1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @section Notes on filtergraph escaping
226 Filtergraph description composition entails several levels of
227 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
228 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
229 information about the employed escaping procedure.
231 A first level escaping affects the content of each filter option
232 value, which may contain the special character @code{:} used to
233 separate values, or one of the escaping characters @code{\'}.
235 A second level escaping affects the whole filter description, which
236 may contain the escaping characters @code{\'} or the special
237 characters @code{[],;} used by the filtergraph description.
239 Finally, when you specify a filtergraph on a shell commandline, you
240 need to perform a third level escaping for the shell special
241 characters contained within it.
243 For example, consider the following string to be embedded in
244 the @ref{drawtext} filter description @option{text} value:
246 this is a 'string': may contain one, or more, special characters
249 This string contains the @code{'} special escaping character, and the
250 @code{:} special character, so it needs to be escaped in this way:
252 text=this is a \'string\'\: may contain one, or more, special characters
255 A second level of escaping is required when embedding the filter
256 description in a filtergraph description, in order to escape all the
257 filtergraph special characters. Thus the example above becomes:
259 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
261 (note that in addition to the @code{\'} escaping special characters,
262 also @code{,} needs to be escaped).
264 Finally an additional level of escaping is needed when writing the
265 filtergraph description in a shell command, which depends on the
266 escaping rules of the adopted shell. For example, assuming that
267 @code{\} is special and needs to be escaped with another @code{\}, the
268 previous string will finally result in:
270 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
273 @chapter Timeline editing
275 Some filters support a generic @option{enable} option. For the filters
276 supporting timeline editing, this option can be set to an expression which is
277 evaluated before sending a frame to the filter. If the evaluation is non-zero,
278 the filter will be enabled, otherwise the frame will be sent unchanged to the
279 next filter in the filtergraph.
281 The expression accepts the following values:
284 timestamp expressed in seconds, NAN if the input timestamp is unknown
287 sequential number of the input frame, starting from 0
290 the position in the file of the input frame, NAN if unknown
294 width and height of the input frame if video
297 Additionally, these filters support an @option{enable} command that can be used
298 to re-define the expression.
300 Like any other filtering option, the @option{enable} option follows the same
303 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
304 minutes, and a @ref{curves} filter starting at 3 seconds:
306 smartblur = enable='between(t,10,3*60)',
307 curves = enable='gte(t,3)' : preset=cross_process
310 See @code{ffmpeg -filters} to view which filters have timeline support.
312 @c man end FILTERGRAPH DESCRIPTION
315 @chapter Options for filters with several inputs (framesync)
316 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
318 Some filters with several inputs support a common set of options.
319 These options can only be set by name, not with the short notation.
323 The action to take when EOF is encountered on the secondary input; it accepts
324 one of the following values:
328 Repeat the last frame (the default).
332 Pass the main input through.
336 If set to 1, force the output to terminate when the shortest input
337 terminates. Default value is 0.
340 If set to 1, force the filter to extend the last frame of secondary streams
341 until the end of the primary stream. A value of 0 disables this behavior.
345 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
347 @chapter Audio Filters
348 @c man begin AUDIO FILTERS
350 When you configure your FFmpeg build, you can disable any of the
351 existing filters using @code{--disable-filters}.
352 The configure output will show the audio filters included in your
355 Below is a description of the currently available audio filters.
359 A compressor is mainly used to reduce the dynamic range of a signal.
360 Especially modern music is mostly compressed at a high ratio to
361 improve the overall loudness. It's done to get the highest attention
362 of a listener, "fatten" the sound and bring more "power" to the track.
363 If a signal is compressed too much it may sound dull or "dead"
364 afterwards or it may start to "pump" (which could be a powerful effect
365 but can also destroy a track completely).
366 The right compression is the key to reach a professional sound and is
367 the high art of mixing and mastering. Because of its complex settings
368 it may take a long time to get the right feeling for this kind of effect.
370 Compression is done by detecting the volume above a chosen level
371 @code{threshold} and dividing it by the factor set with @code{ratio}.
372 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
373 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
374 the signal would cause distortion of the waveform the reduction can be
375 levelled over the time. This is done by setting "Attack" and "Release".
376 @code{attack} determines how long the signal has to rise above the threshold
377 before any reduction will occur and @code{release} sets the time the signal
378 has to fall below the threshold to reduce the reduction again. Shorter signals
379 than the chosen attack time will be left untouched.
380 The overall reduction of the signal can be made up afterwards with the
381 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
382 raising the makeup to this level results in a signal twice as loud than the
383 source. To gain a softer entry in the compression the @code{knee} flattens the
384 hard edge at the threshold in the range of the chosen decibels.
386 The filter accepts the following options:
390 Set input gain. Default is 1. Range is between 0.015625 and 64.
393 If a signal of stream rises above this level it will affect the gain
395 By default it is 0.125. Range is between 0.00097563 and 1.
398 Set a ratio by which the signal is reduced. 1:2 means that if the level
399 rose 4dB above the threshold, it will be only 2dB above after the reduction.
400 Default is 2. Range is between 1 and 20.
403 Amount of milliseconds the signal has to rise above the threshold before gain
404 reduction starts. Default is 20. Range is between 0.01 and 2000.
407 Amount of milliseconds the signal has to fall below the threshold before
408 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
411 Set the amount by how much signal will be amplified after processing.
412 Default is 1. Range is from 1 to 64.
415 Curve the sharp knee around the threshold to enter gain reduction more softly.
416 Default is 2.82843. Range is between 1 and 8.
419 Choose if the @code{average} level between all channels of input stream
420 or the louder(@code{maximum}) channel of input stream affects the
421 reduction. Default is @code{average}.
424 Should the exact signal be taken in case of @code{peak} or an RMS one in case
425 of @code{rms}. Default is @code{rms} which is mostly smoother.
428 How much to use compressed signal in output. Default is 1.
429 Range is between 0 and 1.
433 Simple audio dynamic range commpression/expansion filter.
435 The filter accepts the following options:
439 Set contrast. Default is 33. Allowed range is between 0 and 100.
444 Copy the input audio source unchanged to the output. This is mainly useful for
449 Apply cross fade from one input audio stream to another input audio stream.
450 The cross fade is applied for specified duration near the end of first stream.
452 The filter accepts the following options:
456 Specify the number of samples for which the cross fade effect has to last.
457 At the end of the cross fade effect the first input audio will be completely
458 silent. Default is 44100.
461 Specify the duration of the cross fade effect. See
462 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
463 for the accepted syntax.
464 By default the duration is determined by @var{nb_samples}.
465 If set this option is used instead of @var{nb_samples}.
468 Should first stream end overlap with second stream start. Default is enabled.
471 Set curve for cross fade transition for first stream.
474 Set curve for cross fade transition for second stream.
476 For description of available curve types see @ref{afade} filter description.
483 Cross fade from one input to another:
485 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
489 Cross fade from one input to another but without overlapping:
491 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Reduce audio bit resolution.
499 This filter is bit crusher with enhanced functionality. A bit crusher
500 is used to audibly reduce number of bits an audio signal is sampled
501 with. This doesn't change the bit depth at all, it just produces the
502 effect. Material reduced in bit depth sounds more harsh and "digital".
503 This filter is able to even round to continuous values instead of discrete
505 Additionally it has a D/C offset which results in different crushing of
506 the lower and the upper half of the signal.
507 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
509 Another feature of this filter is the logarithmic mode.
510 This setting switches from linear distances between bits to logarithmic ones.
511 The result is a much more "natural" sounding crusher which doesn't gate low
512 signals for example. The human ear has a logarithmic perception, too
513 so this kind of crushing is much more pleasant.
514 Logarithmic crushing is also able to get anti-aliased.
516 The filter accepts the following options:
532 Can be linear: @code{lin} or logarithmic: @code{log}.
541 Set sample reduction.
544 Enable LFO. By default disabled.
555 Delay one or more audio channels.
557 Samples in delayed channel are filled with silence.
559 The filter accepts the following option:
563 Set list of delays in milliseconds for each channel separated by '|'.
564 Unused delays will be silently ignored. If number of given delays is
565 smaller than number of channels all remaining channels will not be delayed.
566 If you want to delay exact number of samples, append 'S' to number.
573 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
574 the second channel (and any other channels that may be present) unchanged.
580 Delay second channel by 500 samples, the third channel by 700 samples and leave
581 the first channel (and any other channels that may be present) unchanged.
589 Apply echoing to the input audio.
591 Echoes are reflected sound and can occur naturally amongst mountains
592 (and sometimes large buildings) when talking or shouting; digital echo
593 effects emulate this behaviour and are often used to help fill out the
594 sound of a single instrument or vocal. The time difference between the
595 original signal and the reflection is the @code{delay}, and the
596 loudness of the reflected signal is the @code{decay}.
597 Multiple echoes can have different delays and decays.
599 A description of the accepted parameters follows.
603 Set input gain of reflected signal. Default is @code{0.6}.
606 Set output gain of reflected signal. Default is @code{0.3}.
609 Set list of time intervals in milliseconds between original signal and reflections
610 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
611 Default is @code{1000}.
614 Set list of loudness of reflected signals separated by '|'.
615 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
616 Default is @code{0.5}.
623 Make it sound as if there are twice as many instruments as are actually playing:
625 aecho=0.8:0.88:60:0.4
629 If delay is very short, then it sound like a (metallic) robot playing music:
635 A longer delay will sound like an open air concert in the mountains:
637 aecho=0.8:0.9:1000:0.3
641 Same as above but with one more mountain:
643 aecho=0.8:0.9:1000|1800:0.3|0.25
648 Audio emphasis filter creates or restores material directly taken from LPs or
649 emphased CDs with different filter curves. E.g. to store music on vinyl the
650 signal has to be altered by a filter first to even out the disadvantages of
651 this recording medium.
652 Once the material is played back the inverse filter has to be applied to
653 restore the distortion of the frequency response.
655 The filter accepts the following options:
665 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
666 use @code{production} mode. Default is @code{reproduction} mode.
669 Set filter type. Selects medium. Can be one of the following:
681 select Compact Disc (CD).
687 select 50µs (FM-KF).
689 select 75µs (FM-KF).
695 Modify an audio signal according to the specified expressions.
697 This filter accepts one or more expressions (one for each channel),
698 which are evaluated and used to modify a corresponding audio signal.
700 It accepts the following parameters:
704 Set the '|'-separated expressions list for each separate channel. If
705 the number of input channels is greater than the number of
706 expressions, the last specified expression is used for the remaining
709 @item channel_layout, c
710 Set output channel layout. If not specified, the channel layout is
711 specified by the number of expressions. If set to @samp{same}, it will
712 use by default the same input channel layout.
715 Each expression in @var{exprs} can contain the following constants and functions:
719 channel number of the current expression
722 number of the evaluated sample, starting from 0
728 time of the evaluated sample expressed in seconds
731 @item nb_out_channels
732 input and output number of channels
735 the value of input channel with number @var{CH}
738 Note: this filter is slow. For faster processing you should use a
747 aeval=val(ch)/2:c=same
751 Invert phase of the second channel:
760 Apply fade-in/out effect to input audio.
762 A description of the accepted parameters follows.
766 Specify the effect type, can be either @code{in} for fade-in, or
767 @code{out} for a fade-out effect. Default is @code{in}.
769 @item start_sample, ss
770 Specify the number of the start sample for starting to apply the fade
771 effect. Default is 0.
774 Specify the number of samples for which the fade effect has to last. At
775 the end of the fade-in effect the output audio will have the same
776 volume as the input audio, at the end of the fade-out transition
777 the output audio will be silence. Default is 44100.
780 Specify the start time of the fade effect. Default is 0.
781 The value must be specified as a time duration; see
782 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
783 for the accepted syntax.
784 If set this option is used instead of @var{start_sample}.
787 Specify the duration of the fade effect. See
788 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
789 for the accepted syntax.
790 At the end of the fade-in effect the output audio will have the same
791 volume as the input audio, at the end of the fade-out transition
792 the output audio will be silence.
793 By default the duration is determined by @var{nb_samples}.
794 If set this option is used instead of @var{nb_samples}.
797 Set curve for fade transition.
799 It accepts the following values:
802 select triangular, linear slope (default)
804 select quarter of sine wave
806 select half of sine wave
808 select exponential sine wave
812 select inverted parabola
826 select inverted quarter of sine wave
828 select inverted half of sine wave
830 select double-exponential seat
832 select double-exponential sigmoid
840 Fade in first 15 seconds of audio:
846 Fade out last 25 seconds of a 900 seconds audio:
848 afade=t=out:st=875:d=25
853 Apply arbitrary expressions to samples in frequency domain.
857 Set frequency domain real expression for each separate channel separated
858 by '|'. Default is "1".
859 If the number of input channels is greater than the number of
860 expressions, the last specified expression is used for the remaining
864 Set frequency domain imaginary expression for each separate channel
865 separated by '|'. If not set, @var{real} option is used.
867 Each expression in @var{real} and @var{imag} can contain the following
875 current frequency bin number
878 number of available bins
881 channel number of the current expression
893 It accepts the following values:
909 Default is @code{w4096}
912 Set window function. Default is @code{hann}.
915 Set window overlap. If set to 1, the recommended overlap for selected
916 window function will be picked. Default is @code{0.75}.
923 Leave almost only low frequencies in audio:
925 afftfilt="1-clip((b/nb)*b,0,1)"
931 Apply an arbitrary Frequency Impulse Response filter.
933 This filter is designed for applying long FIR filters,
934 up to 30 seconds long.
936 It can be used as component for digital crossover filters,
937 room equalization, cross talk cancellation, wavefield synthesis,
938 auralization, ambiophonics and ambisonics.
940 This filter uses second stream as FIR coefficients.
941 If second stream holds single channel, it will be used
942 for all input channels in first stream, otherwise
943 number of channels in second stream must be same as
944 number of channels in first stream.
946 It accepts the following parameters:
950 Set dry gain. This sets input gain.
953 Set wet gain. This sets final output gain.
956 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
959 Enable applying gain measured from power of IR.
966 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
968 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
975 Set output format constraints for the input audio. The framework will
976 negotiate the most appropriate format to minimize conversions.
978 It accepts the following parameters:
982 A '|'-separated list of requested sample formats.
985 A '|'-separated list of requested sample rates.
987 @item channel_layouts
988 A '|'-separated list of requested channel layouts.
990 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
991 for the required syntax.
994 If a parameter is omitted, all values are allowed.
996 Force the output to either unsigned 8-bit or signed 16-bit stereo
998 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1003 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1004 processing reduces disturbing noise between useful signals.
1006 Gating is done by detecting the volume below a chosen level @var{threshold}
1007 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1008 floor is set via @var{range}. Because an exact manipulation of the signal
1009 would cause distortion of the waveform the reduction can be levelled over
1010 time. This is done by setting @var{attack} and @var{release}.
1012 @var{attack} determines how long the signal has to fall below the threshold
1013 before any reduction will occur and @var{release} sets the time the signal
1014 has to rise above the threshold to reduce the reduction again.
1015 Shorter signals than the chosen attack time will be left untouched.
1019 Set input level before filtering.
1020 Default is 1. Allowed range is from 0.015625 to 64.
1023 Set the level of gain reduction when the signal is below the threshold.
1024 Default is 0.06125. Allowed range is from 0 to 1.
1027 If a signal rises above this level the gain reduction is released.
1028 Default is 0.125. Allowed range is from 0 to 1.
1031 Set a ratio by which the signal is reduced.
1032 Default is 2. Allowed range is from 1 to 9000.
1035 Amount of milliseconds the signal has to rise above the threshold before gain
1037 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1040 Amount of milliseconds the signal has to fall below the threshold before the
1041 reduction is increased again. Default is 250 milliseconds.
1042 Allowed range is from 0.01 to 9000.
1045 Set amount of amplification of signal after processing.
1046 Default is 1. Allowed range is from 1 to 64.
1049 Curve the sharp knee around the threshold to enter gain reduction more softly.
1050 Default is 2.828427125. Allowed range is from 1 to 8.
1053 Choose if exact signal should be taken for detection or an RMS like one.
1054 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1057 Choose if the average level between all channels or the louder channel affects
1059 Default is @code{average}. Can be @code{average} or @code{maximum}.
1064 The limiter prevents an input signal from rising over a desired threshold.
1065 This limiter uses lookahead technology to prevent your signal from distorting.
1066 It means that there is a small delay after the signal is processed. Keep in mind
1067 that the delay it produces is the attack time you set.
1069 The filter accepts the following options:
1073 Set input gain. Default is 1.
1076 Set output gain. Default is 1.
1079 Don't let signals above this level pass the limiter. Default is 1.
1082 The limiter will reach its attenuation level in this amount of time in
1083 milliseconds. Default is 5 milliseconds.
1086 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1087 Default is 50 milliseconds.
1090 When gain reduction is always needed ASC takes care of releasing to an
1091 average reduction level rather than reaching a reduction of 0 in the release
1095 Select how much the release time is affected by ASC, 0 means nearly no changes
1096 in release time while 1 produces higher release times.
1099 Auto level output signal. Default is enabled.
1100 This normalizes audio back to 0dB if enabled.
1103 Depending on picked setting it is recommended to upsample input 2x or 4x times
1104 with @ref{aresample} before applying this filter.
1108 Apply a two-pole all-pass filter with central frequency (in Hz)
1109 @var{frequency}, and filter-width @var{width}.
1110 An all-pass filter changes the audio's frequency to phase relationship
1111 without changing its frequency to amplitude relationship.
1113 The filter accepts the following options:
1117 Set frequency in Hz.
1120 Set method to specify band-width of filter.
1133 Specify the band-width of a filter in width_type units.
1136 Specify which channels to filter, by default all available are filtered.
1143 The filter accepts the following options:
1147 Set the number of loops. Setting this value to -1 will result in infinite loops.
1151 Set maximal number of samples. Default is 0.
1154 Set first sample of loop. Default is 0.
1160 Merge two or more audio streams into a single multi-channel stream.
1162 The filter accepts the following options:
1167 Set the number of inputs. Default is 2.
1171 If the channel layouts of the inputs are disjoint, and therefore compatible,
1172 the channel layout of the output will be set accordingly and the channels
1173 will be reordered as necessary. If the channel layouts of the inputs are not
1174 disjoint, the output will have all the channels of the first input then all
1175 the channels of the second input, in that order, and the channel layout of
1176 the output will be the default value corresponding to the total number of
1179 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1180 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1181 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1182 first input, b1 is the first channel of the second input).
1184 On the other hand, if both input are in stereo, the output channels will be
1185 in the default order: a1, a2, b1, b2, and the channel layout will be
1186 arbitrarily set to 4.0, which may or may not be the expected value.
1188 All inputs must have the same sample rate, and format.
1190 If inputs do not have the same duration, the output will stop with the
1193 @subsection Examples
1197 Merge two mono files into a stereo stream:
1199 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1203 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1205 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
1211 Mixes multiple audio inputs into a single output.
1213 Note that this filter only supports float samples (the @var{amerge}
1214 and @var{pan} audio filters support many formats). If the @var{amix}
1215 input has integer samples then @ref{aresample} will be automatically
1216 inserted to perform the conversion to float samples.
1220 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1222 will mix 3 input audio streams to a single output with the same duration as the
1223 first input and a dropout transition time of 3 seconds.
1225 It accepts the following parameters:
1229 The number of inputs. If unspecified, it defaults to 2.
1232 How to determine the end-of-stream.
1236 The duration of the longest input. (default)
1239 The duration of the shortest input.
1242 The duration of the first input.
1246 @item dropout_transition
1247 The transition time, in seconds, for volume renormalization when an input
1248 stream ends. The default value is 2 seconds.
1252 @section anequalizer
1254 High-order parametric multiband equalizer for each channel.
1256 It accepts the following parameters:
1260 This option string is in format:
1261 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1262 Each equalizer band is separated by '|'.
1266 Set channel number to which equalization will be applied.
1267 If input doesn't have that channel the entry is ignored.
1270 Set central frequency for band.
1271 If input doesn't have that frequency the entry is ignored.
1274 Set band width in hertz.
1277 Set band gain in dB.
1280 Set filter type for band, optional, can be:
1284 Butterworth, this is default.
1295 With this option activated frequency response of anequalizer is displayed
1299 Set video stream size. Only useful if curves option is activated.
1302 Set max gain that will be displayed. Only useful if curves option is activated.
1303 Setting this to a reasonable value makes it possible to display gain which is derived from
1304 neighbour bands which are too close to each other and thus produce higher gain
1305 when both are activated.
1308 Set frequency scale used to draw frequency response in video output.
1309 Can be linear or logarithmic. Default is logarithmic.
1312 Set color for each channel curve which is going to be displayed in video stream.
1313 This is list of color names separated by space or by '|'.
1314 Unrecognised or missing colors will be replaced by white color.
1317 @subsection Examples
1321 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1322 for first 2 channels using Chebyshev type 1 filter:
1324 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1328 @subsection Commands
1330 This filter supports the following commands:
1333 Alter existing filter parameters.
1334 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1336 @var{fN} is existing filter number, starting from 0, if no such filter is available
1338 @var{freq} set new frequency parameter.
1339 @var{width} set new width parameter in herz.
1340 @var{gain} set new gain parameter in dB.
1342 Full filter invocation with asendcmd may look like this:
1343 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1348 Pass the audio source unchanged to the output.
1352 Pad the end of an audio stream with silence.
1354 This can be used together with @command{ffmpeg} @option{-shortest} to
1355 extend audio streams to the same length as the video stream.
1357 A description of the accepted options follows.
1361 Set silence packet size. Default value is 4096.
1364 Set the number of samples of silence to add to the end. After the
1365 value is reached, the stream is terminated. This option is mutually
1366 exclusive with @option{whole_len}.
1369 Set the minimum total number of samples in the output audio stream. If
1370 the value is longer than the input audio length, silence is added to
1371 the end, until the value is reached. This option is mutually exclusive
1372 with @option{pad_len}.
1375 If neither the @option{pad_len} nor the @option{whole_len} option is
1376 set, the filter will add silence to the end of the input stream
1379 @subsection Examples
1383 Add 1024 samples of silence to the end of the input:
1389 Make sure the audio output will contain at least 10000 samples, pad
1390 the input with silence if required:
1392 apad=whole_len=10000
1396 Use @command{ffmpeg} to pad the audio input with silence, so that the
1397 video stream will always result the shortest and will be converted
1398 until the end in the output file when using the @option{shortest}
1401 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1406 Add a phasing effect to the input audio.
1408 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1409 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1411 A description of the accepted parameters follows.
1415 Set input gain. Default is 0.4.
1418 Set output gain. Default is 0.74
1421 Set delay in milliseconds. Default is 3.0.
1424 Set decay. Default is 0.4.
1427 Set modulation speed in Hz. Default is 0.5.
1430 Set modulation type. Default is triangular.
1432 It accepts the following values:
1441 Audio pulsator is something between an autopanner and a tremolo.
1442 But it can produce funny stereo effects as well. Pulsator changes the volume
1443 of the left and right channel based on a LFO (low frequency oscillator) with
1444 different waveforms and shifted phases.
1445 This filter have the ability to define an offset between left and right
1446 channel. An offset of 0 means that both LFO shapes match each other.
1447 The left and right channel are altered equally - a conventional tremolo.
1448 An offset of 50% means that the shape of the right channel is exactly shifted
1449 in phase (or moved backwards about half of the frequency) - pulsator acts as
1450 an autopanner. At 1 both curves match again. Every setting in between moves the
1451 phase shift gapless between all stages and produces some "bypassing" sounds with
1452 sine and triangle waveforms. The more you set the offset near 1 (starting from
1453 the 0.5) the faster the signal passes from the left to the right speaker.
1455 The filter accepts the following options:
1459 Set input gain. By default it is 1. Range is [0.015625 - 64].
1462 Set output gain. By default it is 1. Range is [0.015625 - 64].
1465 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1466 sawup or sawdown. Default is sine.
1469 Set modulation. Define how much of original signal is affected by the LFO.
1472 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1475 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1478 Set pulse width. Default is 1. Allowed range is [0 - 2].
1481 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1484 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1488 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1492 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1493 if timing is set to hz.
1499 Resample the input audio to the specified parameters, using the
1500 libswresample library. If none are specified then the filter will
1501 automatically convert between its input and output.
1503 This filter is also able to stretch/squeeze the audio data to make it match
1504 the timestamps or to inject silence / cut out audio to make it match the
1505 timestamps, do a combination of both or do neither.
1507 The filter accepts the syntax
1508 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1509 expresses a sample rate and @var{resampler_options} is a list of
1510 @var{key}=@var{value} pairs, separated by ":". See the
1511 @ref{Resampler Options,,the "Resampler Options" section in the
1512 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1513 for the complete list of supported options.
1515 @subsection Examples
1519 Resample the input audio to 44100Hz:
1525 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1526 samples per second compensation:
1528 aresample=async=1000
1534 Reverse an audio clip.
1536 Warning: This filter requires memory to buffer the entire clip, so trimming
1539 @subsection Examples
1543 Take the first 5 seconds of a clip, and reverse it.
1545 atrim=end=5,areverse
1549 @section asetnsamples
1551 Set the number of samples per each output audio frame.
1553 The last output packet may contain a different number of samples, as
1554 the filter will flush all the remaining samples when the input audio
1557 The filter accepts the following options:
1561 @item nb_out_samples, n
1562 Set the number of frames per each output audio frame. The number is
1563 intended as the number of samples @emph{per each channel}.
1564 Default value is 1024.
1567 If set to 1, the filter will pad the last audio frame with zeroes, so
1568 that the last frame will contain the same number of samples as the
1569 previous ones. Default value is 1.
1572 For example, to set the number of per-frame samples to 1234 and
1573 disable padding for the last frame, use:
1575 asetnsamples=n=1234:p=0
1580 Set the sample rate without altering the PCM data.
1581 This will result in a change of speed and pitch.
1583 The filter accepts the following options:
1586 @item sample_rate, r
1587 Set the output sample rate. Default is 44100 Hz.
1592 Show a line containing various information for each input audio frame.
1593 The input audio is not modified.
1595 The shown line contains a sequence of key/value pairs of the form
1596 @var{key}:@var{value}.
1598 The following values are shown in the output:
1602 The (sequential) number of the input frame, starting from 0.
1605 The presentation timestamp of the input frame, in time base units; the time base
1606 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1609 The presentation timestamp of the input frame in seconds.
1612 position of the frame in the input stream, -1 if this information in
1613 unavailable and/or meaningless (for example in case of synthetic audio)
1622 The sample rate for the audio frame.
1625 The number of samples (per channel) in the frame.
1628 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1629 audio, the data is treated as if all the planes were concatenated.
1631 @item plane_checksums
1632 A list of Adler-32 checksums for each data plane.
1638 Display time domain statistical information about the audio channels.
1639 Statistics are calculated and displayed for each audio channel and,
1640 where applicable, an overall figure is also given.
1642 It accepts the following option:
1645 Short window length in seconds, used for peak and trough RMS measurement.
1646 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1650 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1651 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1654 Available keys for each channel are:
1688 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1689 this @code{lavfi.astats.Overall.Peak_count}.
1691 For description what each key means read below.
1694 Set number of frame after which stats are going to be recalculated.
1695 Default is disabled.
1698 A description of each shown parameter follows:
1702 Mean amplitude displacement from zero.
1705 Minimal sample level.
1708 Maximal sample level.
1710 @item Min difference
1711 Minimal difference between two consecutive samples.
1713 @item Max difference
1714 Maximal difference between two consecutive samples.
1716 @item Mean difference
1717 Mean difference between two consecutive samples.
1718 The average of each difference between two consecutive samples.
1720 @item RMS difference
1721 Root Mean Square difference between two consecutive samples.
1725 Standard peak and RMS level measured in dBFS.
1729 Peak and trough values for RMS level measured over a short window.
1732 Standard ratio of peak to RMS level (note: not in dB).
1735 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1736 (i.e. either @var{Min level} or @var{Max level}).
1739 Number of occasions (not the number of samples) that the signal attained either
1740 @var{Min level} or @var{Max level}.
1743 Overall bit depth of audio. Number of bits used for each sample.
1746 Measured dynamic range of audio in dB.
1753 The filter accepts exactly one parameter, the audio tempo. If not
1754 specified then the filter will assume nominal 1.0 tempo. Tempo must
1755 be in the [0.5, 2.0] range.
1757 @subsection Examples
1761 Slow down audio to 80% tempo:
1767 To speed up audio to 125% tempo:
1775 Trim the input so that the output contains one continuous subpart of the input.
1777 It accepts the following parameters:
1780 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1781 sample with the timestamp @var{start} will be the first sample in the output.
1784 Specify time of the first audio sample that will be dropped, i.e. the
1785 audio sample immediately preceding the one with the timestamp @var{end} will be
1786 the last sample in the output.
1789 Same as @var{start}, except this option sets the start timestamp in samples
1793 Same as @var{end}, except this option sets the end timestamp in samples instead
1797 The maximum duration of the output in seconds.
1800 The number of the first sample that should be output.
1803 The number of the first sample that should be dropped.
1806 @option{start}, @option{end}, and @option{duration} are expressed as time
1807 duration specifications; see
1808 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1810 Note that the first two sets of the start/end options and the @option{duration}
1811 option look at the frame timestamp, while the _sample options simply count the
1812 samples that pass through the filter. So start/end_pts and start/end_sample will
1813 give different results when the timestamps are wrong, inexact or do not start at
1814 zero. Also note that this filter does not modify the timestamps. If you wish
1815 to have the output timestamps start at zero, insert the asetpts filter after the
1818 If multiple start or end options are set, this filter tries to be greedy and
1819 keep all samples that match at least one of the specified constraints. To keep
1820 only the part that matches all the constraints at once, chain multiple atrim
1823 The defaults are such that all the input is kept. So it is possible to set e.g.
1824 just the end values to keep everything before the specified time.
1829 Drop everything except the second minute of input:
1831 ffmpeg -i INPUT -af atrim=60:120
1835 Keep only the first 1000 samples:
1837 ffmpeg -i INPUT -af atrim=end_sample=1000
1844 Apply a two-pole Butterworth band-pass filter with central
1845 frequency @var{frequency}, and (3dB-point) band-width width.
1846 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1847 instead of the default: constant 0dB peak gain.
1848 The filter roll off at 6dB per octave (20dB per decade).
1850 The filter accepts the following options:
1854 Set the filter's central frequency. Default is @code{3000}.
1857 Constant skirt gain if set to 1. Defaults to 0.
1860 Set method to specify band-width of filter.
1873 Specify the band-width of a filter in width_type units.
1876 Specify which channels to filter, by default all available are filtered.
1881 Apply a two-pole Butterworth band-reject filter with central
1882 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1883 The filter roll off at 6dB per octave (20dB per decade).
1885 The filter accepts the following options:
1889 Set the filter's central frequency. Default is @code{3000}.
1892 Set method to specify band-width of filter.
1905 Specify the band-width of a filter in width_type units.
1908 Specify which channels to filter, by default all available are filtered.
1913 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1914 shelving filter with a response similar to that of a standard
1915 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1917 The filter accepts the following options:
1921 Give the gain at 0 Hz. Its useful range is about -20
1922 (for a large cut) to +20 (for a large boost).
1923 Beware of clipping when using a positive gain.
1926 Set the filter's central frequency and so can be used
1927 to extend or reduce the frequency range to be boosted or cut.
1928 The default value is @code{100} Hz.
1931 Set method to specify band-width of filter.
1944 Determine how steep is the filter's shelf transition.
1947 Specify which channels to filter, by default all available are filtered.
1952 Apply a biquad IIR filter with the given coefficients.
1953 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1954 are the numerator and denominator coefficients respectively.
1955 and @var{channels}, @var{c} specify which channels to filter, by default all
1956 available are filtered.
1959 Bauer stereo to binaural transformation, which improves headphone listening of
1960 stereo audio records.
1962 To enable compilation of this filter you need to configure FFmpeg with
1963 @code{--enable-libbs2b}.
1965 It accepts the following parameters:
1969 Pre-defined crossfeed level.
1973 Default level (fcut=700, feed=50).
1976 Chu Moy circuit (fcut=700, feed=60).
1979 Jan Meier circuit (fcut=650, feed=95).
1984 Cut frequency (in Hz).
1993 Remap input channels to new locations.
1995 It accepts the following parameters:
1998 Map channels from input to output. The argument is a '|'-separated list of
1999 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2000 @var{in_channel} form. @var{in_channel} can be either the name of the input
2001 channel (e.g. FL for front left) or its index in the input channel layout.
2002 @var{out_channel} is the name of the output channel or its index in the output
2003 channel layout. If @var{out_channel} is not given then it is implicitly an
2004 index, starting with zero and increasing by one for each mapping.
2006 @item channel_layout
2007 The channel layout of the output stream.
2010 If no mapping is present, the filter will implicitly map input channels to
2011 output channels, preserving indices.
2013 For example, assuming a 5.1+downmix input MOV file,
2015 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2017 will create an output WAV file tagged as stereo from the downmix channels of
2020 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2022 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2025 @section channelsplit
2027 Split each channel from an input audio stream into a separate output stream.
2029 It accepts the following parameters:
2031 @item channel_layout
2032 The channel layout of the input stream. The default is "stereo".
2035 For example, assuming a stereo input MP3 file,
2037 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2039 will create an output Matroska file with two audio streams, one containing only
2040 the left channel and the other the right channel.
2042 Split a 5.1 WAV file into per-channel files:
2044 ffmpeg -i in.wav -filter_complex
2045 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2046 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2047 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2052 Add a chorus effect to the audio.
2054 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2056 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2057 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2058 The modulation depth defines the range the modulated delay is played before or after
2059 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2060 sound tuned around the original one, like in a chorus where some vocals are slightly
2063 It accepts the following parameters:
2066 Set input gain. Default is 0.4.
2069 Set output gain. Default is 0.4.
2072 Set delays. A typical delay is around 40ms to 60ms.
2084 @subsection Examples
2090 chorus=0.7:0.9:55:0.4:0.25:2
2096 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2100 Fuller sounding chorus with three delays:
2102 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
2107 Compress or expand the audio's dynamic range.
2109 It accepts the following parameters:
2115 A list of times in seconds for each channel over which the instantaneous level
2116 of the input signal is averaged to determine its volume. @var{attacks} refers to
2117 increase of volume and @var{decays} refers to decrease of volume. For most
2118 situations, the attack time (response to the audio getting louder) should be
2119 shorter than the decay time, because the human ear is more sensitive to sudden
2120 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2121 a typical value for decay is 0.8 seconds.
2122 If specified number of attacks & decays is lower than number of channels, the last
2123 set attack/decay will be used for all remaining channels.
2126 A list of points for the transfer function, specified in dB relative to the
2127 maximum possible signal amplitude. Each key points list must be defined using
2128 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2129 @code{x0/y0 x1/y1 x2/y2 ....}
2131 The input values must be in strictly increasing order but the transfer function
2132 does not have to be monotonically rising. The point @code{0/0} is assumed but
2133 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2134 function are @code{-70/-70|-60/-20|1/0}.
2137 Set the curve radius in dB for all joints. It defaults to 0.01.
2140 Set the additional gain in dB to be applied at all points on the transfer
2141 function. This allows for easy adjustment of the overall gain.
2145 Set an initial volume, in dB, to be assumed for each channel when filtering
2146 starts. This permits the user to supply a nominal level initially, so that, for
2147 example, a very large gain is not applied to initial signal levels before the
2148 companding has begun to operate. A typical value for audio which is initially
2149 quiet is -90 dB. It defaults to 0.
2152 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2153 delayed before being fed to the volume adjuster. Specifying a delay
2154 approximately equal to the attack/decay times allows the filter to effectively
2155 operate in predictive rather than reactive mode. It defaults to 0.
2159 @subsection Examples
2163 Make music with both quiet and loud passages suitable for listening to in a
2166 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2169 Another example for audio with whisper and explosion parts:
2171 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2175 A noise gate for when the noise is at a lower level than the signal:
2177 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2181 Here is another noise gate, this time for when the noise is at a higher level
2182 than the signal (making it, in some ways, similar to squelch):
2184 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2188 2:1 compression starting at -6dB:
2190 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2194 2:1 compression starting at -9dB:
2196 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2200 2:1 compression starting at -12dB:
2202 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2206 2:1 compression starting at -18dB:
2208 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2212 3:1 compression starting at -15dB:
2214 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2220 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2226 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
2230 Hard limiter at -6dB:
2232 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2236 Hard limiter at -12dB:
2238 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2242 Hard noise gate at -35 dB:
2244 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2250 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2254 @section compensationdelay
2256 Compensation Delay Line is a metric based delay to compensate differing
2257 positions of microphones or speakers.
2259 For example, you have recorded guitar with two microphones placed in
2260 different location. Because the front of sound wave has fixed speed in
2261 normal conditions, the phasing of microphones can vary and depends on
2262 their location and interposition. The best sound mix can be achieved when
2263 these microphones are in phase (synchronized). Note that distance of
2264 ~30 cm between microphones makes one microphone to capture signal in
2265 antiphase to another microphone. That makes the final mix sounding moody.
2266 This filter helps to solve phasing problems by adding different delays
2267 to each microphone track and make them synchronized.
2269 The best result can be reached when you take one track as base and
2270 synchronize other tracks one by one with it.
2271 Remember that synchronization/delay tolerance depends on sample rate, too.
2272 Higher sample rates will give more tolerance.
2274 It accepts the following parameters:
2278 Set millimeters distance. This is compensation distance for fine tuning.
2282 Set cm distance. This is compensation distance for tightening distance setup.
2286 Set meters distance. This is compensation distance for hard distance setup.
2290 Set dry amount. Amount of unprocessed (dry) signal.
2294 Set wet amount. Amount of processed (wet) signal.
2298 Set temperature degree in Celsius. This is the temperature of the environment.
2303 Apply headphone crossfeed filter.
2305 Crossfeed is the process of blending the left and right channels of stereo
2307 It is mainly used to reduce extreme stereo separation of low frequencies.
2309 The intent is to produce more speaker like sound to the listener.
2311 The filter accepts the following options:
2315 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2316 This sets gain of low shelf filter for side part of stereo image.
2317 Default is -6dB. Max allowed is -30db when strength is set to 1.
2320 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2321 This sets cut off frequency of low shelf filter. Default is cut off near
2322 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2325 Set input gain. Default is 0.9.
2328 Set output gain. Default is 1.
2331 @section crystalizer
2332 Simple algorithm to expand audio dynamic range.
2334 The filter accepts the following options:
2338 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2339 (unchanged sound) to 10.0 (maximum effect).
2342 Enable clipping. By default is enabled.
2346 Apply a DC shift to the audio.
2348 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2349 in the recording chain) from the audio. The effect of a DC offset is reduced
2350 headroom and hence volume. The @ref{astats} filter can be used to determine if
2351 a signal has a DC offset.
2355 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2359 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2360 used to prevent clipping.
2364 Dynamic Audio Normalizer.
2366 This filter applies a certain amount of gain to the input audio in order
2367 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2368 contrast to more "simple" normalization algorithms, the Dynamic Audio
2369 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2370 This allows for applying extra gain to the "quiet" sections of the audio
2371 while avoiding distortions or clipping the "loud" sections. In other words:
2372 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2373 sections, in the sense that the volume of each section is brought to the
2374 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2375 this goal *without* applying "dynamic range compressing". It will retain 100%
2376 of the dynamic range *within* each section of the audio file.
2380 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2381 Default is 500 milliseconds.
2382 The Dynamic Audio Normalizer processes the input audio in small chunks,
2383 referred to as frames. This is required, because a peak magnitude has no
2384 meaning for just a single sample value. Instead, we need to determine the
2385 peak magnitude for a contiguous sequence of sample values. While a "standard"
2386 normalizer would simply use the peak magnitude of the complete file, the
2387 Dynamic Audio Normalizer determines the peak magnitude individually for each
2388 frame. The length of a frame is specified in milliseconds. By default, the
2389 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2390 been found to give good results with most files.
2391 Note that the exact frame length, in number of samples, will be determined
2392 automatically, based on the sampling rate of the individual input audio file.
2395 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2396 number. Default is 31.
2397 Probably the most important parameter of the Dynamic Audio Normalizer is the
2398 @code{window size} of the Gaussian smoothing filter. The filter's window size
2399 is specified in frames, centered around the current frame. For the sake of
2400 simplicity, this must be an odd number. Consequently, the default value of 31
2401 takes into account the current frame, as well as the 15 preceding frames and
2402 the 15 subsequent frames. Using a larger window results in a stronger
2403 smoothing effect and thus in less gain variation, i.e. slower gain
2404 adaptation. Conversely, using a smaller window results in a weaker smoothing
2405 effect and thus in more gain variation, i.e. faster gain adaptation.
2406 In other words, the more you increase this value, the more the Dynamic Audio
2407 Normalizer will behave like a "traditional" normalization filter. On the
2408 contrary, the more you decrease this value, the more the Dynamic Audio
2409 Normalizer will behave like a dynamic range compressor.
2412 Set the target peak value. This specifies the highest permissible magnitude
2413 level for the normalized audio input. This filter will try to approach the
2414 target peak magnitude as closely as possible, but at the same time it also
2415 makes sure that the normalized signal will never exceed the peak magnitude.
2416 A frame's maximum local gain factor is imposed directly by the target peak
2417 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2418 It is not recommended to go above this value.
2421 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2422 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2423 factor for each input frame, i.e. the maximum gain factor that does not
2424 result in clipping or distortion. The maximum gain factor is determined by
2425 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2426 additionally bounds the frame's maximum gain factor by a predetermined
2427 (global) maximum gain factor. This is done in order to avoid excessive gain
2428 factors in "silent" or almost silent frames. By default, the maximum gain
2429 factor is 10.0, For most inputs the default value should be sufficient and
2430 it usually is not recommended to increase this value. Though, for input
2431 with an extremely low overall volume level, it may be necessary to allow even
2432 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2433 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2434 Instead, a "sigmoid" threshold function will be applied. This way, the
2435 gain factors will smoothly approach the threshold value, but never exceed that
2439 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2440 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2441 This means that the maximum local gain factor for each frame is defined
2442 (only) by the frame's highest magnitude sample. This way, the samples can
2443 be amplified as much as possible without exceeding the maximum signal
2444 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2445 Normalizer can also take into account the frame's root mean square,
2446 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2447 determine the power of a time-varying signal. It is therefore considered
2448 that the RMS is a better approximation of the "perceived loudness" than
2449 just looking at the signal's peak magnitude. Consequently, by adjusting all
2450 frames to a constant RMS value, a uniform "perceived loudness" can be
2451 established. If a target RMS value has been specified, a frame's local gain
2452 factor is defined as the factor that would result in exactly that RMS value.
2453 Note, however, that the maximum local gain factor is still restricted by the
2454 frame's highest magnitude sample, in order to prevent clipping.
2457 Enable channels coupling. By default is enabled.
2458 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2459 amount. This means the same gain factor will be applied to all channels, i.e.
2460 the maximum possible gain factor is determined by the "loudest" channel.
2461 However, in some recordings, it may happen that the volume of the different
2462 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2463 In this case, this option can be used to disable the channel coupling. This way,
2464 the gain factor will be determined independently for each channel, depending
2465 only on the individual channel's highest magnitude sample. This allows for
2466 harmonizing the volume of the different channels.
2469 Enable DC bias correction. By default is disabled.
2470 An audio signal (in the time domain) is a sequence of sample values.
2471 In the Dynamic Audio Normalizer these sample values are represented in the
2472 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2473 audio signal, or "waveform", should be centered around the zero point.
2474 That means if we calculate the mean value of all samples in a file, or in a
2475 single frame, then the result should be 0.0 or at least very close to that
2476 value. If, however, there is a significant deviation of the mean value from
2477 0.0, in either positive or negative direction, this is referred to as a
2478 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2479 Audio Normalizer provides optional DC bias correction.
2480 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2481 the mean value, or "DC correction" offset, of each input frame and subtract
2482 that value from all of the frame's sample values which ensures those samples
2483 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2484 boundaries, the DC correction offset values will be interpolated smoothly
2485 between neighbouring frames.
2488 Enable alternative boundary mode. By default is disabled.
2489 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2490 around each frame. This includes the preceding frames as well as the
2491 subsequent frames. However, for the "boundary" frames, located at the very
2492 beginning and at the very end of the audio file, not all neighbouring
2493 frames are available. In particular, for the first few frames in the audio
2494 file, the preceding frames are not known. And, similarly, for the last few
2495 frames in the audio file, the subsequent frames are not known. Thus, the
2496 question arises which gain factors should be assumed for the missing frames
2497 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2498 to deal with this situation. The default boundary mode assumes a gain factor
2499 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2500 "fade out" at the beginning and at the end of the input, respectively.
2503 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2504 By default, the Dynamic Audio Normalizer does not apply "traditional"
2505 compression. This means that signal peaks will not be pruned and thus the
2506 full dynamic range will be retained within each local neighbourhood. However,
2507 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2508 normalization algorithm with a more "traditional" compression.
2509 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2510 (thresholding) function. If (and only if) the compression feature is enabled,
2511 all input frames will be processed by a soft knee thresholding function prior
2512 to the actual normalization process. Put simply, the thresholding function is
2513 going to prune all samples whose magnitude exceeds a certain threshold value.
2514 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2515 value. Instead, the threshold value will be adjusted for each individual
2517 In general, smaller parameters result in stronger compression, and vice versa.
2518 Values below 3.0 are not recommended, because audible distortion may appear.
2523 Make audio easier to listen to on headphones.
2525 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2526 so that when listened to on headphones the stereo image is moved from
2527 inside your head (standard for headphones) to outside and in front of
2528 the listener (standard for speakers).
2534 Apply a two-pole peaking equalisation (EQ) filter. With this
2535 filter, the signal-level at and around a selected frequency can
2536 be increased or decreased, whilst (unlike bandpass and bandreject
2537 filters) that at all other frequencies is unchanged.
2539 In order to produce complex equalisation curves, this filter can
2540 be given several times, each with a different central frequency.
2542 The filter accepts the following options:
2546 Set the filter's central frequency in Hz.
2549 Set method to specify band-width of filter.
2562 Specify the band-width of a filter in width_type units.
2565 Set the required gain or attenuation in dB.
2566 Beware of clipping when using a positive gain.
2569 Specify which channels to filter, by default all available are filtered.
2572 @subsection Examples
2575 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2577 equalizer=f=1000:t=h:width=200:g=-10
2581 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2583 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2587 @section extrastereo
2589 Linearly increases the difference between left and right channels which
2590 adds some sort of "live" effect to playback.
2592 The filter accepts the following options:
2596 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2597 (average of both channels), with 1.0 sound will be unchanged, with
2598 -1.0 left and right channels will be swapped.
2601 Enable clipping. By default is enabled.
2604 @section firequalizer
2605 Apply FIR Equalization using arbitrary frequency response.
2607 The filter accepts the following option:
2611 Set gain curve equation (in dB). The expression can contain variables:
2614 the evaluated frequency
2618 channel number, set to 0 when multichannels evaluation is disabled
2620 channel id, see libavutil/channel_layout.h, set to the first channel id when
2621 multichannels evaluation is disabled
2625 channel_layout, see libavutil/channel_layout.h
2630 @item gain_interpolate(f)
2631 interpolate gain on frequency f based on gain_entry
2632 @item cubic_interpolate(f)
2633 same as gain_interpolate, but smoother
2635 This option is also available as command. Default is @code{gain_interpolate(f)}.
2638 Set gain entry for gain_interpolate function. The expression can
2642 store gain entry at frequency f with value g
2644 This option is also available as command.
2647 Set filter delay in seconds. Higher value means more accurate.
2648 Default is @code{0.01}.
2651 Set filter accuracy in Hz. Lower value means more accurate.
2652 Default is @code{5}.
2655 Set window function. Acceptable values are:
2658 rectangular window, useful when gain curve is already smooth
2660 hann window (default)
2666 3-terms continuous 1st derivative nuttall window
2668 minimum 3-terms discontinuous nuttall window
2670 4-terms continuous 1st derivative nuttall window
2672 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2674 blackman-harris window
2680 If enabled, use fixed number of audio samples. This improves speed when
2681 filtering with large delay. Default is disabled.
2684 Enable multichannels evaluation on gain. Default is disabled.
2687 Enable zero phase mode by subtracting timestamp to compensate delay.
2688 Default is disabled.
2691 Set scale used by gain. Acceptable values are:
2694 linear frequency, linear gain
2696 linear frequency, logarithmic (in dB) gain (default)
2698 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2700 logarithmic frequency, logarithmic gain
2704 Set file for dumping, suitable for gnuplot.
2707 Set scale for dumpfile. Acceptable values are same with scale option.
2711 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2712 Default is disabled.
2715 Enable minimum phase impulse response. Default is disabled.
2718 @subsection Examples
2723 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2726 lowpass at 1000 Hz with gain_entry:
2728 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2731 custom equalization:
2733 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2736 higher delay with zero phase to compensate delay:
2738 firequalizer=delay=0.1:fixed=on:zero_phase=on
2741 lowpass on left channel, highpass on right channel:
2743 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2744 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2749 Apply a flanging effect to the audio.
2751 The filter accepts the following options:
2755 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2758 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
2761 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2765 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2766 Default value is 71.
2769 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2772 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2773 Default value is @var{sinusoidal}.
2776 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2777 Default value is 25.
2780 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2781 Default is @var{linear}.
2785 Apply Haas effect to audio.
2787 Note that this makes most sense to apply on mono signals.
2788 With this filter applied to mono signals it give some directionality and
2789 stretches its stereo image.
2791 The filter accepts the following options:
2795 Set input level. By default is @var{1}, or 0dB
2798 Set output level. By default is @var{1}, or 0dB.
2801 Set gain applied to side part of signal. By default is @var{1}.
2804 Set kind of middle source. Can be one of the following:
2814 Pick middle part signal of stereo image.
2817 Pick side part signal of stereo image.
2821 Change middle phase. By default is disabled.
2824 Set left channel delay. By default is @var{2.05} milliseconds.
2827 Set left channel balance. By default is @var{-1}.
2830 Set left channel gain. By default is @var{1}.
2833 Change left phase. By default is disabled.
2836 Set right channel delay. By defaults is @var{2.12} milliseconds.
2839 Set right channel balance. By default is @var{1}.
2842 Set right channel gain. By default is @var{1}.
2845 Change right phase. By default is enabled.
2850 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2851 embedded HDCD codes is expanded into a 20-bit PCM stream.
2853 The filter supports the Peak Extend and Low-level Gain Adjustment features
2854 of HDCD, and detects the Transient Filter flag.
2857 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2860 When using the filter with wav, note the default encoding for wav is 16-bit,
2861 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2862 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2864 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2865 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
2868 The filter accepts the following options:
2871 @item disable_autoconvert
2872 Disable any automatic format conversion or resampling in the filter graph.
2874 @item process_stereo
2875 Process the stereo channels together. If target_gain does not match between
2876 channels, consider it invalid and use the last valid target_gain.
2879 Set the code detect timer period in ms.
2882 Always extend peaks above -3dBFS even if PE isn't signaled.
2885 Replace audio with a solid tone and adjust the amplitude to signal some
2886 specific aspect of the decoding process. The output file can be loaded in
2887 an audio editor alongside the original to aid analysis.
2889 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2896 Gain adjustment level at each sample
2898 Samples where peak extend occurs
2900 Samples where the code detect timer is active
2902 Samples where the target gain does not match between channels
2908 Apply head-related transfer functions (HRTFs) to create virtual
2909 loudspeakers around the user for binaural listening via headphones.
2910 The HRIRs are provided via additional streams, for each channel
2911 one stereo input stream is needed.
2913 The filter accepts the following options:
2917 Set mapping of input streams for convolution.
2918 The argument is a '|'-separated list of channel names in order as they
2919 are given as additional stream inputs for filter.
2920 This also specify number of input streams. Number of input streams
2921 must be not less than number of channels in first stream plus one.
2924 Set gain applied to audio. Value is in dB. Default is 0.
2927 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
2928 processing audio in time domain which is slow.
2929 @var{freq} is processing audio in frequency domain which is fast.
2930 Default is @var{freq}.
2933 Set custom gain for LFE channels. Value is in dB. Default is 0.
2936 @subsection Examples
2940 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
2941 each amovie filter use stereo file with IR coefficients as input.
2942 The files give coefficients for each position of virtual loudspeaker:
2944 ffmpeg -i input.wav -lavfi-complex "amovie=azi_270_ele_0_DFC.wav[sr],amovie=azi_90_ele_0_DFC.wav[sl],amovie=azi_225_ele_0_DFC.wav[br],amovie=azi_135_ele_0_DFC.wav[bl],amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe],amovie=azi_35_ele_0_DFC.wav[fl],amovie=azi_325_ele_0_DFC.wav[fr],[a:0][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
2951 Apply a high-pass filter with 3dB point frequency.
2952 The filter can be either single-pole, or double-pole (the default).
2953 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2955 The filter accepts the following options:
2959 Set frequency in Hz. Default is 3000.
2962 Set number of poles. Default is 2.
2965 Set method to specify band-width of filter.
2978 Specify the band-width of a filter in width_type units.
2979 Applies only to double-pole filter.
2980 The default is 0.707q and gives a Butterworth response.
2983 Specify which channels to filter, by default all available are filtered.
2988 Join multiple input streams into one multi-channel stream.
2990 It accepts the following parameters:
2994 The number of input streams. It defaults to 2.
2996 @item channel_layout
2997 The desired output channel layout. It defaults to stereo.
3000 Map channels from inputs to output. The argument is a '|'-separated list of
3001 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3002 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3003 can be either the name of the input channel (e.g. FL for front left) or its
3004 index in the specified input stream. @var{out_channel} is the name of the output
3008 The filter will attempt to guess the mappings when they are not specified
3009 explicitly. It does so by first trying to find an unused matching input channel
3010 and if that fails it picks the first unused input channel.
3012 Join 3 inputs (with properly set channel layouts):
3014 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3017 Build a 5.1 output from 6 single-channel streams:
3019 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3020 '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'
3026 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3028 To enable compilation of this filter you need to configure FFmpeg with
3029 @code{--enable-ladspa}.
3033 Specifies the name of LADSPA plugin library to load. If the environment
3034 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3035 each one of the directories specified by the colon separated list in
3036 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3037 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3038 @file{/usr/lib/ladspa/}.
3041 Specifies the plugin within the library. Some libraries contain only
3042 one plugin, but others contain many of them. If this is not set filter
3043 will list all available plugins within the specified library.
3046 Set the '|' separated list of controls which are zero or more floating point
3047 values that determine the behavior of the loaded plugin (for example delay,
3049 Controls need to be defined using the following syntax:
3050 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3051 @var{valuei} is the value set on the @var{i}-th control.
3052 Alternatively they can be also defined using the following syntax:
3053 @var{value0}|@var{value1}|@var{value2}|..., where
3054 @var{valuei} is the value set on the @var{i}-th control.
3055 If @option{controls} is set to @code{help}, all available controls and
3056 their valid ranges are printed.
3058 @item sample_rate, s
3059 Specify the sample rate, default to 44100. Only used if plugin have
3063 Set the number of samples per channel per each output frame, default
3064 is 1024. Only used if plugin have zero inputs.
3067 Set the minimum duration of the sourced audio. See
3068 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3069 for the accepted syntax.
3070 Note that the resulting duration may be greater than the specified duration,
3071 as the generated audio is always cut at the end of a complete frame.
3072 If not specified, or the expressed duration is negative, the audio is
3073 supposed to be generated forever.
3074 Only used if plugin have zero inputs.
3078 @subsection Examples
3082 List all available plugins within amp (LADSPA example plugin) library:
3088 List all available controls and their valid ranges for @code{vcf_notch}
3089 plugin from @code{VCF} library:
3091 ladspa=f=vcf:p=vcf_notch:c=help
3095 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3098 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3102 Add reverberation to the audio using TAP-plugins
3103 (Tom's Audio Processing plugins):
3105 ladspa=file=tap_reverb:tap_reverb
3109 Generate white noise, with 0.2 amplitude:
3111 ladspa=file=cmt:noise_source_white:c=c0=.2
3115 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3116 @code{C* Audio Plugin Suite} (CAPS) library:
3118 ladspa=file=caps:Click:c=c1=20'
3122 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3124 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3128 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3129 @code{SWH Plugins} collection:
3131 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3135 Attenuate low frequencies using Multiband EQ from Steve Harris
3136 @code{SWH Plugins} collection:
3138 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3142 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3145 ladspa=caps:Narrower
3149 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3151 ladspa=caps:White:.2
3155 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3157 ladspa=caps:Fractal:c=c1=1
3161 Dynamic volume normalization using @code{VLevel} plugin:
3163 ladspa=vlevel-ladspa:vlevel_mono
3167 @subsection Commands
3169 This filter supports the following commands:
3172 Modify the @var{N}-th control value.
3174 If the specified value is not valid, it is ignored and prior one is kept.
3179 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3180 Support for both single pass (livestreams, files) and double pass (files) modes.
3181 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3182 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3183 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3185 The filter accepts the following options:
3189 Set integrated loudness target.
3190 Range is -70.0 - -5.0. Default value is -24.0.
3193 Set loudness range target.
3194 Range is 1.0 - 20.0. Default value is 7.0.
3197 Set maximum true peak.
3198 Range is -9.0 - +0.0. Default value is -2.0.
3200 @item measured_I, measured_i
3201 Measured IL of input file.
3202 Range is -99.0 - +0.0.
3204 @item measured_LRA, measured_lra
3205 Measured LRA of input file.
3206 Range is 0.0 - 99.0.
3208 @item measured_TP, measured_tp
3209 Measured true peak of input file.
3210 Range is -99.0 - +99.0.
3212 @item measured_thresh
3213 Measured threshold of input file.
3214 Range is -99.0 - +0.0.
3217 Set offset gain. Gain is applied before the true-peak limiter.
3218 Range is -99.0 - +99.0. Default is +0.0.
3221 Normalize linearly if possible.
3222 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3223 to be specified in order to use this mode.
3224 Options are true or false. Default is true.
3227 Treat mono input files as "dual-mono". If a mono file is intended for playback
3228 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3229 If set to @code{true}, this option will compensate for this effect.
3230 Multi-channel input files are not affected by this option.
3231 Options are true or false. Default is false.
3234 Set print format for stats. Options are summary, json, or none.
3235 Default value is none.
3240 Apply a low-pass filter with 3dB point frequency.
3241 The filter can be either single-pole or double-pole (the default).
3242 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3244 The filter accepts the following options:
3248 Set frequency in Hz. Default is 500.
3251 Set number of poles. Default is 2.
3254 Set method to specify band-width of filter.
3267 Specify the band-width of a filter in width_type units.
3268 Applies only to double-pole filter.
3269 The default is 0.707q and gives a Butterworth response.
3272 Specify which channels to filter, by default all available are filtered.
3275 @subsection Examples
3278 Lowpass only LFE channel, it LFE is not present it does nothing:
3285 Multiband Compress or expand the audio's dynamic range.
3287 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3288 This is akin to the crossover of a loudspeaker, and results in flat frequency
3289 response when absent compander action.
3291 It accepts the following parameters:
3295 This option syntax is:
3296 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3297 For explanation of each item refer to compand filter documentation.
3303 Mix channels with specific gain levels. The filter accepts the output
3304 channel layout followed by a set of channels definitions.
3306 This filter is also designed to efficiently remap the channels of an audio
3309 The filter accepts parameters of the form:
3310 "@var{l}|@var{outdef}|@var{outdef}|..."
3314 output channel layout or number of channels
3317 output channel specification, of the form:
3318 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3321 output channel to define, either a channel name (FL, FR, etc.) or a channel
3322 number (c0, c1, etc.)
3325 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3328 input channel to use, see out_name for details; it is not possible to mix
3329 named and numbered input channels
3332 If the `=' in a channel specification is replaced by `<', then the gains for
3333 that specification will be renormalized so that the total is 1, thus
3334 avoiding clipping noise.
3336 @subsection Mixing examples
3338 For example, if you want to down-mix from stereo to mono, but with a bigger
3339 factor for the left channel:
3341 pan=1c|c0=0.9*c0+0.1*c1
3344 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3345 7-channels surround:
3347 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3350 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3351 that should be preferred (see "-ac" option) unless you have very specific
3354 @subsection Remapping examples
3356 The channel remapping will be effective if, and only if:
3359 @item gain coefficients are zeroes or ones,
3360 @item only one input per channel output,
3363 If all these conditions are satisfied, the filter will notify the user ("Pure
3364 channel mapping detected"), and use an optimized and lossless method to do the
3367 For example, if you have a 5.1 source and want a stereo audio stream by
3368 dropping the extra channels:
3370 pan="stereo| c0=FL | c1=FR"
3373 Given the same source, you can also switch front left and front right channels
3374 and keep the input channel layout:
3376 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3379 If the input is a stereo audio stream, you can mute the front left channel (and
3380 still keep the stereo channel layout) with:
3385 Still with a stereo audio stream input, you can copy the right channel in both
3386 front left and right:
3388 pan="stereo| c0=FR | c1=FR"
3393 ReplayGain scanner filter. This filter takes an audio stream as an input and
3394 outputs it unchanged.
3395 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3399 Convert the audio sample format, sample rate and channel layout. It is
3400 not meant to be used directly.
3403 Apply time-stretching and pitch-shifting with librubberband.
3405 The filter accepts the following options:
3409 Set tempo scale factor.
3412 Set pitch scale factor.
3415 Set transients detector.
3416 Possible values are:
3425 Possible values are:
3434 Possible values are:
3441 Set processing window size.
3442 Possible values are:
3451 Possible values are:
3458 Enable formant preservation when shift pitching.
3459 Possible values are:
3467 Possible values are:
3476 Possible values are:
3483 @section sidechaincompress
3485 This filter acts like normal compressor but has the ability to compress
3486 detected signal using second input signal.
3487 It needs two input streams and returns one output stream.
3488 First input stream will be processed depending on second stream signal.
3489 The filtered signal then can be filtered with other filters in later stages of
3490 processing. See @ref{pan} and @ref{amerge} filter.
3492 The filter accepts the following options:
3496 Set input gain. Default is 1. Range is between 0.015625 and 64.
3499 If a signal of second stream raises above this level it will affect the gain
3500 reduction of first stream.
3501 By default is 0.125. Range is between 0.00097563 and 1.
3504 Set a ratio about which the signal is reduced. 1:2 means that if the level
3505 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3506 Default is 2. Range is between 1 and 20.
3509 Amount of milliseconds the signal has to rise above the threshold before gain
3510 reduction starts. Default is 20. Range is between 0.01 and 2000.
3513 Amount of milliseconds the signal has to fall below the threshold before
3514 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3517 Set the amount by how much signal will be amplified after processing.
3518 Default is 1. Range is from 1 to 64.
3521 Curve the sharp knee around the threshold to enter gain reduction more softly.
3522 Default is 2.82843. Range is between 1 and 8.
3525 Choose if the @code{average} level between all channels of side-chain stream
3526 or the louder(@code{maximum}) channel of side-chain stream affects the
3527 reduction. Default is @code{average}.
3530 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3531 of @code{rms}. Default is @code{rms} which is mainly smoother.
3534 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3537 How much to use compressed signal in output. Default is 1.
3538 Range is between 0 and 1.
3541 @subsection Examples
3545 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3546 depending on the signal of 2nd input and later compressed signal to be
3547 merged with 2nd input:
3549 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3553 @section sidechaingate
3555 A sidechain gate acts like a normal (wideband) gate but has the ability to
3556 filter the detected signal before sending it to the gain reduction stage.
3557 Normally a gate uses the full range signal to detect a level above the
3559 For example: If you cut all lower frequencies from your sidechain signal
3560 the gate will decrease the volume of your track only if not enough highs
3561 appear. With this technique you are able to reduce the resonation of a
3562 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3564 It needs two input streams and returns one output stream.
3565 First input stream will be processed depending on second stream signal.
3567 The filter accepts the following options:
3571 Set input level before filtering.
3572 Default is 1. Allowed range is from 0.015625 to 64.
3575 Set the level of gain reduction when the signal is below the threshold.
3576 Default is 0.06125. Allowed range is from 0 to 1.
3579 If a signal rises above this level the gain reduction is released.
3580 Default is 0.125. Allowed range is from 0 to 1.
3583 Set a ratio about which the signal is reduced.
3584 Default is 2. Allowed range is from 1 to 9000.
3587 Amount of milliseconds the signal has to rise above the threshold before gain
3589 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3592 Amount of milliseconds the signal has to fall below the threshold before the
3593 reduction is increased again. Default is 250 milliseconds.
3594 Allowed range is from 0.01 to 9000.
3597 Set amount of amplification of signal after processing.
3598 Default is 1. Allowed range is from 1 to 64.
3601 Curve the sharp knee around the threshold to enter gain reduction more softly.
3602 Default is 2.828427125. Allowed range is from 1 to 8.
3605 Choose if exact signal should be taken for detection or an RMS like one.
3606 Default is rms. Can be peak or rms.
3609 Choose if the average level between all channels or the louder channel affects
3611 Default is average. Can be average or maximum.
3614 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3617 @section silencedetect
3619 Detect silence in an audio stream.
3621 This filter logs a message when it detects that the input audio volume is less
3622 or equal to a noise tolerance value for a duration greater or equal to the
3623 minimum detected noise duration.
3625 The printed times and duration are expressed in seconds.
3627 The filter accepts the following options:
3631 Set silence duration until notification (default is 2 seconds).
3634 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3635 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3638 @subsection Examples
3642 Detect 5 seconds of silence with -50dB noise tolerance:
3644 silencedetect=n=-50dB:d=5
3648 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3649 tolerance in @file{silence.mp3}:
3651 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3655 @section silenceremove
3657 Remove silence from the beginning, middle or end of the audio.
3659 The filter accepts the following options:
3663 This value is used to indicate if audio should be trimmed at beginning of
3664 the audio. A value of zero indicates no silence should be trimmed from the
3665 beginning. When specifying a non-zero value, it trims audio up until it
3666 finds non-silence. Normally, when trimming silence from beginning of audio
3667 the @var{start_periods} will be @code{1} but it can be increased to higher
3668 values to trim all audio up to specific count of non-silence periods.
3669 Default value is @code{0}.
3671 @item start_duration
3672 Specify the amount of time that non-silence must be detected before it stops
3673 trimming audio. By increasing the duration, bursts of noises can be treated
3674 as silence and trimmed off. Default value is @code{0}.
3676 @item start_threshold
3677 This indicates what sample value should be treated as silence. For digital
3678 audio, a value of @code{0} may be fine but for audio recorded from analog,
3679 you may wish to increase the value to account for background noise.
3680 Can be specified in dB (in case "dB" is appended to the specified value)
3681 or amplitude ratio. Default value is @code{0}.
3684 Set the count for trimming silence from the end of audio.
3685 To remove silence from the middle of a file, specify a @var{stop_periods}
3686 that is negative. This value is then treated as a positive value and is
3687 used to indicate the effect should restart processing as specified by
3688 @var{start_periods}, making it suitable for removing periods of silence
3689 in the middle of the audio.
3690 Default value is @code{0}.
3693 Specify a duration of silence that must exist before audio is not copied any
3694 more. By specifying a higher duration, silence that is wanted can be left in
3696 Default value is @code{0}.
3698 @item stop_threshold
3699 This is the same as @option{start_threshold} but for trimming silence from
3701 Can be specified in dB (in case "dB" is appended to the specified value)
3702 or amplitude ratio. Default value is @code{0}.
3705 This indicates that @var{stop_duration} length of audio should be left intact
3706 at the beginning of each period of silence.
3707 For example, if you want to remove long pauses between words but do not want
3708 to remove the pauses completely. Default value is @code{0}.
3711 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3712 and works better with digital silence which is exactly 0.
3713 Default value is @code{rms}.
3716 Set ratio used to calculate size of window for detecting silence.
3717 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3720 @subsection Examples
3724 The following example shows how this filter can be used to start a recording
3725 that does not contain the delay at the start which usually occurs between
3726 pressing the record button and the start of the performance:
3728 silenceremove=1:5:0.02
3732 Trim all silence encountered from beginning to end where there is more than 1
3733 second of silence in audio:
3735 silenceremove=0:0:0:-1:1:-90dB
3741 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3742 loudspeakers around the user for binaural listening via headphones (audio
3743 formats up to 9 channels supported).
3744 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3745 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3746 Austrian Academy of Sciences.
3748 To enable compilation of this filter you need to configure FFmpeg with
3749 @code{--enable-libmysofa}.
3751 The filter accepts the following options:
3755 Set the SOFA file used for rendering.
3758 Set gain applied to audio. Value is in dB. Default is 0.
3761 Set rotation of virtual loudspeakers in deg. Default is 0.
3764 Set elevation of virtual speakers in deg. Default is 0.
3767 Set distance in meters between loudspeakers and the listener with near-field
3768 HRTFs. Default is 1.
3771 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3772 processing audio in time domain which is slow.
3773 @var{freq} is processing audio in frequency domain which is fast.
3774 Default is @var{freq}.
3777 Set custom positions of virtual loudspeakers. Syntax for this option is:
3778 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3779 Each virtual loudspeaker is described with short channel name following with
3780 azimuth and elevation in degrees.
3781 Each virtual loudspeaker description is separated by '|'.
3782 For example to override front left and front right channel positions use:
3783 'speakers=FL 45 15|FR 345 15'.
3784 Descriptions with unrecognised channel names are ignored.
3787 Set custom gain for LFE channels. Value is in dB. Default is 0.
3790 @subsection Examples
3794 Using ClubFritz6 sofa file:
3796 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3800 Using ClubFritz12 sofa file and bigger radius with small rotation:
3802 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3806 Similar as above but with custom speaker positions for front left, front right, back left and back right
3807 and also with custom gain:
3809 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
3813 @section stereotools
3815 This filter has some handy utilities to manage stereo signals, for converting
3816 M/S stereo recordings to L/R signal while having control over the parameters
3817 or spreading the stereo image of master track.
3819 The filter accepts the following options:
3823 Set input level before filtering for both channels. Defaults is 1.
3824 Allowed range is from 0.015625 to 64.
3827 Set output level after filtering for both channels. Defaults is 1.
3828 Allowed range is from 0.015625 to 64.
3831 Set input balance between both channels. Default is 0.
3832 Allowed range is from -1 to 1.
3835 Set output balance between both channels. Default is 0.
3836 Allowed range is from -1 to 1.
3839 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3840 clipping. Disabled by default.
3843 Mute the left channel. Disabled by default.
3846 Mute the right channel. Disabled by default.
3849 Change the phase of the left channel. Disabled by default.
3852 Change the phase of the right channel. Disabled by default.
3855 Set stereo mode. Available values are:
3859 Left/Right to Left/Right, this is default.
3862 Left/Right to Mid/Side.
3865 Mid/Side to Left/Right.
3868 Left/Right to Left/Left.
3871 Left/Right to Right/Right.
3874 Left/Right to Left + Right.
3877 Left/Right to Right/Left.
3880 Mid/Side to Left/Left.
3883 Mid/Side to Right/Right.
3887 Set level of side signal. Default is 1.
3888 Allowed range is from 0.015625 to 64.
3891 Set balance of side signal. Default is 0.
3892 Allowed range is from -1 to 1.
3895 Set level of the middle signal. Default is 1.
3896 Allowed range is from 0.015625 to 64.
3899 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3902 Set stereo base between mono and inversed channels. Default is 0.
3903 Allowed range is from -1 to 1.
3906 Set delay in milliseconds how much to delay left from right channel and
3907 vice versa. Default is 0. Allowed range is from -20 to 20.
3910 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3913 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3915 @item bmode_in, bmode_out
3916 Set balance mode for balance_in/balance_out option.
3918 Can be one of the following:
3922 Classic balance mode. Attenuate one channel at time.
3923 Gain is raised up to 1.
3926 Similar as classic mode above but gain is raised up to 2.
3929 Equal power distribution, from -6dB to +6dB range.
3933 @subsection Examples
3937 Apply karaoke like effect:
3939 stereotools=mlev=0.015625
3943 Convert M/S signal to L/R:
3945 "stereotools=mode=ms>lr"
3949 @section stereowiden
3951 This filter enhance the stereo effect by suppressing signal common to both
3952 channels and by delaying the signal of left into right and vice versa,
3953 thereby widening the stereo effect.
3955 The filter accepts the following options:
3959 Time in milliseconds of the delay of left signal into right and vice versa.
3960 Default is 20 milliseconds.
3963 Amount of gain in delayed signal into right and vice versa. Gives a delay
3964 effect of left signal in right output and vice versa which gives widening
3965 effect. Default is 0.3.
3968 Cross feed of left into right with inverted phase. This helps in suppressing
3969 the mono. If the value is 1 it will cancel all the signal common to both
3970 channels. Default is 0.3.
3973 Set level of input signal of original channel. Default is 0.8.
3976 @section superequalizer
3977 Apply 18 band equalizer.
3979 The filter accepts the following options:
3986 Set 131Hz band gain.
3988 Set 185Hz band gain.
3990 Set 262Hz band gain.
3992 Set 370Hz band gain.
3994 Set 523Hz band gain.
3996 Set 740Hz band gain.
3998 Set 1047Hz band gain.
4000 Set 1480Hz band gain.
4002 Set 2093Hz band gain.
4004 Set 2960Hz band gain.
4006 Set 4186Hz band gain.
4008 Set 5920Hz band gain.
4010 Set 8372Hz band gain.
4012 Set 11840Hz band gain.
4014 Set 16744Hz band gain.
4016 Set 20000Hz band gain.
4020 Apply audio surround upmix filter.
4022 This filter allows to produce multichannel output from audio stream.
4024 The filter accepts the following options:
4028 Set output channel layout. By default, this is @var{5.1}.
4030 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4031 for the required syntax.
4034 Set input channel layout. By default, this is @var{stereo}.
4036 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4037 for the required syntax.
4040 Set input volume level. By default, this is @var{1}.
4043 Set output volume level. By default, this is @var{1}.
4046 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4049 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4052 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4055 Set front center input volume. By default, this is @var{1}.
4058 Set front center output volume. By default, this is @var{1}.
4061 Set LFE input volume. By default, this is @var{1}.
4064 Set LFE output volume. By default, this is @var{1}.
4069 Boost or cut treble (upper) frequencies of the audio using a two-pole
4070 shelving filter with a response similar to that of a standard
4071 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4073 The filter accepts the following options:
4077 Give the gain at whichever is the lower of ~22 kHz and the
4078 Nyquist frequency. Its useful range is about -20 (for a large cut)
4079 to +20 (for a large boost). Beware of clipping when using a positive gain.
4082 Set the filter's central frequency and so can be used
4083 to extend or reduce the frequency range to be boosted or cut.
4084 The default value is @code{3000} Hz.
4087 Set method to specify band-width of filter.
4100 Determine how steep is the filter's shelf transition.
4103 Specify which channels to filter, by default all available are filtered.
4108 Sinusoidal amplitude modulation.
4110 The filter accepts the following options:
4114 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4115 (20 Hz or lower) will result in a tremolo effect.
4116 This filter may also be used as a ring modulator by specifying
4117 a modulation frequency higher than 20 Hz.
4118 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4121 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4122 Default value is 0.5.
4127 Sinusoidal phase modulation.
4129 The filter accepts the following options:
4133 Modulation frequency in Hertz.
4134 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4137 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4138 Default value is 0.5.
4143 Adjust the input audio volume.
4145 It accepts the following parameters:
4149 Set audio volume expression.
4151 Output values are clipped to the maximum value.
4153 The output audio volume is given by the relation:
4155 @var{output_volume} = @var{volume} * @var{input_volume}
4158 The default value for @var{volume} is "1.0".
4161 This parameter represents the mathematical precision.
4163 It determines which input sample formats will be allowed, which affects the
4164 precision of the volume scaling.
4168 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4170 32-bit floating-point; this limits input sample format to FLT. (default)
4172 64-bit floating-point; this limits input sample format to DBL.
4176 Choose the behaviour on encountering ReplayGain side data in input frames.
4180 Remove ReplayGain side data, ignoring its contents (the default).
4183 Ignore ReplayGain side data, but leave it in the frame.
4186 Prefer the track gain, if present.
4189 Prefer the album gain, if present.
4192 @item replaygain_preamp
4193 Pre-amplification gain in dB to apply to the selected replaygain gain.
4195 Default value for @var{replaygain_preamp} is 0.0.
4198 Set when the volume expression is evaluated.
4200 It accepts the following values:
4203 only evaluate expression once during the filter initialization, or
4204 when the @samp{volume} command is sent
4207 evaluate expression for each incoming frame
4210 Default value is @samp{once}.
4213 The volume expression can contain the following parameters.
4217 frame number (starting at zero)
4220 @item nb_consumed_samples
4221 number of samples consumed by the filter
4223 number of samples in the current frame
4225 original frame position in the file
4231 PTS at start of stream
4233 time at start of stream
4239 last set volume value
4242 Note that when @option{eval} is set to @samp{once} only the
4243 @var{sample_rate} and @var{tb} variables are available, all other
4244 variables will evaluate to NAN.
4246 @subsection Commands
4248 This filter supports the following commands:
4251 Modify the volume expression.
4252 The command accepts the same syntax of the corresponding option.
4254 If the specified expression is not valid, it is kept at its current
4256 @item replaygain_noclip
4257 Prevent clipping by limiting the gain applied.
4259 Default value for @var{replaygain_noclip} is 1.
4263 @subsection Examples
4267 Halve the input audio volume:
4271 volume=volume=-6.0206dB
4274 In all the above example the named key for @option{volume} can be
4275 omitted, for example like in:
4281 Increase input audio power by 6 decibels using fixed-point precision:
4283 volume=volume=6dB:precision=fixed
4287 Fade volume after time 10 with an annihilation period of 5 seconds:
4289 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4293 @section volumedetect
4295 Detect the volume of the input video.
4297 The filter has no parameters. The input is not modified. Statistics about
4298 the volume will be printed in the log when the input stream end is reached.
4300 In particular it will show the mean volume (root mean square), maximum
4301 volume (on a per-sample basis), and the beginning of a histogram of the
4302 registered volume values (from the maximum value to a cumulated 1/1000 of
4305 All volumes are in decibels relative to the maximum PCM value.
4307 @subsection Examples
4309 Here is an excerpt of the output:
4311 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4312 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4313 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4314 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4315 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4316 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4317 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4318 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4319 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4325 The mean square energy is approximately -27 dB, or 10^-2.7.
4327 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4329 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4332 In other words, raising the volume by +4 dB does not cause any clipping,
4333 raising it by +5 dB causes clipping for 6 samples, etc.
4335 @c man end AUDIO FILTERS
4337 @chapter Audio Sources
4338 @c man begin AUDIO SOURCES
4340 Below is a description of the currently available audio sources.
4344 Buffer audio frames, and make them available to the filter chain.
4346 This source is mainly intended for a programmatic use, in particular
4347 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4349 It accepts the following parameters:
4353 The timebase which will be used for timestamps of submitted frames. It must be
4354 either a floating-point number or in @var{numerator}/@var{denominator} form.
4357 The sample rate of the incoming audio buffers.
4360 The sample format of the incoming audio buffers.
4361 Either a sample format name or its corresponding integer representation from
4362 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4364 @item channel_layout
4365 The channel layout of the incoming audio buffers.
4366 Either a channel layout name from channel_layout_map in
4367 @file{libavutil/channel_layout.c} or its corresponding integer representation
4368 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4371 The number of channels of the incoming audio buffers.
4372 If both @var{channels} and @var{channel_layout} are specified, then they
4377 @subsection Examples
4380 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4383 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4384 Since the sample format with name "s16p" corresponds to the number
4385 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4388 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4393 Generate an audio signal specified by an expression.
4395 This source accepts in input one or more expressions (one for each
4396 channel), which are evaluated and used to generate a corresponding
4399 This source accepts the following options:
4403 Set the '|'-separated expressions list for each separate channel. In case the
4404 @option{channel_layout} option is not specified, the selected channel layout
4405 depends on the number of provided expressions. Otherwise the last
4406 specified expression is applied to the remaining output channels.
4408 @item channel_layout, c
4409 Set the channel layout. The number of channels in the specified layout
4410 must be equal to the number of specified expressions.
4413 Set the minimum duration of the sourced audio. See
4414 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4415 for the accepted syntax.
4416 Note that the resulting duration may be greater than the specified
4417 duration, as the generated audio is always cut at the end of a
4420 If not specified, or the expressed duration is negative, the audio is
4421 supposed to be generated forever.
4424 Set the number of samples per channel per each output frame,
4427 @item sample_rate, s
4428 Specify the sample rate, default to 44100.
4431 Each expression in @var{exprs} can contain the following constants:
4435 number of the evaluated sample, starting from 0
4438 time of the evaluated sample expressed in seconds, starting from 0
4445 @subsection Examples
4455 Generate a sin signal with frequency of 440 Hz, set sample rate to
4458 aevalsrc="sin(440*2*PI*t):s=8000"
4462 Generate a two channels signal, specify the channel layout (Front
4463 Center + Back Center) explicitly:
4465 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4469 Generate white noise:
4471 aevalsrc="-2+random(0)"
4475 Generate an amplitude modulated signal:
4477 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4481 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4483 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4490 The null audio source, return unprocessed audio frames. It is mainly useful
4491 as a template and to be employed in analysis / debugging tools, or as
4492 the source for filters which ignore the input data (for example the sox
4495 This source accepts the following options:
4499 @item channel_layout, cl
4501 Specifies the channel layout, and can be either an integer or a string
4502 representing a channel layout. The default value of @var{channel_layout}
4505 Check the channel_layout_map definition in
4506 @file{libavutil/channel_layout.c} for the mapping between strings and
4507 channel layout values.
4509 @item sample_rate, r
4510 Specifies the sample rate, and defaults to 44100.
4513 Set the number of samples per requested frames.
4517 @subsection Examples
4521 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4523 anullsrc=r=48000:cl=4
4527 Do the same operation with a more obvious syntax:
4529 anullsrc=r=48000:cl=mono
4533 All the parameters need to be explicitly defined.
4537 Synthesize a voice utterance using the libflite library.
4539 To enable compilation of this filter you need to configure FFmpeg with
4540 @code{--enable-libflite}.
4542 Note that versions of the flite library prior to 2.0 are not thread-safe.
4544 The filter accepts the following options:
4549 If set to 1, list the names of the available voices and exit
4550 immediately. Default value is 0.
4553 Set the maximum number of samples per frame. Default value is 512.
4556 Set the filename containing the text to speak.
4559 Set the text to speak.
4562 Set the voice to use for the speech synthesis. Default value is
4563 @code{kal}. See also the @var{list_voices} option.
4566 @subsection Examples
4570 Read from file @file{speech.txt}, and synthesize the text using the
4571 standard flite voice:
4573 flite=textfile=speech.txt
4577 Read the specified text selecting the @code{slt} voice:
4579 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4583 Input text to ffmpeg:
4585 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4589 Make @file{ffplay} speak the specified text, using @code{flite} and
4590 the @code{lavfi} device:
4592 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4596 For more information about libflite, check:
4597 @url{http://www.festvox.org/flite/}
4601 Generate a noise audio signal.
4603 The filter accepts the following options:
4606 @item sample_rate, r
4607 Specify the sample rate. Default value is 48000 Hz.
4610 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4614 Specify the duration of the generated audio stream. Not specifying this option
4615 results in noise with an infinite length.
4617 @item color, colour, c
4618 Specify the color of noise. Available noise colors are white, pink, brown,
4619 blue and violet. Default color is white.
4622 Specify a value used to seed the PRNG.
4625 Set the number of samples per each output frame, default is 1024.
4628 @subsection Examples
4633 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4635 anoisesrc=d=60:c=pink:r=44100:a=0.5
4641 Generate an audio signal made of a sine wave with amplitude 1/8.
4643 The audio signal is bit-exact.
4645 The filter accepts the following options:
4650 Set the carrier frequency. Default is 440 Hz.
4652 @item beep_factor, b
4653 Enable a periodic beep every second with frequency @var{beep_factor} times
4654 the carrier frequency. Default is 0, meaning the beep is disabled.
4656 @item sample_rate, r
4657 Specify the sample rate, default is 44100.
4660 Specify the duration of the generated audio stream.
4662 @item samples_per_frame
4663 Set the number of samples per output frame.
4665 The expression can contain the following constants:
4669 The (sequential) number of the output audio frame, starting from 0.
4672 The PTS (Presentation TimeStamp) of the output audio frame,
4673 expressed in @var{TB} units.
4676 The PTS of the output audio frame, expressed in seconds.
4679 The timebase of the output audio frames.
4682 Default is @code{1024}.
4685 @subsection Examples
4690 Generate a simple 440 Hz sine wave:
4696 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4700 sine=frequency=220:beep_factor=4:duration=5
4704 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4707 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4711 @c man end AUDIO SOURCES
4713 @chapter Audio Sinks
4714 @c man begin AUDIO SINKS
4716 Below is a description of the currently available audio sinks.
4718 @section abuffersink
4720 Buffer audio frames, and make them available to the end of filter chain.
4722 This sink is mainly intended for programmatic use, in particular
4723 through the interface defined in @file{libavfilter/buffersink.h}
4724 or the options system.
4726 It accepts a pointer to an AVABufferSinkContext structure, which
4727 defines the incoming buffers' formats, to be passed as the opaque
4728 parameter to @code{avfilter_init_filter} for initialization.
4731 Null audio sink; do absolutely nothing with the input audio. It is
4732 mainly useful as a template and for use in analysis / debugging
4735 @c man end AUDIO SINKS
4737 @chapter Video Filters
4738 @c man begin VIDEO FILTERS
4740 When you configure your FFmpeg build, you can disable any of the
4741 existing filters using @code{--disable-filters}.
4742 The configure output will show the video filters included in your
4745 Below is a description of the currently available video filters.
4747 @section alphaextract
4749 Extract the alpha component from the input as a grayscale video. This
4750 is especially useful with the @var{alphamerge} filter.
4754 Add or replace the alpha component of the primary input with the
4755 grayscale value of a second input. This is intended for use with
4756 @var{alphaextract} to allow the transmission or storage of frame
4757 sequences that have alpha in a format that doesn't support an alpha
4760 For example, to reconstruct full frames from a normal YUV-encoded video
4761 and a separate video created with @var{alphaextract}, you might use:
4763 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4766 Since this filter is designed for reconstruction, it operates on frame
4767 sequences without considering timestamps, and terminates when either
4768 input reaches end of stream. This will cause problems if your encoding
4769 pipeline drops frames. If you're trying to apply an image as an
4770 overlay to a video stream, consider the @var{overlay} filter instead.
4774 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4775 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4776 Substation Alpha) subtitles files.
4778 This filter accepts the following option in addition to the common options from
4779 the @ref{subtitles} filter:
4783 Set the shaping engine
4785 Available values are:
4788 The default libass shaping engine, which is the best available.
4790 Fast, font-agnostic shaper that can do only substitutions
4792 Slower shaper using OpenType for substitutions and positioning
4795 The default is @code{auto}.
4799 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4801 The filter accepts the following options:
4805 Set threshold A for 1st plane. Default is 0.02.
4806 Valid range is 0 to 0.3.
4809 Set threshold B for 1st plane. Default is 0.04.
4810 Valid range is 0 to 5.
4813 Set threshold A for 2nd plane. Default is 0.02.
4814 Valid range is 0 to 0.3.
4817 Set threshold B for 2nd plane. Default is 0.04.
4818 Valid range is 0 to 5.
4821 Set threshold A for 3rd plane. Default is 0.02.
4822 Valid range is 0 to 0.3.
4825 Set threshold B for 3rd plane. Default is 0.04.
4826 Valid range is 0 to 5.
4828 Threshold A is designed to react on abrupt changes in the input signal and
4829 threshold B is designed to react on continuous changes in the input signal.
4832 Set number of frames filter will use for averaging. Default is 33. Must be odd
4833 number in range [5, 129].
4836 Set what planes of frame filter will use for averaging. Default is all.
4841 Apply average blur filter.
4843 The filter accepts the following options:
4847 Set horizontal kernel size.
4850 Set which planes to filter. By default all planes are filtered.
4853 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4854 Default is @code{0}.
4859 Compute the bounding box for the non-black pixels in the input frame
4862 This filter computes the bounding box containing all the pixels with a
4863 luminance value greater than the minimum allowed value.
4864 The parameters describing the bounding box are printed on the filter
4867 The filter accepts the following option:
4871 Set the minimal luminance value. Default is @code{16}.
4874 @section bitplanenoise
4876 Show and measure bit plane noise.
4878 The filter accepts the following options:
4882 Set which plane to analyze. Default is @code{1}.
4885 Filter out noisy pixels from @code{bitplane} set above.
4886 Default is disabled.
4889 @section blackdetect
4891 Detect video intervals that are (almost) completely black. Can be
4892 useful to detect chapter transitions, commercials, or invalid
4893 recordings. Output lines contains the time for the start, end and
4894 duration of the detected black interval expressed in seconds.
4896 In order to display the output lines, you need to set the loglevel at
4897 least to the AV_LOG_INFO value.
4899 The filter accepts the following options:
4902 @item black_min_duration, d
4903 Set the minimum detected black duration expressed in seconds. It must
4904 be a non-negative floating point number.
4906 Default value is 2.0.
4908 @item picture_black_ratio_th, pic_th
4909 Set the threshold for considering a picture "black".
4910 Express the minimum value for the ratio:
4912 @var{nb_black_pixels} / @var{nb_pixels}
4915 for which a picture is considered black.
4916 Default value is 0.98.
4918 @item pixel_black_th, pix_th
4919 Set the threshold for considering a pixel "black".
4921 The threshold expresses the maximum pixel luminance value for which a
4922 pixel is considered "black". The provided value is scaled according to
4923 the following equation:
4925 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4928 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4929 the input video format, the range is [0-255] for YUV full-range
4930 formats and [16-235] for YUV non full-range formats.
4932 Default value is 0.10.
4935 The following example sets the maximum pixel threshold to the minimum
4936 value, and detects only black intervals of 2 or more seconds:
4938 blackdetect=d=2:pix_th=0.00
4943 Detect frames that are (almost) completely black. Can be useful to
4944 detect chapter transitions or commercials. Output lines consist of
4945 the frame number of the detected frame, the percentage of blackness,
4946 the position in the file if known or -1 and the timestamp in seconds.
4948 In order to display the output lines, you need to set the loglevel at
4949 least to the AV_LOG_INFO value.
4951 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
4952 The value represents the percentage of pixels in the picture that
4953 are below the threshold value.
4955 It accepts the following parameters:
4960 The percentage of the pixels that have to be below the threshold; it defaults to
4963 @item threshold, thresh
4964 The threshold below which a pixel value is considered black; it defaults to
4969 @section blend, tblend
4971 Blend two video frames into each other.
4973 The @code{blend} filter takes two input streams and outputs one
4974 stream, the first input is the "top" layer and second input is
4975 "bottom" layer. By default, the output terminates when the longest input terminates.
4977 The @code{tblend} (time blend) filter takes two consecutive frames
4978 from one single stream, and outputs the result obtained by blending
4979 the new frame on top of the old frame.
4981 A description of the accepted options follows.
4989 Set blend mode for specific pixel component or all pixel components in case
4990 of @var{all_mode}. Default value is @code{normal}.
4992 Available values for component modes are:
5034 Set blend opacity for specific pixel component or all pixel components in case
5035 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5042 Set blend expression for specific pixel component or all pixel components in case
5043 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5045 The expressions can use the following variables:
5049 The sequential number of the filtered frame, starting from @code{0}.
5053 the coordinates of the current sample
5057 the width and height of currently filtered plane
5061 Width and height scale depending on the currently filtered plane. It is the
5062 ratio between the corresponding luma plane number of pixels and the current
5063 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5064 @code{0.5,0.5} for chroma planes.
5067 Time of the current frame, expressed in seconds.
5070 Value of pixel component at current location for first video frame (top layer).
5073 Value of pixel component at current location for second video frame (bottom layer).
5077 The @code{blend} filter also supports the @ref{framesync} options.
5079 @subsection Examples
5083 Apply transition from bottom layer to top layer in first 10 seconds:
5085 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5089 Apply linear horizontal transition from top layer to bottom layer:
5091 blend=all_expr='A*(X/W)+B*(1-X/W)'
5095 Apply 1x1 checkerboard effect:
5097 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5101 Apply uncover left effect:
5103 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5107 Apply uncover down effect:
5109 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5113 Apply uncover up-left effect:
5115 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5119 Split diagonally video and shows top and bottom layer on each side:
5121 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5125 Display differences between the current and the previous frame:
5127 tblend=all_mode=grainextract
5133 Apply a boxblur algorithm to the input video.
5135 It accepts the following parameters:
5139 @item luma_radius, lr
5140 @item luma_power, lp
5141 @item chroma_radius, cr
5142 @item chroma_power, cp
5143 @item alpha_radius, ar
5144 @item alpha_power, ap
5148 A description of the accepted options follows.
5151 @item luma_radius, lr
5152 @item chroma_radius, cr
5153 @item alpha_radius, ar
5154 Set an expression for the box radius in pixels used for blurring the
5155 corresponding input plane.
5157 The radius value must be a non-negative number, and must not be
5158 greater than the value of the expression @code{min(w,h)/2} for the
5159 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5162 Default value for @option{luma_radius} is "2". If not specified,
5163 @option{chroma_radius} and @option{alpha_radius} default to the
5164 corresponding value set for @option{luma_radius}.
5166 The expressions can contain the following constants:
5170 The input width and height in pixels.
5174 The input chroma image width and height in pixels.
5178 The horizontal and vertical chroma subsample values. For example, for the
5179 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5182 @item luma_power, lp
5183 @item chroma_power, cp
5184 @item alpha_power, ap
5185 Specify how many times the boxblur filter is applied to the
5186 corresponding plane.
5188 Default value for @option{luma_power} is 2. If not specified,
5189 @option{chroma_power} and @option{alpha_power} default to the
5190 corresponding value set for @option{luma_power}.
5192 A value of 0 will disable the effect.
5195 @subsection Examples
5199 Apply a boxblur filter with the luma, chroma, and alpha radii
5202 boxblur=luma_radius=2:luma_power=1
5207 Set the luma radius to 2, and alpha and chroma radius to 0:
5209 boxblur=2:1:cr=0:ar=0
5213 Set the luma and chroma radii to a fraction of the video dimension:
5215 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5221 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5222 Deinterlacing Filter").
5224 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5225 interpolation algorithms.
5226 It accepts the following parameters:
5230 The interlacing mode to adopt. It accepts one of the following values:
5234 Output one frame for each frame.
5236 Output one frame for each field.
5239 The default value is @code{send_field}.
5242 The picture field parity assumed for the input interlaced video. It accepts one
5243 of the following values:
5247 Assume the top field is first.
5249 Assume the bottom field is first.
5251 Enable automatic detection of field parity.
5254 The default value is @code{auto}.
5255 If the interlacing is unknown or the decoder does not export this information,
5256 top field first will be assumed.
5259 Specify which frames to deinterlace. Accept one of the following
5264 Deinterlace all frames.
5266 Only deinterlace frames marked as interlaced.
5269 The default value is @code{all}.
5273 YUV colorspace color/chroma keying.
5275 The filter accepts the following options:
5279 The color which will be replaced with transparency.
5282 Similarity percentage with the key color.
5284 0.01 matches only the exact key color, while 1.0 matches everything.
5289 0.0 makes pixels either fully transparent, or not transparent at all.
5291 Higher values result in semi-transparent pixels, with a higher transparency
5292 the more similar the pixels color is to the key color.
5295 Signals that the color passed is already in YUV instead of RGB.
5297 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5298 This can be used to pass exact YUV values as hexadecimal numbers.
5301 @subsection Examples
5305 Make every green pixel in the input image transparent:
5307 ffmpeg -i input.png -vf chromakey=green out.png
5311 Overlay a greenscreen-video on top of a static black background.
5313 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
5319 Display CIE color diagram with pixels overlaid onto it.
5321 The filter accepts the following options:
5336 @item uhdtv, rec2020
5349 Set what gamuts to draw.
5351 See @code{system} option for available values.
5354 Set ciescope size, by default set to 512.
5357 Set intensity used to map input pixel values to CIE diagram.
5360 Set contrast used to draw tongue colors that are out of active color system gamut.
5363 Correct gamma displayed on scope, by default enabled.
5366 Show white point on CIE diagram, by default disabled.
5369 Set input gamma. Used only with XYZ input color space.
5374 Visualize information exported by some codecs.
5376 Some codecs can export information through frames using side-data or other
5377 means. For example, some MPEG based codecs export motion vectors through the
5378 @var{export_mvs} flag in the codec @option{flags2} option.
5380 The filter accepts the following option:
5384 Set motion vectors to visualize.
5386 Available flags for @var{mv} are:
5390 forward predicted MVs of P-frames
5392 forward predicted MVs of B-frames
5394 backward predicted MVs of B-frames
5398 Display quantization parameters using the chroma planes.
5401 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5403 Available flags for @var{mv_type} are:
5407 forward predicted MVs
5409 backward predicted MVs
5412 @item frame_type, ft
5413 Set frame type to visualize motion vectors of.
5415 Available flags for @var{frame_type} are:
5419 intra-coded frames (I-frames)
5421 predicted frames (P-frames)
5423 bi-directionally predicted frames (B-frames)
5427 @subsection Examples
5431 Visualize forward predicted MVs of all frames using @command{ffplay}:
5433 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5437 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5439 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5443 @section colorbalance
5444 Modify intensity of primary colors (red, green and blue) of input frames.
5446 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5447 regions for the red-cyan, green-magenta or blue-yellow balance.
5449 A positive adjustment value shifts the balance towards the primary color, a negative
5450 value towards the complementary color.
5452 The filter accepts the following options:
5458 Adjust red, green and blue shadows (darkest pixels).
5463 Adjust red, green and blue midtones (medium pixels).
5468 Adjust red, green and blue highlights (brightest pixels).
5470 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5473 @subsection Examples
5477 Add red color cast to shadows:
5484 RGB colorspace color keying.
5486 The filter accepts the following options:
5490 The color which will be replaced with transparency.
5493 Similarity percentage with the key color.
5495 0.01 matches only the exact key color, while 1.0 matches everything.
5500 0.0 makes pixels either fully transparent, or not transparent at all.
5502 Higher values result in semi-transparent pixels, with a higher transparency
5503 the more similar the pixels color is to the key color.
5506 @subsection Examples
5510 Make every green pixel in the input image transparent:
5512 ffmpeg -i input.png -vf colorkey=green out.png
5516 Overlay a greenscreen-video on top of a static background image.
5518 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
5522 @section colorlevels
5524 Adjust video input frames using levels.
5526 The filter accepts the following options:
5533 Adjust red, green, blue and alpha input black point.
5534 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5540 Adjust red, green, blue and alpha input white point.
5541 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5543 Input levels are used to lighten highlights (bright tones), darken shadows
5544 (dark tones), change the balance of bright and dark tones.
5550 Adjust red, green, blue and alpha output black point.
5551 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5557 Adjust red, green, blue and alpha output white point.
5558 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5560 Output levels allows manual selection of a constrained output level range.
5563 @subsection Examples
5567 Make video output darker:
5569 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5575 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5579 Make video output lighter:
5581 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5585 Increase brightness:
5587 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5591 @section colorchannelmixer
5593 Adjust video input frames by re-mixing color channels.
5595 This filter modifies a color channel by adding the values associated to
5596 the other channels of the same pixels. For example if the value to
5597 modify is red, the output value will be:
5599 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5602 The filter accepts the following options:
5609 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5610 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5616 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5617 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5623 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5624 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5630 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5631 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5633 Allowed ranges for options are @code{[-2.0, 2.0]}.
5636 @subsection Examples
5640 Convert source to grayscale:
5642 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5645 Simulate sepia tones:
5647 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5651 @section colormatrix
5653 Convert color matrix.
5655 The filter accepts the following options:
5660 Specify the source and destination color matrix. Both values must be
5663 The accepted values are:
5691 For example to convert from BT.601 to SMPTE-240M, use the command:
5693 colormatrix=bt601:smpte240m
5698 Convert colorspace, transfer characteristics or color primaries.
5699 Input video needs to have an even size.
5701 The filter accepts the following options:
5706 Specify all color properties at once.
5708 The accepted values are:
5738 Specify output colorspace.
5740 The accepted values are:
5749 BT.470BG or BT.601-6 625
5752 SMPTE-170M or BT.601-6 525
5761 BT.2020 with non-constant luminance
5767 Specify output transfer characteristics.
5769 The accepted values are:
5781 Constant gamma of 2.2
5784 Constant gamma of 2.8
5787 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5805 BT.2020 for 10-bits content
5808 BT.2020 for 12-bits content
5814 Specify output color primaries.
5816 The accepted values are:
5825 BT.470BG or BT.601-6 625
5828 SMPTE-170M or BT.601-6 525
5852 Specify output color range.
5854 The accepted values are:
5857 TV (restricted) range
5860 MPEG (restricted) range
5871 Specify output color format.
5873 The accepted values are:
5876 YUV 4:2:0 planar 8-bits
5879 YUV 4:2:0 planar 10-bits
5882 YUV 4:2:0 planar 12-bits
5885 YUV 4:2:2 planar 8-bits
5888 YUV 4:2:2 planar 10-bits
5891 YUV 4:2:2 planar 12-bits
5894 YUV 4:4:4 planar 8-bits
5897 YUV 4:4:4 planar 10-bits
5900 YUV 4:4:4 planar 12-bits
5905 Do a fast conversion, which skips gamma/primary correction. This will take
5906 significantly less CPU, but will be mathematically incorrect. To get output
5907 compatible with that produced by the colormatrix filter, use fast=1.
5910 Specify dithering mode.
5912 The accepted values are:
5918 Floyd-Steinberg dithering
5922 Whitepoint adaptation mode.
5924 The accepted values are:
5927 Bradford whitepoint adaptation
5930 von Kries whitepoint adaptation
5933 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5937 Override all input properties at once. Same accepted values as @ref{all}.
5940 Override input colorspace. Same accepted values as @ref{space}.
5943 Override input color primaries. Same accepted values as @ref{primaries}.
5946 Override input transfer characteristics. Same accepted values as @ref{trc}.
5949 Override input color range. Same accepted values as @ref{range}.
5953 The filter converts the transfer characteristics, color space and color
5954 primaries to the specified user values. The output value, if not specified,
5955 is set to a default value based on the "all" property. If that property is
5956 also not specified, the filter will log an error. The output color range and
5957 format default to the same value as the input color range and format. The
5958 input transfer characteristics, color space, color primaries and color range
5959 should be set on the input data. If any of these are missing, the filter will
5960 log an error and no conversion will take place.
5962 For example to convert the input to SMPTE-240M, use the command:
5964 colorspace=smpte240m
5967 @section convolution
5969 Apply convolution 3x3 or 5x5 filter.
5971 The filter accepts the following options:
5978 Set matrix for each plane.
5979 Matrix is sequence of 9 or 25 signed integers.
5985 Set multiplier for calculated value for each plane.
5991 Set bias for each plane. This value is added to the result of the multiplication.
5992 Useful for making the overall image brighter or darker. Default is 0.0.
5995 @subsection Examples
6001 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"
6007 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"
6013 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"
6019 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"
6023 Apply laplacian edge detector which includes diagonals:
6025 convolution="1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0"
6031 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"
6037 Apply 2D convolution of video stream in frequency domain using second stream
6040 The filter accepts the following options:
6044 Set which planes to process.
6047 Set which impulse video frames will be processed, can be @var{first}
6048 or @var{all}. Default is @var{all}.
6051 The @code{convolve} filter also supports the @ref{framesync} options.
6055 Copy the input video source unchanged to the output. This is mainly useful for
6060 Video filtering on GPU using Apple's CoreImage API on OSX.
6062 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6063 processed by video hardware. However, software-based OpenGL implementations
6064 exist which means there is no guarantee for hardware processing. It depends on
6067 There are many filters and image generators provided by Apple that come with a
6068 large variety of options. The filter has to be referenced by its name along
6071 The coreimage filter accepts the following options:
6074 List all available filters and generators along with all their respective
6075 options as well as possible minimum and maximum values along with the default
6082 Specify all filters by their respective name and options.
6083 Use @var{list_filters} to determine all valid filter names and options.
6084 Numerical options are specified by a float value and are automatically clamped
6085 to their respective value range. Vector and color options have to be specified
6086 by a list of space separated float values. Character escaping has to be done.
6087 A special option name @code{default} is available to use default options for a
6090 It is required to specify either @code{default} or at least one of the filter options.
6091 All omitted options are used with their default values.
6092 The syntax of the filter string is as follows:
6094 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6098 Specify a rectangle where the output of the filter chain is copied into the
6099 input image. It is given by a list of space separated float values:
6101 output_rect=x\ y\ width\ height
6103 If not given, the output rectangle equals the dimensions of the input image.
6104 The output rectangle is automatically cropped at the borders of the input
6105 image. Negative values are valid for each component.
6107 output_rect=25\ 25\ 100\ 100
6111 Several filters can be chained for successive processing without GPU-HOST
6112 transfers allowing for fast processing of complex filter chains.
6113 Currently, only filters with zero (generators) or exactly one (filters) input
6114 image and one output image are supported. Also, transition filters are not yet
6117 Some filters generate output images with additional padding depending on the
6118 respective filter kernel. The padding is automatically removed to ensure the
6119 filter output has the same size as the input image.
6121 For image generators, the size of the output image is determined by the
6122 previous output image of the filter chain or the input image of the whole
6123 filterchain, respectively. The generators do not use the pixel information of
6124 this image to generate their output. However, the generated output is
6125 blended onto this image, resulting in partial or complete coverage of the
6128 The @ref{coreimagesrc} video source can be used for generating input images
6129 which are directly fed into the filter chain. By using it, providing input
6130 images by another video source or an input video is not required.
6132 @subsection Examples
6137 List all filters available:
6139 coreimage=list_filters=true
6143 Use the CIBoxBlur filter with default options to blur an image:
6145 coreimage=filter=CIBoxBlur@@default
6149 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6150 its center at 100x100 and a radius of 50 pixels:
6152 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6156 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6157 given as complete and escaped command-line for Apple's standard bash shell:
6159 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6165 Crop the input video to given dimensions.
6167 It accepts the following parameters:
6171 The width of the output video. It defaults to @code{iw}.
6172 This expression is evaluated only once during the filter
6173 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6176 The height of the output video. It defaults to @code{ih}.
6177 This expression is evaluated only once during the filter
6178 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6181 The horizontal position, in the input video, of the left edge of the output
6182 video. It defaults to @code{(in_w-out_w)/2}.
6183 This expression is evaluated per-frame.
6186 The vertical position, in the input video, of the top edge of the output video.
6187 It defaults to @code{(in_h-out_h)/2}.
6188 This expression is evaluated per-frame.
6191 If set to 1 will force the output display aspect ratio
6192 to be the same of the input, by changing the output sample aspect
6193 ratio. It defaults to 0.
6196 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6197 width/height/x/y as specified and will not be rounded to nearest smaller value.
6201 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6202 expressions containing the following constants:
6207 The computed values for @var{x} and @var{y}. They are evaluated for
6212 The input width and height.
6216 These are the same as @var{in_w} and @var{in_h}.
6220 The output (cropped) width and height.
6224 These are the same as @var{out_w} and @var{out_h}.
6227 same as @var{iw} / @var{ih}
6230 input sample aspect ratio
6233 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6237 horizontal and vertical chroma subsample values. For example for the
6238 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6241 The number of the input frame, starting from 0.
6244 the position in the file of the input frame, NAN if unknown
6247 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6251 The expression for @var{out_w} may depend on the value of @var{out_h},
6252 and the expression for @var{out_h} may depend on @var{out_w}, but they
6253 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6254 evaluated after @var{out_w} and @var{out_h}.
6256 The @var{x} and @var{y} parameters specify the expressions for the
6257 position of the top-left corner of the output (non-cropped) area. They
6258 are evaluated for each frame. If the evaluated value is not valid, it
6259 is approximated to the nearest valid value.
6261 The expression for @var{x} may depend on @var{y}, and the expression
6262 for @var{y} may depend on @var{x}.
6264 @subsection Examples
6268 Crop area with size 100x100 at position (12,34).
6273 Using named options, the example above becomes:
6275 crop=w=100:h=100:x=12:y=34
6279 Crop the central input area with size 100x100:
6285 Crop the central input area with size 2/3 of the input video:
6287 crop=2/3*in_w:2/3*in_h
6291 Crop the input video central square:
6298 Delimit the rectangle with the top-left corner placed at position
6299 100:100 and the right-bottom corner corresponding to the right-bottom
6300 corner of the input image.
6302 crop=in_w-100:in_h-100:100:100
6306 Crop 10 pixels from the left and right borders, and 20 pixels from
6307 the top and bottom borders
6309 crop=in_w-2*10:in_h-2*20
6313 Keep only the bottom right quarter of the input image:
6315 crop=in_w/2:in_h/2:in_w/2:in_h/2
6319 Crop height for getting Greek harmony:
6321 crop=in_w:1/PHI*in_w
6325 Apply trembling effect:
6327 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)
6331 Apply erratic camera effect depending on timestamp:
6333 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)"
6337 Set x depending on the value of y:
6339 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6343 @subsection Commands
6345 This filter supports the following commands:
6351 Set width/height of the output video and the horizontal/vertical position
6353 The command accepts the same syntax of the corresponding option.
6355 If the specified expression is not valid, it is kept at its current
6361 Auto-detect the crop size.
6363 It calculates the necessary cropping parameters and prints the
6364 recommended parameters via the logging system. The detected dimensions
6365 correspond to the non-black area of the input video.
6367 It accepts the following parameters:
6372 Set higher black value threshold, which can be optionally specified
6373 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6374 value greater to the set value is considered non-black. It defaults to 24.
6375 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6376 on the bitdepth of the pixel format.
6379 The value which the width/height should be divisible by. It defaults to
6380 16. The offset is automatically adjusted to center the video. Use 2 to
6381 get only even dimensions (needed for 4:2:2 video). 16 is best when
6382 encoding to most video codecs.
6384 @item reset_count, reset
6385 Set the counter that determines after how many frames cropdetect will
6386 reset the previously detected largest video area and start over to
6387 detect the current optimal crop area. Default value is 0.
6389 This can be useful when channel logos distort the video area. 0
6390 indicates 'never reset', and returns the largest area encountered during
6397 Apply color adjustments using curves.
6399 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6400 component (red, green and blue) has its values defined by @var{N} key points
6401 tied from each other using a smooth curve. The x-axis represents the pixel
6402 values from the input frame, and the y-axis the new pixel values to be set for
6405 By default, a component curve is defined by the two points @var{(0;0)} and
6406 @var{(1;1)}. This creates a straight line where each original pixel value is
6407 "adjusted" to its own value, which means no change to the image.
6409 The filter allows you to redefine these two points and add some more. A new
6410 curve (using a natural cubic spline interpolation) will be define to pass
6411 smoothly through all these new coordinates. The new defined points needs to be
6412 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6413 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6414 the vector spaces, the values will be clipped accordingly.
6416 The filter accepts the following options:
6420 Select one of the available color presets. This option can be used in addition
6421 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6422 options takes priority on the preset values.
6423 Available presets are:
6426 @item color_negative
6429 @item increase_contrast
6431 @item linear_contrast
6432 @item medium_contrast
6434 @item strong_contrast
6437 Default is @code{none}.
6439 Set the master key points. These points will define a second pass mapping. It
6440 is sometimes called a "luminance" or "value" mapping. It can be used with
6441 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6442 post-processing LUT.
6444 Set the key points for the red component.
6446 Set the key points for the green component.
6448 Set the key points for the blue component.
6450 Set the key points for all components (not including master).
6451 Can be used in addition to the other key points component
6452 options. In this case, the unset component(s) will fallback on this
6453 @option{all} setting.
6455 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6457 Save Gnuplot script of the curves in specified file.
6460 To avoid some filtergraph syntax conflicts, each key points list need to be
6461 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6463 @subsection Examples
6467 Increase slightly the middle level of blue:
6469 curves=blue='0/0 0.5/0.58 1/1'
6475 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'
6477 Here we obtain the following coordinates for each components:
6480 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6482 @code{(0;0) (0.50;0.48) (1;1)}
6484 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6488 The previous example can also be achieved with the associated built-in preset:
6490 curves=preset=vintage
6500 Use a Photoshop preset and redefine the points of the green component:
6502 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6506 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6507 and @command{gnuplot}:
6509 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6510 gnuplot -p /tmp/curves.plt
6516 Video data analysis filter.
6518 This filter shows hexadecimal pixel values of part of video.
6520 The filter accepts the following options:
6524 Set output video size.
6527 Set x offset from where to pick pixels.
6530 Set y offset from where to pick pixels.
6533 Set scope mode, can be one of the following:
6536 Draw hexadecimal pixel values with white color on black background.
6539 Draw hexadecimal pixel values with input video pixel color on black
6543 Draw hexadecimal pixel values on color background picked from input video,
6544 the text color is picked in such way so its always visible.
6548 Draw rows and columns numbers on left and top of video.
6551 Set background opacity.
6556 Denoise frames using 2D DCT (frequency domain filtering).
6558 This filter is not designed for real time.
6560 The filter accepts the following options:
6564 Set the noise sigma constant.
6566 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6567 coefficient (absolute value) below this threshold with be dropped.
6569 If you need a more advanced filtering, see @option{expr}.
6571 Default is @code{0}.
6574 Set number overlapping pixels for each block. Since the filter can be slow, you
6575 may want to reduce this value, at the cost of a less effective filter and the
6576 risk of various artefacts.
6578 If the overlapping value doesn't permit processing the whole input width or
6579 height, a warning will be displayed and according borders won't be denoised.
6581 Default value is @var{blocksize}-1, which is the best possible setting.
6584 Set the coefficient factor expression.
6586 For each coefficient of a DCT block, this expression will be evaluated as a
6587 multiplier value for the coefficient.
6589 If this is option is set, the @option{sigma} option will be ignored.
6591 The absolute value of the coefficient can be accessed through the @var{c}
6595 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6596 @var{blocksize}, which is the width and height of the processed blocks.
6598 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6599 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6600 on the speed processing. Also, a larger block size does not necessarily means a
6604 @subsection Examples
6606 Apply a denoise with a @option{sigma} of @code{4.5}:
6611 The same operation can be achieved using the expression system:
6613 dctdnoiz=e='gte(c, 4.5*3)'
6616 Violent denoise using a block size of @code{16x16}:
6623 Remove banding artifacts from input video.
6624 It works by replacing banded pixels with average value of referenced pixels.
6626 The filter accepts the following options:
6633 Set banding detection threshold for each plane. Default is 0.02.
6634 Valid range is 0.00003 to 0.5.
6635 If difference between current pixel and reference pixel is less than threshold,
6636 it will be considered as banded.
6639 Banding detection range in pixels. Default is 16. If positive, random number
6640 in range 0 to set value will be used. If negative, exact absolute value
6642 The range defines square of four pixels around current pixel.
6645 Set direction in radians from which four pixel will be compared. If positive,
6646 random direction from 0 to set direction will be picked. If negative, exact of
6647 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6648 will pick only pixels on same row and -PI/2 will pick only pixels on same
6652 If enabled, current pixel is compared with average value of all four
6653 surrounding pixels. The default is enabled. If disabled current pixel is
6654 compared with all four surrounding pixels. The pixel is considered banded
6655 if only all four differences with surrounding pixels are less than threshold.
6658 If enabled, current pixel is changed if and only if all pixel components are banded,
6659 e.g. banding detection threshold is triggered for all color components.
6660 The default is disabled.
6666 Drop duplicated frames at regular intervals.
6668 The filter accepts the following options:
6672 Set the number of frames from which one will be dropped. Setting this to
6673 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6674 Default is @code{5}.
6677 Set the threshold for duplicate detection. If the difference metric for a frame
6678 is less than or equal to this value, then it is declared as duplicate. Default
6682 Set scene change threshold. Default is @code{15}.
6686 Set the size of the x and y-axis blocks used during metric calculations.
6687 Larger blocks give better noise suppression, but also give worse detection of
6688 small movements. Must be a power of two. Default is @code{32}.
6691 Mark main input as a pre-processed input and activate clean source input
6692 stream. This allows the input to be pre-processed with various filters to help
6693 the metrics calculation while keeping the frame selection lossless. When set to
6694 @code{1}, the first stream is for the pre-processed input, and the second
6695 stream is the clean source from where the kept frames are chosen. Default is
6699 Set whether or not chroma is considered in the metric calculations. Default is
6705 Apply deflate effect to the video.
6707 This filter replaces the pixel by the local(3x3) average by taking into account
6708 only values lower than the pixel.
6710 It accepts the following options:
6717 Limit the maximum change for each plane, default is 65535.
6718 If 0, plane will remain unchanged.
6723 Remove temporal frame luminance variations.
6725 It accepts the following options:
6729 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
6732 Set averaging mode to smooth temporal luminance variations.
6734 Available values are:
6759 Do not actually modify frame. Useful when one only wants metadata.
6764 Remove judder produced by partially interlaced telecined content.
6766 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6767 source was partially telecined content then the output of @code{pullup,dejudder}
6768 will have a variable frame rate. May change the recorded frame rate of the
6769 container. Aside from that change, this filter will not affect constant frame
6772 The option available in this filter is:
6776 Specify the length of the window over which the judder repeats.
6778 Accepts any integer greater than 1. Useful values are:
6782 If the original was telecined from 24 to 30 fps (Film to NTSC).
6785 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6788 If a mixture of the two.
6791 The default is @samp{4}.
6796 Suppress a TV station logo by a simple interpolation of the surrounding
6797 pixels. Just set a rectangle covering the logo and watch it disappear
6798 (and sometimes something even uglier appear - your mileage may vary).
6800 It accepts the following parameters:
6805 Specify the top left corner coordinates of the logo. They must be
6810 Specify the width and height of the logo to clear. They must be
6814 Specify the thickness of the fuzzy edge of the rectangle (added to
6815 @var{w} and @var{h}). The default value is 1. This option is
6816 deprecated, setting higher values should no longer be necessary and
6820 When set to 1, a green rectangle is drawn on the screen to simplify
6821 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6822 The default value is 0.
6824 The rectangle is drawn on the outermost pixels which will be (partly)
6825 replaced with interpolated values. The values of the next pixels
6826 immediately outside this rectangle in each direction will be used to
6827 compute the interpolated pixel values inside the rectangle.
6831 @subsection Examples
6835 Set a rectangle covering the area with top left corner coordinates 0,0
6836 and size 100x77, and a band of size 10:
6838 delogo=x=0:y=0:w=100:h=77:band=10
6845 Attempt to fix small changes in horizontal and/or vertical shift. This
6846 filter helps remove camera shake from hand-holding a camera, bumping a
6847 tripod, moving on a vehicle, etc.
6849 The filter accepts the following options:
6857 Specify a rectangular area where to limit the search for motion
6859 If desired the search for motion vectors can be limited to a
6860 rectangular area of the frame defined by its top left corner, width
6861 and height. These parameters have the same meaning as the drawbox
6862 filter which can be used to visualise the position of the bounding
6865 This is useful when simultaneous movement of subjects within the frame
6866 might be confused for camera motion by the motion vector search.
6868 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6869 then the full frame is used. This allows later options to be set
6870 without specifying the bounding box for the motion vector search.
6872 Default - search the whole frame.
6876 Specify the maximum extent of movement in x and y directions in the
6877 range 0-64 pixels. Default 16.
6880 Specify how to generate pixels to fill blanks at the edge of the
6881 frame. Available values are:
6884 Fill zeroes at blank locations
6886 Original image at blank locations
6888 Extruded edge value at blank locations
6890 Mirrored edge at blank locations
6892 Default value is @samp{mirror}.
6895 Specify the blocksize to use for motion search. Range 4-128 pixels,
6899 Specify the contrast threshold for blocks. Only blocks with more than
6900 the specified contrast (difference between darkest and lightest
6901 pixels) will be considered. Range 1-255, default 125.
6904 Specify the search strategy. Available values are:
6907 Set exhaustive search
6909 Set less exhaustive search.
6911 Default value is @samp{exhaustive}.
6914 If set then a detailed log of the motion search is written to the
6921 Remove unwanted contamination of foreground colors, caused by reflected color of
6922 greenscreen or bluescreen.
6924 This filter accepts the following options:
6928 Set what type of despill to use.
6931 Set how spillmap will be generated.
6934 Set how much to get rid of still remaining spill.
6937 Controls amount of red in spill area.
6940 Controls amount of green in spill area.
6941 Should be -1 for greenscreen.
6944 Controls amount of blue in spill area.
6945 Should be -1 for bluescreen.
6948 Controls brightness of spill area, preserving colors.
6951 Modify alpha from generated spillmap.
6956 Apply an exact inverse of the telecine operation. It requires a predefined
6957 pattern specified using the pattern option which must be the same as that passed
6958 to the telecine filter.
6960 This filter accepts the following options:
6969 The default value is @code{top}.
6973 A string of numbers representing the pulldown pattern you wish to apply.
6974 The default value is @code{23}.
6977 A number representing position of the first frame with respect to the telecine
6978 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6983 Apply dilation effect to the video.
6985 This filter replaces the pixel by the local(3x3) maximum.
6987 It accepts the following options:
6994 Limit the maximum change for each plane, default is 65535.
6995 If 0, plane will remain unchanged.
6998 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7001 Flags to local 3x3 coordinates maps like this:
7010 Displace pixels as indicated by second and third input stream.
7012 It takes three input streams and outputs one stream, the first input is the
7013 source, and second and third input are displacement maps.
7015 The second input specifies how much to displace pixels along the
7016 x-axis, while the third input specifies how much to displace pixels
7018 If one of displacement map streams terminates, last frame from that
7019 displacement map will be used.
7021 Note that once generated, displacements maps can be reused over and over again.
7023 A description of the accepted options follows.
7027 Set displace behavior for pixels that are out of range.
7029 Available values are:
7032 Missing pixels are replaced by black pixels.
7035 Adjacent pixels will spread out to replace missing pixels.
7038 Out of range pixels are wrapped so they point to pixels of other side.
7041 Out of range pixels will be replaced with mirrored pixels.
7043 Default is @samp{smear}.
7047 @subsection Examples
7051 Add ripple effect to rgb input of video size hd720:
7053 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
7057 Add wave effect to rgb input of video size hd720:
7059 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
7065 Draw a colored box on the input image.
7067 It accepts the following parameters:
7072 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7076 The expressions which specify the width and height of the box; if 0 they are interpreted as
7077 the input width and height. It defaults to 0.
7080 Specify the color of the box to write. For the general syntax of this option,
7081 check the "Color" section in the ffmpeg-utils manual. If the special
7082 value @code{invert} is used, the box edge color is the same as the
7083 video with inverted luma.
7086 The expression which sets the thickness of the box edge. Default value is @code{3}.
7088 See below for the list of accepted constants.
7091 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7092 following constants:
7096 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7100 horizontal and vertical chroma subsample values. For example for the
7101 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7105 The input width and height.
7108 The input sample aspect ratio.
7112 The x and y offset coordinates where the box is drawn.
7116 The width and height of the drawn box.
7119 The thickness of the drawn box.
7121 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7122 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7126 @subsection Examples
7130 Draw a black box around the edge of the input image:
7136 Draw a box with color red and an opacity of 50%:
7138 drawbox=10:20:200:60:red@@0.5
7141 The previous example can be specified as:
7143 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7147 Fill the box with pink color:
7149 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
7153 Draw a 2-pixel red 2.40:1 mask:
7155 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
7161 Draw a grid on the input image.
7163 It accepts the following parameters:
7168 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7172 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7173 input width and height, respectively, minus @code{thickness}, so image gets
7174 framed. Default to 0.
7177 Specify the color of the grid. For the general syntax of this option,
7178 check the "Color" section in the ffmpeg-utils manual. If the special
7179 value @code{invert} is used, the grid color is the same as the
7180 video with inverted luma.
7183 The expression which sets the thickness of the grid line. Default value is @code{1}.
7185 See below for the list of accepted constants.
7188 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7189 following constants:
7193 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7197 horizontal and vertical chroma subsample values. For example for the
7198 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7202 The input grid cell width and height.
7205 The input sample aspect ratio.
7209 The x and y coordinates of some point of grid intersection (meant to configure offset).
7213 The width and height of the drawn cell.
7216 The thickness of the drawn cell.
7218 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7219 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7223 @subsection Examples
7227 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7229 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7233 Draw a white 3x3 grid with an opacity of 50%:
7235 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7242 Draw a text string or text from a specified file on top of a video, using the
7243 libfreetype library.
7245 To enable compilation of this filter, you need to configure FFmpeg with
7246 @code{--enable-libfreetype}.
7247 To enable default font fallback and the @var{font} option you need to
7248 configure FFmpeg with @code{--enable-libfontconfig}.
7249 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7250 @code{--enable-libfribidi}.
7254 It accepts the following parameters:
7259 Used to draw a box around text using the background color.
7260 The value must be either 1 (enable) or 0 (disable).
7261 The default value of @var{box} is 0.
7264 Set the width of the border to be drawn around the box using @var{boxcolor}.
7265 The default value of @var{boxborderw} is 0.
7268 The color to be used for drawing box around text. For the syntax of this
7269 option, check the "Color" section in the ffmpeg-utils manual.
7271 The default value of @var{boxcolor} is "white".
7274 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7275 The default value of @var{line_spacing} is 0.
7278 Set the width of the border to be drawn around the text using @var{bordercolor}.
7279 The default value of @var{borderw} is 0.
7282 Set the color to be used for drawing border around text. For the syntax of this
7283 option, check the "Color" section in the ffmpeg-utils manual.
7285 The default value of @var{bordercolor} is "black".
7288 Select how the @var{text} is expanded. Can be either @code{none},
7289 @code{strftime} (deprecated) or
7290 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7294 Set a start time for the count. Value is in microseconds. Only applied
7295 in the deprecated strftime expansion mode. To emulate in normal expansion
7296 mode use the @code{pts} function, supplying the start time (in seconds)
7297 as the second argument.
7300 If true, check and fix text coords to avoid clipping.
7303 The color to be used for drawing fonts. For the syntax of this option, check
7304 the "Color" section in the ffmpeg-utils manual.
7306 The default value of @var{fontcolor} is "black".
7308 @item fontcolor_expr
7309 String which is expanded the same way as @var{text} to obtain dynamic
7310 @var{fontcolor} value. By default this option has empty value and is not
7311 processed. When this option is set, it overrides @var{fontcolor} option.
7314 The font family to be used for drawing text. By default Sans.
7317 The font file to be used for drawing text. The path must be included.
7318 This parameter is mandatory if the fontconfig support is disabled.
7321 Draw the text applying alpha blending. The value can
7322 be a number between 0.0 and 1.0.
7323 The expression accepts the same variables @var{x, y} as well.
7324 The default value is 1.
7325 Please see @var{fontcolor_expr}.
7328 The font size to be used for drawing text.
7329 The default value of @var{fontsize} is 16.
7332 If set to 1, attempt to shape the text (for example, reverse the order of
7333 right-to-left text and join Arabic characters) before drawing it.
7334 Otherwise, just draw the text exactly as given.
7335 By default 1 (if supported).
7338 The flags to be used for loading the fonts.
7340 The flags map the corresponding flags supported by libfreetype, and are
7341 a combination of the following values:
7348 @item vertical_layout
7349 @item force_autohint
7352 @item ignore_global_advance_width
7354 @item ignore_transform
7360 Default value is "default".
7362 For more information consult the documentation for the FT_LOAD_*
7366 The color to be used for drawing a shadow behind the drawn text. For the
7367 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
7369 The default value of @var{shadowcolor} is "black".
7373 The x and y offsets for the text shadow position with respect to the
7374 position of the text. They can be either positive or negative
7375 values. The default value for both is "0".
7378 The starting frame number for the n/frame_num variable. The default value
7382 The size in number of spaces to use for rendering the tab.
7386 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7387 format. It can be used with or without text parameter. @var{timecode_rate}
7388 option must be specified.
7390 @item timecode_rate, rate, r
7391 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7392 integer. Minimum value is "1".
7393 Drop-frame timecode is supported for frame rates 30 & 60.
7396 If set to 1, the output of the timecode option will wrap around at 24 hours.
7397 Default is 0 (disabled).
7400 The text string to be drawn. The text must be a sequence of UTF-8
7402 This parameter is mandatory if no file is specified with the parameter
7406 A text file containing text to be drawn. The text must be a sequence
7407 of UTF-8 encoded characters.
7409 This parameter is mandatory if no text string is specified with the
7410 parameter @var{text}.
7412 If both @var{text} and @var{textfile} are specified, an error is thrown.
7415 If set to 1, the @var{textfile} will be reloaded before each frame.
7416 Be sure to update it atomically, or it may be read partially, or even fail.
7420 The expressions which specify the offsets where text will be drawn
7421 within the video frame. They are relative to the top/left border of the
7424 The default value of @var{x} and @var{y} is "0".
7426 See below for the list of accepted constants and functions.
7429 The parameters for @var{x} and @var{y} are expressions containing the
7430 following constants and functions:
7434 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7438 horizontal and vertical chroma subsample values. For example for the
7439 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7442 the height of each text line
7450 @item max_glyph_a, ascent
7451 the maximum distance from the baseline to the highest/upper grid
7452 coordinate used to place a glyph outline point, for all the rendered
7454 It is a positive value, due to the grid's orientation with the Y axis
7457 @item max_glyph_d, descent
7458 the maximum distance from the baseline to the lowest grid coordinate
7459 used to place a glyph outline point, for all the rendered glyphs.
7460 This is a negative value, due to the grid's orientation, with the Y axis
7464 maximum glyph height, that is the maximum height for all the glyphs
7465 contained in the rendered text, it is equivalent to @var{ascent} -
7469 maximum glyph width, that is the maximum width for all the glyphs
7470 contained in the rendered text
7473 the number of input frame, starting from 0
7475 @item rand(min, max)
7476 return a random number included between @var{min} and @var{max}
7479 The input sample aspect ratio.
7482 timestamp expressed in seconds, NAN if the input timestamp is unknown
7485 the height of the rendered text
7488 the width of the rendered text
7492 the x and y offset coordinates where the text is drawn.
7494 These parameters allow the @var{x} and @var{y} expressions to refer
7495 each other, so you can for example specify @code{y=x/dar}.
7498 @anchor{drawtext_expansion}
7499 @subsection Text expansion
7501 If @option{expansion} is set to @code{strftime},
7502 the filter recognizes strftime() sequences in the provided text and
7503 expands them accordingly. Check the documentation of strftime(). This
7504 feature is deprecated.
7506 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7508 If @option{expansion} is set to @code{normal} (which is the default),
7509 the following expansion mechanism is used.
7511 The backslash character @samp{\}, followed by any character, always expands to
7512 the second character.
7514 Sequences of the form @code{%@{...@}} are expanded. The text between the
7515 braces is a function name, possibly followed by arguments separated by ':'.
7516 If the arguments contain special characters or delimiters (':' or '@}'),
7517 they should be escaped.
7519 Note that they probably must also be escaped as the value for the
7520 @option{text} option in the filter argument string and as the filter
7521 argument in the filtergraph description, and possibly also for the shell,
7522 that makes up to four levels of escaping; using a text file avoids these
7525 The following functions are available:
7530 The expression evaluation result.
7532 It must take one argument specifying the expression to be evaluated,
7533 which accepts the same constants and functions as the @var{x} and
7534 @var{y} values. Note that not all constants should be used, for
7535 example the text size is not known when evaluating the expression, so
7536 the constants @var{text_w} and @var{text_h} will have an undefined
7539 @item expr_int_format, eif
7540 Evaluate the expression's value and output as formatted integer.
7542 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7543 The second argument specifies the output format. Allowed values are @samp{x},
7544 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7545 @code{printf} function.
7546 The third parameter is optional and sets the number of positions taken by the output.
7547 It can be used to add padding with zeros from the left.
7550 The time at which the filter is running, expressed in UTC.
7551 It can accept an argument: a strftime() format string.
7554 The time at which the filter is running, expressed in the local time zone.
7555 It can accept an argument: a strftime() format string.
7558 Frame metadata. Takes one or two arguments.
7560 The first argument is mandatory and specifies the metadata key.
7562 The second argument is optional and specifies a default value, used when the
7563 metadata key is not found or empty.
7566 The frame number, starting from 0.
7569 A 1 character description of the current picture type.
7572 The timestamp of the current frame.
7573 It can take up to three arguments.
7575 The first argument is the format of the timestamp; it defaults to @code{flt}
7576 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7577 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7578 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7579 @code{localtime} stands for the timestamp of the frame formatted as
7580 local time zone time.
7582 The second argument is an offset added to the timestamp.
7584 If the format is set to @code{localtime} or @code{gmtime},
7585 a third argument may be supplied: a strftime() format string.
7586 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7589 @subsection Examples
7593 Draw "Test Text" with font FreeSerif, using the default values for the
7594 optional parameters.
7597 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7601 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7602 and y=50 (counting from the top-left corner of the screen), text is
7603 yellow with a red box around it. Both the text and the box have an
7607 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7608 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7611 Note that the double quotes are not necessary if spaces are not used
7612 within the parameter list.
7615 Show the text at the center of the video frame:
7617 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7621 Show the text at a random position, switching to a new position every 30 seconds:
7623 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)"
7627 Show a text line sliding from right to left in the last row of the video
7628 frame. The file @file{LONG_LINE} is assumed to contain a single line
7631 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7635 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7637 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7641 Draw a single green letter "g", at the center of the input video.
7642 The glyph baseline is placed at half screen height.
7644 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7648 Show text for 1 second every 3 seconds:
7650 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7654 Use fontconfig to set the font. Note that the colons need to be escaped.
7656 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7660 Print the date of a real-time encoding (see strftime(3)):
7662 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7666 Show text fading in and out (appearing/disappearing):
7669 DS=1.0 # display start
7670 DE=10.0 # display end
7671 FID=1.5 # fade in duration
7672 FOD=5 # fade out duration
7673 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 @}"
7677 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
7678 and the @option{fontsize} value are included in the @option{y} offset.
7680 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
7681 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
7686 For more information about libfreetype, check:
7687 @url{http://www.freetype.org/}.
7689 For more information about fontconfig, check:
7690 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7692 For more information about libfribidi, check:
7693 @url{http://fribidi.org/}.
7697 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7699 The filter accepts the following options:
7704 Set low and high threshold values used by the Canny thresholding
7707 The high threshold selects the "strong" edge pixels, which are then
7708 connected through 8-connectivity with the "weak" edge pixels selected
7709 by the low threshold.
7711 @var{low} and @var{high} threshold values must be chosen in the range
7712 [0,1], and @var{low} should be lesser or equal to @var{high}.
7714 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7718 Define the drawing mode.
7722 Draw white/gray wires on black background.
7725 Mix the colors to create a paint/cartoon effect.
7728 Default value is @var{wires}.
7731 @subsection Examples
7735 Standard edge detection with custom values for the hysteresis thresholding:
7737 edgedetect=low=0.1:high=0.4
7741 Painting effect without thresholding:
7743 edgedetect=mode=colormix:high=0
7748 Set brightness, contrast, saturation and approximate gamma adjustment.
7750 The filter accepts the following options:
7754 Set the contrast expression. The value must be a float value in range
7755 @code{-2.0} to @code{2.0}. The default value is "1".
7758 Set the brightness expression. The value must be a float value in
7759 range @code{-1.0} to @code{1.0}. The default value is "0".
7762 Set the saturation expression. The value must be a float in
7763 range @code{0.0} to @code{3.0}. The default value is "1".
7766 Set the gamma expression. The value must be a float in range
7767 @code{0.1} to @code{10.0}. The default value is "1".
7770 Set the gamma expression for red. The value must be a float in
7771 range @code{0.1} to @code{10.0}. The default value is "1".
7774 Set the gamma expression for green. The value must be a float in range
7775 @code{0.1} to @code{10.0}. The default value is "1".
7778 Set the gamma expression for blue. The value must be a float in range
7779 @code{0.1} to @code{10.0}. The default value is "1".
7782 Set the gamma weight expression. It can be used to reduce the effect
7783 of a high gamma value on bright image areas, e.g. keep them from
7784 getting overamplified and just plain white. The value must be a float
7785 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7786 gamma correction all the way down while @code{1.0} leaves it at its
7787 full strength. Default is "1".
7790 Set when the expressions for brightness, contrast, saturation and
7791 gamma expressions are evaluated.
7793 It accepts the following values:
7796 only evaluate expressions once during the filter initialization or
7797 when a command is processed
7800 evaluate expressions for each incoming frame
7803 Default value is @samp{init}.
7806 The expressions accept the following parameters:
7809 frame count of the input frame starting from 0
7812 byte position of the corresponding packet in the input file, NAN if
7816 frame rate of the input video, NAN if the input frame rate is unknown
7819 timestamp expressed in seconds, NAN if the input timestamp is unknown
7822 @subsection Commands
7823 The filter supports the following commands:
7827 Set the contrast expression.
7830 Set the brightness expression.
7833 Set the saturation expression.
7836 Set the gamma expression.
7839 Set the gamma_r expression.
7842 Set gamma_g expression.
7845 Set gamma_b expression.
7848 Set gamma_weight expression.
7850 The command accepts the same syntax of the corresponding option.
7852 If the specified expression is not valid, it is kept at its current
7859 Apply erosion effect to the video.
7861 This filter replaces the pixel by the local(3x3) minimum.
7863 It accepts the following options:
7870 Limit the maximum change for each plane, default is 65535.
7871 If 0, plane will remain unchanged.
7874 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7877 Flags to local 3x3 coordinates maps like this:
7884 @section extractplanes
7886 Extract color channel components from input video stream into
7887 separate grayscale video streams.
7889 The filter accepts the following option:
7893 Set plane(s) to extract.
7895 Available values for planes are:
7906 Choosing planes not available in the input will result in an error.
7907 That means you cannot select @code{r}, @code{g}, @code{b} planes
7908 with @code{y}, @code{u}, @code{v} planes at same time.
7911 @subsection Examples
7915 Extract luma, u and v color channel component from input video frame
7916 into 3 grayscale outputs:
7918 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
7924 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7926 For each input image, the filter will compute the optimal mapping from
7927 the input to the output given the codebook length, that is the number
7928 of distinct output colors.
7930 This filter accepts the following options.
7933 @item codebook_length, l
7934 Set codebook length. The value must be a positive integer, and
7935 represents the number of distinct output colors. Default value is 256.
7938 Set the maximum number of iterations to apply for computing the optimal
7939 mapping. The higher the value the better the result and the higher the
7940 computation time. Default value is 1.
7943 Set a random seed, must be an integer included between 0 and
7944 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7945 will try to use a good random seed on a best effort basis.
7948 Set pal8 output pixel format. This option does not work with codebook
7949 length greater than 256.
7954 Apply a fade-in/out effect to the input video.
7956 It accepts the following parameters:
7960 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7962 Default is @code{in}.
7964 @item start_frame, s
7965 Specify the number of the frame to start applying the fade
7966 effect at. Default is 0.
7969 The number of frames that the fade effect lasts. At the end of the
7970 fade-in effect, the output video will have the same intensity as the input video.
7971 At the end of the fade-out transition, the output video will be filled with the
7972 selected @option{color}.
7976 If set to 1, fade only alpha channel, if one exists on the input.
7979 @item start_time, st
7980 Specify the timestamp (in seconds) of the frame to start to apply the fade
7981 effect. If both start_frame and start_time are specified, the fade will start at
7982 whichever comes last. Default is 0.
7985 The number of seconds for which the fade effect has to last. At the end of the
7986 fade-in effect the output video will have the same intensity as the input video,
7987 at the end of the fade-out transition the output video will be filled with the
7988 selected @option{color}.
7989 If both duration and nb_frames are specified, duration is used. Default is 0
7990 (nb_frames is used by default).
7993 Specify the color of the fade. Default is "black".
7996 @subsection Examples
8000 Fade in the first 30 frames of video:
8005 The command above is equivalent to:
8011 Fade out the last 45 frames of a 200-frame video:
8014 fade=type=out:start_frame=155:nb_frames=45
8018 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8020 fade=in:0:25, fade=out:975:25
8024 Make the first 5 frames yellow, then fade in from frame 5-24:
8026 fade=in:5:20:color=yellow
8030 Fade in alpha over first 25 frames of video:
8032 fade=in:0:25:alpha=1
8036 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8038 fade=t=in:st=5.5:d=0.5
8044 Apply arbitrary expressions to samples in frequency domain
8048 Adjust the dc value (gain) of the luma plane of the image. The filter
8049 accepts an integer value in range @code{0} to @code{1000}. The default
8050 value is set to @code{0}.
8053 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8054 filter accepts an integer value in range @code{0} to @code{1000}. The
8055 default value is set to @code{0}.
8058 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8059 filter accepts an integer value in range @code{0} to @code{1000}. The
8060 default value is set to @code{0}.
8063 Set the frequency domain weight expression for the luma plane.
8066 Set the frequency domain weight expression for the 1st chroma plane.
8069 Set the frequency domain weight expression for the 2nd chroma plane.
8072 Set when the expressions are evaluated.
8074 It accepts the following values:
8077 Only evaluate expressions once during the filter initialization.
8080 Evaluate expressions for each incoming frame.
8083 Default value is @samp{init}.
8085 The filter accepts the following variables:
8088 The coordinates of the current sample.
8092 The width and height of the image.
8095 The number of input frame, starting from 0.
8098 @subsection Examples
8104 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8110 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8116 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8122 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8129 Extract a single field from an interlaced image using stride
8130 arithmetic to avoid wasting CPU time. The output frames are marked as
8133 The filter accepts the following options:
8137 Specify whether to extract the top (if the value is @code{0} or
8138 @code{top}) or the bottom field (if the value is @code{1} or
8144 Create new frames by copying the top and bottom fields from surrounding frames
8145 supplied as numbers by the hint file.
8149 Set file containing hints: absolute/relative frame numbers.
8151 There must be one line for each frame in a clip. Each line must contain two
8152 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8153 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8154 is current frame number for @code{absolute} mode or out of [-1, 1] range
8155 for @code{relative} mode. First number tells from which frame to pick up top
8156 field and second number tells from which frame to pick up bottom field.
8158 If optionally followed by @code{+} output frame will be marked as interlaced,
8159 else if followed by @code{-} output frame will be marked as progressive, else
8160 it will be marked same as input frame.
8161 If line starts with @code{#} or @code{;} that line is skipped.
8164 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8167 Example of first several lines of @code{hint} file for @code{relative} mode:
8170 1,0 - # second frame, use third's frame top field and second's frame bottom field
8171 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8188 Field matching filter for inverse telecine. It is meant to reconstruct the
8189 progressive frames from a telecined stream. The filter does not drop duplicated
8190 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8191 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8193 The separation of the field matching and the decimation is notably motivated by
8194 the possibility of inserting a de-interlacing filter fallback between the two.
8195 If the source has mixed telecined and real interlaced content,
8196 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8197 But these remaining combed frames will be marked as interlaced, and thus can be
8198 de-interlaced by a later filter such as @ref{yadif} before decimation.
8200 In addition to the various configuration options, @code{fieldmatch} can take an
8201 optional second stream, activated through the @option{ppsrc} option. If
8202 enabled, the frames reconstruction will be based on the fields and frames from
8203 this second stream. This allows the first input to be pre-processed in order to
8204 help the various algorithms of the filter, while keeping the output lossless
8205 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8206 or brightness/contrast adjustments can help.
8208 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8209 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8210 which @code{fieldmatch} is based on. While the semantic and usage are very
8211 close, some behaviour and options names can differ.
8213 The @ref{decimate} filter currently only works for constant frame rate input.
8214 If your input has mixed telecined (30fps) and progressive content with a lower
8215 framerate like 24fps use the following filterchain to produce the necessary cfr
8216 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8218 The filter accepts the following options:
8222 Specify the assumed field order of the input stream. Available values are:
8226 Auto detect parity (use FFmpeg's internal parity value).
8228 Assume bottom field first.
8230 Assume top field first.
8233 Note that it is sometimes recommended not to trust the parity announced by the
8236 Default value is @var{auto}.
8239 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8240 sense that it won't risk creating jerkiness due to duplicate frames when
8241 possible, but if there are bad edits or blended fields it will end up
8242 outputting combed frames when a good match might actually exist. On the other
8243 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8244 but will almost always find a good frame if there is one. The other values are
8245 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8246 jerkiness and creating duplicate frames versus finding good matches in sections
8247 with bad edits, orphaned fields, blended fields, etc.
8249 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8251 Available values are:
8255 2-way matching (p/c)
8257 2-way matching, and trying 3rd match if still combed (p/c + n)
8259 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8261 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8262 still combed (p/c + n + u/b)
8264 3-way matching (p/c/n)
8266 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8267 detected as combed (p/c/n + u/b)
8270 The parenthesis at the end indicate the matches that would be used for that
8271 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8274 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8277 Default value is @var{pc_n}.
8280 Mark the main input stream as a pre-processed input, and enable the secondary
8281 input stream as the clean source to pick the fields from. See the filter
8282 introduction for more details. It is similar to the @option{clip2} feature from
8285 Default value is @code{0} (disabled).
8288 Set the field to match from. It is recommended to set this to the same value as
8289 @option{order} unless you experience matching failures with that setting. In
8290 certain circumstances changing the field that is used to match from can have a
8291 large impact on matching performance. Available values are:
8295 Automatic (same value as @option{order}).
8297 Match from the bottom field.
8299 Match from the top field.
8302 Default value is @var{auto}.
8305 Set whether or not chroma is included during the match comparisons. In most
8306 cases it is recommended to leave this enabled. You should set this to @code{0}
8307 only if your clip has bad chroma problems such as heavy rainbowing or other
8308 artifacts. Setting this to @code{0} could also be used to speed things up at
8309 the cost of some accuracy.
8311 Default value is @code{1}.
8315 These define an exclusion band which excludes the lines between @option{y0} and
8316 @option{y1} from being included in the field matching decision. An exclusion
8317 band can be used to ignore subtitles, a logo, or other things that may
8318 interfere with the matching. @option{y0} sets the starting scan line and
8319 @option{y1} sets the ending line; all lines in between @option{y0} and
8320 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8321 @option{y0} and @option{y1} to the same value will disable the feature.
8322 @option{y0} and @option{y1} defaults to @code{0}.
8325 Set the scene change detection threshold as a percentage of maximum change on
8326 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8327 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8328 @option{scthresh} is @code{[0.0, 100.0]}.
8330 Default value is @code{12.0}.
8333 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8334 account the combed scores of matches when deciding what match to use as the
8335 final match. Available values are:
8339 No final matching based on combed scores.
8341 Combed scores are only used when a scene change is detected.
8343 Use combed scores all the time.
8346 Default is @var{sc}.
8349 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8350 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8351 Available values are:
8355 No forced calculation.
8357 Force p/c/n calculations.
8359 Force p/c/n/u/b calculations.
8362 Default value is @var{none}.
8365 This is the area combing threshold used for combed frame detection. This
8366 essentially controls how "strong" or "visible" combing must be to be detected.
8367 Larger values mean combing must be more visible and smaller values mean combing
8368 can be less visible or strong and still be detected. Valid settings are from
8369 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8370 be detected as combed). This is basically a pixel difference value. A good
8371 range is @code{[8, 12]}.
8373 Default value is @code{9}.
8376 Sets whether or not chroma is considered in the combed frame decision. Only
8377 disable this if your source has chroma problems (rainbowing, etc.) that are
8378 causing problems for the combed frame detection with chroma enabled. Actually,
8379 using @option{chroma}=@var{0} is usually more reliable, except for the case
8380 where there is chroma only combing in the source.
8382 Default value is @code{0}.
8386 Respectively set the x-axis and y-axis size of the window used during combed
8387 frame detection. This has to do with the size of the area in which
8388 @option{combpel} pixels are required to be detected as combed for a frame to be
8389 declared combed. See the @option{combpel} parameter description for more info.
8390 Possible values are any number that is a power of 2 starting at 4 and going up
8393 Default value is @code{16}.
8396 The number of combed pixels inside any of the @option{blocky} by
8397 @option{blockx} size blocks on the frame for the frame to be detected as
8398 combed. While @option{cthresh} controls how "visible" the combing must be, this
8399 setting controls "how much" combing there must be in any localized area (a
8400 window defined by the @option{blockx} and @option{blocky} settings) on the
8401 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8402 which point no frames will ever be detected as combed). This setting is known
8403 as @option{MI} in TFM/VFM vocabulary.
8405 Default value is @code{80}.
8408 @anchor{p/c/n/u/b meaning}
8409 @subsection p/c/n/u/b meaning
8411 @subsubsection p/c/n
8413 We assume the following telecined stream:
8416 Top fields: 1 2 2 3 4
8417 Bottom fields: 1 2 3 4 4
8420 The numbers correspond to the progressive frame the fields relate to. Here, the
8421 first two frames are progressive, the 3rd and 4th are combed, and so on.
8423 When @code{fieldmatch} is configured to run a matching from bottom
8424 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8429 B 1 2 3 4 4 <-- matching reference
8438 As a result of the field matching, we can see that some frames get duplicated.
8439 To perform a complete inverse telecine, you need to rely on a decimation filter
8440 after this operation. See for instance the @ref{decimate} filter.
8442 The same operation now matching from top fields (@option{field}=@var{top})
8447 T 1 2 2 3 4 <-- matching reference
8457 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8458 basically, they refer to the frame and field of the opposite parity:
8461 @item @var{p} matches the field of the opposite parity in the previous frame
8462 @item @var{c} matches the field of the opposite parity in the current frame
8463 @item @var{n} matches the field of the opposite parity in the next frame
8468 The @var{u} and @var{b} matching are a bit special in the sense that they match
8469 from the opposite parity flag. In the following examples, we assume that we are
8470 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8471 'x' is placed above and below each matched fields.
8473 With bottom matching (@option{field}=@var{bottom}):
8478 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8479 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8487 With top matching (@option{field}=@var{top}):
8492 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8493 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8501 @subsection Examples
8503 Simple IVTC of a top field first telecined stream:
8505 fieldmatch=order=tff:combmatch=none, decimate
8508 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8510 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8515 Transform the field order of the input video.
8517 It accepts the following parameters:
8522 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8523 for bottom field first.
8526 The default value is @samp{tff}.
8528 The transformation is done by shifting the picture content up or down
8529 by one line, and filling the remaining line with appropriate picture content.
8530 This method is consistent with most broadcast field order converters.
8532 If the input video is not flagged as being interlaced, or it is already
8533 flagged as being of the required output field order, then this filter does
8534 not alter the incoming video.
8536 It is very useful when converting to or from PAL DV material,
8537 which is bottom field first.
8541 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8544 @section fifo, afifo
8546 Buffer input images and send them when they are requested.
8548 It is mainly useful when auto-inserted by the libavfilter
8551 It does not take parameters.
8555 Find a rectangular object
8557 It accepts the following options:
8561 Filepath of the object image, needs to be in gray8.
8564 Detection threshold, default is 0.5.
8567 Number of mipmaps, default is 3.
8569 @item xmin, ymin, xmax, ymax
8570 Specifies the rectangle in which to search.
8573 @subsection Examples
8577 Generate a representative palette of a given video using @command{ffmpeg}:
8579 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8585 Cover a rectangular object
8587 It accepts the following options:
8591 Filepath of the optional cover image, needs to be in yuv420.
8596 It accepts the following values:
8599 cover it by the supplied image
8601 cover it by interpolating the surrounding pixels
8604 Default value is @var{blur}.
8607 @subsection Examples
8611 Generate a representative palette of a given video using @command{ffmpeg}:
8613 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8619 Flood area with values of same pixel components with another values.
8621 It accepts the following options:
8624 Set pixel x coordinate.
8627 Set pixel y coordinate.
8630 Set source #0 component value.
8633 Set source #1 component value.
8636 Set source #2 component value.
8639 Set source #3 component value.
8642 Set destination #0 component value.
8645 Set destination #1 component value.
8648 Set destination #2 component value.
8651 Set destination #3 component value.
8657 Convert the input video to one of the specified pixel formats.
8658 Libavfilter will try to pick one that is suitable as input to
8661 It accepts the following parameters:
8665 A '|'-separated list of pixel format names, such as
8666 "pix_fmts=yuv420p|monow|rgb24".
8670 @subsection Examples
8674 Convert the input video to the @var{yuv420p} format
8676 format=pix_fmts=yuv420p
8679 Convert the input video to any of the formats in the list
8681 format=pix_fmts=yuv420p|yuv444p|yuv410p
8688 Convert the video to specified constant frame rate by duplicating or dropping
8689 frames as necessary.
8691 It accepts the following parameters:
8695 The desired output frame rate. The default is @code{25}.
8698 Assume the first PTS should be the given value, in seconds. This allows for
8699 padding/trimming at the start of stream. By default, no assumption is made
8700 about the first frame's expected PTS, so no padding or trimming is done.
8701 For example, this could be set to 0 to pad the beginning with duplicates of
8702 the first frame if a video stream starts after the audio stream or to trim any
8703 frames with a negative PTS.
8706 Timestamp (PTS) rounding method.
8708 Possible values are:
8715 round towards -infinity
8717 round towards +infinity
8721 The default is @code{near}.
8724 Action performed when reading the last frame.
8726 Possible values are:
8729 Use same timestamp rounding method as used for other frames.
8731 Pass through last frame if input duration has not been reached yet.
8733 The default is @code{round}.
8737 Alternatively, the options can be specified as a flat string:
8738 @var{fps}[:@var{start_time}[:@var{round}]].
8740 See also the @ref{setpts} filter.
8742 @subsection Examples
8746 A typical usage in order to set the fps to 25:
8752 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8754 fps=fps=film:round=near
8760 Pack two different video streams into a stereoscopic video, setting proper
8761 metadata on supported codecs. The two views should have the same size and
8762 framerate and processing will stop when the shorter video ends. Please note
8763 that you may conveniently adjust view properties with the @ref{scale} and
8766 It accepts the following parameters:
8770 The desired packing format. Supported values are:
8775 The views are next to each other (default).
8778 The views are on top of each other.
8781 The views are packed by line.
8784 The views are packed by column.
8787 The views are temporally interleaved.
8796 # Convert left and right views into a frame-sequential video
8797 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8799 # Convert views into a side-by-side video with the same output resolution as the input
8800 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
8805 Change the frame rate by interpolating new video output frames from the source
8808 This filter is not designed to function correctly with interlaced media. If
8809 you wish to change the frame rate of interlaced media then you are required
8810 to deinterlace before this filter and re-interlace after this filter.
8812 A description of the accepted options follows.
8816 Specify the output frames per second. This option can also be specified
8817 as a value alone. The default is @code{50}.
8820 Specify the start of a range where the output frame will be created as a
8821 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8822 the default is @code{15}.
8825 Specify the end of a range where the output frame will be created as a
8826 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8827 the default is @code{240}.
8830 Specify the level at which a scene change is detected as a value between
8831 0 and 100 to indicate a new scene; a low value reflects a low
8832 probability for the current frame to introduce a new scene, while a higher
8833 value means the current frame is more likely to be one.
8834 The default is @code{7}.
8837 Specify flags influencing the filter process.
8839 Available value for @var{flags} is:
8842 @item scene_change_detect, scd
8843 Enable scene change detection using the value of the option @var{scene}.
8844 This flag is enabled by default.
8850 Select one frame every N-th frame.
8852 This filter accepts the following option:
8855 Select frame after every @code{step} frames.
8856 Allowed values are positive integers higher than 0. Default value is @code{1}.
8862 Apply a frei0r effect to the input video.
8864 To enable the compilation of this filter, you need to install the frei0r
8865 header and configure FFmpeg with @code{--enable-frei0r}.
8867 It accepts the following parameters:
8872 The name of the frei0r effect to load. If the environment variable
8873 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8874 directories specified by the colon-separated list in @env{FREI0R_PATH}.
8875 Otherwise, the standard frei0r paths are searched, in this order:
8876 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8877 @file{/usr/lib/frei0r-1/}.
8880 A '|'-separated list of parameters to pass to the frei0r effect.
8884 A frei0r effect parameter can be a boolean (its value is either
8885 "y" or "n"), a double, a color (specified as
8886 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8887 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8888 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8889 @var{X} and @var{Y} are floating point numbers) and/or a string.
8891 The number and types of parameters depend on the loaded effect. If an
8892 effect parameter is not specified, the default value is set.
8894 @subsection Examples
8898 Apply the distort0r effect, setting the first two double parameters:
8900 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8904 Apply the colordistance effect, taking a color as the first parameter:
8906 frei0r=colordistance:0.2/0.3/0.4
8907 frei0r=colordistance:violet
8908 frei0r=colordistance:0x112233
8912 Apply the perspective effect, specifying the top left and top right image
8915 frei0r=perspective:0.2/0.2|0.8/0.2
8919 For more information, see
8920 @url{http://frei0r.dyne.org}
8924 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8926 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8927 processing filter, one of them is performed once per block, not per pixel.
8928 This allows for much higher speed.
8930 The filter accepts the following options:
8934 Set quality. This option defines the number of levels for averaging. It accepts
8935 an integer in the range 4-5. Default value is @code{4}.
8938 Force a constant quantization parameter. It accepts an integer in range 0-63.
8939 If not set, the filter will use the QP from the video stream (if available).
8942 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8943 more details but also more artifacts, while higher values make the image smoother
8944 but also blurrier. Default value is @code{0} − PSNR optimal.
8947 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8948 option may cause flicker since the B-Frames have often larger QP. Default is
8949 @code{0} (not enabled).
8955 Apply Gaussian blur filter.
8957 The filter accepts the following options:
8961 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8964 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8967 Set which planes to filter. By default all planes are filtered.
8970 Set vertical sigma, if negative it will be same as @code{sigma}.
8971 Default is @code{-1}.
8976 The filter accepts the following options:
8980 Set the luminance expression.
8982 Set the chrominance blue expression.
8984 Set the chrominance red expression.
8986 Set the alpha expression.
8988 Set the red expression.
8990 Set the green expression.
8992 Set the blue expression.
8995 The colorspace is selected according to the specified options. If one
8996 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8997 options is specified, the filter will automatically select a YCbCr
8998 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8999 @option{blue_expr} options is specified, it will select an RGB
9002 If one of the chrominance expression is not defined, it falls back on the other
9003 one. If no alpha expression is specified it will evaluate to opaque value.
9004 If none of chrominance expressions are specified, they will evaluate
9005 to the luminance expression.
9007 The expressions can use the following variables and functions:
9011 The sequential number of the filtered frame, starting from @code{0}.
9015 The coordinates of the current sample.
9019 The width and height of the image.
9023 Width and height scale depending on the currently filtered plane. It is the
9024 ratio between the corresponding luma plane number of pixels and the current
9025 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9026 @code{0.5,0.5} for chroma planes.
9029 Time of the current frame, expressed in seconds.
9032 Return the value of the pixel at location (@var{x},@var{y}) of the current
9036 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9040 Return the value of the pixel at location (@var{x},@var{y}) of the
9041 blue-difference chroma plane. Return 0 if there is no such plane.
9044 Return the value of the pixel at location (@var{x},@var{y}) of the
9045 red-difference chroma plane. Return 0 if there is no such plane.
9050 Return the value of the pixel at location (@var{x},@var{y}) of the
9051 red/green/blue component. Return 0 if there is no such component.
9054 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9055 plane. Return 0 if there is no such plane.
9058 For functions, if @var{x} and @var{y} are outside the area, the value will be
9059 automatically clipped to the closer edge.
9061 @subsection Examples
9065 Flip the image horizontally:
9071 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9072 wavelength of 100 pixels:
9074 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9078 Generate a fancy enigmatic moving light:
9080 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
9084 Generate a quick emboss effect:
9086 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9090 Modify RGB components depending on pixel position:
9092 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9096 Create a radial gradient that is the same size as the input (also see
9097 the @ref{vignette} filter):
9099 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9105 Fix the banding artifacts that are sometimes introduced into nearly flat
9106 regions by truncation to 8-bit color depth.
9107 Interpolate the gradients that should go where the bands are, and
9110 It is designed for playback only. Do not use it prior to
9111 lossy compression, because compression tends to lose the dither and
9112 bring back the bands.
9114 It accepts the following parameters:
9119 The maximum amount by which the filter will change any one pixel. This is also
9120 the threshold for detecting nearly flat regions. Acceptable values range from
9121 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9125 The neighborhood to fit the gradient to. A larger radius makes for smoother
9126 gradients, but also prevents the filter from modifying the pixels near detailed
9127 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9128 values will be clipped to the valid range.
9132 Alternatively, the options can be specified as a flat string:
9133 @var{strength}[:@var{radius}]
9135 @subsection Examples
9139 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9145 Specify radius, omitting the strength (which will fall-back to the default
9156 Apply a Hald CLUT to a video stream.
9158 First input is the video stream to process, and second one is the Hald CLUT.
9159 The Hald CLUT input can be a simple picture or a complete video stream.
9161 The filter accepts the following options:
9165 Force termination when the shortest input terminates. Default is @code{0}.
9167 Continue applying the last CLUT after the end of the stream. A value of
9168 @code{0} disable the filter after the last frame of the CLUT is reached.
9169 Default is @code{1}.
9172 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9173 filters share the same internals).
9175 More information about the Hald CLUT can be found on Eskil Steenberg's website
9176 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9178 @subsection Workflow examples
9180 @subsubsection Hald CLUT video stream
9182 Generate an identity Hald CLUT stream altered with various effects:
9184 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
9187 Note: make sure you use a lossless codec.
9189 Then use it with @code{haldclut} to apply it on some random stream:
9191 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9194 The Hald CLUT will be applied to the 10 first seconds (duration of
9195 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9196 to the remaining frames of the @code{mandelbrot} stream.
9198 @subsubsection Hald CLUT with preview
9200 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9201 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9202 biggest possible square starting at the top left of the picture. The remaining
9203 padding pixels (bottom or right) will be ignored. This area can be used to add
9204 a preview of the Hald CLUT.
9206 Typically, the following generated Hald CLUT will be supported by the
9207 @code{haldclut} filter:
9210 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9211 pad=iw+320 [padded_clut];
9212 smptebars=s=320x256, split [a][b];
9213 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9214 [main][b] overlay=W-320" -frames:v 1 clut.png
9217 It contains the original and a preview of the effect of the CLUT: SMPTE color
9218 bars are displayed on the right-top, and below the same color bars processed by
9221 Then, the effect of this Hald CLUT can be visualized with:
9223 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9228 Flip the input video horizontally.
9230 For example, to horizontally flip the input video with @command{ffmpeg}:
9232 ffmpeg -i in.avi -vf "hflip" out.avi
9236 This filter applies a global color histogram equalization on a
9239 It can be used to correct video that has a compressed range of pixel
9240 intensities. The filter redistributes the pixel intensities to
9241 equalize their distribution across the intensity range. It may be
9242 viewed as an "automatically adjusting contrast filter". This filter is
9243 useful only for correcting degraded or poorly captured source
9246 The filter accepts the following options:
9250 Determine the amount of equalization to be applied. As the strength
9251 is reduced, the distribution of pixel intensities more-and-more
9252 approaches that of the input frame. The value must be a float number
9253 in the range [0,1] and defaults to 0.200.
9256 Set the maximum intensity that can generated and scale the output
9257 values appropriately. The strength should be set as desired and then
9258 the intensity can be limited if needed to avoid washing-out. The value
9259 must be a float number in the range [0,1] and defaults to 0.210.
9262 Set the antibanding level. If enabled the filter will randomly vary
9263 the luminance of output pixels by a small amount to avoid banding of
9264 the histogram. Possible values are @code{none}, @code{weak} or
9265 @code{strong}. It defaults to @code{none}.
9270 Compute and draw a color distribution histogram for the input video.
9272 The computed histogram is a representation of the color component
9273 distribution in an image.
9275 Standard histogram displays the color components distribution in an image.
9276 Displays color graph for each color component. Shows distribution of
9277 the Y, U, V, A or R, G, B components, depending on input format, in the
9278 current frame. Below each graph a color component scale meter is shown.
9280 The filter accepts the following options:
9284 Set height of level. Default value is @code{200}.
9285 Allowed range is [50, 2048].
9288 Set height of color scale. Default value is @code{12}.
9289 Allowed range is [0, 40].
9293 It accepts the following values:
9296 Per color component graphs are placed below each other.
9299 Per color component graphs are placed side by side.
9302 Presents information identical to that in the @code{parade}, except
9303 that the graphs representing color components are superimposed directly
9306 Default is @code{stack}.
9309 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9310 Default is @code{linear}.
9313 Set what color components to display.
9314 Default is @code{7}.
9317 Set foreground opacity. Default is @code{0.7}.
9320 Set background opacity. Default is @code{0.5}.
9323 @subsection Examples
9328 Calculate and draw histogram:
9330 ffplay -i input -vf histogram
9338 This is a high precision/quality 3d denoise filter. It aims to reduce
9339 image noise, producing smooth images and making still images really
9340 still. It should enhance compressibility.
9342 It accepts the following optional parameters:
9346 A non-negative floating point number which specifies spatial luma strength.
9349 @item chroma_spatial
9350 A non-negative floating point number which specifies spatial chroma strength.
9351 It defaults to 3.0*@var{luma_spatial}/4.0.
9354 A floating point number which specifies luma temporal strength. It defaults to
9355 6.0*@var{luma_spatial}/4.0.
9358 A floating point number which specifies chroma temporal strength. It defaults to
9359 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9364 Download hardware frames to system memory.
9366 The input must be in hardware frames, and the output a non-hardware format.
9367 Not all formats will be supported on the output - it may be necessary to insert
9368 an additional @option{format} filter immediately following in the graph to get
9369 the output in a supported format.
9373 Map hardware frames to system memory or to another device.
9375 This filter has several different modes of operation; which one is used depends
9376 on the input and output formats:
9379 Hardware frame input, normal frame output
9381 Map the input frames to system memory and pass them to the output. If the
9382 original hardware frame is later required (for example, after overlaying
9383 something else on part of it), the @option{hwmap} filter can be used again
9384 in the next mode to retrieve it.
9386 Normal frame input, hardware frame output
9388 If the input is actually a software-mapped hardware frame, then unmap it -
9389 that is, return the original hardware frame.
9391 Otherwise, a device must be provided. Create new hardware surfaces on that
9392 device for the output, then map them back to the software format at the input
9393 and give those frames to the preceding filter. This will then act like the
9394 @option{hwupload} filter, but may be able to avoid an additional copy when
9395 the input is already in a compatible format.
9397 Hardware frame input and output
9399 A device must be supplied for the output, either directly or with the
9400 @option{derive_device} option. The input and output devices must be of
9401 different types and compatible - the exact meaning of this is
9402 system-dependent, but typically it means that they must refer to the same
9403 underlying hardware context (for example, refer to the same graphics card).
9405 If the input frames were originally created on the output device, then unmap
9406 to retrieve the original frames.
9408 Otherwise, map the frames to the output device - create new hardware frames
9409 on the output corresponding to the frames on the input.
9412 The following additional parameters are accepted:
9416 Set the frame mapping mode. Some combination of:
9419 The mapped frame should be readable.
9421 The mapped frame should be writeable.
9423 The mapping will always overwrite the entire frame.
9425 This may improve performance in some cases, as the original contents of the
9426 frame need not be loaded.
9428 The mapping must not involve any copying.
9430 Indirect mappings to copies of frames are created in some cases where either
9431 direct mapping is not possible or it would have unexpected properties.
9432 Setting this flag ensures that the mapping is direct and will fail if that is
9435 Defaults to @var{read+write} if not specified.
9437 @item derive_device @var{type}
9438 Rather than using the device supplied at initialisation, instead derive a new
9439 device of type @var{type} from the device the input frames exist on.
9442 In a hardware to hardware mapping, map in reverse - create frames in the sink
9443 and map them back to the source. This may be necessary in some cases where
9444 a mapping in one direction is required but only the opposite direction is
9445 supported by the devices being used.
9447 This option is dangerous - it may break the preceding filter in undefined
9448 ways if there are any additional constraints on that filter's output.
9449 Do not use it without fully understanding the implications of its use.
9454 Upload system memory frames to hardware surfaces.
9456 The device to upload to must be supplied when the filter is initialised. If
9457 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9460 @anchor{hwupload_cuda}
9461 @section hwupload_cuda
9463 Upload system memory frames to a CUDA device.
9465 It accepts the following optional parameters:
9469 The number of the CUDA device to use
9474 Apply a high-quality magnification filter designed for pixel art. This filter
9475 was originally created by Maxim Stepin.
9477 It accepts the following option:
9481 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9482 @code{hq3x} and @code{4} for @code{hq4x}.
9483 Default is @code{3}.
9487 Stack input videos horizontally.
9489 All streams must be of same pixel format and of same height.
9491 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9492 to create same output.
9494 The filter accept the following option:
9498 Set number of input streams. Default is 2.
9501 If set to 1, force the output to terminate when the shortest input
9502 terminates. Default value is 0.
9507 Modify the hue and/or the saturation of the input.
9509 It accepts the following parameters:
9513 Specify the hue angle as a number of degrees. It accepts an expression,
9514 and defaults to "0".
9517 Specify the saturation in the [-10,10] range. It accepts an expression and
9521 Specify the hue angle as a number of radians. It accepts an
9522 expression, and defaults to "0".
9525 Specify the brightness in the [-10,10] range. It accepts an expression and
9529 @option{h} and @option{H} are mutually exclusive, and can't be
9530 specified at the same time.
9532 The @option{b}, @option{h}, @option{H} and @option{s} option values are
9533 expressions containing the following constants:
9537 frame count of the input frame starting from 0
9540 presentation timestamp of the input frame expressed in time base units
9543 frame rate of the input video, NAN if the input frame rate is unknown
9546 timestamp expressed in seconds, NAN if the input timestamp is unknown
9549 time base of the input video
9552 @subsection Examples
9556 Set the hue to 90 degrees and the saturation to 1.0:
9562 Same command but expressing the hue in radians:
9568 Rotate hue and make the saturation swing between 0
9569 and 2 over a period of 1 second:
9571 hue="H=2*PI*t: s=sin(2*PI*t)+1"
9575 Apply a 3 seconds saturation fade-in effect starting at 0:
9580 The general fade-in expression can be written as:
9582 hue="s=min(0\, max((t-START)/DURATION\, 1))"
9586 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
9588 hue="s=max(0\, min(1\, (8-t)/3))"
9591 The general fade-out expression can be written as:
9593 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
9598 @subsection Commands
9600 This filter supports the following commands:
9606 Modify the hue and/or the saturation and/or brightness of the input video.
9607 The command accepts the same syntax of the corresponding option.
9609 If the specified expression is not valid, it is kept at its current
9615 Grow first stream into second stream by connecting components.
9616 This makes it possible to build more robust edge masks.
9618 This filter accepts the following options:
9622 Set which planes will be processed as bitmap, unprocessed planes will be
9623 copied from first stream.
9624 By default value 0xf, all planes will be processed.
9627 Set threshold which is used in filtering. If pixel component value is higher than
9628 this value filter algorithm for connecting components is activated.
9629 By default value is 0.
9634 Detect video interlacing type.
9636 This filter tries to detect if the input frames are interlaced, progressive,
9637 top or bottom field first. It will also try to detect fields that are
9638 repeated between adjacent frames (a sign of telecine).
9640 Single frame detection considers only immediately adjacent frames when classifying each frame.
9641 Multiple frame detection incorporates the classification history of previous frames.
9643 The filter will log these metadata values:
9646 @item single.current_frame
9647 Detected type of current frame using single-frame detection. One of:
9648 ``tff'' (top field first), ``bff'' (bottom field first),
9649 ``progressive'', or ``undetermined''
9652 Cumulative number of frames detected as top field first using single-frame detection.
9655 Cumulative number of frames detected as top field first using multiple-frame detection.
9658 Cumulative number of frames detected as bottom field first using single-frame detection.
9660 @item multiple.current_frame
9661 Detected type of current frame using multiple-frame detection. One of:
9662 ``tff'' (top field first), ``bff'' (bottom field first),
9663 ``progressive'', or ``undetermined''
9666 Cumulative number of frames detected as bottom field first using multiple-frame detection.
9668 @item single.progressive
9669 Cumulative number of frames detected as progressive using single-frame detection.
9671 @item multiple.progressive
9672 Cumulative number of frames detected as progressive using multiple-frame detection.
9674 @item single.undetermined
9675 Cumulative number of frames that could not be classified using single-frame detection.
9677 @item multiple.undetermined
9678 Cumulative number of frames that could not be classified using multiple-frame detection.
9680 @item repeated.current_frame
9681 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
9683 @item repeated.neither
9684 Cumulative number of frames with no repeated field.
9687 Cumulative number of frames with the top field repeated from the previous frame's top field.
9689 @item repeated.bottom
9690 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
9693 The filter accepts the following options:
9697 Set interlacing threshold.
9699 Set progressive threshold.
9701 Threshold for repeated field detection.
9703 Number of frames after which a given frame's contribution to the
9704 statistics is halved (i.e., it contributes only 0.5 to its
9705 classification). The default of 0 means that all frames seen are given
9706 full weight of 1.0 forever.
9707 @item analyze_interlaced_flag
9708 When this is not 0 then idet will use the specified number of frames to determine
9709 if the interlaced flag is accurate, it will not count undetermined frames.
9710 If the flag is found to be accurate it will be used without any further
9711 computations, if it is found to be inaccurate it will be cleared without any
9712 further computations. This allows inserting the idet filter as a low computational
9713 method to clean up the interlaced flag
9718 Deinterleave or interleave fields.
9720 This filter allows one to process interlaced images fields without
9721 deinterlacing them. Deinterleaving splits the input frame into 2
9722 fields (so called half pictures). Odd lines are moved to the top
9723 half of the output image, even lines to the bottom half.
9724 You can process (filter) them independently and then re-interleave them.
9726 The filter accepts the following options:
9730 @item chroma_mode, c
9732 Available values for @var{luma_mode}, @var{chroma_mode} and
9733 @var{alpha_mode} are:
9739 @item deinterleave, d
9740 Deinterleave fields, placing one above the other.
9743 Interleave fields. Reverse the effect of deinterleaving.
9745 Default value is @code{none}.
9748 @item chroma_swap, cs
9749 @item alpha_swap, as
9750 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9755 Apply inflate effect to the video.
9757 This filter replaces the pixel by the local(3x3) average by taking into account
9758 only values higher than the pixel.
9760 It accepts the following options:
9767 Limit the maximum change for each plane, default is 65535.
9768 If 0, plane will remain unchanged.
9773 Simple interlacing filter from progressive contents. This interleaves upper (or
9774 lower) lines from odd frames with lower (or upper) lines from even frames,
9775 halving the frame rate and preserving image height.
9778 Original Original New Frame
9779 Frame 'j' Frame 'j+1' (tff)
9780 ========== =========== ==================
9781 Line 0 --------------------> Frame 'j' Line 0
9782 Line 1 Line 1 ----> Frame 'j+1' Line 1
9783 Line 2 ---------------------> Frame 'j' Line 2
9784 Line 3 Line 3 ----> Frame 'j+1' Line 3
9786 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9789 It accepts the following optional parameters:
9793 This determines whether the interlaced frame is taken from the even
9794 (tff - default) or odd (bff) lines of the progressive frame.
9797 Vertical lowpass filter to avoid twitter interlacing and
9798 reduce moire patterns.
9802 Disable vertical lowpass filter
9805 Enable linear filter (default)
9808 Enable complex filter. This will slightly less reduce twitter and moire
9809 but better retain detail and subjective sharpness impression.
9816 Deinterlace input video by applying Donald Graft's adaptive kernel
9817 deinterling. Work on interlaced parts of a video to produce
9820 The description of the accepted parameters follows.
9824 Set the threshold which affects the filter's tolerance when
9825 determining if a pixel line must be processed. It must be an integer
9826 in the range [0,255] and defaults to 10. A value of 0 will result in
9827 applying the process on every pixels.
9830 Paint pixels exceeding the threshold value to white if set to 1.
9834 Set the fields order. Swap fields if set to 1, leave fields alone if
9838 Enable additional sharpening if set to 1. Default is 0.
9841 Enable twoway sharpening if set to 1. Default is 0.
9844 @subsection Examples
9848 Apply default values:
9850 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9854 Enable additional sharpening:
9860 Paint processed pixels in white:
9866 @section lenscorrection
9868 Correct radial lens distortion
9870 This filter can be used to correct for radial distortion as can result from the use
9871 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9872 one can use tools available for example as part of opencv or simply trial-and-error.
9873 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9874 and extract the k1 and k2 coefficients from the resulting matrix.
9876 Note that effectively the same filter is available in the open-source tools Krita and
9877 Digikam from the KDE project.
9879 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9880 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9881 brightness distribution, so you may want to use both filters together in certain
9882 cases, though you will have to take care of ordering, i.e. whether vignetting should
9883 be applied before or after lens correction.
9887 The filter accepts the following options:
9891 Relative x-coordinate of the focal point of the image, and thereby the center of the
9892 distortion. This value has a range [0,1] and is expressed as fractions of the image
9895 Relative y-coordinate of the focal point of the image, and thereby the center of the
9896 distortion. This value has a range [0,1] and is expressed as fractions of the image
9899 Coefficient of the quadratic correction term. 0.5 means no correction.
9901 Coefficient of the double quadratic correction term. 0.5 means no correction.
9904 The formula that generates the correction is:
9906 @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)
9908 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9909 distances from the focal point in the source and target images, respectively.
9913 Obtain the VMAF (Video Multi-Method Assessment Fusion)
9914 score between two input videos.
9916 The obtained VMAF score is printed through the logging system.
9918 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
9919 After installing the library it can be enabled using:
9920 @code{./configure --enable-libvmaf}.
9921 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
9923 The filter has following options:
9927 Set the model path which is to be used for SVM.
9928 Default value: @code{"vmaf_v0.6.1.pkl"}
9931 Set the file path to be used to store logs.
9934 Set the format of the log file (xml or json).
9936 @item enable_transform
9937 Enables transform for computing vmaf.
9940 Invokes the phone model which will generate VMAF scores higher than in the
9941 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
9944 Enables computing psnr along with vmaf.
9947 Enables computing ssim along with vmaf.
9950 Enables computing ms_ssim along with vmaf.
9953 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
9956 This filter also supports the @ref{framesync} options.
9958 On the below examples the input file @file{main.mpg} being processed is
9959 compared with the reference file @file{ref.mpg}.
9962 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
9965 Example with options:
9967 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
9972 Limits the pixel components values to the specified range [min, max].
9974 The filter accepts the following options:
9978 Lower bound. Defaults to the lowest allowed value for the input.
9981 Upper bound. Defaults to the highest allowed value for the input.
9984 Specify which planes will be processed. Defaults to all available.
9991 The filter accepts the following options:
9995 Set the number of loops. Setting this value to -1 will result in infinite loops.
9999 Set maximal size in number of frames. Default is 0.
10002 Set first frame of loop. Default is 0.
10008 Apply a 3D LUT to an input video.
10010 The filter accepts the following options:
10014 Set the 3D LUT file name.
10016 Currently supported formats:
10028 Select interpolation mode.
10030 Available values are:
10034 Use values from the nearest defined point.
10036 Interpolate values using the 8 points defining a cube.
10038 Interpolate values using a tetrahedron.
10042 This filter also supports the @ref{framesync} options.
10046 Turn certain luma values into transparency.
10048 The filter accepts the following options:
10052 Set the luma which will be used as base for transparency.
10053 Default value is @code{0}.
10056 Set the range of luma values to be keyed out.
10057 Default value is @code{0}.
10060 Set the range of softness. Default value is @code{0}.
10061 Use this to control gradual transition from zero to full transparency.
10064 @section lut, lutrgb, lutyuv
10066 Compute a look-up table for binding each pixel component input value
10067 to an output value, and apply it to the input video.
10069 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10070 to an RGB input video.
10072 These filters accept the following parameters:
10075 set first pixel component expression
10077 set second pixel component expression
10079 set third pixel component expression
10081 set fourth pixel component expression, corresponds to the alpha component
10084 set red component expression
10086 set green component expression
10088 set blue component expression
10090 alpha component expression
10093 set Y/luminance component expression
10095 set U/Cb component expression
10097 set V/Cr component expression
10100 Each of them specifies the expression to use for computing the lookup table for
10101 the corresponding pixel component values.
10103 The exact component associated to each of the @var{c*} options depends on the
10106 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10107 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10109 The expressions can contain the following constants and functions:
10114 The input width and height.
10117 The input value for the pixel component.
10120 The input value, clipped to the @var{minval}-@var{maxval} range.
10123 The maximum value for the pixel component.
10126 The minimum value for the pixel component.
10129 The negated value for the pixel component value, clipped to the
10130 @var{minval}-@var{maxval} range; it corresponds to the expression
10131 "maxval-clipval+minval".
10134 The computed value in @var{val}, clipped to the
10135 @var{minval}-@var{maxval} range.
10137 @item gammaval(gamma)
10138 The computed gamma correction value of the pixel component value,
10139 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10141 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10145 All expressions default to "val".
10147 @subsection Examples
10151 Negate input video:
10153 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10154 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10157 The above is the same as:
10159 lutrgb="r=negval:g=negval:b=negval"
10160 lutyuv="y=negval:u=negval:v=negval"
10170 Remove chroma components, turning the video into a graytone image:
10172 lutyuv="u=128:v=128"
10176 Apply a luma burning effect:
10182 Remove green and blue components:
10188 Set a constant alpha channel value on input:
10190 format=rgba,lutrgb=a="maxval-minval/2"
10194 Correct luminance gamma by a factor of 0.5:
10196 lutyuv=y=gammaval(0.5)
10200 Discard least significant bits of luma:
10202 lutyuv=y='bitand(val, 128+64+32)'
10206 Technicolor like effect:
10208 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10212 @section lut2, tlut2
10214 The @code{lut2} filter takes two input streams and outputs one
10217 The @code{tlut2} (time lut2) filter takes two consecutive frames
10218 from one single stream.
10220 This filter accepts the following parameters:
10223 set first pixel component expression
10225 set second pixel component expression
10227 set third pixel component expression
10229 set fourth pixel component expression, corresponds to the alpha component
10232 Each of them specifies the expression to use for computing the lookup table for
10233 the corresponding pixel component values.
10235 The exact component associated to each of the @var{c*} options depends on the
10238 The expressions can contain the following constants:
10243 The input width and height.
10246 The first input value for the pixel component.
10249 The second input value for the pixel component.
10252 The first input video bit depth.
10255 The second input video bit depth.
10258 All expressions default to "x".
10260 @subsection Examples
10264 Highlight differences between two RGB video streams:
10266 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)'
10270 Highlight differences between two YUV video streams:
10272 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)'
10276 Show max difference between two video streams:
10278 lut2='if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1)))'
10282 @section maskedclamp
10284 Clamp the first input stream with the second input and third input stream.
10286 Returns the value of first stream to be between second input
10287 stream - @code{undershoot} and third input stream + @code{overshoot}.
10289 This filter accepts the following options:
10292 Default value is @code{0}.
10295 Default value is @code{0}.
10298 Set which planes will be processed as bitmap, unprocessed planes will be
10299 copied from first stream.
10300 By default value 0xf, all planes will be processed.
10303 @section maskedmerge
10305 Merge the first input stream with the second input stream using per pixel
10306 weights in the third input stream.
10308 A value of 0 in the third stream pixel component means that pixel component
10309 from first stream is returned unchanged, while maximum value (eg. 255 for
10310 8-bit videos) means that pixel component from second stream is returned
10311 unchanged. Intermediate values define the amount of merging between both
10312 input stream's pixel components.
10314 This filter accepts the following options:
10317 Set which planes will be processed as bitmap, unprocessed planes will be
10318 copied from first stream.
10319 By default value 0xf, all planes will be processed.
10324 Apply motion-compensation deinterlacing.
10326 It needs one field per frame as input and must thus be used together
10327 with yadif=1/3 or equivalent.
10329 This filter accepts the following options:
10332 Set the deinterlacing mode.
10334 It accepts one of the following values:
10339 use iterative motion estimation
10341 like @samp{slow}, but use multiple reference frames.
10343 Default value is @samp{fast}.
10346 Set the picture field parity assumed for the input video. It must be
10347 one of the following values:
10351 assume top field first
10353 assume bottom field first
10356 Default value is @samp{bff}.
10359 Set per-block quantization parameter (QP) used by the internal
10362 Higher values should result in a smoother motion vector field but less
10363 optimal individual vectors. Default value is 1.
10366 @section mergeplanes
10368 Merge color channel components from several video streams.
10370 The filter accepts up to 4 input streams, and merge selected input
10371 planes to the output video.
10373 This filter accepts the following options:
10376 Set input to output plane mapping. Default is @code{0}.
10378 The mappings is specified as a bitmap. It should be specified as a
10379 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10380 mapping for the first plane of the output stream. 'A' sets the number of
10381 the input stream to use (from 0 to 3), and 'a' the plane number of the
10382 corresponding input to use (from 0 to 3). The rest of the mappings is
10383 similar, 'Bb' describes the mapping for the output stream second
10384 plane, 'Cc' describes the mapping for the output stream third plane and
10385 'Dd' describes the mapping for the output stream fourth plane.
10388 Set output pixel format. Default is @code{yuva444p}.
10391 @subsection Examples
10395 Merge three gray video streams of same width and height into single video stream:
10397 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10401 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10403 [a0][a1]mergeplanes=0x00010210:yuva444p
10407 Swap Y and A plane in yuva444p stream:
10409 format=yuva444p,mergeplanes=0x03010200:yuva444p
10413 Swap U and V plane in yuv420p stream:
10415 format=yuv420p,mergeplanes=0x000201:yuv420p
10419 Cast a rgb24 clip to yuv444p:
10421 format=rgb24,mergeplanes=0x000102:yuv444p
10427 Estimate and export motion vectors using block matching algorithms.
10428 Motion vectors are stored in frame side data to be used by other filters.
10430 This filter accepts the following options:
10433 Specify the motion estimation method. Accepts one of the following values:
10437 Exhaustive search algorithm.
10439 Three step search algorithm.
10441 Two dimensional logarithmic search algorithm.
10443 New three step search algorithm.
10445 Four step search algorithm.
10447 Diamond search algorithm.
10449 Hexagon-based search algorithm.
10451 Enhanced predictive zonal search algorithm.
10453 Uneven multi-hexagon search algorithm.
10455 Default value is @samp{esa}.
10458 Macroblock size. Default @code{16}.
10461 Search parameter. Default @code{7}.
10464 @section midequalizer
10466 Apply Midway Image Equalization effect using two video streams.
10468 Midway Image Equalization adjusts a pair of images to have the same
10469 histogram, while maintaining their dynamics as much as possible. It's
10470 useful for e.g. matching exposures from a pair of stereo cameras.
10472 This filter has two inputs and one output, which must be of same pixel format, but
10473 may be of different sizes. The output of filter is first input adjusted with
10474 midway histogram of both inputs.
10476 This filter accepts the following option:
10480 Set which planes to process. Default is @code{15}, which is all available planes.
10483 @section minterpolate
10485 Convert the video to specified frame rate using motion interpolation.
10487 This filter accepts the following options:
10490 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}.
10493 Motion interpolation mode. Following values are accepted:
10496 Duplicate previous or next frame for interpolating new ones.
10498 Blend source frames. Interpolated frame is mean of previous and next frames.
10500 Motion compensated interpolation. Following options are effective when this mode is selected:
10504 Motion compensation mode. Following values are accepted:
10507 Overlapped block motion compensation.
10509 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10511 Default mode is @samp{obmc}.
10514 Motion estimation mode. Following values are accepted:
10517 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
10519 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
10521 Default mode is @samp{bilat}.
10524 The algorithm to be used for motion estimation. Following values are accepted:
10527 Exhaustive search algorithm.
10529 Three step search algorithm.
10531 Two dimensional logarithmic search algorithm.
10533 New three step search algorithm.
10535 Four step search algorithm.
10537 Diamond search algorithm.
10539 Hexagon-based search algorithm.
10541 Enhanced predictive zonal search algorithm.
10543 Uneven multi-hexagon search algorithm.
10545 Default algorithm is @samp{epzs}.
10548 Macroblock size. Default @code{16}.
10551 Motion estimation search parameter. Default @code{32}.
10554 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).
10559 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:
10562 Disable scene change detection.
10564 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
10566 Default method is @samp{fdiff}.
10568 @item scd_threshold
10569 Scene change detection threshold. Default is @code{5.0}.
10574 Mix several video input streams into one video stream.
10576 A description of the accepted options follows.
10580 The number of inputs. If unspecified, it defaults to 2.
10583 Specify weight of each input video stream as sequence.
10584 Each weight is separated by space.
10587 Specify how end of stream is determined.
10590 The duration of the longest input. (default)
10593 The duration of the shortest input.
10596 The duration of the first input.
10600 @section mpdecimate
10602 Drop frames that do not differ greatly from the previous frame in
10603 order to reduce frame rate.
10605 The main use of this filter is for very-low-bitrate encoding
10606 (e.g. streaming over dialup modem), but it could in theory be used for
10607 fixing movies that were inverse-telecined incorrectly.
10609 A description of the accepted options follows.
10613 Set the maximum number of consecutive frames which can be dropped (if
10614 positive), or the minimum interval between dropped frames (if
10615 negative). If the value is 0, the frame is dropped disregarding the
10616 number of previous sequentially dropped frames.
10618 Default value is 0.
10623 Set the dropping threshold values.
10625 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
10626 represent actual pixel value differences, so a threshold of 64
10627 corresponds to 1 unit of difference for each pixel, or the same spread
10628 out differently over the block.
10630 A frame is a candidate for dropping if no 8x8 blocks differ by more
10631 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
10632 meaning the whole image) differ by more than a threshold of @option{lo}.
10634 Default value for @option{hi} is 64*12, default value for @option{lo} is
10635 64*5, and default value for @option{frac} is 0.33.
10641 Negate input video.
10643 It accepts an integer in input; if non-zero it negates the
10644 alpha component (if available). The default value in input is 0.
10648 Denoise frames using Non-Local Means algorithm.
10650 Each pixel is adjusted by looking for other pixels with similar contexts. This
10651 context similarity is defined by comparing their surrounding patches of size
10652 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
10655 Note that the research area defines centers for patches, which means some
10656 patches will be made of pixels outside that research area.
10658 The filter accepts the following options.
10662 Set denoising strength.
10668 Same as @option{p} but for chroma planes.
10670 The default value is @var{0} and means automatic.
10676 Same as @option{r} but for chroma planes.
10678 The default value is @var{0} and means automatic.
10683 Deinterlace video using neural network edge directed interpolation.
10685 This filter accepts the following options:
10689 Mandatory option, without binary file filter can not work.
10690 Currently file can be found here:
10691 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
10694 Set which frames to deinterlace, by default it is @code{all}.
10695 Can be @code{all} or @code{interlaced}.
10698 Set mode of operation.
10700 Can be one of the following:
10704 Use frame flags, both fields.
10706 Use frame flags, single field.
10708 Use top field only.
10710 Use bottom field only.
10712 Use both fields, top first.
10714 Use both fields, bottom first.
10718 Set which planes to process, by default filter process all frames.
10721 Set size of local neighborhood around each pixel, used by the predictor neural
10724 Can be one of the following:
10737 Set the number of neurons in predictor neural network.
10738 Can be one of the following:
10749 Controls the number of different neural network predictions that are blended
10750 together to compute the final output value. Can be @code{fast}, default or
10754 Set which set of weights to use in the predictor.
10755 Can be one of the following:
10759 weights trained to minimize absolute error
10761 weights trained to minimize squared error
10765 Controls whether or not the prescreener neural network is used to decide
10766 which pixels should be processed by the predictor neural network and which
10767 can be handled by simple cubic interpolation.
10768 The prescreener is trained to know whether cubic interpolation will be
10769 sufficient for a pixel or whether it should be predicted by the predictor nn.
10770 The computational complexity of the prescreener nn is much less than that of
10771 the predictor nn. Since most pixels can be handled by cubic interpolation,
10772 using the prescreener generally results in much faster processing.
10773 The prescreener is pretty accurate, so the difference between using it and not
10774 using it is almost always unnoticeable.
10776 Can be one of the following:
10784 Default is @code{new}.
10787 Set various debugging flags.
10792 Force libavfilter not to use any of the specified pixel formats for the
10793 input to the next filter.
10795 It accepts the following parameters:
10799 A '|'-separated list of pixel format names, such as
10800 pix_fmts=yuv420p|monow|rgb24".
10804 @subsection Examples
10808 Force libavfilter to use a format different from @var{yuv420p} for the
10809 input to the vflip filter:
10811 noformat=pix_fmts=yuv420p,vflip
10815 Convert the input video to any of the formats not contained in the list:
10817 noformat=yuv420p|yuv444p|yuv410p
10823 Add noise on video input frame.
10825 The filter accepts the following options:
10833 Set noise seed for specific pixel component or all pixel components in case
10834 of @var{all_seed}. Default value is @code{123457}.
10836 @item all_strength, alls
10837 @item c0_strength, c0s
10838 @item c1_strength, c1s
10839 @item c2_strength, c2s
10840 @item c3_strength, c3s
10841 Set noise strength for specific pixel component or all pixel components in case
10842 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
10844 @item all_flags, allf
10845 @item c0_flags, c0f
10846 @item c1_flags, c1f
10847 @item c2_flags, c2f
10848 @item c3_flags, c3f
10849 Set pixel component flags or set flags for all components if @var{all_flags}.
10850 Available values for component flags are:
10853 averaged temporal noise (smoother)
10855 mix random noise with a (semi)regular pattern
10857 temporal noise (noise pattern changes between frames)
10859 uniform noise (gaussian otherwise)
10863 @subsection Examples
10865 Add temporal and uniform noise to input video:
10867 noise=alls=20:allf=t+u
10872 Normalize RGB video (aka histogram stretching, contrast stretching).
10873 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
10875 For each channel of each frame, the filter computes the input range and maps
10876 it linearly to the user-specified output range. The output range defaults
10877 to the full dynamic range from pure black to pure white.
10879 Temporal smoothing can be used on the input range to reduce flickering (rapid
10880 changes in brightness) caused when small dark or bright objects enter or leave
10881 the scene. This is similar to the auto-exposure (automatic gain control) on a
10882 video camera, and, like a video camera, it may cause a period of over- or
10883 under-exposure of the video.
10885 The R,G,B channels can be normalized independently, which may cause some
10886 color shifting, or linked together as a single channel, which prevents
10887 color shifting. Linked normalization preserves hue. Independent normalization
10888 does not, so it can be used to remove some color casts. Independent and linked
10889 normalization can be combined in any ratio.
10891 The normalize filter accepts the following options:
10896 Colors which define the output range. The minimum input value is mapped to
10897 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
10898 The defaults are black and white respectively. Specifying white for
10899 @var{blackpt} and black for @var{whitept} will give color-inverted,
10900 normalized video. Shades of grey can be used to reduce the dynamic range
10901 (contrast). Specifying saturated colors here can create some interesting
10905 The number of previous frames to use for temporal smoothing. The input range
10906 of each channel is smoothed using a rolling average over the current frame
10907 and the @var{smoothing} previous frames. The default is 0 (no temporal
10911 Controls the ratio of independent (color shifting) channel normalization to
10912 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
10913 independent. Defaults to 1.0 (fully independent).
10916 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
10917 expensive no-op. Defaults to 1.0 (full strength).
10921 @subsection Examples
10923 Stretch video contrast to use the full dynamic range, with no temporal
10924 smoothing; may flicker depending on the source content:
10926 normalize=blackpt=black:whitept=white:smoothing=0
10929 As above, but with 50 frames of temporal smoothing; flicker should be
10930 reduced, depending on the source content:
10932 normalize=blackpt=black:whitept=white:smoothing=50
10935 As above, but with hue-preserving linked channel normalization:
10937 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
10940 As above, but with half strength:
10942 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
10945 Map the darkest input color to red, the brightest input color to cyan:
10947 normalize=blackpt=red:whitept=cyan
10952 Pass the video source unchanged to the output.
10955 Optical Character Recognition
10957 This filter uses Tesseract for optical character recognition.
10959 It accepts the following options:
10963 Set datapath to tesseract data. Default is to use whatever was
10964 set at installation.
10967 Set language, default is "eng".
10970 Set character whitelist.
10973 Set character blacklist.
10976 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
10980 Apply a video transform using libopencv.
10982 To enable this filter, install the libopencv library and headers and
10983 configure FFmpeg with @code{--enable-libopencv}.
10985 It accepts the following parameters:
10990 The name of the libopencv filter to apply.
10992 @item filter_params
10993 The parameters to pass to the libopencv filter. If not specified, the default
10994 values are assumed.
10998 Refer to the official libopencv documentation for more precise
11000 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11002 Several libopencv filters are supported; see the following subsections.
11007 Dilate an image by using a specific structuring element.
11008 It corresponds to the libopencv function @code{cvDilate}.
11010 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11012 @var{struct_el} represents a structuring element, and has the syntax:
11013 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11015 @var{cols} and @var{rows} represent the number of columns and rows of
11016 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11017 point, and @var{shape} the shape for the structuring element. @var{shape}
11018 must be "rect", "cross", "ellipse", or "custom".
11020 If the value for @var{shape} is "custom", it must be followed by a
11021 string of the form "=@var{filename}". The file with name
11022 @var{filename} is assumed to represent a binary image, with each
11023 printable character corresponding to a bright pixel. When a custom
11024 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11025 or columns and rows of the read file are assumed instead.
11027 The default value for @var{struct_el} is "3x3+0x0/rect".
11029 @var{nb_iterations} specifies the number of times the transform is
11030 applied to the image, and defaults to 1.
11034 # Use the default values
11037 # Dilate using a structuring element with a 5x5 cross, iterating two times
11038 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11040 # Read the shape from the file diamond.shape, iterating two times.
11041 # The file diamond.shape may contain a pattern of characters like this
11047 # The specified columns and rows are ignored
11048 # but the anchor point coordinates are not
11049 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11054 Erode an image by using a specific structuring element.
11055 It corresponds to the libopencv function @code{cvErode}.
11057 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11058 with the same syntax and semantics as the @ref{dilate} filter.
11062 Smooth the input video.
11064 The filter takes the following parameters:
11065 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11067 @var{type} is the type of smooth filter to apply, and must be one of
11068 the following values: "blur", "blur_no_scale", "median", "gaussian",
11069 or "bilateral". The default value is "gaussian".
11071 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11072 depend on the smooth type. @var{param1} and
11073 @var{param2} accept integer positive values or 0. @var{param3} and
11074 @var{param4} accept floating point values.
11076 The default value for @var{param1} is 3. The default value for the
11077 other parameters is 0.
11079 These parameters correspond to the parameters assigned to the
11080 libopencv function @code{cvSmooth}.
11082 @section oscilloscope
11084 2D Video Oscilloscope.
11086 Useful to measure spatial impulse, step responses, chroma delays, etc.
11088 It accepts the following parameters:
11092 Set scope center x position.
11095 Set scope center y position.
11098 Set scope size, relative to frame diagonal.
11101 Set scope tilt/rotation.
11107 Set trace center x position.
11110 Set trace center y position.
11113 Set trace width, relative to width of frame.
11116 Set trace height, relative to height of frame.
11119 Set which components to trace. By default it traces first three components.
11122 Draw trace grid. By default is enabled.
11125 Draw some statistics. By default is enabled.
11128 Draw scope. By default is enabled.
11131 @subsection Examples
11135 Inspect full first row of video frame.
11137 oscilloscope=x=0.5:y=0:s=1
11141 Inspect full last row of video frame.
11143 oscilloscope=x=0.5:y=1:s=1
11147 Inspect full 5th line of video frame of height 1080.
11149 oscilloscope=x=0.5:y=5/1080:s=1
11153 Inspect full last column of video frame.
11155 oscilloscope=x=1:y=0.5:s=1:t=1
11163 Overlay one video on top of another.
11165 It takes two inputs and has one output. The first input is the "main"
11166 video on which the second input is overlaid.
11168 It accepts the following parameters:
11170 A description of the accepted options follows.
11175 Set the expression for the x and y coordinates of the overlaid video
11176 on the main video. Default value is "0" for both expressions. In case
11177 the expression is invalid, it is set to a huge value (meaning that the
11178 overlay will not be displayed within the output visible area).
11181 See @ref{framesync}.
11184 Set when the expressions for @option{x}, and @option{y} are evaluated.
11186 It accepts the following values:
11189 only evaluate expressions once during the filter initialization or
11190 when a command is processed
11193 evaluate expressions for each incoming frame
11196 Default value is @samp{frame}.
11199 See @ref{framesync}.
11202 Set the format for the output video.
11204 It accepts the following values:
11207 force YUV420 output
11210 force YUV422 output
11213 force YUV444 output
11216 force packed RGB output
11219 force planar RGB output
11222 automatically pick format
11225 Default value is @samp{yuv420}.
11228 See @ref{framesync}.
11231 The @option{x}, and @option{y} expressions can contain the following
11237 The main input width and height.
11241 The overlay input width and height.
11245 The computed values for @var{x} and @var{y}. They are evaluated for
11250 horizontal and vertical chroma subsample values of the output
11251 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11255 the number of input frame, starting from 0
11258 the position in the file of the input frame, NAN if unknown
11261 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11265 This filter also supports the @ref{framesync} options.
11267 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11268 when evaluation is done @emph{per frame}, and will evaluate to NAN
11269 when @option{eval} is set to @samp{init}.
11271 Be aware that frames are taken from each input video in timestamp
11272 order, hence, if their initial timestamps differ, it is a good idea
11273 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11274 have them begin in the same zero timestamp, as the example for
11275 the @var{movie} filter does.
11277 You can chain together more overlays but you should test the
11278 efficiency of such approach.
11280 @subsection Commands
11282 This filter supports the following commands:
11286 Modify the x and y of the overlay input.
11287 The command accepts the same syntax of the corresponding option.
11289 If the specified expression is not valid, it is kept at its current
11293 @subsection Examples
11297 Draw the overlay at 10 pixels from the bottom right corner of the main
11300 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11303 Using named options the example above becomes:
11305 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11309 Insert a transparent PNG logo in the bottom left corner of the input,
11310 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11312 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11316 Insert 2 different transparent PNG logos (second logo on bottom
11317 right corner) using the @command{ffmpeg} tool:
11319 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
11323 Add a transparent color layer on top of the main video; @code{WxH}
11324 must specify the size of the main input to the overlay filter:
11326 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11330 Play an original video and a filtered version (here with the deshake
11331 filter) side by side using the @command{ffplay} tool:
11333 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11336 The above command is the same as:
11338 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11342 Make a sliding overlay appearing from the left to the right top part of the
11343 screen starting since time 2:
11345 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11349 Compose output by putting two input videos side to side:
11351 ffmpeg -i left.avi -i right.avi -filter_complex "
11352 nullsrc=size=200x100 [background];
11353 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11354 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11355 [background][left] overlay=shortest=1 [background+left];
11356 [background+left][right] overlay=shortest=1:x=100 [left+right]
11361 Mask 10-20 seconds of a video by applying the delogo filter to a section
11363 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11364 -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]'
11369 Chain several overlays in cascade:
11371 nullsrc=s=200x200 [bg];
11372 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11373 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11374 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11375 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11376 [in3] null, [mid2] overlay=100:100 [out0]
11383 Apply Overcomplete Wavelet denoiser.
11385 The filter accepts the following options:
11391 Larger depth values will denoise lower frequency components more, but
11392 slow down filtering.
11394 Must be an int in the range 8-16, default is @code{8}.
11396 @item luma_strength, ls
11399 Must be a double value in the range 0-1000, default is @code{1.0}.
11401 @item chroma_strength, cs
11402 Set chroma strength.
11404 Must be a double value in the range 0-1000, default is @code{1.0}.
11410 Add paddings to the input image, and place the original input at the
11411 provided @var{x}, @var{y} coordinates.
11413 It accepts the following parameters:
11418 Specify an expression for the size of the output image with the
11419 paddings added. If the value for @var{width} or @var{height} is 0, the
11420 corresponding input size is used for the output.
11422 The @var{width} expression can reference the value set by the
11423 @var{height} expression, and vice versa.
11425 The default value of @var{width} and @var{height} is 0.
11429 Specify the offsets to place the input image at within the padded area,
11430 with respect to the top/left border of the output image.
11432 The @var{x} expression can reference the value set by the @var{y}
11433 expression, and vice versa.
11435 The default value of @var{x} and @var{y} is 0.
11437 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11438 so the input image is centered on the padded area.
11441 Specify the color of the padded area. For the syntax of this option,
11442 check the "Color" section in the ffmpeg-utils manual.
11444 The default value of @var{color} is "black".
11447 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11449 It accepts the following values:
11453 Only evaluate expressions once during the filter initialization or when
11454 a command is processed.
11457 Evaluate expressions for each incoming frame.
11461 Default value is @samp{init}.
11464 Pad to aspect instead to a resolution.
11468 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11469 options are expressions containing the following constants:
11474 The input video width and height.
11478 These are the same as @var{in_w} and @var{in_h}.
11482 The output width and height (the size of the padded area), as
11483 specified by the @var{width} and @var{height} expressions.
11487 These are the same as @var{out_w} and @var{out_h}.
11491 The x and y offsets as specified by the @var{x} and @var{y}
11492 expressions, or NAN if not yet specified.
11495 same as @var{iw} / @var{ih}
11498 input sample aspect ratio
11501 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11505 The horizontal and vertical chroma subsample values. For example for the
11506 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11509 @subsection Examples
11513 Add paddings with the color "violet" to the input video. The output video
11514 size is 640x480, and the top-left corner of the input video is placed at
11517 pad=640:480:0:40:violet
11520 The example above is equivalent to the following command:
11522 pad=width=640:height=480:x=0:y=40:color=violet
11526 Pad the input to get an output with dimensions increased by 3/2,
11527 and put the input video at the center of the padded area:
11529 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
11533 Pad the input to get a squared output with size equal to the maximum
11534 value between the input width and height, and put the input video at
11535 the center of the padded area:
11537 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
11541 Pad the input to get a final w/h ratio of 16:9:
11543 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
11547 In case of anamorphic video, in order to set the output display aspect
11548 correctly, it is necessary to use @var{sar} in the expression,
11549 according to the relation:
11551 (ih * X / ih) * sar = output_dar
11552 X = output_dar / sar
11555 Thus the previous example needs to be modified to:
11557 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
11561 Double the output size and put the input video in the bottom-right
11562 corner of the output padded area:
11564 pad="2*iw:2*ih:ow-iw:oh-ih"
11568 @anchor{palettegen}
11569 @section palettegen
11571 Generate one palette for a whole video stream.
11573 It accepts the following options:
11577 Set the maximum number of colors to quantize in the palette.
11578 Note: the palette will still contain 256 colors; the unused palette entries
11581 @item reserve_transparent
11582 Create a palette of 255 colors maximum and reserve the last one for
11583 transparency. Reserving the transparency color is useful for GIF optimization.
11584 If not set, the maximum of colors in the palette will be 256. You probably want
11585 to disable this option for a standalone image.
11588 @item transparency_color
11589 Set the color that will be used as background for transparency.
11592 Set statistics mode.
11594 It accepts the following values:
11597 Compute full frame histograms.
11599 Compute histograms only for the part that differs from previous frame. This
11600 might be relevant to give more importance to the moving part of your input if
11601 the background is static.
11603 Compute new histogram for each frame.
11606 Default value is @var{full}.
11609 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
11610 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
11611 color quantization of the palette. This information is also visible at
11612 @var{info} logging level.
11614 @subsection Examples
11618 Generate a representative palette of a given video using @command{ffmpeg}:
11620 ffmpeg -i input.mkv -vf palettegen palette.png
11624 @section paletteuse
11626 Use a palette to downsample an input video stream.
11628 The filter takes two inputs: one video stream and a palette. The palette must
11629 be a 256 pixels image.
11631 It accepts the following options:
11635 Select dithering mode. Available algorithms are:
11638 Ordered 8x8 bayer dithering (deterministic)
11640 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
11641 Note: this dithering is sometimes considered "wrong" and is included as a
11643 @item floyd_steinberg
11644 Floyd and Steingberg dithering (error diffusion)
11646 Frankie Sierra dithering v2 (error diffusion)
11648 Frankie Sierra dithering v2 "Lite" (error diffusion)
11651 Default is @var{sierra2_4a}.
11654 When @var{bayer} dithering is selected, this option defines the scale of the
11655 pattern (how much the crosshatch pattern is visible). A low value means more
11656 visible pattern for less banding, and higher value means less visible pattern
11657 at the cost of more banding.
11659 The option must be an integer value in the range [0,5]. Default is @var{2}.
11662 If set, define the zone to process
11666 Only the changing rectangle will be reprocessed. This is similar to GIF
11667 cropping/offsetting compression mechanism. This option can be useful for speed
11668 if only a part of the image is changing, and has use cases such as limiting the
11669 scope of the error diffusal @option{dither} to the rectangle that bounds the
11670 moving scene (it leads to more deterministic output if the scene doesn't change
11671 much, and as a result less moving noise and better GIF compression).
11674 Default is @var{none}.
11677 Take new palette for each output frame.
11679 @item alpha_threshold
11680 Sets the alpha threshold for transparency. Alpha values above this threshold
11681 will be treated as completely opaque, and values below this threshold will be
11682 treated as completely transparent.
11684 The option must be an integer value in the range [0,255]. Default is @var{128}.
11687 @subsection Examples
11691 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
11692 using @command{ffmpeg}:
11694 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
11698 @section perspective
11700 Correct perspective of video not recorded perpendicular to the screen.
11702 A description of the accepted parameters follows.
11713 Set coordinates expression for top left, top right, bottom left and bottom right corners.
11714 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
11715 If the @code{sense} option is set to @code{source}, then the specified points will be sent
11716 to the corners of the destination. If the @code{sense} option is set to @code{destination},
11717 then the corners of the source will be sent to the specified coordinates.
11719 The expressions can use the following variables:
11724 the width and height of video frame.
11728 Output frame count.
11731 @item interpolation
11732 Set interpolation for perspective correction.
11734 It accepts the following values:
11740 Default value is @samp{linear}.
11743 Set interpretation of coordinate options.
11745 It accepts the following values:
11749 Send point in the source specified by the given coordinates to
11750 the corners of the destination.
11752 @item 1, destination
11754 Send the corners of the source to the point in the destination specified
11755 by the given coordinates.
11757 Default value is @samp{source}.
11761 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
11763 It accepts the following values:
11766 only evaluate expressions once during the filter initialization or
11767 when a command is processed
11770 evaluate expressions for each incoming frame
11773 Default value is @samp{init}.
11778 Delay interlaced video by one field time so that the field order changes.
11780 The intended use is to fix PAL movies that have been captured with the
11781 opposite field order to the film-to-video transfer.
11783 A description of the accepted parameters follows.
11789 It accepts the following values:
11792 Capture field order top-first, transfer bottom-first.
11793 Filter will delay the bottom field.
11796 Capture field order bottom-first, transfer top-first.
11797 Filter will delay the top field.
11800 Capture and transfer with the same field order. This mode only exists
11801 for the documentation of the other options to refer to, but if you
11802 actually select it, the filter will faithfully do nothing.
11805 Capture field order determined automatically by field flags, transfer
11807 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
11808 basis using field flags. If no field information is available,
11809 then this works just like @samp{u}.
11812 Capture unknown or varying, transfer opposite.
11813 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
11814 analyzing the images and selecting the alternative that produces best
11815 match between the fields.
11818 Capture top-first, transfer unknown or varying.
11819 Filter selects among @samp{t} and @samp{p} using image analysis.
11822 Capture bottom-first, transfer unknown or varying.
11823 Filter selects among @samp{b} and @samp{p} using image analysis.
11826 Capture determined by field flags, transfer unknown or varying.
11827 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
11828 image analysis. If no field information is available, then this works just
11829 like @samp{U}. This is the default mode.
11832 Both capture and transfer unknown or varying.
11833 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
11837 @section pixdesctest
11839 Pixel format descriptor test filter, mainly useful for internal
11840 testing. The output video should be equal to the input video.
11844 format=monow, pixdesctest
11847 can be used to test the monowhite pixel format descriptor definition.
11851 Display sample values of color channels. Mainly useful for checking color
11852 and levels. Minimum supported resolution is 640x480.
11854 The filters accept the following options:
11858 Set scope X position, relative offset on X axis.
11861 Set scope Y position, relative offset on Y axis.
11870 Set window opacity. This window also holds statistics about pixel area.
11873 Set window X position, relative offset on X axis.
11876 Set window Y position, relative offset on Y axis.
11881 Enable the specified chain of postprocessing subfilters using libpostproc. This
11882 library should be automatically selected with a GPL build (@code{--enable-gpl}).
11883 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
11884 Each subfilter and some options have a short and a long name that can be used
11885 interchangeably, i.e. dr/dering are the same.
11887 The filters accept the following options:
11891 Set postprocessing subfilters string.
11894 All subfilters share common options to determine their scope:
11898 Honor the quality commands for this subfilter.
11901 Do chrominance filtering, too (default).
11904 Do luminance filtering only (no chrominance).
11907 Do chrominance filtering only (no luminance).
11910 These options can be appended after the subfilter name, separated by a '|'.
11912 Available subfilters are:
11915 @item hb/hdeblock[|difference[|flatness]]
11916 Horizontal deblocking filter
11919 Difference factor where higher values mean more deblocking (default: @code{32}).
11921 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11924 @item vb/vdeblock[|difference[|flatness]]
11925 Vertical deblocking filter
11928 Difference factor where higher values mean more deblocking (default: @code{32}).
11930 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11933 @item ha/hadeblock[|difference[|flatness]]
11934 Accurate horizontal deblocking filter
11937 Difference factor where higher values mean more deblocking (default: @code{32}).
11939 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11942 @item va/vadeblock[|difference[|flatness]]
11943 Accurate vertical deblocking filter
11946 Difference factor where higher values mean more deblocking (default: @code{32}).
11948 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11952 The horizontal and vertical deblocking filters share the difference and
11953 flatness values so you cannot set different horizontal and vertical
11957 @item h1/x1hdeblock
11958 Experimental horizontal deblocking filter
11960 @item v1/x1vdeblock
11961 Experimental vertical deblocking filter
11966 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
11969 larger -> stronger filtering
11971 larger -> stronger filtering
11973 larger -> stronger filtering
11976 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
11979 Stretch luminance to @code{0-255}.
11982 @item lb/linblenddeint
11983 Linear blend deinterlacing filter that deinterlaces the given block by
11984 filtering all lines with a @code{(1 2 1)} filter.
11986 @item li/linipoldeint
11987 Linear interpolating deinterlacing filter that deinterlaces the given block by
11988 linearly interpolating every second line.
11990 @item ci/cubicipoldeint
11991 Cubic interpolating deinterlacing filter deinterlaces the given block by
11992 cubically interpolating every second line.
11994 @item md/mediandeint
11995 Median deinterlacing filter that deinterlaces the given block by applying a
11996 median filter to every second line.
11998 @item fd/ffmpegdeint
11999 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12000 second line with a @code{(-1 4 2 4 -1)} filter.
12003 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12004 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12006 @item fq/forceQuant[|quantizer]
12007 Overrides the quantizer table from the input with the constant quantizer you
12015 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12018 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12021 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12024 @subsection Examples
12028 Apply horizontal and vertical deblocking, deringing and automatic
12029 brightness/contrast:
12035 Apply default filters without brightness/contrast correction:
12041 Apply default filters and temporal denoiser:
12043 pp=default/tmpnoise|1|2|3
12047 Apply deblocking on luminance only, and switch vertical deblocking on or off
12048 automatically depending on available CPU time:
12055 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12056 similar to spp = 6 with 7 point DCT, where only the center sample is
12059 The filter accepts the following options:
12063 Force a constant quantization parameter. It accepts an integer in range
12064 0 to 63. If not set, the filter will use the QP from the video stream
12068 Set thresholding mode. Available modes are:
12072 Set hard thresholding.
12074 Set soft thresholding (better de-ringing effect, but likely blurrier).
12076 Set medium thresholding (good results, default).
12080 @section premultiply
12081 Apply alpha premultiply effect to input video stream using first plane
12082 of second stream as alpha.
12084 Both streams must have same dimensions and same pixel format.
12086 The filter accepts the following option:
12090 Set which planes will be processed, unprocessed planes will be copied.
12091 By default value 0xf, all planes will be processed.
12094 Do not require 2nd input for processing, instead use alpha plane from input stream.
12098 Apply prewitt operator to input video stream.
12100 The filter accepts the following option:
12104 Set which planes will be processed, unprocessed planes will be copied.
12105 By default value 0xf, all planes will be processed.
12108 Set value which will be multiplied with filtered result.
12111 Set value which will be added to filtered result.
12114 @section pseudocolor
12116 Alter frame colors in video with pseudocolors.
12118 This filter accept the following options:
12122 set pixel first component expression
12125 set pixel second component expression
12128 set pixel third component expression
12131 set pixel fourth component expression, corresponds to the alpha component
12134 set component to use as base for altering colors
12137 Each of them specifies the expression to use for computing the lookup table for
12138 the corresponding pixel component values.
12140 The expressions can contain the following constants and functions:
12145 The input width and height.
12148 The input value for the pixel component.
12150 @item ymin, umin, vmin, amin
12151 The minimum allowed component value.
12153 @item ymax, umax, vmax, amax
12154 The maximum allowed component value.
12157 All expressions default to "val".
12159 @subsection Examples
12163 Change too high luma values to gradient:
12165 pseudocolor="'if(between(val,ymax,amax),lerp(ymin,ymax,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(umax,umin,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(vmin,vmax,(val-ymax)/(amax-ymax)),-1):-1'"
12171 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12172 Ratio) between two input videos.
12174 This filter takes in input two input videos, the first input is
12175 considered the "main" source and is passed unchanged to the
12176 output. The second input is used as a "reference" video for computing
12179 Both video inputs must have the same resolution and pixel format for
12180 this filter to work correctly. Also it assumes that both inputs
12181 have the same number of frames, which are compared one by one.
12183 The obtained average PSNR is printed through the logging system.
12185 The filter stores the accumulated MSE (mean squared error) of each
12186 frame, and at the end of the processing it is averaged across all frames
12187 equally, and the following formula is applied to obtain the PSNR:
12190 PSNR = 10*log10(MAX^2/MSE)
12193 Where MAX is the average of the maximum values of each component of the
12196 The description of the accepted parameters follows.
12199 @item stats_file, f
12200 If specified the filter will use the named file to save the PSNR of
12201 each individual frame. When filename equals "-" the data is sent to
12204 @item stats_version
12205 Specifies which version of the stats file format to use. Details of
12206 each format are written below.
12207 Default value is 1.
12209 @item stats_add_max
12210 Determines whether the max value is output to the stats log.
12211 Default value is 0.
12212 Requires stats_version >= 2. If this is set and stats_version < 2,
12213 the filter will return an error.
12216 This filter also supports the @ref{framesync} options.
12218 The file printed if @var{stats_file} is selected, contains a sequence of
12219 key/value pairs of the form @var{key}:@var{value} for each compared
12222 If a @var{stats_version} greater than 1 is specified, a header line precedes
12223 the list of per-frame-pair stats, with key value pairs following the frame
12224 format with the following parameters:
12227 @item psnr_log_version
12228 The version of the log file format. Will match @var{stats_version}.
12231 A comma separated list of the per-frame-pair parameters included in
12235 A description of each shown per-frame-pair parameter follows:
12239 sequential number of the input frame, starting from 1
12242 Mean Square Error pixel-by-pixel average difference of the compared
12243 frames, averaged over all the image components.
12245 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
12246 Mean Square Error pixel-by-pixel average difference of the compared
12247 frames for the component specified by the suffix.
12249 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12250 Peak Signal to Noise ratio of the compared frames for the component
12251 specified by the suffix.
12253 @item max_avg, max_y, max_u, max_v
12254 Maximum allowed value for each channel, and average over all
12260 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12261 [main][ref] psnr="stats_file=stats.log" [out]
12264 On this example the input file being processed is compared with the
12265 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12266 is stored in @file{stats.log}.
12271 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12272 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12275 The pullup filter is designed to take advantage of future context in making
12276 its decisions. This filter is stateless in the sense that it does not lock
12277 onto a pattern to follow, but it instead looks forward to the following
12278 fields in order to identify matches and rebuild progressive frames.
12280 To produce content with an even framerate, insert the fps filter after
12281 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12282 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12284 The filter accepts the following options:
12291 These options set the amount of "junk" to ignore at the left, right, top, and
12292 bottom of the image, respectively. Left and right are in units of 8 pixels,
12293 while top and bottom are in units of 2 lines.
12294 The default is 8 pixels on each side.
12297 Set the strict breaks. Setting this option to 1 will reduce the chances of
12298 filter generating an occasional mismatched frame, but it may also cause an
12299 excessive number of frames to be dropped during high motion sequences.
12300 Conversely, setting it to -1 will make filter match fields more easily.
12301 This may help processing of video where there is slight blurring between
12302 the fields, but may also cause there to be interlaced frames in the output.
12303 Default value is @code{0}.
12306 Set the metric plane to use. It accepts the following values:
12312 Use chroma blue plane.
12315 Use chroma red plane.
12318 This option may be set to use chroma plane instead of the default luma plane
12319 for doing filter's computations. This may improve accuracy on very clean
12320 source material, but more likely will decrease accuracy, especially if there
12321 is chroma noise (rainbow effect) or any grayscale video.
12322 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12323 load and make pullup usable in realtime on slow machines.
12326 For best results (without duplicated frames in the output file) it is
12327 necessary to change the output frame rate. For example, to inverse
12328 telecine NTSC input:
12330 ffmpeg -i input -vf pullup -r 24000/1001 ...
12335 Change video quantization parameters (QP).
12337 The filter accepts the following option:
12341 Set expression for quantization parameter.
12344 The expression is evaluated through the eval API and can contain, among others,
12345 the following constants:
12349 1 if index is not 129, 0 otherwise.
12352 Sequential index starting from -129 to 128.
12355 @subsection Examples
12359 Some equation like:
12367 Flush video frames from internal cache of frames into a random order.
12368 No frame is discarded.
12369 Inspired by @ref{frei0r} nervous filter.
12373 Set size in number of frames of internal cache, in range from @code{2} to
12374 @code{512}. Default is @code{30}.
12377 Set seed for random number generator, must be an integer included between
12378 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12379 less than @code{0}, the filter will try to use a good random seed on a
12383 @section readeia608
12385 Read closed captioning (EIA-608) information from the top lines of a video frame.
12387 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
12388 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
12389 with EIA-608 data (starting from 0). A description of each metadata value follows:
12392 @item lavfi.readeia608.X.cc
12393 The two bytes stored as EIA-608 data (printed in hexadecimal).
12395 @item lavfi.readeia608.X.line
12396 The number of the line on which the EIA-608 data was identified and read.
12399 This filter accepts the following options:
12403 Set the line to start scanning for EIA-608 data. Default is @code{0}.
12406 Set the line to end scanning for EIA-608 data. Default is @code{29}.
12409 Set minimal acceptable amplitude change for sync codes detection.
12410 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
12413 Set the ratio of width reserved for sync code detection.
12414 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
12417 Set the max peaks height difference for sync code detection.
12418 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12421 Set max peaks period difference for sync code detection.
12422 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12425 Set the first two max start code bits differences.
12426 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
12429 Set the minimum ratio of bits height compared to 3rd start code bit.
12430 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
12433 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
12436 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
12439 Enable checking the parity bit. In the event of a parity error, the filter will output
12440 @code{0x00} for that character. Default is false.
12443 @subsection Examples
12447 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
12449 ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pkt_pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
12455 Read vertical interval timecode (VITC) information from the top lines of a
12458 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
12459 timecode value, if a valid timecode has been detected. Further metadata key
12460 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
12461 timecode data has been found or not.
12463 This filter accepts the following options:
12467 Set the maximum number of lines to scan for VITC data. If the value is set to
12468 @code{-1} the full video frame is scanned. Default is @code{45}.
12471 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
12472 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
12475 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
12476 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
12479 @subsection Examples
12483 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
12484 draw @code{--:--:--:--} as a placeholder:
12486 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
12492 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
12494 Destination pixel at position (X, Y) will be picked from source (x, y) position
12495 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
12496 value for pixel will be used for destination pixel.
12498 Xmap and Ymap input video streams must be of same dimensions. Output video stream
12499 will have Xmap/Ymap video stream dimensions.
12500 Xmap and Ymap input video streams are 16bit depth, single channel.
12502 @section removegrain
12504 The removegrain filter is a spatial denoiser for progressive video.
12508 Set mode for the first plane.
12511 Set mode for the second plane.
12514 Set mode for the third plane.
12517 Set mode for the fourth plane.
12520 Range of mode is from 0 to 24. Description of each mode follows:
12524 Leave input plane unchanged. Default.
12527 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
12530 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
12533 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
12536 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
12537 This is equivalent to a median filter.
12540 Line-sensitive clipping giving the minimal change.
12543 Line-sensitive clipping, intermediate.
12546 Line-sensitive clipping, intermediate.
12549 Line-sensitive clipping, intermediate.
12552 Line-sensitive clipping on a line where the neighbours pixels are the closest.
12555 Replaces the target pixel with the closest neighbour.
12558 [1 2 1] horizontal and vertical kernel blur.
12564 Bob mode, interpolates top field from the line where the neighbours
12565 pixels are the closest.
12568 Bob mode, interpolates bottom field from the line where the neighbours
12569 pixels are the closest.
12572 Bob mode, interpolates top field. Same as 13 but with a more complicated
12573 interpolation formula.
12576 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
12577 interpolation formula.
12580 Clips the pixel with the minimum and maximum of respectively the maximum and
12581 minimum of each pair of opposite neighbour pixels.
12584 Line-sensitive clipping using opposite neighbours whose greatest distance from
12585 the current pixel is minimal.
12588 Replaces the pixel with the average of its 8 neighbours.
12591 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
12594 Clips pixels using the averages of opposite neighbour.
12597 Same as mode 21 but simpler and faster.
12600 Small edge and halo removal, but reputed useless.
12606 @section removelogo
12608 Suppress a TV station logo, using an image file to determine which
12609 pixels comprise the logo. It works by filling in the pixels that
12610 comprise the logo with neighboring pixels.
12612 The filter accepts the following options:
12616 Set the filter bitmap file, which can be any image format supported by
12617 libavformat. The width and height of the image file must match those of the
12618 video stream being processed.
12621 Pixels in the provided bitmap image with a value of zero are not
12622 considered part of the logo, non-zero pixels are considered part of
12623 the logo. If you use white (255) for the logo and black (0) for the
12624 rest, you will be safe. For making the filter bitmap, it is
12625 recommended to take a screen capture of a black frame with the logo
12626 visible, and then using a threshold filter followed by the erode
12627 filter once or twice.
12629 If needed, little splotches can be fixed manually. Remember that if
12630 logo pixels are not covered, the filter quality will be much
12631 reduced. Marking too many pixels as part of the logo does not hurt as
12632 much, but it will increase the amount of blurring needed to cover over
12633 the image and will destroy more information than necessary, and extra
12634 pixels will slow things down on a large logo.
12636 @section repeatfields
12638 This filter uses the repeat_field flag from the Video ES headers and hard repeats
12639 fields based on its value.
12643 Reverse a video clip.
12645 Warning: This filter requires memory to buffer the entire clip, so trimming
12648 @subsection Examples
12652 Take the first 5 seconds of a clip, and reverse it.
12659 Apply roberts cross operator to input video stream.
12661 The filter accepts the following option:
12665 Set which planes will be processed, unprocessed planes will be copied.
12666 By default value 0xf, all planes will be processed.
12669 Set value which will be multiplied with filtered result.
12672 Set value which will be added to filtered result.
12677 Rotate video by an arbitrary angle expressed in radians.
12679 The filter accepts the following options:
12681 A description of the optional parameters follows.
12684 Set an expression for the angle by which to rotate the input video
12685 clockwise, expressed as a number of radians. A negative value will
12686 result in a counter-clockwise rotation. By default it is set to "0".
12688 This expression is evaluated for each frame.
12691 Set the output width expression, default value is "iw".
12692 This expression is evaluated just once during configuration.
12695 Set the output height expression, default value is "ih".
12696 This expression is evaluated just once during configuration.
12699 Enable bilinear interpolation if set to 1, a value of 0 disables
12700 it. Default value is 1.
12703 Set the color used to fill the output area not covered by the rotated
12704 image. For the general syntax of this option, check the "Color" section in the
12705 ffmpeg-utils manual. If the special value "none" is selected then no
12706 background is printed (useful for example if the background is never shown).
12708 Default value is "black".
12711 The expressions for the angle and the output size can contain the
12712 following constants and functions:
12716 sequential number of the input frame, starting from 0. It is always NAN
12717 before the first frame is filtered.
12720 time in seconds of the input frame, it is set to 0 when the filter is
12721 configured. It is always NAN before the first frame is filtered.
12725 horizontal and vertical chroma subsample values. For example for the
12726 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12730 the input video width and height
12734 the output width and height, that is the size of the padded area as
12735 specified by the @var{width} and @var{height} expressions
12739 the minimal width/height required for completely containing the input
12740 video rotated by @var{a} radians.
12742 These are only available when computing the @option{out_w} and
12743 @option{out_h} expressions.
12746 @subsection Examples
12750 Rotate the input by PI/6 radians clockwise:
12756 Rotate the input by PI/6 radians counter-clockwise:
12762 Rotate the input by 45 degrees clockwise:
12768 Apply a constant rotation with period T, starting from an angle of PI/3:
12770 rotate=PI/3+2*PI*t/T
12774 Make the input video rotation oscillating with a period of T
12775 seconds and an amplitude of A radians:
12777 rotate=A*sin(2*PI/T*t)
12781 Rotate the video, output size is chosen so that the whole rotating
12782 input video is always completely contained in the output:
12784 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
12788 Rotate the video, reduce the output size so that no background is ever
12791 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
12795 @subsection Commands
12797 The filter supports the following commands:
12801 Set the angle expression.
12802 The command accepts the same syntax of the corresponding option.
12804 If the specified expression is not valid, it is kept at its current
12810 Apply Shape Adaptive Blur.
12812 The filter accepts the following options:
12815 @item luma_radius, lr
12816 Set luma blur filter strength, must be a value in range 0.1-4.0, default
12817 value is 1.0. A greater value will result in a more blurred image, and
12818 in slower processing.
12820 @item luma_pre_filter_radius, lpfr
12821 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
12824 @item luma_strength, ls
12825 Set luma maximum difference between pixels to still be considered, must
12826 be a value in the 0.1-100.0 range, default value is 1.0.
12828 @item chroma_radius, cr
12829 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
12830 greater value will result in a more blurred image, and in slower
12833 @item chroma_pre_filter_radius, cpfr
12834 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
12836 @item chroma_strength, cs
12837 Set chroma maximum difference between pixels to still be considered,
12838 must be a value in the -0.9-100.0 range.
12841 Each chroma option value, if not explicitly specified, is set to the
12842 corresponding luma option value.
12847 Scale (resize) the input video, using the libswscale library.
12849 The scale filter forces the output display aspect ratio to be the same
12850 of the input, by changing the output sample aspect ratio.
12852 If the input image format is different from the format requested by
12853 the next filter, the scale filter will convert the input to the
12856 @subsection Options
12857 The filter accepts the following options, or any of the options
12858 supported by the libswscale scaler.
12860 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
12861 the complete list of scaler options.
12866 Set the output video dimension expression. Default value is the input
12869 If the @var{width} or @var{w} value is 0, the input width is used for
12870 the output. If the @var{height} or @var{h} value is 0, the input height
12871 is used for the output.
12873 If one and only one of the values is -n with n >= 1, the scale filter
12874 will use a value that maintains the aspect ratio of the input image,
12875 calculated from the other specified dimension. After that it will,
12876 however, make sure that the calculated dimension is divisible by n and
12877 adjust the value if necessary.
12879 If both values are -n with n >= 1, the behavior will be identical to
12880 both values being set to 0 as previously detailed.
12882 See below for the list of accepted constants for use in the dimension
12886 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
12890 Only evaluate expressions once during the filter initialization or when a command is processed.
12893 Evaluate expressions for each incoming frame.
12897 Default value is @samp{init}.
12901 Set the interlacing mode. It accepts the following values:
12905 Force interlaced aware scaling.
12908 Do not apply interlaced scaling.
12911 Select interlaced aware scaling depending on whether the source frames
12912 are flagged as interlaced or not.
12915 Default value is @samp{0}.
12918 Set libswscale scaling flags. See
12919 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12920 complete list of values. If not explicitly specified the filter applies
12924 @item param0, param1
12925 Set libswscale input parameters for scaling algorithms that need them. See
12926 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12927 complete documentation. If not explicitly specified the filter applies
12933 Set the video size. For the syntax of this option, check the
12934 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12936 @item in_color_matrix
12937 @item out_color_matrix
12938 Set in/output YCbCr color space type.
12940 This allows the autodetected value to be overridden as well as allows forcing
12941 a specific value used for the output and encoder.
12943 If not specified, the color space type depends on the pixel format.
12949 Choose automatically.
12952 Format conforming to International Telecommunication Union (ITU)
12953 Recommendation BT.709.
12956 Set color space conforming to the United States Federal Communications
12957 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
12960 Set color space conforming to:
12964 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
12967 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
12970 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
12975 Set color space conforming to SMPTE ST 240:1999.
12980 Set in/output YCbCr sample range.
12982 This allows the autodetected value to be overridden as well as allows forcing
12983 a specific value used for the output and encoder. If not specified, the
12984 range depends on the pixel format. Possible values:
12988 Choose automatically.
12991 Set full range (0-255 in case of 8-bit luma).
12994 Set "MPEG" range (16-235 in case of 8-bit luma).
12997 @item force_original_aspect_ratio
12998 Enable decreasing or increasing output video width or height if necessary to
12999 keep the original aspect ratio. Possible values:
13003 Scale the video as specified and disable this feature.
13006 The output video dimensions will automatically be decreased if needed.
13009 The output video dimensions will automatically be increased if needed.
13013 One useful instance of this option is that when you know a specific device's
13014 maximum allowed resolution, you can use this to limit the output video to
13015 that, while retaining the aspect ratio. For example, device A allows
13016 1280x720 playback, and your video is 1920x800. Using this option (set it to
13017 decrease) and specifying 1280x720 to the command line makes the output
13020 Please note that this is a different thing than specifying -1 for @option{w}
13021 or @option{h}, you still need to specify the output resolution for this option
13026 The values of the @option{w} and @option{h} options are expressions
13027 containing the following constants:
13032 The input width and height
13036 These are the same as @var{in_w} and @var{in_h}.
13040 The output (scaled) width and height
13044 These are the same as @var{out_w} and @var{out_h}
13047 The same as @var{iw} / @var{ih}
13050 input sample aspect ratio
13053 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13057 horizontal and vertical input chroma subsample values. For example for the
13058 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13062 horizontal and vertical output chroma subsample values. For example for the
13063 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13066 @subsection Examples
13070 Scale the input video to a size of 200x100
13075 This is equivalent to:
13086 Specify a size abbreviation for the output size:
13091 which can also be written as:
13097 Scale the input to 2x:
13099 scale=w=2*iw:h=2*ih
13103 The above is the same as:
13105 scale=2*in_w:2*in_h
13109 Scale the input to 2x with forced interlaced scaling:
13111 scale=2*iw:2*ih:interl=1
13115 Scale the input to half size:
13117 scale=w=iw/2:h=ih/2
13121 Increase the width, and set the height to the same size:
13127 Seek Greek harmony:
13134 Increase the height, and set the width to 3/2 of the height:
13136 scale=w=3/2*oh:h=3/5*ih
13140 Increase the size, making the size a multiple of the chroma
13143 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13147 Increase the width to a maximum of 500 pixels,
13148 keeping the same aspect ratio as the input:
13150 scale=w='min(500\, iw*3/2):h=-1'
13154 @subsection Commands
13156 This filter supports the following commands:
13160 Set the output video dimension expression.
13161 The command accepts the same syntax of the corresponding option.
13163 If the specified expression is not valid, it is kept at its current
13169 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13170 format conversion on CUDA video frames. Setting the output width and height
13171 works in the same way as for the @var{scale} filter.
13173 The following additional options are accepted:
13176 The pixel format of the output CUDA frames. If set to the string "same" (the
13177 default), the input format will be kept. Note that automatic format negotiation
13178 and conversion is not yet supported for hardware frames
13181 The interpolation algorithm used for resizing. One of the following:
13188 @item cubic2p_bspline
13189 2-parameter cubic (B=1, C=0)
13191 @item cubic2p_catmullrom
13192 2-parameter cubic (B=0, C=1/2)
13194 @item cubic2p_b05c03
13195 2-parameter cubic (B=1/2, C=3/10)
13207 Scale (resize) the input video, based on a reference video.
13209 See the scale filter for available options, scale2ref supports the same but
13210 uses the reference video instead of the main input as basis. scale2ref also
13211 supports the following additional constants for the @option{w} and
13212 @option{h} options:
13217 The main input video's width and height
13220 The same as @var{main_w} / @var{main_h}
13223 The main input video's sample aspect ratio
13225 @item main_dar, mdar
13226 The main input video's display aspect ratio. Calculated from
13227 @code{(main_w / main_h) * main_sar}.
13231 The main input video's horizontal and vertical chroma subsample values.
13232 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13236 @subsection Examples
13240 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13242 'scale2ref[b][a];[a][b]overlay'
13246 @anchor{selectivecolor}
13247 @section selectivecolor
13249 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13250 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13251 by the "purity" of the color (that is, how saturated it already is).
13253 This filter is similar to the Adobe Photoshop Selective Color tool.
13255 The filter accepts the following options:
13258 @item correction_method
13259 Select color correction method.
13261 Available values are:
13264 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13267 Specified adjustments are relative to the original component value.
13269 Default is @code{absolute}.
13271 Adjustments for red pixels (pixels where the red component is the maximum)
13273 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13275 Adjustments for green pixels (pixels where the green component is the maximum)
13277 Adjustments for cyan pixels (pixels where the red component is the minimum)
13279 Adjustments for blue pixels (pixels where the blue component is the maximum)
13281 Adjustments for magenta pixels (pixels where the green component is the minimum)
13283 Adjustments for white pixels (pixels where all components are greater than 128)
13285 Adjustments for all pixels except pure black and pure white
13287 Adjustments for black pixels (pixels where all components are lesser than 128)
13289 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13292 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13293 4 space separated floating point adjustment values in the [-1,1] range,
13294 respectively to adjust the amount of cyan, magenta, yellow and black for the
13295 pixels of its range.
13297 @subsection Examples
13301 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13302 increase magenta by 27% in blue areas:
13304 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13308 Use a Photoshop selective color preset:
13310 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13314 @anchor{separatefields}
13315 @section separatefields
13317 The @code{separatefields} takes a frame-based video input and splits
13318 each frame into its components fields, producing a new half height clip
13319 with twice the frame rate and twice the frame count.
13321 This filter use field-dominance information in frame to decide which
13322 of each pair of fields to place first in the output.
13323 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13325 @section setdar, setsar
13327 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13330 This is done by changing the specified Sample (aka Pixel) Aspect
13331 Ratio, according to the following equation:
13333 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13336 Keep in mind that the @code{setdar} filter does not modify the pixel
13337 dimensions of the video frame. Also, the display aspect ratio set by
13338 this filter may be changed by later filters in the filterchain,
13339 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13342 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13343 the filter output video.
13345 Note that as a consequence of the application of this filter, the
13346 output display aspect ratio will change according to the equation
13349 Keep in mind that the sample aspect ratio set by the @code{setsar}
13350 filter may be changed by later filters in the filterchain, e.g. if
13351 another "setsar" or a "setdar" filter is applied.
13353 It accepts the following parameters:
13356 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13357 Set the aspect ratio used by the filter.
13359 The parameter can be a floating point number string, an expression, or
13360 a string of the form @var{num}:@var{den}, where @var{num} and
13361 @var{den} are the numerator and denominator of the aspect ratio. If
13362 the parameter is not specified, it is assumed the value "0".
13363 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13367 Set the maximum integer value to use for expressing numerator and
13368 denominator when reducing the expressed aspect ratio to a rational.
13369 Default value is @code{100}.
13373 The parameter @var{sar} is an expression containing
13374 the following constants:
13378 These are approximated values for the mathematical constants e
13379 (Euler's number), pi (Greek pi), and phi (the golden ratio).
13382 The input width and height.
13385 These are the same as @var{w} / @var{h}.
13388 The input sample aspect ratio.
13391 The input display aspect ratio. It is the same as
13392 (@var{w} / @var{h}) * @var{sar}.
13395 Horizontal and vertical chroma subsample values. For example, for the
13396 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13399 @subsection Examples
13404 To change the display aspect ratio to 16:9, specify one of the following:
13411 To change the sample aspect ratio to 10:11, specify:
13417 To set a display aspect ratio of 16:9, and specify a maximum integer value of
13418 1000 in the aspect ratio reduction, use the command:
13420 setdar=ratio=16/9:max=1000
13428 Force field for the output video frame.
13430 The @code{setfield} filter marks the interlace type field for the
13431 output frames. It does not change the input frame, but only sets the
13432 corresponding property, which affects how the frame is treated by
13433 following filters (e.g. @code{fieldorder} or @code{yadif}).
13435 The filter accepts the following options:
13440 Available values are:
13444 Keep the same field property.
13447 Mark the frame as bottom-field-first.
13450 Mark the frame as top-field-first.
13453 Mark the frame as progressive.
13459 Show a line containing various information for each input video frame.
13460 The input video is not modified.
13462 The shown line contains a sequence of key/value pairs of the form
13463 @var{key}:@var{value}.
13465 The following values are shown in the output:
13469 The (sequential) number of the input frame, starting from 0.
13472 The Presentation TimeStamp of the input frame, expressed as a number of
13473 time base units. The time base unit depends on the filter input pad.
13476 The Presentation TimeStamp of the input frame, expressed as a number of
13480 The position of the frame in the input stream, or -1 if this information is
13481 unavailable and/or meaningless (for example in case of synthetic video).
13484 The pixel format name.
13487 The sample aspect ratio of the input frame, expressed in the form
13488 @var{num}/@var{den}.
13491 The size of the input frame. For the syntax of this option, check the
13492 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13495 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
13496 for bottom field first).
13499 This is 1 if the frame is a key frame, 0 otherwise.
13502 The picture type of the input frame ("I" for an I-frame, "P" for a
13503 P-frame, "B" for a B-frame, or "?" for an unknown type).
13504 Also refer to the documentation of the @code{AVPictureType} enum and of
13505 the @code{av_get_picture_type_char} function defined in
13506 @file{libavutil/avutil.h}.
13509 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
13511 @item plane_checksum
13512 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
13513 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
13516 @section showpalette
13518 Displays the 256 colors palette of each frame. This filter is only relevant for
13519 @var{pal8} pixel format frames.
13521 It accepts the following option:
13525 Set the size of the box used to represent one palette color entry. Default is
13526 @code{30} (for a @code{30x30} pixel box).
13529 @section shuffleframes
13531 Reorder and/or duplicate and/or drop video frames.
13533 It accepts the following parameters:
13537 Set the destination indexes of input frames.
13538 This is space or '|' separated list of indexes that maps input frames to output
13539 frames. Number of indexes also sets maximal value that each index may have.
13540 '-1' index have special meaning and that is to drop frame.
13543 The first frame has the index 0. The default is to keep the input unchanged.
13545 @subsection Examples
13549 Swap second and third frame of every three frames of the input:
13551 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
13555 Swap 10th and 1st frame of every ten frames of the input:
13557 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
13561 @section shuffleplanes
13563 Reorder and/or duplicate video planes.
13565 It accepts the following parameters:
13570 The index of the input plane to be used as the first output plane.
13573 The index of the input plane to be used as the second output plane.
13576 The index of the input plane to be used as the third output plane.
13579 The index of the input plane to be used as the fourth output plane.
13583 The first plane has the index 0. The default is to keep the input unchanged.
13585 @subsection Examples
13589 Swap the second and third planes of the input:
13591 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
13595 @anchor{signalstats}
13596 @section signalstats
13597 Evaluate various visual metrics that assist in determining issues associated
13598 with the digitization of analog video media.
13600 By default the filter will log these metadata values:
13604 Display the minimal Y value contained within the input frame. Expressed in
13608 Display the Y value at the 10% percentile within the input frame. Expressed in
13612 Display the average Y value within the input frame. Expressed in range of
13616 Display the Y value at the 90% percentile within the input frame. Expressed in
13620 Display the maximum Y value contained within the input frame. Expressed in
13624 Display the minimal U value contained within the input frame. Expressed in
13628 Display the U value at the 10% percentile within the input frame. Expressed in
13632 Display the average U value within the input frame. Expressed in range of
13636 Display the U value at the 90% percentile within the input frame. Expressed in
13640 Display the maximum U value contained within the input frame. Expressed in
13644 Display the minimal V value contained within the input frame. Expressed in
13648 Display the V value at the 10% percentile within the input frame. Expressed in
13652 Display the average V value within the input frame. Expressed in range of
13656 Display the V value at the 90% percentile within the input frame. Expressed in
13660 Display the maximum V value contained within the input frame. Expressed in
13664 Display the minimal saturation value contained within the input frame.
13665 Expressed in range of [0-~181.02].
13668 Display the saturation value at the 10% percentile within the input frame.
13669 Expressed in range of [0-~181.02].
13672 Display the average saturation value within the input frame. Expressed in range
13676 Display the saturation value at the 90% percentile within the input frame.
13677 Expressed in range of [0-~181.02].
13680 Display the maximum saturation value contained within the input frame.
13681 Expressed in range of [0-~181.02].
13684 Display the median value for hue within the input frame. Expressed in range of
13688 Display the average value for hue within the input frame. Expressed in range of
13692 Display the average of sample value difference between all values of the Y
13693 plane in the current frame and corresponding values of the previous input frame.
13694 Expressed in range of [0-255].
13697 Display the average of sample value difference between all values of the U
13698 plane in the current frame and corresponding values of the previous input frame.
13699 Expressed in range of [0-255].
13702 Display the average of sample value difference between all values of the V
13703 plane in the current frame and corresponding values of the previous input frame.
13704 Expressed in range of [0-255].
13707 Display bit depth of Y plane in current frame.
13708 Expressed in range of [0-16].
13711 Display bit depth of U plane in current frame.
13712 Expressed in range of [0-16].
13715 Display bit depth of V plane in current frame.
13716 Expressed in range of [0-16].
13719 The filter accepts the following options:
13725 @option{stat} specify an additional form of image analysis.
13726 @option{out} output video with the specified type of pixel highlighted.
13728 Both options accept the following values:
13732 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
13733 unlike the neighboring pixels of the same field. Examples of temporal outliers
13734 include the results of video dropouts, head clogs, or tape tracking issues.
13737 Identify @var{vertical line repetition}. Vertical line repetition includes
13738 similar rows of pixels within a frame. In born-digital video vertical line
13739 repetition is common, but this pattern is uncommon in video digitized from an
13740 analog source. When it occurs in video that results from the digitization of an
13741 analog source it can indicate concealment from a dropout compensator.
13744 Identify pixels that fall outside of legal broadcast range.
13748 Set the highlight color for the @option{out} option. The default color is
13752 @subsection Examples
13756 Output data of various video metrics:
13758 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
13762 Output specific data about the minimum and maximum values of the Y plane per frame:
13764 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
13768 Playback video while highlighting pixels that are outside of broadcast range in red.
13770 ffplay example.mov -vf signalstats="out=brng:color=red"
13774 Playback video with signalstats metadata drawn over the frame.
13776 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
13779 The contents of signalstat_drawtext.txt used in the command are:
13782 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
13783 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
13784 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
13785 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
13793 Calculates the MPEG-7 Video Signature. The filter can handle more than one
13794 input. In this case the matching between the inputs can be calculated additionally.
13795 The filter always passes through the first input. The signature of each stream can
13796 be written into a file.
13798 It accepts the following options:
13802 Enable or disable the matching process.
13804 Available values are:
13808 Disable the calculation of a matching (default).
13810 Calculate the matching for the whole video and output whether the whole video
13811 matches or only parts.
13813 Calculate only until a matching is found or the video ends. Should be faster in
13818 Set the number of inputs. The option value must be a non negative integer.
13819 Default value is 1.
13822 Set the path to which the output is written. If there is more than one input,
13823 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
13824 integer), that will be replaced with the input number. If no filename is
13825 specified, no output will be written. This is the default.
13828 Choose the output format.
13830 Available values are:
13834 Use the specified binary representation (default).
13836 Use the specified xml representation.
13840 Set threshold to detect one word as similar. The option value must be an integer
13841 greater than zero. The default value is 9000.
13844 Set threshold to detect all words as similar. The option value must be an integer
13845 greater than zero. The default value is 60000.
13848 Set threshold to detect frames as similar. The option value must be an integer
13849 greater than zero. The default value is 116.
13852 Set the minimum length of a sequence in frames to recognize it as matching
13853 sequence. The option value must be a non negative integer value.
13854 The default value is 0.
13857 Set the minimum relation, that matching frames to all frames must have.
13858 The option value must be a double value between 0 and 1. The default value is 0.5.
13861 @subsection Examples
13865 To calculate the signature of an input video and store it in signature.bin:
13867 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
13871 To detect whether two videos match and store the signatures in XML format in
13872 signature0.xml and signature1.xml:
13874 ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
13882 Blur the input video without impacting the outlines.
13884 It accepts the following options:
13887 @item luma_radius, lr
13888 Set the luma radius. The option value must be a float number in
13889 the range [0.1,5.0] that specifies the variance of the gaussian filter
13890 used to blur the image (slower if larger). Default value is 1.0.
13892 @item luma_strength, ls
13893 Set the luma strength. The option value must be a float number
13894 in the range [-1.0,1.0] that configures the blurring. A value included
13895 in [0.0,1.0] will blur the image whereas a value included in
13896 [-1.0,0.0] will sharpen the image. Default value is 1.0.
13898 @item luma_threshold, lt
13899 Set the luma threshold used as a coefficient to determine
13900 whether a pixel should be blurred or not. The option value must be an
13901 integer in the range [-30,30]. A value of 0 will filter all the image,
13902 a value included in [0,30] will filter flat areas and a value included
13903 in [-30,0] will filter edges. Default value is 0.
13905 @item chroma_radius, cr
13906 Set the chroma radius. The option value must be a float number in
13907 the range [0.1,5.0] that specifies the variance of the gaussian filter
13908 used to blur the image (slower if larger). Default value is @option{luma_radius}.
13910 @item chroma_strength, cs
13911 Set the chroma strength. The option value must be a float number
13912 in the range [-1.0,1.0] that configures the blurring. A value included
13913 in [0.0,1.0] will blur the image whereas a value included in
13914 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
13916 @item chroma_threshold, ct
13917 Set the chroma threshold used as a coefficient to determine
13918 whether a pixel should be blurred or not. The option value must be an
13919 integer in the range [-30,30]. A value of 0 will filter all the image,
13920 a value included in [0,30] will filter flat areas and a value included
13921 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
13924 If a chroma option is not explicitly set, the corresponding luma value
13929 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
13931 This filter takes in input two input videos, the first input is
13932 considered the "main" source and is passed unchanged to the
13933 output. The second input is used as a "reference" video for computing
13936 Both video inputs must have the same resolution and pixel format for
13937 this filter to work correctly. Also it assumes that both inputs
13938 have the same number of frames, which are compared one by one.
13940 The filter stores the calculated SSIM of each frame.
13942 The description of the accepted parameters follows.
13945 @item stats_file, f
13946 If specified the filter will use the named file to save the SSIM of
13947 each individual frame. When filename equals "-" the data is sent to
13951 The file printed if @var{stats_file} is selected, contains a sequence of
13952 key/value pairs of the form @var{key}:@var{value} for each compared
13955 A description of each shown parameter follows:
13959 sequential number of the input frame, starting from 1
13961 @item Y, U, V, R, G, B
13962 SSIM of the compared frames for the component specified by the suffix.
13965 SSIM of the compared frames for the whole frame.
13968 Same as above but in dB representation.
13971 This filter also supports the @ref{framesync} options.
13975 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13976 [main][ref] ssim="stats_file=stats.log" [out]
13979 On this example the input file being processed is compared with the
13980 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
13981 is stored in @file{stats.log}.
13983 Another example with both psnr and ssim at same time:
13985 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
13990 Convert between different stereoscopic image formats.
13992 The filters accept the following options:
13996 Set stereoscopic image format of input.
13998 Available values for input image formats are:
14001 side by side parallel (left eye left, right eye right)
14004 side by side crosseye (right eye left, left eye right)
14007 side by side parallel with half width resolution
14008 (left eye left, right eye right)
14011 side by side crosseye with half width resolution
14012 (right eye left, left eye right)
14015 above-below (left eye above, right eye below)
14018 above-below (right eye above, left eye below)
14021 above-below with half height resolution
14022 (left eye above, right eye below)
14025 above-below with half height resolution
14026 (right eye above, left eye below)
14029 alternating frames (left eye first, right eye second)
14032 alternating frames (right eye first, left eye second)
14035 interleaved rows (left eye has top row, right eye starts on next row)
14038 interleaved rows (right eye has top row, left eye starts on next row)
14041 interleaved columns, left eye first
14044 interleaved columns, right eye first
14046 Default value is @samp{sbsl}.
14050 Set stereoscopic image format of output.
14054 side by side parallel (left eye left, right eye right)
14057 side by side crosseye (right eye left, left eye right)
14060 side by side parallel with half width resolution
14061 (left eye left, right eye right)
14064 side by side crosseye with half width resolution
14065 (right eye left, left eye right)
14068 above-below (left eye above, right eye below)
14071 above-below (right eye above, left eye below)
14074 above-below with half height resolution
14075 (left eye above, right eye below)
14078 above-below with half height resolution
14079 (right eye above, left eye below)
14082 alternating frames (left eye first, right eye second)
14085 alternating frames (right eye first, left eye second)
14088 interleaved rows (left eye has top row, right eye starts on next row)
14091 interleaved rows (right eye has top row, left eye starts on next row)
14094 anaglyph red/blue gray
14095 (red filter on left eye, blue filter on right eye)
14098 anaglyph red/green gray
14099 (red filter on left eye, green filter on right eye)
14102 anaglyph red/cyan gray
14103 (red filter on left eye, cyan filter on right eye)
14106 anaglyph red/cyan half colored
14107 (red filter on left eye, cyan filter on right eye)
14110 anaglyph red/cyan color
14111 (red filter on left eye, cyan filter on right eye)
14114 anaglyph red/cyan color optimized with the least squares projection of dubois
14115 (red filter on left eye, cyan filter on right eye)
14118 anaglyph green/magenta gray
14119 (green filter on left eye, magenta filter on right eye)
14122 anaglyph green/magenta half colored
14123 (green filter on left eye, magenta filter on right eye)
14126 anaglyph green/magenta colored
14127 (green filter on left eye, magenta filter on right eye)
14130 anaglyph green/magenta color optimized with the least squares projection of dubois
14131 (green filter on left eye, magenta filter on right eye)
14134 anaglyph yellow/blue gray
14135 (yellow filter on left eye, blue filter on right eye)
14138 anaglyph yellow/blue half colored
14139 (yellow filter on left eye, blue filter on right eye)
14142 anaglyph yellow/blue colored
14143 (yellow filter on left eye, blue filter on right eye)
14146 anaglyph yellow/blue color optimized with the least squares projection of dubois
14147 (yellow filter on left eye, blue filter on right eye)
14150 mono output (left eye only)
14153 mono output (right eye only)
14156 checkerboard, left eye first
14159 checkerboard, right eye first
14162 interleaved columns, left eye first
14165 interleaved columns, right eye first
14171 Default value is @samp{arcd}.
14174 @subsection Examples
14178 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14184 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14190 @section streamselect, astreamselect
14191 Select video or audio streams.
14193 The filter accepts the following options:
14197 Set number of inputs. Default is 2.
14200 Set input indexes to remap to outputs.
14203 @subsection Commands
14205 The @code{streamselect} and @code{astreamselect} filter supports the following
14210 Set input indexes to remap to outputs.
14213 @subsection Examples
14217 Select first 5 seconds 1st stream and rest of time 2nd stream:
14219 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14223 Same as above, but for audio:
14225 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14230 Apply sobel operator to input video stream.
14232 The filter accepts the following option:
14236 Set which planes will be processed, unprocessed planes will be copied.
14237 By default value 0xf, all planes will be processed.
14240 Set value which will be multiplied with filtered result.
14243 Set value which will be added to filtered result.
14249 Apply a simple postprocessing filter that compresses and decompresses the image
14250 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14251 and average the results.
14253 The filter accepts the following options:
14257 Set quality. This option defines the number of levels for averaging. It accepts
14258 an integer in the range 0-6. If set to @code{0}, the filter will have no
14259 effect. A value of @code{6} means the higher quality. For each increment of
14260 that value the speed drops by a factor of approximately 2. Default value is
14264 Force a constant quantization parameter. If not set, the filter will use the QP
14265 from the video stream (if available).
14268 Set thresholding mode. Available modes are:
14272 Set hard thresholding (default).
14274 Set soft thresholding (better de-ringing effect, but likely blurrier).
14277 @item use_bframe_qp
14278 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14279 option may cause flicker since the B-Frames have often larger QP. Default is
14280 @code{0} (not enabled).
14286 Draw subtitles on top of input video using the libass library.
14288 To enable compilation of this filter you need to configure FFmpeg with
14289 @code{--enable-libass}. This filter also requires a build with libavcodec and
14290 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14291 Alpha) subtitles format.
14293 The filter accepts the following options:
14297 Set the filename of the subtitle file to read. It must be specified.
14299 @item original_size
14300 Specify the size of the original video, the video for which the ASS file
14301 was composed. For the syntax of this option, check the
14302 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14303 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14304 correctly scale the fonts if the aspect ratio has been changed.
14307 Set a directory path containing fonts that can be used by the filter.
14308 These fonts will be used in addition to whatever the font provider uses.
14311 Process alpha channel, by default alpha channel is untouched.
14314 Set subtitles input character encoding. @code{subtitles} filter only. Only
14315 useful if not UTF-8.
14317 @item stream_index, si
14318 Set subtitles stream index. @code{subtitles} filter only.
14321 Override default style or script info parameters of the subtitles. It accepts a
14322 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14325 If the first key is not specified, it is assumed that the first value
14326 specifies the @option{filename}.
14328 For example, to render the file @file{sub.srt} on top of the input
14329 video, use the command:
14334 which is equivalent to:
14336 subtitles=filename=sub.srt
14339 To render the default subtitles stream from file @file{video.mkv}, use:
14341 subtitles=video.mkv
14344 To render the second subtitles stream from that file, use:
14346 subtitles=video.mkv:si=1
14349 To make the subtitles stream from @file{sub.srt} appear in transparent green
14350 @code{DejaVu Serif}, use:
14352 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14355 @section super2xsai
14357 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14358 Interpolate) pixel art scaling algorithm.
14360 Useful for enlarging pixel art images without reducing sharpness.
14364 Swap two rectangular objects in video.
14366 This filter accepts the following options:
14376 Set 1st rect x coordinate.
14379 Set 1st rect y coordinate.
14382 Set 2nd rect x coordinate.
14385 Set 2nd rect y coordinate.
14387 All expressions are evaluated once for each frame.
14390 The all options are expressions containing the following constants:
14395 The input width and height.
14398 same as @var{w} / @var{h}
14401 input sample aspect ratio
14404 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
14407 The number of the input frame, starting from 0.
14410 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
14413 the position in the file of the input frame, NAN if unknown
14421 Apply telecine process to the video.
14423 This filter accepts the following options:
14432 The default value is @code{top}.
14436 A string of numbers representing the pulldown pattern you wish to apply.
14437 The default value is @code{23}.
14441 Some typical patterns:
14446 24p: 2332 (preferred)
14453 24p: 222222222223 ("Euro pulldown")
14460 Apply threshold effect to video stream.
14462 This filter needs four video streams to perform thresholding.
14463 First stream is stream we are filtering.
14464 Second stream is holding threshold values, third stream is holding min values,
14465 and last, fourth stream is holding max values.
14467 The filter accepts the following option:
14471 Set which planes will be processed, unprocessed planes will be copied.
14472 By default value 0xf, all planes will be processed.
14475 For example if first stream pixel's component value is less then threshold value
14476 of pixel component from 2nd threshold stream, third stream value will picked,
14477 otherwise fourth stream pixel component value will be picked.
14479 Using color source filter one can perform various types of thresholding:
14481 @subsection Examples
14485 Binary threshold, using gray color as threshold:
14487 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
14491 Inverted binary threshold, using gray color as threshold:
14493 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
14497 Truncate binary threshold, using gray color as threshold:
14499 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
14503 Threshold to zero, using gray color as threshold:
14505 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
14509 Inverted threshold to zero, using gray color as threshold:
14511 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
14516 Select the most representative frame in a given sequence of consecutive frames.
14518 The filter accepts the following options:
14522 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
14523 will pick one of them, and then handle the next batch of @var{n} frames until
14524 the end. Default is @code{100}.
14527 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
14528 value will result in a higher memory usage, so a high value is not recommended.
14530 @subsection Examples
14534 Extract one picture each 50 frames:
14540 Complete example of a thumbnail creation with @command{ffmpeg}:
14542 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
14548 Tile several successive frames together.
14550 The filter accepts the following options:
14555 Set the grid size (i.e. the number of lines and columns). For the syntax of
14556 this option, check the
14557 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14560 Set the maximum number of frames to render in the given area. It must be less
14561 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
14562 the area will be used.
14565 Set the outer border margin in pixels.
14568 Set the inner border thickness (i.e. the number of pixels between frames). For
14569 more advanced padding options (such as having different values for the edges),
14570 refer to the pad video filter.
14573 Specify the color of the unused area. For the syntax of this option, check the
14574 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
14578 Set the number of frames to overlap when tiling several successive frames together.
14579 The value must be between @code{0} and @var{nb_frames - 1}.
14582 @subsection Examples
14586 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
14588 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
14590 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
14591 duplicating each output frame to accommodate the originally detected frame
14595 Display @code{5} pictures in an area of @code{3x2} frames,
14596 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
14597 mixed flat and named options:
14599 tile=3x2:nb_frames=5:padding=7:margin=2
14603 @section tinterlace
14605 Perform various types of temporal field interlacing.
14607 Frames are counted starting from 1, so the first input frame is
14610 The filter accepts the following options:
14615 Specify the mode of the interlacing. This option can also be specified
14616 as a value alone. See below for a list of values for this option.
14618 Available values are:
14622 Move odd frames into the upper field, even into the lower field,
14623 generating a double height frame at half frame rate.
14627 Frame 1 Frame 2 Frame 3 Frame 4
14629 11111 22222 33333 44444
14630 11111 22222 33333 44444
14631 11111 22222 33333 44444
14632 11111 22222 33333 44444
14646 Only output odd frames, even frames are dropped, generating a frame with
14647 unchanged height at half frame rate.
14652 Frame 1 Frame 2 Frame 3 Frame 4
14654 11111 22222 33333 44444
14655 11111 22222 33333 44444
14656 11111 22222 33333 44444
14657 11111 22222 33333 44444
14667 Only output even frames, odd frames are dropped, generating a frame with
14668 unchanged height at half frame rate.
14673 Frame 1 Frame 2 Frame 3 Frame 4
14675 11111 22222 33333 44444
14676 11111 22222 33333 44444
14677 11111 22222 33333 44444
14678 11111 22222 33333 44444
14688 Expand each frame to full height, but pad alternate lines with black,
14689 generating a frame with double height at the same input frame rate.
14694 Frame 1 Frame 2 Frame 3 Frame 4
14696 11111 22222 33333 44444
14697 11111 22222 33333 44444
14698 11111 22222 33333 44444
14699 11111 22222 33333 44444
14702 11111 ..... 33333 .....
14703 ..... 22222 ..... 44444
14704 11111 ..... 33333 .....
14705 ..... 22222 ..... 44444
14706 11111 ..... 33333 .....
14707 ..... 22222 ..... 44444
14708 11111 ..... 33333 .....
14709 ..... 22222 ..... 44444
14713 @item interleave_top, 4
14714 Interleave the upper field from odd frames with the lower field from
14715 even frames, generating a frame with unchanged height at half frame rate.
14720 Frame 1 Frame 2 Frame 3 Frame 4
14722 11111<- 22222 33333<- 44444
14723 11111 22222<- 33333 44444<-
14724 11111<- 22222 33333<- 44444
14725 11111 22222<- 33333 44444<-
14735 @item interleave_bottom, 5
14736 Interleave the lower field from odd frames with the upper field from
14737 even frames, generating a frame with unchanged height at half frame rate.
14742 Frame 1 Frame 2 Frame 3 Frame 4
14744 11111 22222<- 33333 44444<-
14745 11111<- 22222 33333<- 44444
14746 11111 22222<- 33333 44444<-
14747 11111<- 22222 33333<- 44444
14757 @item interlacex2, 6
14758 Double frame rate with unchanged height. Frames are inserted each
14759 containing the second temporal field from the previous input frame and
14760 the first temporal field from the next input frame. This mode relies on
14761 the top_field_first flag. Useful for interlaced video displays with no
14762 field synchronisation.
14767 Frame 1 Frame 2 Frame 3 Frame 4
14769 11111 22222 33333 44444
14770 11111 22222 33333 44444
14771 11111 22222 33333 44444
14772 11111 22222 33333 44444
14775 11111 22222 22222 33333 33333 44444 44444
14776 11111 11111 22222 22222 33333 33333 44444
14777 11111 22222 22222 33333 33333 44444 44444
14778 11111 11111 22222 22222 33333 33333 44444
14783 Move odd frames into the upper field, even into the lower field,
14784 generating a double height frame at same frame rate.
14789 Frame 1 Frame 2 Frame 3 Frame 4
14791 11111 22222 33333 44444
14792 11111 22222 33333 44444
14793 11111 22222 33333 44444
14794 11111 22222 33333 44444
14797 11111 33333 33333 55555
14798 22222 22222 44444 44444
14799 11111 33333 33333 55555
14800 22222 22222 44444 44444
14801 11111 33333 33333 55555
14802 22222 22222 44444 44444
14803 11111 33333 33333 55555
14804 22222 22222 44444 44444
14809 Numeric values are deprecated but are accepted for backward
14810 compatibility reasons.
14812 Default mode is @code{merge}.
14815 Specify flags influencing the filter process.
14817 Available value for @var{flags} is:
14820 @item low_pass_filter, vlfp
14821 Enable linear vertical low-pass filtering in the filter.
14822 Vertical low-pass filtering is required when creating an interlaced
14823 destination from a progressive source which contains high-frequency
14824 vertical detail. Filtering will reduce interlace 'twitter' and Moire
14827 @item complex_filter, cvlfp
14828 Enable complex vertical low-pass filtering.
14829 This will slightly less reduce interlace 'twitter' and Moire
14830 patterning but better retain detail and subjective sharpness impression.
14834 Vertical low-pass filtering can only be enabled for @option{mode}
14835 @var{interleave_top} and @var{interleave_bottom}.
14840 Tone map colors from different dynamic ranges.
14842 This filter expects data in single precision floating point, as it needs to
14843 operate on (and can output) out-of-range values. Another filter, such as
14844 @ref{zscale}, is needed to convert the resulting frame to a usable format.
14846 The tonemapping algorithms implemented only work on linear light, so input
14847 data should be linearized beforehand (and possibly correctly tagged).
14850 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
14853 @subsection Options
14854 The filter accepts the following options.
14858 Set the tone map algorithm to use.
14860 Possible values are:
14863 Do not apply any tone map, only desaturate overbright pixels.
14866 Hard-clip any out-of-range values. Use it for perfect color accuracy for
14867 in-range values, while distorting out-of-range values.
14870 Stretch the entire reference gamut to a linear multiple of the display.
14873 Fit a logarithmic transfer between the tone curves.
14876 Preserve overall image brightness with a simple curve, using nonlinear
14877 contrast, which results in flattening details and degrading color accuracy.
14880 Preserve both dark and bright details better than @var{reinhard}, at the cost
14881 of slightly darkening everything. Use it when detail preservation is more
14882 important than color and brightness accuracy.
14885 Smoothly map out-of-range values, while retaining contrast and colors for
14886 in-range material as much as possible. Use it when color accuracy is more
14887 important than detail preservation.
14893 Tune the tone mapping algorithm.
14895 This affects the following algorithms:
14901 Specifies the scale factor to use while stretching.
14905 Specifies the exponent of the function.
14909 Specify an extra linear coefficient to multiply into the signal before clipping.
14913 Specify the local contrast coefficient at the display peak.
14914 Default to 0.5, which means that in-gamut values will be about half as bright
14921 Specify the transition point from linear to mobius transform. Every value
14922 below this point is guaranteed to be mapped 1:1. The higher the value, the
14923 more accurate the result will be, at the cost of losing bright details.
14924 Default to 0.3, which due to the steep initial slope still preserves in-range
14925 colors fairly accurately.
14929 Apply desaturation for highlights that exceed this level of brightness. The
14930 higher the parameter, the more color information will be preserved. This
14931 setting helps prevent unnaturally blown-out colors for super-highlights, by
14932 (smoothly) turning into white instead. This makes images feel more natural,
14933 at the cost of reducing information about out-of-range colors.
14935 The default of 2.0 is somewhat conservative and will mostly just apply to
14936 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
14938 This option works only if the input frame has a supported color tag.
14941 Override signal/nominal/reference peak with this value. Useful when the
14942 embedded peak information in display metadata is not reliable or when tone
14943 mapping from a lower range to a higher range.
14948 Transpose rows with columns in the input video and optionally flip it.
14950 It accepts the following parameters:
14955 Specify the transposition direction.
14957 Can assume the following values:
14959 @item 0, 4, cclock_flip
14960 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
14968 Rotate by 90 degrees clockwise, that is:
14976 Rotate by 90 degrees counterclockwise, that is:
14983 @item 3, 7, clock_flip
14984 Rotate by 90 degrees clockwise and vertically flip, that is:
14992 For values between 4-7, the transposition is only done if the input
14993 video geometry is portrait and not landscape. These values are
14994 deprecated, the @code{passthrough} option should be used instead.
14996 Numerical values are deprecated, and should be dropped in favor of
14997 symbolic constants.
15000 Do not apply the transposition if the input geometry matches the one
15001 specified by the specified value. It accepts the following values:
15004 Always apply transposition.
15006 Preserve portrait geometry (when @var{height} >= @var{width}).
15008 Preserve landscape geometry (when @var{width} >= @var{height}).
15011 Default value is @code{none}.
15014 For example to rotate by 90 degrees clockwise and preserve portrait
15017 transpose=dir=1:passthrough=portrait
15020 The command above can also be specified as:
15022 transpose=1:portrait
15026 Trim the input so that the output contains one continuous subpart of the input.
15028 It accepts the following parameters:
15031 Specify the time of the start of the kept section, i.e. the frame with the
15032 timestamp @var{start} will be the first frame in the output.
15035 Specify the time of the first frame that will be dropped, i.e. the frame
15036 immediately preceding the one with the timestamp @var{end} will be the last
15037 frame in the output.
15040 This is the same as @var{start}, except this option sets the start timestamp
15041 in timebase units instead of seconds.
15044 This is the same as @var{end}, except this option sets the end timestamp
15045 in timebase units instead of seconds.
15048 The maximum duration of the output in seconds.
15051 The number of the first frame that should be passed to the output.
15054 The number of the first frame that should be dropped.
15057 @option{start}, @option{end}, and @option{duration} are expressed as time
15058 duration specifications; see
15059 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15060 for the accepted syntax.
15062 Note that the first two sets of the start/end options and the @option{duration}
15063 option look at the frame timestamp, while the _frame variants simply count the
15064 frames that pass through the filter. Also note that this filter does not modify
15065 the timestamps. If you wish for the output timestamps to start at zero, insert a
15066 setpts filter after the trim filter.
15068 If multiple start or end options are set, this filter tries to be greedy and
15069 keep all the frames that match at least one of the specified constraints. To keep
15070 only the part that matches all the constraints at once, chain multiple trim
15073 The defaults are such that all the input is kept. So it is possible to set e.g.
15074 just the end values to keep everything before the specified time.
15079 Drop everything except the second minute of input:
15081 ffmpeg -i INPUT -vf trim=60:120
15085 Keep only the first second:
15087 ffmpeg -i INPUT -vf trim=duration=1
15092 @section unpremultiply
15093 Apply alpha unpremultiply effect to input video stream using first plane
15094 of second stream as alpha.
15096 Both streams must have same dimensions and same pixel format.
15098 The filter accepts the following option:
15102 Set which planes will be processed, unprocessed planes will be copied.
15103 By default value 0xf, all planes will be processed.
15105 If the format has 1 or 2 components, then luma is bit 0.
15106 If the format has 3 or 4 components:
15107 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15108 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15109 If present, the alpha channel is always the last bit.
15112 Do not require 2nd input for processing, instead use alpha plane from input stream.
15118 Sharpen or blur the input video.
15120 It accepts the following parameters:
15123 @item luma_msize_x, lx
15124 Set the luma matrix horizontal size. It must be an odd integer between
15125 3 and 23. The default value is 5.
15127 @item luma_msize_y, ly
15128 Set the luma matrix vertical size. It must be an odd integer between 3
15129 and 23. The default value is 5.
15131 @item luma_amount, la
15132 Set the luma effect strength. It must be a floating point number, reasonable
15133 values lay between -1.5 and 1.5.
15135 Negative values will blur the input video, while positive values will
15136 sharpen it, a value of zero will disable the effect.
15138 Default value is 1.0.
15140 @item chroma_msize_x, cx
15141 Set the chroma matrix horizontal size. It must be an odd integer
15142 between 3 and 23. The default value is 5.
15144 @item chroma_msize_y, cy
15145 Set the chroma matrix vertical size. It must be an odd integer
15146 between 3 and 23. The default value is 5.
15148 @item chroma_amount, ca
15149 Set the chroma effect strength. It must be a floating point number, reasonable
15150 values lay between -1.5 and 1.5.
15152 Negative values will blur the input video, while positive values will
15153 sharpen it, a value of zero will disable the effect.
15155 Default value is 0.0.
15159 All parameters are optional and default to the equivalent of the
15160 string '5:5:1.0:5:5:0.0'.
15162 @subsection Examples
15166 Apply strong luma sharpen effect:
15168 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15172 Apply a strong blur of both luma and chroma parameters:
15174 unsharp=7:7:-2:7:7:-2
15180 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15181 the image at several (or - in the case of @option{quality} level @code{8} - all)
15182 shifts and average the results.
15184 The way this differs from the behavior of spp is that uspp actually encodes &
15185 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15186 DCT similar to MJPEG.
15188 The filter accepts the following options:
15192 Set quality. This option defines the number of levels for averaging. It accepts
15193 an integer in the range 0-8. If set to @code{0}, the filter will have no
15194 effect. A value of @code{8} means the higher quality. For each increment of
15195 that value the speed drops by a factor of approximately 2. Default value is
15199 Force a constant quantization parameter. If not set, the filter will use the QP
15200 from the video stream (if available).
15203 @section vaguedenoiser
15205 Apply a wavelet based denoiser.
15207 It transforms each frame from the video input into the wavelet domain,
15208 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15209 the obtained coefficients. It does an inverse wavelet transform after.
15210 Due to wavelet properties, it should give a nice smoothed result, and
15211 reduced noise, without blurring picture features.
15213 This filter accepts the following options:
15217 The filtering strength. The higher, the more filtered the video will be.
15218 Hard thresholding can use a higher threshold than soft thresholding
15219 before the video looks overfiltered. Default value is 2.
15222 The filtering method the filter will use.
15224 It accepts the following values:
15227 All values under the threshold will be zeroed.
15230 All values under the threshold will be zeroed. All values above will be
15231 reduced by the threshold.
15234 Scales or nullifies coefficients - intermediary between (more) soft and
15235 (less) hard thresholding.
15238 Default is garrote.
15241 Number of times, the wavelet will decompose the picture. Picture can't
15242 be decomposed beyond a particular point (typically, 8 for a 640x480
15243 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15246 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15249 A list of the planes to process. By default all planes are processed.
15252 @section vectorscope
15254 Display 2 color component values in the two dimensional graph (which is called
15257 This filter accepts the following options:
15261 Set vectorscope mode.
15263 It accepts the following values:
15266 Gray values are displayed on graph, higher brightness means more pixels have
15267 same component color value on location in graph. This is the default mode.
15270 Gray values are displayed on graph. Surrounding pixels values which are not
15271 present in video frame are drawn in gradient of 2 color components which are
15272 set by option @code{x} and @code{y}. The 3rd color component is static.
15275 Actual color components values present in video frame are displayed on graph.
15278 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15279 on graph increases value of another color component, which is luminance by
15280 default values of @code{x} and @code{y}.
15283 Actual colors present in video frame are displayed on graph. If two different
15284 colors map to same position on graph then color with higher value of component
15285 not present in graph is picked.
15288 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15289 component picked from radial gradient.
15293 Set which color component will be represented on X-axis. Default is @code{1}.
15296 Set which color component will be represented on Y-axis. Default is @code{2}.
15299 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15300 of color component which represents frequency of (X, Y) location in graph.
15305 No envelope, this is default.
15308 Instant envelope, even darkest single pixel will be clearly highlighted.
15311 Hold maximum and minimum values presented in graph over time. This way you
15312 can still spot out of range values without constantly looking at vectorscope.
15315 Peak and instant envelope combined together.
15319 Set what kind of graticule to draw.
15327 Set graticule opacity.
15330 Set graticule flags.
15334 Draw graticule for white point.
15337 Draw graticule for black point.
15340 Draw color points short names.
15344 Set background opacity.
15346 @item lthreshold, l
15347 Set low threshold for color component not represented on X or Y axis.
15348 Values lower than this value will be ignored. Default is 0.
15349 Note this value is multiplied with actual max possible value one pixel component
15350 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15353 @item hthreshold, h
15354 Set high threshold for color component not represented on X or Y axis.
15355 Values higher than this value will be ignored. Default is 1.
15356 Note this value is multiplied with actual max possible value one pixel component
15357 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15358 is 0.9 * 255 = 230.
15360 @item colorspace, c
15361 Set what kind of colorspace to use when drawing graticule.
15370 @anchor{vidstabdetect}
15371 @section vidstabdetect
15373 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
15374 @ref{vidstabtransform} for pass 2.
15376 This filter generates a file with relative translation and rotation
15377 transform information about subsequent frames, which is then used by
15378 the @ref{vidstabtransform} filter.
15380 To enable compilation of this filter you need to configure FFmpeg with
15381 @code{--enable-libvidstab}.
15383 This filter accepts the following options:
15387 Set the path to the file used to write the transforms information.
15388 Default value is @file{transforms.trf}.
15391 Set how shaky the video is and how quick the camera is. It accepts an
15392 integer in the range 1-10, a value of 1 means little shakiness, a
15393 value of 10 means strong shakiness. Default value is 5.
15396 Set the accuracy of the detection process. It must be a value in the
15397 range 1-15. A value of 1 means low accuracy, a value of 15 means high
15398 accuracy. Default value is 15.
15401 Set stepsize of the search process. The region around minimum is
15402 scanned with 1 pixel resolution. Default value is 6.
15405 Set minimum contrast. Below this value a local measurement field is
15406 discarded. Must be a floating point value in the range 0-1. Default
15410 Set reference frame number for tripod mode.
15412 If enabled, the motion of the frames is compared to a reference frame
15413 in the filtered stream, identified by the specified number. The idea
15414 is to compensate all movements in a more-or-less static scene and keep
15415 the camera view absolutely still.
15417 If set to 0, it is disabled. The frames are counted starting from 1.
15420 Show fields and transforms in the resulting frames. It accepts an
15421 integer in the range 0-2. Default value is 0, which disables any
15425 @subsection Examples
15429 Use default values:
15435 Analyze strongly shaky movie and put the results in file
15436 @file{mytransforms.trf}:
15438 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
15442 Visualize the result of internal transformations in the resulting
15445 vidstabdetect=show=1
15449 Analyze a video with medium shakiness using @command{ffmpeg}:
15451 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
15455 @anchor{vidstabtransform}
15456 @section vidstabtransform
15458 Video stabilization/deshaking: pass 2 of 2,
15459 see @ref{vidstabdetect} for pass 1.
15461 Read a file with transform information for each frame and
15462 apply/compensate them. Together with the @ref{vidstabdetect}
15463 filter this can be used to deshake videos. See also
15464 @url{http://public.hronopik.de/vid.stab}. It is important to also use
15465 the @ref{unsharp} filter, see below.
15467 To enable compilation of this filter you need to configure FFmpeg with
15468 @code{--enable-libvidstab}.
15470 @subsection Options
15474 Set path to the file used to read the transforms. Default value is
15475 @file{transforms.trf}.
15478 Set the number of frames (value*2 + 1) used for lowpass filtering the
15479 camera movements. Default value is 10.
15481 For example a number of 10 means that 21 frames are used (10 in the
15482 past and 10 in the future) to smoothen the motion in the video. A
15483 larger value leads to a smoother video, but limits the acceleration of
15484 the camera (pan/tilt movements). 0 is a special case where a static
15485 camera is simulated.
15488 Set the camera path optimization algorithm.
15490 Accepted values are:
15493 gaussian kernel low-pass filter on camera motion (default)
15495 averaging on transformations
15499 Set maximal number of pixels to translate frames. Default value is -1,
15503 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
15504 value is -1, meaning no limit.
15507 Specify how to deal with borders that may be visible due to movement
15510 Available values are:
15513 keep image information from previous frame (default)
15515 fill the border black
15519 Invert transforms if set to 1. Default value is 0.
15522 Consider transforms as relative to previous frame if set to 1,
15523 absolute if set to 0. Default value is 0.
15526 Set percentage to zoom. A positive value will result in a zoom-in
15527 effect, a negative value in a zoom-out effect. Default value is 0 (no
15531 Set optimal zooming to avoid borders.
15533 Accepted values are:
15538 optimal static zoom value is determined (only very strong movements
15539 will lead to visible borders) (default)
15541 optimal adaptive zoom value is determined (no borders will be
15542 visible), see @option{zoomspeed}
15545 Note that the value given at zoom is added to the one calculated here.
15548 Set percent to zoom maximally each frame (enabled when
15549 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
15553 Specify type of interpolation.
15555 Available values are:
15560 linear only horizontal
15562 linear in both directions (default)
15564 cubic in both directions (slow)
15568 Enable virtual tripod mode if set to 1, which is equivalent to
15569 @code{relative=0:smoothing=0}. Default value is 0.
15571 Use also @code{tripod} option of @ref{vidstabdetect}.
15574 Increase log verbosity if set to 1. Also the detected global motions
15575 are written to the temporary file @file{global_motions.trf}. Default
15579 @subsection Examples
15583 Use @command{ffmpeg} for a typical stabilization with default values:
15585 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
15588 Note the use of the @ref{unsharp} filter which is always recommended.
15591 Zoom in a bit more and load transform data from a given file:
15593 vidstabtransform=zoom=5:input="mytransforms.trf"
15597 Smoothen the video even more:
15599 vidstabtransform=smoothing=30
15605 Flip the input video vertically.
15607 For example, to vertically flip a video with @command{ffmpeg}:
15609 ffmpeg -i in.avi -vf "vflip" out.avi
15615 Make or reverse a natural vignetting effect.
15617 The filter accepts the following options:
15621 Set lens angle expression as a number of radians.
15623 The value is clipped in the @code{[0,PI/2]} range.
15625 Default value: @code{"PI/5"}
15629 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
15633 Set forward/backward mode.
15635 Available modes are:
15638 The larger the distance from the central point, the darker the image becomes.
15641 The larger the distance from the central point, the brighter the image becomes.
15642 This can be used to reverse a vignette effect, though there is no automatic
15643 detection to extract the lens @option{angle} and other settings (yet). It can
15644 also be used to create a burning effect.
15647 Default value is @samp{forward}.
15650 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
15652 It accepts the following values:
15655 Evaluate expressions only once during the filter initialization.
15658 Evaluate expressions for each incoming frame. This is way slower than the
15659 @samp{init} mode since it requires all the scalers to be re-computed, but it
15660 allows advanced dynamic expressions.
15663 Default value is @samp{init}.
15666 Set dithering to reduce the circular banding effects. Default is @code{1}
15670 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
15671 Setting this value to the SAR of the input will make a rectangular vignetting
15672 following the dimensions of the video.
15674 Default is @code{1/1}.
15677 @subsection Expressions
15679 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
15680 following parameters.
15685 input width and height
15688 the number of input frame, starting from 0
15691 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
15692 @var{TB} units, NAN if undefined
15695 frame rate of the input video, NAN if the input frame rate is unknown
15698 the PTS (Presentation TimeStamp) of the filtered video frame,
15699 expressed in seconds, NAN if undefined
15702 time base of the input video
15706 @subsection Examples
15710 Apply simple strong vignetting effect:
15716 Make a flickering vignetting:
15718 vignette='PI/4+random(1)*PI/50':eval=frame
15723 @section vmafmotion
15725 Obtain the average vmaf motion score of a video.
15726 It is one of the component filters of VMAF.
15728 The obtained average motion score is printed through the logging system.
15730 In the below example the input file @file{ref.mpg} is being processed and score
15734 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
15738 Stack input videos vertically.
15740 All streams must be of same pixel format and of same width.
15742 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
15743 to create same output.
15745 The filter accept the following option:
15749 Set number of input streams. Default is 2.
15752 If set to 1, force the output to terminate when the shortest input
15753 terminates. Default value is 0.
15758 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
15759 Deinterlacing Filter").
15761 Based on the process described by Martin Weston for BBC R&D, and
15762 implemented based on the de-interlace algorithm written by Jim
15763 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
15764 uses filter coefficients calculated by BBC R&D.
15766 There are two sets of filter coefficients, so called "simple":
15767 and "complex". Which set of filter coefficients is used can
15768 be set by passing an optional parameter:
15772 Set the interlacing filter coefficients. Accepts one of the following values:
15776 Simple filter coefficient set.
15778 More-complex filter coefficient set.
15780 Default value is @samp{complex}.
15783 Specify which frames to deinterlace. Accept one of the following values:
15787 Deinterlace all frames,
15789 Only deinterlace frames marked as interlaced.
15792 Default value is @samp{all}.
15796 Video waveform monitor.
15798 The waveform monitor plots color component intensity. By default luminance
15799 only. Each column of the waveform corresponds to a column of pixels in the
15802 It accepts the following options:
15806 Can be either @code{row}, or @code{column}. Default is @code{column}.
15807 In row mode, the graph on the left side represents color component value 0 and
15808 the right side represents value = 255. In column mode, the top side represents
15809 color component value = 0 and bottom side represents value = 255.
15812 Set intensity. Smaller values are useful to find out how many values of the same
15813 luminance are distributed across input rows/columns.
15814 Default value is @code{0.04}. Allowed range is [0, 1].
15817 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
15818 In mirrored mode, higher values will be represented on the left
15819 side for @code{row} mode and at the top for @code{column} mode. Default is
15820 @code{1} (mirrored).
15824 It accepts the following values:
15827 Presents information identical to that in the @code{parade}, except
15828 that the graphs representing color components are superimposed directly
15831 This display mode makes it easier to spot relative differences or similarities
15832 in overlapping areas of the color components that are supposed to be identical,
15833 such as neutral whites, grays, or blacks.
15836 Display separate graph for the color components side by side in
15837 @code{row} mode or one below the other in @code{column} mode.
15840 Display separate graph for the color components side by side in
15841 @code{column} mode or one below the other in @code{row} mode.
15843 Using this display mode makes it easy to spot color casts in the highlights
15844 and shadows of an image, by comparing the contours of the top and the bottom
15845 graphs of each waveform. Since whites, grays, and blacks are characterized
15846 by exactly equal amounts of red, green, and blue, neutral areas of the picture
15847 should display three waveforms of roughly equal width/height. If not, the
15848 correction is easy to perform by making level adjustments the three waveforms.
15850 Default is @code{stack}.
15852 @item components, c
15853 Set which color components to display. Default is 1, which means only luminance
15854 or red color component if input is in RGB colorspace. If is set for example to
15855 7 it will display all 3 (if) available color components.
15860 No envelope, this is default.
15863 Instant envelope, minimum and maximum values presented in graph will be easily
15864 visible even with small @code{step} value.
15867 Hold minimum and maximum values presented in graph across time. This way you
15868 can still spot out of range values without constantly looking at waveforms.
15871 Peak and instant envelope combined together.
15877 No filtering, this is default.
15880 Luma and chroma combined together.
15883 Similar as above, but shows difference between blue and red chroma.
15886 Displays only chroma.
15889 Displays actual color value on waveform.
15892 Similar as above, but with luma showing frequency of chroma values.
15896 Set which graticule to display.
15900 Do not display graticule.
15903 Display green graticule showing legal broadcast ranges.
15907 Set graticule opacity.
15910 Set graticule flags.
15914 Draw numbers above lines. By default enabled.
15917 Draw dots instead of lines.
15921 Set scale used for displaying graticule.
15928 Default is digital.
15931 Set background opacity.
15934 @section weave, doubleweave
15936 The @code{weave} takes a field-based video input and join
15937 each two sequential fields into single frame, producing a new double
15938 height clip with half the frame rate and half the frame count.
15940 The @code{doubleweave} works same as @code{weave} but without
15941 halving frame rate and frame count.
15943 It accepts the following option:
15947 Set first field. Available values are:
15951 Set the frame as top-field-first.
15954 Set the frame as bottom-field-first.
15958 @subsection Examples
15962 Interlace video using @ref{select} and @ref{separatefields} filter:
15964 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
15969 Apply the xBR high-quality magnification filter which is designed for pixel
15970 art. It follows a set of edge-detection rules, see
15971 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
15973 It accepts the following option:
15977 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
15978 @code{3xBR} and @code{4} for @code{4xBR}.
15979 Default is @code{3}.
15985 Deinterlace the input video ("yadif" means "yet another deinterlacing
15988 It accepts the following parameters:
15994 The interlacing mode to adopt. It accepts one of the following values:
15997 @item 0, send_frame
15998 Output one frame for each frame.
15999 @item 1, send_field
16000 Output one frame for each field.
16001 @item 2, send_frame_nospatial
16002 Like @code{send_frame}, but it skips the spatial interlacing check.
16003 @item 3, send_field_nospatial
16004 Like @code{send_field}, but it skips the spatial interlacing check.
16007 The default value is @code{send_frame}.
16010 The picture field parity assumed for the input interlaced video. It accepts one
16011 of the following values:
16015 Assume the top field is first.
16017 Assume the bottom field is first.
16019 Enable automatic detection of field parity.
16022 The default value is @code{auto}.
16023 If the interlacing is unknown or the decoder does not export this information,
16024 top field first will be assumed.
16027 Specify which frames to deinterlace. Accept one of the following
16032 Deinterlace all frames.
16033 @item 1, interlaced
16034 Only deinterlace frames marked as interlaced.
16037 The default value is @code{all}.
16042 Apply Zoom & Pan effect.
16044 This filter accepts the following options:
16048 Set the zoom expression. Default is 1.
16052 Set the x and y expression. Default is 0.
16055 Set the duration expression in number of frames.
16056 This sets for how many number of frames effect will last for
16057 single input image.
16060 Set the output image size, default is 'hd720'.
16063 Set the output frame rate, default is '25'.
16066 Each expression can contain the following constants:
16085 Output frame count.
16089 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16090 for current input frame.
16094 'x' and 'y' of last output frame of previous input frame or 0 when there was
16095 not yet such frame (first input frame).
16098 Last calculated zoom from 'z' expression for current input frame.
16101 Last calculated zoom of last output frame of previous input frame.
16104 Number of output frames for current input frame. Calculated from 'd' expression
16105 for each input frame.
16108 number of output frames created for previous input frame
16111 Rational number: input width / input height
16114 sample aspect ratio
16117 display aspect ratio
16121 @subsection Examples
16125 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16127 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
16131 Zoom-in up to 1.5 and pan always at center of picture:
16133 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16137 Same as above but without pausing:
16139 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16145 Scale (resize) the input video, using the z.lib library:
16146 https://github.com/sekrit-twc/zimg.
16148 The zscale filter forces the output display aspect ratio to be the same
16149 as the input, by changing the output sample aspect ratio.
16151 If the input image format is different from the format requested by
16152 the next filter, the zscale filter will convert the input to the
16155 @subsection Options
16156 The filter accepts the following options.
16161 Set the output video dimension expression. Default value is the input
16164 If the @var{width} or @var{w} value is 0, the input width is used for
16165 the output. If the @var{height} or @var{h} value is 0, the input height
16166 is used for the output.
16168 If one and only one of the values is -n with n >= 1, the zscale filter
16169 will use a value that maintains the aspect ratio of the input image,
16170 calculated from the other specified dimension. After that it will,
16171 however, make sure that the calculated dimension is divisible by n and
16172 adjust the value if necessary.
16174 If both values are -n with n >= 1, the behavior will be identical to
16175 both values being set to 0 as previously detailed.
16177 See below for the list of accepted constants for use in the dimension
16181 Set the video size. For the syntax of this option, check the
16182 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16185 Set the dither type.
16187 Possible values are:
16192 @item error_diffusion
16198 Set the resize filter type.
16200 Possible values are:
16210 Default is bilinear.
16213 Set the color range.
16215 Possible values are:
16222 Default is same as input.
16225 Set the color primaries.
16227 Possible values are:
16237 Default is same as input.
16240 Set the transfer characteristics.
16242 Possible values are:
16256 Default is same as input.
16259 Set the colorspace matrix.
16261 Possible value are:
16272 Default is same as input.
16275 Set the input color range.
16277 Possible values are:
16284 Default is same as input.
16286 @item primariesin, pin
16287 Set the input color primaries.
16289 Possible values are:
16299 Default is same as input.
16301 @item transferin, tin
16302 Set the input transfer characteristics.
16304 Possible values are:
16315 Default is same as input.
16317 @item matrixin, min
16318 Set the input colorspace matrix.
16320 Possible value are:
16332 Set the output chroma location.
16334 Possible values are:
16345 @item chromalin, cin
16346 Set the input chroma location.
16348 Possible values are:
16360 Set the nominal peak luminance.
16363 The values of the @option{w} and @option{h} options are expressions
16364 containing the following constants:
16369 The input width and height
16373 These are the same as @var{in_w} and @var{in_h}.
16377 The output (scaled) width and height
16381 These are the same as @var{out_w} and @var{out_h}
16384 The same as @var{iw} / @var{ih}
16387 input sample aspect ratio
16390 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16394 horizontal and vertical input chroma subsample values. For example for the
16395 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16399 horizontal and vertical output chroma subsample values. For example for the
16400 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16406 @c man end VIDEO FILTERS
16408 @chapter Video Sources
16409 @c man begin VIDEO SOURCES
16411 Below is a description of the currently available video sources.
16415 Buffer video frames, and make them available to the filter chain.
16417 This source is mainly intended for a programmatic use, in particular
16418 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
16420 It accepts the following parameters:
16425 Specify the size (width and height) of the buffered video frames. For the
16426 syntax of this option, check the
16427 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16430 The input video width.
16433 The input video height.
16436 A string representing the pixel format of the buffered video frames.
16437 It may be a number corresponding to a pixel format, or a pixel format
16441 Specify the timebase assumed by the timestamps of the buffered frames.
16444 Specify the frame rate expected for the video stream.
16446 @item pixel_aspect, sar
16447 The sample (pixel) aspect ratio of the input video.
16450 Specify the optional parameters to be used for the scale filter which
16451 is automatically inserted when an input change is detected in the
16452 input size or format.
16454 @item hw_frames_ctx
16455 When using a hardware pixel format, this should be a reference to an
16456 AVHWFramesContext describing input frames.
16461 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
16464 will instruct the source to accept video frames with size 320x240 and
16465 with format "yuv410p", assuming 1/24 as the timestamps timebase and
16466 square pixels (1:1 sample aspect ratio).
16467 Since the pixel format with name "yuv410p" corresponds to the number 6
16468 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
16469 this example corresponds to:
16471 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
16474 Alternatively, the options can be specified as a flat string, but this
16475 syntax is deprecated:
16477 @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}]
16481 Create a pattern generated by an elementary cellular automaton.
16483 The initial state of the cellular automaton can be defined through the
16484 @option{filename} and @option{pattern} options. If such options are
16485 not specified an initial state is created randomly.
16487 At each new frame a new row in the video is filled with the result of
16488 the cellular automaton next generation. The behavior when the whole
16489 frame is filled is defined by the @option{scroll} option.
16491 This source accepts the following options:
16495 Read the initial cellular automaton state, i.e. the starting row, from
16496 the specified file.
16497 In the file, each non-whitespace character is considered an alive
16498 cell, a newline will terminate the row, and further characters in the
16499 file will be ignored.
16502 Read the initial cellular automaton state, i.e. the starting row, from
16503 the specified string.
16505 Each non-whitespace character in the string is considered an alive
16506 cell, a newline will terminate the row, and further characters in the
16507 string will be ignored.
16510 Set the video rate, that is the number of frames generated per second.
16513 @item random_fill_ratio, ratio
16514 Set the random fill ratio for the initial cellular automaton row. It
16515 is a floating point number value ranging from 0 to 1, defaults to
16518 This option is ignored when a file or a pattern is specified.
16520 @item random_seed, seed
16521 Set the seed for filling randomly the initial row, must be an integer
16522 included between 0 and UINT32_MAX. If not specified, or if explicitly
16523 set to -1, the filter will try to use a good random seed on a best
16527 Set the cellular automaton rule, it is a number ranging from 0 to 255.
16528 Default value is 110.
16531 Set the size of the output video. For the syntax of this option, check the
16532 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16534 If @option{filename} or @option{pattern} is specified, the size is set
16535 by default to the width of the specified initial state row, and the
16536 height is set to @var{width} * PHI.
16538 If @option{size} is set, it must contain the width of the specified
16539 pattern string, and the specified pattern will be centered in the
16542 If a filename or a pattern string is not specified, the size value
16543 defaults to "320x518" (used for a randomly generated initial state).
16546 If set to 1, scroll the output upward when all the rows in the output
16547 have been already filled. If set to 0, the new generated row will be
16548 written over the top row just after the bottom row is filled.
16551 @item start_full, full
16552 If set to 1, completely fill the output with generated rows before
16553 outputting the first frame.
16554 This is the default behavior, for disabling set the value to 0.
16557 If set to 1, stitch the left and right row edges together.
16558 This is the default behavior, for disabling set the value to 0.
16561 @subsection Examples
16565 Read the initial state from @file{pattern}, and specify an output of
16568 cellauto=f=pattern:s=200x400
16572 Generate a random initial row with a width of 200 cells, with a fill
16575 cellauto=ratio=2/3:s=200x200
16579 Create a pattern generated by rule 18 starting by a single alive cell
16580 centered on an initial row with width 100:
16582 cellauto=p=@@:s=100x400:full=0:rule=18
16586 Specify a more elaborated initial pattern:
16588 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
16593 @anchor{coreimagesrc}
16594 @section coreimagesrc
16595 Video source generated on GPU using Apple's CoreImage API on OSX.
16597 This video source is a specialized version of the @ref{coreimage} video filter.
16598 Use a core image generator at the beginning of the applied filterchain to
16599 generate the content.
16601 The coreimagesrc video source accepts the following options:
16603 @item list_generators
16604 List all available generators along with all their respective options as well as
16605 possible minimum and maximum values along with the default values.
16607 list_generators=true
16611 Specify the size of the sourced video. For the syntax of this option, check the
16612 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16613 The default value is @code{320x240}.
16616 Specify the frame rate of the sourced video, as the number of frames
16617 generated per second. It has to be a string in the format
16618 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16619 number or a valid video frame rate abbreviation. The default value is
16623 Set the sample aspect ratio of the sourced video.
16626 Set the duration of the sourced video. See
16627 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16628 for the accepted syntax.
16630 If not specified, or the expressed duration is negative, the video is
16631 supposed to be generated forever.
16634 Additionally, all options of the @ref{coreimage} video filter are accepted.
16635 A complete filterchain can be used for further processing of the
16636 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
16637 and examples for details.
16639 @subsection Examples
16644 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
16645 given as complete and escaped command-line for Apple's standard bash shell:
16647 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
16649 This example is equivalent to the QRCode example of @ref{coreimage} without the
16650 need for a nullsrc video source.
16654 @section mandelbrot
16656 Generate a Mandelbrot set fractal, and progressively zoom towards the
16657 point specified with @var{start_x} and @var{start_y}.
16659 This source accepts the following options:
16664 Set the terminal pts value. Default value is 400.
16667 Set the terminal scale value.
16668 Must be a floating point value. Default value is 0.3.
16671 Set the inner coloring mode, that is the algorithm used to draw the
16672 Mandelbrot fractal internal region.
16674 It shall assume one of the following values:
16679 Show time until convergence.
16681 Set color based on point closest to the origin of the iterations.
16686 Default value is @var{mincol}.
16689 Set the bailout value. Default value is 10.0.
16692 Set the maximum of iterations performed by the rendering
16693 algorithm. Default value is 7189.
16696 Set outer coloring mode.
16697 It shall assume one of following values:
16699 @item iteration_count
16700 Set iteration cound mode.
16701 @item normalized_iteration_count
16702 set normalized iteration count mode.
16704 Default value is @var{normalized_iteration_count}.
16707 Set frame rate, expressed as number of frames per second. Default
16711 Set frame size. For the syntax of this option, check the "Video
16712 size" section in the ffmpeg-utils manual. Default value is "640x480".
16715 Set the initial scale value. Default value is 3.0.
16718 Set the initial x position. Must be a floating point value between
16719 -100 and 100. Default value is -0.743643887037158704752191506114774.
16722 Set the initial y position. Must be a floating point value between
16723 -100 and 100. Default value is -0.131825904205311970493132056385139.
16728 Generate various test patterns, as generated by the MPlayer test filter.
16730 The size of the generated video is fixed, and is 256x256.
16731 This source is useful in particular for testing encoding features.
16733 This source accepts the following options:
16738 Specify the frame rate of the sourced video, as the number of frames
16739 generated per second. It has to be a string in the format
16740 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16741 number or a valid video frame rate abbreviation. The default value is
16745 Set the duration of the sourced video. See
16746 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16747 for the accepted syntax.
16749 If not specified, or the expressed duration is negative, the video is
16750 supposed to be generated forever.
16754 Set the number or the name of the test to perform. Supported tests are:
16770 Default value is "all", which will cycle through the list of all tests.
16775 mptestsrc=t=dc_luma
16778 will generate a "dc_luma" test pattern.
16780 @section frei0r_src
16782 Provide a frei0r source.
16784 To enable compilation of this filter you need to install the frei0r
16785 header and configure FFmpeg with @code{--enable-frei0r}.
16787 This source accepts the following parameters:
16792 The size of the video to generate. For the syntax of this option, check the
16793 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16796 The framerate of the generated video. It may be a string of the form
16797 @var{num}/@var{den} or a frame rate abbreviation.
16800 The name to the frei0r source to load. For more information regarding frei0r and
16801 how to set the parameters, read the @ref{frei0r} section in the video filters
16804 @item filter_params
16805 A '|'-separated list of parameters to pass to the frei0r source.
16809 For example, to generate a frei0r partik0l source with size 200x200
16810 and frame rate 10 which is overlaid on the overlay filter main input:
16812 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
16817 Generate a life pattern.
16819 This source is based on a generalization of John Conway's life game.
16821 The sourced input represents a life grid, each pixel represents a cell
16822 which can be in one of two possible states, alive or dead. Every cell
16823 interacts with its eight neighbours, which are the cells that are
16824 horizontally, vertically, or diagonally adjacent.
16826 At each interaction the grid evolves according to the adopted rule,
16827 which specifies the number of neighbor alive cells which will make a
16828 cell stay alive or born. The @option{rule} option allows one to specify
16831 This source accepts the following options:
16835 Set the file from which to read the initial grid state. In the file,
16836 each non-whitespace character is considered an alive cell, and newline
16837 is used to delimit the end of each row.
16839 If this option is not specified, the initial grid is generated
16843 Set the video rate, that is the number of frames generated per second.
16846 @item random_fill_ratio, ratio
16847 Set the random fill ratio for the initial random grid. It is a
16848 floating point number value ranging from 0 to 1, defaults to 1/PHI.
16849 It is ignored when a file is specified.
16851 @item random_seed, seed
16852 Set the seed for filling the initial random grid, must be an integer
16853 included between 0 and UINT32_MAX. If not specified, or if explicitly
16854 set to -1, the filter will try to use a good random seed on a best
16860 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
16861 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
16862 @var{NS} specifies the number of alive neighbor cells which make a
16863 live cell stay alive, and @var{NB} the number of alive neighbor cells
16864 which make a dead cell to become alive (i.e. to "born").
16865 "s" and "b" can be used in place of "S" and "B", respectively.
16867 Alternatively a rule can be specified by an 18-bits integer. The 9
16868 high order bits are used to encode the next cell state if it is alive
16869 for each number of neighbor alive cells, the low order bits specify
16870 the rule for "borning" new cells. Higher order bits encode for an
16871 higher number of neighbor cells.
16872 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
16873 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
16875 Default value is "S23/B3", which is the original Conway's game of life
16876 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
16877 cells, and will born a new cell if there are three alive cells around
16881 Set the size of the output video. For the syntax of this option, check the
16882 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16884 If @option{filename} is specified, the size is set by default to the
16885 same size of the input file. If @option{size} is set, it must contain
16886 the size specified in the input file, and the initial grid defined in
16887 that file is centered in the larger resulting area.
16889 If a filename is not specified, the size value defaults to "320x240"
16890 (used for a randomly generated initial grid).
16893 If set to 1, stitch the left and right grid edges together, and the
16894 top and bottom edges also. Defaults to 1.
16897 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
16898 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
16899 value from 0 to 255.
16902 Set the color of living (or new born) cells.
16905 Set the color of dead cells. If @option{mold} is set, this is the first color
16906 used to represent a dead cell.
16909 Set mold color, for definitely dead and moldy cells.
16911 For the syntax of these 3 color options, check the "Color" section in the
16912 ffmpeg-utils manual.
16915 @subsection Examples
16919 Read a grid from @file{pattern}, and center it on a grid of size
16922 life=f=pattern:s=300x300
16926 Generate a random grid of size 200x200, with a fill ratio of 2/3:
16928 life=ratio=2/3:s=200x200
16932 Specify a custom rule for evolving a randomly generated grid:
16938 Full example with slow death effect (mold) using @command{ffplay}:
16940 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
16947 @anchor{haldclutsrc}
16949 @anchor{rgbtestsrc}
16951 @anchor{smptehdbars}
16954 @anchor{yuvtestsrc}
16955 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
16957 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
16959 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
16961 The @code{color} source provides an uniformly colored input.
16963 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
16964 @ref{haldclut} filter.
16966 The @code{nullsrc} source returns unprocessed video frames. It is
16967 mainly useful to be employed in analysis / debugging tools, or as the
16968 source for filters which ignore the input data.
16970 The @code{rgbtestsrc} source generates an RGB test pattern useful for
16971 detecting RGB vs BGR issues. You should see a red, green and blue
16972 stripe from top to bottom.
16974 The @code{smptebars} source generates a color bars pattern, based on
16975 the SMPTE Engineering Guideline EG 1-1990.
16977 The @code{smptehdbars} source generates a color bars pattern, based on
16978 the SMPTE RP 219-2002.
16980 The @code{testsrc} source generates a test video pattern, showing a
16981 color pattern, a scrolling gradient and a timestamp. This is mainly
16982 intended for testing purposes.
16984 The @code{testsrc2} source is similar to testsrc, but supports more
16985 pixel formats instead of just @code{rgb24}. This allows using it as an
16986 input for other tests without requiring a format conversion.
16988 The @code{yuvtestsrc} source generates an YUV test pattern. You should
16989 see a y, cb and cr stripe from top to bottom.
16991 The sources accept the following parameters:
16996 Specify the alpha (opacity) of the background, only available in the
16997 @code{testsrc2} source. The value must be between 0 (fully transparent) and
16998 255 (fully opaque, the default).
17001 Specify the color of the source, only available in the @code{color}
17002 source. For the syntax of this option, check the "Color" section in the
17003 ffmpeg-utils manual.
17006 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17007 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17008 pixels to be used as identity matrix for 3D lookup tables. Each component is
17009 coded on a @code{1/(N*N)} scale.
17012 Specify the size of the sourced video. For the syntax of this option, check the
17013 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17014 The default value is @code{320x240}.
17016 This option is not available with the @code{haldclutsrc} filter.
17019 Specify the frame rate of the sourced video, as the number of frames
17020 generated per second. It has to be a string in the format
17021 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17022 number or a valid video frame rate abbreviation. The default value is
17026 Set the sample aspect ratio of the sourced video.
17029 Set the duration of the sourced video. See
17030 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17031 for the accepted syntax.
17033 If not specified, or the expressed duration is negative, the video is
17034 supposed to be generated forever.
17037 Set the number of decimals to show in the timestamp, only available in the
17038 @code{testsrc} source.
17040 The displayed timestamp value will correspond to the original
17041 timestamp value multiplied by the power of 10 of the specified
17042 value. Default value is 0.
17045 For example the following:
17047 testsrc=duration=5.3:size=qcif:rate=10
17050 will generate a video with a duration of 5.3 seconds, with size
17051 176x144 and a frame rate of 10 frames per second.
17053 The following graph description will generate a red source
17054 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17057 color=c=red@@0.2:s=qcif:r=10
17060 If the input content is to be ignored, @code{nullsrc} can be used. The
17061 following command generates noise in the luminance plane by employing
17062 the @code{geq} filter:
17064 nullsrc=s=256x256, geq=random(1)*255:128:128
17067 @subsection Commands
17069 The @code{color} source supports the following commands:
17073 Set the color of the created image. Accepts the same syntax of the
17074 corresponding @option{color} option.
17077 @c man end VIDEO SOURCES
17079 @chapter Video Sinks
17080 @c man begin VIDEO SINKS
17082 Below is a description of the currently available video sinks.
17084 @section buffersink
17086 Buffer video frames, and make them available to the end of the filter
17089 This sink is mainly intended for programmatic use, in particular
17090 through the interface defined in @file{libavfilter/buffersink.h}
17091 or the options system.
17093 It accepts a pointer to an AVBufferSinkContext structure, which
17094 defines the incoming buffers' formats, to be passed as the opaque
17095 parameter to @code{avfilter_init_filter} for initialization.
17099 Null video sink: do absolutely nothing with the input video. It is
17100 mainly useful as a template and for use in analysis / debugging
17103 @c man end VIDEO SINKS
17105 @chapter Multimedia Filters
17106 @c man begin MULTIMEDIA FILTERS
17108 Below is a description of the currently available multimedia filters.
17112 Convert input audio to a video output, displaying the audio bit scope.
17114 The filter accepts the following options:
17118 Set frame rate, expressed as number of frames per second. Default
17122 Specify the video size for the output. For the syntax of this option, check the
17123 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17124 Default value is @code{1024x256}.
17127 Specify list of colors separated by space or by '|' which will be used to
17128 draw channels. Unrecognized or missing colors will be replaced
17132 @section ahistogram
17134 Convert input audio to a video output, displaying the volume histogram.
17136 The filter accepts the following options:
17140 Specify how histogram is calculated.
17142 It accepts the following values:
17145 Use single histogram for all channels.
17147 Use separate histogram for each channel.
17149 Default is @code{single}.
17152 Set frame rate, expressed as number of frames per second. Default
17156 Specify the video size for the output. For the syntax of this option, check the
17157 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17158 Default value is @code{hd720}.
17163 It accepts the following values:
17174 reverse logarithmic
17176 Default is @code{log}.
17179 Set amplitude scale.
17181 It accepts the following values:
17188 Default is @code{log}.
17191 Set how much frames to accumulate in histogram.
17192 Defauls is 1. Setting this to -1 accumulates all frames.
17195 Set histogram ratio of window height.
17198 Set sonogram sliding.
17200 It accepts the following values:
17203 replace old rows with new ones.
17205 scroll from top to bottom.
17207 Default is @code{replace}.
17210 @section aphasemeter
17212 Convert input audio to a video output, displaying the audio phase.
17214 The filter accepts the following options:
17218 Set the output frame rate. Default value is @code{25}.
17221 Set the video size for the output. For the syntax of this option, check the
17222 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17223 Default value is @code{800x400}.
17228 Specify the red, green, blue contrast. Default values are @code{2},
17229 @code{7} and @code{1}.
17230 Allowed range is @code{[0, 255]}.
17233 Set color which will be used for drawing median phase. If color is
17234 @code{none} which is default, no median phase value will be drawn.
17237 Enable video output. Default is enabled.
17240 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17241 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17242 The @code{-1} means left and right channels are completely out of phase and
17243 @code{1} means channels are in phase.
17245 @section avectorscope
17247 Convert input audio to a video output, representing the audio vector
17250 The filter is used to measure the difference between channels of stereo
17251 audio stream. A monoaural signal, consisting of identical left and right
17252 signal, results in straight vertical line. Any stereo separation is visible
17253 as a deviation from this line, creating a Lissajous figure.
17254 If the straight (or deviation from it) but horizontal line appears this
17255 indicates that the left and right channels are out of phase.
17257 The filter accepts the following options:
17261 Set the vectorscope mode.
17263 Available values are:
17266 Lissajous rotated by 45 degrees.
17269 Same as above but not rotated.
17272 Shape resembling half of circle.
17275 Default value is @samp{lissajous}.
17278 Set the video size for the output. For the syntax of this option, check the
17279 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17280 Default value is @code{400x400}.
17283 Set the output frame rate. Default value is @code{25}.
17289 Specify the red, green, blue and alpha contrast. Default values are @code{40},
17290 @code{160}, @code{80} and @code{255}.
17291 Allowed range is @code{[0, 255]}.
17297 Specify the red, green, blue and alpha fade. Default values are @code{15},
17298 @code{10}, @code{5} and @code{5}.
17299 Allowed range is @code{[0, 255]}.
17302 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
17303 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
17306 Set the vectorscope drawing mode.
17308 Available values are:
17311 Draw dot for each sample.
17314 Draw line between previous and current sample.
17317 Default value is @samp{dot}.
17320 Specify amplitude scale of audio samples.
17322 Available values are:
17338 Swap left channel axis with right channel axis.
17348 Mirror only x axis.
17351 Mirror only y axis.
17359 @subsection Examples
17363 Complete example using @command{ffplay}:
17365 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17366 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
17370 @section bench, abench
17372 Benchmark part of a filtergraph.
17374 The filter accepts the following options:
17378 Start or stop a timer.
17380 Available values are:
17383 Get the current time, set it as frame metadata (using the key
17384 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
17387 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
17388 the input frame metadata to get the time difference. Time difference, average,
17389 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
17390 @code{min}) are then printed. The timestamps are expressed in seconds.
17394 @subsection Examples
17398 Benchmark @ref{selectivecolor} filter:
17400 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
17406 Concatenate audio and video streams, joining them together one after the
17409 The filter works on segments of synchronized video and audio streams. All
17410 segments must have the same number of streams of each type, and that will
17411 also be the number of streams at output.
17413 The filter accepts the following options:
17418 Set the number of segments. Default is 2.
17421 Set the number of output video streams, that is also the number of video
17422 streams in each segment. Default is 1.
17425 Set the number of output audio streams, that is also the number of audio
17426 streams in each segment. Default is 0.
17429 Activate unsafe mode: do not fail if segments have a different format.
17433 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
17434 @var{a} audio outputs.
17436 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
17437 segment, in the same order as the outputs, then the inputs for the second
17440 Related streams do not always have exactly the same duration, for various
17441 reasons including codec frame size or sloppy authoring. For that reason,
17442 related synchronized streams (e.g. a video and its audio track) should be
17443 concatenated at once. The concat filter will use the duration of the longest
17444 stream in each segment (except the last one), and if necessary pad shorter
17445 audio streams with silence.
17447 For this filter to work correctly, all segments must start at timestamp 0.
17449 All corresponding streams must have the same parameters in all segments; the
17450 filtering system will automatically select a common pixel format for video
17451 streams, and a common sample format, sample rate and channel layout for
17452 audio streams, but other settings, such as resolution, must be converted
17453 explicitly by the user.
17455 Different frame rates are acceptable but will result in variable frame rate
17456 at output; be sure to configure the output file to handle it.
17458 @subsection Examples
17462 Concatenate an opening, an episode and an ending, all in bilingual version
17463 (video in stream 0, audio in streams 1 and 2):
17465 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
17466 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
17467 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
17468 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
17472 Concatenate two parts, handling audio and video separately, using the
17473 (a)movie sources, and adjusting the resolution:
17475 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
17476 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
17477 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
17479 Note that a desync will happen at the stitch if the audio and video streams
17480 do not have exactly the same duration in the first file.
17484 @section drawgraph, adrawgraph
17486 Draw a graph using input video or audio metadata.
17488 It accepts the following parameters:
17492 Set 1st frame metadata key from which metadata values will be used to draw a graph.
17495 Set 1st foreground color expression.
17498 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
17501 Set 2nd foreground color expression.
17504 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
17507 Set 3rd foreground color expression.
17510 Set 4th frame metadata key from which metadata values will be used to draw a graph.
17513 Set 4th foreground color expression.
17516 Set minimal value of metadata value.
17519 Set maximal value of metadata value.
17522 Set graph background color. Default is white.
17527 Available values for mode is:
17534 Default is @code{line}.
17539 Available values for slide is:
17542 Draw new frame when right border is reached.
17545 Replace old columns with new ones.
17548 Scroll from right to left.
17551 Scroll from left to right.
17554 Draw single picture.
17557 Default is @code{frame}.
17560 Set size of graph video. For the syntax of this option, check the
17561 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17562 The default value is @code{900x256}.
17564 The foreground color expressions can use the following variables:
17567 Minimal value of metadata value.
17570 Maximal value of metadata value.
17573 Current metadata key value.
17576 The color is defined as 0xAABBGGRR.
17579 Example using metadata from @ref{signalstats} filter:
17581 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
17584 Example using metadata from @ref{ebur128} filter:
17586 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
17592 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
17593 it unchanged. By default, it logs a message at a frequency of 10Hz with the
17594 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
17595 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
17597 The filter also has a video output (see the @var{video} option) with a real
17598 time graph to observe the loudness evolution. The graphic contains the logged
17599 message mentioned above, so it is not printed anymore when this option is set,
17600 unless the verbose logging is set. The main graphing area contains the
17601 short-term loudness (3 seconds of analysis), and the gauge on the right is for
17602 the momentary loudness (400 milliseconds).
17604 More information about the Loudness Recommendation EBU R128 on
17605 @url{http://tech.ebu.ch/loudness}.
17607 The filter accepts the following options:
17612 Activate the video output. The audio stream is passed unchanged whether this
17613 option is set or no. The video stream will be the first output stream if
17614 activated. Default is @code{0}.
17617 Set the video size. This option is for video only. For the syntax of this
17619 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17620 Default and minimum resolution is @code{640x480}.
17623 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
17624 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
17625 other integer value between this range is allowed.
17628 Set metadata injection. If set to @code{1}, the audio input will be segmented
17629 into 100ms output frames, each of them containing various loudness information
17630 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
17632 Default is @code{0}.
17635 Force the frame logging level.
17637 Available values are:
17640 information logging level
17642 verbose logging level
17645 By default, the logging level is set to @var{info}. If the @option{video} or
17646 the @option{metadata} options are set, it switches to @var{verbose}.
17651 Available modes can be cumulated (the option is a @code{flag} type). Possible
17655 Disable any peak mode (default).
17657 Enable sample-peak mode.
17659 Simple peak mode looking for the higher sample value. It logs a message
17660 for sample-peak (identified by @code{SPK}).
17662 Enable true-peak mode.
17664 If enabled, the peak lookup is done on an over-sampled version of the input
17665 stream for better peak accuracy. It logs a message for true-peak.
17666 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
17667 This mode requires a build with @code{libswresample}.
17671 Treat mono input files as "dual mono". If a mono file is intended for playback
17672 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
17673 If set to @code{true}, this option will compensate for this effect.
17674 Multi-channel input files are not affected by this option.
17677 Set a specific pan law to be used for the measurement of dual mono files.
17678 This parameter is optional, and has a default value of -3.01dB.
17681 @subsection Examples
17685 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
17687 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
17691 Run an analysis with @command{ffmpeg}:
17693 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
17697 @section interleave, ainterleave
17699 Temporally interleave frames from several inputs.
17701 @code{interleave} works with video inputs, @code{ainterleave} with audio.
17703 These filters read frames from several inputs and send the oldest
17704 queued frame to the output.
17706 Input streams must have well defined, monotonically increasing frame
17709 In order to submit one frame to output, these filters need to enqueue
17710 at least one frame for each input, so they cannot work in case one
17711 input is not yet terminated and will not receive incoming frames.
17713 For example consider the case when one input is a @code{select} filter
17714 which always drops input frames. The @code{interleave} filter will keep
17715 reading from that input, but it will never be able to send new frames
17716 to output until the input sends an end-of-stream signal.
17718 Also, depending on inputs synchronization, the filters will drop
17719 frames in case one input receives more frames than the other ones, and
17720 the queue is already filled.
17722 These filters accept the following options:
17726 Set the number of different inputs, it is 2 by default.
17729 @subsection Examples
17733 Interleave frames belonging to different streams using @command{ffmpeg}:
17735 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
17739 Add flickering blur effect:
17741 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
17745 @section metadata, ametadata
17747 Manipulate frame metadata.
17749 This filter accepts the following options:
17753 Set mode of operation of the filter.
17755 Can be one of the following:
17759 If both @code{value} and @code{key} is set, select frames
17760 which have such metadata. If only @code{key} is set, select
17761 every frame that has such key in metadata.
17764 Add new metadata @code{key} and @code{value}. If key is already available
17768 Modify value of already present key.
17771 If @code{value} is set, delete only keys that have such value.
17772 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
17776 Print key and its value if metadata was found. If @code{key} is not set print all
17777 metadata values available in frame.
17781 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
17784 Set metadata value which will be used. This option is mandatory for
17785 @code{modify} and @code{add} mode.
17788 Which function to use when comparing metadata value and @code{value}.
17790 Can be one of following:
17794 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
17797 Values are interpreted as strings, returns true if metadata value starts with
17798 the @code{value} option string.
17801 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
17804 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
17807 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
17810 Values are interpreted as floats, returns true if expression from option @code{expr}
17815 Set expression which is used when @code{function} is set to @code{expr}.
17816 The expression is evaluated through the eval API and can contain the following
17821 Float representation of @code{value} from metadata key.
17824 Float representation of @code{value} as supplied by user in @code{value} option.
17828 If specified in @code{print} mode, output is written to the named file. Instead of
17829 plain filename any writable url can be specified. Filename ``-'' is a shorthand
17830 for standard output. If @code{file} option is not set, output is written to the log
17831 with AV_LOG_INFO loglevel.
17835 @subsection Examples
17839 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
17842 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
17845 Print silencedetect output to file @file{metadata.txt}.
17847 silencedetect,ametadata=mode=print:file=metadata.txt
17850 Direct all metadata to a pipe with file descriptor 4.
17852 metadata=mode=print:file='pipe\:4'
17856 @section perms, aperms
17858 Set read/write permissions for the output frames.
17860 These filters are mainly aimed at developers to test direct path in the
17861 following filter in the filtergraph.
17863 The filters accept the following options:
17867 Select the permissions mode.
17869 It accepts the following values:
17872 Do nothing. This is the default.
17874 Set all the output frames read-only.
17876 Set all the output frames directly writable.
17878 Make the frame read-only if writable, and writable if read-only.
17880 Set each output frame read-only or writable randomly.
17884 Set the seed for the @var{random} mode, must be an integer included between
17885 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
17886 @code{-1}, the filter will try to use a good random seed on a best effort
17890 Note: in case of auto-inserted filter between the permission filter and the
17891 following one, the permission might not be received as expected in that
17892 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
17893 perms/aperms filter can avoid this problem.
17895 @section realtime, arealtime
17897 Slow down filtering to match real time approximately.
17899 These filters will pause the filtering for a variable amount of time to
17900 match the output rate with the input timestamps.
17901 They are similar to the @option{re} option to @code{ffmpeg}.
17903 They accept the following options:
17907 Time limit for the pauses. Any pause longer than that will be considered
17908 a timestamp discontinuity and reset the timer. Default is 2 seconds.
17912 @section select, aselect
17914 Select frames to pass in output.
17916 This filter accepts the following options:
17921 Set expression, which is evaluated for each input frame.
17923 If the expression is evaluated to zero, the frame is discarded.
17925 If the evaluation result is negative or NaN, the frame is sent to the
17926 first output; otherwise it is sent to the output with index
17927 @code{ceil(val)-1}, assuming that the input index starts from 0.
17929 For example a value of @code{1.2} corresponds to the output with index
17930 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
17933 Set the number of outputs. The output to which to send the selected
17934 frame is based on the result of the evaluation. Default value is 1.
17937 The expression can contain the following constants:
17941 The (sequential) number of the filtered frame, starting from 0.
17944 The (sequential) number of the selected frame, starting from 0.
17946 @item prev_selected_n
17947 The sequential number of the last selected frame. It's NAN if undefined.
17950 The timebase of the input timestamps.
17953 The PTS (Presentation TimeStamp) of the filtered video frame,
17954 expressed in @var{TB} units. It's NAN if undefined.
17957 The PTS of the filtered video frame,
17958 expressed in seconds. It's NAN if undefined.
17961 The PTS of the previously filtered video frame. It's NAN if undefined.
17963 @item prev_selected_pts
17964 The PTS of the last previously filtered video frame. It's NAN if undefined.
17966 @item prev_selected_t
17967 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
17970 The PTS of the first video frame in the video. It's NAN if undefined.
17973 The time of the first video frame in the video. It's NAN if undefined.
17975 @item pict_type @emph{(video only)}
17976 The type of the filtered frame. It can assume one of the following
17988 @item interlace_type @emph{(video only)}
17989 The frame interlace type. It can assume one of the following values:
17992 The frame is progressive (not interlaced).
17994 The frame is top-field-first.
17996 The frame is bottom-field-first.
17999 @item consumed_sample_n @emph{(audio only)}
18000 the number of selected samples before the current frame
18002 @item samples_n @emph{(audio only)}
18003 the number of samples in the current frame
18005 @item sample_rate @emph{(audio only)}
18006 the input sample rate
18009 This is 1 if the filtered frame is a key-frame, 0 otherwise.
18012 the position in the file of the filtered frame, -1 if the information
18013 is not available (e.g. for synthetic video)
18015 @item scene @emph{(video only)}
18016 value between 0 and 1 to indicate a new scene; a low value reflects a low
18017 probability for the current frame to introduce a new scene, while a higher
18018 value means the current frame is more likely to be one (see the example below)
18020 @item concatdec_select
18021 The concat demuxer can select only part of a concat input file by setting an
18022 inpoint and an outpoint, but the output packets may not be entirely contained
18023 in the selected interval. By using this variable, it is possible to skip frames
18024 generated by the concat demuxer which are not exactly contained in the selected
18027 This works by comparing the frame pts against the @var{lavf.concat.start_time}
18028 and the @var{lavf.concat.duration} packet metadata values which are also
18029 present in the decoded frames.
18031 The @var{concatdec_select} variable is -1 if the frame pts is at least
18032 start_time and either the duration metadata is missing or the frame pts is less
18033 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
18036 That basically means that an input frame is selected if its pts is within the
18037 interval set by the concat demuxer.
18041 The default value of the select expression is "1".
18043 @subsection Examples
18047 Select all frames in input:
18052 The example above is the same as:
18064 Select only I-frames:
18066 select='eq(pict_type\,I)'
18070 Select one frame every 100:
18072 select='not(mod(n\,100))'
18076 Select only frames contained in the 10-20 time interval:
18078 select=between(t\,10\,20)
18082 Select only I-frames contained in the 10-20 time interval:
18084 select=between(t\,10\,20)*eq(pict_type\,I)
18088 Select frames with a minimum distance of 10 seconds:
18090 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18094 Use aselect to select only audio frames with samples number > 100:
18096 aselect='gt(samples_n\,100)'
18100 Create a mosaic of the first scenes:
18102 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18105 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18109 Send even and odd frames to separate outputs, and compose them:
18111 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18115 Select useful frames from an ffconcat file which is using inpoints and
18116 outpoints but where the source files are not intra frame only.
18118 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18122 @section sendcmd, asendcmd
18124 Send commands to filters in the filtergraph.
18126 These filters read commands to be sent to other filters in the
18129 @code{sendcmd} must be inserted between two video filters,
18130 @code{asendcmd} must be inserted between two audio filters, but apart
18131 from that they act the same way.
18133 The specification of commands can be provided in the filter arguments
18134 with the @var{commands} option, or in a file specified by the
18135 @var{filename} option.
18137 These filters accept the following options:
18140 Set the commands to be read and sent to the other filters.
18142 Set the filename of the commands to be read and sent to the other
18146 @subsection Commands syntax
18148 A commands description consists of a sequence of interval
18149 specifications, comprising a list of commands to be executed when a
18150 particular event related to that interval occurs. The occurring event
18151 is typically the current frame time entering or leaving a given time
18154 An interval is specified by the following syntax:
18156 @var{START}[-@var{END}] @var{COMMANDS};
18159 The time interval is specified by the @var{START} and @var{END} times.
18160 @var{END} is optional and defaults to the maximum time.
18162 The current frame time is considered within the specified interval if
18163 it is included in the interval [@var{START}, @var{END}), that is when
18164 the time is greater or equal to @var{START} and is lesser than
18167 @var{COMMANDS} consists of a sequence of one or more command
18168 specifications, separated by ",", relating to that interval. The
18169 syntax of a command specification is given by:
18171 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18174 @var{FLAGS} is optional and specifies the type of events relating to
18175 the time interval which enable sending the specified command, and must
18176 be a non-null sequence of identifier flags separated by "+" or "|" and
18177 enclosed between "[" and "]".
18179 The following flags are recognized:
18182 The command is sent when the current frame timestamp enters the
18183 specified interval. In other words, the command is sent when the
18184 previous frame timestamp was not in the given interval, and the
18188 The command is sent when the current frame timestamp leaves the
18189 specified interval. In other words, the command is sent when the
18190 previous frame timestamp was in the given interval, and the
18194 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18197 @var{TARGET} specifies the target of the command, usually the name of
18198 the filter class or a specific filter instance name.
18200 @var{COMMAND} specifies the name of the command for the target filter.
18202 @var{ARG} is optional and specifies the optional list of argument for
18203 the given @var{COMMAND}.
18205 Between one interval specification and another, whitespaces, or
18206 sequences of characters starting with @code{#} until the end of line,
18207 are ignored and can be used to annotate comments.
18209 A simplified BNF description of the commands specification syntax
18212 @var{COMMAND_FLAG} ::= "enter" | "leave"
18213 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18214 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18215 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18216 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18217 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18220 @subsection Examples
18224 Specify audio tempo change at second 4:
18226 asendcmd=c='4.0 atempo tempo 1.5',atempo
18230 Target a specific filter instance:
18232 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18236 Specify a list of drawtext and hue commands in a file.
18238 # show text in the interval 5-10
18239 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18240 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18242 # desaturate the image in the interval 15-20
18243 15.0-20.0 [enter] hue s 0,
18244 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18246 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18248 # apply an exponential saturation fade-out effect, starting from time 25
18249 25 [enter] hue s exp(25-t)
18252 A filtergraph allowing to read and process the above command list
18253 stored in a file @file{test.cmd}, can be specified with:
18255 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18260 @section setpts, asetpts
18262 Change the PTS (presentation timestamp) of the input frames.
18264 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18266 This filter accepts the following options:
18271 The expression which is evaluated for each frame to construct its timestamp.
18275 The expression is evaluated through the eval API and can contain the following
18280 frame rate, only defined for constant frame-rate video
18283 The presentation timestamp in input
18286 The count of the input frame for video or the number of consumed samples,
18287 not including the current frame for audio, starting from 0.
18289 @item NB_CONSUMED_SAMPLES
18290 The number of consumed samples, not including the current frame (only
18293 @item NB_SAMPLES, S
18294 The number of samples in the current frame (only audio)
18296 @item SAMPLE_RATE, SR
18297 The audio sample rate.
18300 The PTS of the first frame.
18303 the time in seconds of the first frame
18306 State whether the current frame is interlaced.
18309 the time in seconds of the current frame
18312 original position in the file of the frame, or undefined if undefined
18313 for the current frame
18316 The previous input PTS.
18319 previous input time in seconds
18322 The previous output PTS.
18325 previous output time in seconds
18328 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
18332 The wallclock (RTC) time at the start of the movie in microseconds.
18335 The timebase of the input timestamps.
18339 @subsection Examples
18343 Start counting PTS from zero
18345 setpts=PTS-STARTPTS
18349 Apply fast motion effect:
18355 Apply slow motion effect:
18361 Set fixed rate of 25 frames per second:
18367 Set fixed rate 25 fps with some jitter:
18369 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
18373 Apply an offset of 10 seconds to the input PTS:
18379 Generate timestamps from a "live source" and rebase onto the current timebase:
18381 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
18385 Generate timestamps by counting samples:
18392 @section settb, asettb
18394 Set the timebase to use for the output frames timestamps.
18395 It is mainly useful for testing timebase configuration.
18397 It accepts the following parameters:
18402 The expression which is evaluated into the output timebase.
18406 The value for @option{tb} is an arithmetic expression representing a
18407 rational. The expression can contain the constants "AVTB" (the default
18408 timebase), "intb" (the input timebase) and "sr" (the sample rate,
18409 audio only). Default value is "intb".
18411 @subsection Examples
18415 Set the timebase to 1/25:
18421 Set the timebase to 1/10:
18427 Set the timebase to 1001/1000:
18433 Set the timebase to 2*intb:
18439 Set the default timebase value:
18446 Convert input audio to a video output representing frequency spectrum
18447 logarithmically using Brown-Puckette constant Q transform algorithm with
18448 direct frequency domain coefficient calculation (but the transform itself
18449 is not really constant Q, instead the Q factor is actually variable/clamped),
18450 with musical tone scale, from E0 to D#10.
18452 The filter accepts the following options:
18456 Specify the video size for the output. It must be even. For the syntax of this option,
18457 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18458 Default value is @code{1920x1080}.
18461 Set the output frame rate. Default value is @code{25}.
18464 Set the bargraph height. It must be even. Default value is @code{-1} which
18465 computes the bargraph height automatically.
18468 Set the axis height. It must be even. Default value is @code{-1} which computes
18469 the axis height automatically.
18472 Set the sonogram height. It must be even. Default value is @code{-1} which
18473 computes the sonogram height automatically.
18476 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
18477 instead. Default value is @code{1}.
18479 @item sono_v, volume
18480 Specify the sonogram volume expression. It can contain variables:
18483 the @var{bar_v} evaluated expression
18484 @item frequency, freq, f
18485 the frequency where it is evaluated
18486 @item timeclamp, tc
18487 the value of @var{timeclamp} option
18491 @item a_weighting(f)
18492 A-weighting of equal loudness
18493 @item b_weighting(f)
18494 B-weighting of equal loudness
18495 @item c_weighting(f)
18496 C-weighting of equal loudness.
18498 Default value is @code{16}.
18500 @item bar_v, volume2
18501 Specify the bargraph volume expression. It can contain variables:
18504 the @var{sono_v} evaluated expression
18505 @item frequency, freq, f
18506 the frequency where it is evaluated
18507 @item timeclamp, tc
18508 the value of @var{timeclamp} option
18512 @item a_weighting(f)
18513 A-weighting of equal loudness
18514 @item b_weighting(f)
18515 B-weighting of equal loudness
18516 @item c_weighting(f)
18517 C-weighting of equal loudness.
18519 Default value is @code{sono_v}.
18521 @item sono_g, gamma
18522 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
18523 higher gamma makes the spectrum having more range. Default value is @code{3}.
18524 Acceptable range is @code{[1, 7]}.
18526 @item bar_g, gamma2
18527 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
18531 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
18532 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
18534 @item timeclamp, tc
18535 Specify the transform timeclamp. At low frequency, there is trade-off between
18536 accuracy in time domain and frequency domain. If timeclamp is lower,
18537 event in time domain is represented more accurately (such as fast bass drum),
18538 otherwise event in frequency domain is represented more accurately
18539 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
18542 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
18543 limits future samples by applying asymmetric windowing in time domain, useful
18544 when low latency is required. Accepted range is @code{[0, 1]}.
18547 Specify the transform base frequency. Default value is @code{20.01523126408007475},
18548 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
18551 Specify the transform end frequency. Default value is @code{20495.59681441799654},
18552 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
18555 This option is deprecated and ignored.
18558 Specify the transform length in time domain. Use this option to control accuracy
18559 trade-off between time domain and frequency domain at every frequency sample.
18560 It can contain variables:
18562 @item frequency, freq, f
18563 the frequency where it is evaluated
18564 @item timeclamp, tc
18565 the value of @var{timeclamp} option.
18567 Default value is @code{384*tc/(384+tc*f)}.
18570 Specify the transform count for every video frame. Default value is @code{6}.
18571 Acceptable range is @code{[1, 30]}.
18574 Specify the transform count for every single pixel. Default value is @code{0},
18575 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
18578 Specify font file for use with freetype to draw the axis. If not specified,
18579 use embedded font. Note that drawing with font file or embedded font is not
18580 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
18584 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
18585 The : in the pattern may be replaced by | to avoid unnecessary escaping.
18588 Specify font color expression. This is arithmetic expression that should return
18589 integer value 0xRRGGBB. It can contain variables:
18591 @item frequency, freq, f
18592 the frequency where it is evaluated
18593 @item timeclamp, tc
18594 the value of @var{timeclamp} option
18599 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
18600 @item r(x), g(x), b(x)
18601 red, green, and blue value of intensity x.
18603 Default value is @code{st(0, (midi(f)-59.5)/12);
18604 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
18605 r(1-ld(1)) + b(ld(1))}.
18608 Specify image file to draw the axis. This option override @var{fontfile} and
18609 @var{fontcolor} option.
18612 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
18613 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
18614 Default value is @code{1}.
18617 Set colorspace. The accepted values are:
18620 Unspecified (default)
18629 BT.470BG or BT.601-6 625
18632 SMPTE-170M or BT.601-6 525
18638 BT.2020 with non-constant luminance
18643 Set spectrogram color scheme. This is list of floating point values with format
18644 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
18645 The default is @code{1|0.5|0|0|0.5|1}.
18649 @subsection Examples
18653 Playing audio while showing the spectrum:
18655 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
18659 Same as above, but with frame rate 30 fps:
18661 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
18665 Playing at 1280x720:
18667 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
18671 Disable sonogram display:
18677 A1 and its harmonics: A1, A2, (near)E3, A3:
18679 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),
18680 asplit[a][out1]; [a] showcqt [out0]'
18684 Same as above, but with more accuracy in frequency domain:
18686 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),
18687 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
18693 bar_v=10:sono_v=bar_v*a_weighting(f)
18697 Custom gamma, now spectrum is linear to the amplitude.
18703 Custom tlength equation:
18705 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)))'
18709 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
18711 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
18715 Custom font using fontconfig:
18717 font='Courier New,Monospace,mono|bold'
18721 Custom frequency range with custom axis using image file:
18723 axisfile=myaxis.png:basefreq=40:endfreq=10000
18729 Convert input audio to video output representing the audio power spectrum.
18730 Audio amplitude is on Y-axis while frequency is on X-axis.
18732 The filter accepts the following options:
18736 Specify size of video. For the syntax of this option, check the
18737 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18738 Default is @code{1024x512}.
18742 This set how each frequency bin will be represented.
18744 It accepts the following values:
18750 Default is @code{bar}.
18753 Set amplitude scale.
18755 It accepts the following values:
18769 Default is @code{log}.
18772 Set frequency scale.
18774 It accepts the following values:
18783 Reverse logarithmic scale.
18785 Default is @code{lin}.
18790 It accepts the following values:
18806 Default is @code{w2048}
18809 Set windowing function.
18811 It accepts the following values:
18833 Default is @code{hanning}.
18836 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
18837 which means optimal overlap for selected window function will be picked.
18840 Set time averaging. Setting this to 0 will display current maximal peaks.
18841 Default is @code{1}, which means time averaging is disabled.
18844 Specify list of colors separated by space or by '|' which will be used to
18845 draw channel frequencies. Unrecognized or missing colors will be replaced
18849 Set channel display mode.
18851 It accepts the following values:
18856 Default is @code{combined}.
18859 Set minimum amplitude used in @code{log} amplitude scaler.
18863 @anchor{showspectrum}
18864 @section showspectrum
18866 Convert input audio to a video output, representing the audio frequency
18869 The filter accepts the following options:
18873 Specify the video size for the output. For the syntax of this option, check the
18874 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18875 Default value is @code{640x512}.
18878 Specify how the spectrum should slide along the window.
18880 It accepts the following values:
18883 the samples start again on the left when they reach the right
18885 the samples scroll from right to left
18887 frames are only produced when the samples reach the right
18889 the samples scroll from left to right
18892 Default value is @code{replace}.
18895 Specify display mode.
18897 It accepts the following values:
18900 all channels are displayed in the same row
18902 all channels are displayed in separate rows
18905 Default value is @samp{combined}.
18908 Specify display color mode.
18910 It accepts the following values:
18913 each channel is displayed in a separate color
18915 each channel is displayed using the same color scheme
18917 each channel is displayed using the rainbow color scheme
18919 each channel is displayed using the moreland color scheme
18921 each channel is displayed using the nebulae color scheme
18923 each channel is displayed using the fire color scheme
18925 each channel is displayed using the fiery color scheme
18927 each channel is displayed using the fruit color scheme
18929 each channel is displayed using the cool color scheme
18932 Default value is @samp{channel}.
18935 Specify scale used for calculating intensity color values.
18937 It accepts the following values:
18942 square root, default
18953 Default value is @samp{sqrt}.
18956 Set saturation modifier for displayed colors. Negative values provide
18957 alternative color scheme. @code{0} is no saturation at all.
18958 Saturation must be in [-10.0, 10.0] range.
18959 Default value is @code{1}.
18962 Set window function.
18964 It accepts the following values:
18988 Default value is @code{hann}.
18991 Set orientation of time vs frequency axis. Can be @code{vertical} or
18992 @code{horizontal}. Default is @code{vertical}.
18995 Set ratio of overlap window. Default value is @code{0}.
18996 When value is @code{1} overlap is set to recommended size for specific
18997 window function currently used.
19000 Set scale gain for calculating intensity color values.
19001 Default value is @code{1}.
19004 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
19007 Set color rotation, must be in [-1.0, 1.0] range.
19008 Default value is @code{0}.
19011 The usage is very similar to the showwaves filter; see the examples in that
19014 @subsection Examples
19018 Large window with logarithmic color scaling:
19020 showspectrum=s=1280x480:scale=log
19024 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
19026 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19027 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
19031 @section showspectrumpic
19033 Convert input audio to a single video frame, representing the audio frequency
19036 The filter accepts the following options:
19040 Specify the video size for the output. For the syntax of this option, check the
19041 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19042 Default value is @code{4096x2048}.
19045 Specify display mode.
19047 It accepts the following values:
19050 all channels are displayed in the same row
19052 all channels are displayed in separate rows
19054 Default value is @samp{combined}.
19057 Specify display color mode.
19059 It accepts the following values:
19062 each channel is displayed in a separate color
19064 each channel is displayed using the same color scheme
19066 each channel is displayed using the rainbow color scheme
19068 each channel is displayed using the moreland color scheme
19070 each channel is displayed using the nebulae color scheme
19072 each channel is displayed using the fire color scheme
19074 each channel is displayed using the fiery color scheme
19076 each channel is displayed using the fruit color scheme
19078 each channel is displayed using the cool color scheme
19080 Default value is @samp{intensity}.
19083 Specify scale used for calculating intensity color values.
19085 It accepts the following values:
19090 square root, default
19100 Default value is @samp{log}.
19103 Set saturation modifier for displayed colors. Negative values provide
19104 alternative color scheme. @code{0} is no saturation at all.
19105 Saturation must be in [-10.0, 10.0] range.
19106 Default value is @code{1}.
19109 Set window function.
19111 It accepts the following values:
19134 Default value is @code{hann}.
19137 Set orientation of time vs frequency axis. Can be @code{vertical} or
19138 @code{horizontal}. Default is @code{vertical}.
19141 Set scale gain for calculating intensity color values.
19142 Default value is @code{1}.
19145 Draw time and frequency axes and legends. Default is enabled.
19148 Set color rotation, must be in [-1.0, 1.0] range.
19149 Default value is @code{0}.
19152 @subsection Examples
19156 Extract an audio spectrogram of a whole audio track
19157 in a 1024x1024 picture using @command{ffmpeg}:
19159 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19163 @section showvolume
19165 Convert input audio volume to a video output.
19167 The filter accepts the following options:
19174 Set border width, allowed range is [0, 5]. Default is 1.
19177 Set channel width, allowed range is [80, 8192]. Default is 400.
19180 Set channel height, allowed range is [1, 900]. Default is 20.
19183 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19186 Set volume color expression.
19188 The expression can use the following variables:
19192 Current max volume of channel in dB.
19198 Current channel number, starting from 0.
19202 If set, displays channel names. Default is enabled.
19205 If set, displays volume values. Default is enabled.
19208 Set orientation, can be @code{horizontal} or @code{vertical},
19209 default is @code{horizontal}.
19212 Set step size, allowed range s [0, 5]. Default is 0, which means
19218 Convert input audio to a video output, representing the samples waves.
19220 The filter accepts the following options:
19224 Specify the video size for the output. For the syntax of this option, check the
19225 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19226 Default value is @code{600x240}.
19231 Available values are:
19234 Draw a point for each sample.
19237 Draw a vertical line for each sample.
19240 Draw a point for each sample and a line between them.
19243 Draw a centered vertical line for each sample.
19246 Default value is @code{point}.
19249 Set the number of samples which are printed on the same column. A
19250 larger value will decrease the frame rate. Must be a positive
19251 integer. This option can be set only if the value for @var{rate}
19252 is not explicitly specified.
19255 Set the (approximate) output frame rate. This is done by setting the
19256 option @var{n}. Default value is "25".
19258 @item split_channels
19259 Set if channels should be drawn separately or overlap. Default value is 0.
19262 Set colors separated by '|' which are going to be used for drawing of each channel.
19265 Set amplitude scale.
19267 Available values are:
19285 @subsection Examples
19289 Output the input file audio and the corresponding video representation
19292 amovie=a.mp3,asplit[out0],showwaves[out1]
19296 Create a synthetic signal and show it with showwaves, forcing a
19297 frame rate of 30 frames per second:
19299 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
19303 @section showwavespic
19305 Convert input audio to a single video frame, representing the samples waves.
19307 The filter accepts the following options:
19311 Specify the video size for the output. For the syntax of this option, check the
19312 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19313 Default value is @code{600x240}.
19315 @item split_channels
19316 Set if channels should be drawn separately or overlap. Default value is 0.
19319 Set colors separated by '|' which are going to be used for drawing of each channel.
19322 Set amplitude scale.
19324 Available values are:
19342 @subsection Examples
19346 Extract a channel split representation of the wave form of a whole audio track
19347 in a 1024x800 picture using @command{ffmpeg}:
19349 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
19353 @section sidedata, asidedata
19355 Delete frame side data, or select frames based on it.
19357 This filter accepts the following options:
19361 Set mode of operation of the filter.
19363 Can be one of the following:
19367 Select every frame with side data of @code{type}.
19370 Delete side data of @code{type}. If @code{type} is not set, delete all side
19376 Set side data type used with all modes. Must be set for @code{select} mode. For
19377 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
19378 in @file{libavutil/frame.h}. For example, to choose
19379 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
19383 @section spectrumsynth
19385 Sythesize audio from 2 input video spectrums, first input stream represents
19386 magnitude across time and second represents phase across time.
19387 The filter will transform from frequency domain as displayed in videos back
19388 to time domain as presented in audio output.
19390 This filter is primarily created for reversing processed @ref{showspectrum}
19391 filter outputs, but can synthesize sound from other spectrograms too.
19392 But in such case results are going to be poor if the phase data is not
19393 available, because in such cases phase data need to be recreated, usually
19394 its just recreated from random noise.
19395 For best results use gray only output (@code{channel} color mode in
19396 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
19397 @code{lin} scale for phase video. To produce phase, for 2nd video, use
19398 @code{data} option. Inputs videos should generally use @code{fullframe}
19399 slide mode as that saves resources needed for decoding video.
19401 The filter accepts the following options:
19405 Specify sample rate of output audio, the sample rate of audio from which
19406 spectrum was generated may differ.
19409 Set number of channels represented in input video spectrums.
19412 Set scale which was used when generating magnitude input spectrum.
19413 Can be @code{lin} or @code{log}. Default is @code{log}.
19416 Set slide which was used when generating inputs spectrums.
19417 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
19418 Default is @code{fullframe}.
19421 Set window function used for resynthesis.
19424 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19425 which means optimal overlap for selected window function will be picked.
19428 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
19429 Default is @code{vertical}.
19432 @subsection Examples
19436 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
19437 then resynthesize videos back to audio with spectrumsynth:
19439 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
19440 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
19441 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
19445 @section split, asplit
19447 Split input into several identical outputs.
19449 @code{asplit} works with audio input, @code{split} with video.
19451 The filter accepts a single parameter which specifies the number of outputs. If
19452 unspecified, it defaults to 2.
19454 @subsection Examples
19458 Create two separate outputs from the same input:
19460 [in] split [out0][out1]
19464 To create 3 or more outputs, you need to specify the number of
19467 [in] asplit=3 [out0][out1][out2]
19471 Create two separate outputs from the same input, one cropped and
19474 [in] split [splitout1][splitout2];
19475 [splitout1] crop=100:100:0:0 [cropout];
19476 [splitout2] pad=200:200:100:100 [padout];
19480 Create 5 copies of the input audio with @command{ffmpeg}:
19482 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
19488 Receive commands sent through a libzmq client, and forward them to
19489 filters in the filtergraph.
19491 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
19492 must be inserted between two video filters, @code{azmq} between two
19495 To enable these filters you need to install the libzmq library and
19496 headers and configure FFmpeg with @code{--enable-libzmq}.
19498 For more information about libzmq see:
19499 @url{http://www.zeromq.org/}
19501 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
19502 receives messages sent through a network interface defined by the
19503 @option{bind_address} option.
19505 The received message must be in the form:
19507 @var{TARGET} @var{COMMAND} [@var{ARG}]
19510 @var{TARGET} specifies the target of the command, usually the name of
19511 the filter class or a specific filter instance name.
19513 @var{COMMAND} specifies the name of the command for the target filter.
19515 @var{ARG} is optional and specifies the optional argument list for the
19516 given @var{COMMAND}.
19518 Upon reception, the message is processed and the corresponding command
19519 is injected into the filtergraph. Depending on the result, the filter
19520 will send a reply to the client, adopting the format:
19522 @var{ERROR_CODE} @var{ERROR_REASON}
19526 @var{MESSAGE} is optional.
19528 @subsection Examples
19530 Look at @file{tools/zmqsend} for an example of a zmq client which can
19531 be used to send commands processed by these filters.
19533 Consider the following filtergraph generated by @command{ffplay}
19535 ffplay -dumpgraph 1 -f lavfi "
19536 color=s=100x100:c=red [l];
19537 color=s=100x100:c=blue [r];
19538 nullsrc=s=200x100, zmq [bg];
19539 [bg][l] overlay [bg+l];
19540 [bg+l][r] overlay=x=100 "
19543 To change the color of the left side of the video, the following
19544 command can be used:
19546 echo Parsed_color_0 c yellow | tools/zmqsend
19549 To change the right side:
19551 echo Parsed_color_1 c pink | tools/zmqsend
19554 @c man end MULTIMEDIA FILTERS
19556 @chapter Multimedia Sources
19557 @c man begin MULTIMEDIA SOURCES
19559 Below is a description of the currently available multimedia sources.
19563 This is the same as @ref{movie} source, except it selects an audio
19569 Read audio and/or video stream(s) from a movie container.
19571 It accepts the following parameters:
19575 The name of the resource to read (not necessarily a file; it can also be a
19576 device or a stream accessed through some protocol).
19578 @item format_name, f
19579 Specifies the format assumed for the movie to read, and can be either
19580 the name of a container or an input device. If not specified, the
19581 format is guessed from @var{movie_name} or by probing.
19583 @item seek_point, sp
19584 Specifies the seek point in seconds. The frames will be output
19585 starting from this seek point. The parameter is evaluated with
19586 @code{av_strtod}, so the numerical value may be suffixed by an IS
19587 postfix. The default value is "0".
19590 Specifies the streams to read. Several streams can be specified,
19591 separated by "+". The source will then have as many outputs, in the
19592 same order. The syntax is explained in the ``Stream specifiers''
19593 section in the ffmpeg manual. Two special names, "dv" and "da" specify
19594 respectively the default (best suited) video and audio stream. Default
19595 is "dv", or "da" if the filter is called as "amovie".
19597 @item stream_index, si
19598 Specifies the index of the video stream to read. If the value is -1,
19599 the most suitable video stream will be automatically selected. The default
19600 value is "-1". Deprecated. If the filter is called "amovie", it will select
19601 audio instead of video.
19604 Specifies how many times to read the stream in sequence.
19605 If the value is 0, the stream will be looped infinitely.
19606 Default value is "1".
19608 Note that when the movie is looped the source timestamps are not
19609 changed, so it will generate non monotonically increasing timestamps.
19611 @item discontinuity
19612 Specifies the time difference between frames above which the point is
19613 considered a timestamp discontinuity which is removed by adjusting the later
19617 It allows overlaying a second video on top of the main input of
19618 a filtergraph, as shown in this graph:
19620 input -----------> deltapts0 --> overlay --> output
19623 movie --> scale--> deltapts1 -------+
19625 @subsection Examples
19629 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
19630 on top of the input labelled "in":
19632 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
19633 [in] setpts=PTS-STARTPTS [main];
19634 [main][over] overlay=16:16 [out]
19638 Read from a video4linux2 device, and overlay it on top of the input
19641 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
19642 [in] setpts=PTS-STARTPTS [main];
19643 [main][over] overlay=16:16 [out]
19647 Read the first video stream and the audio stream with id 0x81 from
19648 dvd.vob; the video is connected to the pad named "video" and the audio is
19649 connected to the pad named "audio":
19651 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
19655 @subsection Commands
19657 Both movie and amovie support the following commands:
19660 Perform seek using "av_seek_frame".
19661 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
19664 @var{stream_index}: If stream_index is -1, a default
19665 stream is selected, and @var{timestamp} is automatically converted
19666 from AV_TIME_BASE units to the stream specific time_base.
19668 @var{timestamp}: Timestamp in AVStream.time_base units
19669 or, if no stream is specified, in AV_TIME_BASE units.
19671 @var{flags}: Flags which select direction and seeking mode.
19675 Get movie duration in AV_TIME_BASE units.
19679 @c man end MULTIMEDIA SOURCES