed, seed
23769 Set the seed for filling the initial random grid, must be an integer
23770 included between 0 and UINT32_MAX. If not specified, or if explicitly
23771 set to -1, the filter will try to use a good random seed on a best
23777 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
23778 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
23779 @var{NS} specifies the number of alive neighbor cells which make a
23780 live cell stay alive, and @var{NB} the number of alive neighbor cells
23781 which make a dead cell to become alive (i.e. to "born").
23782 "s" and "b" can be used in place of "S" and "B", respectively.
23784 Alternatively a rule can be specified by an 18-bits integer. The 9
23785 high order bits are used to encode the next cell state if it is alive
23786 for each number of neighbor alive cells, the low order bits specify
23787 the rule for "borning" new cells. Higher order bits encode for an
23788 higher number of neighbor cells.
23789 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
23790 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
23792 Default value is "S23/B3", which is the original Conway's game of life
23793 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
23794 cells, and will born a new cell if there are three alive cells around
23798 Set the size of the output video. For the syntax of this option, check the
23799 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23801 If @option{filename} is specified, the size is set by default to the
23802 same size of the input file. If @option{size} is set, it must contain
23803 the size specified in the input file, and the initial grid defined in
23804 that file is centered in the larger resulting area.
23806 If a filename is not specified, the size value defaults to "320x240"
23807 (used for a randomly generated initial grid).
23810 If set to 1, stitch the left and right grid edges together, and the
23811 top and bottom edges also. Defaults to 1.
23814 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
23815 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
23816 value from 0 to 255.
23819 Set the color of living (or new born) cells.
23822 Set the color of dead cells. If @option{mold} is set, this is the first color
23823 used to represent a dead cell.
23826 Set mold color, for definitely dead and moldy cells.
23828 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
23829 ffmpeg-utils manual,ffmpeg-utils}.
23832 @subsection Examples
23836 Read a grid from @file{pattern}, and center it on a grid of size
23839 life=f=pattern:s=300x300
23843 Generate a random grid of size 200x200, with a fill ratio of 2/3:
23845 life=ratio=2/3:s=200x200
23849 Specify a custom rule for evolving a randomly generated grid:
23855 Full example with slow death effect (mold) using @command{ffplay}:
23857 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
23864 @anchor{haldclutsrc}
23867 @anchor{pal100bars}
23868 @anchor{rgbtestsrc}
23870 @anchor{smptehdbars}
23873 @anchor{yuvtestsrc}
23874 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
23876 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
23878 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
23880 The @code{color} source provides an uniformly colored input.
23882 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
23883 @ref{haldclut} filter.
23885 The @code{nullsrc} source returns unprocessed video frames. It is
23886 mainly useful to be employed in analysis / debugging tools, or as the
23887 source for filters which ignore the input data.
23889 The @code{pal75bars} source generates a color bars pattern, based on
23890 EBU PAL recommendations with 75% color levels.
23892 The @code{pal100bars} source generates a color bars pattern, based on
23893 EBU PAL recommendations with 100% color levels.
23895 The @code{rgbtestsrc} source generates an RGB test pattern useful for
23896 detecting RGB vs BGR issues. You should see a red, green and blue
23897 stripe from top to bottom.
23899 The @code{smptebars} source generates a color bars pattern, based on
23900 the SMPTE Engineering Guideline EG 1-1990.
23902 The @code{smptehdbars} source generates a color bars pattern, based on
23903 the SMPTE RP 219-2002.
23905 The @code{testsrc} source generates a test video pattern, showing a
23906 color pattern, a scrolling gradient and a timestamp. This is mainly
23907 intended for testing purposes.
23909 The @code{testsrc2} source is similar to testsrc, but supports more
23910 pixel formats instead of just @code{rgb24}. This allows using it as an
23911 input for other tests without requiring a format conversion.
23913 The @code{yuvtestsrc} source generates an YUV test pattern. You should
23914 see a y, cb and cr stripe from top to bottom.
23916 The sources accept the following parameters:
23921 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
23922 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
23923 pixels to be used as identity matrix for 3D lookup tables. Each component is
23924 coded on a @code{1/(N*N)} scale.
23927 Specify the color of the source, only available in the @code{color}
23928 source. For the syntax of this option, check the
23929 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
23932 Specify the size of the sourced video. For the syntax of this option, check the
23933 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23934 The default value is @code{320x240}.
23936 This option is not available with the @code{allrgb}, @code{allyuv}, and
23937 @code{haldclutsrc} filters.
23940 Specify the frame rate of the sourced video, as the number of frames
23941 generated per second. It has to be a string in the format
23942 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23943 number or a valid video frame rate abbreviation. The default value is
23947 Set the duration of the sourced video. See
23948 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23949 for the accepted syntax.
23951 If not specified, or the expressed duration is negative, the video is
23952 supposed to be generated forever.
23954 Since the frame rate is used as time base, all frames including the last one
23955 will have their full duration. If the specified duration is not a multiple
23956 of the frame duration, it will be rounded up.
23959 Set the sample aspect ratio of the sourced video.
23962 Specify the alpha (opacity) of the background, only available in the
23963 @code{testsrc2} source. The value must be between 0 (fully transparent) and
23964 255 (fully opaque, the default).
23967 Set the number of decimals to show in the timestamp, only available in the
23968 @code{testsrc} source.
23970 The displayed timestamp value will correspond to the original
23971 timestamp value multiplied by the power of 10 of the specified
23972 value. Default value is 0.
23975 @subsection Examples
23979 Generate a video with a duration of 5.3 seconds, with size
23980 176x144 and a frame rate of 10 frames per second:
23982 testsrc=duration=5.3:size=qcif:rate=10
23986 The following graph description will generate a red source
23987 with an opacity of 0.2, with size "qcif" and a frame rate of 10
23990 color=c=red@@0.2:s=qcif:r=10
23994 If the input content is to be ignored, @code{nullsrc} can be used. The
23995 following command generates noise in the luminance plane by employing
23996 the @code{geq} filter:
23998 nullsrc=s=256x256, geq=random(1)*255:128:128
24002 @subsection Commands
24004 The @code{color} source supports the following commands:
24008 Set the color of the created image. Accepts the same syntax of the
24009 corresponding @option{color} option.
24014 Generate video using an OpenCL program.
24019 OpenCL program source file.
24022 Kernel name in program.
24025 Size of frames to generate. This must be set.
24028 Pixel format to use for the generated frames. This must be set.
24031 Number of frames generated every second. Default value is '25'.
24035 For details of how the program loading works, see the @ref{program_opencl}
24042 Generate a colour ramp by setting pixel values from the position of the pixel
24043 in the output image. (Note that this will work with all pixel formats, but
24044 the generated output will not be the same.)
24046 __kernel void ramp(__write_only image2d_t dst,
24047 unsigned int index)
24049 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24052 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24054 write_imagef(dst, loc, val);
24059 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24061 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24062 unsigned int index)
24064 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24066 float4 value = 0.0f;
24067 int x = loc.x + index;
24068 int y = loc.y + index;
24069 while (x > 0 || y > 0) {
24070 if (x % 3 == 1 && y % 3 == 1) {
24078 write_imagef(dst, loc, value);
24084 @section sierpinski
24086 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24088 This source accepts the following options:
24092 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24093 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24096 Set frame rate, expressed as number of frames per second. Default
24100 Set seed which is used for random panning.
24103 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24106 Set fractal type, can be default @code{carpet} or @code{triangle}.
24109 @c man end VIDEO SOURCES
24111 @chapter Video Sinks
24112 @c man begin VIDEO SINKS
24114 Below is a description of the currently available video sinks.
24116 @section buffersink
24118 Buffer video frames, and make them available to the end of the filter
24121 This sink is mainly intended for programmatic use, in particular
24122 through the interface defined in @file{libavfilter/buffersink.h}
24123 or the options system.
24125 It accepts a pointer to an AVBufferSinkContext structure, which
24126 defines the incoming buffers' formats, to be passed as the opaque
24127 parameter to @code{avfilter_init_filter} for initialization.
24131 Null video sink: do absolutely nothing with the input video. It is
24132 mainly useful as a template and for use in analysis / debugging
24135 @c man end VIDEO SINKS
24137 @chapter Multimedia Filters
24138 @c man begin MULTIMEDIA FILTERS
24140 Below is a description of the currently available multimedia filters.
24144 Convert input audio to a video output, displaying the audio bit scope.
24146 The filter accepts the following options:
24150 Set frame rate, expressed as number of frames per second. Default
24154 Specify the video size for the output. For the syntax of this option, check the
24155 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24156 Default value is @code{1024x256}.
24159 Specify list of colors separated by space or by '|' which will be used to
24160 draw channels. Unrecognized or missing colors will be replaced
24164 @section adrawgraph
24165 Draw a graph using input audio metadata.
24167 See @ref{drawgraph}
24169 @section agraphmonitor
24171 See @ref{graphmonitor}.
24173 @section ahistogram
24175 Convert input audio to a video output, displaying the volume histogram.
24177 The filter accepts the following options:
24181 Specify how histogram is calculated.
24183 It accepts the following values:
24186 Use single histogram for all channels.
24188 Use separate histogram for each channel.
24190 Default is @code{single}.
24193 Set frame rate, expressed as number of frames per second. Default
24197 Specify the video size for the output. For the syntax of this option, check the
24198 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24199 Default value is @code{hd720}.
24204 It accepts the following values:
24215 reverse logarithmic
24217 Default is @code{log}.
24220 Set amplitude scale.
24222 It accepts the following values:
24229 Default is @code{log}.
24232 Set how much frames to accumulate in histogram.
24233 Default is 1. Setting this to -1 accumulates all frames.
24236 Set histogram ratio of window height.
24239 Set sonogram sliding.
24241 It accepts the following values:
24244 replace old rows with new ones.
24246 scroll from top to bottom.
24248 Default is @code{replace}.
24251 @section aphasemeter
24253 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24254 representing mean phase of current audio frame. A video output can also be produced and is
24255 enabled by default. The audio is passed through as first output.
24257 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24258 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24259 and @code{1} means channels are in phase.
24261 The filter accepts the following options, all related to its video output:
24265 Set the output frame rate. Default value is @code{25}.
24268 Set the video size for the output. For the syntax of this option, check the
24269 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24270 Default value is @code{800x400}.
24275 Specify the red, green, blue contrast. Default values are @code{2},
24276 @code{7} and @code{1}.
24277 Allowed range is @code{[0, 255]}.
24280 Set color which will be used for drawing median phase. If color is
24281 @code{none} which is default, no median phase value will be drawn.
24284 Enable video output. Default is enabled.
24287 @subsection phasing detection
24289 The filter also detects out of phase and mono sequences in stereo streams.
24290 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24292 The filter accepts the following options for this detection:
24296 Enable mono and out of phase detection. Default is disabled.
24299 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24300 Allowed range is @code{[0, 1]}.
24303 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24304 Allowed range is @code{[90, 180]}.
24307 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24310 @subsection Examples
24314 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24316 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24320 @section avectorscope
24322 Convert input audio to a video output, representing the audio vector
24325 The filter is used to measure the difference between channels of stereo
24326 audio stream. A monaural signal, consisting of identical left and right
24327 signal, results in straight vertical line. Any stereo separation is visible
24328 as a deviation from this line, creating a Lissajous figure.
24329 If the straight (or deviation from it) but horizontal line appears this
24330 indicates that the left and right channels are out of phase.
24332 The filter accepts the following options:
24336 Set the vectorscope mode.
24338 Available values are:
24341 Lissajous rotated by 45 degrees.
24344 Same as above but not rotated.
24347 Shape resembling half of circle.
24350 Default value is @samp{lissajous}.
24353 Set the video size for the output. For the syntax of this option, check the
24354 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24355 Default value is @code{400x400}.
24358 Set the output frame rate. Default value is @code{25}.
24364 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24365 @code{160}, @code{80} and @code{255}.
24366 Allowed range is @code{[0, 255]}.
24372 Specify the red, green, blue and alpha fade. Default values are @code{15},
24373 @code{10}, @code{5} and @code{5}.
24374 Allowed range is @code{[0, 255]}.
24377 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24378 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24381 Set the vectorscope drawing mode.
24383 Available values are:
24386 Draw dot for each sample.
24389 Draw line between previous and current sample.
24392 Default value is @samp{dot}.
24395 Specify amplitude scale of audio samples.
24397 Available values are:
24413 Swap left channel axis with right channel axis.
24423 Mirror only x axis.
24426 Mirror only y axis.
24434 @subsection Examples
24438 Complete example using @command{ffplay}:
24440 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24441 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24445 @section bench, abench
24447 Benchmark part of a filtergraph.
24449 The filter accepts the following options:
24453 Start or stop a timer.
24455 Available values are:
24458 Get the current time, set it as frame metadata (using the key
24459 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24462 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24463 the input frame metadata to get the time difference. Time difference, average,
24464 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24465 @code{min}) are then printed. The timestamps are expressed in seconds.
24469 @subsection Examples
24473 Benchmark @ref{selectivecolor} filter:
24475 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24481 Concatenate audio and video streams, joining them together one after the
24484 The filter works on segments of synchronized video and audio streams. All
24485 segments must have the same number of streams of each type, and that will
24486 also be the number of streams at output.
24488 The filter accepts the following options:
24493 Set the number of segments. Default is 2.
24496 Set the number of output video streams, that is also the number of video
24497 streams in each segment. Default is 1.
24500 Set the number of output audio streams, that is also the number of audio
24501 streams in each segment. Default is 0.
24504 Activate unsafe mode: do not fail if segments have a different format.
24508 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24509 @var{a} audio outputs.
24511 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24512 segment, in the same order as the outputs, then the inputs for the second
24515 Related streams do not always have exactly the same duration, for various
24516 reasons including codec frame size or sloppy authoring. For that reason,
24517 related synchronized streams (e.g. a video and its audio track) should be
24518 concatenated at once. The concat filter will use the duration of the longest
24519 stream in each segment (except the last one), and if necessary pad shorter
24520 audio streams with silence.
24522 For this filter to work correctly, all segments must start at timestamp 0.
24524 All corresponding streams must have the same parameters in all segments; the
24525 filtering system will automatically select a common pixel format for video
24526 streams, and a common sample format, sample rate and channel layout for
24527 audio streams, but other settings, such as resolution, must be converted
24528 explicitly by the user.
24530 Different frame rates are acceptable but will result in variable frame rate
24531 at output; be sure to configure the output file to handle it.
24533 @subsection Examples
24537 Concatenate an opening, an episode and an ending, all in bilingual version
24538 (video in stream 0, audio in streams 1 and 2):
24540 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24541 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24542 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24543 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24547 Concatenate two parts, handling audio and video separately, using the
24548 (a)movie sources, and adjusting the resolution:
24550 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24551 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24552 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24554 Note that a desync will happen at the stitch if the audio and video streams
24555 do not have exactly the same duration in the first file.
24559 @subsection Commands
24561 This filter supports the following commands:
24564 Close the current segment and step to the next one
24570 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24571 level. By default, it logs a message at a frequency of 10Hz with the
24572 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24573 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24575 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24576 sample format is double-precision floating point. The input stream will be converted to
24577 this specification, if needed. Users may need to insert aformat and/or aresample filters
24578 after this filter to obtain the original parameters.
24580 The filter also has a video output (see the @var{video} option) with a real
24581 time graph to observe the loudness evolution. The graphic contains the logged
24582 message mentioned above, so it is not printed anymore when this option is set,
24583 unless the verbose logging is set. The main graphing area contains the
24584 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24585 the momentary loudness (400 milliseconds), but can optionally be configured
24586 to instead display short-term loudness (see @var{gauge}).
24588 The green area marks a +/- 1LU target range around the target loudness
24589 (-23LUFS by default, unless modified through @var{target}).
24591 More information about the Loudness Recommendation EBU R128 on
24592 @url{http://tech.ebu.ch/loudness}.
24594 The filter accepts the following options:
24599 Activate the video output. The audio stream is passed unchanged whether this
24600 option is set or no. The video stream will be the first output stream if
24601 activated. Default is @code{0}.
24604 Set the video size. This option is for video only. For the syntax of this
24606 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24607 Default and minimum resolution is @code{640x480}.
24610 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24611 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24612 other integer value between this range is allowed.
24615 Set metadata injection. If set to @code{1}, the audio input will be segmented
24616 into 100ms output frames, each of them containing various loudness information
24617 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24619 Default is @code{0}.
24622 Force the frame logging level.
24624 Available values are:
24627 information logging level
24629 verbose logging level
24632 By default, the logging level is set to @var{info}. If the @option{video} or
24633 the @option{metadata} options are set, it switches to @var{verbose}.
24638 Available modes can be cumulated (the option is a @code{flag} type). Possible
24642 Disable any peak mode (default).
24644 Enable sample-peak mode.
24646 Simple peak mode looking for the higher sample value. It logs a message
24647 for sample-peak (identified by @code{SPK}).
24649 Enable true-peak mode.
24651 If enabled, the peak lookup is done on an over-sampled version of the input
24652 stream for better peak accuracy. It logs a message for true-peak.
24653 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24654 This mode requires a build with @code{libswresample}.
24658 Treat mono input files as "dual mono". If a mono file is intended for playback
24659 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24660 If set to @code{true}, this option will compensate for this effect.
24661 Multi-channel input files are not affected by this option.
24664 Set a specific pan law to be used for the measurement of dual mono files.
24665 This parameter is optional, and has a default value of -3.01dB.
24668 Set a specific target level (in LUFS) used as relative zero in the visualization.
24669 This parameter is optional and has a default value of -23LUFS as specified
24670 by EBU R128. However, material published online may prefer a level of -16LUFS
24671 (e.g. for use with podcasts or video platforms).
24674 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24675 @code{shortterm}. By default the momentary value will be used, but in certain
24676 scenarios it may be more useful to observe the short term value instead (e.g.
24680 Sets the display scale for the loudness. Valid parameters are @code{absolute}
24681 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
24682 video output, not the summary or continuous log output.
24685 @subsection Examples
24689 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
24691 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
24695 Run an analysis with @command{ffmpeg}:
24697 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
24701 @section interleave, ainterleave
24703 Temporally interleave frames from several inputs.
24705 @code{interleave} works with video inputs, @code{ainterleave} with audio.
24707 These filters read frames from several inputs and send the oldest
24708 queued frame to the output.
24710 Input streams must have well defined, monotonically increasing frame
24713 In order to submit one frame to output, these filters need to enqueue
24714 at least one frame for each input, so they cannot work in case one
24715 input is not yet terminated and will not receive incoming frames.
24717 For example consider the case when one input is a @code{select} filter
24718 which always drops input frames. The @code{interleave} filter will keep
24719 reading from that input, but it will never be able to send new frames
24720 to output until the input sends an end-of-stream signal.
24722 Also, depending on inputs synchronization, the filters will drop
24723 frames in case one input receives more frames than the other ones, and
24724 the queue is already filled.
24726 These filters accept the following options:
24730 Set the number of different inputs, it is 2 by default.
24733 How to determine the end-of-stream.
24737 The duration of the longest input. (default)
24740 The duration of the shortest input.
24743 The duration of the first input.
24748 @subsection Examples
24752 Interleave frames belonging to different streams using @command{ffmpeg}:
24754 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
24758 Add flickering blur effect:
24760 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
24764 @section metadata, ametadata
24766 Manipulate frame metadata.
24768 This filter accepts the following options:
24772 Set mode of operation of the filter.
24774 Can be one of the following:
24778 If both @code{value} and @code{key} is set, select frames
24779 which have such metadata. If only @code{key} is set, select
24780 every frame that has such key in metadata.
24783 Add new metadata @code{key} and @code{value}. If key is already available
24787 Modify value of already present key.
24790 If @code{value} is set, delete only keys that have such value.
24791 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
24795 Print key and its value if metadata was found. If @code{key} is not set print all
24796 metadata values available in frame.
24800 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
24803 Set metadata value which will be used. This option is mandatory for
24804 @code{modify} and @code{add} mode.
24807 Which function to use when comparing metadata value and @code{value}.
24809 Can be one of following:
24813 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
24816 Values are interpreted as strings, returns true if metadata value starts with
24817 the @code{value} option string.
24820 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
24823 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
24826 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
24829 Values are interpreted as floats, returns true if expression from option @code{expr}
24833 Values are interpreted as strings, returns true if metadata value ends with
24834 the @code{value} option string.
24838 Set expression which is used when @code{function} is set to @code{expr}.
24839 The expression is evaluated through the eval API and can contain the following
24844 Float representation of @code{value} from metadata key.
24847 Float representation of @code{value} as supplied by user in @code{value} option.
24851 If specified in @code{print} mode, output is written to the named file. Instead of
24852 plain filename any writable url can be specified. Filename ``-'' is a shorthand
24853 for standard output. If @code{file} option is not set, output is written to the log
24854 with AV_LOG_INFO loglevel.
24857 Reduces buffering in print mode when output is written to a URL set using @var{file}.
24861 @subsection Examples
24865 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
24868 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
24871 Print silencedetect output to file @file{metadata.txt}.
24873 silencedetect,ametadata=mode=print:file=metadata.txt
24876 Direct all metadata to a pipe with file descriptor 4.
24878 metadata=mode=print:file='pipe\:4'
24882 @section perms, aperms
24884 Set read/write permissions for the output frames.
24886 These filters are mainly aimed at developers to test direct path in the
24887 following filter in the filtergraph.
24889 The filters accept the following options:
24893 Select the permissions mode.
24895 It accepts the following values:
24898 Do nothing. This is the default.
24900 Set all the output frames read-only.
24902 Set all the output frames directly writable.
24904 Make the frame read-only if writable, and writable if read-only.
24906 Set each output frame read-only or writable randomly.
24910 Set the seed for the @var{random} mode, must be an integer included between
24911 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
24912 @code{-1}, the filter will try to use a good random seed on a best effort
24916 Note: in case of auto-inserted filter between the permission filter and the
24917 following one, the permission might not be received as expected in that
24918 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
24919 perms/aperms filter can avoid this problem.
24921 @section realtime, arealtime
24923 Slow down filtering to match real time approximately.
24925 These filters will pause the filtering for a variable amount of time to
24926 match the output rate with the input timestamps.
24927 They are similar to the @option{re} option to @code{ffmpeg}.
24929 They accept the following options:
24933 Time limit for the pauses. Any pause longer than that will be considered
24934 a timestamp discontinuity and reset the timer. Default is 2 seconds.
24936 Speed factor for processing. The value must be a float larger than zero.
24937 Values larger than 1.0 will result in faster than realtime processing,
24938 smaller will slow processing down. The @var{limit} is automatically adapted
24939 accordingly. Default is 1.0.
24941 A processing speed faster than what is possible without these filters cannot
24946 @section select, aselect
24948 Select frames to pass in output.
24950 This filter accepts the following options:
24955 Set expression, which is evaluated for each input frame.
24957 If the expression is evaluated to zero, the frame is discarded.
24959 If the evaluation result is negative or NaN, the frame is sent to the
24960 first output; otherwise it is sent to the output with index
24961 @code{ceil(val)-1}, assuming that the input index starts from 0.
24963 For example a value of @code{1.2} corresponds to the output with index
24964 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
24967 Set the number of outputs. The output to which to send the selected
24968 frame is based on the result of the evaluation. Default value is 1.
24971 The expression can contain the following constants:
24975 The (sequential) number of the filtered frame, starting from 0.
24978 The (sequential) number of the selected frame, starting from 0.
24980 @item prev_selected_n
24981 The sequential number of the last selected frame. It's NAN if undefined.
24984 The timebase of the input timestamps.
24987 The PTS (Presentation TimeStamp) of the filtered video frame,
24988 expressed in @var{TB} units. It's NAN if undefined.
24991 The PTS of the filtered video frame,
24992 expressed in seconds. It's NAN if undefined.
24995 The PTS of the previously filtered video frame. It's NAN if undefined.
24997 @item prev_selected_pts
24998 The PTS of the last previously filtered video frame. It's NAN if undefined.
25000 @item prev_selected_t
25001 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25004 The PTS of the first video frame in the video. It's NAN if undefined.
25007 The time of the first video frame in the video. It's NAN if undefined.
25009 @item pict_type @emph{(video only)}
25010 The type of the filtered frame. It can assume one of the following
25022 @item interlace_type @emph{(video only)}
25023 The frame interlace type. It can assume one of the following values:
25026 The frame is progressive (not interlaced).
25028 The frame is top-field-first.
25030 The frame is bottom-field-first.
25033 @item consumed_sample_n @emph{(audio only)}
25034 the number of selected samples before the current frame
25036 @item samples_n @emph{(audio only)}
25037 the number of samples in the current frame
25039 @item sample_rate @emph{(audio only)}
25040 the input sample rate
25043 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25046 the position in the file of the filtered frame, -1 if the information
25047 is not available (e.g. for synthetic video)
25049 @item scene @emph{(video only)}
25050 value between 0 and 1 to indicate a new scene; a low value reflects a low
25051 probability for the current frame to introduce a new scene, while a higher
25052 value means the current frame is more likely to be one (see the example below)
25054 @item concatdec_select
25055 The concat demuxer can select only part of a concat input file by setting an
25056 inpoint and an outpoint, but the output packets may not be entirely contained
25057 in the selected interval. By using this variable, it is possible to skip frames
25058 generated by the concat demuxer which are not exactly contained in the selected
25061 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25062 and the @var{lavf.concat.duration} packet metadata values which are also
25063 present in the decoded frames.
25065 The @var{concatdec_select} variable is -1 if the frame pts is at least
25066 start_time and either the duration metadata is missing or the frame pts is less
25067 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25070 That basically means that an input frame is selected if its pts is within the
25071 interval set by the concat demuxer.
25075 The default value of the select expression is "1".
25077 @subsection Examples
25081 Select all frames in input:
25086 The example above is the same as:
25098 Select only I-frames:
25100 select='eq(pict_type\,I)'
25104 Select one frame every 100:
25106 select='not(mod(n\,100))'
25110 Select only frames contained in the 10-20 time interval:
25112 select=between(t\,10\,20)
25116 Select only I-frames contained in the 10-20 time interval:
25118 select=between(t\,10\,20)*eq(pict_type\,I)
25122 Select frames with a minimum distance of 10 seconds:
25124 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25128 Use aselect to select only audio frames with samples number > 100:
25130 aselect='gt(samples_n\,100)'
25134 Create a mosaic of the first scenes:
25136 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25139 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25143 Send even and odd frames to separate outputs, and compose them:
25145 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25149 Select useful frames from an ffconcat file which is using inpoints and
25150 outpoints but where the source files are not intra frame only.
25152 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25156 @section sendcmd, asendcmd
25158 Send commands to filters in the filtergraph.
25160 These filters read commands to be sent to other filters in the
25163 @code{sendcmd} must be inserted between two video filters,
25164 @code{asendcmd} must be inserted between two audio filters, but apart
25165 from that they act the same way.
25167 The specification of commands can be provided in the filter arguments
25168 with the @var{commands} option, or in a file specified by the
25169 @var{filename} option.
25171 These filters accept the following options:
25174 Set the commands to be read and sent to the other filters.
25176 Set the filename of the commands to be read and sent to the other
25180 @subsection Commands syntax
25182 A commands description consists of a sequence of interval
25183 specifications, comprising a list of commands to be executed when a
25184 particular event related to that interval occurs. The occurring event
25185 is typically the current frame time entering or leaving a given time
25188 An interval is specified by the following syntax:
25190 @var{START}[-@var{END}] @var{COMMANDS};
25193 The time interval is specified by the @var{START} and @var{END} times.
25194 @var{END} is optional and defaults to the maximum time.
25196 The current frame time is considered within the specified interval if
25197 it is included in the interval [@var{START}, @var{END}), that is when
25198 the time is greater or equal to @var{START} and is lesser than
25201 @var{COMMANDS} consists of a sequence of one or more command
25202 specifications, separated by ",", relating to that interval. The
25203 syntax of a command specification is given by:
25205 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25208 @var{FLAGS} is optional and specifies the type of events relating to
25209 the time interval which enable sending the specified command, and must
25210 be a non-null sequence of identifier flags separated by "+" or "|" and
25211 enclosed between "[" and "]".
25213 The following flags are recognized:
25216 The command is sent when the current frame timestamp enters the
25217 specified interval. In other words, the command is sent when the
25218 previous frame timestamp was not in the given interval, and the
25222 The command is sent when the current frame timestamp leaves the
25223 specified interval. In other words, the command is sent when the
25224 previous frame timestamp was in the given interval, and the
25228 The command @var{ARG} is interpreted as expression and result of
25229 expression is passed as @var{ARG}.
25231 The expression is evaluated through the eval API and can contain the following
25236 Original position in the file of the frame, or undefined if undefined
25237 for the current frame.
25240 The presentation timestamp in input.
25243 The count of the input frame for video or audio, starting from 0.
25246 The time in seconds of the current frame.
25249 The start time in seconds of the current command interval.
25252 The end time in seconds of the current command interval.
25255 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25260 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25263 @var{TARGET} specifies the target of the command, usually the name of
25264 the filter class or a specific filter instance name.
25266 @var{COMMAND} specifies the name of the command for the target filter.
25268 @var{ARG} is optional and specifies the optional list of argument for
25269 the given @var{COMMAND}.
25271 Between one interval specification and another, whitespaces, or
25272 sequences of characters starting with @code{#} until the end of line,
25273 are ignored and can be used to annotate comments.
25275 A simplified BNF description of the commands specification syntax
25278 @var{COMMAND_FLAG} ::= "enter" | "leave"
25279 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25280 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25281 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25282 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25283 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25286 @subsection Examples
25290 Specify audio tempo change at second 4:
25292 asendcmd=c='4.0 atempo tempo 1.5',atempo
25296 Target a specific filter instance:
25298 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25302 Specify a list of drawtext and hue commands in a file.
25304 # show text in the interval 5-10
25305 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25306 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25308 # desaturate the image in the interval 15-20
25309 15.0-20.0 [enter] hue s 0,
25310 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25312 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25314 # apply an exponential saturation fade-out effect, starting from time 25
25315 25 [enter] hue s exp(25-t)
25318 A filtergraph allowing to read and process the above command list
25319 stored in a file @file{test.cmd}, can be specified with:
25321 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25326 @section setpts, asetpts
25328 Change the PTS (presentation timestamp) of the input frames.
25330 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25332 This filter accepts the following options:
25337 The expression which is evaluated for each frame to construct its timestamp.
25341 The expression is evaluated through the eval API and can contain the following
25345 @item FRAME_RATE, FR
25346 frame rate, only defined for constant frame-rate video
25349 The presentation timestamp in input
25352 The count of the input frame for video or the number of consumed samples,
25353 not including the current frame for audio, starting from 0.
25355 @item NB_CONSUMED_SAMPLES
25356 The number of consumed samples, not including the current frame (only
25359 @item NB_SAMPLES, S
25360 The number of samples in the current frame (only audio)
25362 @item SAMPLE_RATE, SR
25363 The audio sample rate.
25366 The PTS of the first frame.
25369 the time in seconds of the first frame
25372 State whether the current frame is interlaced.
25375 the time in seconds of the current frame
25378 original position in the file of the frame, or undefined if undefined
25379 for the current frame
25382 The previous input PTS.
25385 previous input time in seconds
25388 The previous output PTS.
25391 previous output time in seconds
25394 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25398 The wallclock (RTC) time at the start of the movie in microseconds.
25401 The timebase of the input timestamps.
25405 @subsection Examples
25409 Start counting PTS from zero
25411 setpts=PTS-STARTPTS
25415 Apply fast motion effect:
25421 Apply slow motion effect:
25427 Set fixed rate of 25 frames per second:
25433 Set fixed rate 25 fps with some jitter:
25435 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25439 Apply an offset of 10 seconds to the input PTS:
25445 Generate timestamps from a "live source" and rebase onto the current timebase:
25447 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25451 Generate timestamps by counting samples:
25460 Force color range for the output video frame.
25462 The @code{setrange} filter marks the color range property for the
25463 output frames. It does not change the input frame, but only sets the
25464 corresponding property, which affects how the frame is treated by
25467 The filter accepts the following options:
25472 Available values are:
25476 Keep the same color range property.
25478 @item unspecified, unknown
25479 Set the color range as unspecified.
25481 @item limited, tv, mpeg
25482 Set the color range as limited.
25484 @item full, pc, jpeg
25485 Set the color range as full.
25489 @section settb, asettb
25491 Set the timebase to use for the output frames timestamps.
25492 It is mainly useful for testing timebase configuration.
25494 It accepts the following parameters:
25499 The expression which is evaluated into the output timebase.
25503 The value for @option{tb} is an arithmetic expression representing a
25504 rational. The expression can contain the constants "AVTB" (the default
25505 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25506 audio only). Default value is "intb".
25508 @subsection Examples
25512 Set the timebase to 1/25:
25518 Set the timebase to 1/10:
25524 Set the timebase to 1001/1000:
25530 Set the timebase to 2*intb:
25536 Set the default timebase value:
25543 Convert input audio to a video output representing frequency spectrum
25544 logarithmically using Brown-Puckette constant Q transform algorithm with
25545 direct frequency domain coefficient calculation (but the transform itself
25546 is not really constant Q, instead the Q factor is actually variable/clamped),
25547 with musical tone scale, from E0 to D#10.
25549 The filter accepts the following options:
25553 Specify the video size for the output. It must be even. For the syntax of this option,
25554 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25555 Default value is @code{1920x1080}.
25558 Set the output frame rate. Default value is @code{25}.
25561 Set the bargraph height. It must be even. Default value is @code{-1} which
25562 computes the bargraph height automatically.
25565 Set the axis height. It must be even. Default value is @code{-1} which computes
25566 the axis height automatically.
25569 Set the sonogram height. It must be even. Default value is @code{-1} which
25570 computes the sonogram height automatically.
25573 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25574 instead. Default value is @code{1}.
25576 @item sono_v, volume
25577 Specify the sonogram volume expression. It can contain variables:
25580 the @var{bar_v} evaluated expression
25581 @item frequency, freq, f
25582 the frequency where it is evaluated
25583 @item timeclamp, tc
25584 the value of @var{timeclamp} option
25588 @item a_weighting(f)
25589 A-weighting of equal loudness
25590 @item b_weighting(f)
25591 B-weighting of equal loudness
25592 @item c_weighting(f)
25593 C-weighting of equal loudness.
25595 Default value is @code{16}.
25597 @item bar_v, volume2
25598 Specify the bargraph volume expression. It can contain variables:
25601 the @var{sono_v} evaluated expression
25602 @item frequency, freq, f
25603 the frequency where it is evaluated
25604 @item timeclamp, tc
25605 the value of @var{timeclamp} option
25609 @item a_weighting(f)
25610 A-weighting of equal loudness
25611 @item b_weighting(f)
25612 B-weighting of equal loudness
25613 @item c_weighting(f)
25614 C-weighting of equal loudness.
25616 Default value is @code{sono_v}.
25618 @item sono_g, gamma
25619 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25620 higher gamma makes the spectrum having more range. Default value is @code{3}.
25621 Acceptable range is @code{[1, 7]}.
25623 @item bar_g, gamma2
25624 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25628 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25629 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25631 @item timeclamp, tc
25632 Specify the transform timeclamp. At low frequency, there is trade-off between
25633 accuracy in time domain and frequency domain. If timeclamp is lower,
25634 event in time domain is represented more accurately (such as fast bass drum),
25635 otherwise event in frequency domain is represented more accurately
25636 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25639 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25640 limits future samples by applying asymmetric windowing in time domain, useful
25641 when low latency is required. Accepted range is @code{[0, 1]}.
25644 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25645 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25648 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25649 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25652 This option is deprecated and ignored.
25655 Specify the transform length in time domain. Use this option to control accuracy
25656 trade-off between time domain and frequency domain at every frequency sample.
25657 It can contain variables:
25659 @item frequency, freq, f
25660 the frequency where it is evaluated
25661 @item timeclamp, tc
25662 the value of @var{timeclamp} option.
25664 Default value is @code{384*tc/(384+tc*f)}.
25667 Specify the transform count for every video frame. Default value is @code{6}.
25668 Acceptable range is @code{[1, 30]}.
25671 Specify the transform count for every single pixel. Default value is @code{0},
25672 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25675 Specify font file for use with freetype to draw the axis. If not specified,
25676 use embedded font. Note that drawing with font file or embedded font is not
25677 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
25681 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
25682 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
25686 Specify font color expression. This is arithmetic expression that should return
25687 integer value 0xRRGGBB. It can contain variables:
25689 @item frequency, freq, f
25690 the frequency where it is evaluated
25691 @item timeclamp, tc
25692 the value of @var{timeclamp} option
25697 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
25698 @item r(x), g(x), b(x)
25699 red, green, and blue value of intensity x.
25701 Default value is @code{st(0, (midi(f)-59.5)/12);
25702 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
25703 r(1-ld(1)) + b(ld(1))}.
25706 Specify image file to draw the axis. This option override @var{fontfile} and
25707 @var{fontcolor} option.
25710 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
25711 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
25712 Default value is @code{1}.
25715 Set colorspace. The accepted values are:
25718 Unspecified (default)
25727 BT.470BG or BT.601-6 625
25730 SMPTE-170M or BT.601-6 525
25736 BT.2020 with non-constant luminance
25741 Set spectrogram color scheme. This is list of floating point values with format
25742 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
25743 The default is @code{1|0.5|0|0|0.5|1}.
25747 @subsection Examples
25751 Playing audio while showing the spectrum:
25753 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
25757 Same as above, but with frame rate 30 fps:
25759 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
25763 Playing at 1280x720:
25765 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
25769 Disable sonogram display:
25775 A1 and its harmonics: A1, A2, (near)E3, A3:
25777 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),
25778 asplit[a][out1]; [a] showcqt [out0]'
25782 Same as above, but with more accuracy in frequency domain:
25784 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),
25785 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
25791 bar_v=10:sono_v=bar_v*a_weighting(f)
25795 Custom gamma, now spectrum is linear to the amplitude.
25801 Custom tlength equation:
25803 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)))'
25807 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
25809 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
25813 Custom font using fontconfig:
25815 font='Courier New,Monospace,mono|bold'
25819 Custom frequency range with custom axis using image file:
25821 axisfile=myaxis.png:basefreq=40:endfreq=10000
25827 Convert input audio to video output representing the audio power spectrum.
25828 Audio amplitude is on Y-axis while frequency is on X-axis.
25830 The filter accepts the following options:
25834 Specify size of video. For the syntax of this option, check the
25835 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25836 Default is @code{1024x512}.
25840 This set how each frequency bin will be represented.
25842 It accepts the following values:
25848 Default is @code{bar}.
25851 Set amplitude scale.
25853 It accepts the following values:
25867 Default is @code{log}.
25870 Set frequency scale.
25872 It accepts the following values:
25881 Reverse logarithmic scale.
25883 Default is @code{lin}.
25886 Set window size. Allowed range is from 16 to 65536.
25888 Default is @code{2048}
25891 Set windowing function.
25893 It accepts the following values:
25916 Default is @code{hanning}.
25919 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
25920 which means optimal overlap for selected window function will be picked.
25923 Set time averaging. Setting this to 0 will display current maximal peaks.
25924 Default is @code{1}, which means time averaging is disabled.
25927 Specify list of colors separated by space or by '|' which will be used to
25928 draw channel frequencies. Unrecognized or missing colors will be replaced
25932 Set channel display mode.
25934 It accepts the following values:
25939 Default is @code{combined}.
25942 Set minimum amplitude used in @code{log} amplitude scaler.
25945 Set data display mode.
25947 It accepts the following values:
25953 Default is @code{magnitude}.
25956 @section showspatial
25958 Convert stereo input audio to a video output, representing the spatial relationship
25959 between two channels.
25961 The filter accepts the following options:
25965 Specify the video size for the output. For the syntax of this option, check the
25966 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25967 Default value is @code{512x512}.
25970 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
25973 Set window function.
25975 It accepts the following values:
26000 Default value is @code{hann}.
26003 Set ratio of overlap window. Default value is @code{0.5}.
26004 When value is @code{1} overlap is set to recommended size for specific
26005 window function currently used.
26008 @anchor{showspectrum}
26009 @section showspectrum
26011 Convert input audio to a video output, representing the audio frequency
26014 The filter accepts the following options:
26018 Specify the video size for the output. For the syntax of this option, check the
26019 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26020 Default value is @code{640x512}.
26023 Specify how the spectrum should slide along the window.
26025 It accepts the following values:
26028 the samples start again on the left when they reach the right
26030 the samples scroll from right to left
26032 frames are only produced when the samples reach the right
26034 the samples scroll from left to right
26037 Default value is @code{replace}.
26040 Specify display mode.
26042 It accepts the following values:
26045 all channels are displayed in the same row
26047 all channels are displayed in separate rows
26050 Default value is @samp{combined}.
26053 Specify display color mode.
26055 It accepts the following values:
26058 each channel is displayed in a separate color
26060 each channel is displayed using the same color scheme
26062 each channel is displayed using the rainbow color scheme
26064 each channel is displayed using the moreland color scheme
26066 each channel is displayed using the nebulae color scheme
26068 each channel is displayed using the fire color scheme
26070 each channel is displayed using the fiery color scheme
26072 each channel is displayed using the fruit color scheme
26074 each channel is displayed using the cool color scheme
26076 each channel is displayed using the magma color scheme
26078 each channel is displayed using the green color scheme
26080 each channel is displayed using the viridis color scheme
26082 each channel is displayed using the plasma color scheme
26084 each channel is displayed using the cividis color scheme
26086 each channel is displayed using the terrain color scheme
26089 Default value is @samp{channel}.
26092 Specify scale used for calculating intensity color values.
26094 It accepts the following values:
26099 square root, default
26110 Default value is @samp{sqrt}.
26113 Specify frequency scale.
26115 It accepts the following values:
26123 Default value is @samp{lin}.
26126 Set saturation modifier for displayed colors. Negative values provide
26127 alternative color scheme. @code{0} is no saturation at all.
26128 Saturation must be in [-10.0, 10.0] range.
26129 Default value is @code{1}.
26132 Set window function.
26134 It accepts the following values:
26159 Default value is @code{hann}.
26162 Set orientation of time vs frequency axis. Can be @code{vertical} or
26163 @code{horizontal}. Default is @code{vertical}.
26166 Set ratio of overlap window. Default value is @code{0}.
26167 When value is @code{1} overlap is set to recommended size for specific
26168 window function currently used.
26171 Set scale gain for calculating intensity color values.
26172 Default value is @code{1}.
26175 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26178 Set color rotation, must be in [-1.0, 1.0] range.
26179 Default value is @code{0}.
26182 Set start frequency from which to display spectrogram. Default is @code{0}.
26185 Set stop frequency to which to display spectrogram. Default is @code{0}.
26188 Set upper frame rate limit. Default is @code{auto}, unlimited.
26191 Draw time and frequency axes and legends. Default is disabled.
26194 The usage is very similar to the showwaves filter; see the examples in that
26197 @subsection Examples
26201 Large window with logarithmic color scaling:
26203 showspectrum=s=1280x480:scale=log
26207 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26209 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26210 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26214 @section showspectrumpic
26216 Convert input audio to a single video frame, representing the audio frequency
26219 The filter accepts the following options:
26223 Specify the video size for the output. For the syntax of this option, check the
26224 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26225 Default value is @code{4096x2048}.
26228 Specify display mode.
26230 It accepts the following values:
26233 all channels are displayed in the same row
26235 all channels are displayed in separate rows
26237 Default value is @samp{combined}.
26240 Specify display color mode.
26242 It accepts the following values:
26245 each channel is displayed in a separate color
26247 each channel is displayed using the same color scheme
26249 each channel is displayed using the rainbow color scheme
26251 each channel is displayed using the moreland color scheme
26253 each channel is displayed using the nebulae color scheme
26255 each channel is displayed using the fire color scheme
26257 each channel is displayed using the fiery color scheme
26259 each channel is displayed using the fruit color scheme
26261 each channel is displayed using the cool color scheme
26263 each channel is displayed using the magma color scheme
26265 each channel is displayed using the green color scheme
26267 each channel is displayed using the viridis color scheme
26269 each channel is displayed using the plasma color scheme
26271 each channel is displayed using the cividis color scheme
26273 each channel is displayed using the terrain color scheme
26275 Default value is @samp{intensity}.
26278 Specify scale used for calculating intensity color values.
26280 It accepts the following values:
26285 square root, default
26295 Default value is @samp{log}.
26298 Specify frequency scale.
26300 It accepts the following values:
26308 Default value is @samp{lin}.
26311 Set saturation modifier for displayed colors. Negative values provide
26312 alternative color scheme. @code{0} is no saturation at all.
26313 Saturation must be in [-10.0, 10.0] range.
26314 Default value is @code{1}.
26317 Set window function.
26319 It accepts the following values:
26343 Default value is @code{hann}.
26346 Set orientation of time vs frequency axis. Can be @code{vertical} or
26347 @code{horizontal}. Default is @code{vertical}.
26350 Set scale gain for calculating intensity color values.
26351 Default value is @code{1}.
26354 Draw time and frequency axes and legends. Default is enabled.
26357 Set color rotation, must be in [-1.0, 1.0] range.
26358 Default value is @code{0}.
26361 Set start frequency from which to display spectrogram. Default is @code{0}.
26364 Set stop frequency to which to display spectrogram. Default is @code{0}.
26367 @subsection Examples
26371 Extract an audio spectrogram of a whole audio track
26372 in a 1024x1024 picture using @command{ffmpeg}:
26374 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26378 @section showvolume
26380 Convert input audio volume to a video output.
26382 The filter accepts the following options:
26389 Set border width, allowed range is [0, 5]. Default is 1.
26392 Set channel width, allowed range is [80, 8192]. Default is 400.
26395 Set channel height, allowed range is [1, 900]. Default is 20.
26398 Set fade, allowed range is [0, 1]. Default is 0.95.
26401 Set volume color expression.
26403 The expression can use the following variables:
26407 Current max volume of channel in dB.
26413 Current channel number, starting from 0.
26417 If set, displays channel names. Default is enabled.
26420 If set, displays volume values. Default is enabled.
26423 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26424 default is @code{h}.
26427 Set step size, allowed range is [0, 5]. Default is 0, which means
26431 Set background opacity, allowed range is [0, 1]. Default is 0.
26434 Set metering mode, can be peak: @code{p} or rms: @code{r},
26435 default is @code{p}.
26438 Set display scale, can be linear: @code{lin} or log: @code{log},
26439 default is @code{lin}.
26443 If set to > 0., display a line for the max level
26444 in the previous seconds.
26445 default is disabled: @code{0.}
26448 The color of the max line. Use when @code{dm} option is set to > 0.
26449 default is: @code{orange}
26454 Convert input audio to a video output, representing the samples waves.
26456 The filter accepts the following options:
26460 Specify the video size for the output. For the syntax of this option, check the
26461 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26462 Default value is @code{600x240}.
26467 Available values are:
26470 Draw a point for each sample.
26473 Draw a vertical line for each sample.
26476 Draw a point for each sample and a line between them.
26479 Draw a centered vertical line for each sample.
26482 Default value is @code{point}.
26485 Set the number of samples which are printed on the same column. A
26486 larger value will decrease the frame rate. Must be a positive
26487 integer. This option can be set only if the value for @var{rate}
26488 is not explicitly specified.
26491 Set the (approximate) output frame rate. This is done by setting the
26492 option @var{n}. Default value is "25".
26494 @item split_channels
26495 Set if channels should be drawn separately or overlap. Default value is 0.
26498 Set colors separated by '|' which are going to be used for drawing of each channel.
26501 Set amplitude scale.
26503 Available values are:
26521 Set the draw mode. This is mostly useful to set for high @var{n}.
26523 Available values are:
26526 Scale pixel values for each drawn sample.
26529 Draw every sample directly.
26532 Default value is @code{scale}.
26535 @subsection Examples
26539 Output the input file audio and the corresponding video representation
26542 amovie=a.mp3,asplit[out0],showwaves[out1]
26546 Create a synthetic signal and show it with showwaves, forcing a
26547 frame rate of 30 frames per second:
26549 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26553 @section showwavespic
26555 Convert input audio to a single video frame, representing the samples waves.
26557 The filter accepts the following options:
26561 Specify the video size for the output. For the syntax of this option, check the
26562 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26563 Default value is @code{600x240}.
26565 @item split_channels
26566 Set if channels should be drawn separately or overlap. Default value is 0.
26569 Set colors separated by '|' which are going to be used for drawing of each channel.
26572 Set amplitude scale.
26574 Available values are:
26594 Available values are:
26597 Scale pixel values for each drawn sample.
26600 Draw every sample directly.
26603 Default value is @code{scale}.
26606 Set the filter mode.
26608 Available values are:
26611 Use average samples values for each drawn sample.
26614 Use peak samples values for each drawn sample.
26617 Default value is @code{average}.
26620 @subsection Examples
26624 Extract a channel split representation of the wave form of a whole audio track
26625 in a 1024x800 picture using @command{ffmpeg}:
26627 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26631 @section sidedata, asidedata
26633 Delete frame side data, or select frames based on it.
26635 This filter accepts the following options:
26639 Set mode of operation of the filter.
26641 Can be one of the following:
26645 Select every frame with side data of @code{type}.
26648 Delete side data of @code{type}. If @code{type} is not set, delete all side
26654 Set side data type used with all modes. Must be set for @code{select} mode. For
26655 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26656 in @file{libavutil/frame.h}. For example, to choose
26657 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26661 @section spectrumsynth
26663 Synthesize audio from 2 input video spectrums, first input stream represents
26664 magnitude across time and second represents phase across time.
26665 The filter will transform from frequency domain as displayed in videos back
26666 to time domain as presented in audio output.
26668 This filter is primarily created for reversing processed @ref{showspectrum}
26669 filter outputs, but can synthesize sound from other spectrograms too.
26670 But in such case results are going to be poor if the phase data is not
26671 available, because in such cases phase data need to be recreated, usually
26672 it's just recreated from random noise.
26673 For best results use gray only output (@code{channel} color mode in
26674 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26675 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26676 @code{data} option. Inputs videos should generally use @code{fullframe}
26677 slide mode as that saves resources needed for decoding video.
26679 The filter accepts the following options:
26683 Specify sample rate of output audio, the sample rate of audio from which
26684 spectrum was generated may differ.
26687 Set number of channels represented in input video spectrums.
26690 Set scale which was used when generating magnitude input spectrum.
26691 Can be @code{lin} or @code{log}. Default is @code{log}.
26694 Set slide which was used when generating inputs spectrums.
26695 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
26696 Default is @code{fullframe}.
26699 Set window function used for resynthesis.
26702 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26703 which means optimal overlap for selected window function will be picked.
26706 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
26707 Default is @code{vertical}.
26710 @subsection Examples
26714 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
26715 then resynthesize videos back to audio with spectrumsynth:
26717 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
26718 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
26719 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
26723 @section split, asplit
26725 Split input into several identical outputs.
26727 @code{asplit} works with audio input, @code{split} with video.
26729 The filter accepts a single parameter which specifies the number of outputs. If
26730 unspecified, it defaults to 2.
26732 @subsection Examples
26736 Create two separate outputs from the same input:
26738 [in] split [out0][out1]
26742 To create 3 or more outputs, you need to specify the number of
26745 [in] asplit=3 [out0][out1][out2]
26749 Create two separate outputs from the same input, one cropped and
26752 [in] split [splitout1][splitout2];
26753 [splitout1] crop=100:100:0:0 [cropout];
26754 [splitout2] pad=200:200:100:100 [padout];
26758 Create 5 copies of the input audio with @command{ffmpeg}:
26760 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
26766 Receive commands sent through a libzmq client, and forward them to
26767 filters in the filtergraph.
26769 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
26770 must be inserted between two video filters, @code{azmq} between two
26771 audio filters. Both are capable to send messages to any filter type.
26773 To enable these filters you need to install the libzmq library and
26774 headers and configure FFmpeg with @code{--enable-libzmq}.
26776 For more information about libzmq see:
26777 @url{http://www.zeromq.org/}
26779 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
26780 receives messages sent through a network interface defined by the
26781 @option{bind_address} (or the abbreviation "@option{b}") option.
26782 Default value of this option is @file{tcp://localhost:5555}. You may
26783 want to alter this value to your needs, but do not forget to escape any
26784 ':' signs (see @ref{filtergraph escaping}).
26786 The received message must be in the form:
26788 @var{TARGET} @var{COMMAND} [@var{ARG}]
26791 @var{TARGET} specifies the target of the command, usually the name of
26792 the filter class or a specific filter instance name. The default
26793 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
26794 but you can override this by using the @samp{filter_name@@id} syntax
26795 (see @ref{Filtergraph syntax}).
26797 @var{COMMAND} specifies the name of the command for the target filter.
26799 @var{ARG} is optional and specifies the optional argument list for the
26800 given @var{COMMAND}.
26802 Upon reception, the message is processed and the corresponding command
26803 is injected into the filtergraph. Depending on the result, the filter
26804 will send a reply to the client, adopting the format:
26806 @var{ERROR_CODE} @var{ERROR_REASON}
26810 @var{MESSAGE} is optional.
26812 @subsection Examples
26814 Look at @file{tools/zmqsend} for an example of a zmq client which can
26815 be used to send commands processed by these filters.
26817 Consider the following filtergraph generated by @command{ffplay}.
26818 In this example the last overlay filter has an instance name. All other
26819 filters will have default instance names.
26822 ffplay -dumpgraph 1 -f lavfi "
26823 color=s=100x100:c=red [l];
26824 color=s=100x100:c=blue [r];
26825 nullsrc=s=200x100, zmq [bg];
26826 [bg][l] overlay [bg+l];
26827 [bg+l][r] overlay@@my=x=100 "
26830 To change the color of the left side of the video, the following
26831 command can be used:
26833 echo Parsed_color_0 c yellow | tools/zmqsend
26836 To change the right side:
26838 echo Parsed_color_1 c pink | tools/zmqsend
26841 To change the position of the right side:
26843 echo overlay@@my x 150 | tools/zmqsend
26847 @c man end MULTIMEDIA FILTERS
26849 @chapter Multimedia Sources
26850 @c man begin MULTIMEDIA SOURCES
26852 Below is a description of the currently available multimedia sources.
26856 This is the same as @ref{movie} source, except it selects an audio
26862 Read audio and/or video stream(s) from a movie container.
26864 It accepts the following parameters:
26868 The name of the resource to read (not necessarily a file; it can also be a
26869 device or a stream accessed through some protocol).
26871 @item format_name, f
26872 Specifies the format assumed for the movie to read, and can be either
26873 the name of a container or an input device. If not specified, the
26874 format is guessed from @var{movie_name} or by probing.
26876 @item seek_point, sp
26877 Specifies the seek point in seconds. The frames will be output
26878 starting from this seek point. The parameter is evaluated with
26879 @code{av_strtod}, so the numerical value may be suffixed by an IS
26880 postfix. The default value is "0".
26883 Specifies the streams to read. Several streams can be specified,
26884 separated by "+". The source will then have as many outputs, in the
26885 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
26886 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
26887 respectively the default (best suited) video and audio stream. Default
26888 is "dv", or "da" if the filter is called as "amovie".
26890 @item stream_index, si
26891 Specifies the index of the video stream to read. If the value is -1,
26892 the most suitable video stream will be automatically selected. The default
26893 value is "-1". Deprecated. If the filter is called "amovie", it will select
26894 audio instead of video.
26897 Specifies how many times to read the stream in sequence.
26898 If the value is 0, the stream will be looped infinitely.
26899 Default value is "1".
26901 Note that when the movie is looped the source timestamps are not
26902 changed, so it will generate non monotonically increasing timestamps.
26904 @item discontinuity
26905 Specifies the time difference between frames above which the point is
26906 considered a timestamp discontinuity which is removed by adjusting the later
26910 It allows overlaying a second video on top of the main input of
26911 a filtergraph, as shown in this graph:
26913 input -----------> deltapts0 --> overlay --> output
26916 movie --> scale--> deltapts1 -------+
26918 @subsection Examples
26922 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
26923 on top of the input labelled "in":
26925 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
26926 [in] setpts=PTS-STARTPTS [main];
26927 [main][over] overlay=16:16 [out]
26931 Read from a video4linux2 device, and overlay it on top of the input
26934 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
26935 [in] setpts=PTS-STARTPTS [main];
26936 [main][over] overlay=16:16 [out]
26940 Read the first video stream and the audio stream with id 0x81 from
26941 dvd.vob; the video is connected to the pad named "video" and the audio is
26942 connected to the pad named "audio":
26944 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
26948 @subsection Commands
26950 Both movie and amovie support the following commands:
26953 Perform seek using "av_seek_frame".
26954 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
26957 @var{stream_index}: If stream_index is -1, a default
26958 stream is selected, and @var{timestamp} is automatically converted
26959 from AV_TIME_BASE units to the stream specific time_base.
26961 @var{timestamp}: Timestamp in AVStream.time_base units
26962 or, if no stream is specified, in AV_TIME_BASE units.
26964 @var{flags}: Flags which select direction and seeking mode.
26968 Get movie duration in AV_TIME_BASE units.
26972 @c man end MULTIMEDIA SOURCES