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{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.
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{LINKLABEL} ::= "[" @var{NAME} "]"
216 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
217 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
218 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
219 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
220 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
223 @section Notes on filtergraph escaping
225 Filtergraph description composition entails several levels of
226 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
227 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
228 information about the employed escaping procedure.
230 A first level escaping affects the content of each filter option
231 value, which may contain the special character @code{:} used to
232 separate values, or one of the escaping characters @code{\'}.
234 A second level escaping affects the whole filter description, which
235 may contain the escaping characters @code{\'} or the special
236 characters @code{[],;} used by the filtergraph description.
238 Finally, when you specify a filtergraph on a shell commandline, you
239 need to perform a third level escaping for the shell special
240 characters contained within it.
242 For example, consider the following string to be embedded in
243 the @ref{drawtext} filter description @option{text} value:
245 this is a 'string': may contain one, or more, special characters
248 This string contains the @code{'} special escaping character, and the
249 @code{:} special character, so it needs to be escaped in this way:
251 text=this is a \'string\'\: may contain one, or more, special characters
254 A second level of escaping is required when embedding the filter
255 description in a filtergraph description, in order to escape all the
256 filtergraph special characters. Thus the example above becomes:
258 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
260 (note that in addition to the @code{\'} escaping special characters,
261 also @code{,} needs to be escaped).
263 Finally an additional level of escaping is needed when writing the
264 filtergraph description in a shell command, which depends on the
265 escaping rules of the adopted shell. For example, assuming that
266 @code{\} is special and needs to be escaped with another @code{\}, the
267 previous string will finally result in:
269 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
272 @chapter Timeline editing
274 Some filters support a generic @option{enable} option. For the filters
275 supporting timeline editing, this option can be set to an expression which is
276 evaluated before sending a frame to the filter. If the evaluation is non-zero,
277 the filter will be enabled, otherwise the frame will be sent unchanged to the
278 next filter in the filtergraph.
280 The expression accepts the following values:
283 timestamp expressed in seconds, NAN if the input timestamp is unknown
286 sequential number of the input frame, starting from 0
289 the position in the file of the input frame, NAN if unknown
293 width and height of the input frame if video
296 Additionally, these filters support an @option{enable} command that can be used
297 to re-define the expression.
299 Like any other filtering option, the @option{enable} option follows the same
302 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
303 minutes, and a @ref{curves} filter starting at 3 seconds:
305 smartblur = enable='between(t,10,3*60)',
306 curves = enable='gte(t,3)' : preset=cross_process
309 @c man end FILTERGRAPH DESCRIPTION
311 @chapter Audio Filters
312 @c man begin AUDIO FILTERS
314 When you configure your FFmpeg build, you can disable any of the
315 existing filters using @code{--disable-filters}.
316 The configure output will show the audio filters included in your
319 Below is a description of the currently available audio filters.
323 A compressor is mainly used to reduce the dynamic range of a signal.
324 Especially modern music is mostly compressed at a high ratio to
325 improve the overall loudness. It's done to get the highest attention
326 of a listener, "fatten" the sound and bring more "power" to the track.
327 If a signal is compressed too much it may sound dull or "dead"
328 afterwards or it may start to "pump" (which could be a powerful effect
329 but can also destroy a track completely).
330 The right compression is the key to reach a professional sound and is
331 the high art of mixing and mastering. Because of its complex settings
332 it may take a long time to get the right feeling for this kind of effect.
334 Compression is done by detecting the volume above a chosen level
335 @code{threshold} and dividing it by the factor set with @code{ratio}.
336 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
337 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
338 the signal would cause distortion of the waveform the reduction can be
339 levelled over the time. This is done by setting "Attack" and "Release".
340 @code{attack} determines how long the signal has to rise above the threshold
341 before any reduction will occur and @code{release} sets the time the signal
342 has to fall below the threshold to reduce the reduction again. Shorter signals
343 than the chosen attack time will be left untouched.
344 The overall reduction of the signal can be made up afterwards with the
345 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
346 raising the makeup to this level results in a signal twice as loud than the
347 source. To gain a softer entry in the compression the @code{knee} flattens the
348 hard edge at the threshold in the range of the chosen decibels.
350 The filter accepts the following options:
354 Set input gain. Default is 1. Range is between 0.015625 and 64.
357 If a signal of second stream rises above this level it will affect the gain
358 reduction of the first stream.
359 By default it is 0.125. Range is between 0.00097563 and 1.
362 Set a ratio by which the signal is reduced. 1:2 means that if the level
363 rose 4dB above the threshold, it will be only 2dB above after the reduction.
364 Default is 2. Range is between 1 and 20.
367 Amount of milliseconds the signal has to rise above the threshold before gain
368 reduction starts. Default is 20. Range is between 0.01 and 2000.
371 Amount of milliseconds the signal has to fall below the threshold before
372 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
375 Set the amount by how much signal will be amplified after processing.
376 Default is 2. Range is from 1 and 64.
379 Curve the sharp knee around the threshold to enter gain reduction more softly.
380 Default is 2.82843. Range is between 1 and 8.
383 Choose if the @code{average} level between all channels of input stream
384 or the louder(@code{maximum}) channel of input stream affects the
385 reduction. Default is @code{average}.
388 Should the exact signal be taken in case of @code{peak} or an RMS one in case
389 of @code{rms}. Default is @code{rms} which is mostly smoother.
392 How much to use compressed signal in output. Default is 1.
393 Range is between 0 and 1.
398 Apply cross fade from one input audio stream to another input audio stream.
399 The cross fade is applied for specified duration near the end of first stream.
401 The filter accepts the following options:
405 Specify the number of samples for which the cross fade effect has to last.
406 At the end of the cross fade effect the first input audio will be completely
407 silent. Default is 44100.
410 Specify the duration of the cross fade effect. See
411 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
412 for the accepted syntax.
413 By default the duration is determined by @var{nb_samples}.
414 If set this option is used instead of @var{nb_samples}.
417 Should first stream end overlap with second stream start. Default is enabled.
420 Set curve for cross fade transition for first stream.
423 Set curve for cross fade transition for second stream.
425 For description of available curve types see @ref{afade} filter description.
432 Cross fade from one input to another:
434 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
438 Cross fade from one input to another but without overlapping:
440 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
446 Reduce audio bit resolution.
448 This filter is bit crusher with enhanced functionality. A bit crusher
449 is used to audibly reduce number of bits an audio signal is sampled
450 with. This doesn't change the bit depth at all, it just produces the
451 effect. Material reduced in bit depth sounds more harsh and "digital".
452 This filter is able to even round to continuous values instead of discrete
454 Additionally it has a D/C offset which results in different crushing of
455 the lower and the upper half of the signal.
456 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
458 Another feature of this filter is the logarithmic mode.
459 This setting switches from linear distances between bits to logarithmic ones.
460 The result is a much more "natural" sounding crusher which doesn't gate low
461 signals for example. The human ear has a logarithmic perception, too
462 so this kind of crushing is much more pleasant.
463 Logarithmic crushing is also able to get anti-aliased.
465 The filter accepts the following options:
481 Can be linear: @code{lin} or logarithmic: @code{log}.
490 Set sample reduction.
493 Enable LFO. By default disabled.
504 Delay one or more audio channels.
506 Samples in delayed channel are filled with silence.
508 The filter accepts the following option:
512 Set list of delays in milliseconds for each channel separated by '|'.
513 At least one delay greater than 0 should be provided.
514 Unused delays will be silently ignored. If number of given delays is
515 smaller than number of channels all remaining channels will not be delayed.
516 If you want to delay exact number of samples, append 'S' to number.
523 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
524 the second channel (and any other channels that may be present) unchanged.
530 Delay second channel by 500 samples, the third channel by 700 samples and leave
531 the first channel (and any other channels that may be present) unchanged.
539 Apply echoing to the input audio.
541 Echoes are reflected sound and can occur naturally amongst mountains
542 (and sometimes large buildings) when talking or shouting; digital echo
543 effects emulate this behaviour and are often used to help fill out the
544 sound of a single instrument or vocal. The time difference between the
545 original signal and the reflection is the @code{delay}, and the
546 loudness of the reflected signal is the @code{decay}.
547 Multiple echoes can have different delays and decays.
549 A description of the accepted parameters follows.
553 Set input gain of reflected signal. Default is @code{0.6}.
556 Set output gain of reflected signal. Default is @code{0.3}.
559 Set list of time intervals in milliseconds between original signal and reflections
560 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
561 Default is @code{1000}.
564 Set list of loudnesses of reflected signals separated by '|'.
565 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
566 Default is @code{0.5}.
573 Make it sound as if there are twice as many instruments as are actually playing:
575 aecho=0.8:0.88:60:0.4
579 If delay is very short, then it sound like a (metallic) robot playing music:
585 A longer delay will sound like an open air concert in the mountains:
587 aecho=0.8:0.9:1000:0.3
591 Same as above but with one more mountain:
593 aecho=0.8:0.9:1000|1800:0.3|0.25
598 Audio emphasis filter creates or restores material directly taken from LPs or
599 emphased CDs with different filter curves. E.g. to store music on vinyl the
600 signal has to be altered by a filter first to even out the disadvantages of
601 this recording medium.
602 Once the material is played back the inverse filter has to be applied to
603 restore the distortion of the frequency response.
605 The filter accepts the following options:
615 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
616 use @code{production} mode. Default is @code{reproduction} mode.
619 Set filter type. Selects medium. Can be one of the following:
631 select Compact Disc (CD).
637 select 50µs (FM-KF).
639 select 75µs (FM-KF).
645 Modify an audio signal according to the specified expressions.
647 This filter accepts one or more expressions (one for each channel),
648 which are evaluated and used to modify a corresponding audio signal.
650 It accepts the following parameters:
654 Set the '|'-separated expressions list for each separate channel. If
655 the number of input channels is greater than the number of
656 expressions, the last specified expression is used for the remaining
659 @item channel_layout, c
660 Set output channel layout. If not specified, the channel layout is
661 specified by the number of expressions. If set to @samp{same}, it will
662 use by default the same input channel layout.
665 Each expression in @var{exprs} can contain the following constants and functions:
669 channel number of the current expression
672 number of the evaluated sample, starting from 0
678 time of the evaluated sample expressed in seconds
681 @item nb_out_channels
682 input and output number of channels
685 the value of input channel with number @var{CH}
688 Note: this filter is slow. For faster processing you should use a
697 aeval=val(ch)/2:c=same
701 Invert phase of the second channel:
710 Apply fade-in/out effect to input audio.
712 A description of the accepted parameters follows.
716 Specify the effect type, can be either @code{in} for fade-in, or
717 @code{out} for a fade-out effect. Default is @code{in}.
719 @item start_sample, ss
720 Specify the number of the start sample for starting to apply the fade
721 effect. Default is 0.
724 Specify the number of samples for which the fade effect has to last. At
725 the end of the fade-in effect the output audio will have the same
726 volume as the input audio, at the end of the fade-out transition
727 the output audio will be silence. Default is 44100.
730 Specify the start time of the fade effect. Default is 0.
731 The value must be specified as a time duration; see
732 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
733 for the accepted syntax.
734 If set this option is used instead of @var{start_sample}.
737 Specify the duration of the fade effect. See
738 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
739 for the accepted syntax.
740 At the end of the fade-in effect the output audio will have the same
741 volume as the input audio, at the end of the fade-out transition
742 the output audio will be silence.
743 By default the duration is determined by @var{nb_samples}.
744 If set this option is used instead of @var{nb_samples}.
747 Set curve for fade transition.
749 It accepts the following values:
752 select triangular, linear slope (default)
754 select quarter of sine wave
756 select half of sine wave
758 select exponential sine wave
762 select inverted parabola
776 select inverted quarter of sine wave
778 select inverted half of sine wave
780 select double-exponential seat
782 select double-exponential sigmoid
790 Fade in first 15 seconds of audio:
796 Fade out last 25 seconds of a 900 seconds audio:
798 afade=t=out:st=875:d=25
803 Apply arbitrary expressions to samples in frequency domain.
807 Set frequency domain real expression for each separate channel separated
808 by '|'. Default is "1".
809 If the number of input channels is greater than the number of
810 expressions, the last specified expression is used for the remaining
814 Set frequency domain imaginary expression for each separate channel
815 separated by '|'. If not set, @var{real} option is used.
817 Each expression in @var{real} and @var{imag} can contain the following
825 current frequency bin number
828 number of available bins
831 channel number of the current expression
843 It accepts the following values:
859 Default is @code{w4096}
862 Set window function. Default is @code{hann}.
865 Set window overlap. If set to 1, the recommended overlap for selected
866 window function will be picked. Default is @code{0.75}.
873 Leave almost only low frequencies in audio:
875 afftfilt="1-clip((b/nb)*b,0,1)"
882 Set output format constraints for the input audio. The framework will
883 negotiate the most appropriate format to minimize conversions.
885 It accepts the following parameters:
889 A '|'-separated list of requested sample formats.
892 A '|'-separated list of requested sample rates.
894 @item channel_layouts
895 A '|'-separated list of requested channel layouts.
897 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
898 for the required syntax.
901 If a parameter is omitted, all values are allowed.
903 Force the output to either unsigned 8-bit or signed 16-bit stereo
905 aformat=sample_fmts=u8|s16:channel_layouts=stereo
910 A gate is mainly used to reduce lower parts of a signal. This kind of signal
911 processing reduces disturbing noise between useful signals.
913 Gating is done by detecting the volume below a chosen level @var{threshold}
914 and dividing it by the factor set with @var{ratio}. The bottom of the noise
915 floor is set via @var{range}. Because an exact manipulation of the signal
916 would cause distortion of the waveform the reduction can be levelled over
917 time. This is done by setting @var{attack} and @var{release}.
919 @var{attack} determines how long the signal has to fall below the threshold
920 before any reduction will occur and @var{release} sets the time the signal
921 has to rise above the threshold to reduce the reduction again.
922 Shorter signals than the chosen attack time will be left untouched.
926 Set input level before filtering.
927 Default is 1. Allowed range is from 0.015625 to 64.
930 Set the level of gain reduction when the signal is below the threshold.
931 Default is 0.06125. Allowed range is from 0 to 1.
934 If a signal rises above this level the gain reduction is released.
935 Default is 0.125. Allowed range is from 0 to 1.
938 Set a ratio by which the signal is reduced.
939 Default is 2. Allowed range is from 1 to 9000.
942 Amount of milliseconds the signal has to rise above the threshold before gain
944 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
947 Amount of milliseconds the signal has to fall below the threshold before the
948 reduction is increased again. Default is 250 milliseconds.
949 Allowed range is from 0.01 to 9000.
952 Set amount of amplification of signal after processing.
953 Default is 1. Allowed range is from 1 to 64.
956 Curve the sharp knee around the threshold to enter gain reduction more softly.
957 Default is 2.828427125. Allowed range is from 1 to 8.
960 Choose if exact signal should be taken for detection or an RMS like one.
961 Default is @code{rms}. Can be @code{peak} or @code{rms}.
964 Choose if the average level between all channels or the louder channel affects
966 Default is @code{average}. Can be @code{average} or @code{maximum}.
971 The limiter prevents an input signal from rising over a desired threshold.
972 This limiter uses lookahead technology to prevent your signal from distorting.
973 It means that there is a small delay after the signal is processed. Keep in mind
974 that the delay it produces is the attack time you set.
976 The filter accepts the following options:
980 Set input gain. Default is 1.
983 Set output gain. Default is 1.
986 Don't let signals above this level pass the limiter. Default is 1.
989 The limiter will reach its attenuation level in this amount of time in
990 milliseconds. Default is 5 milliseconds.
993 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
994 Default is 50 milliseconds.
997 When gain reduction is always needed ASC takes care of releasing to an
998 average reduction level rather than reaching a reduction of 0 in the release
1002 Select how much the release time is affected by ASC, 0 means nearly no changes
1003 in release time while 1 produces higher release times.
1006 Auto level output signal. Default is enabled.
1007 This normalizes audio back to 0dB if enabled.
1010 Depending on picked setting it is recommended to upsample input 2x or 4x times
1011 with @ref{aresample} before applying this filter.
1015 Apply a two-pole all-pass filter with central frequency (in Hz)
1016 @var{frequency}, and filter-width @var{width}.
1017 An all-pass filter changes the audio's frequency to phase relationship
1018 without changing its frequency to amplitude relationship.
1020 The filter accepts the following options:
1024 Set frequency in Hz.
1027 Set method to specify band-width of filter.
1040 Specify the band-width of a filter in width_type units.
1047 The filter accepts the following options:
1051 Set the number of loops.
1054 Set maximal number of samples.
1057 Set first sample of loop.
1063 Merge two or more audio streams into a single multi-channel stream.
1065 The filter accepts the following options:
1070 Set the number of inputs. Default is 2.
1074 If the channel layouts of the inputs are disjoint, and therefore compatible,
1075 the channel layout of the output will be set accordingly and the channels
1076 will be reordered as necessary. If the channel layouts of the inputs are not
1077 disjoint, the output will have all the channels of the first input then all
1078 the channels of the second input, in that order, and the channel layout of
1079 the output will be the default value corresponding to the total number of
1082 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1083 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1084 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1085 first input, b1 is the first channel of the second input).
1087 On the other hand, if both input are in stereo, the output channels will be
1088 in the default order: a1, a2, b1, b2, and the channel layout will be
1089 arbitrarily set to 4.0, which may or may not be the expected value.
1091 All inputs must have the same sample rate, and format.
1093 If inputs do not have the same duration, the output will stop with the
1096 @subsection Examples
1100 Merge two mono files into a stereo stream:
1102 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1106 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1108 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
1114 Mixes multiple audio inputs into a single output.
1116 Note that this filter only supports float samples (the @var{amerge}
1117 and @var{pan} audio filters support many formats). If the @var{amix}
1118 input has integer samples then @ref{aresample} will be automatically
1119 inserted to perform the conversion to float samples.
1123 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1125 will mix 3 input audio streams to a single output with the same duration as the
1126 first input and a dropout transition time of 3 seconds.
1128 It accepts the following parameters:
1132 The number of inputs. If unspecified, it defaults to 2.
1135 How to determine the end-of-stream.
1139 The duration of the longest input. (default)
1142 The duration of the shortest input.
1145 The duration of the first input.
1149 @item dropout_transition
1150 The transition time, in seconds, for volume renormalization when an input
1151 stream ends. The default value is 2 seconds.
1155 @section anequalizer
1157 High-order parametric multiband equalizer for each channel.
1159 It accepts the following parameters:
1163 This option string is in format:
1164 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1165 Each equalizer band is separated by '|'.
1169 Set channel number to which equalization will be applied.
1170 If input doesn't have that channel the entry is ignored.
1173 Set central frequency for band.
1174 If input doesn't have that frequency the entry is ignored.
1177 Set band width in hertz.
1180 Set band gain in dB.
1183 Set filter type for band, optional, can be:
1187 Butterworth, this is default.
1198 With this option activated frequency response of anequalizer is displayed
1202 Set video stream size. Only useful if curves option is activated.
1205 Set max gain that will be displayed. Only useful if curves option is activated.
1206 Setting this to a reasonable value makes it possible to display gain which is derived from
1207 neighbour bands which are too close to each other and thus produce higher gain
1208 when both are activated.
1211 Set frequency scale used to draw frequency response in video output.
1212 Can be linear or logarithmic. Default is logarithmic.
1215 Set color for each channel curve which is going to be displayed in video stream.
1216 This is list of color names separated by space or by '|'.
1217 Unrecognised or missing colors will be replaced by white color.
1220 @subsection Examples
1224 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1225 for first 2 channels using Chebyshev type 1 filter:
1227 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1231 @subsection Commands
1233 This filter supports the following commands:
1236 Alter existing filter parameters.
1237 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1239 @var{fN} is existing filter number, starting from 0, if no such filter is available
1241 @var{freq} set new frequency parameter.
1242 @var{width} set new width parameter in herz.
1243 @var{gain} set new gain parameter in dB.
1245 Full filter invocation with asendcmd may look like this:
1246 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1251 Pass the audio source unchanged to the output.
1255 Pad the end of an audio stream with silence.
1257 This can be used together with @command{ffmpeg} @option{-shortest} to
1258 extend audio streams to the same length as the video stream.
1260 A description of the accepted options follows.
1264 Set silence packet size. Default value is 4096.
1267 Set the number of samples of silence to add to the end. After the
1268 value is reached, the stream is terminated. This option is mutually
1269 exclusive with @option{whole_len}.
1272 Set the minimum total number of samples in the output audio stream. If
1273 the value is longer than the input audio length, silence is added to
1274 the end, until the value is reached. This option is mutually exclusive
1275 with @option{pad_len}.
1278 If neither the @option{pad_len} nor the @option{whole_len} option is
1279 set, the filter will add silence to the end of the input stream
1282 @subsection Examples
1286 Add 1024 samples of silence to the end of the input:
1292 Make sure the audio output will contain at least 10000 samples, pad
1293 the input with silence if required:
1295 apad=whole_len=10000
1299 Use @command{ffmpeg} to pad the audio input with silence, so that the
1300 video stream will always result the shortest and will be converted
1301 until the end in the output file when using the @option{shortest}
1304 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1309 Add a phasing effect to the input audio.
1311 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1312 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1314 A description of the accepted parameters follows.
1318 Set input gain. Default is 0.4.
1321 Set output gain. Default is 0.74
1324 Set delay in milliseconds. Default is 3.0.
1327 Set decay. Default is 0.4.
1330 Set modulation speed in Hz. Default is 0.5.
1333 Set modulation type. Default is triangular.
1335 It accepts the following values:
1344 Audio pulsator is something between an autopanner and a tremolo.
1345 But it can produce funny stereo effects as well. Pulsator changes the volume
1346 of the left and right channel based on a LFO (low frequency oscillator) with
1347 different waveforms and shifted phases.
1348 This filter have the ability to define an offset between left and right
1349 channel. An offset of 0 means that both LFO shapes match each other.
1350 The left and right channel are altered equally - a conventional tremolo.
1351 An offset of 50% means that the shape of the right channel is exactly shifted
1352 in phase (or moved backwards about half of the frequency) - pulsator acts as
1353 an autopanner. At 1 both curves match again. Every setting in between moves the
1354 phase shift gapless between all stages and produces some "bypassing" sounds with
1355 sine and triangle waveforms. The more you set the offset near 1 (starting from
1356 the 0.5) the faster the signal passes from the left to the right speaker.
1358 The filter accepts the following options:
1362 Set input gain. By default it is 1. Range is [0.015625 - 64].
1365 Set output gain. By default it is 1. Range is [0.015625 - 64].
1368 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1369 sawup or sawdown. Default is sine.
1372 Set modulation. Define how much of original signal is affected by the LFO.
1375 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1378 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1381 Set pulse width. Default is 1. Allowed range is [0 - 2].
1384 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1387 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1391 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1395 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1396 if timing is set to hz.
1402 Resample the input audio to the specified parameters, using the
1403 libswresample library. If none are specified then the filter will
1404 automatically convert between its input and output.
1406 This filter is also able to stretch/squeeze the audio data to make it match
1407 the timestamps or to inject silence / cut out audio to make it match the
1408 timestamps, do a combination of both or do neither.
1410 The filter accepts the syntax
1411 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1412 expresses a sample rate and @var{resampler_options} is a list of
1413 @var{key}=@var{value} pairs, separated by ":". See the
1414 ffmpeg-resampler manual for the complete list of supported options.
1416 @subsection Examples
1420 Resample the input audio to 44100Hz:
1426 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1427 samples per second compensation:
1429 aresample=async=1000
1435 Reverse an audio clip.
1437 Warning: This filter requires memory to buffer the entire clip, so trimming
1440 @subsection Examples
1444 Take the first 5 seconds of a clip, and reverse it.
1446 atrim=end=5,areverse
1450 @section asetnsamples
1452 Set the number of samples per each output audio frame.
1454 The last output packet may contain a different number of samples, as
1455 the filter will flush all the remaining samples when the input audio
1458 The filter accepts the following options:
1462 @item nb_out_samples, n
1463 Set the number of frames per each output audio frame. The number is
1464 intended as the number of samples @emph{per each channel}.
1465 Default value is 1024.
1468 If set to 1, the filter will pad the last audio frame with zeroes, so
1469 that the last frame will contain the same number of samples as the
1470 previous ones. Default value is 1.
1473 For example, to set the number of per-frame samples to 1234 and
1474 disable padding for the last frame, use:
1476 asetnsamples=n=1234:p=0
1481 Set the sample rate without altering the PCM data.
1482 This will result in a change of speed and pitch.
1484 The filter accepts the following options:
1487 @item sample_rate, r
1488 Set the output sample rate. Default is 44100 Hz.
1493 Show a line containing various information for each input audio frame.
1494 The input audio is not modified.
1496 The shown line contains a sequence of key/value pairs of the form
1497 @var{key}:@var{value}.
1499 The following values are shown in the output:
1503 The (sequential) number of the input frame, starting from 0.
1506 The presentation timestamp of the input frame, in time base units; the time base
1507 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1510 The presentation timestamp of the input frame in seconds.
1513 position of the frame in the input stream, -1 if this information in
1514 unavailable and/or meaningless (for example in case of synthetic audio)
1523 The sample rate for the audio frame.
1526 The number of samples (per channel) in the frame.
1529 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1530 audio, the data is treated as if all the planes were concatenated.
1532 @item plane_checksums
1533 A list of Adler-32 checksums for each data plane.
1539 Display time domain statistical information about the audio channels.
1540 Statistics are calculated and displayed for each audio channel and,
1541 where applicable, an overall figure is also given.
1543 It accepts the following option:
1546 Short window length in seconds, used for peak and trough RMS measurement.
1547 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1551 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1552 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1555 Available keys for each channel are:
1586 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1587 this @code{lavfi.astats.Overall.Peak_count}.
1589 For description what each key means read below.
1592 Set number of frame after which stats are going to be recalculated.
1593 Default is disabled.
1596 A description of each shown parameter follows:
1600 Mean amplitude displacement from zero.
1603 Minimal sample level.
1606 Maximal sample level.
1608 @item Min difference
1609 Minimal difference between two consecutive samples.
1611 @item Max difference
1612 Maximal difference between two consecutive samples.
1614 @item Mean difference
1615 Mean difference between two consecutive samples.
1616 The average of each difference between two consecutive samples.
1620 Standard peak and RMS level measured in dBFS.
1624 Peak and trough values for RMS level measured over a short window.
1627 Standard ratio of peak to RMS level (note: not in dB).
1630 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1631 (i.e. either @var{Min level} or @var{Max level}).
1634 Number of occasions (not the number of samples) that the signal attained either
1635 @var{Min level} or @var{Max level}.
1638 Overall bit depth of audio. Number of bits used for each sample.
1643 Synchronize audio data with timestamps by squeezing/stretching it and/or
1644 dropping samples/adding silence when needed.
1646 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
1648 It accepts the following parameters:
1652 Enable stretching/squeezing the data to make it match the timestamps. Disabled
1653 by default. When disabled, time gaps are covered with silence.
1656 The minimum difference between timestamps and audio data (in seconds) to trigger
1657 adding/dropping samples. The default value is 0.1. If you get an imperfect
1658 sync with this filter, try setting this parameter to 0.
1661 The maximum compensation in samples per second. Only relevant with compensate=1.
1662 The default value is 500.
1665 Assume that the first PTS should be this value. The time base is 1 / sample
1666 rate. This allows for padding/trimming at the start of the stream. By default,
1667 no assumption is made about the first frame's expected PTS, so no padding or
1668 trimming is done. For example, this could be set to 0 to pad the beginning with
1669 silence if an audio stream starts after the video stream or to trim any samples
1670 with a negative PTS due to encoder delay.
1678 The filter accepts exactly one parameter, the audio tempo. If not
1679 specified then the filter will assume nominal 1.0 tempo. Tempo must
1680 be in the [0.5, 2.0] range.
1682 @subsection Examples
1686 Slow down audio to 80% tempo:
1692 To speed up audio to 125% tempo:
1700 Trim the input so that the output contains one continuous subpart of the input.
1702 It accepts the following parameters:
1705 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1706 sample with the timestamp @var{start} will be the first sample in the output.
1709 Specify time of the first audio sample that will be dropped, i.e. the
1710 audio sample immediately preceding the one with the timestamp @var{end} will be
1711 the last sample in the output.
1714 Same as @var{start}, except this option sets the start timestamp in samples
1718 Same as @var{end}, except this option sets the end timestamp in samples instead
1722 The maximum duration of the output in seconds.
1725 The number of the first sample that should be output.
1728 The number of the first sample that should be dropped.
1731 @option{start}, @option{end}, and @option{duration} are expressed as time
1732 duration specifications; see
1733 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1735 Note that the first two sets of the start/end options and the @option{duration}
1736 option look at the frame timestamp, while the _sample options simply count the
1737 samples that pass through the filter. So start/end_pts and start/end_sample will
1738 give different results when the timestamps are wrong, inexact or do not start at
1739 zero. Also note that this filter does not modify the timestamps. If you wish
1740 to have the output timestamps start at zero, insert the asetpts filter after the
1743 If multiple start or end options are set, this filter tries to be greedy and
1744 keep all samples that match at least one of the specified constraints. To keep
1745 only the part that matches all the constraints at once, chain multiple atrim
1748 The defaults are such that all the input is kept. So it is possible to set e.g.
1749 just the end values to keep everything before the specified time.
1754 Drop everything except the second minute of input:
1756 ffmpeg -i INPUT -af atrim=60:120
1760 Keep only the first 1000 samples:
1762 ffmpeg -i INPUT -af atrim=end_sample=1000
1769 Apply a two-pole Butterworth band-pass filter with central
1770 frequency @var{frequency}, and (3dB-point) band-width width.
1771 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1772 instead of the default: constant 0dB peak gain.
1773 The filter roll off at 6dB per octave (20dB per decade).
1775 The filter accepts the following options:
1779 Set the filter's central frequency. Default is @code{3000}.
1782 Constant skirt gain if set to 1. Defaults to 0.
1785 Set method to specify band-width of filter.
1798 Specify the band-width of a filter in width_type units.
1803 Apply a two-pole Butterworth band-reject filter with central
1804 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1805 The filter roll off at 6dB per octave (20dB per decade).
1807 The filter accepts the following options:
1811 Set the filter's central frequency. Default is @code{3000}.
1814 Set method to specify band-width of filter.
1827 Specify the band-width of a filter in width_type units.
1832 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1833 shelving filter with a response similar to that of a standard
1834 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1836 The filter accepts the following options:
1840 Give the gain at 0 Hz. Its useful range is about -20
1841 (for a large cut) to +20 (for a large boost).
1842 Beware of clipping when using a positive gain.
1845 Set the filter's central frequency and so can be used
1846 to extend or reduce the frequency range to be boosted or cut.
1847 The default value is @code{100} Hz.
1850 Set method to specify band-width of filter.
1863 Determine how steep is the filter's shelf transition.
1868 Apply a biquad IIR filter with the given coefficients.
1869 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1870 are the numerator and denominator coefficients respectively.
1873 Bauer stereo to binaural transformation, which improves headphone listening of
1874 stereo audio records.
1876 It accepts the following parameters:
1880 Pre-defined crossfeed level.
1884 Default level (fcut=700, feed=50).
1887 Chu Moy circuit (fcut=700, feed=60).
1890 Jan Meier circuit (fcut=650, feed=95).
1895 Cut frequency (in Hz).
1904 Remap input channels to new locations.
1906 It accepts the following parameters:
1908 @item channel_layout
1909 The channel layout of the output stream.
1912 Map channels from input to output. The argument is a '|'-separated list of
1913 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1914 @var{in_channel} form. @var{in_channel} can be either the name of the input
1915 channel (e.g. FL for front left) or its index in the input channel layout.
1916 @var{out_channel} is the name of the output channel or its index in the output
1917 channel layout. If @var{out_channel} is not given then it is implicitly an
1918 index, starting with zero and increasing by one for each mapping.
1921 If no mapping is present, the filter will implicitly map input channels to
1922 output channels, preserving indices.
1924 For example, assuming a 5.1+downmix input MOV file,
1926 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1928 will create an output WAV file tagged as stereo from the downmix channels of
1931 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1933 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
1936 @section channelsplit
1938 Split each channel from an input audio stream into a separate output stream.
1940 It accepts the following parameters:
1942 @item channel_layout
1943 The channel layout of the input stream. The default is "stereo".
1946 For example, assuming a stereo input MP3 file,
1948 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1950 will create an output Matroska file with two audio streams, one containing only
1951 the left channel and the other the right channel.
1953 Split a 5.1 WAV file into per-channel files:
1955 ffmpeg -i in.wav -filter_complex
1956 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1957 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1958 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1963 Add a chorus effect to the audio.
1965 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
1967 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
1968 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
1969 The modulation depth defines the range the modulated delay is played before or after
1970 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
1971 sound tuned around the original one, like in a chorus where some vocals are slightly
1974 It accepts the following parameters:
1977 Set input gain. Default is 0.4.
1980 Set output gain. Default is 0.4.
1983 Set delays. A typical delay is around 40ms to 60ms.
1995 @subsection Examples
2001 chorus=0.7:0.9:55:0.4:0.25:2
2007 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2011 Fuller sounding chorus with three delays:
2013 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
2018 Compress or expand the audio's dynamic range.
2020 It accepts the following parameters:
2026 A list of times in seconds for each channel over which the instantaneous level
2027 of the input signal is averaged to determine its volume. @var{attacks} refers to
2028 increase of volume and @var{decays} refers to decrease of volume. For most
2029 situations, the attack time (response to the audio getting louder) should be
2030 shorter than the decay time, because the human ear is more sensitive to sudden
2031 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2032 a typical value for decay is 0.8 seconds.
2033 If specified number of attacks & decays is lower than number of channels, the last
2034 set attack/decay will be used for all remaining channels.
2037 A list of points for the transfer function, specified in dB relative to the
2038 maximum possible signal amplitude. Each key points list must be defined using
2039 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2040 @code{x0/y0 x1/y1 x2/y2 ....}
2042 The input values must be in strictly increasing order but the transfer function
2043 does not have to be monotonically rising. The point @code{0/0} is assumed but
2044 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2045 function are @code{-70/-70|-60/-20}.
2048 Set the curve radius in dB for all joints. It defaults to 0.01.
2051 Set the additional gain in dB to be applied at all points on the transfer
2052 function. This allows for easy adjustment of the overall gain.
2056 Set an initial volume, in dB, to be assumed for each channel when filtering
2057 starts. This permits the user to supply a nominal level initially, so that, for
2058 example, a very large gain is not applied to initial signal levels before the
2059 companding has begun to operate. A typical value for audio which is initially
2060 quiet is -90 dB. It defaults to 0.
2063 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2064 delayed before being fed to the volume adjuster. Specifying a delay
2065 approximately equal to the attack/decay times allows the filter to effectively
2066 operate in predictive rather than reactive mode. It defaults to 0.
2070 @subsection Examples
2074 Make music with both quiet and loud passages suitable for listening to in a
2077 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2080 Another example for audio with whisper and explosion parts:
2082 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2086 A noise gate for when the noise is at a lower level than the signal:
2088 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2092 Here is another noise gate, this time for when the noise is at a higher level
2093 than the signal (making it, in some ways, similar to squelch):
2095 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2099 2:1 compression starting at -6dB:
2101 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2105 2:1 compression starting at -9dB:
2107 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2111 2:1 compression starting at -12dB:
2113 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2117 2:1 compression starting at -18dB:
2119 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2123 3:1 compression starting at -15dB:
2125 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2131 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2137 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
2141 Hard limiter at -6dB:
2143 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2147 Hard limiter at -12dB:
2149 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2153 Hard noise gate at -35 dB:
2155 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2161 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2165 @section compensationdelay
2167 Compensation Delay Line is a metric based delay to compensate differing
2168 positions of microphones or speakers.
2170 For example, you have recorded guitar with two microphones placed in
2171 different location. Because the front of sound wave has fixed speed in
2172 normal conditions, the phasing of microphones can vary and depends on
2173 their location and interposition. The best sound mix can be achieved when
2174 these microphones are in phase (synchronized). Note that distance of
2175 ~30 cm between microphones makes one microphone to capture signal in
2176 antiphase to another microphone. That makes the final mix sounding moody.
2177 This filter helps to solve phasing problems by adding different delays
2178 to each microphone track and make them synchronized.
2180 The best result can be reached when you take one track as base and
2181 synchronize other tracks one by one with it.
2182 Remember that synchronization/delay tolerance depends on sample rate, too.
2183 Higher sample rates will give more tolerance.
2185 It accepts the following parameters:
2189 Set millimeters distance. This is compensation distance for fine tuning.
2193 Set cm distance. This is compensation distance for tightening distance setup.
2197 Set meters distance. This is compensation distance for hard distance setup.
2201 Set dry amount. Amount of unprocessed (dry) signal.
2205 Set wet amount. Amount of processed (wet) signal.
2209 Set temperature degree in Celsius. This is the temperature of the environment.
2213 @section crystalizer
2214 Simple algorithm to expand audio dynamic range.
2216 The filter accepts the following options:
2220 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2221 (unchanged sound) to 10.0 (maximum effect).
2224 Enable clipping. By default is enabled.
2228 Apply a DC shift to the audio.
2230 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2231 in the recording chain) from the audio. The effect of a DC offset is reduced
2232 headroom and hence volume. The @ref{astats} filter can be used to determine if
2233 a signal has a DC offset.
2237 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2241 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2242 used to prevent clipping.
2246 Dynamic Audio Normalizer.
2248 This filter applies a certain amount of gain to the input audio in order
2249 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2250 contrast to more "simple" normalization algorithms, the Dynamic Audio
2251 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2252 This allows for applying extra gain to the "quiet" sections of the audio
2253 while avoiding distortions or clipping the "loud" sections. In other words:
2254 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2255 sections, in the sense that the volume of each section is brought to the
2256 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2257 this goal *without* applying "dynamic range compressing". It will retain 100%
2258 of the dynamic range *within* each section of the audio file.
2262 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2263 Default is 500 milliseconds.
2264 The Dynamic Audio Normalizer processes the input audio in small chunks,
2265 referred to as frames. This is required, because a peak magnitude has no
2266 meaning for just a single sample value. Instead, we need to determine the
2267 peak magnitude for a contiguous sequence of sample values. While a "standard"
2268 normalizer would simply use the peak magnitude of the complete file, the
2269 Dynamic Audio Normalizer determines the peak magnitude individually for each
2270 frame. The length of a frame is specified in milliseconds. By default, the
2271 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2272 been found to give good results with most files.
2273 Note that the exact frame length, in number of samples, will be determined
2274 automatically, based on the sampling rate of the individual input audio file.
2277 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2278 number. Default is 31.
2279 Probably the most important parameter of the Dynamic Audio Normalizer is the
2280 @code{window size} of the Gaussian smoothing filter. The filter's window size
2281 is specified in frames, centered around the current frame. For the sake of
2282 simplicity, this must be an odd number. Consequently, the default value of 31
2283 takes into account the current frame, as well as the 15 preceding frames and
2284 the 15 subsequent frames. Using a larger window results in a stronger
2285 smoothing effect and thus in less gain variation, i.e. slower gain
2286 adaptation. Conversely, using a smaller window results in a weaker smoothing
2287 effect and thus in more gain variation, i.e. faster gain adaptation.
2288 In other words, the more you increase this value, the more the Dynamic Audio
2289 Normalizer will behave like a "traditional" normalization filter. On the
2290 contrary, the more you decrease this value, the more the Dynamic Audio
2291 Normalizer will behave like a dynamic range compressor.
2294 Set the target peak value. This specifies the highest permissible magnitude
2295 level for the normalized audio input. This filter will try to approach the
2296 target peak magnitude as closely as possible, but at the same time it also
2297 makes sure that the normalized signal will never exceed the peak magnitude.
2298 A frame's maximum local gain factor is imposed directly by the target peak
2299 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2300 It is not recommended to go above this value.
2303 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2304 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2305 factor for each input frame, i.e. the maximum gain factor that does not
2306 result in clipping or distortion. The maximum gain factor is determined by
2307 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2308 additionally bounds the frame's maximum gain factor by a predetermined
2309 (global) maximum gain factor. This is done in order to avoid excessive gain
2310 factors in "silent" or almost silent frames. By default, the maximum gain
2311 factor is 10.0, For most inputs the default value should be sufficient and
2312 it usually is not recommended to increase this value. Though, for input
2313 with an extremely low overall volume level, it may be necessary to allow even
2314 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2315 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2316 Instead, a "sigmoid" threshold function will be applied. This way, the
2317 gain factors will smoothly approach the threshold value, but never exceed that
2321 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2322 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2323 This means that the maximum local gain factor for each frame is defined
2324 (only) by the frame's highest magnitude sample. This way, the samples can
2325 be amplified as much as possible without exceeding the maximum signal
2326 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2327 Normalizer can also take into account the frame's root mean square,
2328 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2329 determine the power of a time-varying signal. It is therefore considered
2330 that the RMS is a better approximation of the "perceived loudness" than
2331 just looking at the signal's peak magnitude. Consequently, by adjusting all
2332 frames to a constant RMS value, a uniform "perceived loudness" can be
2333 established. If a target RMS value has been specified, a frame's local gain
2334 factor is defined as the factor that would result in exactly that RMS value.
2335 Note, however, that the maximum local gain factor is still restricted by the
2336 frame's highest magnitude sample, in order to prevent clipping.
2339 Enable channels coupling. By default is enabled.
2340 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2341 amount. This means the same gain factor will be applied to all channels, i.e.
2342 the maximum possible gain factor is determined by the "loudest" channel.
2343 However, in some recordings, it may happen that the volume of the different
2344 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2345 In this case, this option can be used to disable the channel coupling. This way,
2346 the gain factor will be determined independently for each channel, depending
2347 only on the individual channel's highest magnitude sample. This allows for
2348 harmonizing the volume of the different channels.
2351 Enable DC bias correction. By default is disabled.
2352 An audio signal (in the time domain) is a sequence of sample values.
2353 In the Dynamic Audio Normalizer these sample values are represented in the
2354 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2355 audio signal, or "waveform", should be centered around the zero point.
2356 That means if we calculate the mean value of all samples in a file, or in a
2357 single frame, then the result should be 0.0 or at least very close to that
2358 value. If, however, there is a significant deviation of the mean value from
2359 0.0, in either positive or negative direction, this is referred to as a
2360 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2361 Audio Normalizer provides optional DC bias correction.
2362 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2363 the mean value, or "DC correction" offset, of each input frame and subtract
2364 that value from all of the frame's sample values which ensures those samples
2365 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2366 boundaries, the DC correction offset values will be interpolated smoothly
2367 between neighbouring frames.
2370 Enable alternative boundary mode. By default is disabled.
2371 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2372 around each frame. This includes the preceding frames as well as the
2373 subsequent frames. However, for the "boundary" frames, located at the very
2374 beginning and at the very end of the audio file, not all neighbouring
2375 frames are available. In particular, for the first few frames in the audio
2376 file, the preceding frames are not known. And, similarly, for the last few
2377 frames in the audio file, the subsequent frames are not known. Thus, the
2378 question arises which gain factors should be assumed for the missing frames
2379 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2380 to deal with this situation. The default boundary mode assumes a gain factor
2381 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2382 "fade out" at the beginning and at the end of the input, respectively.
2385 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2386 By default, the Dynamic Audio Normalizer does not apply "traditional"
2387 compression. This means that signal peaks will not be pruned and thus the
2388 full dynamic range will be retained within each local neighbourhood. However,
2389 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2390 normalization algorithm with a more "traditional" compression.
2391 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2392 (thresholding) function. If (and only if) the compression feature is enabled,
2393 all input frames will be processed by a soft knee thresholding function prior
2394 to the actual normalization process. Put simply, the thresholding function is
2395 going to prune all samples whose magnitude exceeds a certain threshold value.
2396 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2397 value. Instead, the threshold value will be adjusted for each individual
2399 In general, smaller parameters result in stronger compression, and vice versa.
2400 Values below 3.0 are not recommended, because audible distortion may appear.
2405 Make audio easier to listen to on headphones.
2407 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2408 so that when listened to on headphones the stereo image is moved from
2409 inside your head (standard for headphones) to outside and in front of
2410 the listener (standard for speakers).
2416 Apply a two-pole peaking equalisation (EQ) filter. With this
2417 filter, the signal-level at and around a selected frequency can
2418 be increased or decreased, whilst (unlike bandpass and bandreject
2419 filters) that at all other frequencies is unchanged.
2421 In order to produce complex equalisation curves, this filter can
2422 be given several times, each with a different central frequency.
2424 The filter accepts the following options:
2428 Set the filter's central frequency in Hz.
2431 Set method to specify band-width of filter.
2444 Specify the band-width of a filter in width_type units.
2447 Set the required gain or attenuation in dB.
2448 Beware of clipping when using a positive gain.
2451 @subsection Examples
2454 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2456 equalizer=f=1000:width_type=h:width=200:g=-10
2460 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2462 equalizer=f=1000:width_type=q:width=1:g=2,equalizer=f=100:width_type=q:width=2:g=-5
2466 @section extrastereo
2468 Linearly increases the difference between left and right channels which
2469 adds some sort of "live" effect to playback.
2471 The filter accepts the following options:
2475 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2476 (average of both channels), with 1.0 sound will be unchanged, with
2477 -1.0 left and right channels will be swapped.
2480 Enable clipping. By default is enabled.
2483 @section firequalizer
2484 Apply FIR Equalization using arbitrary frequency response.
2486 The filter accepts the following option:
2490 Set gain curve equation (in dB). The expression can contain variables:
2493 the evaluated frequency
2497 channel number, set to 0 when multichannels evaluation is disabled
2499 channel id, see libavutil/channel_layout.h, set to the first channel id when
2500 multichannels evaluation is disabled
2504 channel_layout, see libavutil/channel_layout.h
2509 @item gain_interpolate(f)
2510 interpolate gain on frequency f based on gain_entry
2511 @item cubic_interpolate(f)
2512 same as gain_interpolate, but smoother
2514 This option is also available as command. Default is @code{gain_interpolate(f)}.
2517 Set gain entry for gain_interpolate function. The expression can
2521 store gain entry at frequency f with value g
2523 This option is also available as command.
2526 Set filter delay in seconds. Higher value means more accurate.
2527 Default is @code{0.01}.
2530 Set filter accuracy in Hz. Lower value means more accurate.
2531 Default is @code{5}.
2534 Set window function. Acceptable values are:
2537 rectangular window, useful when gain curve is already smooth
2539 hann window (default)
2545 3-terms continuous 1st derivative nuttall window
2547 minimum 3-terms discontinuous nuttall window
2549 4-terms continuous 1st derivative nuttall window
2551 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2553 blackman-harris window
2559 If enabled, use fixed number of audio samples. This improves speed when
2560 filtering with large delay. Default is disabled.
2563 Enable multichannels evaluation on gain. Default is disabled.
2566 Enable zero phase mode by subtracting timestamp to compensate delay.
2567 Default is disabled.
2570 Set scale used by gain. Acceptable values are:
2573 linear frequency, linear gain
2575 linear frequency, logarithmic (in dB) gain (default)
2577 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2579 logarithmic frequency, logarithmic gain
2583 Set file for dumping, suitable for gnuplot.
2586 Set scale for dumpfile. Acceptable values are same with scale option.
2590 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2591 Default is disabled.
2594 @subsection Examples
2599 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2602 lowpass at 1000 Hz with gain_entry:
2604 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2607 custom equalization:
2609 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2612 higher delay with zero phase to compensate delay:
2614 firequalizer=delay=0.1:fixed=on:zero_phase=on
2617 lowpass on left channel, highpass on right channel:
2619 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2620 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2625 Apply a flanging effect to the audio.
2627 The filter accepts the following options:
2631 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2634 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2637 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2641 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2642 Default value is 71.
2645 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2648 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2649 Default value is @var{sinusoidal}.
2652 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2653 Default value is 25.
2656 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2657 Default is @var{linear}.
2662 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2663 embedded HDCD codes is expanded into a 20-bit PCM stream.
2665 The filter supports the Peak Extend and Low-level Gain Adjustment features
2666 of HDCD, and detects the Transient Filter flag.
2669 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2672 When using the filter with wav, note the default encoding for wav is 16-bit,
2673 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2674 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2676 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2677 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2680 The filter accepts the following options:
2683 @item disable_autoconvert
2684 Disable any automatic format conversion or resampling in the filter graph.
2686 @item process_stereo
2687 Process the stereo channels together. If target_gain does not match between
2688 channels, consider it invalid and use the last valid target_gain.
2691 Set the code detect timer period in ms.
2694 Always extend peaks above -3dBFS even if PE isn't signaled.
2697 Replace audio with a solid tone and adjust the amplitude to signal some
2698 specific aspect of the decoding process. The output file can be loaded in
2699 an audio editor alongside the original to aid analysis.
2701 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2708 Gain adjustment level at each sample
2710 Samples where peak extend occurs
2712 Samples where the code detect timer is active
2714 Samples where the target gain does not match between channels
2720 Apply a high-pass filter with 3dB point frequency.
2721 The filter can be either single-pole, or double-pole (the default).
2722 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2724 The filter accepts the following options:
2728 Set frequency in Hz. Default is 3000.
2731 Set number of poles. Default is 2.
2734 Set method to specify band-width of filter.
2747 Specify the band-width of a filter in width_type units.
2748 Applies only to double-pole filter.
2749 The default is 0.707q and gives a Butterworth response.
2754 Join multiple input streams into one multi-channel stream.
2756 It accepts the following parameters:
2760 The number of input streams. It defaults to 2.
2762 @item channel_layout
2763 The desired output channel layout. It defaults to stereo.
2766 Map channels from inputs to output. The argument is a '|'-separated list of
2767 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2768 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2769 can be either the name of the input channel (e.g. FL for front left) or its
2770 index in the specified input stream. @var{out_channel} is the name of the output
2774 The filter will attempt to guess the mappings when they are not specified
2775 explicitly. It does so by first trying to find an unused matching input channel
2776 and if that fails it picks the first unused input channel.
2778 Join 3 inputs (with properly set channel layouts):
2780 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2783 Build a 5.1 output from 6 single-channel streams:
2785 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2786 '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'
2792 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2794 To enable compilation of this filter you need to configure FFmpeg with
2795 @code{--enable-ladspa}.
2799 Specifies the name of LADSPA plugin library to load. If the environment
2800 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2801 each one of the directories specified by the colon separated list in
2802 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2803 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2804 @file{/usr/lib/ladspa/}.
2807 Specifies the plugin within the library. Some libraries contain only
2808 one plugin, but others contain many of them. If this is not set filter
2809 will list all available plugins within the specified library.
2812 Set the '|' separated list of controls which are zero or more floating point
2813 values that determine the behavior of the loaded plugin (for example delay,
2815 Controls need to be defined using the following syntax:
2816 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2817 @var{valuei} is the value set on the @var{i}-th control.
2818 Alternatively they can be also defined using the following syntax:
2819 @var{value0}|@var{value1}|@var{value2}|..., where
2820 @var{valuei} is the value set on the @var{i}-th control.
2821 If @option{controls} is set to @code{help}, all available controls and
2822 their valid ranges are printed.
2824 @item sample_rate, s
2825 Specify the sample rate, default to 44100. Only used if plugin have
2829 Set the number of samples per channel per each output frame, default
2830 is 1024. Only used if plugin have zero inputs.
2833 Set the minimum duration of the sourced audio. See
2834 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2835 for the accepted syntax.
2836 Note that the resulting duration may be greater than the specified duration,
2837 as the generated audio is always cut at the end of a complete frame.
2838 If not specified, or the expressed duration is negative, the audio is
2839 supposed to be generated forever.
2840 Only used if plugin have zero inputs.
2844 @subsection Examples
2848 List all available plugins within amp (LADSPA example plugin) library:
2854 List all available controls and their valid ranges for @code{vcf_notch}
2855 plugin from @code{VCF} library:
2857 ladspa=f=vcf:p=vcf_notch:c=help
2861 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2864 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2868 Add reverberation to the audio using TAP-plugins
2869 (Tom's Audio Processing plugins):
2871 ladspa=file=tap_reverb:tap_reverb
2875 Generate white noise, with 0.2 amplitude:
2877 ladspa=file=cmt:noise_source_white:c=c0=.2
2881 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2882 @code{C* Audio Plugin Suite} (CAPS) library:
2884 ladspa=file=caps:Click:c=c1=20'
2888 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2890 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2894 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2895 @code{SWH Plugins} collection:
2897 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2901 Attenuate low frequencies using Multiband EQ from Steve Harris
2902 @code{SWH Plugins} collection:
2904 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2908 @subsection Commands
2910 This filter supports the following commands:
2913 Modify the @var{N}-th control value.
2915 If the specified value is not valid, it is ignored and prior one is kept.
2920 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2921 Support for both single pass (livestreams, files) and double pass (files) modes.
2922 This algorithm can target IL, LRA, and maximum true peak.
2924 To enable compilation of this filter you need to configure FFmpeg with
2925 @code{--enable-libebur128}.
2927 The filter accepts the following options:
2931 Set integrated loudness target.
2932 Range is -70.0 - -5.0. Default value is -24.0.
2935 Set loudness range target.
2936 Range is 1.0 - 20.0. Default value is 7.0.
2939 Set maximum true peak.
2940 Range is -9.0 - +0.0. Default value is -2.0.
2942 @item measured_I, measured_i
2943 Measured IL of input file.
2944 Range is -99.0 - +0.0.
2946 @item measured_LRA, measured_lra
2947 Measured LRA of input file.
2948 Range is 0.0 - 99.0.
2950 @item measured_TP, measured_tp
2951 Measured true peak of input file.
2952 Range is -99.0 - +99.0.
2954 @item measured_thresh
2955 Measured threshold of input file.
2956 Range is -99.0 - +0.0.
2959 Set offset gain. Gain is applied before the true-peak limiter.
2960 Range is -99.0 - +99.0. Default is +0.0.
2963 Normalize linearly if possible.
2964 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2965 to be specified in order to use this mode.
2966 Options are true or false. Default is true.
2969 Treat mono input files as "dual-mono". If a mono file is intended for playback
2970 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2971 If set to @code{true}, this option will compensate for this effect.
2972 Multi-channel input files are not affected by this option.
2973 Options are true or false. Default is false.
2976 Set print format for stats. Options are summary, json, or none.
2977 Default value is none.
2982 Apply a low-pass filter with 3dB point frequency.
2983 The filter can be either single-pole or double-pole (the default).
2984 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2986 The filter accepts the following options:
2990 Set frequency in Hz. Default is 500.
2993 Set number of poles. Default is 2.
2996 Set method to specify band-width of filter.
3009 Specify the band-width of a filter in width_type units.
3010 Applies only to double-pole filter.
3011 The default is 0.707q and gives a Butterworth response.
3017 Mix channels with specific gain levels. The filter accepts the output
3018 channel layout followed by a set of channels definitions.
3020 This filter is also designed to efficiently remap the channels of an audio
3023 The filter accepts parameters of the form:
3024 "@var{l}|@var{outdef}|@var{outdef}|..."
3028 output channel layout or number of channels
3031 output channel specification, of the form:
3032 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3035 output channel to define, either a channel name (FL, FR, etc.) or a channel
3036 number (c0, c1, etc.)
3039 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3042 input channel to use, see out_name for details; it is not possible to mix
3043 named and numbered input channels
3046 If the `=' in a channel specification is replaced by `<', then the gains for
3047 that specification will be renormalized so that the total is 1, thus
3048 avoiding clipping noise.
3050 @subsection Mixing examples
3052 For example, if you want to down-mix from stereo to mono, but with a bigger
3053 factor for the left channel:
3055 pan=1c|c0=0.9*c0+0.1*c1
3058 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3059 7-channels surround:
3061 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3064 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3065 that should be preferred (see "-ac" option) unless you have very specific
3068 @subsection Remapping examples
3070 The channel remapping will be effective if, and only if:
3073 @item gain coefficients are zeroes or ones,
3074 @item only one input per channel output,
3077 If all these conditions are satisfied, the filter will notify the user ("Pure
3078 channel mapping detected"), and use an optimized and lossless method to do the
3081 For example, if you have a 5.1 source and want a stereo audio stream by
3082 dropping the extra channels:
3084 pan="stereo| c0=FL | c1=FR"
3087 Given the same source, you can also switch front left and front right channels
3088 and keep the input channel layout:
3090 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3093 If the input is a stereo audio stream, you can mute the front left channel (and
3094 still keep the stereo channel layout) with:
3099 Still with a stereo audio stream input, you can copy the right channel in both
3100 front left and right:
3102 pan="stereo| c0=FR | c1=FR"
3107 ReplayGain scanner filter. This filter takes an audio stream as an input and
3108 outputs it unchanged.
3109 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3113 Convert the audio sample format, sample rate and channel layout. It is
3114 not meant to be used directly.
3117 Apply time-stretching and pitch-shifting with librubberband.
3119 The filter accepts the following options:
3123 Set tempo scale factor.
3126 Set pitch scale factor.
3129 Set transients detector.
3130 Possible values are:
3139 Possible values are:
3148 Possible values are:
3155 Set processing window size.
3156 Possible values are:
3165 Possible values are:
3172 Enable formant preservation when shift pitching.
3173 Possible values are:
3181 Possible values are:
3190 Possible values are:
3197 @section sidechaincompress
3199 This filter acts like normal compressor but has the ability to compress
3200 detected signal using second input signal.
3201 It needs two input streams and returns one output stream.
3202 First input stream will be processed depending on second stream signal.
3203 The filtered signal then can be filtered with other filters in later stages of
3204 processing. See @ref{pan} and @ref{amerge} filter.
3206 The filter accepts the following options:
3210 Set input gain. Default is 1. Range is between 0.015625 and 64.
3213 If a signal of second stream raises above this level it will affect the gain
3214 reduction of first stream.
3215 By default is 0.125. Range is between 0.00097563 and 1.
3218 Set a ratio about which the signal is reduced. 1:2 means that if the level
3219 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3220 Default is 2. Range is between 1 and 20.
3223 Amount of milliseconds the signal has to rise above the threshold before gain
3224 reduction starts. Default is 20. Range is between 0.01 and 2000.
3227 Amount of milliseconds the signal has to fall below the threshold before
3228 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3231 Set the amount by how much signal will be amplified after processing.
3232 Default is 2. Range is from 1 and 64.
3235 Curve the sharp knee around the threshold to enter gain reduction more softly.
3236 Default is 2.82843. Range is between 1 and 8.
3239 Choose if the @code{average} level between all channels of side-chain stream
3240 or the louder(@code{maximum}) channel of side-chain stream affects the
3241 reduction. Default is @code{average}.
3244 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3245 of @code{rms}. Default is @code{rms} which is mainly smoother.
3248 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3251 How much to use compressed signal in output. Default is 1.
3252 Range is between 0 and 1.
3255 @subsection Examples
3259 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3260 depending on the signal of 2nd input and later compressed signal to be
3261 merged with 2nd input:
3263 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3267 @section sidechaingate
3269 A sidechain gate acts like a normal (wideband) gate but has the ability to
3270 filter the detected signal before sending it to the gain reduction stage.
3271 Normally a gate uses the full range signal to detect a level above the
3273 For example: If you cut all lower frequencies from your sidechain signal
3274 the gate will decrease the volume of your track only if not enough highs
3275 appear. With this technique you are able to reduce the resonation of a
3276 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3278 It needs two input streams and returns one output stream.
3279 First input stream will be processed depending on second stream signal.
3281 The filter accepts the following options:
3285 Set input level before filtering.
3286 Default is 1. Allowed range is from 0.015625 to 64.
3289 Set the level of gain reduction when the signal is below the threshold.
3290 Default is 0.06125. Allowed range is from 0 to 1.
3293 If a signal rises above this level the gain reduction is released.
3294 Default is 0.125. Allowed range is from 0 to 1.
3297 Set a ratio about which the signal is reduced.
3298 Default is 2. Allowed range is from 1 to 9000.
3301 Amount of milliseconds the signal has to rise above the threshold before gain
3303 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3306 Amount of milliseconds the signal has to fall below the threshold before the
3307 reduction is increased again. Default is 250 milliseconds.
3308 Allowed range is from 0.01 to 9000.
3311 Set amount of amplification of signal after processing.
3312 Default is 1. Allowed range is from 1 to 64.
3315 Curve the sharp knee around the threshold to enter gain reduction more softly.
3316 Default is 2.828427125. Allowed range is from 1 to 8.
3319 Choose if exact signal should be taken for detection or an RMS like one.
3320 Default is rms. Can be peak or rms.
3323 Choose if the average level between all channels or the louder channel affects
3325 Default is average. Can be average or maximum.
3328 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3331 @section silencedetect
3333 Detect silence in an audio stream.
3335 This filter logs a message when it detects that the input audio volume is less
3336 or equal to a noise tolerance value for a duration greater or equal to the
3337 minimum detected noise duration.
3339 The printed times and duration are expressed in seconds.
3341 The filter accepts the following options:
3345 Set silence duration until notification (default is 2 seconds).
3348 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3349 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3352 @subsection Examples
3356 Detect 5 seconds of silence with -50dB noise tolerance:
3358 silencedetect=n=-50dB:d=5
3362 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3363 tolerance in @file{silence.mp3}:
3365 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3369 @section silenceremove
3371 Remove silence from the beginning, middle or end of the audio.
3373 The filter accepts the following options:
3377 This value is used to indicate if audio should be trimmed at beginning of
3378 the audio. A value of zero indicates no silence should be trimmed from the
3379 beginning. When specifying a non-zero value, it trims audio up until it
3380 finds non-silence. Normally, when trimming silence from beginning of audio
3381 the @var{start_periods} will be @code{1} but it can be increased to higher
3382 values to trim all audio up to specific count of non-silence periods.
3383 Default value is @code{0}.
3385 @item start_duration
3386 Specify the amount of time that non-silence must be detected before it stops
3387 trimming audio. By increasing the duration, bursts of noises can be treated
3388 as silence and trimmed off. Default value is @code{0}.
3390 @item start_threshold
3391 This indicates what sample value should be treated as silence. For digital
3392 audio, a value of @code{0} may be fine but for audio recorded from analog,
3393 you may wish to increase the value to account for background noise.
3394 Can be specified in dB (in case "dB" is appended to the specified value)
3395 or amplitude ratio. Default value is @code{0}.
3398 Set the count for trimming silence from the end of audio.
3399 To remove silence from the middle of a file, specify a @var{stop_periods}
3400 that is negative. This value is then treated as a positive value and is
3401 used to indicate the effect should restart processing as specified by
3402 @var{start_periods}, making it suitable for removing periods of silence
3403 in the middle of the audio.
3404 Default value is @code{0}.
3407 Specify a duration of silence that must exist before audio is not copied any
3408 more. By specifying a higher duration, silence that is wanted can be left in
3410 Default value is @code{0}.
3412 @item stop_threshold
3413 This is the same as @option{start_threshold} but for trimming silence from
3415 Can be specified in dB (in case "dB" is appended to the specified value)
3416 or amplitude ratio. Default value is @code{0}.
3419 This indicates that @var{stop_duration} length of audio should be left intact
3420 at the beginning of each period of silence.
3421 For example, if you want to remove long pauses between words but do not want
3422 to remove the pauses completely. Default value is @code{0}.
3425 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3426 and works better with digital silence which is exactly 0.
3427 Default value is @code{rms}.
3430 Set ratio used to calculate size of window for detecting silence.
3431 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3434 @subsection Examples
3438 The following example shows how this filter can be used to start a recording
3439 that does not contain the delay at the start which usually occurs between
3440 pressing the record button and the start of the performance:
3442 silenceremove=1:5:0.02
3446 Trim all silence encountered from beginning to end where there is more than 1
3447 second of silence in audio:
3449 silenceremove=0:0:0:-1:1:-90dB
3455 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3456 loudspeakers around the user for binaural listening via headphones (audio
3457 formats up to 9 channels supported).
3458 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3459 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3460 Austrian Academy of Sciences.
3462 To enable compilation of this filter you need to configure FFmpeg with
3463 @code{--enable-netcdf}.
3465 The filter accepts the following options:
3469 Set the SOFA file used for rendering.
3472 Set gain applied to audio. Value is in dB. Default is 0.
3475 Set rotation of virtual loudspeakers in deg. Default is 0.
3478 Set elevation of virtual speakers in deg. Default is 0.
3481 Set distance in meters between loudspeakers and the listener with near-field
3482 HRTFs. Default is 1.
3485 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3486 processing audio in time domain which is slow.
3487 @var{freq} is processing audio in frequency domain which is fast.
3488 Default is @var{freq}.
3491 Set custom positions of virtual loudspeakers. Syntax for this option is:
3492 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3493 Each virtual loudspeaker is described with short channel name following with
3494 azimuth and elevation in degreees.
3495 Each virtual loudspeaker description is separated by '|'.
3496 For example to override front left and front right channel positions use:
3497 'speakers=FL 45 15|FR 345 15'.
3498 Descriptions with unrecognised channel names are ignored.
3501 @subsection Examples
3505 Using ClubFritz6 sofa file:
3507 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3511 Using ClubFritz12 sofa file and bigger radius with small rotation:
3513 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3517 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3518 and also with custom gain:
3520 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3524 @section stereotools
3526 This filter has some handy utilities to manage stereo signals, for converting
3527 M/S stereo recordings to L/R signal while having control over the parameters
3528 or spreading the stereo image of master track.
3530 The filter accepts the following options:
3534 Set input level before filtering for both channels. Defaults is 1.
3535 Allowed range is from 0.015625 to 64.
3538 Set output level after filtering for both channels. Defaults is 1.
3539 Allowed range is from 0.015625 to 64.
3542 Set input balance between both channels. Default is 0.
3543 Allowed range is from -1 to 1.
3546 Set output balance between both channels. Default is 0.
3547 Allowed range is from -1 to 1.
3550 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3551 clipping. Disabled by default.
3554 Mute the left channel. Disabled by default.
3557 Mute the right channel. Disabled by default.
3560 Change the phase of the left channel. Disabled by default.
3563 Change the phase of the right channel. Disabled by default.
3566 Set stereo mode. Available values are:
3570 Left/Right to Left/Right, this is default.
3573 Left/Right to Mid/Side.
3576 Mid/Side to Left/Right.
3579 Left/Right to Left/Left.
3582 Left/Right to Right/Right.
3585 Left/Right to Left + Right.
3588 Left/Right to Right/Left.
3592 Set level of side signal. Default is 1.
3593 Allowed range is from 0.015625 to 64.
3596 Set balance of side signal. Default is 0.
3597 Allowed range is from -1 to 1.
3600 Set level of the middle signal. Default is 1.
3601 Allowed range is from 0.015625 to 64.
3604 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3607 Set stereo base between mono and inversed channels. Default is 0.
3608 Allowed range is from -1 to 1.
3611 Set delay in milliseconds how much to delay left from right channel and
3612 vice versa. Default is 0. Allowed range is from -20 to 20.
3615 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3618 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3621 @subsection Examples
3625 Apply karaoke like effect:
3627 stereotools=mlev=0.015625
3631 Convert M/S signal to L/R:
3633 "stereotools=mode=ms>lr"
3637 @section stereowiden
3639 This filter enhance the stereo effect by suppressing signal common to both
3640 channels and by delaying the signal of left into right and vice versa,
3641 thereby widening the stereo effect.
3643 The filter accepts the following options:
3647 Time in milliseconds of the delay of left signal into right and vice versa.
3648 Default is 20 milliseconds.
3651 Amount of gain in delayed signal into right and vice versa. Gives a delay
3652 effect of left signal in right output and vice versa which gives widening
3653 effect. Default is 0.3.
3656 Cross feed of left into right with inverted phase. This helps in suppressing
3657 the mono. If the value is 1 it will cancel all the signal common to both
3658 channels. Default is 0.3.
3661 Set level of input signal of original channel. Default is 0.8.
3666 Boost or cut treble (upper) frequencies of the audio using a two-pole
3667 shelving filter with a response similar to that of a standard
3668 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3670 The filter accepts the following options:
3674 Give the gain at whichever is the lower of ~22 kHz and the
3675 Nyquist frequency. Its useful range is about -20 (for a large cut)
3676 to +20 (for a large boost). Beware of clipping when using a positive gain.
3679 Set the filter's central frequency and so can be used
3680 to extend or reduce the frequency range to be boosted or cut.
3681 The default value is @code{3000} Hz.
3684 Set method to specify band-width of filter.
3697 Determine how steep is the filter's shelf transition.
3702 Sinusoidal amplitude modulation.
3704 The filter accepts the following options:
3708 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3709 (20 Hz or lower) will result in a tremolo effect.
3710 This filter may also be used as a ring modulator by specifying
3711 a modulation frequency higher than 20 Hz.
3712 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3715 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3716 Default value is 0.5.
3721 Sinusoidal phase modulation.
3723 The filter accepts the following options:
3727 Modulation frequency in Hertz.
3728 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3731 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3732 Default value is 0.5.
3737 Adjust the input audio volume.
3739 It accepts the following parameters:
3743 Set audio volume expression.
3745 Output values are clipped to the maximum value.
3747 The output audio volume is given by the relation:
3749 @var{output_volume} = @var{volume} * @var{input_volume}
3752 The default value for @var{volume} is "1.0".
3755 This parameter represents the mathematical precision.
3757 It determines which input sample formats will be allowed, which affects the
3758 precision of the volume scaling.
3762 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3764 32-bit floating-point; this limits input sample format to FLT. (default)
3766 64-bit floating-point; this limits input sample format to DBL.
3770 Choose the behaviour on encountering ReplayGain side data in input frames.
3774 Remove ReplayGain side data, ignoring its contents (the default).
3777 Ignore ReplayGain side data, but leave it in the frame.
3780 Prefer the track gain, if present.
3783 Prefer the album gain, if present.
3786 @item replaygain_preamp
3787 Pre-amplification gain in dB to apply to the selected replaygain gain.
3789 Default value for @var{replaygain_preamp} is 0.0.
3792 Set when the volume expression is evaluated.
3794 It accepts the following values:
3797 only evaluate expression once during the filter initialization, or
3798 when the @samp{volume} command is sent
3801 evaluate expression for each incoming frame
3804 Default value is @samp{once}.
3807 The volume expression can contain the following parameters.
3811 frame number (starting at zero)
3814 @item nb_consumed_samples
3815 number of samples consumed by the filter
3817 number of samples in the current frame
3819 original frame position in the file
3825 PTS at start of stream
3827 time at start of stream
3833 last set volume value
3836 Note that when @option{eval} is set to @samp{once} only the
3837 @var{sample_rate} and @var{tb} variables are available, all other
3838 variables will evaluate to NAN.
3840 @subsection Commands
3842 This filter supports the following commands:
3845 Modify the volume expression.
3846 The command accepts the same syntax of the corresponding option.
3848 If the specified expression is not valid, it is kept at its current
3850 @item replaygain_noclip
3851 Prevent clipping by limiting the gain applied.
3853 Default value for @var{replaygain_noclip} is 1.
3857 @subsection Examples
3861 Halve the input audio volume:
3865 volume=volume=-6.0206dB
3868 In all the above example the named key for @option{volume} can be
3869 omitted, for example like in:
3875 Increase input audio power by 6 decibels using fixed-point precision:
3877 volume=volume=6dB:precision=fixed
3881 Fade volume after time 10 with an annihilation period of 5 seconds:
3883 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3887 @section volumedetect
3889 Detect the volume of the input video.
3891 The filter has no parameters. The input is not modified. Statistics about
3892 the volume will be printed in the log when the input stream end is reached.
3894 In particular it will show the mean volume (root mean square), maximum
3895 volume (on a per-sample basis), and the beginning of a histogram of the
3896 registered volume values (from the maximum value to a cumulated 1/1000 of
3899 All volumes are in decibels relative to the maximum PCM value.
3901 @subsection Examples
3903 Here is an excerpt of the output:
3905 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3906 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3907 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3908 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3909 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3910 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3911 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3912 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3913 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3919 The mean square energy is approximately -27 dB, or 10^-2.7.
3921 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3923 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3926 In other words, raising the volume by +4 dB does not cause any clipping,
3927 raising it by +5 dB causes clipping for 6 samples, etc.
3929 @c man end AUDIO FILTERS
3931 @chapter Audio Sources
3932 @c man begin AUDIO SOURCES
3934 Below is a description of the currently available audio sources.
3938 Buffer audio frames, and make them available to the filter chain.
3940 This source is mainly intended for a programmatic use, in particular
3941 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3943 It accepts the following parameters:
3947 The timebase which will be used for timestamps of submitted frames. It must be
3948 either a floating-point number or in @var{numerator}/@var{denominator} form.
3951 The sample rate of the incoming audio buffers.
3954 The sample format of the incoming audio buffers.
3955 Either a sample format name or its corresponding integer representation from
3956 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3958 @item channel_layout
3959 The channel layout of the incoming audio buffers.
3960 Either a channel layout name from channel_layout_map in
3961 @file{libavutil/channel_layout.c} or its corresponding integer representation
3962 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3965 The number of channels of the incoming audio buffers.
3966 If both @var{channels} and @var{channel_layout} are specified, then they
3971 @subsection Examples
3974 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3977 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3978 Since the sample format with name "s16p" corresponds to the number
3979 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3982 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3987 Generate an audio signal specified by an expression.
3989 This source accepts in input one or more expressions (one for each
3990 channel), which are evaluated and used to generate a corresponding
3993 This source accepts the following options:
3997 Set the '|'-separated expressions list for each separate channel. In case the
3998 @option{channel_layout} option is not specified, the selected channel layout
3999 depends on the number of provided expressions. Otherwise the last
4000 specified expression is applied to the remaining output channels.
4002 @item channel_layout, c
4003 Set the channel layout. The number of channels in the specified layout
4004 must be equal to the number of specified expressions.
4007 Set the minimum duration of the sourced audio. See
4008 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4009 for the accepted syntax.
4010 Note that the resulting duration may be greater than the specified
4011 duration, as the generated audio is always cut at the end of a
4014 If not specified, or the expressed duration is negative, the audio is
4015 supposed to be generated forever.
4018 Set the number of samples per channel per each output frame,
4021 @item sample_rate, s
4022 Specify the sample rate, default to 44100.
4025 Each expression in @var{exprs} can contain the following constants:
4029 number of the evaluated sample, starting from 0
4032 time of the evaluated sample expressed in seconds, starting from 0
4039 @subsection Examples
4049 Generate a sin signal with frequency of 440 Hz, set sample rate to
4052 aevalsrc="sin(440*2*PI*t):s=8000"
4056 Generate a two channels signal, specify the channel layout (Front
4057 Center + Back Center) explicitly:
4059 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4063 Generate white noise:
4065 aevalsrc="-2+random(0)"
4069 Generate an amplitude modulated signal:
4071 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4075 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4077 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4084 The null audio source, return unprocessed audio frames. It is mainly useful
4085 as a template and to be employed in analysis / debugging tools, or as
4086 the source for filters which ignore the input data (for example the sox
4089 This source accepts the following options:
4093 @item channel_layout, cl
4095 Specifies the channel layout, and can be either an integer or a string
4096 representing a channel layout. The default value of @var{channel_layout}
4099 Check the channel_layout_map definition in
4100 @file{libavutil/channel_layout.c} for the mapping between strings and
4101 channel layout values.
4103 @item sample_rate, r
4104 Specifies the sample rate, and defaults to 44100.
4107 Set the number of samples per requested frames.
4111 @subsection Examples
4115 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4117 anullsrc=r=48000:cl=4
4121 Do the same operation with a more obvious syntax:
4123 anullsrc=r=48000:cl=mono
4127 All the parameters need to be explicitly defined.
4131 Synthesize a voice utterance using the libflite library.
4133 To enable compilation of this filter you need to configure FFmpeg with
4134 @code{--enable-libflite}.
4136 Note that the flite library is not thread-safe.
4138 The filter accepts the following options:
4143 If set to 1, list the names of the available voices and exit
4144 immediately. Default value is 0.
4147 Set the maximum number of samples per frame. Default value is 512.
4150 Set the filename containing the text to speak.
4153 Set the text to speak.
4156 Set the voice to use for the speech synthesis. Default value is
4157 @code{kal}. See also the @var{list_voices} option.
4160 @subsection Examples
4164 Read from file @file{speech.txt}, and synthesize the text using the
4165 standard flite voice:
4167 flite=textfile=speech.txt
4171 Read the specified text selecting the @code{slt} voice:
4173 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4177 Input text to ffmpeg:
4179 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4183 Make @file{ffplay} speak the specified text, using @code{flite} and
4184 the @code{lavfi} device:
4186 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4190 For more information about libflite, check:
4191 @url{http://www.speech.cs.cmu.edu/flite/}
4195 Generate a noise audio signal.
4197 The filter accepts the following options:
4200 @item sample_rate, r
4201 Specify the sample rate. Default value is 48000 Hz.
4204 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4208 Specify the duration of the generated audio stream. Not specifying this option
4209 results in noise with an infinite length.
4211 @item color, colour, c
4212 Specify the color of noise. Available noise colors are white, pink, and brown.
4213 Default color is white.
4216 Specify a value used to seed the PRNG.
4219 Set the number of samples per each output frame, default is 1024.
4222 @subsection Examples
4227 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4229 anoisesrc=d=60:c=pink:r=44100:a=0.5
4235 Generate an audio signal made of a sine wave with amplitude 1/8.
4237 The audio signal is bit-exact.
4239 The filter accepts the following options:
4244 Set the carrier frequency. Default is 440 Hz.
4246 @item beep_factor, b
4247 Enable a periodic beep every second with frequency @var{beep_factor} times
4248 the carrier frequency. Default is 0, meaning the beep is disabled.
4250 @item sample_rate, r
4251 Specify the sample rate, default is 44100.
4254 Specify the duration of the generated audio stream.
4256 @item samples_per_frame
4257 Set the number of samples per output frame.
4259 The expression can contain the following constants:
4263 The (sequential) number of the output audio frame, starting from 0.
4266 The PTS (Presentation TimeStamp) of the output audio frame,
4267 expressed in @var{TB} units.
4270 The PTS of the output audio frame, expressed in seconds.
4273 The timebase of the output audio frames.
4276 Default is @code{1024}.
4279 @subsection Examples
4284 Generate a simple 440 Hz sine wave:
4290 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4294 sine=frequency=220:beep_factor=4:duration=5
4298 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4301 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4305 @c man end AUDIO SOURCES
4307 @chapter Audio Sinks
4308 @c man begin AUDIO SINKS
4310 Below is a description of the currently available audio sinks.
4312 @section abuffersink
4314 Buffer audio frames, and make them available to the end of filter chain.
4316 This sink is mainly intended for programmatic use, in particular
4317 through the interface defined in @file{libavfilter/buffersink.h}
4318 or the options system.
4320 It accepts a pointer to an AVABufferSinkContext structure, which
4321 defines the incoming buffers' formats, to be passed as the opaque
4322 parameter to @code{avfilter_init_filter} for initialization.
4325 Null audio sink; do absolutely nothing with the input audio. It is
4326 mainly useful as a template and for use in analysis / debugging
4329 @c man end AUDIO SINKS
4331 @chapter Video Filters
4332 @c man begin VIDEO FILTERS
4334 When you configure your FFmpeg build, you can disable any of the
4335 existing filters using @code{--disable-filters}.
4336 The configure output will show the video filters included in your
4339 Below is a description of the currently available video filters.
4341 @section alphaextract
4343 Extract the alpha component from the input as a grayscale video. This
4344 is especially useful with the @var{alphamerge} filter.
4348 Add or replace the alpha component of the primary input with the
4349 grayscale value of a second input. This is intended for use with
4350 @var{alphaextract} to allow the transmission or storage of frame
4351 sequences that have alpha in a format that doesn't support an alpha
4354 For example, to reconstruct full frames from a normal YUV-encoded video
4355 and a separate video created with @var{alphaextract}, you might use:
4357 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4360 Since this filter is designed for reconstruction, it operates on frame
4361 sequences without considering timestamps, and terminates when either
4362 input reaches end of stream. This will cause problems if your encoding
4363 pipeline drops frames. If you're trying to apply an image as an
4364 overlay to a video stream, consider the @var{overlay} filter instead.
4368 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4369 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4370 Substation Alpha) subtitles files.
4372 This filter accepts the following option in addition to the common options from
4373 the @ref{subtitles} filter:
4377 Set the shaping engine
4379 Available values are:
4382 The default libass shaping engine, which is the best available.
4384 Fast, font-agnostic shaper that can do only substitutions
4386 Slower shaper using OpenType for substitutions and positioning
4389 The default is @code{auto}.
4393 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4395 The filter accepts the following options:
4399 Set threshold A for 1st plane. Default is 0.02.
4400 Valid range is 0 to 0.3.
4403 Set threshold B for 1st plane. Default is 0.04.
4404 Valid range is 0 to 5.
4407 Set threshold A for 2nd plane. Default is 0.02.
4408 Valid range is 0 to 0.3.
4411 Set threshold B for 2nd plane. Default is 0.04.
4412 Valid range is 0 to 5.
4415 Set threshold A for 3rd plane. Default is 0.02.
4416 Valid range is 0 to 0.3.
4419 Set threshold B for 3rd plane. Default is 0.04.
4420 Valid range is 0 to 5.
4422 Threshold A is designed to react on abrupt changes in the input signal and
4423 threshold B is designed to react on continuous changes in the input signal.
4426 Set number of frames filter will use for averaging. Default is 33. Must be odd
4427 number in range [5, 129].
4430 Set what planes of frame filter will use for averaging. Default is all.
4435 Apply average blur filter.
4437 The filter accepts the following options:
4441 Set horizontal kernel size.
4444 Set which planes to filter. By default all planes are filtered.
4447 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4448 Default is @code{0}.
4453 Compute the bounding box for the non-black pixels in the input frame
4456 This filter computes the bounding box containing all the pixels with a
4457 luminance value greater than the minimum allowed value.
4458 The parameters describing the bounding box are printed on the filter
4461 The filter accepts the following option:
4465 Set the minimal luminance value. Default is @code{16}.
4468 @section bitplanenoise
4470 Show and measure bit plane noise.
4472 The filter accepts the following options:
4476 Set which plane to analyze. Default is @code{1}.
4479 Filter out noisy pixels from @code{bitplane} set above.
4480 Default is disabled.
4483 @section blackdetect
4485 Detect video intervals that are (almost) completely black. Can be
4486 useful to detect chapter transitions, commercials, or invalid
4487 recordings. Output lines contains the time for the start, end and
4488 duration of the detected black interval expressed in seconds.
4490 In order to display the output lines, you need to set the loglevel at
4491 least to the AV_LOG_INFO value.
4493 The filter accepts the following options:
4496 @item black_min_duration, d
4497 Set the minimum detected black duration expressed in seconds. It must
4498 be a non-negative floating point number.
4500 Default value is 2.0.
4502 @item picture_black_ratio_th, pic_th
4503 Set the threshold for considering a picture "black".
4504 Express the minimum value for the ratio:
4506 @var{nb_black_pixels} / @var{nb_pixels}
4509 for which a picture is considered black.
4510 Default value is 0.98.
4512 @item pixel_black_th, pix_th
4513 Set the threshold for considering a pixel "black".
4515 The threshold expresses the maximum pixel luminance value for which a
4516 pixel is considered "black". The provided value is scaled according to
4517 the following equation:
4519 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4522 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4523 the input video format, the range is [0-255] for YUV full-range
4524 formats and [16-235] for YUV non full-range formats.
4526 Default value is 0.10.
4529 The following example sets the maximum pixel threshold to the minimum
4530 value, and detects only black intervals of 2 or more seconds:
4532 blackdetect=d=2:pix_th=0.00
4537 Detect frames that are (almost) completely black. Can be useful to
4538 detect chapter transitions or commercials. Output lines consist of
4539 the frame number of the detected frame, the percentage of blackness,
4540 the position in the file if known or -1 and the timestamp in seconds.
4542 In order to display the output lines, you need to set the loglevel at
4543 least to the AV_LOG_INFO value.
4545 It accepts the following parameters:
4550 The percentage of the pixels that have to be below the threshold; it defaults to
4553 @item threshold, thresh
4554 The threshold below which a pixel value is considered black; it defaults to
4559 @section blend, tblend
4561 Blend two video frames into each other.
4563 The @code{blend} filter takes two input streams and outputs one
4564 stream, the first input is the "top" layer and second input is
4565 "bottom" layer. By default, the output terminates when the longest input terminates.
4567 The @code{tblend} (time blend) filter takes two consecutive frames
4568 from one single stream, and outputs the result obtained by blending
4569 the new frame on top of the old frame.
4571 A description of the accepted options follows.
4579 Set blend mode for specific pixel component or all pixel components in case
4580 of @var{all_mode}. Default value is @code{normal}.
4582 Available values for component modes are:
4623 Set blend opacity for specific pixel component or all pixel components in case
4624 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4631 Set blend expression for specific pixel component or all pixel components in case
4632 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4634 The expressions can use the following variables:
4638 The sequential number of the filtered frame, starting from @code{0}.
4642 the coordinates of the current sample
4646 the width and height of currently filtered plane
4650 Width and height scale depending on the currently filtered plane. It is the
4651 ratio between the corresponding luma plane number of pixels and the current
4652 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4653 @code{0.5,0.5} for chroma planes.
4656 Time of the current frame, expressed in seconds.
4659 Value of pixel component at current location for first video frame (top layer).
4662 Value of pixel component at current location for second video frame (bottom layer).
4666 Force termination when the shortest input terminates. Default is
4667 @code{0}. This option is only defined for the @code{blend} filter.
4670 Continue applying the last bottom frame after the end of the stream. A value of
4671 @code{0} disable the filter after the last frame of the bottom layer is reached.
4672 Default is @code{1}. This option is only defined for the @code{blend} filter.
4675 @subsection Examples
4679 Apply transition from bottom layer to top layer in first 10 seconds:
4681 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4685 Apply 1x1 checkerboard effect:
4687 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4691 Apply uncover left effect:
4693 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4697 Apply uncover down effect:
4699 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4703 Apply uncover up-left effect:
4705 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4709 Split diagonally video and shows top and bottom layer on each side:
4711 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4715 Display differences between the current and the previous frame:
4717 tblend=all_mode=difference128
4723 Apply a boxblur algorithm to the input video.
4725 It accepts the following parameters:
4729 @item luma_radius, lr
4730 @item luma_power, lp
4731 @item chroma_radius, cr
4732 @item chroma_power, cp
4733 @item alpha_radius, ar
4734 @item alpha_power, ap
4738 A description of the accepted options follows.
4741 @item luma_radius, lr
4742 @item chroma_radius, cr
4743 @item alpha_radius, ar
4744 Set an expression for the box radius in pixels used for blurring the
4745 corresponding input plane.
4747 The radius value must be a non-negative number, and must not be
4748 greater than the value of the expression @code{min(w,h)/2} for the
4749 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4752 Default value for @option{luma_radius} is "2". If not specified,
4753 @option{chroma_radius} and @option{alpha_radius} default to the
4754 corresponding value set for @option{luma_radius}.
4756 The expressions can contain the following constants:
4760 The input width and height in pixels.
4764 The input chroma image width and height in pixels.
4768 The horizontal and vertical chroma subsample values. For example, for the
4769 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4772 @item luma_power, lp
4773 @item chroma_power, cp
4774 @item alpha_power, ap
4775 Specify how many times the boxblur filter is applied to the
4776 corresponding plane.
4778 Default value for @option{luma_power} is 2. If not specified,
4779 @option{chroma_power} and @option{alpha_power} default to the
4780 corresponding value set for @option{luma_power}.
4782 A value of 0 will disable the effect.
4785 @subsection Examples
4789 Apply a boxblur filter with the luma, chroma, and alpha radii
4792 boxblur=luma_radius=2:luma_power=1
4797 Set the luma radius to 2, and alpha and chroma radius to 0:
4799 boxblur=2:1:cr=0:ar=0
4803 Set the luma and chroma radii to a fraction of the video dimension:
4805 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4811 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4812 Deinterlacing Filter").
4814 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4815 interpolation algorithms.
4816 It accepts the following parameters:
4820 The interlacing mode to adopt. It accepts one of the following values:
4824 Output one frame for each frame.
4826 Output one frame for each field.
4829 The default value is @code{send_field}.
4832 The picture field parity assumed for the input interlaced video. It accepts one
4833 of the following values:
4837 Assume the top field is first.
4839 Assume the bottom field is first.
4841 Enable automatic detection of field parity.
4844 The default value is @code{auto}.
4845 If the interlacing is unknown or the decoder does not export this information,
4846 top field first will be assumed.
4849 Specify which frames to deinterlace. Accept one of the following
4854 Deinterlace all frames.
4856 Only deinterlace frames marked as interlaced.
4859 The default value is @code{all}.
4863 YUV colorspace color/chroma keying.
4865 The filter accepts the following options:
4869 The color which will be replaced with transparency.
4872 Similarity percentage with the key color.
4874 0.01 matches only the exact key color, while 1.0 matches everything.
4879 0.0 makes pixels either fully transparent, or not transparent at all.
4881 Higher values result in semi-transparent pixels, with a higher transparency
4882 the more similar the pixels color is to the key color.
4885 Signals that the color passed is already in YUV instead of RGB.
4887 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4888 This can be used to pass exact YUV values as hexadecimal numbers.
4891 @subsection Examples
4895 Make every green pixel in the input image transparent:
4897 ffmpeg -i input.png -vf chromakey=green out.png
4901 Overlay a greenscreen-video on top of a static black background.
4903 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
4909 Display CIE color diagram with pixels overlaid onto it.
4911 The filter accepts the following options:
4926 @item uhdtv, rec2020
4939 Set what gamuts to draw.
4941 See @code{system} option for available values.
4944 Set ciescope size, by default set to 512.
4947 Set intensity used to map input pixel values to CIE diagram.
4950 Set contrast used to draw tongue colors that are out of active color system gamut.
4953 Correct gamma displayed on scope, by default enabled.
4956 Show white point on CIE diagram, by default disabled.
4959 Set input gamma. Used only with XYZ input color space.
4964 Visualize information exported by some codecs.
4966 Some codecs can export information through frames using side-data or other
4967 means. For example, some MPEG based codecs export motion vectors through the
4968 @var{export_mvs} flag in the codec @option{flags2} option.
4970 The filter accepts the following option:
4974 Set motion vectors to visualize.
4976 Available flags for @var{mv} are:
4980 forward predicted MVs of P-frames
4982 forward predicted MVs of B-frames
4984 backward predicted MVs of B-frames
4988 Display quantization parameters using the chroma planes.
4991 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4993 Available flags for @var{mv_type} are:
4997 forward predicted MVs
4999 backward predicted MVs
5002 @item frame_type, ft
5003 Set frame type to visualize motion vectors of.
5005 Available flags for @var{frame_type} are:
5009 intra-coded frames (I-frames)
5011 predicted frames (P-frames)
5013 bi-directionally predicted frames (B-frames)
5017 @subsection Examples
5021 Visualize forward predicted MVs of all frames using @command{ffplay}:
5023 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5027 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5029 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5033 @section colorbalance
5034 Modify intensity of primary colors (red, green and blue) of input frames.
5036 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5037 regions for the red-cyan, green-magenta or blue-yellow balance.
5039 A positive adjustment value shifts the balance towards the primary color, a negative
5040 value towards the complementary color.
5042 The filter accepts the following options:
5048 Adjust red, green and blue shadows (darkest pixels).
5053 Adjust red, green and blue midtones (medium pixels).
5058 Adjust red, green and blue highlights (brightest pixels).
5060 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5063 @subsection Examples
5067 Add red color cast to shadows:
5074 RGB colorspace color keying.
5076 The filter accepts the following options:
5080 The color which will be replaced with transparency.
5083 Similarity percentage with the key color.
5085 0.01 matches only the exact key color, while 1.0 matches everything.
5090 0.0 makes pixels either fully transparent, or not transparent at all.
5092 Higher values result in semi-transparent pixels, with a higher transparency
5093 the more similar the pixels color is to the key color.
5096 @subsection Examples
5100 Make every green pixel in the input image transparent:
5102 ffmpeg -i input.png -vf colorkey=green out.png
5106 Overlay a greenscreen-video on top of a static background image.
5108 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
5112 @section colorlevels
5114 Adjust video input frames using levels.
5116 The filter accepts the following options:
5123 Adjust red, green, blue and alpha input black point.
5124 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5130 Adjust red, green, blue and alpha input white point.
5131 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5133 Input levels are used to lighten highlights (bright tones), darken shadows
5134 (dark tones), change the balance of bright and dark tones.
5140 Adjust red, green, blue and alpha output black point.
5141 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5147 Adjust red, green, blue and alpha output white point.
5148 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5150 Output levels allows manual selection of a constrained output level range.
5153 @subsection Examples
5157 Make video output darker:
5159 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5165 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5169 Make video output lighter:
5171 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5175 Increase brightness:
5177 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5181 @section colorchannelmixer
5183 Adjust video input frames by re-mixing color channels.
5185 This filter modifies a color channel by adding the values associated to
5186 the other channels of the same pixels. For example if the value to
5187 modify is red, the output value will be:
5189 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5192 The filter accepts the following options:
5199 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5200 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5206 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5207 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5213 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5214 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5220 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5221 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5223 Allowed ranges for options are @code{[-2.0, 2.0]}.
5226 @subsection Examples
5230 Convert source to grayscale:
5232 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5235 Simulate sepia tones:
5237 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5241 @section colormatrix
5243 Convert color matrix.
5245 The filter accepts the following options:
5250 Specify the source and destination color matrix. Both values must be
5253 The accepted values are:
5272 For example to convert from BT.601 to SMPTE-240M, use the command:
5274 colormatrix=bt601:smpte240m
5279 Convert colorspace, transfer characteristics or color primaries.
5281 The filter accepts the following options:
5286 Specify all color properties at once.
5288 The accepted values are:
5318 Specify output colorspace.
5320 The accepted values are:
5329 BT.470BG or BT.601-6 625
5332 SMPTE-170M or BT.601-6 525
5338 BT.2020 with non-constant luminance
5344 Specify output transfer characteristics.
5346 The accepted values are:
5352 Constant gamma of 2.2
5355 Constant gamma of 2.8
5358 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5364 BT.2020 for 10-bits content
5367 BT.2020 for 12-bits content
5373 Specify output color primaries.
5375 The accepted values are:
5384 BT.470BG or BT.601-6 625
5387 SMPTE-170M or BT.601-6 525
5399 Specify output color range.
5401 The accepted values are:
5404 MPEG (restricted) range
5412 Specify output color format.
5414 The accepted values are:
5417 YUV 4:2:0 planar 8-bits
5420 YUV 4:2:0 planar 10-bits
5423 YUV 4:2:0 planar 12-bits
5426 YUV 4:2:2 planar 8-bits
5429 YUV 4:2:2 planar 10-bits
5432 YUV 4:2:2 planar 12-bits
5435 YUV 4:4:4 planar 8-bits
5438 YUV 4:4:4 planar 10-bits
5441 YUV 4:4:4 planar 12-bits
5446 Do a fast conversion, which skips gamma/primary correction. This will take
5447 significantly less CPU, but will be mathematically incorrect. To get output
5448 compatible with that produced by the colormatrix filter, use fast=1.
5451 Specify dithering mode.
5453 The accepted values are:
5459 Floyd-Steinberg dithering
5463 Whitepoint adaptation mode.
5465 The accepted values are:
5468 Bradford whitepoint adaptation
5471 von Kries whitepoint adaptation
5474 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5478 Override all input properties at once. Same accepted values as @ref{all}.
5481 Override input colorspace. Same accepted values as @ref{space}.
5484 Override input color primaries. Same accepted values as @ref{primaries}.
5487 Override input transfer characteristics. Same accepted values as @ref{trc}.
5490 Override input color range. Same accepted values as @ref{range}.
5494 The filter converts the transfer characteristics, color space and color
5495 primaries to the specified user values. The output value, if not specified,
5496 is set to a default value based on the "all" property. If that property is
5497 also not specified, the filter will log an error. The output color range and
5498 format default to the same value as the input color range and format. The
5499 input transfer characteristics, color space, color primaries and color range
5500 should be set on the input data. If any of these are missing, the filter will
5501 log an error and no conversion will take place.
5503 For example to convert the input to SMPTE-240M, use the command:
5505 colorspace=smpte240m
5508 @section convolution
5510 Apply convolution 3x3 or 5x5 filter.
5512 The filter accepts the following options:
5519 Set matrix for each plane.
5520 Matrix is sequence of 9 or 25 signed integers.
5526 Set multiplier for calculated value for each plane.
5532 Set bias for each plane. This value is added to the result of the multiplication.
5533 Useful for making the overall image brighter or darker. Default is 0.0.
5536 @subsection Examples
5542 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"
5548 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"
5554 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"
5560 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"
5566 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"
5572 Copy the input source unchanged to the output. This is mainly useful for
5577 Video filtering on GPU using Apple's CoreImage API on OSX.
5579 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5580 processed by video hardware. However, software-based OpenGL implementations
5581 exist which means there is no guarantee for hardware processing. It depends on
5584 There are many filters and image generators provided by Apple that come with a
5585 large variety of options. The filter has to be referenced by its name along
5588 The coreimage filter accepts the following options:
5591 List all available filters and generators along with all their respective
5592 options as well as possible minimum and maximum values along with the default
5599 Specify all filters by their respective name and options.
5600 Use @var{list_filters} to determine all valid filter names and options.
5601 Numerical options are specified by a float value and are automatically clamped
5602 to their respective value range. Vector and color options have to be specified
5603 by a list of space separated float values. Character escaping has to be done.
5604 A special option name @code{default} is available to use default options for a
5607 It is required to specify either @code{default} or at least one of the filter options.
5608 All omitted options are used with their default values.
5609 The syntax of the filter string is as follows:
5611 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5615 Specify a rectangle where the output of the filter chain is copied into the
5616 input image. It is given by a list of space separated float values:
5618 output_rect=x\ y\ width\ height
5620 If not given, the output rectangle equals the dimensions of the input image.
5621 The output rectangle is automatically cropped at the borders of the input
5622 image. Negative values are valid for each component.
5624 output_rect=25\ 25\ 100\ 100
5628 Several filters can be chained for successive processing without GPU-HOST
5629 transfers allowing for fast processing of complex filter chains.
5630 Currently, only filters with zero (generators) or exactly one (filters) input
5631 image and one output image are supported. Also, transition filters are not yet
5634 Some filters generate output images with additional padding depending on the
5635 respective filter kernel. The padding is automatically removed to ensure the
5636 filter output has the same size as the input image.
5638 For image generators, the size of the output image is determined by the
5639 previous output image of the filter chain or the input image of the whole
5640 filterchain, respectively. The generators do not use the pixel information of
5641 this image to generate their output. However, the generated output is
5642 blended onto this image, resulting in partial or complete coverage of the
5645 The @ref{coreimagesrc} video source can be used for generating input images
5646 which are directly fed into the filter chain. By using it, providing input
5647 images by another video source or an input video is not required.
5649 @subsection Examples
5654 List all filters available:
5656 coreimage=list_filters=true
5660 Use the CIBoxBlur filter with default options to blur an image:
5662 coreimage=filter=CIBoxBlur@@default
5666 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5667 its center at 100x100 and a radius of 50 pixels:
5669 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5673 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5674 given as complete and escaped command-line for Apple's standard bash shell:
5676 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5682 Crop the input video to given dimensions.
5684 It accepts the following parameters:
5688 The width of the output video. It defaults to @code{iw}.
5689 This expression is evaluated only once during the filter
5690 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5693 The height of the output video. It defaults to @code{ih}.
5694 This expression is evaluated only once during the filter
5695 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5698 The horizontal position, in the input video, of the left edge of the output
5699 video. It defaults to @code{(in_w-out_w)/2}.
5700 This expression is evaluated per-frame.
5703 The vertical position, in the input video, of the top edge of the output video.
5704 It defaults to @code{(in_h-out_h)/2}.
5705 This expression is evaluated per-frame.
5708 If set to 1 will force the output display aspect ratio
5709 to be the same of the input, by changing the output sample aspect
5710 ratio. It defaults to 0.
5713 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5714 width/height/x/y as specified and will not be rounded to nearest smaller value.
5718 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5719 expressions containing the following constants:
5724 The computed values for @var{x} and @var{y}. They are evaluated for
5729 The input width and height.
5733 These are the same as @var{in_w} and @var{in_h}.
5737 The output (cropped) width and height.
5741 These are the same as @var{out_w} and @var{out_h}.
5744 same as @var{iw} / @var{ih}
5747 input sample aspect ratio
5750 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5754 horizontal and vertical chroma subsample values. For example for the
5755 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5758 The number of the input frame, starting from 0.
5761 the position in the file of the input frame, NAN if unknown
5764 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5768 The expression for @var{out_w} may depend on the value of @var{out_h},
5769 and the expression for @var{out_h} may depend on @var{out_w}, but they
5770 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5771 evaluated after @var{out_w} and @var{out_h}.
5773 The @var{x} and @var{y} parameters specify the expressions for the
5774 position of the top-left corner of the output (non-cropped) area. They
5775 are evaluated for each frame. If the evaluated value is not valid, it
5776 is approximated to the nearest valid value.
5778 The expression for @var{x} may depend on @var{y}, and the expression
5779 for @var{y} may depend on @var{x}.
5781 @subsection Examples
5785 Crop area with size 100x100 at position (12,34).
5790 Using named options, the example above becomes:
5792 crop=w=100:h=100:x=12:y=34
5796 Crop the central input area with size 100x100:
5802 Crop the central input area with size 2/3 of the input video:
5804 crop=2/3*in_w:2/3*in_h
5808 Crop the input video central square:
5815 Delimit the rectangle with the top-left corner placed at position
5816 100:100 and the right-bottom corner corresponding to the right-bottom
5817 corner of the input image.
5819 crop=in_w-100:in_h-100:100:100
5823 Crop 10 pixels from the left and right borders, and 20 pixels from
5824 the top and bottom borders
5826 crop=in_w-2*10:in_h-2*20
5830 Keep only the bottom right quarter of the input image:
5832 crop=in_w/2:in_h/2:in_w/2:in_h/2
5836 Crop height for getting Greek harmony:
5838 crop=in_w:1/PHI*in_w
5842 Apply trembling effect:
5844 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)
5848 Apply erratic camera effect depending on timestamp:
5850 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)"
5854 Set x depending on the value of y:
5856 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5860 @subsection Commands
5862 This filter supports the following commands:
5868 Set width/height of the output video and the horizontal/vertical position
5870 The command accepts the same syntax of the corresponding option.
5872 If the specified expression is not valid, it is kept at its current
5878 Auto-detect the crop size.
5880 It calculates the necessary cropping parameters and prints the
5881 recommended parameters via the logging system. The detected dimensions
5882 correspond to the non-black area of the input video.
5884 It accepts the following parameters:
5889 Set higher black value threshold, which can be optionally specified
5890 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5891 value greater to the set value is considered non-black. It defaults to 24.
5892 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5893 on the bitdepth of the pixel format.
5896 The value which the width/height should be divisible by. It defaults to
5897 16. The offset is automatically adjusted to center the video. Use 2 to
5898 get only even dimensions (needed for 4:2:2 video). 16 is best when
5899 encoding to most video codecs.
5901 @item reset_count, reset
5902 Set the counter that determines after how many frames cropdetect will
5903 reset the previously detected largest video area and start over to
5904 detect the current optimal crop area. Default value is 0.
5906 This can be useful when channel logos distort the video area. 0
5907 indicates 'never reset', and returns the largest area encountered during
5914 Apply color adjustments using curves.
5916 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5917 component (red, green and blue) has its values defined by @var{N} key points
5918 tied from each other using a smooth curve. The x-axis represents the pixel
5919 values from the input frame, and the y-axis the new pixel values to be set for
5922 By default, a component curve is defined by the two points @var{(0;0)} and
5923 @var{(1;1)}. This creates a straight line where each original pixel value is
5924 "adjusted" to its own value, which means no change to the image.
5926 The filter allows you to redefine these two points and add some more. A new
5927 curve (using a natural cubic spline interpolation) will be define to pass
5928 smoothly through all these new coordinates. The new defined points needs to be
5929 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5930 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5931 the vector spaces, the values will be clipped accordingly.
5933 The filter accepts the following options:
5937 Select one of the available color presets. This option can be used in addition
5938 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5939 options takes priority on the preset values.
5940 Available presets are:
5943 @item color_negative
5946 @item increase_contrast
5948 @item linear_contrast
5949 @item medium_contrast
5951 @item strong_contrast
5954 Default is @code{none}.
5956 Set the master key points. These points will define a second pass mapping. It
5957 is sometimes called a "luminance" or "value" mapping. It can be used with
5958 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5959 post-processing LUT.
5961 Set the key points for the red component.
5963 Set the key points for the green component.
5965 Set the key points for the blue component.
5967 Set the key points for all components (not including master).
5968 Can be used in addition to the other key points component
5969 options. In this case, the unset component(s) will fallback on this
5970 @option{all} setting.
5972 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5974 Save Gnuplot script of the curves in specified file.
5977 To avoid some filtergraph syntax conflicts, each key points list need to be
5978 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5980 @subsection Examples
5984 Increase slightly the middle level of blue:
5986 curves=blue='0/0 0.5/0.58 1/1'
5992 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'
5994 Here we obtain the following coordinates for each components:
5997 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5999 @code{(0;0) (0.50;0.48) (1;1)}
6001 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6005 The previous example can also be achieved with the associated built-in preset:
6007 curves=preset=vintage
6017 Use a Photoshop preset and redefine the points of the green component:
6019 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6023 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6024 and @command{gnuplot}:
6026 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6027 gnuplot -p /tmp/curves.plt
6033 Video data analysis filter.
6035 This filter shows hexadecimal pixel values of part of video.
6037 The filter accepts the following options:
6041 Set output video size.
6044 Set x offset from where to pick pixels.
6047 Set y offset from where to pick pixels.
6050 Set scope mode, can be one of the following:
6053 Draw hexadecimal pixel values with white color on black background.
6056 Draw hexadecimal pixel values with input video pixel color on black
6060 Draw hexadecimal pixel values on color background picked from input video,
6061 the text color is picked in such way so its always visible.
6065 Draw rows and columns numbers on left and top of video.
6068 Set background opacity.
6073 Denoise frames using 2D DCT (frequency domain filtering).
6075 This filter is not designed for real time.
6077 The filter accepts the following options:
6081 Set the noise sigma constant.
6083 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6084 coefficient (absolute value) below this threshold with be dropped.
6086 If you need a more advanced filtering, see @option{expr}.
6088 Default is @code{0}.
6091 Set number overlapping pixels for each block. Since the filter can be slow, you
6092 may want to reduce this value, at the cost of a less effective filter and the
6093 risk of various artefacts.
6095 If the overlapping value doesn't permit processing the whole input width or
6096 height, a warning will be displayed and according borders won't be denoised.
6098 Default value is @var{blocksize}-1, which is the best possible setting.
6101 Set the coefficient factor expression.
6103 For each coefficient of a DCT block, this expression will be evaluated as a
6104 multiplier value for the coefficient.
6106 If this is option is set, the @option{sigma} option will be ignored.
6108 The absolute value of the coefficient can be accessed through the @var{c}
6112 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6113 @var{blocksize}, which is the width and height of the processed blocks.
6115 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6116 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6117 on the speed processing. Also, a larger block size does not necessarily means a
6121 @subsection Examples
6123 Apply a denoise with a @option{sigma} of @code{4.5}:
6128 The same operation can be achieved using the expression system:
6130 dctdnoiz=e='gte(c, 4.5*3)'
6133 Violent denoise using a block size of @code{16x16}:
6140 Remove banding artifacts from input video.
6141 It works by replacing banded pixels with average value of referenced pixels.
6143 The filter accepts the following options:
6150 Set banding detection threshold for each plane. Default is 0.02.
6151 Valid range is 0.00003 to 0.5.
6152 If difference between current pixel and reference pixel is less than threshold,
6153 it will be considered as banded.
6156 Banding detection range in pixels. Default is 16. If positive, random number
6157 in range 0 to set value will be used. If negative, exact absolute value
6159 The range defines square of four pixels around current pixel.
6162 Set direction in radians from which four pixel will be compared. If positive,
6163 random direction from 0 to set direction will be picked. If negative, exact of
6164 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6165 will pick only pixels on same row and -PI/2 will pick only pixels on same
6169 If enabled, current pixel is compared with average value of all four
6170 surrounding pixels. The default is enabled. If disabled current pixel is
6171 compared with all four surrounding pixels. The pixel is considered banded
6172 if only all four differences with surrounding pixels are less than threshold.
6178 Drop duplicated frames at regular intervals.
6180 The filter accepts the following options:
6184 Set the number of frames from which one will be dropped. Setting this to
6185 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6186 Default is @code{5}.
6189 Set the threshold for duplicate detection. If the difference metric for a frame
6190 is less than or equal to this value, then it is declared as duplicate. Default
6194 Set scene change threshold. Default is @code{15}.
6198 Set the size of the x and y-axis blocks used during metric calculations.
6199 Larger blocks give better noise suppression, but also give worse detection of
6200 small movements. Must be a power of two. Default is @code{32}.
6203 Mark main input as a pre-processed input and activate clean source input
6204 stream. This allows the input to be pre-processed with various filters to help
6205 the metrics calculation while keeping the frame selection lossless. When set to
6206 @code{1}, the first stream is for the pre-processed input, and the second
6207 stream is the clean source from where the kept frames are chosen. Default is
6211 Set whether or not chroma is considered in the metric calculations. Default is
6217 Apply deflate effect to the video.
6219 This filter replaces the pixel by the local(3x3) average by taking into account
6220 only values lower than the pixel.
6222 It accepts the following options:
6229 Limit the maximum change for each plane, default is 65535.
6230 If 0, plane will remain unchanged.
6235 Remove judder produced by partially interlaced telecined content.
6237 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6238 source was partially telecined content then the output of @code{pullup,dejudder}
6239 will have a variable frame rate. May change the recorded frame rate of the
6240 container. Aside from that change, this filter will not affect constant frame
6243 The option available in this filter is:
6247 Specify the length of the window over which the judder repeats.
6249 Accepts any integer greater than 1. Useful values are:
6253 If the original was telecined from 24 to 30 fps (Film to NTSC).
6256 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6259 If a mixture of the two.
6262 The default is @samp{4}.
6267 Suppress a TV station logo by a simple interpolation of the surrounding
6268 pixels. Just set a rectangle covering the logo and watch it disappear
6269 (and sometimes something even uglier appear - your mileage may vary).
6271 It accepts the following parameters:
6276 Specify the top left corner coordinates of the logo. They must be
6281 Specify the width and height of the logo to clear. They must be
6285 Specify the thickness of the fuzzy edge of the rectangle (added to
6286 @var{w} and @var{h}). The default value is 1. This option is
6287 deprecated, setting higher values should no longer be necessary and
6291 When set to 1, a green rectangle is drawn on the screen to simplify
6292 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6293 The default value is 0.
6295 The rectangle is drawn on the outermost pixels which will be (partly)
6296 replaced with interpolated values. The values of the next pixels
6297 immediately outside this rectangle in each direction will be used to
6298 compute the interpolated pixel values inside the rectangle.
6302 @subsection Examples
6306 Set a rectangle covering the area with top left corner coordinates 0,0
6307 and size 100x77, and a band of size 10:
6309 delogo=x=0:y=0:w=100:h=77:band=10
6316 Attempt to fix small changes in horizontal and/or vertical shift. This
6317 filter helps remove camera shake from hand-holding a camera, bumping a
6318 tripod, moving on a vehicle, etc.
6320 The filter accepts the following options:
6328 Specify a rectangular area where to limit the search for motion
6330 If desired the search for motion vectors can be limited to a
6331 rectangular area of the frame defined by its top left corner, width
6332 and height. These parameters have the same meaning as the drawbox
6333 filter which can be used to visualise the position of the bounding
6336 This is useful when simultaneous movement of subjects within the frame
6337 might be confused for camera motion by the motion vector search.
6339 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6340 then the full frame is used. This allows later options to be set
6341 without specifying the bounding box for the motion vector search.
6343 Default - search the whole frame.
6347 Specify the maximum extent of movement in x and y directions in the
6348 range 0-64 pixels. Default 16.
6351 Specify how to generate pixels to fill blanks at the edge of the
6352 frame. Available values are:
6355 Fill zeroes at blank locations
6357 Original image at blank locations
6359 Extruded edge value at blank locations
6361 Mirrored edge at blank locations
6363 Default value is @samp{mirror}.
6366 Specify the blocksize to use for motion search. Range 4-128 pixels,
6370 Specify the contrast threshold for blocks. Only blocks with more than
6371 the specified contrast (difference between darkest and lightest
6372 pixels) will be considered. Range 1-255, default 125.
6375 Specify the search strategy. Available values are:
6378 Set exhaustive search
6380 Set less exhaustive search.
6382 Default value is @samp{exhaustive}.
6385 If set then a detailed log of the motion search is written to the
6389 If set to 1, specify using OpenCL capabilities, only available if
6390 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6396 Apply an exact inverse of the telecine operation. It requires a predefined
6397 pattern specified using the pattern option which must be the same as that passed
6398 to the telecine filter.
6400 This filter accepts the following options:
6409 The default value is @code{top}.
6413 A string of numbers representing the pulldown pattern you wish to apply.
6414 The default value is @code{23}.
6417 A number representing position of the first frame with respect to the telecine
6418 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6423 Apply dilation effect to the video.
6425 This filter replaces the pixel by the local(3x3) maximum.
6427 It accepts the following options:
6434 Limit the maximum change for each plane, default is 65535.
6435 If 0, plane will remain unchanged.
6438 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6441 Flags to local 3x3 coordinates maps like this:
6450 Displace pixels as indicated by second and third input stream.
6452 It takes three input streams and outputs one stream, the first input is the
6453 source, and second and third input are displacement maps.
6455 The second input specifies how much to displace pixels along the
6456 x-axis, while the third input specifies how much to displace pixels
6458 If one of displacement map streams terminates, last frame from that
6459 displacement map will be used.
6461 Note that once generated, displacements maps can be reused over and over again.
6463 A description of the accepted options follows.
6467 Set displace behavior for pixels that are out of range.
6469 Available values are:
6472 Missing pixels are replaced by black pixels.
6475 Adjacent pixels will spread out to replace missing pixels.
6478 Out of range pixels are wrapped so they point to pixels of other side.
6480 Default is @samp{smear}.
6484 @subsection Examples
6488 Add ripple effect to rgb input of video size hd720:
6490 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
6494 Add wave effect to rgb input of video size hd720:
6496 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
6502 Draw a colored box on the input image.
6504 It accepts the following parameters:
6509 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6513 The expressions which specify the width and height of the box; if 0 they are interpreted as
6514 the input width and height. It defaults to 0.
6517 Specify the color of the box to write. For the general syntax of this option,
6518 check the "Color" section in the ffmpeg-utils manual. If the special
6519 value @code{invert} is used, the box edge color is the same as the
6520 video with inverted luma.
6523 The expression which sets the thickness of the box edge. Default value is @code{3}.
6525 See below for the list of accepted constants.
6528 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6529 following constants:
6533 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6537 horizontal and vertical chroma subsample values. For example for the
6538 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6542 The input width and height.
6545 The input sample aspect ratio.
6549 The x and y offset coordinates where the box is drawn.
6553 The width and height of the drawn box.
6556 The thickness of the drawn box.
6558 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6559 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6563 @subsection Examples
6567 Draw a black box around the edge of the input image:
6573 Draw a box with color red and an opacity of 50%:
6575 drawbox=10:20:200:60:red@@0.5
6578 The previous example can be specified as:
6580 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6584 Fill the box with pink color:
6586 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6590 Draw a 2-pixel red 2.40:1 mask:
6592 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
6598 Draw a grid on the input image.
6600 It accepts the following parameters:
6605 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6609 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6610 input width and height, respectively, minus @code{thickness}, so image gets
6611 framed. Default to 0.
6614 Specify the color of the grid. For the general syntax of this option,
6615 check the "Color" section in the ffmpeg-utils manual. If the special
6616 value @code{invert} is used, the grid color is the same as the
6617 video with inverted luma.
6620 The expression which sets the thickness of the grid line. Default value is @code{1}.
6622 See below for the list of accepted constants.
6625 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6626 following constants:
6630 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6634 horizontal and vertical chroma subsample values. For example for the
6635 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6639 The input grid cell width and height.
6642 The input sample aspect ratio.
6646 The x and y coordinates of some point of grid intersection (meant to configure offset).
6650 The width and height of the drawn cell.
6653 The thickness of the drawn cell.
6655 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6656 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6660 @subsection Examples
6664 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6666 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6670 Draw a white 3x3 grid with an opacity of 50%:
6672 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6679 Draw a text string or text from a specified file on top of a video, using the
6680 libfreetype library.
6682 To enable compilation of this filter, you need to configure FFmpeg with
6683 @code{--enable-libfreetype}.
6684 To enable default font fallback and the @var{font} option you need to
6685 configure FFmpeg with @code{--enable-libfontconfig}.
6686 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6687 @code{--enable-libfribidi}.
6691 It accepts the following parameters:
6696 Used to draw a box around text using the background color.
6697 The value must be either 1 (enable) or 0 (disable).
6698 The default value of @var{box} is 0.
6701 Set the width of the border to be drawn around the box using @var{boxcolor}.
6702 The default value of @var{boxborderw} is 0.
6705 The color to be used for drawing box around text. For the syntax of this
6706 option, check the "Color" section in the ffmpeg-utils manual.
6708 The default value of @var{boxcolor} is "white".
6711 Set the width of the border to be drawn around the text using @var{bordercolor}.
6712 The default value of @var{borderw} is 0.
6715 Set the color to be used for drawing border around text. For the syntax of this
6716 option, check the "Color" section in the ffmpeg-utils manual.
6718 The default value of @var{bordercolor} is "black".
6721 Select how the @var{text} is expanded. Can be either @code{none},
6722 @code{strftime} (deprecated) or
6723 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6727 If true, check and fix text coords to avoid clipping.
6730 The color to be used for drawing fonts. For the syntax of this option, check
6731 the "Color" section in the ffmpeg-utils manual.
6733 The default value of @var{fontcolor} is "black".
6735 @item fontcolor_expr
6736 String which is expanded the same way as @var{text} to obtain dynamic
6737 @var{fontcolor} value. By default this option has empty value and is not
6738 processed. When this option is set, it overrides @var{fontcolor} option.
6741 The font family to be used for drawing text. By default Sans.
6744 The font file to be used for drawing text. The path must be included.
6745 This parameter is mandatory if the fontconfig support is disabled.
6748 This option does not exist, please see the timeline system
6751 Draw the text applying alpha blending. The value can
6752 be a number between 0.0 and 1.0.
6753 The expression accepts the same variables @var{x, y} as well.
6754 The default value is 1.
6755 Please see @var{fontcolor_expr}.
6758 The font size to be used for drawing text.
6759 The default value of @var{fontsize} is 16.
6762 If set to 1, attempt to shape the text (for example, reverse the order of
6763 right-to-left text and join Arabic characters) before drawing it.
6764 Otherwise, just draw the text exactly as given.
6765 By default 1 (if supported).
6768 The flags to be used for loading the fonts.
6770 The flags map the corresponding flags supported by libfreetype, and are
6771 a combination of the following values:
6778 @item vertical_layout
6779 @item force_autohint
6782 @item ignore_global_advance_width
6784 @item ignore_transform
6790 Default value is "default".
6792 For more information consult the documentation for the FT_LOAD_*
6796 The color to be used for drawing a shadow behind the drawn text. For the
6797 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6799 The default value of @var{shadowcolor} is "black".
6803 The x and y offsets for the text shadow position with respect to the
6804 position of the text. They can be either positive or negative
6805 values. The default value for both is "0".
6808 The starting frame number for the n/frame_num variable. The default value
6812 The size in number of spaces to use for rendering the tab.
6816 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6817 format. It can be used with or without text parameter. @var{timecode_rate}
6818 option must be specified.
6820 @item timecode_rate, rate, r
6821 Set the timecode frame rate (timecode only).
6824 The text string to be drawn. The text must be a sequence of UTF-8
6826 This parameter is mandatory if no file is specified with the parameter
6830 A text file containing text to be drawn. The text must be a sequence
6831 of UTF-8 encoded characters.
6833 This parameter is mandatory if no text string is specified with the
6834 parameter @var{text}.
6836 If both @var{text} and @var{textfile} are specified, an error is thrown.
6839 If set to 1, the @var{textfile} will be reloaded before each frame.
6840 Be sure to update it atomically, or it may be read partially, or even fail.
6844 The expressions which specify the offsets where text will be drawn
6845 within the video frame. They are relative to the top/left border of the
6848 The default value of @var{x} and @var{y} is "0".
6850 See below for the list of accepted constants and functions.
6853 The parameters for @var{x} and @var{y} are expressions containing the
6854 following constants and functions:
6858 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6862 horizontal and vertical chroma subsample values. For example for the
6863 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6866 the height of each text line
6874 @item max_glyph_a, ascent
6875 the maximum distance from the baseline to the highest/upper grid
6876 coordinate used to place a glyph outline point, for all the rendered
6878 It is a positive value, due to the grid's orientation with the Y axis
6881 @item max_glyph_d, descent
6882 the maximum distance from the baseline to the lowest grid coordinate
6883 used to place a glyph outline point, for all the rendered glyphs.
6884 This is a negative value, due to the grid's orientation, with the Y axis
6888 maximum glyph height, that is the maximum height for all the glyphs
6889 contained in the rendered text, it is equivalent to @var{ascent} -
6893 maximum glyph width, that is the maximum width for all the glyphs
6894 contained in the rendered text
6897 the number of input frame, starting from 0
6899 @item rand(min, max)
6900 return a random number included between @var{min} and @var{max}
6903 The input sample aspect ratio.
6906 timestamp expressed in seconds, NAN if the input timestamp is unknown
6909 the height of the rendered text
6912 the width of the rendered text
6916 the x and y offset coordinates where the text is drawn.
6918 These parameters allow the @var{x} and @var{y} expressions to refer
6919 each other, so you can for example specify @code{y=x/dar}.
6922 @anchor{drawtext_expansion}
6923 @subsection Text expansion
6925 If @option{expansion} is set to @code{strftime},
6926 the filter recognizes strftime() sequences in the provided text and
6927 expands them accordingly. Check the documentation of strftime(). This
6928 feature is deprecated.
6930 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6932 If @option{expansion} is set to @code{normal} (which is the default),
6933 the following expansion mechanism is used.
6935 The backslash character @samp{\}, followed by any character, always expands to
6936 the second character.
6938 Sequences of the form @code{%@{...@}} are expanded. The text between the
6939 braces is a function name, possibly followed by arguments separated by ':'.
6940 If the arguments contain special characters or delimiters (':' or '@}'),
6941 they should be escaped.
6943 Note that they probably must also be escaped as the value for the
6944 @option{text} option in the filter argument string and as the filter
6945 argument in the filtergraph description, and possibly also for the shell,
6946 that makes up to four levels of escaping; using a text file avoids these
6949 The following functions are available:
6954 The expression evaluation result.
6956 It must take one argument specifying the expression to be evaluated,
6957 which accepts the same constants and functions as the @var{x} and
6958 @var{y} values. Note that not all constants should be used, for
6959 example the text size is not known when evaluating the expression, so
6960 the constants @var{text_w} and @var{text_h} will have an undefined
6963 @item expr_int_format, eif
6964 Evaluate the expression's value and output as formatted integer.
6966 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6967 The second argument specifies the output format. Allowed values are @samp{x},
6968 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6969 @code{printf} function.
6970 The third parameter is optional and sets the number of positions taken by the output.
6971 It can be used to add padding with zeros from the left.
6974 The time at which the filter is running, expressed in UTC.
6975 It can accept an argument: a strftime() format string.
6978 The time at which the filter is running, expressed in the local time zone.
6979 It can accept an argument: a strftime() format string.
6982 Frame metadata. Takes one or two arguments.
6984 The first argument is mandatory and specifies the metadata key.
6986 The second argument is optional and specifies a default value, used when the
6987 metadata key is not found or empty.
6990 The frame number, starting from 0.
6993 A 1 character description of the current picture type.
6996 The timestamp of the current frame.
6997 It can take up to three arguments.
6999 The first argument is the format of the timestamp; it defaults to @code{flt}
7000 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7001 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7002 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7003 @code{localtime} stands for the timestamp of the frame formatted as
7004 local time zone time.
7006 The second argument is an offset added to the timestamp.
7008 If the format is set to @code{localtime} or @code{gmtime},
7009 a third argument may be supplied: a strftime() format string.
7010 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7013 @subsection Examples
7017 Draw "Test Text" with font FreeSerif, using the default values for the
7018 optional parameters.
7021 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7025 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7026 and y=50 (counting from the top-left corner of the screen), text is
7027 yellow with a red box around it. Both the text and the box have an
7031 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7032 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7035 Note that the double quotes are not necessary if spaces are not used
7036 within the parameter list.
7039 Show the text at the center of the video frame:
7041 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7045 Show the text at a random position, switching to a new position every 30 seconds:
7047 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)"
7051 Show a text line sliding from right to left in the last row of the video
7052 frame. The file @file{LONG_LINE} is assumed to contain a single line
7055 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7059 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7061 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7065 Draw a single green letter "g", at the center of the input video.
7066 The glyph baseline is placed at half screen height.
7068 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7072 Show text for 1 second every 3 seconds:
7074 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7078 Use fontconfig to set the font. Note that the colons need to be escaped.
7080 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7084 Print the date of a real-time encoding (see strftime(3)):
7086 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7090 Show text fading in and out (appearing/disappearing):
7093 DS=1.0 # display start
7094 DE=10.0 # display end
7095 FID=1.5 # fade in duration
7096 FOD=5 # fade out duration
7097 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 @}"
7102 For more information about libfreetype, check:
7103 @url{http://www.freetype.org/}.
7105 For more information about fontconfig, check:
7106 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7108 For more information about libfribidi, check:
7109 @url{http://fribidi.org/}.
7113 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7115 The filter accepts the following options:
7120 Set low and high threshold values used by the Canny thresholding
7123 The high threshold selects the "strong" edge pixels, which are then
7124 connected through 8-connectivity with the "weak" edge pixels selected
7125 by the low threshold.
7127 @var{low} and @var{high} threshold values must be chosen in the range
7128 [0,1], and @var{low} should be lesser or equal to @var{high}.
7130 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7134 Define the drawing mode.
7138 Draw white/gray wires on black background.
7141 Mix the colors to create a paint/cartoon effect.
7144 Default value is @var{wires}.
7147 @subsection Examples
7151 Standard edge detection with custom values for the hysteresis thresholding:
7153 edgedetect=low=0.1:high=0.4
7157 Painting effect without thresholding:
7159 edgedetect=mode=colormix:high=0
7164 Set brightness, contrast, saturation and approximate gamma adjustment.
7166 The filter accepts the following options:
7170 Set the contrast expression. The value must be a float value in range
7171 @code{-2.0} to @code{2.0}. The default value is "1".
7174 Set the brightness expression. The value must be a float value in
7175 range @code{-1.0} to @code{1.0}. The default value is "0".
7178 Set the saturation expression. The value must be a float in
7179 range @code{0.0} to @code{3.0}. The default value is "1".
7182 Set the gamma expression. The value must be a float in range
7183 @code{0.1} to @code{10.0}. The default value is "1".
7186 Set the gamma expression for red. The value must be a float in
7187 range @code{0.1} to @code{10.0}. The default value is "1".
7190 Set the gamma expression for green. The value must be a float in range
7191 @code{0.1} to @code{10.0}. The default value is "1".
7194 Set the gamma expression for blue. The value must be a float in range
7195 @code{0.1} to @code{10.0}. The default value is "1".
7198 Set the gamma weight expression. It can be used to reduce the effect
7199 of a high gamma value on bright image areas, e.g. keep them from
7200 getting overamplified and just plain white. The value must be a float
7201 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7202 gamma correction all the way down while @code{1.0} leaves it at its
7203 full strength. Default is "1".
7206 Set when the expressions for brightness, contrast, saturation and
7207 gamma expressions are evaluated.
7209 It accepts the following values:
7212 only evaluate expressions once during the filter initialization or
7213 when a command is processed
7216 evaluate expressions for each incoming frame
7219 Default value is @samp{init}.
7222 The expressions accept the following parameters:
7225 frame count of the input frame starting from 0
7228 byte position of the corresponding packet in the input file, NAN if
7232 frame rate of the input video, NAN if the input frame rate is unknown
7235 timestamp expressed in seconds, NAN if the input timestamp is unknown
7238 @subsection Commands
7239 The filter supports the following commands:
7243 Set the contrast expression.
7246 Set the brightness expression.
7249 Set the saturation expression.
7252 Set the gamma expression.
7255 Set the gamma_r expression.
7258 Set gamma_g expression.
7261 Set gamma_b expression.
7264 Set gamma_weight expression.
7266 The command accepts the same syntax of the corresponding option.
7268 If the specified expression is not valid, it is kept at its current
7275 Apply erosion effect to the video.
7277 This filter replaces the pixel by the local(3x3) minimum.
7279 It accepts the following options:
7286 Limit the maximum change for each plane, default is 65535.
7287 If 0, plane will remain unchanged.
7290 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7293 Flags to local 3x3 coordinates maps like this:
7300 @section extractplanes
7302 Extract color channel components from input video stream into
7303 separate grayscale video streams.
7305 The filter accepts the following option:
7309 Set plane(s) to extract.
7311 Available values for planes are:
7322 Choosing planes not available in the input will result in an error.
7323 That means you cannot select @code{r}, @code{g}, @code{b} planes
7324 with @code{y}, @code{u}, @code{v} planes at same time.
7327 @subsection Examples
7331 Extract luma, u and v color channel component from input video frame
7332 into 3 grayscale outputs:
7334 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
7340 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7342 For each input image, the filter will compute the optimal mapping from
7343 the input to the output given the codebook length, that is the number
7344 of distinct output colors.
7346 This filter accepts the following options.
7349 @item codebook_length, l
7350 Set codebook length. The value must be a positive integer, and
7351 represents the number of distinct output colors. Default value is 256.
7354 Set the maximum number of iterations to apply for computing the optimal
7355 mapping. The higher the value the better the result and the higher the
7356 computation time. Default value is 1.
7359 Set a random seed, must be an integer included between 0 and
7360 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7361 will try to use a good random seed on a best effort basis.
7364 Set pal8 output pixel format. This option does not work with codebook
7365 length greater than 256.
7370 Apply a fade-in/out effect to the input video.
7372 It accepts the following parameters:
7376 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7378 Default is @code{in}.
7380 @item start_frame, s
7381 Specify the number of the frame to start applying the fade
7382 effect at. Default is 0.
7385 The number of frames that the fade effect lasts. At the end of the
7386 fade-in effect, the output video will have the same intensity as the input video.
7387 At the end of the fade-out transition, the output video will be filled with the
7388 selected @option{color}.
7392 If set to 1, fade only alpha channel, if one exists on the input.
7395 @item start_time, st
7396 Specify the timestamp (in seconds) of the frame to start to apply the fade
7397 effect. If both start_frame and start_time are specified, the fade will start at
7398 whichever comes last. Default is 0.
7401 The number of seconds for which the fade effect has to last. At the end of the
7402 fade-in effect the output video will have the same intensity as the input video,
7403 at the end of the fade-out transition the output video will be filled with the
7404 selected @option{color}.
7405 If both duration and nb_frames are specified, duration is used. Default is 0
7406 (nb_frames is used by default).
7409 Specify the color of the fade. Default is "black".
7412 @subsection Examples
7416 Fade in the first 30 frames of video:
7421 The command above is equivalent to:
7427 Fade out the last 45 frames of a 200-frame video:
7430 fade=type=out:start_frame=155:nb_frames=45
7434 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7436 fade=in:0:25, fade=out:975:25
7440 Make the first 5 frames yellow, then fade in from frame 5-24:
7442 fade=in:5:20:color=yellow
7446 Fade in alpha over first 25 frames of video:
7448 fade=in:0:25:alpha=1
7452 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7454 fade=t=in:st=5.5:d=0.5
7460 Apply arbitrary expressions to samples in frequency domain
7464 Adjust the dc value (gain) of the luma plane of the image. The filter
7465 accepts an integer value in range @code{0} to @code{1000}. The default
7466 value is set to @code{0}.
7469 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7470 filter accepts an integer value in range @code{0} to @code{1000}. The
7471 default value is set to @code{0}.
7474 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7475 filter accepts an integer value in range @code{0} to @code{1000}. The
7476 default value is set to @code{0}.
7479 Set the frequency domain weight expression for the luma plane.
7482 Set the frequency domain weight expression for the 1st chroma plane.
7485 Set the frequency domain weight expression for the 2nd chroma plane.
7487 The filter accepts the following variables:
7490 The coordinates of the current sample.
7494 The width and height of the image.
7497 @subsection Examples
7503 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7509 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7515 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7521 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7528 Extract a single field from an interlaced image using stride
7529 arithmetic to avoid wasting CPU time. The output frames are marked as
7532 The filter accepts the following options:
7536 Specify whether to extract the top (if the value is @code{0} or
7537 @code{top}) or the bottom field (if the value is @code{1} or
7543 Create new frames by copying the top and bottom fields from surrounding frames
7544 supplied as numbers by the hint file.
7548 Set file containing hints: absolute/relative frame numbers.
7550 There must be one line for each frame in a clip. Each line must contain two
7551 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7552 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7553 is current frame number for @code{absolute} mode or out of [-1, 1] range
7554 for @code{relative} mode. First number tells from which frame to pick up top
7555 field and second number tells from which frame to pick up bottom field.
7557 If optionally followed by @code{+} output frame will be marked as interlaced,
7558 else if followed by @code{-} output frame will be marked as progressive, else
7559 it will be marked same as input frame.
7560 If line starts with @code{#} or @code{;} that line is skipped.
7563 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7566 Example of first several lines of @code{hint} file for @code{relative} mode:
7569 1,0 - # second frame, use third's frame top field and second's frame bottom field
7570 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7587 Field matching filter for inverse telecine. It is meant to reconstruct the
7588 progressive frames from a telecined stream. The filter does not drop duplicated
7589 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7590 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7592 The separation of the field matching and the decimation is notably motivated by
7593 the possibility of inserting a de-interlacing filter fallback between the two.
7594 If the source has mixed telecined and real interlaced content,
7595 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7596 But these remaining combed frames will be marked as interlaced, and thus can be
7597 de-interlaced by a later filter such as @ref{yadif} before decimation.
7599 In addition to the various configuration options, @code{fieldmatch} can take an
7600 optional second stream, activated through the @option{ppsrc} option. If
7601 enabled, the frames reconstruction will be based on the fields and frames from
7602 this second stream. This allows the first input to be pre-processed in order to
7603 help the various algorithms of the filter, while keeping the output lossless
7604 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7605 or brightness/contrast adjustments can help.
7607 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7608 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7609 which @code{fieldmatch} is based on. While the semantic and usage are very
7610 close, some behaviour and options names can differ.
7612 The @ref{decimate} filter currently only works for constant frame rate input.
7613 If your input has mixed telecined (30fps) and progressive content with a lower
7614 framerate like 24fps use the following filterchain to produce the necessary cfr
7615 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7617 The filter accepts the following options:
7621 Specify the assumed field order of the input stream. Available values are:
7625 Auto detect parity (use FFmpeg's internal parity value).
7627 Assume bottom field first.
7629 Assume top field first.
7632 Note that it is sometimes recommended not to trust the parity announced by the
7635 Default value is @var{auto}.
7638 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7639 sense that it won't risk creating jerkiness due to duplicate frames when
7640 possible, but if there are bad edits or blended fields it will end up
7641 outputting combed frames when a good match might actually exist. On the other
7642 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7643 but will almost always find a good frame if there is one. The other values are
7644 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7645 jerkiness and creating duplicate frames versus finding good matches in sections
7646 with bad edits, orphaned fields, blended fields, etc.
7648 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7650 Available values are:
7654 2-way matching (p/c)
7656 2-way matching, and trying 3rd match if still combed (p/c + n)
7658 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7660 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7661 still combed (p/c + n + u/b)
7663 3-way matching (p/c/n)
7665 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7666 detected as combed (p/c/n + u/b)
7669 The parenthesis at the end indicate the matches that would be used for that
7670 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7673 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7676 Default value is @var{pc_n}.
7679 Mark the main input stream as a pre-processed input, and enable the secondary
7680 input stream as the clean source to pick the fields from. See the filter
7681 introduction for more details. It is similar to the @option{clip2} feature from
7684 Default value is @code{0} (disabled).
7687 Set the field to match from. It is recommended to set this to the same value as
7688 @option{order} unless you experience matching failures with that setting. In
7689 certain circumstances changing the field that is used to match from can have a
7690 large impact on matching performance. Available values are:
7694 Automatic (same value as @option{order}).
7696 Match from the bottom field.
7698 Match from the top field.
7701 Default value is @var{auto}.
7704 Set whether or not chroma is included during the match comparisons. In most
7705 cases it is recommended to leave this enabled. You should set this to @code{0}
7706 only if your clip has bad chroma problems such as heavy rainbowing or other
7707 artifacts. Setting this to @code{0} could also be used to speed things up at
7708 the cost of some accuracy.
7710 Default value is @code{1}.
7714 These define an exclusion band which excludes the lines between @option{y0} and
7715 @option{y1} from being included in the field matching decision. An exclusion
7716 band can be used to ignore subtitles, a logo, or other things that may
7717 interfere with the matching. @option{y0} sets the starting scan line and
7718 @option{y1} sets the ending line; all lines in between @option{y0} and
7719 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7720 @option{y0} and @option{y1} to the same value will disable the feature.
7721 @option{y0} and @option{y1} defaults to @code{0}.
7724 Set the scene change detection threshold as a percentage of maximum change on
7725 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7726 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7727 @option{scthresh} is @code{[0.0, 100.0]}.
7729 Default value is @code{12.0}.
7732 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7733 account the combed scores of matches when deciding what match to use as the
7734 final match. Available values are:
7738 No final matching based on combed scores.
7740 Combed scores are only used when a scene change is detected.
7742 Use combed scores all the time.
7745 Default is @var{sc}.
7748 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7749 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7750 Available values are:
7754 No forced calculation.
7756 Force p/c/n calculations.
7758 Force p/c/n/u/b calculations.
7761 Default value is @var{none}.
7764 This is the area combing threshold used for combed frame detection. This
7765 essentially controls how "strong" or "visible" combing must be to be detected.
7766 Larger values mean combing must be more visible and smaller values mean combing
7767 can be less visible or strong and still be detected. Valid settings are from
7768 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7769 be detected as combed). This is basically a pixel difference value. A good
7770 range is @code{[8, 12]}.
7772 Default value is @code{9}.
7775 Sets whether or not chroma is considered in the combed frame decision. Only
7776 disable this if your source has chroma problems (rainbowing, etc.) that are
7777 causing problems for the combed frame detection with chroma enabled. Actually,
7778 using @option{chroma}=@var{0} is usually more reliable, except for the case
7779 where there is chroma only combing in the source.
7781 Default value is @code{0}.
7785 Respectively set the x-axis and y-axis size of the window used during combed
7786 frame detection. This has to do with the size of the area in which
7787 @option{combpel} pixels are required to be detected as combed for a frame to be
7788 declared combed. See the @option{combpel} parameter description for more info.
7789 Possible values are any number that is a power of 2 starting at 4 and going up
7792 Default value is @code{16}.
7795 The number of combed pixels inside any of the @option{blocky} by
7796 @option{blockx} size blocks on the frame for the frame to be detected as
7797 combed. While @option{cthresh} controls how "visible" the combing must be, this
7798 setting controls "how much" combing there must be in any localized area (a
7799 window defined by the @option{blockx} and @option{blocky} settings) on the
7800 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7801 which point no frames will ever be detected as combed). This setting is known
7802 as @option{MI} in TFM/VFM vocabulary.
7804 Default value is @code{80}.
7807 @anchor{p/c/n/u/b meaning}
7808 @subsection p/c/n/u/b meaning
7810 @subsubsection p/c/n
7812 We assume the following telecined stream:
7815 Top fields: 1 2 2 3 4
7816 Bottom fields: 1 2 3 4 4
7819 The numbers correspond to the progressive frame the fields relate to. Here, the
7820 first two frames are progressive, the 3rd and 4th are combed, and so on.
7822 When @code{fieldmatch} is configured to run a matching from bottom
7823 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7828 B 1 2 3 4 4 <-- matching reference
7837 As a result of the field matching, we can see that some frames get duplicated.
7838 To perform a complete inverse telecine, you need to rely on a decimation filter
7839 after this operation. See for instance the @ref{decimate} filter.
7841 The same operation now matching from top fields (@option{field}=@var{top})
7846 T 1 2 2 3 4 <-- matching reference
7856 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7857 basically, they refer to the frame and field of the opposite parity:
7860 @item @var{p} matches the field of the opposite parity in the previous frame
7861 @item @var{c} matches the field of the opposite parity in the current frame
7862 @item @var{n} matches the field of the opposite parity in the next frame
7867 The @var{u} and @var{b} matching are a bit special in the sense that they match
7868 from the opposite parity flag. In the following examples, we assume that we are
7869 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7870 'x' is placed above and below each matched fields.
7872 With bottom matching (@option{field}=@var{bottom}):
7877 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7878 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7886 With top matching (@option{field}=@var{top}):
7891 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7892 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7900 @subsection Examples
7902 Simple IVTC of a top field first telecined stream:
7904 fieldmatch=order=tff:combmatch=none, decimate
7907 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7909 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7914 Transform the field order of the input video.
7916 It accepts the following parameters:
7921 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7922 for bottom field first.
7925 The default value is @samp{tff}.
7927 The transformation is done by shifting the picture content up or down
7928 by one line, and filling the remaining line with appropriate picture content.
7929 This method is consistent with most broadcast field order converters.
7931 If the input video is not flagged as being interlaced, or it is already
7932 flagged as being of the required output field order, then this filter does
7933 not alter the incoming video.
7935 It is very useful when converting to or from PAL DV material,
7936 which is bottom field first.
7940 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7943 @section fifo, afifo
7945 Buffer input images and send them when they are requested.
7947 It is mainly useful when auto-inserted by the libavfilter
7950 It does not take parameters.
7954 Find a rectangular object
7956 It accepts the following options:
7960 Filepath of the object image, needs to be in gray8.
7963 Detection threshold, default is 0.5.
7966 Number of mipmaps, default is 3.
7968 @item xmin, ymin, xmax, ymax
7969 Specifies the rectangle in which to search.
7972 @subsection Examples
7976 Generate a representative palette of a given video using @command{ffmpeg}:
7978 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7984 Cover a rectangular object
7986 It accepts the following options:
7990 Filepath of the optional cover image, needs to be in yuv420.
7995 It accepts the following values:
7998 cover it by the supplied image
8000 cover it by interpolating the surrounding pixels
8003 Default value is @var{blur}.
8006 @subsection Examples
8010 Generate a representative palette of a given video using @command{ffmpeg}:
8012 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8019 Convert the input video to one of the specified pixel formats.
8020 Libavfilter will try to pick one that is suitable as input to
8023 It accepts the following parameters:
8027 A '|'-separated list of pixel format names, such as
8028 "pix_fmts=yuv420p|monow|rgb24".
8032 @subsection Examples
8036 Convert the input video to the @var{yuv420p} format
8038 format=pix_fmts=yuv420p
8041 Convert the input video to any of the formats in the list
8043 format=pix_fmts=yuv420p|yuv444p|yuv410p
8050 Convert the video to specified constant frame rate by duplicating or dropping
8051 frames as necessary.
8053 It accepts the following parameters:
8057 The desired output frame rate. The default is @code{25}.
8062 Possible values are:
8065 zero round towards 0
8069 round towards -infinity
8071 round towards +infinity
8075 The default is @code{near}.
8078 Assume the first PTS should be the given value, in seconds. This allows for
8079 padding/trimming at the start of stream. By default, no assumption is made
8080 about the first frame's expected PTS, so no padding or trimming is done.
8081 For example, this could be set to 0 to pad the beginning with duplicates of
8082 the first frame if a video stream starts after the audio stream or to trim any
8083 frames with a negative PTS.
8087 Alternatively, the options can be specified as a flat string:
8088 @var{fps}[:@var{round}].
8090 See also the @ref{setpts} filter.
8092 @subsection Examples
8096 A typical usage in order to set the fps to 25:
8102 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8104 fps=fps=film:round=near
8110 Pack two different video streams into a stereoscopic video, setting proper
8111 metadata on supported codecs. The two views should have the same size and
8112 framerate and processing will stop when the shorter video ends. Please note
8113 that you may conveniently adjust view properties with the @ref{scale} and
8116 It accepts the following parameters:
8120 The desired packing format. Supported values are:
8125 The views are next to each other (default).
8128 The views are on top of each other.
8131 The views are packed by line.
8134 The views are packed by column.
8137 The views are temporally interleaved.
8146 # Convert left and right views into a frame-sequential video
8147 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8149 # Convert views into a side-by-side video with the same output resolution as the input
8150 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
8155 Change the frame rate by interpolating new video output frames from the source
8158 This filter is not designed to function correctly with interlaced media. If
8159 you wish to change the frame rate of interlaced media then you are required
8160 to deinterlace before this filter and re-interlace after this filter.
8162 A description of the accepted options follows.
8166 Specify the output frames per second. This option can also be specified
8167 as a value alone. The default is @code{50}.
8170 Specify the start of a range where the output frame will be created as a
8171 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8172 the default is @code{15}.
8175 Specify the end of a range where the output frame will be created as a
8176 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8177 the default is @code{240}.
8180 Specify the level at which a scene change is detected as a value between
8181 0 and 100 to indicate a new scene; a low value reflects a low
8182 probability for the current frame to introduce a new scene, while a higher
8183 value means the current frame is more likely to be one.
8184 The default is @code{7}.
8187 Specify flags influencing the filter process.
8189 Available value for @var{flags} is:
8192 @item scene_change_detect, scd
8193 Enable scene change detection using the value of the option @var{scene}.
8194 This flag is enabled by default.
8200 Select one frame every N-th frame.
8202 This filter accepts the following option:
8205 Select frame after every @code{step} frames.
8206 Allowed values are positive integers higher than 0. Default value is @code{1}.
8212 Apply a frei0r effect to the input video.
8214 To enable the compilation of this filter, you need to install the frei0r
8215 header and configure FFmpeg with @code{--enable-frei0r}.
8217 It accepts the following parameters:
8222 The name of the frei0r effect to load. If the environment variable
8223 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8224 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8225 Otherwise, the standard frei0r paths are searched, in this order:
8226 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8227 @file{/usr/lib/frei0r-1/}.
8230 A '|'-separated list of parameters to pass to the frei0r effect.
8234 A frei0r effect parameter can be a boolean (its value is either
8235 "y" or "n"), a double, a color (specified as
8236 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8237 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8238 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8239 @var{X} and @var{Y} are floating point numbers) and/or a string.
8241 The number and types of parameters depend on the loaded effect. If an
8242 effect parameter is not specified, the default value is set.
8244 @subsection Examples
8248 Apply the distort0r effect, setting the first two double parameters:
8250 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8254 Apply the colordistance effect, taking a color as the first parameter:
8256 frei0r=colordistance:0.2/0.3/0.4
8257 frei0r=colordistance:violet
8258 frei0r=colordistance:0x112233
8262 Apply the perspective effect, specifying the top left and top right image
8265 frei0r=perspective:0.2/0.2|0.8/0.2
8269 For more information, see
8270 @url{http://frei0r.dyne.org}
8274 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8276 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8277 processing filter, one of them is performed once per block, not per pixel.
8278 This allows for much higher speed.
8280 The filter accepts the following options:
8284 Set quality. This option defines the number of levels for averaging. It accepts
8285 an integer in the range 4-5. Default value is @code{4}.
8288 Force a constant quantization parameter. It accepts an integer in range 0-63.
8289 If not set, the filter will use the QP from the video stream (if available).
8292 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8293 more details but also more artifacts, while higher values make the image smoother
8294 but also blurrier. Default value is @code{0} − PSNR optimal.
8297 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8298 option may cause flicker since the B-Frames have often larger QP. Default is
8299 @code{0} (not enabled).
8305 Apply Gaussian blur filter.
8307 The filter accepts the following options:
8311 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8314 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8317 Set which planes to filter. By default all planes are filtered.
8320 Set vertical sigma, if negative it will be same as @code{sigma}.
8321 Default is @code{-1}.
8326 The filter accepts the following options:
8330 Set the luminance expression.
8332 Set the chrominance blue expression.
8334 Set the chrominance red expression.
8336 Set the alpha expression.
8338 Set the red expression.
8340 Set the green expression.
8342 Set the blue expression.
8345 The colorspace is selected according to the specified options. If one
8346 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8347 options is specified, the filter will automatically select a YCbCr
8348 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8349 @option{blue_expr} options is specified, it will select an RGB
8352 If one of the chrominance expression is not defined, it falls back on the other
8353 one. If no alpha expression is specified it will evaluate to opaque value.
8354 If none of chrominance expressions are specified, they will evaluate
8355 to the luminance expression.
8357 The expressions can use the following variables and functions:
8361 The sequential number of the filtered frame, starting from @code{0}.
8365 The coordinates of the current sample.
8369 The width and height of the image.
8373 Width and height scale depending on the currently filtered plane. It is the
8374 ratio between the corresponding luma plane number of pixels and the current
8375 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8376 @code{0.5,0.5} for chroma planes.
8379 Time of the current frame, expressed in seconds.
8382 Return the value of the pixel at location (@var{x},@var{y}) of the current
8386 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8390 Return the value of the pixel at location (@var{x},@var{y}) of the
8391 blue-difference chroma plane. Return 0 if there is no such plane.
8394 Return the value of the pixel at location (@var{x},@var{y}) of the
8395 red-difference chroma plane. Return 0 if there is no such plane.
8400 Return the value of the pixel at location (@var{x},@var{y}) of the
8401 red/green/blue component. Return 0 if there is no such component.
8404 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8405 plane. Return 0 if there is no such plane.
8408 For functions, if @var{x} and @var{y} are outside the area, the value will be
8409 automatically clipped to the closer edge.
8411 @subsection Examples
8415 Flip the image horizontally:
8421 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8422 wavelength of 100 pixels:
8424 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8428 Generate a fancy enigmatic moving light:
8430 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
8434 Generate a quick emboss effect:
8436 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8440 Modify RGB components depending on pixel position:
8442 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8446 Create a radial gradient that is the same size as the input (also see
8447 the @ref{vignette} filter):
8449 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8455 Fix the banding artifacts that are sometimes introduced into nearly flat
8456 regions by truncation to 8-bit color depth.
8457 Interpolate the gradients that should go where the bands are, and
8460 It is designed for playback only. Do not use it prior to
8461 lossy compression, because compression tends to lose the dither and
8462 bring back the bands.
8464 It accepts the following parameters:
8469 The maximum amount by which the filter will change any one pixel. This is also
8470 the threshold for detecting nearly flat regions. Acceptable values range from
8471 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8475 The neighborhood to fit the gradient to. A larger radius makes for smoother
8476 gradients, but also prevents the filter from modifying the pixels near detailed
8477 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8478 values will be clipped to the valid range.
8482 Alternatively, the options can be specified as a flat string:
8483 @var{strength}[:@var{radius}]
8485 @subsection Examples
8489 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8495 Specify radius, omitting the strength (which will fall-back to the default
8506 Apply a Hald CLUT to a video stream.
8508 First input is the video stream to process, and second one is the Hald CLUT.
8509 The Hald CLUT input can be a simple picture or a complete video stream.
8511 The filter accepts the following options:
8515 Force termination when the shortest input terminates. Default is @code{0}.
8517 Continue applying the last CLUT after the end of the stream. A value of
8518 @code{0} disable the filter after the last frame of the CLUT is reached.
8519 Default is @code{1}.
8522 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8523 filters share the same internals).
8525 More information about the Hald CLUT can be found on Eskil Steenberg's website
8526 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8528 @subsection Workflow examples
8530 @subsubsection Hald CLUT video stream
8532 Generate an identity Hald CLUT stream altered with various effects:
8534 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
8537 Note: make sure you use a lossless codec.
8539 Then use it with @code{haldclut} to apply it on some random stream:
8541 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8544 The Hald CLUT will be applied to the 10 first seconds (duration of
8545 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8546 to the remaining frames of the @code{mandelbrot} stream.
8548 @subsubsection Hald CLUT with preview
8550 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8551 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8552 biggest possible square starting at the top left of the picture. The remaining
8553 padding pixels (bottom or right) will be ignored. This area can be used to add
8554 a preview of the Hald CLUT.
8556 Typically, the following generated Hald CLUT will be supported by the
8557 @code{haldclut} filter:
8560 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8561 pad=iw+320 [padded_clut];
8562 smptebars=s=320x256, split [a][b];
8563 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8564 [main][b] overlay=W-320" -frames:v 1 clut.png
8567 It contains the original and a preview of the effect of the CLUT: SMPTE color
8568 bars are displayed on the right-top, and below the same color bars processed by
8571 Then, the effect of this Hald CLUT can be visualized with:
8573 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8578 Flip the input video horizontally.
8580 For example, to horizontally flip the input video with @command{ffmpeg}:
8582 ffmpeg -i in.avi -vf "hflip" out.avi
8586 This filter applies a global color histogram equalization on a
8589 It can be used to correct video that has a compressed range of pixel
8590 intensities. The filter redistributes the pixel intensities to
8591 equalize their distribution across the intensity range. It may be
8592 viewed as an "automatically adjusting contrast filter". This filter is
8593 useful only for correcting degraded or poorly captured source
8596 The filter accepts the following options:
8600 Determine the amount of equalization to be applied. As the strength
8601 is reduced, the distribution of pixel intensities more-and-more
8602 approaches that of the input frame. The value must be a float number
8603 in the range [0,1] and defaults to 0.200.
8606 Set the maximum intensity that can generated and scale the output
8607 values appropriately. The strength should be set as desired and then
8608 the intensity can be limited if needed to avoid washing-out. The value
8609 must be a float number in the range [0,1] and defaults to 0.210.
8612 Set the antibanding level. If enabled the filter will randomly vary
8613 the luminance of output pixels by a small amount to avoid banding of
8614 the histogram. Possible values are @code{none}, @code{weak} or
8615 @code{strong}. It defaults to @code{none}.
8620 Compute and draw a color distribution histogram for the input video.
8622 The computed histogram is a representation of the color component
8623 distribution in an image.
8625 Standard histogram displays the color components distribution in an image.
8626 Displays color graph for each color component. Shows distribution of
8627 the Y, U, V, A or R, G, B components, depending on input format, in the
8628 current frame. Below each graph a color component scale meter is shown.
8630 The filter accepts the following options:
8634 Set height of level. Default value is @code{200}.
8635 Allowed range is [50, 2048].
8638 Set height of color scale. Default value is @code{12}.
8639 Allowed range is [0, 40].
8643 It accepts the following values:
8646 Per color component graphs are placed below each other.
8649 Presents information identical to that in the @code{parade}, except
8650 that the graphs representing color components are superimposed directly
8653 Default is @code{parade}.
8656 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8657 Default is @code{linear}.
8660 Set what color components to display.
8661 Default is @code{7}.
8664 Set foreground opacity. Default is @code{0.7}.
8667 Set background opacity. Default is @code{0.5}.
8670 @subsection Examples
8675 Calculate and draw histogram:
8677 ffplay -i input -vf histogram
8685 This is a high precision/quality 3d denoise filter. It aims to reduce
8686 image noise, producing smooth images and making still images really
8687 still. It should enhance compressibility.
8689 It accepts the following optional parameters:
8693 A non-negative floating point number which specifies spatial luma strength.
8696 @item chroma_spatial
8697 A non-negative floating point number which specifies spatial chroma strength.
8698 It defaults to 3.0*@var{luma_spatial}/4.0.
8701 A floating point number which specifies luma temporal strength. It defaults to
8702 6.0*@var{luma_spatial}/4.0.
8705 A floating point number which specifies chroma temporal strength. It defaults to
8706 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8709 @anchor{hwupload_cuda}
8710 @section hwupload_cuda
8712 Upload system memory frames to a CUDA device.
8714 It accepts the following optional parameters:
8718 The number of the CUDA device to use
8723 Apply a high-quality magnification filter designed for pixel art. This filter
8724 was originally created by Maxim Stepin.
8726 It accepts the following option:
8730 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8731 @code{hq3x} and @code{4} for @code{hq4x}.
8732 Default is @code{3}.
8736 Stack input videos horizontally.
8738 All streams must be of same pixel format and of same height.
8740 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8741 to create same output.
8743 The filter accept the following option:
8747 Set number of input streams. Default is 2.
8750 If set to 1, force the output to terminate when the shortest input
8751 terminates. Default value is 0.
8756 Modify the hue and/or the saturation of the input.
8758 It accepts the following parameters:
8762 Specify the hue angle as a number of degrees. It accepts an expression,
8763 and defaults to "0".
8766 Specify the saturation in the [-10,10] range. It accepts an expression and
8770 Specify the hue angle as a number of radians. It accepts an
8771 expression, and defaults to "0".
8774 Specify the brightness in the [-10,10] range. It accepts an expression and
8778 @option{h} and @option{H} are mutually exclusive, and can't be
8779 specified at the same time.
8781 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8782 expressions containing the following constants:
8786 frame count of the input frame starting from 0
8789 presentation timestamp of the input frame expressed in time base units
8792 frame rate of the input video, NAN if the input frame rate is unknown
8795 timestamp expressed in seconds, NAN if the input timestamp is unknown
8798 time base of the input video
8801 @subsection Examples
8805 Set the hue to 90 degrees and the saturation to 1.0:
8811 Same command but expressing the hue in radians:
8817 Rotate hue and make the saturation swing between 0
8818 and 2 over a period of 1 second:
8820 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8824 Apply a 3 seconds saturation fade-in effect starting at 0:
8829 The general fade-in expression can be written as:
8831 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8835 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8837 hue="s=max(0\, min(1\, (8-t)/3))"
8840 The general fade-out expression can be written as:
8842 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8847 @subsection Commands
8849 This filter supports the following commands:
8855 Modify the hue and/or the saturation and/or brightness of the input video.
8856 The command accepts the same syntax of the corresponding option.
8858 If the specified expression is not valid, it is kept at its current
8864 Grow first stream into second stream by connecting components.
8865 This makes it possible to build more robust edge masks.
8867 This filter accepts the following options:
8871 Set which planes will be processed as bitmap, unprocessed planes will be
8872 copied from first stream.
8873 By default value 0xf, all planes will be processed.
8876 Set threshold which is used in filtering. If pixel component value is higher than
8877 this value filter algorithm for connecting components is activated.
8878 By default value is 0.
8883 Detect video interlacing type.
8885 This filter tries to detect if the input frames are interlaced, progressive,
8886 top or bottom field first. It will also try to detect fields that are
8887 repeated between adjacent frames (a sign of telecine).
8889 Single frame detection considers only immediately adjacent frames when classifying each frame.
8890 Multiple frame detection incorporates the classification history of previous frames.
8892 The filter will log these metadata values:
8895 @item single.current_frame
8896 Detected type of current frame using single-frame detection. One of:
8897 ``tff'' (top field first), ``bff'' (bottom field first),
8898 ``progressive'', or ``undetermined''
8901 Cumulative number of frames detected as top field first using single-frame detection.
8904 Cumulative number of frames detected as top field first using multiple-frame detection.
8907 Cumulative number of frames detected as bottom field first using single-frame detection.
8909 @item multiple.current_frame
8910 Detected type of current frame using multiple-frame detection. One of:
8911 ``tff'' (top field first), ``bff'' (bottom field first),
8912 ``progressive'', or ``undetermined''
8915 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8917 @item single.progressive
8918 Cumulative number of frames detected as progressive using single-frame detection.
8920 @item multiple.progressive
8921 Cumulative number of frames detected as progressive using multiple-frame detection.
8923 @item single.undetermined
8924 Cumulative number of frames that could not be classified using single-frame detection.
8926 @item multiple.undetermined
8927 Cumulative number of frames that could not be classified using multiple-frame detection.
8929 @item repeated.current_frame
8930 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8932 @item repeated.neither
8933 Cumulative number of frames with no repeated field.
8936 Cumulative number of frames with the top field repeated from the previous frame's top field.
8938 @item repeated.bottom
8939 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8942 The filter accepts the following options:
8946 Set interlacing threshold.
8948 Set progressive threshold.
8950 Threshold for repeated field detection.
8952 Number of frames after which a given frame's contribution to the
8953 statistics is halved (i.e., it contributes only 0.5 to its
8954 classification). The default of 0 means that all frames seen are given
8955 full weight of 1.0 forever.
8956 @item analyze_interlaced_flag
8957 When this is not 0 then idet will use the specified number of frames to determine
8958 if the interlaced flag is accurate, it will not count undetermined frames.
8959 If the flag is found to be accurate it will be used without any further
8960 computations, if it is found to be inaccurate it will be cleared without any
8961 further computations. This allows inserting the idet filter as a low computational
8962 method to clean up the interlaced flag
8967 Deinterleave or interleave fields.
8969 This filter allows one to process interlaced images fields without
8970 deinterlacing them. Deinterleaving splits the input frame into 2
8971 fields (so called half pictures). Odd lines are moved to the top
8972 half of the output image, even lines to the bottom half.
8973 You can process (filter) them independently and then re-interleave them.
8975 The filter accepts the following options:
8979 @item chroma_mode, c
8981 Available values for @var{luma_mode}, @var{chroma_mode} and
8982 @var{alpha_mode} are:
8988 @item deinterleave, d
8989 Deinterleave fields, placing one above the other.
8992 Interleave fields. Reverse the effect of deinterleaving.
8994 Default value is @code{none}.
8997 @item chroma_swap, cs
8998 @item alpha_swap, as
8999 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9004 Apply inflate effect to the video.
9006 This filter replaces the pixel by the local(3x3) average by taking into account
9007 only values higher than the pixel.
9009 It accepts the following options:
9016 Limit the maximum change for each plane, default is 65535.
9017 If 0, plane will remain unchanged.
9022 Simple interlacing filter from progressive contents. This interleaves upper (or
9023 lower) lines from odd frames with lower (or upper) lines from even frames,
9024 halving the frame rate and preserving image height.
9027 Original Original New Frame
9028 Frame 'j' Frame 'j+1' (tff)
9029 ========== =========== ==================
9030 Line 0 --------------------> Frame 'j' Line 0
9031 Line 1 Line 1 ----> Frame 'j+1' Line 1
9032 Line 2 ---------------------> Frame 'j' Line 2
9033 Line 3 Line 3 ----> Frame 'j+1' Line 3
9035 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9038 It accepts the following optional parameters:
9042 This determines whether the interlaced frame is taken from the even
9043 (tff - default) or odd (bff) lines of the progressive frame.
9046 Enable (default) or disable the vertical lowpass filter to avoid twitter
9047 interlacing and reduce moire patterns.
9052 Deinterlace input video by applying Donald Graft's adaptive kernel
9053 deinterling. Work on interlaced parts of a video to produce
9056 The description of the accepted parameters follows.
9060 Set the threshold which affects the filter's tolerance when
9061 determining if a pixel line must be processed. It must be an integer
9062 in the range [0,255] and defaults to 10. A value of 0 will result in
9063 applying the process on every pixels.
9066 Paint pixels exceeding the threshold value to white if set to 1.
9070 Set the fields order. Swap fields if set to 1, leave fields alone if
9074 Enable additional sharpening if set to 1. Default is 0.
9077 Enable twoway sharpening if set to 1. Default is 0.
9080 @subsection Examples
9084 Apply default values:
9086 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9090 Enable additional sharpening:
9096 Paint processed pixels in white:
9102 @section lenscorrection
9104 Correct radial lens distortion
9106 This filter can be used to correct for radial distortion as can result from the use
9107 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9108 one can use tools available for example as part of opencv or simply trial-and-error.
9109 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9110 and extract the k1 and k2 coefficients from the resulting matrix.
9112 Note that effectively the same filter is available in the open-source tools Krita and
9113 Digikam from the KDE project.
9115 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9116 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9117 brightness distribution, so you may want to use both filters together in certain
9118 cases, though you will have to take care of ordering, i.e. whether vignetting should
9119 be applied before or after lens correction.
9123 The filter accepts the following options:
9127 Relative x-coordinate of the focal point of the image, and thereby the center of the
9128 distortion. This value has a range [0,1] and is expressed as fractions of the image
9131 Relative y-coordinate of the focal point of the image, and thereby the center of the
9132 distortion. This value has a range [0,1] and is expressed as fractions of the image
9135 Coefficient of the quadratic correction term. 0.5 means no correction.
9137 Coefficient of the double quadratic correction term. 0.5 means no correction.
9140 The formula that generates the correction is:
9142 @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)
9144 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9145 distances from the focal point in the source and target images, respectively.
9151 The filter accepts the following options:
9155 Set the number of loops.
9158 Set maximal size in number of frames.
9161 Set first frame of loop.
9167 Apply a 3D LUT to an input video.
9169 The filter accepts the following options:
9173 Set the 3D LUT file name.
9175 Currently supported formats:
9187 Select interpolation mode.
9189 Available values are:
9193 Use values from the nearest defined point.
9195 Interpolate values using the 8 points defining a cube.
9197 Interpolate values using a tetrahedron.
9201 @section lut, lutrgb, lutyuv
9203 Compute a look-up table for binding each pixel component input value
9204 to an output value, and apply it to the input video.
9206 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9207 to an RGB input video.
9209 These filters accept the following parameters:
9212 set first pixel component expression
9214 set second pixel component expression
9216 set third pixel component expression
9218 set fourth pixel component expression, corresponds to the alpha component
9221 set red component expression
9223 set green component expression
9225 set blue component expression
9227 alpha component expression
9230 set Y/luminance component expression
9232 set U/Cb component expression
9234 set V/Cr component expression
9237 Each of them specifies the expression to use for computing the lookup table for
9238 the corresponding pixel component values.
9240 The exact component associated to each of the @var{c*} options depends on the
9243 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9244 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9246 The expressions can contain the following constants and functions:
9251 The input width and height.
9254 The input value for the pixel component.
9257 The input value, clipped to the @var{minval}-@var{maxval} range.
9260 The maximum value for the pixel component.
9263 The minimum value for the pixel component.
9266 The negated value for the pixel component value, clipped to the
9267 @var{minval}-@var{maxval} range; it corresponds to the expression
9268 "maxval-clipval+minval".
9271 The computed value in @var{val}, clipped to the
9272 @var{minval}-@var{maxval} range.
9274 @item gammaval(gamma)
9275 The computed gamma correction value of the pixel component value,
9276 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9278 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9282 All expressions default to "val".
9284 @subsection Examples
9290 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9291 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9294 The above is the same as:
9296 lutrgb="r=negval:g=negval:b=negval"
9297 lutyuv="y=negval:u=negval:v=negval"
9307 Remove chroma components, turning the video into a graytone image:
9309 lutyuv="u=128:v=128"
9313 Apply a luma burning effect:
9319 Remove green and blue components:
9325 Set a constant alpha channel value on input:
9327 format=rgba,lutrgb=a="maxval-minval/2"
9331 Correct luminance gamma by a factor of 0.5:
9333 lutyuv=y=gammaval(0.5)
9337 Discard least significant bits of luma:
9339 lutyuv=y='bitand(val, 128+64+32)'
9343 Technicolor like effect:
9345 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9351 Compute and apply a lookup table from two video inputs.
9353 This filter accepts the following parameters:
9356 set first pixel component expression
9358 set second pixel component expression
9360 set third pixel component expression
9362 set fourth pixel component expression, corresponds to the alpha component
9365 Each of them specifies the expression to use for computing the lookup table for
9366 the corresponding pixel component values.
9368 The exact component associated to each of the @var{c*} options depends on the
9371 The expressions can contain the following constants:
9376 The input width and height.
9379 The first input value for the pixel component.
9382 The second input value for the pixel component.
9385 The first input video bit depth.
9388 The second input video bit depth.
9391 All expressions default to "x".
9393 @subsection Examples
9397 Highlight differences between two RGB video streams:
9399 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)'
9403 Highlight differences between two YUV video streams:
9405 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)'
9409 @section maskedclamp
9411 Clamp the first input stream with the second input and third input stream.
9413 Returns the value of first stream to be between second input
9414 stream - @code{undershoot} and third input stream + @code{overshoot}.
9416 This filter accepts the following options:
9419 Default value is @code{0}.
9422 Default value is @code{0}.
9425 Set which planes will be processed as bitmap, unprocessed planes will be
9426 copied from first stream.
9427 By default value 0xf, all planes will be processed.
9430 @section maskedmerge
9432 Merge the first input stream with the second input stream using per pixel
9433 weights in the third input stream.
9435 A value of 0 in the third stream pixel component means that pixel component
9436 from first stream is returned unchanged, while maximum value (eg. 255 for
9437 8-bit videos) means that pixel component from second stream is returned
9438 unchanged. Intermediate values define the amount of merging between both
9439 input stream's pixel components.
9441 This filter accepts the following options:
9444 Set which planes will be processed as bitmap, unprocessed planes will be
9445 copied from first stream.
9446 By default value 0xf, all planes will be processed.
9451 Apply motion-compensation deinterlacing.
9453 It needs one field per frame as input and must thus be used together
9454 with yadif=1/3 or equivalent.
9456 This filter accepts the following options:
9459 Set the deinterlacing mode.
9461 It accepts one of the following values:
9466 use iterative motion estimation
9468 like @samp{slow}, but use multiple reference frames.
9470 Default value is @samp{fast}.
9473 Set the picture field parity assumed for the input video. It must be
9474 one of the following values:
9478 assume top field first
9480 assume bottom field first
9483 Default value is @samp{bff}.
9486 Set per-block quantization parameter (QP) used by the internal
9489 Higher values should result in a smoother motion vector field but less
9490 optimal individual vectors. Default value is 1.
9493 @section mergeplanes
9495 Merge color channel components from several video streams.
9497 The filter accepts up to 4 input streams, and merge selected input
9498 planes to the output video.
9500 This filter accepts the following options:
9503 Set input to output plane mapping. Default is @code{0}.
9505 The mappings is specified as a bitmap. It should be specified as a
9506 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9507 mapping for the first plane of the output stream. 'A' sets the number of
9508 the input stream to use (from 0 to 3), and 'a' the plane number of the
9509 corresponding input to use (from 0 to 3). The rest of the mappings is
9510 similar, 'Bb' describes the mapping for the output stream second
9511 plane, 'Cc' describes the mapping for the output stream third plane and
9512 'Dd' describes the mapping for the output stream fourth plane.
9515 Set output pixel format. Default is @code{yuva444p}.
9518 @subsection Examples
9522 Merge three gray video streams of same width and height into single video stream:
9524 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9528 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9530 [a0][a1]mergeplanes=0x00010210:yuva444p
9534 Swap Y and A plane in yuva444p stream:
9536 format=yuva444p,mergeplanes=0x03010200:yuva444p
9540 Swap U and V plane in yuv420p stream:
9542 format=yuv420p,mergeplanes=0x000201:yuv420p
9546 Cast a rgb24 clip to yuv444p:
9548 format=rgb24,mergeplanes=0x000102:yuv444p
9554 Estimate and export motion vectors using block matching algorithms.
9555 Motion vectors are stored in frame side data to be used by other filters.
9557 This filter accepts the following options:
9560 Specify the motion estimation method. Accepts one of the following values:
9564 Exhaustive search algorithm.
9566 Three step search algorithm.
9568 Two dimensional logarithmic search algorithm.
9570 New three step search algorithm.
9572 Four step search algorithm.
9574 Diamond search algorithm.
9576 Hexagon-based search algorithm.
9578 Enhanced predictive zonal search algorithm.
9580 Uneven multi-hexagon search algorithm.
9582 Default value is @samp{esa}.
9585 Macroblock size. Default @code{16}.
9588 Search parameter. Default @code{7}.
9591 @section minterpolate
9593 Convert the video to specified frame rate using motion interpolation.
9595 This filter accepts the following options:
9598 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}.
9601 Motion interpolation mode. Following values are accepted:
9604 Duplicate previous or next frame for interpolating new ones.
9606 Blend source frames. Interpolated frame is mean of previous and next frames.
9608 Motion compensated interpolation. Following options are effective when this mode is selected:
9612 Motion compensation mode. Following values are accepted:
9615 Overlapped block motion compensation.
9617 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9619 Default mode is @samp{obmc}.
9622 Motion estimation mode. Following values are accepted:
9625 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9627 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9629 Default mode is @samp{bilat}.
9632 The algorithm to be used for motion estimation. Following values are accepted:
9635 Exhaustive search algorithm.
9637 Three step search algorithm.
9639 Two dimensional logarithmic search algorithm.
9641 New three step search algorithm.
9643 Four step search algorithm.
9645 Diamond search algorithm.
9647 Hexagon-based search algorithm.
9649 Enhanced predictive zonal search algorithm.
9651 Uneven multi-hexagon search algorithm.
9653 Default algorithm is @samp{epzs}.
9656 Macroblock size. Default @code{16}.
9659 Motion estimation search parameter. Default @code{32}.
9662 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).
9667 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:
9670 Disable scene change detection.
9672 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
9674 Default method is @samp{fdiff}.
9677 Scene change detection threshold. Default is @code{5.0}.
9682 Drop frames that do not differ greatly from the previous frame in
9683 order to reduce frame rate.
9685 The main use of this filter is for very-low-bitrate encoding
9686 (e.g. streaming over dialup modem), but it could in theory be used for
9687 fixing movies that were inverse-telecined incorrectly.
9689 A description of the accepted options follows.
9693 Set the maximum number of consecutive frames which can be dropped (if
9694 positive), or the minimum interval between dropped frames (if
9695 negative). If the value is 0, the frame is dropped unregarding the
9696 number of previous sequentially dropped frames.
9703 Set the dropping threshold values.
9705 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9706 represent actual pixel value differences, so a threshold of 64
9707 corresponds to 1 unit of difference for each pixel, or the same spread
9708 out differently over the block.
9710 A frame is a candidate for dropping if no 8x8 blocks differ by more
9711 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9712 meaning the whole image) differ by more than a threshold of @option{lo}.
9714 Default value for @option{hi} is 64*12, default value for @option{lo} is
9715 64*5, and default value for @option{frac} is 0.33.
9723 It accepts an integer in input; if non-zero it negates the
9724 alpha component (if available). The default value in input is 0.
9728 Denoise frames using Non-Local Means algorithm.
9730 Each pixel is adjusted by looking for other pixels with similar contexts. This
9731 context similarity is defined by comparing their surrounding patches of size
9732 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
9735 Note that the research area defines centers for patches, which means some
9736 patches will be made of pixels outside that research area.
9738 The filter accepts the following options.
9742 Set denoising strength.
9748 Same as @option{p} but for chroma planes.
9750 The default value is @var{0} and means automatic.
9756 Same as @option{r} but for chroma planes.
9758 The default value is @var{0} and means automatic.
9763 Deinterlace video using neural network edge directed interpolation.
9765 This filter accepts the following options:
9769 Mandatory option, without binary file filter can not work.
9770 Currently file can be found here:
9771 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9774 Set which frames to deinterlace, by default it is @code{all}.
9775 Can be @code{all} or @code{interlaced}.
9778 Set mode of operation.
9780 Can be one of the following:
9784 Use frame flags, both fields.
9786 Use frame flags, single field.
9790 Use bottom field only.
9792 Use both fields, top first.
9794 Use both fields, bottom first.
9798 Set which planes to process, by default filter process all frames.
9801 Set size of local neighborhood around each pixel, used by the predictor neural
9804 Can be one of the following:
9817 Set the number of neurons in predicctor neural network.
9818 Can be one of the following:
9829 Controls the number of different neural network predictions that are blended
9830 together to compute the final output value. Can be @code{fast}, default or
9834 Set which set of weights to use in the predictor.
9835 Can be one of the following:
9839 weights trained to minimize absolute error
9841 weights trained to minimize squared error
9845 Controls whether or not the prescreener neural network is used to decide
9846 which pixels should be processed by the predictor neural network and which
9847 can be handled by simple cubic interpolation.
9848 The prescreener is trained to know whether cubic interpolation will be
9849 sufficient for a pixel or whether it should be predicted by the predictor nn.
9850 The computational complexity of the prescreener nn is much less than that of
9851 the predictor nn. Since most pixels can be handled by cubic interpolation,
9852 using the prescreener generally results in much faster processing.
9853 The prescreener is pretty accurate, so the difference between using it and not
9854 using it is almost always unnoticeable.
9856 Can be one of the following:
9864 Default is @code{new}.
9867 Set various debugging flags.
9872 Force libavfilter not to use any of the specified pixel formats for the
9873 input to the next filter.
9875 It accepts the following parameters:
9879 A '|'-separated list of pixel format names, such as
9880 apix_fmts=yuv420p|monow|rgb24".
9884 @subsection Examples
9888 Force libavfilter to use a format different from @var{yuv420p} for the
9889 input to the vflip filter:
9891 noformat=pix_fmts=yuv420p,vflip
9895 Convert the input video to any of the formats not contained in the list:
9897 noformat=yuv420p|yuv444p|yuv410p
9903 Add noise on video input frame.
9905 The filter accepts the following options:
9913 Set noise seed for specific pixel component or all pixel components in case
9914 of @var{all_seed}. Default value is @code{123457}.
9916 @item all_strength, alls
9917 @item c0_strength, c0s
9918 @item c1_strength, c1s
9919 @item c2_strength, c2s
9920 @item c3_strength, c3s
9921 Set noise strength for specific pixel component or all pixel components in case
9922 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9924 @item all_flags, allf
9929 Set pixel component flags or set flags for all components if @var{all_flags}.
9930 Available values for component flags are:
9933 averaged temporal noise (smoother)
9935 mix random noise with a (semi)regular pattern
9937 temporal noise (noise pattern changes between frames)
9939 uniform noise (gaussian otherwise)
9943 @subsection Examples
9945 Add temporal and uniform noise to input video:
9947 noise=alls=20:allf=t+u
9952 Pass the video source unchanged to the output.
9955 Optical Character Recognition
9957 This filter uses Tesseract for optical character recognition.
9959 It accepts the following options:
9963 Set datapath to tesseract data. Default is to use whatever was
9964 set at installation.
9967 Set language, default is "eng".
9970 Set character whitelist.
9973 Set character blacklist.
9976 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9980 Apply a video transform using libopencv.
9982 To enable this filter, install the libopencv library and headers and
9983 configure FFmpeg with @code{--enable-libopencv}.
9985 It accepts the following parameters:
9990 The name of the libopencv filter to apply.
9993 The parameters to pass to the libopencv filter. If not specified, the default
9998 Refer to the official libopencv documentation for more precise
10000 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
10002 Several libopencv filters are supported; see the following subsections.
10007 Dilate an image by using a specific structuring element.
10008 It corresponds to the libopencv function @code{cvDilate}.
10010 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10012 @var{struct_el} represents a structuring element, and has the syntax:
10013 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10015 @var{cols} and @var{rows} represent the number of columns and rows of
10016 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10017 point, and @var{shape} the shape for the structuring element. @var{shape}
10018 must be "rect", "cross", "ellipse", or "custom".
10020 If the value for @var{shape} is "custom", it must be followed by a
10021 string of the form "=@var{filename}". The file with name
10022 @var{filename} is assumed to represent a binary image, with each
10023 printable character corresponding to a bright pixel. When a custom
10024 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10025 or columns and rows of the read file are assumed instead.
10027 The default value for @var{struct_el} is "3x3+0x0/rect".
10029 @var{nb_iterations} specifies the number of times the transform is
10030 applied to the image, and defaults to 1.
10034 # Use the default values
10037 # Dilate using a structuring element with a 5x5 cross, iterating two times
10038 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10040 # Read the shape from the file diamond.shape, iterating two times.
10041 # The file diamond.shape may contain a pattern of characters like this
10047 # The specified columns and rows are ignored
10048 # but the anchor point coordinates are not
10049 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10054 Erode an image by using a specific structuring element.
10055 It corresponds to the libopencv function @code{cvErode}.
10057 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10058 with the same syntax and semantics as the @ref{dilate} filter.
10062 Smooth the input video.
10064 The filter takes the following parameters:
10065 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10067 @var{type} is the type of smooth filter to apply, and must be one of
10068 the following values: "blur", "blur_no_scale", "median", "gaussian",
10069 or "bilateral". The default value is "gaussian".
10071 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10072 depend on the smooth type. @var{param1} and
10073 @var{param2} accept integer positive values or 0. @var{param3} and
10074 @var{param4} accept floating point values.
10076 The default value for @var{param1} is 3. The default value for the
10077 other parameters is 0.
10079 These parameters correspond to the parameters assigned to the
10080 libopencv function @code{cvSmooth}.
10085 Overlay one video on top of another.
10087 It takes two inputs and has one output. The first input is the "main"
10088 video on which the second input is overlaid.
10090 It accepts the following parameters:
10092 A description of the accepted options follows.
10097 Set the expression for the x and y coordinates of the overlaid video
10098 on the main video. Default value is "0" for both expressions. In case
10099 the expression is invalid, it is set to a huge value (meaning that the
10100 overlay will not be displayed within the output visible area).
10103 The action to take when EOF is encountered on the secondary input; it accepts
10104 one of the following values:
10108 Repeat the last frame (the default).
10112 Pass the main input through.
10116 Set when the expressions for @option{x}, and @option{y} are evaluated.
10118 It accepts the following values:
10121 only evaluate expressions once during the filter initialization or
10122 when a command is processed
10125 evaluate expressions for each incoming frame
10128 Default value is @samp{frame}.
10131 If set to 1, force the output to terminate when the shortest input
10132 terminates. Default value is 0.
10135 Set the format for the output video.
10137 It accepts the following values:
10140 force YUV420 output
10143 force YUV422 output
10146 force YUV444 output
10152 Default value is @samp{yuv420}.
10154 @item rgb @emph{(deprecated)}
10155 If set to 1, force the filter to accept inputs in the RGB
10156 color space. Default value is 0. This option is deprecated, use
10157 @option{format} instead.
10160 If set to 1, force the filter to draw the last overlay frame over the
10161 main input until the end of the stream. A value of 0 disables this
10162 behavior. Default value is 1.
10165 The @option{x}, and @option{y} expressions can contain the following
10171 The main input width and height.
10175 The overlay input width and height.
10179 The computed values for @var{x} and @var{y}. They are evaluated for
10184 horizontal and vertical chroma subsample values of the output
10185 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10189 the number of input frame, starting from 0
10192 the position in the file of the input frame, NAN if unknown
10195 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10199 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10200 when evaluation is done @emph{per frame}, and will evaluate to NAN
10201 when @option{eval} is set to @samp{init}.
10203 Be aware that frames are taken from each input video in timestamp
10204 order, hence, if their initial timestamps differ, it is a good idea
10205 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10206 have them begin in the same zero timestamp, as the example for
10207 the @var{movie} filter does.
10209 You can chain together more overlays but you should test the
10210 efficiency of such approach.
10212 @subsection Commands
10214 This filter supports the following commands:
10218 Modify the x and y of the overlay input.
10219 The command accepts the same syntax of the corresponding option.
10221 If the specified expression is not valid, it is kept at its current
10225 @subsection Examples
10229 Draw the overlay at 10 pixels from the bottom right corner of the main
10232 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10235 Using named options the example above becomes:
10237 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10241 Insert a transparent PNG logo in the bottom left corner of the input,
10242 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10244 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10248 Insert 2 different transparent PNG logos (second logo on bottom
10249 right corner) using the @command{ffmpeg} tool:
10251 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
10255 Add a transparent color layer on top of the main video; @code{WxH}
10256 must specify the size of the main input to the overlay filter:
10258 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10262 Play an original video and a filtered version (here with the deshake
10263 filter) side by side using the @command{ffplay} tool:
10265 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10268 The above command is the same as:
10270 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10274 Make a sliding overlay appearing from the left to the right top part of the
10275 screen starting since time 2:
10277 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10281 Compose output by putting two input videos side to side:
10283 ffmpeg -i left.avi -i right.avi -filter_complex "
10284 nullsrc=size=200x100 [background];
10285 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10286 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10287 [background][left] overlay=shortest=1 [background+left];
10288 [background+left][right] overlay=shortest=1:x=100 [left+right]
10293 Mask 10-20 seconds of a video by applying the delogo filter to a section
10295 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10296 -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]'
10301 Chain several overlays in cascade:
10303 nullsrc=s=200x200 [bg];
10304 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10305 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10306 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10307 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10308 [in3] null, [mid2] overlay=100:100 [out0]
10315 Apply Overcomplete Wavelet denoiser.
10317 The filter accepts the following options:
10323 Larger depth values will denoise lower frequency components more, but
10324 slow down filtering.
10326 Must be an int in the range 8-16, default is @code{8}.
10328 @item luma_strength, ls
10331 Must be a double value in the range 0-1000, default is @code{1.0}.
10333 @item chroma_strength, cs
10334 Set chroma strength.
10336 Must be a double value in the range 0-1000, default is @code{1.0}.
10342 Add paddings to the input image, and place the original input at the
10343 provided @var{x}, @var{y} coordinates.
10345 It accepts the following parameters:
10350 Specify an expression for the size of the output image with the
10351 paddings added. If the value for @var{width} or @var{height} is 0, the
10352 corresponding input size is used for the output.
10354 The @var{width} expression can reference the value set by the
10355 @var{height} expression, and vice versa.
10357 The default value of @var{width} and @var{height} is 0.
10361 Specify the offsets to place the input image at within the padded area,
10362 with respect to the top/left border of the output image.
10364 The @var{x} expression can reference the value set by the @var{y}
10365 expression, and vice versa.
10367 The default value of @var{x} and @var{y} is 0.
10370 Specify the color of the padded area. For the syntax of this option,
10371 check the "Color" section in the ffmpeg-utils manual.
10373 The default value of @var{color} is "black".
10376 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10377 options are expressions containing the following constants:
10382 The input video width and height.
10386 These are the same as @var{in_w} and @var{in_h}.
10390 The output width and height (the size of the padded area), as
10391 specified by the @var{width} and @var{height} expressions.
10395 These are the same as @var{out_w} and @var{out_h}.
10399 The x and y offsets as specified by the @var{x} and @var{y}
10400 expressions, or NAN if not yet specified.
10403 same as @var{iw} / @var{ih}
10406 input sample aspect ratio
10409 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10413 The horizontal and vertical chroma subsample values. For example for the
10414 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10417 @subsection Examples
10421 Add paddings with the color "violet" to the input video. The output video
10422 size is 640x480, and the top-left corner of the input video is placed at
10425 pad=640:480:0:40:violet
10428 The example above is equivalent to the following command:
10430 pad=width=640:height=480:x=0:y=40:color=violet
10434 Pad the input to get an output with dimensions increased by 3/2,
10435 and put the input video at the center of the padded area:
10437 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10441 Pad the input to get a squared output with size equal to the maximum
10442 value between the input width and height, and put the input video at
10443 the center of the padded area:
10445 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10449 Pad the input to get a final w/h ratio of 16:9:
10451 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10455 In case of anamorphic video, in order to set the output display aspect
10456 correctly, it is necessary to use @var{sar} in the expression,
10457 according to the relation:
10459 (ih * X / ih) * sar = output_dar
10460 X = output_dar / sar
10463 Thus the previous example needs to be modified to:
10465 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10469 Double the output size and put the input video in the bottom-right
10470 corner of the output padded area:
10472 pad="2*iw:2*ih:ow-iw:oh-ih"
10476 @anchor{palettegen}
10477 @section palettegen
10479 Generate one palette for a whole video stream.
10481 It accepts the following options:
10485 Set the maximum number of colors to quantize in the palette.
10486 Note: the palette will still contain 256 colors; the unused palette entries
10489 @item reserve_transparent
10490 Create a palette of 255 colors maximum and reserve the last one for
10491 transparency. Reserving the transparency color is useful for GIF optimization.
10492 If not set, the maximum of colors in the palette will be 256. You probably want
10493 to disable this option for a standalone image.
10497 Set statistics mode.
10499 It accepts the following values:
10502 Compute full frame histograms.
10504 Compute histograms only for the part that differs from previous frame. This
10505 might be relevant to give more importance to the moving part of your input if
10506 the background is static.
10508 Compute new histogram for each frame.
10511 Default value is @var{full}.
10514 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10515 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10516 color quantization of the palette. This information is also visible at
10517 @var{info} logging level.
10519 @subsection Examples
10523 Generate a representative palette of a given video using @command{ffmpeg}:
10525 ffmpeg -i input.mkv -vf palettegen palette.png
10529 @section paletteuse
10531 Use a palette to downsample an input video stream.
10533 The filter takes two inputs: one video stream and a palette. The palette must
10534 be a 256 pixels image.
10536 It accepts the following options:
10540 Select dithering mode. Available algorithms are:
10543 Ordered 8x8 bayer dithering (deterministic)
10545 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10546 Note: this dithering is sometimes considered "wrong" and is included as a
10548 @item floyd_steinberg
10549 Floyd and Steingberg dithering (error diffusion)
10551 Frankie Sierra dithering v2 (error diffusion)
10553 Frankie Sierra dithering v2 "Lite" (error diffusion)
10556 Default is @var{sierra2_4a}.
10559 When @var{bayer} dithering is selected, this option defines the scale of the
10560 pattern (how much the crosshatch pattern is visible). A low value means more
10561 visible pattern for less banding, and higher value means less visible pattern
10562 at the cost of more banding.
10564 The option must be an integer value in the range [0,5]. Default is @var{2}.
10567 If set, define the zone to process
10571 Only the changing rectangle will be reprocessed. This is similar to GIF
10572 cropping/offsetting compression mechanism. This option can be useful for speed
10573 if only a part of the image is changing, and has use cases such as limiting the
10574 scope of the error diffusal @option{dither} to the rectangle that bounds the
10575 moving scene (it leads to more deterministic output if the scene doesn't change
10576 much, and as a result less moving noise and better GIF compression).
10579 Default is @var{none}.
10582 Take new palette for each output frame.
10585 @subsection Examples
10589 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10590 using @command{ffmpeg}:
10592 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10596 @section perspective
10598 Correct perspective of video not recorded perpendicular to the screen.
10600 A description of the accepted parameters follows.
10611 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10612 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10613 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10614 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10615 then the corners of the source will be sent to the specified coordinates.
10617 The expressions can use the following variables:
10622 the width and height of video frame.
10626 Output frame count.
10629 @item interpolation
10630 Set interpolation for perspective correction.
10632 It accepts the following values:
10638 Default value is @samp{linear}.
10641 Set interpretation of coordinate options.
10643 It accepts the following values:
10647 Send point in the source specified by the given coordinates to
10648 the corners of the destination.
10650 @item 1, destination
10652 Send the corners of the source to the point in the destination specified
10653 by the given coordinates.
10655 Default value is @samp{source}.
10659 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10661 It accepts the following values:
10664 only evaluate expressions once during the filter initialization or
10665 when a command is processed
10668 evaluate expressions for each incoming frame
10671 Default value is @samp{init}.
10676 Delay interlaced video by one field time so that the field order changes.
10678 The intended use is to fix PAL movies that have been captured with the
10679 opposite field order to the film-to-video transfer.
10681 A description of the accepted parameters follows.
10687 It accepts the following values:
10690 Capture field order top-first, transfer bottom-first.
10691 Filter will delay the bottom field.
10694 Capture field order bottom-first, transfer top-first.
10695 Filter will delay the top field.
10698 Capture and transfer with the same field order. This mode only exists
10699 for the documentation of the other options to refer to, but if you
10700 actually select it, the filter will faithfully do nothing.
10703 Capture field order determined automatically by field flags, transfer
10705 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10706 basis using field flags. If no field information is available,
10707 then this works just like @samp{u}.
10710 Capture unknown or varying, transfer opposite.
10711 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10712 analyzing the images and selecting the alternative that produces best
10713 match between the fields.
10716 Capture top-first, transfer unknown or varying.
10717 Filter selects among @samp{t} and @samp{p} using image analysis.
10720 Capture bottom-first, transfer unknown or varying.
10721 Filter selects among @samp{b} and @samp{p} using image analysis.
10724 Capture determined by field flags, transfer unknown or varying.
10725 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10726 image analysis. If no field information is available, then this works just
10727 like @samp{U}. This is the default mode.
10730 Both capture and transfer unknown or varying.
10731 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10735 @section pixdesctest
10737 Pixel format descriptor test filter, mainly useful for internal
10738 testing. The output video should be equal to the input video.
10742 format=monow, pixdesctest
10745 can be used to test the monowhite pixel format descriptor definition.
10749 Enable the specified chain of postprocessing subfilters using libpostproc. This
10750 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10751 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10752 Each subfilter and some options have a short and a long name that can be used
10753 interchangeably, i.e. dr/dering are the same.
10755 The filters accept the following options:
10759 Set postprocessing subfilters string.
10762 All subfilters share common options to determine their scope:
10766 Honor the quality commands for this subfilter.
10769 Do chrominance filtering, too (default).
10772 Do luminance filtering only (no chrominance).
10775 Do chrominance filtering only (no luminance).
10778 These options can be appended after the subfilter name, separated by a '|'.
10780 Available subfilters are:
10783 @item hb/hdeblock[|difference[|flatness]]
10784 Horizontal deblocking filter
10787 Difference factor where higher values mean more deblocking (default: @code{32}).
10789 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10792 @item vb/vdeblock[|difference[|flatness]]
10793 Vertical deblocking filter
10796 Difference factor where higher values mean more deblocking (default: @code{32}).
10798 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10801 @item ha/hadeblock[|difference[|flatness]]
10802 Accurate horizontal deblocking filter
10805 Difference factor where higher values mean more deblocking (default: @code{32}).
10807 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10810 @item va/vadeblock[|difference[|flatness]]
10811 Accurate vertical deblocking filter
10814 Difference factor where higher values mean more deblocking (default: @code{32}).
10816 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10820 The horizontal and vertical deblocking filters share the difference and
10821 flatness values so you cannot set different horizontal and vertical
10825 @item h1/x1hdeblock
10826 Experimental horizontal deblocking filter
10828 @item v1/x1vdeblock
10829 Experimental vertical deblocking filter
10834 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10837 larger -> stronger filtering
10839 larger -> stronger filtering
10841 larger -> stronger filtering
10844 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10847 Stretch luminance to @code{0-255}.
10850 @item lb/linblenddeint
10851 Linear blend deinterlacing filter that deinterlaces the given block by
10852 filtering all lines with a @code{(1 2 1)} filter.
10854 @item li/linipoldeint
10855 Linear interpolating deinterlacing filter that deinterlaces the given block by
10856 linearly interpolating every second line.
10858 @item ci/cubicipoldeint
10859 Cubic interpolating deinterlacing filter deinterlaces the given block by
10860 cubically interpolating every second line.
10862 @item md/mediandeint
10863 Median deinterlacing filter that deinterlaces the given block by applying a
10864 median filter to every second line.
10866 @item fd/ffmpegdeint
10867 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10868 second line with a @code{(-1 4 2 4 -1)} filter.
10871 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10872 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10874 @item fq/forceQuant[|quantizer]
10875 Overrides the quantizer table from the input with the constant quantizer you
10883 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10886 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10889 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10892 @subsection Examples
10896 Apply horizontal and vertical deblocking, deringing and automatic
10897 brightness/contrast:
10903 Apply default filters without brightness/contrast correction:
10909 Apply default filters and temporal denoiser:
10911 pp=default/tmpnoise|1|2|3
10915 Apply deblocking on luminance only, and switch vertical deblocking on or off
10916 automatically depending on available CPU time:
10923 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10924 similar to spp = 6 with 7 point DCT, where only the center sample is
10927 The filter accepts the following options:
10931 Force a constant quantization parameter. It accepts an integer in range
10932 0 to 63. If not set, the filter will use the QP from the video stream
10936 Set thresholding mode. Available modes are:
10940 Set hard thresholding.
10942 Set soft thresholding (better de-ringing effect, but likely blurrier).
10944 Set medium thresholding (good results, default).
10949 Apply prewitt operator to input video stream.
10951 The filter accepts the following option:
10955 Set which planes will be processed, unprocessed planes will be copied.
10956 By default value 0xf, all planes will be processed.
10959 Set value which will be multiplied with filtered result.
10962 Set value which will be added to filtered result.
10967 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10968 Ratio) between two input videos.
10970 This filter takes in input two input videos, the first input is
10971 considered the "main" source and is passed unchanged to the
10972 output. The second input is used as a "reference" video for computing
10975 Both video inputs must have the same resolution and pixel format for
10976 this filter to work correctly. Also it assumes that both inputs
10977 have the same number of frames, which are compared one by one.
10979 The obtained average PSNR is printed through the logging system.
10981 The filter stores the accumulated MSE (mean squared error) of each
10982 frame, and at the end of the processing it is averaged across all frames
10983 equally, and the following formula is applied to obtain the PSNR:
10986 PSNR = 10*log10(MAX^2/MSE)
10989 Where MAX is the average of the maximum values of each component of the
10992 The description of the accepted parameters follows.
10995 @item stats_file, f
10996 If specified the filter will use the named file to save the PSNR of
10997 each individual frame. When filename equals "-" the data is sent to
11000 @item stats_version
11001 Specifies which version of the stats file format to use. Details of
11002 each format are written below.
11003 Default value is 1.
11005 @item stats_add_max
11006 Determines whether the max value is output to the stats log.
11007 Default value is 0.
11008 Requires stats_version >= 2. If this is set and stats_version < 2,
11009 the filter will return an error.
11012 The file printed if @var{stats_file} is selected, contains a sequence of
11013 key/value pairs of the form @var{key}:@var{value} for each compared
11016 If a @var{stats_version} greater than 1 is specified, a header line precedes
11017 the list of per-frame-pair stats, with key value pairs following the frame
11018 format with the following parameters:
11021 @item psnr_log_version
11022 The version of the log file format. Will match @var{stats_version}.
11025 A comma separated list of the per-frame-pair parameters included in
11029 A description of each shown per-frame-pair parameter follows:
11033 sequential number of the input frame, starting from 1
11036 Mean Square Error pixel-by-pixel average difference of the compared
11037 frames, averaged over all the image components.
11039 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
11040 Mean Square Error pixel-by-pixel average difference of the compared
11041 frames for the component specified by the suffix.
11043 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
11044 Peak Signal to Noise ratio of the compared frames for the component
11045 specified by the suffix.
11047 @item max_avg, max_y, max_u, max_v
11048 Maximum allowed value for each channel, and average over all
11054 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
11055 [main][ref] psnr="stats_file=stats.log" [out]
11058 On this example the input file being processed is compared with the
11059 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
11060 is stored in @file{stats.log}.
11065 Pulldown reversal (inverse telecine) filter, capable of handling mixed
11066 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
11069 The pullup filter is designed to take advantage of future context in making
11070 its decisions. This filter is stateless in the sense that it does not lock
11071 onto a pattern to follow, but it instead looks forward to the following
11072 fields in order to identify matches and rebuild progressive frames.
11074 To produce content with an even framerate, insert the fps filter after
11075 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
11076 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
11078 The filter accepts the following options:
11085 These options set the amount of "junk" to ignore at the left, right, top, and
11086 bottom of the image, respectively. Left and right are in units of 8 pixels,
11087 while top and bottom are in units of 2 lines.
11088 The default is 8 pixels on each side.
11091 Set the strict breaks. Setting this option to 1 will reduce the chances of
11092 filter generating an occasional mismatched frame, but it may also cause an
11093 excessive number of frames to be dropped during high motion sequences.
11094 Conversely, setting it to -1 will make filter match fields more easily.
11095 This may help processing of video where there is slight blurring between
11096 the fields, but may also cause there to be interlaced frames in the output.
11097 Default value is @code{0}.
11100 Set the metric plane to use. It accepts the following values:
11106 Use chroma blue plane.
11109 Use chroma red plane.
11112 This option may be set to use chroma plane instead of the default luma plane
11113 for doing filter's computations. This may improve accuracy on very clean
11114 source material, but more likely will decrease accuracy, especially if there
11115 is chroma noise (rainbow effect) or any grayscale video.
11116 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
11117 load and make pullup usable in realtime on slow machines.
11120 For best results (without duplicated frames in the output file) it is
11121 necessary to change the output frame rate. For example, to inverse
11122 telecine NTSC input:
11124 ffmpeg -i input -vf pullup -r 24000/1001 ...
11129 Change video quantization parameters (QP).
11131 The filter accepts the following option:
11135 Set expression for quantization parameter.
11138 The expression is evaluated through the eval API and can contain, among others,
11139 the following constants:
11143 1 if index is not 129, 0 otherwise.
11146 Sequentional index starting from -129 to 128.
11149 @subsection Examples
11153 Some equation like:
11161 Flush video frames from internal cache of frames into a random order.
11162 No frame is discarded.
11163 Inspired by @ref{frei0r} nervous filter.
11167 Set size in number of frames of internal cache, in range from @code{2} to
11168 @code{512}. Default is @code{30}.
11171 Set seed for random number generator, must be an integer included between
11172 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11173 less than @code{0}, the filter will try to use a good random seed on a
11179 Read vertical interval timecode (VITC) information from the top lines of a
11182 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11183 timecode value, if a valid timecode has been detected. Further metadata key
11184 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11185 timecode data has been found or not.
11187 This filter accepts the following options:
11191 Set the maximum number of lines to scan for VITC data. If the value is set to
11192 @code{-1} the full video frame is scanned. Default is @code{45}.
11195 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11196 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11199 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11200 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11203 @subsection Examples
11207 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11208 draw @code{--:--:--:--} as a placeholder:
11210 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11216 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11218 Destination pixel at position (X, Y) will be picked from source (x, y) position
11219 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11220 value for pixel will be used for destination pixel.
11222 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11223 will have Xmap/Ymap video stream dimensions.
11224 Xmap and Ymap input video streams are 16bit depth, single channel.
11226 @section removegrain
11228 The removegrain filter is a spatial denoiser for progressive video.
11232 Set mode for the first plane.
11235 Set mode for the second plane.
11238 Set mode for the third plane.
11241 Set mode for the fourth plane.
11244 Range of mode is from 0 to 24. Description of each mode follows:
11248 Leave input plane unchanged. Default.
11251 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11254 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11257 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11260 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11261 This is equivalent to a median filter.
11264 Line-sensitive clipping giving the minimal change.
11267 Line-sensitive clipping, intermediate.
11270 Line-sensitive clipping, intermediate.
11273 Line-sensitive clipping, intermediate.
11276 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11279 Replaces the target pixel with the closest neighbour.
11282 [1 2 1] horizontal and vertical kernel blur.
11288 Bob mode, interpolates top field from the line where the neighbours
11289 pixels are the closest.
11292 Bob mode, interpolates bottom field from the line where the neighbours
11293 pixels are the closest.
11296 Bob mode, interpolates top field. Same as 13 but with a more complicated
11297 interpolation formula.
11300 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11301 interpolation formula.
11304 Clips the pixel with the minimum and maximum of respectively the maximum and
11305 minimum of each pair of opposite neighbour pixels.
11308 Line-sensitive clipping using opposite neighbours whose greatest distance from
11309 the current pixel is minimal.
11312 Replaces the pixel with the average of its 8 neighbours.
11315 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11318 Clips pixels using the averages of opposite neighbour.
11321 Same as mode 21 but simpler and faster.
11324 Small edge and halo removal, but reputed useless.
11330 @section removelogo
11332 Suppress a TV station logo, using an image file to determine which
11333 pixels comprise the logo. It works by filling in the pixels that
11334 comprise the logo with neighboring pixels.
11336 The filter accepts the following options:
11340 Set the filter bitmap file, which can be any image format supported by
11341 libavformat. The width and height of the image file must match those of the
11342 video stream being processed.
11345 Pixels in the provided bitmap image with a value of zero are not
11346 considered part of the logo, non-zero pixels are considered part of
11347 the logo. If you use white (255) for the logo and black (0) for the
11348 rest, you will be safe. For making the filter bitmap, it is
11349 recommended to take a screen capture of a black frame with the logo
11350 visible, and then using a threshold filter followed by the erode
11351 filter once or twice.
11353 If needed, little splotches can be fixed manually. Remember that if
11354 logo pixels are not covered, the filter quality will be much
11355 reduced. Marking too many pixels as part of the logo does not hurt as
11356 much, but it will increase the amount of blurring needed to cover over
11357 the image and will destroy more information than necessary, and extra
11358 pixels will slow things down on a large logo.
11360 @section repeatfields
11362 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11363 fields based on its value.
11367 Reverse a video clip.
11369 Warning: This filter requires memory to buffer the entire clip, so trimming
11372 @subsection Examples
11376 Take the first 5 seconds of a clip, and reverse it.
11384 Rotate video by an arbitrary angle expressed in radians.
11386 The filter accepts the following options:
11388 A description of the optional parameters follows.
11391 Set an expression for the angle by which to rotate the input video
11392 clockwise, expressed as a number of radians. A negative value will
11393 result in a counter-clockwise rotation. By default it is set to "0".
11395 This expression is evaluated for each frame.
11398 Set the output width expression, default value is "iw".
11399 This expression is evaluated just once during configuration.
11402 Set the output height expression, default value is "ih".
11403 This expression is evaluated just once during configuration.
11406 Enable bilinear interpolation if set to 1, a value of 0 disables
11407 it. Default value is 1.
11410 Set the color used to fill the output area not covered by the rotated
11411 image. For the general syntax of this option, check the "Color" section in the
11412 ffmpeg-utils manual. If the special value "none" is selected then no
11413 background is printed (useful for example if the background is never shown).
11415 Default value is "black".
11418 The expressions for the angle and the output size can contain the
11419 following constants and functions:
11423 sequential number of the input frame, starting from 0. It is always NAN
11424 before the first frame is filtered.
11427 time in seconds of the input frame, it is set to 0 when the filter is
11428 configured. It is always NAN before the first frame is filtered.
11432 horizontal and vertical chroma subsample values. For example for the
11433 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11437 the input video width and height
11441 the output width and height, that is the size of the padded area as
11442 specified by the @var{width} and @var{height} expressions
11446 the minimal width/height required for completely containing the input
11447 video rotated by @var{a} radians.
11449 These are only available when computing the @option{out_w} and
11450 @option{out_h} expressions.
11453 @subsection Examples
11457 Rotate the input by PI/6 radians clockwise:
11463 Rotate the input by PI/6 radians counter-clockwise:
11469 Rotate the input by 45 degrees clockwise:
11475 Apply a constant rotation with period T, starting from an angle of PI/3:
11477 rotate=PI/3+2*PI*t/T
11481 Make the input video rotation oscillating with a period of T
11482 seconds and an amplitude of A radians:
11484 rotate=A*sin(2*PI/T*t)
11488 Rotate the video, output size is chosen so that the whole rotating
11489 input video is always completely contained in the output:
11491 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11495 Rotate the video, reduce the output size so that no background is ever
11498 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11502 @subsection Commands
11504 The filter supports the following commands:
11508 Set the angle expression.
11509 The command accepts the same syntax of the corresponding option.
11511 If the specified expression is not valid, it is kept at its current
11517 Apply Shape Adaptive Blur.
11519 The filter accepts the following options:
11522 @item luma_radius, lr
11523 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11524 value is 1.0. A greater value will result in a more blurred image, and
11525 in slower processing.
11527 @item luma_pre_filter_radius, lpfr
11528 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11531 @item luma_strength, ls
11532 Set luma maximum difference between pixels to still be considered, must
11533 be a value in the 0.1-100.0 range, default value is 1.0.
11535 @item chroma_radius, cr
11536 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11537 greater value will result in a more blurred image, and in slower
11540 @item chroma_pre_filter_radius, cpfr
11541 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11543 @item chroma_strength, cs
11544 Set chroma maximum difference between pixels to still be considered,
11545 must be a value in the -0.9-100.0 range.
11548 Each chroma option value, if not explicitly specified, is set to the
11549 corresponding luma option value.
11554 Scale (resize) the input video, using the libswscale library.
11556 The scale filter forces the output display aspect ratio to be the same
11557 of the input, by changing the output sample aspect ratio.
11559 If the input image format is different from the format requested by
11560 the next filter, the scale filter will convert the input to the
11563 @subsection Options
11564 The filter accepts the following options, or any of the options
11565 supported by the libswscale scaler.
11567 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11568 the complete list of scaler options.
11573 Set the output video dimension expression. Default value is the input
11576 If the value is 0, the input width is used for the output.
11578 If one of the values is -1, the scale filter will use a value that
11579 maintains the aspect ratio of the input image, calculated from the
11580 other specified dimension. If both of them are -1, the input size is
11583 If one of the values is -n with n > 1, the scale filter will also use a value
11584 that maintains the aspect ratio of the input image, calculated from the other
11585 specified dimension. After that it will, however, make sure that the calculated
11586 dimension is divisible by n and adjust the value if necessary.
11588 See below for the list of accepted constants for use in the dimension
11592 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11596 Only evaluate expressions once during the filter initialization or when a command is processed.
11599 Evaluate expressions for each incoming frame.
11603 Default value is @samp{init}.
11607 Set the interlacing mode. It accepts the following values:
11611 Force interlaced aware scaling.
11614 Do not apply interlaced scaling.
11617 Select interlaced aware scaling depending on whether the source frames
11618 are flagged as interlaced or not.
11621 Default value is @samp{0}.
11624 Set libswscale scaling flags. See
11625 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11626 complete list of values. If not explicitly specified the filter applies
11630 @item param0, param1
11631 Set libswscale input parameters for scaling algorithms that need them. See
11632 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11633 complete documentation. If not explicitly specified the filter applies
11639 Set the video size. For the syntax of this option, check the
11640 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11642 @item in_color_matrix
11643 @item out_color_matrix
11644 Set in/output YCbCr color space type.
11646 This allows the autodetected value to be overridden as well as allows forcing
11647 a specific value used for the output and encoder.
11649 If not specified, the color space type depends on the pixel format.
11655 Choose automatically.
11658 Format conforming to International Telecommunication Union (ITU)
11659 Recommendation BT.709.
11662 Set color space conforming to the United States Federal Communications
11663 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11666 Set color space conforming to:
11670 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11673 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11676 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11681 Set color space conforming to SMPTE ST 240:1999.
11686 Set in/output YCbCr sample range.
11688 This allows the autodetected value to be overridden as well as allows forcing
11689 a specific value used for the output and encoder. If not specified, the
11690 range depends on the pixel format. Possible values:
11694 Choose automatically.
11697 Set full range (0-255 in case of 8-bit luma).
11700 Set "MPEG" range (16-235 in case of 8-bit luma).
11703 @item force_original_aspect_ratio
11704 Enable decreasing or increasing output video width or height if necessary to
11705 keep the original aspect ratio. Possible values:
11709 Scale the video as specified and disable this feature.
11712 The output video dimensions will automatically be decreased if needed.
11715 The output video dimensions will automatically be increased if needed.
11719 One useful instance of this option is that when you know a specific device's
11720 maximum allowed resolution, you can use this to limit the output video to
11721 that, while retaining the aspect ratio. For example, device A allows
11722 1280x720 playback, and your video is 1920x800. Using this option (set it to
11723 decrease) and specifying 1280x720 to the command line makes the output
11726 Please note that this is a different thing than specifying -1 for @option{w}
11727 or @option{h}, you still need to specify the output resolution for this option
11732 The values of the @option{w} and @option{h} options are expressions
11733 containing the following constants:
11738 The input width and height
11742 These are the same as @var{in_w} and @var{in_h}.
11746 The output (scaled) width and height
11750 These are the same as @var{out_w} and @var{out_h}
11753 The same as @var{iw} / @var{ih}
11756 input sample aspect ratio
11759 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11763 horizontal and vertical input chroma subsample values. For example for the
11764 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11768 horizontal and vertical output chroma subsample values. For example for the
11769 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11772 @subsection Examples
11776 Scale the input video to a size of 200x100
11781 This is equivalent to:
11792 Specify a size abbreviation for the output size:
11797 which can also be written as:
11803 Scale the input to 2x:
11805 scale=w=2*iw:h=2*ih
11809 The above is the same as:
11811 scale=2*in_w:2*in_h
11815 Scale the input to 2x with forced interlaced scaling:
11817 scale=2*iw:2*ih:interl=1
11821 Scale the input to half size:
11823 scale=w=iw/2:h=ih/2
11827 Increase the width, and set the height to the same size:
11833 Seek Greek harmony:
11840 Increase the height, and set the width to 3/2 of the height:
11842 scale=w=3/2*oh:h=3/5*ih
11846 Increase the size, making the size a multiple of the chroma
11849 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11853 Increase the width to a maximum of 500 pixels,
11854 keeping the same aspect ratio as the input:
11856 scale=w='min(500\, iw*3/2):h=-1'
11860 @subsection Commands
11862 This filter supports the following commands:
11866 Set the output video dimension expression.
11867 The command accepts the same syntax of the corresponding option.
11869 If the specified expression is not valid, it is kept at its current
11875 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11876 format conversion on CUDA video frames. Setting the output width and height
11877 works in the same way as for the @var{scale} filter.
11879 The following additional options are accepted:
11882 The pixel format of the output CUDA frames. If set to the string "same" (the
11883 default), the input format will be kept. Note that automatic format negotiation
11884 and conversion is not yet supported for hardware frames
11887 The interpolation algorithm used for resizing. One of the following:
11894 @item cubic2p_bspline
11895 2-parameter cubic (B=1, C=0)
11897 @item cubic2p_catmullrom
11898 2-parameter cubic (B=0, C=1/2)
11900 @item cubic2p_b05c03
11901 2-parameter cubic (B=1/2, C=3/10)
11913 Scale (resize) the input video, based on a reference video.
11915 See the scale filter for available options, scale2ref supports the same but
11916 uses the reference video instead of the main input as basis.
11918 @subsection Examples
11922 Scale a subtitle stream to match the main video in size before overlaying
11924 'scale2ref[b][a];[a][b]overlay'
11928 @anchor{selectivecolor}
11929 @section selectivecolor
11931 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11932 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11933 by the "purity" of the color (that is, how saturated it already is).
11935 This filter is similar to the Adobe Photoshop Selective Color tool.
11937 The filter accepts the following options:
11940 @item correction_method
11941 Select color correction method.
11943 Available values are:
11946 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11949 Specified adjustments are relative to the original component value.
11951 Default is @code{absolute}.
11953 Adjustments for red pixels (pixels where the red component is the maximum)
11955 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11957 Adjustments for green pixels (pixels where the green component is the maximum)
11959 Adjustments for cyan pixels (pixels where the red component is the minimum)
11961 Adjustments for blue pixels (pixels where the blue component is the maximum)
11963 Adjustments for magenta pixels (pixels where the green component is the minimum)
11965 Adjustments for white pixels (pixels where all components are greater than 128)
11967 Adjustments for all pixels except pure black and pure white
11969 Adjustments for black pixels (pixels where all components are lesser than 128)
11971 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11974 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11975 4 space separated floating point adjustment values in the [-1,1] range,
11976 respectively to adjust the amount of cyan, magenta, yellow and black for the
11977 pixels of its range.
11979 @subsection Examples
11983 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11984 increase magenta by 27% in blue areas:
11986 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11990 Use a Photoshop selective color preset:
11992 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11996 @anchor{separatefields}
11997 @section separatefields
11999 The @code{separatefields} takes a frame-based video input and splits
12000 each frame into its components fields, producing a new half height clip
12001 with twice the frame rate and twice the frame count.
12003 This filter use field-dominance information in frame to decide which
12004 of each pair of fields to place first in the output.
12005 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
12007 @section setdar, setsar
12009 The @code{setdar} filter sets the Display Aspect Ratio for the filter
12012 This is done by changing the specified Sample (aka Pixel) Aspect
12013 Ratio, according to the following equation:
12015 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
12018 Keep in mind that the @code{setdar} filter does not modify the pixel
12019 dimensions of the video frame. Also, the display aspect ratio set by
12020 this filter may be changed by later filters in the filterchain,
12021 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
12024 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
12025 the filter output video.
12027 Note that as a consequence of the application of this filter, the
12028 output display aspect ratio will change according to the equation
12031 Keep in mind that the sample aspect ratio set by the @code{setsar}
12032 filter may be changed by later filters in the filterchain, e.g. if
12033 another "setsar" or a "setdar" filter is applied.
12035 It accepts the following parameters:
12038 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
12039 Set the aspect ratio used by the filter.
12041 The parameter can be a floating point number string, an expression, or
12042 a string of the form @var{num}:@var{den}, where @var{num} and
12043 @var{den} are the numerator and denominator of the aspect ratio. If
12044 the parameter is not specified, it is assumed the value "0".
12045 In case the form "@var{num}:@var{den}" is used, the @code{:} character
12049 Set the maximum integer value to use for expressing numerator and
12050 denominator when reducing the expressed aspect ratio to a rational.
12051 Default value is @code{100}.
12055 The parameter @var{sar} is an expression containing
12056 the following constants:
12060 These are approximated values for the mathematical constants e
12061 (Euler's number), pi (Greek pi), and phi (the golden ratio).
12064 The input width and height.
12067 These are the same as @var{w} / @var{h}.
12070 The input sample aspect ratio.
12073 The input display aspect ratio. It is the same as
12074 (@var{w} / @var{h}) * @var{sar}.
12077 Horizontal and vertical chroma subsample values. For example, for the
12078 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12081 @subsection Examples
12086 To change the display aspect ratio to 16:9, specify one of the following:
12093 To change the sample aspect ratio to 10:11, specify:
12099 To set a display aspect ratio of 16:9, and specify a maximum integer value of
12100 1000 in the aspect ratio reduction, use the command:
12102 setdar=ratio=16/9:max=1000
12110 Force field for the output video frame.
12112 The @code{setfield} filter marks the interlace type field for the
12113 output frames. It does not change the input frame, but only sets the
12114 corresponding property, which affects how the frame is treated by
12115 following filters (e.g. @code{fieldorder} or @code{yadif}).
12117 The filter accepts the following options:
12122 Available values are:
12126 Keep the same field property.
12129 Mark the frame as bottom-field-first.
12132 Mark the frame as top-field-first.
12135 Mark the frame as progressive.
12141 Show a line containing various information for each input video frame.
12142 The input video is not modified.
12144 The shown line contains a sequence of key/value pairs of the form
12145 @var{key}:@var{value}.
12147 The following values are shown in the output:
12151 The (sequential) number of the input frame, starting from 0.
12154 The Presentation TimeStamp of the input frame, expressed as a number of
12155 time base units. The time base unit depends on the filter input pad.
12158 The Presentation TimeStamp of the input frame, expressed as a number of
12162 The position of the frame in the input stream, or -1 if this information is
12163 unavailable and/or meaningless (for example in case of synthetic video).
12166 The pixel format name.
12169 The sample aspect ratio of the input frame, expressed in the form
12170 @var{num}/@var{den}.
12173 The size of the input frame. For the syntax of this option, check the
12174 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12177 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12178 for bottom field first).
12181 This is 1 if the frame is a key frame, 0 otherwise.
12184 The picture type of the input frame ("I" for an I-frame, "P" for a
12185 P-frame, "B" for a B-frame, or "?" for an unknown type).
12186 Also refer to the documentation of the @code{AVPictureType} enum and of
12187 the @code{av_get_picture_type_char} function defined in
12188 @file{libavutil/avutil.h}.
12191 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12193 @item plane_checksum
12194 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12195 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12198 @section showpalette
12200 Displays the 256 colors palette of each frame. This filter is only relevant for
12201 @var{pal8} pixel format frames.
12203 It accepts the following option:
12207 Set the size of the box used to represent one palette color entry. Default is
12208 @code{30} (for a @code{30x30} pixel box).
12211 @section shuffleframes
12213 Reorder and/or duplicate video frames.
12215 It accepts the following parameters:
12219 Set the destination indexes of input frames.
12220 This is space or '|' separated list of indexes that maps input frames to output
12221 frames. Number of indexes also sets maximal value that each index may have.
12224 The first frame has the index 0. The default is to keep the input unchanged.
12226 @subsection Examples
12230 Swap second and third frame of every three frames of the input:
12232 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12236 Swap 10th and 1st frame of every ten frames of the input:
12238 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
12242 @section shuffleplanes
12244 Reorder and/or duplicate video planes.
12246 It accepts the following parameters:
12251 The index of the input plane to be used as the first output plane.
12254 The index of the input plane to be used as the second output plane.
12257 The index of the input plane to be used as the third output plane.
12260 The index of the input plane to be used as the fourth output plane.
12264 The first plane has the index 0. The default is to keep the input unchanged.
12266 @subsection Examples
12270 Swap the second and third planes of the input:
12272 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12276 @anchor{signalstats}
12277 @section signalstats
12278 Evaluate various visual metrics that assist in determining issues associated
12279 with the digitization of analog video media.
12281 By default the filter will log these metadata values:
12285 Display the minimal Y value contained within the input frame. Expressed in
12289 Display the Y value at the 10% percentile within the input frame. Expressed in
12293 Display the average Y value within the input frame. Expressed in range of
12297 Display the Y value at the 90% percentile within the input frame. Expressed in
12301 Display the maximum Y value contained within the input frame. Expressed in
12305 Display the minimal U value contained within the input frame. Expressed in
12309 Display the U value at the 10% percentile within the input frame. Expressed in
12313 Display the average U value within the input frame. Expressed in range of
12317 Display the U value at the 90% percentile within the input frame. Expressed in
12321 Display the maximum U value contained within the input frame. Expressed in
12325 Display the minimal V value contained within the input frame. Expressed in
12329 Display the V value at the 10% percentile within the input frame. Expressed in
12333 Display the average V value within the input frame. Expressed in range of
12337 Display the V value at the 90% percentile within the input frame. Expressed in
12341 Display the maximum V value contained within the input frame. Expressed in
12345 Display the minimal saturation value contained within the input frame.
12346 Expressed in range of [0-~181.02].
12349 Display the saturation value at the 10% percentile within the input frame.
12350 Expressed in range of [0-~181.02].
12353 Display the average saturation value within the input frame. Expressed in range
12357 Display the saturation value at the 90% percentile within the input frame.
12358 Expressed in range of [0-~181.02].
12361 Display the maximum saturation value contained within the input frame.
12362 Expressed in range of [0-~181.02].
12365 Display the median value for hue within the input frame. Expressed in range of
12369 Display the average value for hue within the input frame. Expressed in range of
12373 Display the average of sample value difference between all values of the Y
12374 plane in the current frame and corresponding values of the previous input frame.
12375 Expressed in range of [0-255].
12378 Display the average of sample value difference between all values of the U
12379 plane in the current frame and corresponding values of the previous input frame.
12380 Expressed in range of [0-255].
12383 Display the average of sample value difference between all values of the V
12384 plane in the current frame and corresponding values of the previous input frame.
12385 Expressed in range of [0-255].
12388 Display bit depth of Y plane in current frame.
12389 Expressed in range of [0-16].
12392 Display bit depth of U plane in current frame.
12393 Expressed in range of [0-16].
12396 Display bit depth of V plane in current frame.
12397 Expressed in range of [0-16].
12400 The filter accepts the following options:
12406 @option{stat} specify an additional form of image analysis.
12407 @option{out} output video with the specified type of pixel highlighted.
12409 Both options accept the following values:
12413 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12414 unlike the neighboring pixels of the same field. Examples of temporal outliers
12415 include the results of video dropouts, head clogs, or tape tracking issues.
12418 Identify @var{vertical line repetition}. Vertical line repetition includes
12419 similar rows of pixels within a frame. In born-digital video vertical line
12420 repetition is common, but this pattern is uncommon in video digitized from an
12421 analog source. When it occurs in video that results from the digitization of an
12422 analog source it can indicate concealment from a dropout compensator.
12425 Identify pixels that fall outside of legal broadcast range.
12429 Set the highlight color for the @option{out} option. The default color is
12433 @subsection Examples
12437 Output data of various video metrics:
12439 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12443 Output specific data about the minimum and maximum values of the Y plane per frame:
12445 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12449 Playback video while highlighting pixels that are outside of broadcast range in red.
12451 ffplay example.mov -vf signalstats="out=brng:color=red"
12455 Playback video with signalstats metadata drawn over the frame.
12457 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12460 The contents of signalstat_drawtext.txt used in the command are:
12463 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12464 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12465 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12466 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12474 Blur the input video without impacting the outlines.
12476 It accepts the following options:
12479 @item luma_radius, lr
12480 Set the luma radius. The option value must be a float number in
12481 the range [0.1,5.0] that specifies the variance of the gaussian filter
12482 used to blur the image (slower if larger). Default value is 1.0.
12484 @item luma_strength, ls
12485 Set the luma strength. The option value must be a float number
12486 in the range [-1.0,1.0] that configures the blurring. A value included
12487 in [0.0,1.0] will blur the image whereas a value included in
12488 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12490 @item luma_threshold, lt
12491 Set the luma threshold used as a coefficient to determine
12492 whether a pixel should be blurred or not. The option value must be an
12493 integer in the range [-30,30]. A value of 0 will filter all the image,
12494 a value included in [0,30] will filter flat areas and a value included
12495 in [-30,0] will filter edges. Default value is 0.
12497 @item chroma_radius, cr
12498 Set the chroma radius. The option value must be a float number in
12499 the range [0.1,5.0] that specifies the variance of the gaussian filter
12500 used to blur the image (slower if larger). Default value is 1.0.
12502 @item chroma_strength, cs
12503 Set the chroma strength. The option value must be a float number
12504 in the range [-1.0,1.0] that configures the blurring. A value included
12505 in [0.0,1.0] will blur the image whereas a value included in
12506 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12508 @item chroma_threshold, ct
12509 Set the chroma threshold used as a coefficient to determine
12510 whether a pixel should be blurred or not. The option value must be an
12511 integer in the range [-30,30]. A value of 0 will filter all the image,
12512 a value included in [0,30] will filter flat areas and a value included
12513 in [-30,0] will filter edges. Default value is 0.
12516 If a chroma option is not explicitly set, the corresponding luma value
12521 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12523 This filter takes in input two input videos, the first input is
12524 considered the "main" source and is passed unchanged to the
12525 output. The second input is used as a "reference" video for computing
12528 Both video inputs must have the same resolution and pixel format for
12529 this filter to work correctly. Also it assumes that both inputs
12530 have the same number of frames, which are compared one by one.
12532 The filter stores the calculated SSIM of each frame.
12534 The description of the accepted parameters follows.
12537 @item stats_file, f
12538 If specified the filter will use the named file to save the SSIM of
12539 each individual frame. When filename equals "-" the data is sent to
12543 The file printed if @var{stats_file} is selected, contains a sequence of
12544 key/value pairs of the form @var{key}:@var{value} for each compared
12547 A description of each shown parameter follows:
12551 sequential number of the input frame, starting from 1
12553 @item Y, U, V, R, G, B
12554 SSIM of the compared frames for the component specified by the suffix.
12557 SSIM of the compared frames for the whole frame.
12560 Same as above but in dB representation.
12565 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12566 [main][ref] ssim="stats_file=stats.log" [out]
12569 On this example the input file being processed is compared with the
12570 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12571 is stored in @file{stats.log}.
12573 Another example with both psnr and ssim at same time:
12575 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12580 Convert between different stereoscopic image formats.
12582 The filters accept the following options:
12586 Set stereoscopic image format of input.
12588 Available values for input image formats are:
12591 side by side parallel (left eye left, right eye right)
12594 side by side crosseye (right eye left, left eye right)
12597 side by side parallel with half width resolution
12598 (left eye left, right eye right)
12601 side by side crosseye with half width resolution
12602 (right eye left, left eye right)
12605 above-below (left eye above, right eye below)
12608 above-below (right eye above, left eye below)
12611 above-below with half height resolution
12612 (left eye above, right eye below)
12615 above-below with half height resolution
12616 (right eye above, left eye below)
12619 alternating frames (left eye first, right eye second)
12622 alternating frames (right eye first, left eye second)
12625 interleaved rows (left eye has top row, right eye starts on next row)
12628 interleaved rows (right eye has top row, left eye starts on next row)
12631 interleaved columns, left eye first
12634 interleaved columns, right eye first
12636 Default value is @samp{sbsl}.
12640 Set stereoscopic image format of output.
12644 side by side parallel (left eye left, right eye right)
12647 side by side crosseye (right eye left, left eye right)
12650 side by side parallel with half width resolution
12651 (left eye left, right eye right)
12654 side by side crosseye with half width resolution
12655 (right eye left, left eye right)
12658 above-below (left eye above, right eye below)
12661 above-below (right eye above, left eye below)
12664 above-below with half height resolution
12665 (left eye above, right eye below)
12668 above-below with half height resolution
12669 (right eye above, left eye below)
12672 alternating frames (left eye first, right eye second)
12675 alternating frames (right eye first, left eye second)
12678 interleaved rows (left eye has top row, right eye starts on next row)
12681 interleaved rows (right eye has top row, left eye starts on next row)
12684 anaglyph red/blue gray
12685 (red filter on left eye, blue filter on right eye)
12688 anaglyph red/green gray
12689 (red filter on left eye, green filter on right eye)
12692 anaglyph red/cyan gray
12693 (red filter on left eye, cyan filter on right eye)
12696 anaglyph red/cyan half colored
12697 (red filter on left eye, cyan filter on right eye)
12700 anaglyph red/cyan color
12701 (red filter on left eye, cyan filter on right eye)
12704 anaglyph red/cyan color optimized with the least squares projection of dubois
12705 (red filter on left eye, cyan filter on right eye)
12708 anaglyph green/magenta gray
12709 (green filter on left eye, magenta filter on right eye)
12712 anaglyph green/magenta half colored
12713 (green filter on left eye, magenta filter on right eye)
12716 anaglyph green/magenta colored
12717 (green filter on left eye, magenta filter on right eye)
12720 anaglyph green/magenta color optimized with the least squares projection of dubois
12721 (green filter on left eye, magenta filter on right eye)
12724 anaglyph yellow/blue gray
12725 (yellow filter on left eye, blue filter on right eye)
12728 anaglyph yellow/blue half colored
12729 (yellow filter on left eye, blue filter on right eye)
12732 anaglyph yellow/blue colored
12733 (yellow filter on left eye, blue filter on right eye)
12736 anaglyph yellow/blue color optimized with the least squares projection of dubois
12737 (yellow filter on left eye, blue filter on right eye)
12740 mono output (left eye only)
12743 mono output (right eye only)
12746 checkerboard, left eye first
12749 checkerboard, right eye first
12752 interleaved columns, left eye first
12755 interleaved columns, right eye first
12761 Default value is @samp{arcd}.
12764 @subsection Examples
12768 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12774 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12780 @section streamselect, astreamselect
12781 Select video or audio streams.
12783 The filter accepts the following options:
12787 Set number of inputs. Default is 2.
12790 Set input indexes to remap to outputs.
12793 @subsection Commands
12795 The @code{streamselect} and @code{astreamselect} filter supports the following
12800 Set input indexes to remap to outputs.
12803 @subsection Examples
12807 Select first 5 seconds 1st stream and rest of time 2nd stream:
12809 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12813 Same as above, but for audio:
12815 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12820 Apply sobel operator to input video stream.
12822 The filter accepts the following option:
12826 Set which planes will be processed, unprocessed planes will be copied.
12827 By default value 0xf, all planes will be processed.
12830 Set value which will be multiplied with filtered result.
12833 Set value which will be added to filtered result.
12839 Apply a simple postprocessing filter that compresses and decompresses the image
12840 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12841 and average the results.
12843 The filter accepts the following options:
12847 Set quality. This option defines the number of levels for averaging. It accepts
12848 an integer in the range 0-6. If set to @code{0}, the filter will have no
12849 effect. A value of @code{6} means the higher quality. For each increment of
12850 that value the speed drops by a factor of approximately 2. Default value is
12854 Force a constant quantization parameter. If not set, the filter will use the QP
12855 from the video stream (if available).
12858 Set thresholding mode. Available modes are:
12862 Set hard thresholding (default).
12864 Set soft thresholding (better de-ringing effect, but likely blurrier).
12867 @item use_bframe_qp
12868 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12869 option may cause flicker since the B-Frames have often larger QP. Default is
12870 @code{0} (not enabled).
12876 Draw subtitles on top of input video using the libass library.
12878 To enable compilation of this filter you need to configure FFmpeg with
12879 @code{--enable-libass}. This filter also requires a build with libavcodec and
12880 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12881 Alpha) subtitles format.
12883 The filter accepts the following options:
12887 Set the filename of the subtitle file to read. It must be specified.
12889 @item original_size
12890 Specify the size of the original video, the video for which the ASS file
12891 was composed. For the syntax of this option, check the
12892 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12893 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12894 correctly scale the fonts if the aspect ratio has been changed.
12897 Set a directory path containing fonts that can be used by the filter.
12898 These fonts will be used in addition to whatever the font provider uses.
12901 Set subtitles input character encoding. @code{subtitles} filter only. Only
12902 useful if not UTF-8.
12904 @item stream_index, si
12905 Set subtitles stream index. @code{subtitles} filter only.
12908 Override default style or script info parameters of the subtitles. It accepts a
12909 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12912 If the first key is not specified, it is assumed that the first value
12913 specifies the @option{filename}.
12915 For example, to render the file @file{sub.srt} on top of the input
12916 video, use the command:
12921 which is equivalent to:
12923 subtitles=filename=sub.srt
12926 To render the default subtitles stream from file @file{video.mkv}, use:
12928 subtitles=video.mkv
12931 To render the second subtitles stream from that file, use:
12933 subtitles=video.mkv:si=1
12936 To make the subtitles stream from @file{sub.srt} appear in transparent green
12937 @code{DejaVu Serif}, use:
12939 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12942 @section super2xsai
12944 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12945 Interpolate) pixel art scaling algorithm.
12947 Useful for enlarging pixel art images without reducing sharpness.
12951 Swap two rectangular objects in video.
12953 This filter accepts the following options:
12963 Set 1st rect x coordinate.
12966 Set 1st rect y coordinate.
12969 Set 2nd rect x coordinate.
12972 Set 2nd rect y coordinate.
12974 All expressions are evaluated once for each frame.
12977 The all options are expressions containing the following constants:
12982 The input width and height.
12985 same as @var{w} / @var{h}
12988 input sample aspect ratio
12991 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12994 The number of the input frame, starting from 0.
12997 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
13000 the position in the file of the input frame, NAN if unknown
13008 Apply telecine process to the video.
13010 This filter accepts the following options:
13019 The default value is @code{top}.
13023 A string of numbers representing the pulldown pattern you wish to apply.
13024 The default value is @code{23}.
13028 Some typical patterns:
13033 24p: 2332 (preferred)
13040 24p: 222222222223 ("Euro pulldown")
13046 Select the most representative frame in a given sequence of consecutive frames.
13048 The filter accepts the following options:
13052 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
13053 will pick one of them, and then handle the next batch of @var{n} frames until
13054 the end. Default is @code{100}.
13057 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
13058 value will result in a higher memory usage, so a high value is not recommended.
13060 @subsection Examples
13064 Extract one picture each 50 frames:
13070 Complete example of a thumbnail creation with @command{ffmpeg}:
13072 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
13078 Tile several successive frames together.
13080 The filter accepts the following options:
13085 Set the grid size (i.e. the number of lines and columns). For the syntax of
13086 this option, check the
13087 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13090 Set the maximum number of frames to render in the given area. It must be less
13091 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
13092 the area will be used.
13095 Set the outer border margin in pixels.
13098 Set the inner border thickness (i.e. the number of pixels between frames). For
13099 more advanced padding options (such as having different values for the edges),
13100 refer to the pad video filter.
13103 Specify the color of the unused area. For the syntax of this option, check the
13104 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
13108 @subsection Examples
13112 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
13114 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
13116 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
13117 duplicating each output frame to accommodate the originally detected frame
13121 Display @code{5} pictures in an area of @code{3x2} frames,
13122 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
13123 mixed flat and named options:
13125 tile=3x2:nb_frames=5:padding=7:margin=2
13129 @section tinterlace
13131 Perform various types of temporal field interlacing.
13133 Frames are counted starting from 1, so the first input frame is
13136 The filter accepts the following options:
13141 Specify the mode of the interlacing. This option can also be specified
13142 as a value alone. See below for a list of values for this option.
13144 Available values are:
13148 Move odd frames into the upper field, even into the lower field,
13149 generating a double height frame at half frame rate.
13153 Frame 1 Frame 2 Frame 3 Frame 4
13155 11111 22222 33333 44444
13156 11111 22222 33333 44444
13157 11111 22222 33333 44444
13158 11111 22222 33333 44444
13172 Only output odd frames, even frames are dropped, generating a frame with
13173 unchanged height at half frame rate.
13178 Frame 1 Frame 2 Frame 3 Frame 4
13180 11111 22222 33333 44444
13181 11111 22222 33333 44444
13182 11111 22222 33333 44444
13183 11111 22222 33333 44444
13193 Only output even frames, odd frames are dropped, generating a frame with
13194 unchanged height at half frame rate.
13199 Frame 1 Frame 2 Frame 3 Frame 4
13201 11111 22222 33333 44444
13202 11111 22222 33333 44444
13203 11111 22222 33333 44444
13204 11111 22222 33333 44444
13214 Expand each frame to full height, but pad alternate lines with black,
13215 generating a frame with double height at the same input frame rate.
13220 Frame 1 Frame 2 Frame 3 Frame 4
13222 11111 22222 33333 44444
13223 11111 22222 33333 44444
13224 11111 22222 33333 44444
13225 11111 22222 33333 44444
13228 11111 ..... 33333 .....
13229 ..... 22222 ..... 44444
13230 11111 ..... 33333 .....
13231 ..... 22222 ..... 44444
13232 11111 ..... 33333 .....
13233 ..... 22222 ..... 44444
13234 11111 ..... 33333 .....
13235 ..... 22222 ..... 44444
13239 @item interleave_top, 4
13240 Interleave the upper field from odd frames with the lower field from
13241 even frames, generating a frame with unchanged height at half frame rate.
13246 Frame 1 Frame 2 Frame 3 Frame 4
13248 11111<- 22222 33333<- 44444
13249 11111 22222<- 33333 44444<-
13250 11111<- 22222 33333<- 44444
13251 11111 22222<- 33333 44444<-
13261 @item interleave_bottom, 5
13262 Interleave the lower field from odd frames with the upper field from
13263 even frames, generating a frame with unchanged height at half frame rate.
13268 Frame 1 Frame 2 Frame 3 Frame 4
13270 11111 22222<- 33333 44444<-
13271 11111<- 22222 33333<- 44444
13272 11111 22222<- 33333 44444<-
13273 11111<- 22222 33333<- 44444
13283 @item interlacex2, 6
13284 Double frame rate with unchanged height. Frames are inserted each
13285 containing the second temporal field from the previous input frame and
13286 the first temporal field from the next input frame. This mode relies on
13287 the top_field_first flag. Useful for interlaced video displays with no
13288 field synchronisation.
13293 Frame 1 Frame 2 Frame 3 Frame 4
13295 11111 22222 33333 44444
13296 11111 22222 33333 44444
13297 11111 22222 33333 44444
13298 11111 22222 33333 44444
13301 11111 22222 22222 33333 33333 44444 44444
13302 11111 11111 22222 22222 33333 33333 44444
13303 11111 22222 22222 33333 33333 44444 44444
13304 11111 11111 22222 22222 33333 33333 44444
13309 Move odd frames into the upper field, even into the lower field,
13310 generating a double height frame at same frame rate.
13315 Frame 1 Frame 2 Frame 3 Frame 4
13317 11111 22222 33333 44444
13318 11111 22222 33333 44444
13319 11111 22222 33333 44444
13320 11111 22222 33333 44444
13323 11111 33333 33333 55555
13324 22222 22222 44444 44444
13325 11111 33333 33333 55555
13326 22222 22222 44444 44444
13327 11111 33333 33333 55555
13328 22222 22222 44444 44444
13329 11111 33333 33333 55555
13330 22222 22222 44444 44444
13335 Numeric values are deprecated but are accepted for backward
13336 compatibility reasons.
13338 Default mode is @code{merge}.
13341 Specify flags influencing the filter process.
13343 Available value for @var{flags} is:
13346 @item low_pass_filter, vlfp
13347 Enable vertical low-pass filtering in the filter.
13348 Vertical low-pass filtering is required when creating an interlaced
13349 destination from a progressive source which contains high-frequency
13350 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13353 Vertical low-pass filtering can only be enabled for @option{mode}
13354 @var{interleave_top} and @var{interleave_bottom}.
13361 Transpose rows with columns in the input video and optionally flip it.
13363 It accepts the following parameters:
13368 Specify the transposition direction.
13370 Can assume the following values:
13372 @item 0, 4, cclock_flip
13373 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13381 Rotate by 90 degrees clockwise, that is:
13389 Rotate by 90 degrees counterclockwise, that is:
13396 @item 3, 7, clock_flip
13397 Rotate by 90 degrees clockwise and vertically flip, that is:
13405 For values between 4-7, the transposition is only done if the input
13406 video geometry is portrait and not landscape. These values are
13407 deprecated, the @code{passthrough} option should be used instead.
13409 Numerical values are deprecated, and should be dropped in favor of
13410 symbolic constants.
13413 Do not apply the transposition if the input geometry matches the one
13414 specified by the specified value. It accepts the following values:
13417 Always apply transposition.
13419 Preserve portrait geometry (when @var{height} >= @var{width}).
13421 Preserve landscape geometry (when @var{width} >= @var{height}).
13424 Default value is @code{none}.
13427 For example to rotate by 90 degrees clockwise and preserve portrait
13430 transpose=dir=1:passthrough=portrait
13433 The command above can also be specified as:
13435 transpose=1:portrait
13439 Trim the input so that the output contains one continuous subpart of the input.
13441 It accepts the following parameters:
13444 Specify the time of the start of the kept section, i.e. the frame with the
13445 timestamp @var{start} will be the first frame in the output.
13448 Specify the time of the first frame that will be dropped, i.e. the frame
13449 immediately preceding the one with the timestamp @var{end} will be the last
13450 frame in the output.
13453 This is the same as @var{start}, except this option sets the start timestamp
13454 in timebase units instead of seconds.
13457 This is the same as @var{end}, except this option sets the end timestamp
13458 in timebase units instead of seconds.
13461 The maximum duration of the output in seconds.
13464 The number of the first frame that should be passed to the output.
13467 The number of the first frame that should be dropped.
13470 @option{start}, @option{end}, and @option{duration} are expressed as time
13471 duration specifications; see
13472 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13473 for the accepted syntax.
13475 Note that the first two sets of the start/end options and the @option{duration}
13476 option look at the frame timestamp, while the _frame variants simply count the
13477 frames that pass through the filter. Also note that this filter does not modify
13478 the timestamps. If you wish for the output timestamps to start at zero, insert a
13479 setpts filter after the trim filter.
13481 If multiple start or end options are set, this filter tries to be greedy and
13482 keep all the frames that match at least one of the specified constraints. To keep
13483 only the part that matches all the constraints at once, chain multiple trim
13486 The defaults are such that all the input is kept. So it is possible to set e.g.
13487 just the end values to keep everything before the specified time.
13492 Drop everything except the second minute of input:
13494 ffmpeg -i INPUT -vf trim=60:120
13498 Keep only the first second:
13500 ffmpeg -i INPUT -vf trim=duration=1
13509 Sharpen or blur the input video.
13511 It accepts the following parameters:
13514 @item luma_msize_x, lx
13515 Set the luma matrix horizontal size. It must be an odd integer between
13516 3 and 23. The default value is 5.
13518 @item luma_msize_y, ly
13519 Set the luma matrix vertical size. It must be an odd integer between 3
13520 and 23. The default value is 5.
13522 @item luma_amount, la
13523 Set the luma effect strength. It must be a floating point number, reasonable
13524 values lay between -1.5 and 1.5.
13526 Negative values will blur the input video, while positive values will
13527 sharpen it, a value of zero will disable the effect.
13529 Default value is 1.0.
13531 @item chroma_msize_x, cx
13532 Set the chroma matrix horizontal size. It must be an odd integer
13533 between 3 and 23. The default value is 5.
13535 @item chroma_msize_y, cy
13536 Set the chroma matrix vertical size. It must be an odd integer
13537 between 3 and 23. The default value is 5.
13539 @item chroma_amount, ca
13540 Set the chroma effect strength. It must be a floating point number, reasonable
13541 values lay between -1.5 and 1.5.
13543 Negative values will blur the input video, while positive values will
13544 sharpen it, a value of zero will disable the effect.
13546 Default value is 0.0.
13549 If set to 1, specify using OpenCL capabilities, only available if
13550 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13554 All parameters are optional and default to the equivalent of the
13555 string '5:5:1.0:5:5:0.0'.
13557 @subsection Examples
13561 Apply strong luma sharpen effect:
13563 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13567 Apply a strong blur of both luma and chroma parameters:
13569 unsharp=7:7:-2:7:7:-2
13575 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13576 the image at several (or - in the case of @option{quality} level @code{8} - all)
13577 shifts and average the results.
13579 The way this differs from the behavior of spp is that uspp actually encodes &
13580 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13581 DCT similar to MJPEG.
13583 The filter accepts the following options:
13587 Set quality. This option defines the number of levels for averaging. It accepts
13588 an integer in the range 0-8. If set to @code{0}, the filter will have no
13589 effect. A value of @code{8} means the higher quality. For each increment of
13590 that value the speed drops by a factor of approximately 2. Default value is
13594 Force a constant quantization parameter. If not set, the filter will use the QP
13595 from the video stream (if available).
13598 @section vaguedenoiser
13600 Apply a wavelet based denoiser.
13602 It transforms each frame from the video input into the wavelet domain,
13603 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
13604 the obtained coefficients. It does an inverse wavelet transform after.
13605 Due to wavelet properties, it should give a nice smoothed result, and
13606 reduced noise, without blurring picture features.
13608 This filter accepts the following options:
13612 The filtering strength. The higher, the more filtered the video will be.
13613 Hard thresholding can use a higher threshold than soft thresholding
13614 before the video looks overfiltered.
13617 The filtering method the filter will use.
13619 It accepts the following values:
13622 All values under the threshold will be zeroed.
13625 All values under the threshold will be zeroed. All values above will be
13626 reduced by the threshold.
13629 Scales or nullifies coefficients - intermediary between (more) soft and
13630 (less) hard thresholding.
13634 Number of times, the wavelet will decompose the picture. Picture can't
13635 be decomposed beyond a particular point (typically, 8 for a 640x480
13636 frame - as 2^9 = 512 > 480)
13639 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
13642 A list of the planes to process. By default all planes are processed.
13645 @section vectorscope
13647 Display 2 color component values in the two dimensional graph (which is called
13650 This filter accepts the following options:
13654 Set vectorscope mode.
13656 It accepts the following values:
13659 Gray values are displayed on graph, higher brightness means more pixels have
13660 same component color value on location in graph. This is the default mode.
13663 Gray values are displayed on graph. Surrounding pixels values which are not
13664 present in video frame are drawn in gradient of 2 color components which are
13665 set by option @code{x} and @code{y}. The 3rd color component is static.
13668 Actual color components values present in video frame are displayed on graph.
13671 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13672 on graph increases value of another color component, which is luminance by
13673 default values of @code{x} and @code{y}.
13676 Actual colors present in video frame are displayed on graph. If two different
13677 colors map to same position on graph then color with higher value of component
13678 not present in graph is picked.
13681 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13682 component picked from radial gradient.
13686 Set which color component will be represented on X-axis. Default is @code{1}.
13689 Set which color component will be represented on Y-axis. Default is @code{2}.
13692 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13693 of color component which represents frequency of (X, Y) location in graph.
13698 No envelope, this is default.
13701 Instant envelope, even darkest single pixel will be clearly highlighted.
13704 Hold maximum and minimum values presented in graph over time. This way you
13705 can still spot out of range values without constantly looking at vectorscope.
13708 Peak and instant envelope combined together.
13712 Set what kind of graticule to draw.
13720 Set graticule opacity.
13723 Set graticule flags.
13727 Draw graticule for white point.
13730 Draw graticule for black point.
13733 Draw color points short names.
13737 Set background opacity.
13739 @item lthreshold, l
13740 Set low threshold for color component not represented on X or Y axis.
13741 Values lower than this value will be ignored. Default is 0.
13742 Note this value is multiplied with actual max possible value one pixel component
13743 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13746 @item hthreshold, h
13747 Set high threshold for color component not represented on X or Y axis.
13748 Values higher than this value will be ignored. Default is 1.
13749 Note this value is multiplied with actual max possible value one pixel component
13750 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13751 is 0.9 * 255 = 230.
13753 @item colorspace, c
13754 Set what kind of colorspace to use when drawing graticule.
13763 @anchor{vidstabdetect}
13764 @section vidstabdetect
13766 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13767 @ref{vidstabtransform} for pass 2.
13769 This filter generates a file with relative translation and rotation
13770 transform information about subsequent frames, which is then used by
13771 the @ref{vidstabtransform} filter.
13773 To enable compilation of this filter you need to configure FFmpeg with
13774 @code{--enable-libvidstab}.
13776 This filter accepts the following options:
13780 Set the path to the file used to write the transforms information.
13781 Default value is @file{transforms.trf}.
13784 Set how shaky the video is and how quick the camera is. It accepts an
13785 integer in the range 1-10, a value of 1 means little shakiness, a
13786 value of 10 means strong shakiness. Default value is 5.
13789 Set the accuracy of the detection process. It must be a value in the
13790 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13791 accuracy. Default value is 15.
13794 Set stepsize of the search process. The region around minimum is
13795 scanned with 1 pixel resolution. Default value is 6.
13798 Set minimum contrast. Below this value a local measurement field is
13799 discarded. Must be a floating point value in the range 0-1. Default
13803 Set reference frame number for tripod mode.
13805 If enabled, the motion of the frames is compared to a reference frame
13806 in the filtered stream, identified by the specified number. The idea
13807 is to compensate all movements in a more-or-less static scene and keep
13808 the camera view absolutely still.
13810 If set to 0, it is disabled. The frames are counted starting from 1.
13813 Show fields and transforms in the resulting frames. It accepts an
13814 integer in the range 0-2. Default value is 0, which disables any
13818 @subsection Examples
13822 Use default values:
13828 Analyze strongly shaky movie and put the results in file
13829 @file{mytransforms.trf}:
13831 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13835 Visualize the result of internal transformations in the resulting
13838 vidstabdetect=show=1
13842 Analyze a video with medium shakiness using @command{ffmpeg}:
13844 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13848 @anchor{vidstabtransform}
13849 @section vidstabtransform
13851 Video stabilization/deshaking: pass 2 of 2,
13852 see @ref{vidstabdetect} for pass 1.
13854 Read a file with transform information for each frame and
13855 apply/compensate them. Together with the @ref{vidstabdetect}
13856 filter this can be used to deshake videos. See also
13857 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13858 the @ref{unsharp} filter, see below.
13860 To enable compilation of this filter you need to configure FFmpeg with
13861 @code{--enable-libvidstab}.
13863 @subsection Options
13867 Set path to the file used to read the transforms. Default value is
13868 @file{transforms.trf}.
13871 Set the number of frames (value*2 + 1) used for lowpass filtering the
13872 camera movements. Default value is 10.
13874 For example a number of 10 means that 21 frames are used (10 in the
13875 past and 10 in the future) to smoothen the motion in the video. A
13876 larger value leads to a smoother video, but limits the acceleration of
13877 the camera (pan/tilt movements). 0 is a special case where a static
13878 camera is simulated.
13881 Set the camera path optimization algorithm.
13883 Accepted values are:
13886 gaussian kernel low-pass filter on camera motion (default)
13888 averaging on transformations
13892 Set maximal number of pixels to translate frames. Default value is -1,
13896 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13897 value is -1, meaning no limit.
13900 Specify how to deal with borders that may be visible due to movement
13903 Available values are:
13906 keep image information from previous frame (default)
13908 fill the border black
13912 Invert transforms if set to 1. Default value is 0.
13915 Consider transforms as relative to previous frame if set to 1,
13916 absolute if set to 0. Default value is 0.
13919 Set percentage to zoom. A positive value will result in a zoom-in
13920 effect, a negative value in a zoom-out effect. Default value is 0 (no
13924 Set optimal zooming to avoid borders.
13926 Accepted values are:
13931 optimal static zoom value is determined (only very strong movements
13932 will lead to visible borders) (default)
13934 optimal adaptive zoom value is determined (no borders will be
13935 visible), see @option{zoomspeed}
13938 Note that the value given at zoom is added to the one calculated here.
13941 Set percent to zoom maximally each frame (enabled when
13942 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13946 Specify type of interpolation.
13948 Available values are:
13953 linear only horizontal
13955 linear in both directions (default)
13957 cubic in both directions (slow)
13961 Enable virtual tripod mode if set to 1, which is equivalent to
13962 @code{relative=0:smoothing=0}. Default value is 0.
13964 Use also @code{tripod} option of @ref{vidstabdetect}.
13967 Increase log verbosity if set to 1. Also the detected global motions
13968 are written to the temporary file @file{global_motions.trf}. Default
13972 @subsection Examples
13976 Use @command{ffmpeg} for a typical stabilization with default values:
13978 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13981 Note the use of the @ref{unsharp} filter which is always recommended.
13984 Zoom in a bit more and load transform data from a given file:
13986 vidstabtransform=zoom=5:input="mytransforms.trf"
13990 Smoothen the video even more:
13992 vidstabtransform=smoothing=30
13998 Flip the input video vertically.
14000 For example, to vertically flip a video with @command{ffmpeg}:
14002 ffmpeg -i in.avi -vf "vflip" out.avi
14008 Make or reverse a natural vignetting effect.
14010 The filter accepts the following options:
14014 Set lens angle expression as a number of radians.
14016 The value is clipped in the @code{[0,PI/2]} range.
14018 Default value: @code{"PI/5"}
14022 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
14026 Set forward/backward mode.
14028 Available modes are:
14031 The larger the distance from the central point, the darker the image becomes.
14034 The larger the distance from the central point, the brighter the image becomes.
14035 This can be used to reverse a vignette effect, though there is no automatic
14036 detection to extract the lens @option{angle} and other settings (yet). It can
14037 also be used to create a burning effect.
14040 Default value is @samp{forward}.
14043 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
14045 It accepts the following values:
14048 Evaluate expressions only once during the filter initialization.
14051 Evaluate expressions for each incoming frame. This is way slower than the
14052 @samp{init} mode since it requires all the scalers to be re-computed, but it
14053 allows advanced dynamic expressions.
14056 Default value is @samp{init}.
14059 Set dithering to reduce the circular banding effects. Default is @code{1}
14063 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
14064 Setting this value to the SAR of the input will make a rectangular vignetting
14065 following the dimensions of the video.
14067 Default is @code{1/1}.
14070 @subsection Expressions
14072 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
14073 following parameters.
14078 input width and height
14081 the number of input frame, starting from 0
14084 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
14085 @var{TB} units, NAN if undefined
14088 frame rate of the input video, NAN if the input frame rate is unknown
14091 the PTS (Presentation TimeStamp) of the filtered video frame,
14092 expressed in seconds, NAN if undefined
14095 time base of the input video
14099 @subsection Examples
14103 Apply simple strong vignetting effect:
14109 Make a flickering vignetting:
14111 vignette='PI/4+random(1)*PI/50':eval=frame
14117 Stack input videos vertically.
14119 All streams must be of same pixel format and of same width.
14121 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
14122 to create same output.
14124 The filter accept the following option:
14128 Set number of input streams. Default is 2.
14131 If set to 1, force the output to terminate when the shortest input
14132 terminates. Default value is 0.
14137 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
14138 Deinterlacing Filter").
14140 Based on the process described by Martin Weston for BBC R&D, and
14141 implemented based on the de-interlace algorithm written by Jim
14142 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
14143 uses filter coefficients calculated by BBC R&D.
14145 There are two sets of filter coefficients, so called "simple":
14146 and "complex". Which set of filter coefficients is used can
14147 be set by passing an optional parameter:
14151 Set the interlacing filter coefficients. Accepts one of the following values:
14155 Simple filter coefficient set.
14157 More-complex filter coefficient set.
14159 Default value is @samp{complex}.
14162 Specify which frames to deinterlace. Accept one of the following values:
14166 Deinterlace all frames,
14168 Only deinterlace frames marked as interlaced.
14171 Default value is @samp{all}.
14175 Video waveform monitor.
14177 The waveform monitor plots color component intensity. By default luminance
14178 only. Each column of the waveform corresponds to a column of pixels in the
14181 It accepts the following options:
14185 Can be either @code{row}, or @code{column}. Default is @code{column}.
14186 In row mode, the graph on the left side represents color component value 0 and
14187 the right side represents value = 255. In column mode, the top side represents
14188 color component value = 0 and bottom side represents value = 255.
14191 Set intensity. Smaller values are useful to find out how many values of the same
14192 luminance are distributed across input rows/columns.
14193 Default value is @code{0.04}. Allowed range is [0, 1].
14196 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
14197 In mirrored mode, higher values will be represented on the left
14198 side for @code{row} mode and at the top for @code{column} mode. Default is
14199 @code{1} (mirrored).
14203 It accepts the following values:
14206 Presents information identical to that in the @code{parade}, except
14207 that the graphs representing color components are superimposed directly
14210 This display mode makes it easier to spot relative differences or similarities
14211 in overlapping areas of the color components that are supposed to be identical,
14212 such as neutral whites, grays, or blacks.
14215 Display separate graph for the color components side by side in
14216 @code{row} mode or one below the other in @code{column} mode.
14219 Display separate graph for the color components side by side in
14220 @code{column} mode or one below the other in @code{row} mode.
14222 Using this display mode makes it easy to spot color casts in the highlights
14223 and shadows of an image, by comparing the contours of the top and the bottom
14224 graphs of each waveform. Since whites, grays, and blacks are characterized
14225 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14226 should display three waveforms of roughly equal width/height. If not, the
14227 correction is easy to perform by making level adjustments the three waveforms.
14229 Default is @code{stack}.
14231 @item components, c
14232 Set which color components to display. Default is 1, which means only luminance
14233 or red color component if input is in RGB colorspace. If is set for example to
14234 7 it will display all 3 (if) available color components.
14239 No envelope, this is default.
14242 Instant envelope, minimum and maximum values presented in graph will be easily
14243 visible even with small @code{step} value.
14246 Hold minimum and maximum values presented in graph across time. This way you
14247 can still spot out of range values without constantly looking at waveforms.
14250 Peak and instant envelope combined together.
14256 No filtering, this is default.
14259 Luma and chroma combined together.
14262 Similar as above, but shows difference between blue and red chroma.
14265 Displays only chroma.
14268 Displays actual color value on waveform.
14271 Similar as above, but with luma showing frequency of chroma values.
14275 Set which graticule to display.
14279 Do not display graticule.
14282 Display green graticule showing legal broadcast ranges.
14286 Set graticule opacity.
14289 Set graticule flags.
14293 Draw numbers above lines. By default enabled.
14296 Draw dots instead of lines.
14300 Set scale used for displaying graticule.
14307 Default is digital.
14310 Set background opacity.
14315 The @code{weave} takes a field-based video input and join
14316 each two sequential fields into single frame, producing a new double
14317 height clip with half the frame rate and half the frame count.
14319 It accepts the following option:
14323 Set first field. Available values are:
14327 Set the frame as top-field-first.
14330 Set the frame as bottom-field-first.
14334 @subsection Examples
14338 Interlace video using @ref{select} and @ref{separatefields} filter:
14340 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
14345 Apply the xBR high-quality magnification filter which is designed for pixel
14346 art. It follows a set of edge-detection rules, see
14347 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14349 It accepts the following option:
14353 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14354 @code{3xBR} and @code{4} for @code{4xBR}.
14355 Default is @code{3}.
14361 Deinterlace the input video ("yadif" means "yet another deinterlacing
14364 It accepts the following parameters:
14370 The interlacing mode to adopt. It accepts one of the following values:
14373 @item 0, send_frame
14374 Output one frame for each frame.
14375 @item 1, send_field
14376 Output one frame for each field.
14377 @item 2, send_frame_nospatial
14378 Like @code{send_frame}, but it skips the spatial interlacing check.
14379 @item 3, send_field_nospatial
14380 Like @code{send_field}, but it skips the spatial interlacing check.
14383 The default value is @code{send_frame}.
14386 The picture field parity assumed for the input interlaced video. It accepts one
14387 of the following values:
14391 Assume the top field is first.
14393 Assume the bottom field is first.
14395 Enable automatic detection of field parity.
14398 The default value is @code{auto}.
14399 If the interlacing is unknown or the decoder does not export this information,
14400 top field first will be assumed.
14403 Specify which frames to deinterlace. Accept one of the following
14408 Deinterlace all frames.
14409 @item 1, interlaced
14410 Only deinterlace frames marked as interlaced.
14413 The default value is @code{all}.
14418 Apply Zoom & Pan effect.
14420 This filter accepts the following options:
14424 Set the zoom expression. Default is 1.
14428 Set the x and y expression. Default is 0.
14431 Set the duration expression in number of frames.
14432 This sets for how many number of frames effect will last for
14433 single input image.
14436 Set the output image size, default is 'hd720'.
14439 Set the output frame rate, default is '25'.
14442 Each expression can contain the following constants:
14461 Output frame count.
14465 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14466 for current input frame.
14470 'x' and 'y' of last output frame of previous input frame or 0 when there was
14471 not yet such frame (first input frame).
14474 Last calculated zoom from 'z' expression for current input frame.
14477 Last calculated zoom of last output frame of previous input frame.
14480 Number of output frames for current input frame. Calculated from 'd' expression
14481 for each input frame.
14484 number of output frames created for previous input frame
14487 Rational number: input width / input height
14490 sample aspect ratio
14493 display aspect ratio
14497 @subsection Examples
14501 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14503 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
14507 Zoom-in up to 1.5 and pan always at center of picture:
14509 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14513 Same as above but without pausing:
14515 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14520 Scale (resize) the input video, using the z.lib library:
14521 https://github.com/sekrit-twc/zimg.
14523 The zscale filter forces the output display aspect ratio to be the same
14524 as the input, by changing the output sample aspect ratio.
14526 If the input image format is different from the format requested by
14527 the next filter, the zscale filter will convert the input to the
14530 @subsection Options
14531 The filter accepts the following options.
14536 Set the output video dimension expression. Default value is the input
14539 If the @var{width} or @var{w} is 0, the input width is used for the output.
14540 If the @var{height} or @var{h} is 0, the input height is used for the output.
14542 If one of the values is -1, the zscale filter will use a value that
14543 maintains the aspect ratio of the input image, calculated from the
14544 other specified dimension. If both of them are -1, the input size is
14547 If one of the values is -n with n > 1, the zscale filter will also use a value
14548 that maintains the aspect ratio of the input image, calculated from the other
14549 specified dimension. After that it will, however, make sure that the calculated
14550 dimension is divisible by n and adjust the value if necessary.
14552 See below for the list of accepted constants for use in the dimension
14556 Set the video size. For the syntax of this option, check the
14557 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14560 Set the dither type.
14562 Possible values are:
14567 @item error_diffusion
14573 Set the resize filter type.
14575 Possible values are:
14585 Default is bilinear.
14588 Set the color range.
14590 Possible values are:
14597 Default is same as input.
14600 Set the color primaries.
14602 Possible values are:
14612 Default is same as input.
14615 Set the transfer characteristics.
14617 Possible values are:
14628 Default is same as input.
14631 Set the colorspace matrix.
14633 Possible value are:
14644 Default is same as input.
14647 Set the input color range.
14649 Possible values are:
14656 Default is same as input.
14658 @item primariesin, pin
14659 Set the input color primaries.
14661 Possible values are:
14671 Default is same as input.
14673 @item transferin, tin
14674 Set the input transfer characteristics.
14676 Possible values are:
14687 Default is same as input.
14689 @item matrixin, min
14690 Set the input colorspace matrix.
14692 Possible value are:
14704 Set the output chroma location.
14706 Possible values are:
14717 @item chromalin, cin
14718 Set the input chroma location.
14720 Possible values are:
14732 The values of the @option{w} and @option{h} options are expressions
14733 containing the following constants:
14738 The input width and height
14742 These are the same as @var{in_w} and @var{in_h}.
14746 The output (scaled) width and height
14750 These are the same as @var{out_w} and @var{out_h}
14753 The same as @var{iw} / @var{ih}
14756 input sample aspect ratio
14759 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14763 horizontal and vertical input chroma subsample values. For example for the
14764 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14768 horizontal and vertical output chroma subsample values. For example for the
14769 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14775 @c man end VIDEO FILTERS
14777 @chapter Video Sources
14778 @c man begin VIDEO SOURCES
14780 Below is a description of the currently available video sources.
14784 Buffer video frames, and make them available to the filter chain.
14786 This source is mainly intended for a programmatic use, in particular
14787 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14789 It accepts the following parameters:
14794 Specify the size (width and height) of the buffered video frames. For the
14795 syntax of this option, check the
14796 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14799 The input video width.
14802 The input video height.
14805 A string representing the pixel format of the buffered video frames.
14806 It may be a number corresponding to a pixel format, or a pixel format
14810 Specify the timebase assumed by the timestamps of the buffered frames.
14813 Specify the frame rate expected for the video stream.
14815 @item pixel_aspect, sar
14816 The sample (pixel) aspect ratio of the input video.
14819 Specify the optional parameters to be used for the scale filter which
14820 is automatically inserted when an input change is detected in the
14821 input size or format.
14823 @item hw_frames_ctx
14824 When using a hardware pixel format, this should be a reference to an
14825 AVHWFramesContext describing input frames.
14830 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14833 will instruct the source to accept video frames with size 320x240 and
14834 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14835 square pixels (1:1 sample aspect ratio).
14836 Since the pixel format with name "yuv410p" corresponds to the number 6
14837 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14838 this example corresponds to:
14840 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14843 Alternatively, the options can be specified as a flat string, but this
14844 syntax is deprecated:
14846 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}[:@var{sws_param}]
14850 Create a pattern generated by an elementary cellular automaton.
14852 The initial state of the cellular automaton can be defined through the
14853 @option{filename} and @option{pattern} options. If such options are
14854 not specified an initial state is created randomly.
14856 At each new frame a new row in the video is filled with the result of
14857 the cellular automaton next generation. The behavior when the whole
14858 frame is filled is defined by the @option{scroll} option.
14860 This source accepts the following options:
14864 Read the initial cellular automaton state, i.e. the starting row, from
14865 the specified file.
14866 In the file, each non-whitespace character is considered an alive
14867 cell, a newline will terminate the row, and further characters in the
14868 file will be ignored.
14871 Read the initial cellular automaton state, i.e. the starting row, from
14872 the specified string.
14874 Each non-whitespace character in the string is considered an alive
14875 cell, a newline will terminate the row, and further characters in the
14876 string will be ignored.
14879 Set the video rate, that is the number of frames generated per second.
14882 @item random_fill_ratio, ratio
14883 Set the random fill ratio for the initial cellular automaton row. It
14884 is a floating point number value ranging from 0 to 1, defaults to
14887 This option is ignored when a file or a pattern is specified.
14889 @item random_seed, seed
14890 Set the seed for filling randomly the initial row, must be an integer
14891 included between 0 and UINT32_MAX. If not specified, or if explicitly
14892 set to -1, the filter will try to use a good random seed on a best
14896 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14897 Default value is 110.
14900 Set the size of the output video. For the syntax of this option, check the
14901 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14903 If @option{filename} or @option{pattern} is specified, the size is set
14904 by default to the width of the specified initial state row, and the
14905 height is set to @var{width} * PHI.
14907 If @option{size} is set, it must contain the width of the specified
14908 pattern string, and the specified pattern will be centered in the
14911 If a filename or a pattern string is not specified, the size value
14912 defaults to "320x518" (used for a randomly generated initial state).
14915 If set to 1, scroll the output upward when all the rows in the output
14916 have been already filled. If set to 0, the new generated row will be
14917 written over the top row just after the bottom row is filled.
14920 @item start_full, full
14921 If set to 1, completely fill the output with generated rows before
14922 outputting the first frame.
14923 This is the default behavior, for disabling set the value to 0.
14926 If set to 1, stitch the left and right row edges together.
14927 This is the default behavior, for disabling set the value to 0.
14930 @subsection Examples
14934 Read the initial state from @file{pattern}, and specify an output of
14937 cellauto=f=pattern:s=200x400
14941 Generate a random initial row with a width of 200 cells, with a fill
14944 cellauto=ratio=2/3:s=200x200
14948 Create a pattern generated by rule 18 starting by a single alive cell
14949 centered on an initial row with width 100:
14951 cellauto=p=@@:s=100x400:full=0:rule=18
14955 Specify a more elaborated initial pattern:
14957 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14962 @anchor{coreimagesrc}
14963 @section coreimagesrc
14964 Video source generated on GPU using Apple's CoreImage API on OSX.
14966 This video source is a specialized version of the @ref{coreimage} video filter.
14967 Use a core image generator at the beginning of the applied filterchain to
14968 generate the content.
14970 The coreimagesrc video source accepts the following options:
14972 @item list_generators
14973 List all available generators along with all their respective options as well as
14974 possible minimum and maximum values along with the default values.
14976 list_generators=true
14980 Specify the size of the sourced video. For the syntax of this option, check the
14981 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14982 The default value is @code{320x240}.
14985 Specify the frame rate of the sourced video, as the number of frames
14986 generated per second. It has to be a string in the format
14987 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14988 number or a valid video frame rate abbreviation. The default value is
14992 Set the sample aspect ratio of the sourced video.
14995 Set the duration of the sourced video. See
14996 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14997 for the accepted syntax.
14999 If not specified, or the expressed duration is negative, the video is
15000 supposed to be generated forever.
15003 Additionally, all options of the @ref{coreimage} video filter are accepted.
15004 A complete filterchain can be used for further processing of the
15005 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
15006 and examples for details.
15008 @subsection Examples
15013 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
15014 given as complete and escaped command-line for Apple's standard bash shell:
15016 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
15018 This example is equivalent to the QRCode example of @ref{coreimage} without the
15019 need for a nullsrc video source.
15023 @section mandelbrot
15025 Generate a Mandelbrot set fractal, and progressively zoom towards the
15026 point specified with @var{start_x} and @var{start_y}.
15028 This source accepts the following options:
15033 Set the terminal pts value. Default value is 400.
15036 Set the terminal scale value.
15037 Must be a floating point value. Default value is 0.3.
15040 Set the inner coloring mode, that is the algorithm used to draw the
15041 Mandelbrot fractal internal region.
15043 It shall assume one of the following values:
15048 Show time until convergence.
15050 Set color based on point closest to the origin of the iterations.
15055 Default value is @var{mincol}.
15058 Set the bailout value. Default value is 10.0.
15061 Set the maximum of iterations performed by the rendering
15062 algorithm. Default value is 7189.
15065 Set outer coloring mode.
15066 It shall assume one of following values:
15068 @item iteration_count
15069 Set iteration cound mode.
15070 @item normalized_iteration_count
15071 set normalized iteration count mode.
15073 Default value is @var{normalized_iteration_count}.
15076 Set frame rate, expressed as number of frames per second. Default
15080 Set frame size. For the syntax of this option, check the "Video
15081 size" section in the ffmpeg-utils manual. Default value is "640x480".
15084 Set the initial scale value. Default value is 3.0.
15087 Set the initial x position. Must be a floating point value between
15088 -100 and 100. Default value is -0.743643887037158704752191506114774.
15091 Set the initial y position. Must be a floating point value between
15092 -100 and 100. Default value is -0.131825904205311970493132056385139.
15097 Generate various test patterns, as generated by the MPlayer test filter.
15099 The size of the generated video is fixed, and is 256x256.
15100 This source is useful in particular for testing encoding features.
15102 This source accepts the following options:
15107 Specify the frame rate of the sourced video, as the number of frames
15108 generated per second. It has to be a string in the format
15109 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15110 number or a valid video frame rate abbreviation. The default value is
15114 Set the duration of the sourced video. See
15115 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15116 for the accepted syntax.
15118 If not specified, or the expressed duration is negative, the video is
15119 supposed to be generated forever.
15123 Set the number or the name of the test to perform. Supported tests are:
15139 Default value is "all", which will cycle through the list of all tests.
15144 mptestsrc=t=dc_luma
15147 will generate a "dc_luma" test pattern.
15149 @section frei0r_src
15151 Provide a frei0r source.
15153 To enable compilation of this filter you need to install the frei0r
15154 header and configure FFmpeg with @code{--enable-frei0r}.
15156 This source accepts the following parameters:
15161 The size of the video to generate. For the syntax of this option, check the
15162 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15165 The framerate of the generated video. It may be a string of the form
15166 @var{num}/@var{den} or a frame rate abbreviation.
15169 The name to the frei0r source to load. For more information regarding frei0r and
15170 how to set the parameters, read the @ref{frei0r} section in the video filters
15173 @item filter_params
15174 A '|'-separated list of parameters to pass to the frei0r source.
15178 For example, to generate a frei0r partik0l source with size 200x200
15179 and frame rate 10 which is overlaid on the overlay filter main input:
15181 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
15186 Generate a life pattern.
15188 This source is based on a generalization of John Conway's life game.
15190 The sourced input represents a life grid, each pixel represents a cell
15191 which can be in one of two possible states, alive or dead. Every cell
15192 interacts with its eight neighbours, which are the cells that are
15193 horizontally, vertically, or diagonally adjacent.
15195 At each interaction the grid evolves according to the adopted rule,
15196 which specifies the number of neighbor alive cells which will make a
15197 cell stay alive or born. The @option{rule} option allows one to specify
15200 This source accepts the following options:
15204 Set the file from which to read the initial grid state. In the file,
15205 each non-whitespace character is considered an alive cell, and newline
15206 is used to delimit the end of each row.
15208 If this option is not specified, the initial grid is generated
15212 Set the video rate, that is the number of frames generated per second.
15215 @item random_fill_ratio, ratio
15216 Set the random fill ratio for the initial random grid. It is a
15217 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15218 It is ignored when a file is specified.
15220 @item random_seed, seed
15221 Set the seed for filling the initial random grid, must be an integer
15222 included between 0 and UINT32_MAX. If not specified, or if explicitly
15223 set to -1, the filter will try to use a good random seed on a best
15229 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15230 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15231 @var{NS} specifies the number of alive neighbor cells which make a
15232 live cell stay alive, and @var{NB} the number of alive neighbor cells
15233 which make a dead cell to become alive (i.e. to "born").
15234 "s" and "b" can be used in place of "S" and "B", respectively.
15236 Alternatively a rule can be specified by an 18-bits integer. The 9
15237 high order bits are used to encode the next cell state if it is alive
15238 for each number of neighbor alive cells, the low order bits specify
15239 the rule for "borning" new cells. Higher order bits encode for an
15240 higher number of neighbor cells.
15241 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15242 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15244 Default value is "S23/B3", which is the original Conway's game of life
15245 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15246 cells, and will born a new cell if there are three alive cells around
15250 Set the size of the output video. For the syntax of this option, check the
15251 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15253 If @option{filename} is specified, the size is set by default to the
15254 same size of the input file. If @option{size} is set, it must contain
15255 the size specified in the input file, and the initial grid defined in
15256 that file is centered in the larger resulting area.
15258 If a filename is not specified, the size value defaults to "320x240"
15259 (used for a randomly generated initial grid).
15262 If set to 1, stitch the left and right grid edges together, and the
15263 top and bottom edges also. Defaults to 1.
15266 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15267 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15268 value from 0 to 255.
15271 Set the color of living (or new born) cells.
15274 Set the color of dead cells. If @option{mold} is set, this is the first color
15275 used to represent a dead cell.
15278 Set mold color, for definitely dead and moldy cells.
15280 For the syntax of these 3 color options, check the "Color" section in the
15281 ffmpeg-utils manual.
15284 @subsection Examples
15288 Read a grid from @file{pattern}, and center it on a grid of size
15291 life=f=pattern:s=300x300
15295 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15297 life=ratio=2/3:s=200x200
15301 Specify a custom rule for evolving a randomly generated grid:
15307 Full example with slow death effect (mold) using @command{ffplay}:
15309 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15316 @anchor{haldclutsrc}
15318 @anchor{rgbtestsrc}
15320 @anchor{smptehdbars}
15323 @anchor{yuvtestsrc}
15324 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15326 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15328 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15330 The @code{color} source provides an uniformly colored input.
15332 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15333 @ref{haldclut} filter.
15335 The @code{nullsrc} source returns unprocessed video frames. It is
15336 mainly useful to be employed in analysis / debugging tools, or as the
15337 source for filters which ignore the input data.
15339 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15340 detecting RGB vs BGR issues. You should see a red, green and blue
15341 stripe from top to bottom.
15343 The @code{smptebars} source generates a color bars pattern, based on
15344 the SMPTE Engineering Guideline EG 1-1990.
15346 The @code{smptehdbars} source generates a color bars pattern, based on
15347 the SMPTE RP 219-2002.
15349 The @code{testsrc} source generates a test video pattern, showing a
15350 color pattern, a scrolling gradient and a timestamp. This is mainly
15351 intended for testing purposes.
15353 The @code{testsrc2} source is similar to testsrc, but supports more
15354 pixel formats instead of just @code{rgb24}. This allows using it as an
15355 input for other tests without requiring a format conversion.
15357 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15358 see a y, cb and cr stripe from top to bottom.
15360 The sources accept the following parameters:
15365 Specify the color of the source, only available in the @code{color}
15366 source. For the syntax of this option, check the "Color" section in the
15367 ffmpeg-utils manual.
15370 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15371 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15372 pixels to be used as identity matrix for 3D lookup tables. Each component is
15373 coded on a @code{1/(N*N)} scale.
15376 Specify the size of the sourced video. For the syntax of this option, check the
15377 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15378 The default value is @code{320x240}.
15380 This option is not available with the @code{haldclutsrc} filter.
15383 Specify the frame rate of the sourced video, as the number of frames
15384 generated per second. It has to be a string in the format
15385 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15386 number or a valid video frame rate abbreviation. The default value is
15390 Set the sample aspect ratio of the sourced video.
15393 Set the duration of the sourced video. See
15394 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15395 for the accepted syntax.
15397 If not specified, or the expressed duration is negative, the video is
15398 supposed to be generated forever.
15401 Set the number of decimals to show in the timestamp, only available in the
15402 @code{testsrc} source.
15404 The displayed timestamp value will correspond to the original
15405 timestamp value multiplied by the power of 10 of the specified
15406 value. Default value is 0.
15409 For example the following:
15411 testsrc=duration=5.3:size=qcif:rate=10
15414 will generate a video with a duration of 5.3 seconds, with size
15415 176x144 and a frame rate of 10 frames per second.
15417 The following graph description will generate a red source
15418 with an opacity of 0.2, with size "qcif" and a frame rate of 10
15421 color=c=red@@0.2:s=qcif:r=10
15424 If the input content is to be ignored, @code{nullsrc} can be used. The
15425 following command generates noise in the luminance plane by employing
15426 the @code{geq} filter:
15428 nullsrc=s=256x256, geq=random(1)*255:128:128
15431 @subsection Commands
15433 The @code{color} source supports the following commands:
15437 Set the color of the created image. Accepts the same syntax of the
15438 corresponding @option{color} option.
15441 @c man end VIDEO SOURCES
15443 @chapter Video Sinks
15444 @c man begin VIDEO SINKS
15446 Below is a description of the currently available video sinks.
15448 @section buffersink
15450 Buffer video frames, and make them available to the end of the filter
15453 This sink is mainly intended for programmatic use, in particular
15454 through the interface defined in @file{libavfilter/buffersink.h}
15455 or the options system.
15457 It accepts a pointer to an AVBufferSinkContext structure, which
15458 defines the incoming buffers' formats, to be passed as the opaque
15459 parameter to @code{avfilter_init_filter} for initialization.
15463 Null video sink: do absolutely nothing with the input video. It is
15464 mainly useful as a template and for use in analysis / debugging
15467 @c man end VIDEO SINKS
15469 @chapter Multimedia Filters
15470 @c man begin MULTIMEDIA FILTERS
15472 Below is a description of the currently available multimedia filters.
15474 @section ahistogram
15476 Convert input audio to a video output, displaying the volume histogram.
15478 The filter accepts the following options:
15482 Specify how histogram is calculated.
15484 It accepts the following values:
15487 Use single histogram for all channels.
15489 Use separate histogram for each channel.
15491 Default is @code{single}.
15494 Set frame rate, expressed as number of frames per second. Default
15498 Specify the video size for the output. For the syntax of this option, check the
15499 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15500 Default value is @code{hd720}.
15505 It accepts the following values:
15516 reverse logarithmic
15518 Default is @code{log}.
15521 Set amplitude scale.
15523 It accepts the following values:
15530 Default is @code{log}.
15533 Set how much frames to accumulate in histogram.
15534 Defauls is 1. Setting this to -1 accumulates all frames.
15537 Set histogram ratio of window height.
15540 Set sonogram sliding.
15542 It accepts the following values:
15545 replace old rows with new ones.
15547 scroll from top to bottom.
15549 Default is @code{replace}.
15552 @section aphasemeter
15554 Convert input audio to a video output, displaying the audio phase.
15556 The filter accepts the following options:
15560 Set the output frame rate. Default value is @code{25}.
15563 Set the video size for the output. For the syntax of this option, check the
15564 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15565 Default value is @code{800x400}.
15570 Specify the red, green, blue contrast. Default values are @code{2},
15571 @code{7} and @code{1}.
15572 Allowed range is @code{[0, 255]}.
15575 Set color which will be used for drawing median phase. If color is
15576 @code{none} which is default, no median phase value will be drawn.
15579 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15580 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15581 The @code{-1} means left and right channels are completely out of phase and
15582 @code{1} means channels are in phase.
15584 @section avectorscope
15586 Convert input audio to a video output, representing the audio vector
15589 The filter is used to measure the difference between channels of stereo
15590 audio stream. A monoaural signal, consisting of identical left and right
15591 signal, results in straight vertical line. Any stereo separation is visible
15592 as a deviation from this line, creating a Lissajous figure.
15593 If the straight (or deviation from it) but horizontal line appears this
15594 indicates that the left and right channels are out of phase.
15596 The filter accepts the following options:
15600 Set the vectorscope mode.
15602 Available values are:
15605 Lissajous rotated by 45 degrees.
15608 Same as above but not rotated.
15611 Shape resembling half of circle.
15614 Default value is @samp{lissajous}.
15617 Set the video size for the output. For the syntax of this option, check the
15618 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15619 Default value is @code{400x400}.
15622 Set the output frame rate. Default value is @code{25}.
15628 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15629 @code{160}, @code{80} and @code{255}.
15630 Allowed range is @code{[0, 255]}.
15636 Specify the red, green, blue and alpha fade. Default values are @code{15},
15637 @code{10}, @code{5} and @code{5}.
15638 Allowed range is @code{[0, 255]}.
15641 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15644 Set the vectorscope drawing mode.
15646 Available values are:
15649 Draw dot for each sample.
15652 Draw line between previous and current sample.
15655 Default value is @samp{dot}.
15658 Specify amplitude scale of audio samples.
15660 Available values are:
15677 @subsection Examples
15681 Complete example using @command{ffplay}:
15683 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15684 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15688 @section bench, abench
15690 Benchmark part of a filtergraph.
15692 The filter accepts the following options:
15696 Start or stop a timer.
15698 Available values are:
15701 Get the current time, set it as frame metadata (using the key
15702 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15705 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15706 the input frame metadata to get the time difference. Time difference, average,
15707 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15708 @code{min}) are then printed. The timestamps are expressed in seconds.
15712 @subsection Examples
15716 Benchmark @ref{selectivecolor} filter:
15718 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15724 Concatenate audio and video streams, joining them together one after the
15727 The filter works on segments of synchronized video and audio streams. All
15728 segments must have the same number of streams of each type, and that will
15729 also be the number of streams at output.
15731 The filter accepts the following options:
15736 Set the number of segments. Default is 2.
15739 Set the number of output video streams, that is also the number of video
15740 streams in each segment. Default is 1.
15743 Set the number of output audio streams, that is also the number of audio
15744 streams in each segment. Default is 0.
15747 Activate unsafe mode: do not fail if segments have a different format.
15751 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15752 @var{a} audio outputs.
15754 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15755 segment, in the same order as the outputs, then the inputs for the second
15758 Related streams do not always have exactly the same duration, for various
15759 reasons including codec frame size or sloppy authoring. For that reason,
15760 related synchronized streams (e.g. a video and its audio track) should be
15761 concatenated at once. The concat filter will use the duration of the longest
15762 stream in each segment (except the last one), and if necessary pad shorter
15763 audio streams with silence.
15765 For this filter to work correctly, all segments must start at timestamp 0.
15767 All corresponding streams must have the same parameters in all segments; the
15768 filtering system will automatically select a common pixel format for video
15769 streams, and a common sample format, sample rate and channel layout for
15770 audio streams, but other settings, such as resolution, must be converted
15771 explicitly by the user.
15773 Different frame rates are acceptable but will result in variable frame rate
15774 at output; be sure to configure the output file to handle it.
15776 @subsection Examples
15780 Concatenate an opening, an episode and an ending, all in bilingual version
15781 (video in stream 0, audio in streams 1 and 2):
15783 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15784 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15785 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15786 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15790 Concatenate two parts, handling audio and video separately, using the
15791 (a)movie sources, and adjusting the resolution:
15793 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15794 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15795 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15797 Note that a desync will happen at the stitch if the audio and video streams
15798 do not have exactly the same duration in the first file.
15802 @section drawgraph, adrawgraph
15804 Draw a graph using input video or audio metadata.
15806 It accepts the following parameters:
15810 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15813 Set 1st foreground color expression.
15816 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15819 Set 2nd foreground color expression.
15822 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15825 Set 3rd foreground color expression.
15828 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15831 Set 4th foreground color expression.
15834 Set minimal value of metadata value.
15837 Set maximal value of metadata value.
15840 Set graph background color. Default is white.
15845 Available values for mode is:
15852 Default is @code{line}.
15857 Available values for slide is:
15860 Draw new frame when right border is reached.
15863 Replace old columns with new ones.
15866 Scroll from right to left.
15869 Scroll from left to right.
15872 Draw single picture.
15875 Default is @code{frame}.
15878 Set size of graph video. For the syntax of this option, check the
15879 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15880 The default value is @code{900x256}.
15882 The foreground color expressions can use the following variables:
15885 Minimal value of metadata value.
15888 Maximal value of metadata value.
15891 Current metadata key value.
15894 The color is defined as 0xAABBGGRR.
15897 Example using metadata from @ref{signalstats} filter:
15899 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15902 Example using metadata from @ref{ebur128} filter:
15904 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15910 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15911 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15912 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15913 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15915 The filter also has a video output (see the @var{video} option) with a real
15916 time graph to observe the loudness evolution. The graphic contains the logged
15917 message mentioned above, so it is not printed anymore when this option is set,
15918 unless the verbose logging is set. The main graphing area contains the
15919 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15920 the momentary loudness (400 milliseconds).
15922 More information about the Loudness Recommendation EBU R128 on
15923 @url{http://tech.ebu.ch/loudness}.
15925 The filter accepts the following options:
15930 Activate the video output. The audio stream is passed unchanged whether this
15931 option is set or no. The video stream will be the first output stream if
15932 activated. Default is @code{0}.
15935 Set the video size. This option is for video only. For the syntax of this
15937 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15938 Default and minimum resolution is @code{640x480}.
15941 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15942 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15943 other integer value between this range is allowed.
15946 Set metadata injection. If set to @code{1}, the audio input will be segmented
15947 into 100ms output frames, each of them containing various loudness information
15948 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15950 Default is @code{0}.
15953 Force the frame logging level.
15955 Available values are:
15958 information logging level
15960 verbose logging level
15963 By default, the logging level is set to @var{info}. If the @option{video} or
15964 the @option{metadata} options are set, it switches to @var{verbose}.
15969 Available modes can be cumulated (the option is a @code{flag} type). Possible
15973 Disable any peak mode (default).
15975 Enable sample-peak mode.
15977 Simple peak mode looking for the higher sample value. It logs a message
15978 for sample-peak (identified by @code{SPK}).
15980 Enable true-peak mode.
15982 If enabled, the peak lookup is done on an over-sampled version of the input
15983 stream for better peak accuracy. It logs a message for true-peak.
15984 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15985 This mode requires a build with @code{libswresample}.
15989 Treat mono input files as "dual mono". If a mono file is intended for playback
15990 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15991 If set to @code{true}, this option will compensate for this effect.
15992 Multi-channel input files are not affected by this option.
15995 Set a specific pan law to be used for the measurement of dual mono files.
15996 This parameter is optional, and has a default value of -3.01dB.
15999 @subsection Examples
16003 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
16005 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
16009 Run an analysis with @command{ffmpeg}:
16011 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
16015 @section interleave, ainterleave
16017 Temporally interleave frames from several inputs.
16019 @code{interleave} works with video inputs, @code{ainterleave} with audio.
16021 These filters read frames from several inputs and send the oldest
16022 queued frame to the output.
16024 Input streams must have well defined, monotonically increasing frame
16027 In order to submit one frame to output, these filters need to enqueue
16028 at least one frame for each input, so they cannot work in case one
16029 input is not yet terminated and will not receive incoming frames.
16031 For example consider the case when one input is a @code{select} filter
16032 which always drops input frames. The @code{interleave} filter will keep
16033 reading from that input, but it will never be able to send new frames
16034 to output until the input sends an end-of-stream signal.
16036 Also, depending on inputs synchronization, the filters will drop
16037 frames in case one input receives more frames than the other ones, and
16038 the queue is already filled.
16040 These filters accept the following options:
16044 Set the number of different inputs, it is 2 by default.
16047 @subsection Examples
16051 Interleave frames belonging to different streams using @command{ffmpeg}:
16053 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
16057 Add flickering blur effect:
16059 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
16063 @section metadata, ametadata
16065 Manipulate frame metadata.
16067 This filter accepts the following options:
16071 Set mode of operation of the filter.
16073 Can be one of the following:
16077 If both @code{value} and @code{key} is set, select frames
16078 which have such metadata. If only @code{key} is set, select
16079 every frame that has such key in metadata.
16082 Add new metadata @code{key} and @code{value}. If key is already available
16086 Modify value of already present key.
16089 If @code{value} is set, delete only keys that have such value.
16090 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
16094 Print key and its value if metadata was found. If @code{key} is not set print all
16095 metadata values available in frame.
16099 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
16102 Set metadata value which will be used. This option is mandatory for
16103 @code{modify} and @code{add} mode.
16106 Which function to use when comparing metadata value and @code{value}.
16108 Can be one of following:
16112 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
16115 Values are interpreted as strings, returns true if metadata value starts with
16116 the @code{value} option string.
16119 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
16122 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
16125 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
16128 Values are interpreted as floats, returns true if expression from option @code{expr}
16133 Set expression which is used when @code{function} is set to @code{expr}.
16134 The expression is evaluated through the eval API and can contain the following
16139 Float representation of @code{value} from metadata key.
16142 Float representation of @code{value} as supplied by user in @code{value} option.
16145 If specified in @code{print} mode, output is written to the named file. Instead of
16146 plain filename any writable url can be specified. Filename ``-'' is a shorthand
16147 for standard output. If @code{file} option is not set, output is written to the log
16148 with AV_LOG_INFO loglevel.
16153 @subsection Examples
16157 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
16160 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
16163 Print silencedetect output to file @file{metadata.txt}.
16165 silencedetect,ametadata=mode=print:file=metadata.txt
16168 Direct all metadata to a pipe with file descriptor 4.
16170 metadata=mode=print:file='pipe\:4'
16174 @section perms, aperms
16176 Set read/write permissions for the output frames.
16178 These filters are mainly aimed at developers to test direct path in the
16179 following filter in the filtergraph.
16181 The filters accept the following options:
16185 Select the permissions mode.
16187 It accepts the following values:
16190 Do nothing. This is the default.
16192 Set all the output frames read-only.
16194 Set all the output frames directly writable.
16196 Make the frame read-only if writable, and writable if read-only.
16198 Set each output frame read-only or writable randomly.
16202 Set the seed for the @var{random} mode, must be an integer included between
16203 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16204 @code{-1}, the filter will try to use a good random seed on a best effort
16208 Note: in case of auto-inserted filter between the permission filter and the
16209 following one, the permission might not be received as expected in that
16210 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
16211 perms/aperms filter can avoid this problem.
16213 @section realtime, arealtime
16215 Slow down filtering to match real time approximatively.
16217 These filters will pause the filtering for a variable amount of time to
16218 match the output rate with the input timestamps.
16219 They are similar to the @option{re} option to @code{ffmpeg}.
16221 They accept the following options:
16225 Time limit for the pauses. Any pause longer than that will be considered
16226 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16230 @section select, aselect
16232 Select frames to pass in output.
16234 This filter accepts the following options:
16239 Set expression, which is evaluated for each input frame.
16241 If the expression is evaluated to zero, the frame is discarded.
16243 If the evaluation result is negative or NaN, the frame is sent to the
16244 first output; otherwise it is sent to the output with index
16245 @code{ceil(val)-1}, assuming that the input index starts from 0.
16247 For example a value of @code{1.2} corresponds to the output with index
16248 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16251 Set the number of outputs. The output to which to send the selected
16252 frame is based on the result of the evaluation. Default value is 1.
16255 The expression can contain the following constants:
16259 The (sequential) number of the filtered frame, starting from 0.
16262 The (sequential) number of the selected frame, starting from 0.
16264 @item prev_selected_n
16265 The sequential number of the last selected frame. It's NAN if undefined.
16268 The timebase of the input timestamps.
16271 The PTS (Presentation TimeStamp) of the filtered video frame,
16272 expressed in @var{TB} units. It's NAN if undefined.
16275 The PTS of the filtered video frame,
16276 expressed in seconds. It's NAN if undefined.
16279 The PTS of the previously filtered video frame. It's NAN if undefined.
16281 @item prev_selected_pts
16282 The PTS of the last previously filtered video frame. It's NAN if undefined.
16284 @item prev_selected_t
16285 The PTS of the last previously selected video frame. It's NAN if undefined.
16288 The PTS of the first video frame in the video. It's NAN if undefined.
16291 The time of the first video frame in the video. It's NAN if undefined.
16293 @item pict_type @emph{(video only)}
16294 The type of the filtered frame. It can assume one of the following
16306 @item interlace_type @emph{(video only)}
16307 The frame interlace type. It can assume one of the following values:
16310 The frame is progressive (not interlaced).
16312 The frame is top-field-first.
16314 The frame is bottom-field-first.
16317 @item consumed_sample_n @emph{(audio only)}
16318 the number of selected samples before the current frame
16320 @item samples_n @emph{(audio only)}
16321 the number of samples in the current frame
16323 @item sample_rate @emph{(audio only)}
16324 the input sample rate
16327 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16330 the position in the file of the filtered frame, -1 if the information
16331 is not available (e.g. for synthetic video)
16333 @item scene @emph{(video only)}
16334 value between 0 and 1 to indicate a new scene; a low value reflects a low
16335 probability for the current frame to introduce a new scene, while a higher
16336 value means the current frame is more likely to be one (see the example below)
16338 @item concatdec_select
16339 The concat demuxer can select only part of a concat input file by setting an
16340 inpoint and an outpoint, but the output packets may not be entirely contained
16341 in the selected interval. By using this variable, it is possible to skip frames
16342 generated by the concat demuxer which are not exactly contained in the selected
16345 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16346 and the @var{lavf.concat.duration} packet metadata values which are also
16347 present in the decoded frames.
16349 The @var{concatdec_select} variable is -1 if the frame pts is at least
16350 start_time and either the duration metadata is missing or the frame pts is less
16351 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16354 That basically means that an input frame is selected if its pts is within the
16355 interval set by the concat demuxer.
16359 The default value of the select expression is "1".
16361 @subsection Examples
16365 Select all frames in input:
16370 The example above is the same as:
16382 Select only I-frames:
16384 select='eq(pict_type\,I)'
16388 Select one frame every 100:
16390 select='not(mod(n\,100))'
16394 Select only frames contained in the 10-20 time interval:
16396 select=between(t\,10\,20)
16400 Select only I-frames contained in the 10-20 time interval:
16402 select=between(t\,10\,20)*eq(pict_type\,I)
16406 Select frames with a minimum distance of 10 seconds:
16408 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
16412 Use aselect to select only audio frames with samples number > 100:
16414 aselect='gt(samples_n\,100)'
16418 Create a mosaic of the first scenes:
16420 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
16423 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
16427 Send even and odd frames to separate outputs, and compose them:
16429 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
16433 Select useful frames from an ffconcat file which is using inpoints and
16434 outpoints but where the source files are not intra frame only.
16436 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
16440 @section sendcmd, asendcmd
16442 Send commands to filters in the filtergraph.
16444 These filters read commands to be sent to other filters in the
16447 @code{sendcmd} must be inserted between two video filters,
16448 @code{asendcmd} must be inserted between two audio filters, but apart
16449 from that they act the same way.
16451 The specification of commands can be provided in the filter arguments
16452 with the @var{commands} option, or in a file specified by the
16453 @var{filename} option.
16455 These filters accept the following options:
16458 Set the commands to be read and sent to the other filters.
16460 Set the filename of the commands to be read and sent to the other
16464 @subsection Commands syntax
16466 A commands description consists of a sequence of interval
16467 specifications, comprising a list of commands to be executed when a
16468 particular event related to that interval occurs. The occurring event
16469 is typically the current frame time entering or leaving a given time
16472 An interval is specified by the following syntax:
16474 @var{START}[-@var{END}] @var{COMMANDS};
16477 The time interval is specified by the @var{START} and @var{END} times.
16478 @var{END} is optional and defaults to the maximum time.
16480 The current frame time is considered within the specified interval if
16481 it is included in the interval [@var{START}, @var{END}), that is when
16482 the time is greater or equal to @var{START} and is lesser than
16485 @var{COMMANDS} consists of a sequence of one or more command
16486 specifications, separated by ",", relating to that interval. The
16487 syntax of a command specification is given by:
16489 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
16492 @var{FLAGS} is optional and specifies the type of events relating to
16493 the time interval which enable sending the specified command, and must
16494 be a non-null sequence of identifier flags separated by "+" or "|" and
16495 enclosed between "[" and "]".
16497 The following flags are recognized:
16500 The command is sent when the current frame timestamp enters the
16501 specified interval. In other words, the command is sent when the
16502 previous frame timestamp was not in the given interval, and the
16506 The command is sent when the current frame timestamp leaves the
16507 specified interval. In other words, the command is sent when the
16508 previous frame timestamp was in the given interval, and the
16512 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16515 @var{TARGET} specifies the target of the command, usually the name of
16516 the filter class or a specific filter instance name.
16518 @var{COMMAND} specifies the name of the command for the target filter.
16520 @var{ARG} is optional and specifies the optional list of argument for
16521 the given @var{COMMAND}.
16523 Between one interval specification and another, whitespaces, or
16524 sequences of characters starting with @code{#} until the end of line,
16525 are ignored and can be used to annotate comments.
16527 A simplified BNF description of the commands specification syntax
16530 @var{COMMAND_FLAG} ::= "enter" | "leave"
16531 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16532 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16533 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16534 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16535 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16538 @subsection Examples
16542 Specify audio tempo change at second 4:
16544 asendcmd=c='4.0 atempo tempo 1.5',atempo
16548 Specify a list of drawtext and hue commands in a file.
16550 # show text in the interval 5-10
16551 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16552 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16554 # desaturate the image in the interval 15-20
16555 15.0-20.0 [enter] hue s 0,
16556 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16558 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16560 # apply an exponential saturation fade-out effect, starting from time 25
16561 25 [enter] hue s exp(25-t)
16564 A filtergraph allowing to read and process the above command list
16565 stored in a file @file{test.cmd}, can be specified with:
16567 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16572 @section setpts, asetpts
16574 Change the PTS (presentation timestamp) of the input frames.
16576 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16578 This filter accepts the following options:
16583 The expression which is evaluated for each frame to construct its timestamp.
16587 The expression is evaluated through the eval API and can contain the following
16592 frame rate, only defined for constant frame-rate video
16595 The presentation timestamp in input
16598 The count of the input frame for video or the number of consumed samples,
16599 not including the current frame for audio, starting from 0.
16601 @item NB_CONSUMED_SAMPLES
16602 The number of consumed samples, not including the current frame (only
16605 @item NB_SAMPLES, S
16606 The number of samples in the current frame (only audio)
16608 @item SAMPLE_RATE, SR
16609 The audio sample rate.
16612 The PTS of the first frame.
16615 the time in seconds of the first frame
16618 State whether the current frame is interlaced.
16621 the time in seconds of the current frame
16624 original position in the file of the frame, or undefined if undefined
16625 for the current frame
16628 The previous input PTS.
16631 previous input time in seconds
16634 The previous output PTS.
16637 previous output time in seconds
16640 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16644 The wallclock (RTC) time at the start of the movie in microseconds.
16647 The timebase of the input timestamps.
16651 @subsection Examples
16655 Start counting PTS from zero
16657 setpts=PTS-STARTPTS
16661 Apply fast motion effect:
16667 Apply slow motion effect:
16673 Set fixed rate of 25 frames per second:
16679 Set fixed rate 25 fps with some jitter:
16681 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16685 Apply an offset of 10 seconds to the input PTS:
16691 Generate timestamps from a "live source" and rebase onto the current timebase:
16693 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16697 Generate timestamps by counting samples:
16704 @section settb, asettb
16706 Set the timebase to use for the output frames timestamps.
16707 It is mainly useful for testing timebase configuration.
16709 It accepts the following parameters:
16714 The expression which is evaluated into the output timebase.
16718 The value for @option{tb} is an arithmetic expression representing a
16719 rational. The expression can contain the constants "AVTB" (the default
16720 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16721 audio only). Default value is "intb".
16723 @subsection Examples
16727 Set the timebase to 1/25:
16733 Set the timebase to 1/10:
16739 Set the timebase to 1001/1000:
16745 Set the timebase to 2*intb:
16751 Set the default timebase value:
16758 Convert input audio to a video output representing frequency spectrum
16759 logarithmically using Brown-Puckette constant Q transform algorithm with
16760 direct frequency domain coefficient calculation (but the transform itself
16761 is not really constant Q, instead the Q factor is actually variable/clamped),
16762 with musical tone scale, from E0 to D#10.
16764 The filter accepts the following options:
16768 Specify the video size for the output. It must be even. For the syntax of this option,
16769 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16770 Default value is @code{1920x1080}.
16773 Set the output frame rate. Default value is @code{25}.
16776 Set the bargraph height. It must be even. Default value is @code{-1} which
16777 computes the bargraph height automatically.
16780 Set the axis height. It must be even. Default value is @code{-1} which computes
16781 the axis height automatically.
16784 Set the sonogram height. It must be even. Default value is @code{-1} which
16785 computes the sonogram height automatically.
16788 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16789 instead. Default value is @code{1}.
16791 @item sono_v, volume
16792 Specify the sonogram volume expression. It can contain variables:
16795 the @var{bar_v} evaluated expression
16796 @item frequency, freq, f
16797 the frequency where it is evaluated
16798 @item timeclamp, tc
16799 the value of @var{timeclamp} option
16803 @item a_weighting(f)
16804 A-weighting of equal loudness
16805 @item b_weighting(f)
16806 B-weighting of equal loudness
16807 @item c_weighting(f)
16808 C-weighting of equal loudness.
16810 Default value is @code{16}.
16812 @item bar_v, volume2
16813 Specify the bargraph volume expression. It can contain variables:
16816 the @var{sono_v} evaluated expression
16817 @item frequency, freq, f
16818 the frequency where it is evaluated
16819 @item timeclamp, tc
16820 the value of @var{timeclamp} option
16824 @item a_weighting(f)
16825 A-weighting of equal loudness
16826 @item b_weighting(f)
16827 B-weighting of equal loudness
16828 @item c_weighting(f)
16829 C-weighting of equal loudness.
16831 Default value is @code{sono_v}.
16833 @item sono_g, gamma
16834 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16835 higher gamma makes the spectrum having more range. Default value is @code{3}.
16836 Acceptable range is @code{[1, 7]}.
16838 @item bar_g, gamma2
16839 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16843 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
16844 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
16846 @item timeclamp, tc
16847 Specify the transform timeclamp. At low frequency, there is trade-off between
16848 accuracy in time domain and frequency domain. If timeclamp is lower,
16849 event in time domain is represented more accurately (such as fast bass drum),
16850 otherwise event in frequency domain is represented more accurately
16851 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16854 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16855 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16858 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16859 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16862 This option is deprecated and ignored.
16865 Specify the transform length in time domain. Use this option to control accuracy
16866 trade-off between time domain and frequency domain at every frequency sample.
16867 It can contain variables:
16869 @item frequency, freq, f
16870 the frequency where it is evaluated
16871 @item timeclamp, tc
16872 the value of @var{timeclamp} option.
16874 Default value is @code{384*tc/(384+tc*f)}.
16877 Specify the transform count for every video frame. Default value is @code{6}.
16878 Acceptable range is @code{[1, 30]}.
16881 Specify the transform count for every single pixel. Default value is @code{0},
16882 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16885 Specify font file for use with freetype to draw the axis. If not specified,
16886 use embedded font. Note that drawing with font file or embedded font is not
16887 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16891 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
16892 The : in the pattern may be replaced by | to avoid unnecessary escaping.
16895 Specify font color expression. This is arithmetic expression that should return
16896 integer value 0xRRGGBB. It can contain variables:
16898 @item frequency, freq, f
16899 the frequency where it is evaluated
16900 @item timeclamp, tc
16901 the value of @var{timeclamp} option
16906 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16907 @item r(x), g(x), b(x)
16908 red, green, and blue value of intensity x.
16910 Default value is @code{st(0, (midi(f)-59.5)/12);
16911 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16912 r(1-ld(1)) + b(ld(1))}.
16915 Specify image file to draw the axis. This option override @var{fontfile} and
16916 @var{fontcolor} option.
16919 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16920 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16921 Default value is @code{1}.
16924 Set colorspace. The accepted values are:
16927 Unspecified (default)
16936 BT.470BG or BT.601-6 625
16939 SMPTE-170M or BT.601-6 525
16945 BT.2020 with non-constant luminance
16950 Set spectrogram color scheme. This is list of floating point values with format
16951 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
16952 The default is @code{1|0.5|0|0|0.5|1}.
16956 @subsection Examples
16960 Playing audio while showing the spectrum:
16962 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16966 Same as above, but with frame rate 30 fps:
16968 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16972 Playing at 1280x720:
16974 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16978 Disable sonogram display:
16984 A1 and its harmonics: A1, A2, (near)E3, A3:
16986 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),
16987 asplit[a][out1]; [a] showcqt [out0]'
16991 Same as above, but with more accuracy in frequency domain:
16993 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),
16994 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
17000 bar_v=10:sono_v=bar_v*a_weighting(f)
17004 Custom gamma, now spectrum is linear to the amplitude.
17010 Custom tlength equation:
17012 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)))'
17016 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
17018 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
17022 Custom font using fontconfig:
17024 font='Courier New,Monospace,mono|bold'
17028 Custom frequency range with custom axis using image file:
17030 axisfile=myaxis.png:basefreq=40:endfreq=10000
17036 Convert input audio to video output representing the audio power spectrum.
17037 Audio amplitude is on Y-axis while frequency is on X-axis.
17039 The filter accepts the following options:
17043 Specify size of video. For the syntax of this option, check the
17044 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17045 Default is @code{1024x512}.
17049 This set how each frequency bin will be represented.
17051 It accepts the following values:
17057 Default is @code{bar}.
17060 Set amplitude scale.
17062 It accepts the following values:
17076 Default is @code{log}.
17079 Set frequency scale.
17081 It accepts the following values:
17090 Reverse logarithmic scale.
17092 Default is @code{lin}.
17097 It accepts the following values:
17113 Default is @code{w2048}
17116 Set windowing function.
17118 It accepts the following values:
17140 Default is @code{hanning}.
17143 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17144 which means optimal overlap for selected window function will be picked.
17147 Set time averaging. Setting this to 0 will display current maximal peaks.
17148 Default is @code{1}, which means time averaging is disabled.
17151 Specify list of colors separated by space or by '|' which will be used to
17152 draw channel frequencies. Unrecognized or missing colors will be replaced
17156 Set channel display mode.
17158 It accepts the following values:
17163 Default is @code{combined}.
17166 Set minimum amplitude used in @code{log} amplitude scaler.
17170 @anchor{showspectrum}
17171 @section showspectrum
17173 Convert input audio to a video output, representing the audio frequency
17176 The filter accepts the following options:
17180 Specify the video size for the output. For the syntax of this option, check the
17181 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17182 Default value is @code{640x512}.
17185 Specify how the spectrum should slide along the window.
17187 It accepts the following values:
17190 the samples start again on the left when they reach the right
17192 the samples scroll from right to left
17194 frames are only produced when the samples reach the right
17196 the samples scroll from left to right
17199 Default value is @code{replace}.
17202 Specify display mode.
17204 It accepts the following values:
17207 all channels are displayed in the same row
17209 all channels are displayed in separate rows
17212 Default value is @samp{combined}.
17215 Specify display color mode.
17217 It accepts the following values:
17220 each channel is displayed in a separate color
17222 each channel is displayed using the same color scheme
17224 each channel is displayed using the rainbow color scheme
17226 each channel is displayed using the moreland color scheme
17228 each channel is displayed using the nebulae color scheme
17230 each channel is displayed using the fire color scheme
17232 each channel is displayed using the fiery color scheme
17234 each channel is displayed using the fruit color scheme
17236 each channel is displayed using the cool color scheme
17239 Default value is @samp{channel}.
17242 Specify scale used for calculating intensity color values.
17244 It accepts the following values:
17249 square root, default
17260 Default value is @samp{sqrt}.
17263 Set saturation modifier for displayed colors. Negative values provide
17264 alternative color scheme. @code{0} is no saturation at all.
17265 Saturation must be in [-10.0, 10.0] range.
17266 Default value is @code{1}.
17269 Set window function.
17271 It accepts the following values:
17295 Default value is @code{hann}.
17298 Set orientation of time vs frequency axis. Can be @code{vertical} or
17299 @code{horizontal}. Default is @code{vertical}.
17302 Set ratio of overlap window. Default value is @code{0}.
17303 When value is @code{1} overlap is set to recommended size for specific
17304 window function currently used.
17307 Set scale gain for calculating intensity color values.
17308 Default value is @code{1}.
17311 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17314 Set color rotation, must be in [-1.0, 1.0] range.
17315 Default value is @code{0}.
17318 The usage is very similar to the showwaves filter; see the examples in that
17321 @subsection Examples
17325 Large window with logarithmic color scaling:
17327 showspectrum=s=1280x480:scale=log
17331 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17333 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17334 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17338 @section showspectrumpic
17340 Convert input audio to a single video frame, representing the audio frequency
17343 The filter accepts the following options:
17347 Specify the video size for the output. For the syntax of this option, check the
17348 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17349 Default value is @code{4096x2048}.
17352 Specify display mode.
17354 It accepts the following values:
17357 all channels are displayed in the same row
17359 all channels are displayed in separate rows
17361 Default value is @samp{combined}.
17364 Specify display color mode.
17366 It accepts the following values:
17369 each channel is displayed in a separate color
17371 each channel is displayed using the same color scheme
17373 each channel is displayed using the rainbow color scheme
17375 each channel is displayed using the moreland color scheme
17377 each channel is displayed using the nebulae color scheme
17379 each channel is displayed using the fire color scheme
17381 each channel is displayed using the fiery color scheme
17383 each channel is displayed using the fruit color scheme
17385 each channel is displayed using the cool color scheme
17387 Default value is @samp{intensity}.
17390 Specify scale used for calculating intensity color values.
17392 It accepts the following values:
17397 square root, default
17407 Default value is @samp{log}.
17410 Set saturation modifier for displayed colors. Negative values provide
17411 alternative color scheme. @code{0} is no saturation at all.
17412 Saturation must be in [-10.0, 10.0] range.
17413 Default value is @code{1}.
17416 Set window function.
17418 It accepts the following values:
17441 Default value is @code{hann}.
17444 Set orientation of time vs frequency axis. Can be @code{vertical} or
17445 @code{horizontal}. Default is @code{vertical}.
17448 Set scale gain for calculating intensity color values.
17449 Default value is @code{1}.
17452 Draw time and frequency axes and legends. Default is enabled.
17455 Set color rotation, must be in [-1.0, 1.0] range.
17456 Default value is @code{0}.
17459 @subsection Examples
17463 Extract an audio spectrogram of a whole audio track
17464 in a 1024x1024 picture using @command{ffmpeg}:
17466 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
17470 @section showvolume
17472 Convert input audio volume to a video output.
17474 The filter accepts the following options:
17481 Set border width, allowed range is [0, 5]. Default is 1.
17484 Set channel width, allowed range is [80, 8192]. Default is 400.
17487 Set channel height, allowed range is [1, 900]. Default is 20.
17490 Set fade, allowed range is [0.001, 1]. Default is 0.95.
17493 Set volume color expression.
17495 The expression can use the following variables:
17499 Current max volume of channel in dB.
17505 Current channel number, starting from 0.
17509 If set, displays channel names. Default is enabled.
17512 If set, displays volume values. Default is enabled.
17515 Set orientation, can be @code{horizontal} or @code{vertical},
17516 default is @code{horizontal}.
17519 Set step size, allowed range s [0, 5]. Default is 0, which means
17525 Convert input audio to a video output, representing the samples waves.
17527 The filter accepts the following options:
17531 Specify the video size for the output. For the syntax of this option, check the
17532 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17533 Default value is @code{600x240}.
17538 Available values are:
17541 Draw a point for each sample.
17544 Draw a vertical line for each sample.
17547 Draw a point for each sample and a line between them.
17550 Draw a centered vertical line for each sample.
17553 Default value is @code{point}.
17556 Set the number of samples which are printed on the same column. A
17557 larger value will decrease the frame rate. Must be a positive
17558 integer. This option can be set only if the value for @var{rate}
17559 is not explicitly specified.
17562 Set the (approximate) output frame rate. This is done by setting the
17563 option @var{n}. Default value is "25".
17565 @item split_channels
17566 Set if channels should be drawn separately or overlap. Default value is 0.
17569 Set colors separated by '|' which are going to be used for drawing of each channel.
17572 Set amplitude scale.
17574 Available values are:
17592 @subsection Examples
17596 Output the input file audio and the corresponding video representation
17599 amovie=a.mp3,asplit[out0],showwaves[out1]
17603 Create a synthetic signal and show it with showwaves, forcing a
17604 frame rate of 30 frames per second:
17606 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17610 @section showwavespic
17612 Convert input audio to a single video frame, representing the samples waves.
17614 The filter accepts the following options:
17618 Specify the video size for the output. For the syntax of this option, check the
17619 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17620 Default value is @code{600x240}.
17622 @item split_channels
17623 Set if channels should be drawn separately or overlap. Default value is 0.
17626 Set colors separated by '|' which are going to be used for drawing of each channel.
17629 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17633 @subsection Examples
17637 Extract a channel split representation of the wave form of a whole audio track
17638 in a 1024x800 picture using @command{ffmpeg}:
17640 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17644 @section sidedata, asidedata
17646 Delete frame side data, or select frames based on it.
17648 This filter accepts the following options:
17652 Set mode of operation of the filter.
17654 Can be one of the following:
17658 Select every frame with side data of @code{type}.
17661 Delete side data of @code{type}. If @code{type} is not set, delete all side
17667 Set side data type used with all modes. Must be set for @code{select} mode. For
17668 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
17669 in @file{libavutil/frame.h}. For example, to choose
17670 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
17674 @section spectrumsynth
17676 Sythesize audio from 2 input video spectrums, first input stream represents
17677 magnitude across time and second represents phase across time.
17678 The filter will transform from frequency domain as displayed in videos back
17679 to time domain as presented in audio output.
17681 This filter is primarily created for reversing processed @ref{showspectrum}
17682 filter outputs, but can synthesize sound from other spectrograms too.
17683 But in such case results are going to be poor if the phase data is not
17684 available, because in such cases phase data need to be recreated, usually
17685 its just recreated from random noise.
17686 For best results use gray only output (@code{channel} color mode in
17687 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17688 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17689 @code{data} option. Inputs videos should generally use @code{fullframe}
17690 slide mode as that saves resources needed for decoding video.
17692 The filter accepts the following options:
17696 Specify sample rate of output audio, the sample rate of audio from which
17697 spectrum was generated may differ.
17700 Set number of channels represented in input video spectrums.
17703 Set scale which was used when generating magnitude input spectrum.
17704 Can be @code{lin} or @code{log}. Default is @code{log}.
17707 Set slide which was used when generating inputs spectrums.
17708 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17709 Default is @code{fullframe}.
17712 Set window function used for resynthesis.
17715 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17716 which means optimal overlap for selected window function will be picked.
17719 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17720 Default is @code{vertical}.
17723 @subsection Examples
17727 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17728 then resynthesize videos back to audio with spectrumsynth:
17730 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
17731 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
17732 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17736 @section split, asplit
17738 Split input into several identical outputs.
17740 @code{asplit} works with audio input, @code{split} with video.
17742 The filter accepts a single parameter which specifies the number of outputs. If
17743 unspecified, it defaults to 2.
17745 @subsection Examples
17749 Create two separate outputs from the same input:
17751 [in] split [out0][out1]
17755 To create 3 or more outputs, you need to specify the number of
17758 [in] asplit=3 [out0][out1][out2]
17762 Create two separate outputs from the same input, one cropped and
17765 [in] split [splitout1][splitout2];
17766 [splitout1] crop=100:100:0:0 [cropout];
17767 [splitout2] pad=200:200:100:100 [padout];
17771 Create 5 copies of the input audio with @command{ffmpeg}:
17773 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17779 Receive commands sent through a libzmq client, and forward them to
17780 filters in the filtergraph.
17782 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17783 must be inserted between two video filters, @code{azmq} between two
17786 To enable these filters you need to install the libzmq library and
17787 headers and configure FFmpeg with @code{--enable-libzmq}.
17789 For more information about libzmq see:
17790 @url{http://www.zeromq.org/}
17792 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17793 receives messages sent through a network interface defined by the
17794 @option{bind_address} option.
17796 The received message must be in the form:
17798 @var{TARGET} @var{COMMAND} [@var{ARG}]
17801 @var{TARGET} specifies the target of the command, usually the name of
17802 the filter class or a specific filter instance name.
17804 @var{COMMAND} specifies the name of the command for the target filter.
17806 @var{ARG} is optional and specifies the optional argument list for the
17807 given @var{COMMAND}.
17809 Upon reception, the message is processed and the corresponding command
17810 is injected into the filtergraph. Depending on the result, the filter
17811 will send a reply to the client, adopting the format:
17813 @var{ERROR_CODE} @var{ERROR_REASON}
17817 @var{MESSAGE} is optional.
17819 @subsection Examples
17821 Look at @file{tools/zmqsend} for an example of a zmq client which can
17822 be used to send commands processed by these filters.
17824 Consider the following filtergraph generated by @command{ffplay}
17826 ffplay -dumpgraph 1 -f lavfi "
17827 color=s=100x100:c=red [l];
17828 color=s=100x100:c=blue [r];
17829 nullsrc=s=200x100, zmq [bg];
17830 [bg][l] overlay [bg+l];
17831 [bg+l][r] overlay=x=100 "
17834 To change the color of the left side of the video, the following
17835 command can be used:
17837 echo Parsed_color_0 c yellow | tools/zmqsend
17840 To change the right side:
17842 echo Parsed_color_1 c pink | tools/zmqsend
17845 @c man end MULTIMEDIA FILTERS
17847 @chapter Multimedia Sources
17848 @c man begin MULTIMEDIA SOURCES
17850 Below is a description of the currently available multimedia sources.
17854 This is the same as @ref{movie} source, except it selects an audio
17860 Read audio and/or video stream(s) from a movie container.
17862 It accepts the following parameters:
17866 The name of the resource to read (not necessarily a file; it can also be a
17867 device or a stream accessed through some protocol).
17869 @item format_name, f
17870 Specifies the format assumed for the movie to read, and can be either
17871 the name of a container or an input device. If not specified, the
17872 format is guessed from @var{movie_name} or by probing.
17874 @item seek_point, sp
17875 Specifies the seek point in seconds. The frames will be output
17876 starting from this seek point. The parameter is evaluated with
17877 @code{av_strtod}, so the numerical value may be suffixed by an IS
17878 postfix. The default value is "0".
17881 Specifies the streams to read. Several streams can be specified,
17882 separated by "+". The source will then have as many outputs, in the
17883 same order. The syntax is explained in the ``Stream specifiers''
17884 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17885 respectively the default (best suited) video and audio stream. Default
17886 is "dv", or "da" if the filter is called as "amovie".
17888 @item stream_index, si
17889 Specifies the index of the video stream to read. If the value is -1,
17890 the most suitable video stream will be automatically selected. The default
17891 value is "-1". Deprecated. If the filter is called "amovie", it will select
17892 audio instead of video.
17895 Specifies how many times to read the stream in sequence.
17896 If the value is less than 1, the stream will be read again and again.
17897 Default value is "1".
17899 Note that when the movie is looped the source timestamps are not
17900 changed, so it will generate non monotonically increasing timestamps.
17902 @item discontinuity
17903 Specifies the time difference between frames above which the point is
17904 considered a timestamp discontinuity which is removed by adjusting the later
17908 It allows overlaying a second video on top of the main input of
17909 a filtergraph, as shown in this graph:
17911 input -----------> deltapts0 --> overlay --> output
17914 movie --> scale--> deltapts1 -------+
17916 @subsection Examples
17920 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17921 on top of the input labelled "in":
17923 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17924 [in] setpts=PTS-STARTPTS [main];
17925 [main][over] overlay=16:16 [out]
17929 Read from a video4linux2 device, and overlay it on top of the input
17932 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17933 [in] setpts=PTS-STARTPTS [main];
17934 [main][over] overlay=16:16 [out]
17938 Read the first video stream and the audio stream with id 0x81 from
17939 dvd.vob; the video is connected to the pad named "video" and the audio is
17940 connected to the pad named "audio":
17942 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17946 @subsection Commands
17948 Both movie and amovie support the following commands:
17951 Perform seek using "av_seek_frame".
17952 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17955 @var{stream_index}: If stream_index is -1, a default
17956 stream is selected, and @var{timestamp} is automatically converted
17957 from AV_TIME_BASE units to the stream specific time_base.
17959 @var{timestamp}: Timestamp in AVStream.time_base units
17960 or, if no stream is specified, in AV_TIME_BASE units.
17962 @var{flags}: Flags which select direction and seeking mode.
17966 Get movie duration in AV_TIME_BASE units.
17970 @c man end MULTIMEDIA SOURCES