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 Pass the video source unchanged to the output.
10875 Optical Character Recognition
10877 This filter uses Tesseract for optical character recognition.
10879 It accepts the following options:
10883 Set datapath to tesseract data. Default is to use whatever was
10884 set at installation.
10887 Set language, default is "eng".
10890 Set character whitelist.
10893 Set character blacklist.
10896 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
10900 Apply a video transform using libopencv.
10902 To enable this filter, install the libopencv library and headers and
10903 configure FFmpeg with @code{--enable-libopencv}.
10905 It accepts the following parameters:
10910 The name of the libopencv filter to apply.
10912 @item filter_params
10913 The parameters to pass to the libopencv filter. If not specified, the default
10914 values are assumed.
10918 Refer to the official libopencv documentation for more precise
10920 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
10922 Several libopencv filters are supported; see the following subsections.
10927 Dilate an image by using a specific structuring element.
10928 It corresponds to the libopencv function @code{cvDilate}.
10930 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10932 @var{struct_el} represents a structuring element, and has the syntax:
10933 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10935 @var{cols} and @var{rows} represent the number of columns and rows of
10936 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10937 point, and @var{shape} the shape for the structuring element. @var{shape}
10938 must be "rect", "cross", "ellipse", or "custom".
10940 If the value for @var{shape} is "custom", it must be followed by a
10941 string of the form "=@var{filename}". The file with name
10942 @var{filename} is assumed to represent a binary image, with each
10943 printable character corresponding to a bright pixel. When a custom
10944 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10945 or columns and rows of the read file are assumed instead.
10947 The default value for @var{struct_el} is "3x3+0x0/rect".
10949 @var{nb_iterations} specifies the number of times the transform is
10950 applied to the image, and defaults to 1.
10954 # Use the default values
10957 # Dilate using a structuring element with a 5x5 cross, iterating two times
10958 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10960 # Read the shape from the file diamond.shape, iterating two times.
10961 # The file diamond.shape may contain a pattern of characters like this
10967 # The specified columns and rows are ignored
10968 # but the anchor point coordinates are not
10969 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10974 Erode an image by using a specific structuring element.
10975 It corresponds to the libopencv function @code{cvErode}.
10977 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10978 with the same syntax and semantics as the @ref{dilate} filter.
10982 Smooth the input video.
10984 The filter takes the following parameters:
10985 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10987 @var{type} is the type of smooth filter to apply, and must be one of
10988 the following values: "blur", "blur_no_scale", "median", "gaussian",
10989 or "bilateral". The default value is "gaussian".
10991 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10992 depend on the smooth type. @var{param1} and
10993 @var{param2} accept integer positive values or 0. @var{param3} and
10994 @var{param4} accept floating point values.
10996 The default value for @var{param1} is 3. The default value for the
10997 other parameters is 0.
10999 These parameters correspond to the parameters assigned to the
11000 libopencv function @code{cvSmooth}.
11002 @section oscilloscope
11004 2D Video Oscilloscope.
11006 Useful to measure spatial impulse, step responses, chroma delays, etc.
11008 It accepts the following parameters:
11012 Set scope center x position.
11015 Set scope center y position.
11018 Set scope size, relative to frame diagonal.
11021 Set scope tilt/rotation.
11027 Set trace center x position.
11030 Set trace center y position.
11033 Set trace width, relative to width of frame.
11036 Set trace height, relative to height of frame.
11039 Set which components to trace. By default it traces first three components.
11042 Draw trace grid. By default is enabled.
11045 Draw some statistics. By default is enabled.
11048 Draw scope. By default is enabled.
11051 @subsection Examples
11055 Inspect full first row of video frame.
11057 oscilloscope=x=0.5:y=0:s=1
11061 Inspect full last row of video frame.
11063 oscilloscope=x=0.5:y=1:s=1
11067 Inspect full 5th line of video frame of height 1080.
11069 oscilloscope=x=0.5:y=5/1080:s=1
11073 Inspect full last column of video frame.
11075 oscilloscope=x=1:y=0.5:s=1:t=1
11083 Overlay one video on top of another.
11085 It takes two inputs and has one output. The first input is the "main"
11086 video on which the second input is overlaid.
11088 It accepts the following parameters:
11090 A description of the accepted options follows.
11095 Set the expression for the x and y coordinates of the overlaid video
11096 on the main video. Default value is "0" for both expressions. In case
11097 the expression is invalid, it is set to a huge value (meaning that the
11098 overlay will not be displayed within the output visible area).
11101 See @ref{framesync}.
11104 Set when the expressions for @option{x}, and @option{y} are evaluated.
11106 It accepts the following values:
11109 only evaluate expressions once during the filter initialization or
11110 when a command is processed
11113 evaluate expressions for each incoming frame
11116 Default value is @samp{frame}.
11119 See @ref{framesync}.
11122 Set the format for the output video.
11124 It accepts the following values:
11127 force YUV420 output
11130 force YUV422 output
11133 force YUV444 output
11136 force packed RGB output
11139 force planar RGB output
11142 automatically pick format
11145 Default value is @samp{yuv420}.
11148 See @ref{framesync}.
11151 The @option{x}, and @option{y} expressions can contain the following
11157 The main input width and height.
11161 The overlay input width and height.
11165 The computed values for @var{x} and @var{y}. They are evaluated for
11170 horizontal and vertical chroma subsample values of the output
11171 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11175 the number of input frame, starting from 0
11178 the position in the file of the input frame, NAN if unknown
11181 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11185 This filter also supports the @ref{framesync} options.
11187 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11188 when evaluation is done @emph{per frame}, and will evaluate to NAN
11189 when @option{eval} is set to @samp{init}.
11191 Be aware that frames are taken from each input video in timestamp
11192 order, hence, if their initial timestamps differ, it is a good idea
11193 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11194 have them begin in the same zero timestamp, as the example for
11195 the @var{movie} filter does.
11197 You can chain together more overlays but you should test the
11198 efficiency of such approach.
11200 @subsection Commands
11202 This filter supports the following commands:
11206 Modify the x and y of the overlay input.
11207 The command accepts the same syntax of the corresponding option.
11209 If the specified expression is not valid, it is kept at its current
11213 @subsection Examples
11217 Draw the overlay at 10 pixels from the bottom right corner of the main
11220 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11223 Using named options the example above becomes:
11225 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11229 Insert a transparent PNG logo in the bottom left corner of the input,
11230 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11232 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11236 Insert 2 different transparent PNG logos (second logo on bottom
11237 right corner) using the @command{ffmpeg} tool:
11239 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
11243 Add a transparent color layer on top of the main video; @code{WxH}
11244 must specify the size of the main input to the overlay filter:
11246 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11250 Play an original video and a filtered version (here with the deshake
11251 filter) side by side using the @command{ffplay} tool:
11253 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11256 The above command is the same as:
11258 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11262 Make a sliding overlay appearing from the left to the right top part of the
11263 screen starting since time 2:
11265 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11269 Compose output by putting two input videos side to side:
11271 ffmpeg -i left.avi -i right.avi -filter_complex "
11272 nullsrc=size=200x100 [background];
11273 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11274 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11275 [background][left] overlay=shortest=1 [background+left];
11276 [background+left][right] overlay=shortest=1:x=100 [left+right]
11281 Mask 10-20 seconds of a video by applying the delogo filter to a section
11283 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11284 -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]'
11289 Chain several overlays in cascade:
11291 nullsrc=s=200x200 [bg];
11292 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11293 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11294 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11295 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11296 [in3] null, [mid2] overlay=100:100 [out0]
11303 Apply Overcomplete Wavelet denoiser.
11305 The filter accepts the following options:
11311 Larger depth values will denoise lower frequency components more, but
11312 slow down filtering.
11314 Must be an int in the range 8-16, default is @code{8}.
11316 @item luma_strength, ls
11319 Must be a double value in the range 0-1000, default is @code{1.0}.
11321 @item chroma_strength, cs
11322 Set chroma strength.
11324 Must be a double value in the range 0-1000, default is @code{1.0}.
11330 Add paddings to the input image, and place the original input at the
11331 provided @var{x}, @var{y} coordinates.
11333 It accepts the following parameters:
11338 Specify an expression for the size of the output image with the
11339 paddings added. If the value for @var{width} or @var{height} is 0, the
11340 corresponding input size is used for the output.
11342 The @var{width} expression can reference the value set by the
11343 @var{height} expression, and vice versa.
11345 The default value of @var{width} and @var{height} is 0.
11349 Specify the offsets to place the input image at within the padded area,
11350 with respect to the top/left border of the output image.
11352 The @var{x} expression can reference the value set by the @var{y}
11353 expression, and vice versa.
11355 The default value of @var{x} and @var{y} is 0.
11357 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11358 so the input image is centered on the padded area.
11361 Specify the color of the padded area. For the syntax of this option,
11362 check the "Color" section in the ffmpeg-utils manual.
11364 The default value of @var{color} is "black".
11367 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11369 It accepts the following values:
11373 Only evaluate expressions once during the filter initialization or when
11374 a command is processed.
11377 Evaluate expressions for each incoming frame.
11381 Default value is @samp{init}.
11384 Pad to aspect instead to a resolution.
11388 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11389 options are expressions containing the following constants:
11394 The input video width and height.
11398 These are the same as @var{in_w} and @var{in_h}.
11402 The output width and height (the size of the padded area), as
11403 specified by the @var{width} and @var{height} expressions.
11407 These are the same as @var{out_w} and @var{out_h}.
11411 The x and y offsets as specified by the @var{x} and @var{y}
11412 expressions, or NAN if not yet specified.
11415 same as @var{iw} / @var{ih}
11418 input sample aspect ratio
11421 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11425 The horizontal and vertical chroma subsample values. For example for the
11426 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11429 @subsection Examples
11433 Add paddings with the color "violet" to the input video. The output video
11434 size is 640x480, and the top-left corner of the input video is placed at
11437 pad=640:480:0:40:violet
11440 The example above is equivalent to the following command:
11442 pad=width=640:height=480:x=0:y=40:color=violet
11446 Pad the input to get an output with dimensions increased by 3/2,
11447 and put the input video at the center of the padded area:
11449 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
11453 Pad the input to get a squared output with size equal to the maximum
11454 value between the input width and height, and put the input video at
11455 the center of the padded area:
11457 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
11461 Pad the input to get a final w/h ratio of 16:9:
11463 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
11467 In case of anamorphic video, in order to set the output display aspect
11468 correctly, it is necessary to use @var{sar} in the expression,
11469 according to the relation:
11471 (ih * X / ih) * sar = output_dar
11472 X = output_dar / sar
11475 Thus the previous example needs to be modified to:
11477 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
11481 Double the output size and put the input video in the bottom-right
11482 corner of the output padded area:
11484 pad="2*iw:2*ih:ow-iw:oh-ih"
11488 @anchor{palettegen}
11489 @section palettegen
11491 Generate one palette for a whole video stream.
11493 It accepts the following options:
11497 Set the maximum number of colors to quantize in the palette.
11498 Note: the palette will still contain 256 colors; the unused palette entries
11501 @item reserve_transparent
11502 Create a palette of 255 colors maximum and reserve the last one for
11503 transparency. Reserving the transparency color is useful for GIF optimization.
11504 If not set, the maximum of colors in the palette will be 256. You probably want
11505 to disable this option for a standalone image.
11508 @item transparency_color
11509 Set the color that will be used as background for transparency.
11512 Set statistics mode.
11514 It accepts the following values:
11517 Compute full frame histograms.
11519 Compute histograms only for the part that differs from previous frame. This
11520 might be relevant to give more importance to the moving part of your input if
11521 the background is static.
11523 Compute new histogram for each frame.
11526 Default value is @var{full}.
11529 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
11530 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
11531 color quantization of the palette. This information is also visible at
11532 @var{info} logging level.
11534 @subsection Examples
11538 Generate a representative palette of a given video using @command{ffmpeg}:
11540 ffmpeg -i input.mkv -vf palettegen palette.png
11544 @section paletteuse
11546 Use a palette to downsample an input video stream.
11548 The filter takes two inputs: one video stream and a palette. The palette must
11549 be a 256 pixels image.
11551 It accepts the following options:
11555 Select dithering mode. Available algorithms are:
11558 Ordered 8x8 bayer dithering (deterministic)
11560 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
11561 Note: this dithering is sometimes considered "wrong" and is included as a
11563 @item floyd_steinberg
11564 Floyd and Steingberg dithering (error diffusion)
11566 Frankie Sierra dithering v2 (error diffusion)
11568 Frankie Sierra dithering v2 "Lite" (error diffusion)
11571 Default is @var{sierra2_4a}.
11574 When @var{bayer} dithering is selected, this option defines the scale of the
11575 pattern (how much the crosshatch pattern is visible). A low value means more
11576 visible pattern for less banding, and higher value means less visible pattern
11577 at the cost of more banding.
11579 The option must be an integer value in the range [0,5]. Default is @var{2}.
11582 If set, define the zone to process
11586 Only the changing rectangle will be reprocessed. This is similar to GIF
11587 cropping/offsetting compression mechanism. This option can be useful for speed
11588 if only a part of the image is changing, and has use cases such as limiting the
11589 scope of the error diffusal @option{dither} to the rectangle that bounds the
11590 moving scene (it leads to more deterministic output if the scene doesn't change
11591 much, and as a result less moving noise and better GIF compression).
11594 Default is @var{none}.
11597 Take new palette for each output frame.
11599 @item alpha_threshold
11600 Sets the alpha threshold for transparency. Alpha values above this threshold
11601 will be treated as completely opaque, and values below this threshold will be
11602 treated as completely transparent.
11604 The option must be an integer value in the range [0,255]. Default is @var{128}.
11607 @subsection Examples
11611 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
11612 using @command{ffmpeg}:
11614 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
11618 @section perspective
11620 Correct perspective of video not recorded perpendicular to the screen.
11622 A description of the accepted parameters follows.
11633 Set coordinates expression for top left, top right, bottom left and bottom right corners.
11634 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
11635 If the @code{sense} option is set to @code{source}, then the specified points will be sent
11636 to the corners of the destination. If the @code{sense} option is set to @code{destination},
11637 then the corners of the source will be sent to the specified coordinates.
11639 The expressions can use the following variables:
11644 the width and height of video frame.
11648 Output frame count.
11651 @item interpolation
11652 Set interpolation for perspective correction.
11654 It accepts the following values:
11660 Default value is @samp{linear}.
11663 Set interpretation of coordinate options.
11665 It accepts the following values:
11669 Send point in the source specified by the given coordinates to
11670 the corners of the destination.
11672 @item 1, destination
11674 Send the corners of the source to the point in the destination specified
11675 by the given coordinates.
11677 Default value is @samp{source}.
11681 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
11683 It accepts the following values:
11686 only evaluate expressions once during the filter initialization or
11687 when a command is processed
11690 evaluate expressions for each incoming frame
11693 Default value is @samp{init}.
11698 Delay interlaced video by one field time so that the field order changes.
11700 The intended use is to fix PAL movies that have been captured with the
11701 opposite field order to the film-to-video transfer.
11703 A description of the accepted parameters follows.
11709 It accepts the following values:
11712 Capture field order top-first, transfer bottom-first.
11713 Filter will delay the bottom field.
11716 Capture field order bottom-first, transfer top-first.
11717 Filter will delay the top field.
11720 Capture and transfer with the same field order. This mode only exists
11721 for the documentation of the other options to refer to, but if you
11722 actually select it, the filter will faithfully do nothing.
11725 Capture field order determined automatically by field flags, transfer
11727 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
11728 basis using field flags. If no field information is available,
11729 then this works just like @samp{u}.
11732 Capture unknown or varying, transfer opposite.
11733 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
11734 analyzing the images and selecting the alternative that produces best
11735 match between the fields.
11738 Capture top-first, transfer unknown or varying.
11739 Filter selects among @samp{t} and @samp{p} using image analysis.
11742 Capture bottom-first, transfer unknown or varying.
11743 Filter selects among @samp{b} and @samp{p} using image analysis.
11746 Capture determined by field flags, transfer unknown or varying.
11747 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
11748 image analysis. If no field information is available, then this works just
11749 like @samp{U}. This is the default mode.
11752 Both capture and transfer unknown or varying.
11753 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
11757 @section pixdesctest
11759 Pixel format descriptor test filter, mainly useful for internal
11760 testing. The output video should be equal to the input video.
11764 format=monow, pixdesctest
11767 can be used to test the monowhite pixel format descriptor definition.
11771 Display sample values of color channels. Mainly useful for checking color
11772 and levels. Minimum supported resolution is 640x480.
11774 The filters accept the following options:
11778 Set scope X position, relative offset on X axis.
11781 Set scope Y position, relative offset on Y axis.
11790 Set window opacity. This window also holds statistics about pixel area.
11793 Set window X position, relative offset on X axis.
11796 Set window Y position, relative offset on Y axis.
11801 Enable the specified chain of postprocessing subfilters using libpostproc. This
11802 library should be automatically selected with a GPL build (@code{--enable-gpl}).
11803 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
11804 Each subfilter and some options have a short and a long name that can be used
11805 interchangeably, i.e. dr/dering are the same.
11807 The filters accept the following options:
11811 Set postprocessing subfilters string.
11814 All subfilters share common options to determine their scope:
11818 Honor the quality commands for this subfilter.
11821 Do chrominance filtering, too (default).
11824 Do luminance filtering only (no chrominance).
11827 Do chrominance filtering only (no luminance).
11830 These options can be appended after the subfilter name, separated by a '|'.
11832 Available subfilters are:
11835 @item hb/hdeblock[|difference[|flatness]]
11836 Horizontal deblocking filter
11839 Difference factor where higher values mean more deblocking (default: @code{32}).
11841 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11844 @item vb/vdeblock[|difference[|flatness]]
11845 Vertical deblocking filter
11848 Difference factor where higher values mean more deblocking (default: @code{32}).
11850 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11853 @item ha/hadeblock[|difference[|flatness]]
11854 Accurate horizontal deblocking filter
11857 Difference factor where higher values mean more deblocking (default: @code{32}).
11859 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11862 @item va/vadeblock[|difference[|flatness]]
11863 Accurate vertical deblocking filter
11866 Difference factor where higher values mean more deblocking (default: @code{32}).
11868 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11872 The horizontal and vertical deblocking filters share the difference and
11873 flatness values so you cannot set different horizontal and vertical
11877 @item h1/x1hdeblock
11878 Experimental horizontal deblocking filter
11880 @item v1/x1vdeblock
11881 Experimental vertical deblocking filter
11886 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
11889 larger -> stronger filtering
11891 larger -> stronger filtering
11893 larger -> stronger filtering
11896 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
11899 Stretch luminance to @code{0-255}.
11902 @item lb/linblenddeint
11903 Linear blend deinterlacing filter that deinterlaces the given block by
11904 filtering all lines with a @code{(1 2 1)} filter.
11906 @item li/linipoldeint
11907 Linear interpolating deinterlacing filter that deinterlaces the given block by
11908 linearly interpolating every second line.
11910 @item ci/cubicipoldeint
11911 Cubic interpolating deinterlacing filter deinterlaces the given block by
11912 cubically interpolating every second line.
11914 @item md/mediandeint
11915 Median deinterlacing filter that deinterlaces the given block by applying a
11916 median filter to every second line.
11918 @item fd/ffmpegdeint
11919 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
11920 second line with a @code{(-1 4 2 4 -1)} filter.
11923 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
11924 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
11926 @item fq/forceQuant[|quantizer]
11927 Overrides the quantizer table from the input with the constant quantizer you
11935 Default pp filter combination (@code{hb|a,vb|a,dr|a})
11938 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
11941 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
11944 @subsection Examples
11948 Apply horizontal and vertical deblocking, deringing and automatic
11949 brightness/contrast:
11955 Apply default filters without brightness/contrast correction:
11961 Apply default filters and temporal denoiser:
11963 pp=default/tmpnoise|1|2|3
11967 Apply deblocking on luminance only, and switch vertical deblocking on or off
11968 automatically depending on available CPU time:
11975 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
11976 similar to spp = 6 with 7 point DCT, where only the center sample is
11979 The filter accepts the following options:
11983 Force a constant quantization parameter. It accepts an integer in range
11984 0 to 63. If not set, the filter will use the QP from the video stream
11988 Set thresholding mode. Available modes are:
11992 Set hard thresholding.
11994 Set soft thresholding (better de-ringing effect, but likely blurrier).
11996 Set medium thresholding (good results, default).
12000 @section premultiply
12001 Apply alpha premultiply effect to input video stream using first plane
12002 of second stream as alpha.
12004 Both streams must have same dimensions and same pixel format.
12006 The filter accepts the following option:
12010 Set which planes will be processed, unprocessed planes will be copied.
12011 By default value 0xf, all planes will be processed.
12014 Do not require 2nd input for processing, instead use alpha plane from input stream.
12018 Apply prewitt operator to input video stream.
12020 The filter accepts the following option:
12024 Set which planes will be processed, unprocessed planes will be copied.
12025 By default value 0xf, all planes will be processed.
12028 Set value which will be multiplied with filtered result.
12031 Set value which will be added to filtered result.
12034 @section pseudocolor
12036 Alter frame colors in video with pseudocolors.
12038 This filter accept the following options:
12042 set pixel first component expression
12045 set pixel second component expression
12048 set pixel third component expression
12051 set pixel fourth component expression, corresponds to the alpha component
12054 set component to use as base for altering colors
12057 Each of them specifies the expression to use for computing the lookup table for
12058 the corresponding pixel component values.
12060 The expressions can contain the following constants and functions:
12065 The input width and height.
12068 The input value for the pixel component.
12070 @item ymin, umin, vmin, amin
12071 The minimum allowed component value.
12073 @item ymax, umax, vmax, amax
12074 The maximum allowed component value.
12077 All expressions default to "val".
12079 @subsection Examples
12083 Change too high luma values to gradient:
12085 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'"
12091 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12092 Ratio) between two input videos.
12094 This filter takes in input two input videos, the first input is
12095 considered the "main" source and is passed unchanged to the
12096 output. The second input is used as a "reference" video for computing
12099 Both video inputs must have the same resolution and pixel format for
12100 this filter to work correctly. Also it assumes that both inputs
12101 have the same number of frames, which are compared one by one.
12103 The obtained average PSNR is printed through the logging system.
12105 The filter stores the accumulated MSE (mean squared error) of each
12106 frame, and at the end of the processing it is averaged across all frames
12107 equally, and the following formula is applied to obtain the PSNR:
12110 PSNR = 10*log10(MAX^2/MSE)
12113 Where MAX is the average of the maximum values of each component of the
12116 The description of the accepted parameters follows.
12119 @item stats_file, f
12120 If specified the filter will use the named file to save the PSNR of
12121 each individual frame. When filename equals "-" the data is sent to
12124 @item stats_version
12125 Specifies which version of the stats file format to use. Details of
12126 each format are written below.
12127 Default value is 1.
12129 @item stats_add_max
12130 Determines whether the max value is output to the stats log.
12131 Default value is 0.
12132 Requires stats_version >= 2. If this is set and stats_version < 2,
12133 the filter will return an error.
12136 This filter also supports the @ref{framesync} options.
12138 The file printed if @var{stats_file} is selected, contains a sequence of
12139 key/value pairs of the form @var{key}:@var{value} for each compared
12142 If a @var{stats_version} greater than 1 is specified, a header line precedes
12143 the list of per-frame-pair stats, with key value pairs following the frame
12144 format with the following parameters:
12147 @item psnr_log_version
12148 The version of the log file format. Will match @var{stats_version}.
12151 A comma separated list of the per-frame-pair parameters included in
12155 A description of each shown per-frame-pair parameter follows:
12159 sequential number of the input frame, starting from 1
12162 Mean Square Error pixel-by-pixel average difference of the compared
12163 frames, averaged over all the image components.
12165 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
12166 Mean Square Error pixel-by-pixel average difference of the compared
12167 frames for the component specified by the suffix.
12169 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12170 Peak Signal to Noise ratio of the compared frames for the component
12171 specified by the suffix.
12173 @item max_avg, max_y, max_u, max_v
12174 Maximum allowed value for each channel, and average over all
12180 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12181 [main][ref] psnr="stats_file=stats.log" [out]
12184 On this example the input file being processed is compared with the
12185 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12186 is stored in @file{stats.log}.
12191 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12192 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12195 The pullup filter is designed to take advantage of future context in making
12196 its decisions. This filter is stateless in the sense that it does not lock
12197 onto a pattern to follow, but it instead looks forward to the following
12198 fields in order to identify matches and rebuild progressive frames.
12200 To produce content with an even framerate, insert the fps filter after
12201 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12202 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12204 The filter accepts the following options:
12211 These options set the amount of "junk" to ignore at the left, right, top, and
12212 bottom of the image, respectively. Left and right are in units of 8 pixels,
12213 while top and bottom are in units of 2 lines.
12214 The default is 8 pixels on each side.
12217 Set the strict breaks. Setting this option to 1 will reduce the chances of
12218 filter generating an occasional mismatched frame, but it may also cause an
12219 excessive number of frames to be dropped during high motion sequences.
12220 Conversely, setting it to -1 will make filter match fields more easily.
12221 This may help processing of video where there is slight blurring between
12222 the fields, but may also cause there to be interlaced frames in the output.
12223 Default value is @code{0}.
12226 Set the metric plane to use. It accepts the following values:
12232 Use chroma blue plane.
12235 Use chroma red plane.
12238 This option may be set to use chroma plane instead of the default luma plane
12239 for doing filter's computations. This may improve accuracy on very clean
12240 source material, but more likely will decrease accuracy, especially if there
12241 is chroma noise (rainbow effect) or any grayscale video.
12242 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12243 load and make pullup usable in realtime on slow machines.
12246 For best results (without duplicated frames in the output file) it is
12247 necessary to change the output frame rate. For example, to inverse
12248 telecine NTSC input:
12250 ffmpeg -i input -vf pullup -r 24000/1001 ...
12255 Change video quantization parameters (QP).
12257 The filter accepts the following option:
12261 Set expression for quantization parameter.
12264 The expression is evaluated through the eval API and can contain, among others,
12265 the following constants:
12269 1 if index is not 129, 0 otherwise.
12272 Sequential index starting from -129 to 128.
12275 @subsection Examples
12279 Some equation like:
12287 Flush video frames from internal cache of frames into a random order.
12288 No frame is discarded.
12289 Inspired by @ref{frei0r} nervous filter.
12293 Set size in number of frames of internal cache, in range from @code{2} to
12294 @code{512}. Default is @code{30}.
12297 Set seed for random number generator, must be an integer included between
12298 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12299 less than @code{0}, the filter will try to use a good random seed on a
12303 @section readeia608
12305 Read closed captioning (EIA-608) information from the top lines of a video frame.
12307 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
12308 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
12309 with EIA-608 data (starting from 0). A description of each metadata value follows:
12312 @item lavfi.readeia608.X.cc
12313 The two bytes stored as EIA-608 data (printed in hexadecimal).
12315 @item lavfi.readeia608.X.line
12316 The number of the line on which the EIA-608 data was identified and read.
12319 This filter accepts the following options:
12323 Set the line to start scanning for EIA-608 data. Default is @code{0}.
12326 Set the line to end scanning for EIA-608 data. Default is @code{29}.
12329 Set minimal acceptable amplitude change for sync codes detection.
12330 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
12333 Set the ratio of width reserved for sync code detection.
12334 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
12337 Set the max peaks height difference for sync code detection.
12338 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12341 Set max peaks period difference for sync code detection.
12342 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12345 Set the first two max start code bits differences.
12346 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
12349 Set the minimum ratio of bits height compared to 3rd start code bit.
12350 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
12353 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
12356 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
12359 Enable checking the parity bit. In the event of a parity error, the filter will output
12360 @code{0x00} for that character. Default is false.
12363 @subsection Examples
12367 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
12369 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
12375 Read vertical interval timecode (VITC) information from the top lines of a
12378 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
12379 timecode value, if a valid timecode has been detected. Further metadata key
12380 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
12381 timecode data has been found or not.
12383 This filter accepts the following options:
12387 Set the maximum number of lines to scan for VITC data. If the value is set to
12388 @code{-1} the full video frame is scanned. Default is @code{45}.
12391 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
12392 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
12395 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
12396 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
12399 @subsection Examples
12403 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
12404 draw @code{--:--:--:--} as a placeholder:
12406 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
12412 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
12414 Destination pixel at position (X, Y) will be picked from source (x, y) position
12415 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
12416 value for pixel will be used for destination pixel.
12418 Xmap and Ymap input video streams must be of same dimensions. Output video stream
12419 will have Xmap/Ymap video stream dimensions.
12420 Xmap and Ymap input video streams are 16bit depth, single channel.
12422 @section removegrain
12424 The removegrain filter is a spatial denoiser for progressive video.
12428 Set mode for the first plane.
12431 Set mode for the second plane.
12434 Set mode for the third plane.
12437 Set mode for the fourth plane.
12440 Range of mode is from 0 to 24. Description of each mode follows:
12444 Leave input plane unchanged. Default.
12447 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
12450 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
12453 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
12456 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
12457 This is equivalent to a median filter.
12460 Line-sensitive clipping giving the minimal change.
12463 Line-sensitive clipping, intermediate.
12466 Line-sensitive clipping, intermediate.
12469 Line-sensitive clipping, intermediate.
12472 Line-sensitive clipping on a line where the neighbours pixels are the closest.
12475 Replaces the target pixel with the closest neighbour.
12478 [1 2 1] horizontal and vertical kernel blur.
12484 Bob mode, interpolates top field from the line where the neighbours
12485 pixels are the closest.
12488 Bob mode, interpolates bottom field from the line where the neighbours
12489 pixels are the closest.
12492 Bob mode, interpolates top field. Same as 13 but with a more complicated
12493 interpolation formula.
12496 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
12497 interpolation formula.
12500 Clips the pixel with the minimum and maximum of respectively the maximum and
12501 minimum of each pair of opposite neighbour pixels.
12504 Line-sensitive clipping using opposite neighbours whose greatest distance from
12505 the current pixel is minimal.
12508 Replaces the pixel with the average of its 8 neighbours.
12511 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
12514 Clips pixels using the averages of opposite neighbour.
12517 Same as mode 21 but simpler and faster.
12520 Small edge and halo removal, but reputed useless.
12526 @section removelogo
12528 Suppress a TV station logo, using an image file to determine which
12529 pixels comprise the logo. It works by filling in the pixels that
12530 comprise the logo with neighboring pixels.
12532 The filter accepts the following options:
12536 Set the filter bitmap file, which can be any image format supported by
12537 libavformat. The width and height of the image file must match those of the
12538 video stream being processed.
12541 Pixels in the provided bitmap image with a value of zero are not
12542 considered part of the logo, non-zero pixels are considered part of
12543 the logo. If you use white (255) for the logo and black (0) for the
12544 rest, you will be safe. For making the filter bitmap, it is
12545 recommended to take a screen capture of a black frame with the logo
12546 visible, and then using a threshold filter followed by the erode
12547 filter once or twice.
12549 If needed, little splotches can be fixed manually. Remember that if
12550 logo pixels are not covered, the filter quality will be much
12551 reduced. Marking too many pixels as part of the logo does not hurt as
12552 much, but it will increase the amount of blurring needed to cover over
12553 the image and will destroy more information than necessary, and extra
12554 pixels will slow things down on a large logo.
12556 @section repeatfields
12558 This filter uses the repeat_field flag from the Video ES headers and hard repeats
12559 fields based on its value.
12563 Reverse a video clip.
12565 Warning: This filter requires memory to buffer the entire clip, so trimming
12568 @subsection Examples
12572 Take the first 5 seconds of a clip, and reverse it.
12579 Apply roberts cross operator to input video stream.
12581 The filter accepts the following option:
12585 Set which planes will be processed, unprocessed planes will be copied.
12586 By default value 0xf, all planes will be processed.
12589 Set value which will be multiplied with filtered result.
12592 Set value which will be added to filtered result.
12597 Rotate video by an arbitrary angle expressed in radians.
12599 The filter accepts the following options:
12601 A description of the optional parameters follows.
12604 Set an expression for the angle by which to rotate the input video
12605 clockwise, expressed as a number of radians. A negative value will
12606 result in a counter-clockwise rotation. By default it is set to "0".
12608 This expression is evaluated for each frame.
12611 Set the output width expression, default value is "iw".
12612 This expression is evaluated just once during configuration.
12615 Set the output height expression, default value is "ih".
12616 This expression is evaluated just once during configuration.
12619 Enable bilinear interpolation if set to 1, a value of 0 disables
12620 it. Default value is 1.
12623 Set the color used to fill the output area not covered by the rotated
12624 image. For the general syntax of this option, check the "Color" section in the
12625 ffmpeg-utils manual. If the special value "none" is selected then no
12626 background is printed (useful for example if the background is never shown).
12628 Default value is "black".
12631 The expressions for the angle and the output size can contain the
12632 following constants and functions:
12636 sequential number of the input frame, starting from 0. It is always NAN
12637 before the first frame is filtered.
12640 time in seconds of the input frame, it is set to 0 when the filter is
12641 configured. It is always NAN before the first frame is filtered.
12645 horizontal and vertical chroma subsample values. For example for the
12646 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12650 the input video width and height
12654 the output width and height, that is the size of the padded area as
12655 specified by the @var{width} and @var{height} expressions
12659 the minimal width/height required for completely containing the input
12660 video rotated by @var{a} radians.
12662 These are only available when computing the @option{out_w} and
12663 @option{out_h} expressions.
12666 @subsection Examples
12670 Rotate the input by PI/6 radians clockwise:
12676 Rotate the input by PI/6 radians counter-clockwise:
12682 Rotate the input by 45 degrees clockwise:
12688 Apply a constant rotation with period T, starting from an angle of PI/3:
12690 rotate=PI/3+2*PI*t/T
12694 Make the input video rotation oscillating with a period of T
12695 seconds and an amplitude of A radians:
12697 rotate=A*sin(2*PI/T*t)
12701 Rotate the video, output size is chosen so that the whole rotating
12702 input video is always completely contained in the output:
12704 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
12708 Rotate the video, reduce the output size so that no background is ever
12711 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
12715 @subsection Commands
12717 The filter supports the following commands:
12721 Set the angle expression.
12722 The command accepts the same syntax of the corresponding option.
12724 If the specified expression is not valid, it is kept at its current
12730 Apply Shape Adaptive Blur.
12732 The filter accepts the following options:
12735 @item luma_radius, lr
12736 Set luma blur filter strength, must be a value in range 0.1-4.0, default
12737 value is 1.0. A greater value will result in a more blurred image, and
12738 in slower processing.
12740 @item luma_pre_filter_radius, lpfr
12741 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
12744 @item luma_strength, ls
12745 Set luma maximum difference between pixels to still be considered, must
12746 be a value in the 0.1-100.0 range, default value is 1.0.
12748 @item chroma_radius, cr
12749 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
12750 greater value will result in a more blurred image, and in slower
12753 @item chroma_pre_filter_radius, cpfr
12754 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
12756 @item chroma_strength, cs
12757 Set chroma maximum difference between pixels to still be considered,
12758 must be a value in the -0.9-100.0 range.
12761 Each chroma option value, if not explicitly specified, is set to the
12762 corresponding luma option value.
12767 Scale (resize) the input video, using the libswscale library.
12769 The scale filter forces the output display aspect ratio to be the same
12770 of the input, by changing the output sample aspect ratio.
12772 If the input image format is different from the format requested by
12773 the next filter, the scale filter will convert the input to the
12776 @subsection Options
12777 The filter accepts the following options, or any of the options
12778 supported by the libswscale scaler.
12780 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
12781 the complete list of scaler options.
12786 Set the output video dimension expression. Default value is the input
12789 If the @var{width} or @var{w} value is 0, the input width is used for
12790 the output. If the @var{height} or @var{h} value is 0, the input height
12791 is used for the output.
12793 If one and only one of the values is -n with n >= 1, the scale filter
12794 will use a value that maintains the aspect ratio of the input image,
12795 calculated from the other specified dimension. After that it will,
12796 however, make sure that the calculated dimension is divisible by n and
12797 adjust the value if necessary.
12799 If both values are -n with n >= 1, the behavior will be identical to
12800 both values being set to 0 as previously detailed.
12802 See below for the list of accepted constants for use in the dimension
12806 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
12810 Only evaluate expressions once during the filter initialization or when a command is processed.
12813 Evaluate expressions for each incoming frame.
12817 Default value is @samp{init}.
12821 Set the interlacing mode. It accepts the following values:
12825 Force interlaced aware scaling.
12828 Do not apply interlaced scaling.
12831 Select interlaced aware scaling depending on whether the source frames
12832 are flagged as interlaced or not.
12835 Default value is @samp{0}.
12838 Set libswscale scaling flags. See
12839 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12840 complete list of values. If not explicitly specified the filter applies
12844 @item param0, param1
12845 Set libswscale input parameters for scaling algorithms that need them. See
12846 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12847 complete documentation. If not explicitly specified the filter applies
12853 Set the video size. For the syntax of this option, check the
12854 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12856 @item in_color_matrix
12857 @item out_color_matrix
12858 Set in/output YCbCr color space type.
12860 This allows the autodetected value to be overridden as well as allows forcing
12861 a specific value used for the output and encoder.
12863 If not specified, the color space type depends on the pixel format.
12869 Choose automatically.
12872 Format conforming to International Telecommunication Union (ITU)
12873 Recommendation BT.709.
12876 Set color space conforming to the United States Federal Communications
12877 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
12880 Set color space conforming to:
12884 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
12887 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
12890 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
12895 Set color space conforming to SMPTE ST 240:1999.
12900 Set in/output YCbCr sample range.
12902 This allows the autodetected value to be overridden as well as allows forcing
12903 a specific value used for the output and encoder. If not specified, the
12904 range depends on the pixel format. Possible values:
12908 Choose automatically.
12911 Set full range (0-255 in case of 8-bit luma).
12914 Set "MPEG" range (16-235 in case of 8-bit luma).
12917 @item force_original_aspect_ratio
12918 Enable decreasing or increasing output video width or height if necessary to
12919 keep the original aspect ratio. Possible values:
12923 Scale the video as specified and disable this feature.
12926 The output video dimensions will automatically be decreased if needed.
12929 The output video dimensions will automatically be increased if needed.
12933 One useful instance of this option is that when you know a specific device's
12934 maximum allowed resolution, you can use this to limit the output video to
12935 that, while retaining the aspect ratio. For example, device A allows
12936 1280x720 playback, and your video is 1920x800. Using this option (set it to
12937 decrease) and specifying 1280x720 to the command line makes the output
12940 Please note that this is a different thing than specifying -1 for @option{w}
12941 or @option{h}, you still need to specify the output resolution for this option
12946 The values of the @option{w} and @option{h} options are expressions
12947 containing the following constants:
12952 The input width and height
12956 These are the same as @var{in_w} and @var{in_h}.
12960 The output (scaled) width and height
12964 These are the same as @var{out_w} and @var{out_h}
12967 The same as @var{iw} / @var{ih}
12970 input sample aspect ratio
12973 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
12977 horizontal and vertical input chroma subsample values. For example for the
12978 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12982 horizontal and vertical output chroma subsample values. For example for the
12983 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12986 @subsection Examples
12990 Scale the input video to a size of 200x100
12995 This is equivalent to:
13006 Specify a size abbreviation for the output size:
13011 which can also be written as:
13017 Scale the input to 2x:
13019 scale=w=2*iw:h=2*ih
13023 The above is the same as:
13025 scale=2*in_w:2*in_h
13029 Scale the input to 2x with forced interlaced scaling:
13031 scale=2*iw:2*ih:interl=1
13035 Scale the input to half size:
13037 scale=w=iw/2:h=ih/2
13041 Increase the width, and set the height to the same size:
13047 Seek Greek harmony:
13054 Increase the height, and set the width to 3/2 of the height:
13056 scale=w=3/2*oh:h=3/5*ih
13060 Increase the size, making the size a multiple of the chroma
13063 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13067 Increase the width to a maximum of 500 pixels,
13068 keeping the same aspect ratio as the input:
13070 scale=w='min(500\, iw*3/2):h=-1'
13074 @subsection Commands
13076 This filter supports the following commands:
13080 Set the output video dimension expression.
13081 The command accepts the same syntax of the corresponding option.
13083 If the specified expression is not valid, it is kept at its current
13089 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13090 format conversion on CUDA video frames. Setting the output width and height
13091 works in the same way as for the @var{scale} filter.
13093 The following additional options are accepted:
13096 The pixel format of the output CUDA frames. If set to the string "same" (the
13097 default), the input format will be kept. Note that automatic format negotiation
13098 and conversion is not yet supported for hardware frames
13101 The interpolation algorithm used for resizing. One of the following:
13108 @item cubic2p_bspline
13109 2-parameter cubic (B=1, C=0)
13111 @item cubic2p_catmullrom
13112 2-parameter cubic (B=0, C=1/2)
13114 @item cubic2p_b05c03
13115 2-parameter cubic (B=1/2, C=3/10)
13127 Scale (resize) the input video, based on a reference video.
13129 See the scale filter for available options, scale2ref supports the same but
13130 uses the reference video instead of the main input as basis. scale2ref also
13131 supports the following additional constants for the @option{w} and
13132 @option{h} options:
13137 The main input video's width and height
13140 The same as @var{main_w} / @var{main_h}
13143 The main input video's sample aspect ratio
13145 @item main_dar, mdar
13146 The main input video's display aspect ratio. Calculated from
13147 @code{(main_w / main_h) * main_sar}.
13151 The main input video's horizontal and vertical chroma subsample values.
13152 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13156 @subsection Examples
13160 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13162 'scale2ref[b][a];[a][b]overlay'
13166 @anchor{selectivecolor}
13167 @section selectivecolor
13169 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13170 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13171 by the "purity" of the color (that is, how saturated it already is).
13173 This filter is similar to the Adobe Photoshop Selective Color tool.
13175 The filter accepts the following options:
13178 @item correction_method
13179 Select color correction method.
13181 Available values are:
13184 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13187 Specified adjustments are relative to the original component value.
13189 Default is @code{absolute}.
13191 Adjustments for red pixels (pixels where the red component is the maximum)
13193 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13195 Adjustments for green pixels (pixels where the green component is the maximum)
13197 Adjustments for cyan pixels (pixels where the red component is the minimum)
13199 Adjustments for blue pixels (pixels where the blue component is the maximum)
13201 Adjustments for magenta pixels (pixels where the green component is the minimum)
13203 Adjustments for white pixels (pixels where all components are greater than 128)
13205 Adjustments for all pixels except pure black and pure white
13207 Adjustments for black pixels (pixels where all components are lesser than 128)
13209 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13212 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13213 4 space separated floating point adjustment values in the [-1,1] range,
13214 respectively to adjust the amount of cyan, magenta, yellow and black for the
13215 pixels of its range.
13217 @subsection Examples
13221 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13222 increase magenta by 27% in blue areas:
13224 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13228 Use a Photoshop selective color preset:
13230 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13234 @anchor{separatefields}
13235 @section separatefields
13237 The @code{separatefields} takes a frame-based video input and splits
13238 each frame into its components fields, producing a new half height clip
13239 with twice the frame rate and twice the frame count.
13241 This filter use field-dominance information in frame to decide which
13242 of each pair of fields to place first in the output.
13243 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13245 @section setdar, setsar
13247 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13250 This is done by changing the specified Sample (aka Pixel) Aspect
13251 Ratio, according to the following equation:
13253 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13256 Keep in mind that the @code{setdar} filter does not modify the pixel
13257 dimensions of the video frame. Also, the display aspect ratio set by
13258 this filter may be changed by later filters in the filterchain,
13259 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13262 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13263 the filter output video.
13265 Note that as a consequence of the application of this filter, the
13266 output display aspect ratio will change according to the equation
13269 Keep in mind that the sample aspect ratio set by the @code{setsar}
13270 filter may be changed by later filters in the filterchain, e.g. if
13271 another "setsar" or a "setdar" filter is applied.
13273 It accepts the following parameters:
13276 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13277 Set the aspect ratio used by the filter.
13279 The parameter can be a floating point number string, an expression, or
13280 a string of the form @var{num}:@var{den}, where @var{num} and
13281 @var{den} are the numerator and denominator of the aspect ratio. If
13282 the parameter is not specified, it is assumed the value "0".
13283 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13287 Set the maximum integer value to use for expressing numerator and
13288 denominator when reducing the expressed aspect ratio to a rational.
13289 Default value is @code{100}.
13293 The parameter @var{sar} is an expression containing
13294 the following constants:
13298 These are approximated values for the mathematical constants e
13299 (Euler's number), pi (Greek pi), and phi (the golden ratio).
13302 The input width and height.
13305 These are the same as @var{w} / @var{h}.
13308 The input sample aspect ratio.
13311 The input display aspect ratio. It is the same as
13312 (@var{w} / @var{h}) * @var{sar}.
13315 Horizontal and vertical chroma subsample values. For example, for the
13316 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13319 @subsection Examples
13324 To change the display aspect ratio to 16:9, specify one of the following:
13331 To change the sample aspect ratio to 10:11, specify:
13337 To set a display aspect ratio of 16:9, and specify a maximum integer value of
13338 1000 in the aspect ratio reduction, use the command:
13340 setdar=ratio=16/9:max=1000
13348 Force field for the output video frame.
13350 The @code{setfield} filter marks the interlace type field for the
13351 output frames. It does not change the input frame, but only sets the
13352 corresponding property, which affects how the frame is treated by
13353 following filters (e.g. @code{fieldorder} or @code{yadif}).
13355 The filter accepts the following options:
13360 Available values are:
13364 Keep the same field property.
13367 Mark the frame as bottom-field-first.
13370 Mark the frame as top-field-first.
13373 Mark the frame as progressive.
13379 Show a line containing various information for each input video frame.
13380 The input video is not modified.
13382 The shown line contains a sequence of key/value pairs of the form
13383 @var{key}:@var{value}.
13385 The following values are shown in the output:
13389 The (sequential) number of the input frame, starting from 0.
13392 The Presentation TimeStamp of the input frame, expressed as a number of
13393 time base units. The time base unit depends on the filter input pad.
13396 The Presentation TimeStamp of the input frame, expressed as a number of
13400 The position of the frame in the input stream, or -1 if this information is
13401 unavailable and/or meaningless (for example in case of synthetic video).
13404 The pixel format name.
13407 The sample aspect ratio of the input frame, expressed in the form
13408 @var{num}/@var{den}.
13411 The size of the input frame. For the syntax of this option, check the
13412 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13415 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
13416 for bottom field first).
13419 This is 1 if the frame is a key frame, 0 otherwise.
13422 The picture type of the input frame ("I" for an I-frame, "P" for a
13423 P-frame, "B" for a B-frame, or "?" for an unknown type).
13424 Also refer to the documentation of the @code{AVPictureType} enum and of
13425 the @code{av_get_picture_type_char} function defined in
13426 @file{libavutil/avutil.h}.
13429 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
13431 @item plane_checksum
13432 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
13433 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
13436 @section showpalette
13438 Displays the 256 colors palette of each frame. This filter is only relevant for
13439 @var{pal8} pixel format frames.
13441 It accepts the following option:
13445 Set the size of the box used to represent one palette color entry. Default is
13446 @code{30} (for a @code{30x30} pixel box).
13449 @section shuffleframes
13451 Reorder and/or duplicate and/or drop video frames.
13453 It accepts the following parameters:
13457 Set the destination indexes of input frames.
13458 This is space or '|' separated list of indexes that maps input frames to output
13459 frames. Number of indexes also sets maximal value that each index may have.
13460 '-1' index have special meaning and that is to drop frame.
13463 The first frame has the index 0. The default is to keep the input unchanged.
13465 @subsection Examples
13469 Swap second and third frame of every three frames of the input:
13471 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
13475 Swap 10th and 1st frame of every ten frames of the input:
13477 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
13481 @section shuffleplanes
13483 Reorder and/or duplicate video planes.
13485 It accepts the following parameters:
13490 The index of the input plane to be used as the first output plane.
13493 The index of the input plane to be used as the second output plane.
13496 The index of the input plane to be used as the third output plane.
13499 The index of the input plane to be used as the fourth output plane.
13503 The first plane has the index 0. The default is to keep the input unchanged.
13505 @subsection Examples
13509 Swap the second and third planes of the input:
13511 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
13515 @anchor{signalstats}
13516 @section signalstats
13517 Evaluate various visual metrics that assist in determining issues associated
13518 with the digitization of analog video media.
13520 By default the filter will log these metadata values:
13524 Display the minimal Y value contained within the input frame. Expressed in
13528 Display the Y value at the 10% percentile within the input frame. Expressed in
13532 Display the average Y value within the input frame. Expressed in range of
13536 Display the Y value at the 90% percentile within the input frame. Expressed in
13540 Display the maximum Y value contained within the input frame. Expressed in
13544 Display the minimal U value contained within the input frame. Expressed in
13548 Display the U value at the 10% percentile within the input frame. Expressed in
13552 Display the average U value within the input frame. Expressed in range of
13556 Display the U value at the 90% percentile within the input frame. Expressed in
13560 Display the maximum U value contained within the input frame. Expressed in
13564 Display the minimal V value contained within the input frame. Expressed in
13568 Display the V value at the 10% percentile within the input frame. Expressed in
13572 Display the average V value within the input frame. Expressed in range of
13576 Display the V value at the 90% percentile within the input frame. Expressed in
13580 Display the maximum V value contained within the input frame. Expressed in
13584 Display the minimal saturation value contained within the input frame.
13585 Expressed in range of [0-~181.02].
13588 Display the saturation value at the 10% percentile within the input frame.
13589 Expressed in range of [0-~181.02].
13592 Display the average saturation value within the input frame. Expressed in range
13596 Display the saturation value at the 90% percentile within the input frame.
13597 Expressed in range of [0-~181.02].
13600 Display the maximum saturation value contained within the input frame.
13601 Expressed in range of [0-~181.02].
13604 Display the median value for hue within the input frame. Expressed in range of
13608 Display the average value for hue within the input frame. Expressed in range of
13612 Display the average of sample value difference between all values of the Y
13613 plane in the current frame and corresponding values of the previous input frame.
13614 Expressed in range of [0-255].
13617 Display the average of sample value difference between all values of the U
13618 plane in the current frame and corresponding values of the previous input frame.
13619 Expressed in range of [0-255].
13622 Display the average of sample value difference between all values of the V
13623 plane in the current frame and corresponding values of the previous input frame.
13624 Expressed in range of [0-255].
13627 Display bit depth of Y plane in current frame.
13628 Expressed in range of [0-16].
13631 Display bit depth of U plane in current frame.
13632 Expressed in range of [0-16].
13635 Display bit depth of V plane in current frame.
13636 Expressed in range of [0-16].
13639 The filter accepts the following options:
13645 @option{stat} specify an additional form of image analysis.
13646 @option{out} output video with the specified type of pixel highlighted.
13648 Both options accept the following values:
13652 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
13653 unlike the neighboring pixels of the same field. Examples of temporal outliers
13654 include the results of video dropouts, head clogs, or tape tracking issues.
13657 Identify @var{vertical line repetition}. Vertical line repetition includes
13658 similar rows of pixels within a frame. In born-digital video vertical line
13659 repetition is common, but this pattern is uncommon in video digitized from an
13660 analog source. When it occurs in video that results from the digitization of an
13661 analog source it can indicate concealment from a dropout compensator.
13664 Identify pixels that fall outside of legal broadcast range.
13668 Set the highlight color for the @option{out} option. The default color is
13672 @subsection Examples
13676 Output data of various video metrics:
13678 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
13682 Output specific data about the minimum and maximum values of the Y plane per frame:
13684 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
13688 Playback video while highlighting pixels that are outside of broadcast range in red.
13690 ffplay example.mov -vf signalstats="out=brng:color=red"
13694 Playback video with signalstats metadata drawn over the frame.
13696 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
13699 The contents of signalstat_drawtext.txt used in the command are:
13702 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
13703 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
13704 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
13705 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
13713 Calculates the MPEG-7 Video Signature. The filter can handle more than one
13714 input. In this case the matching between the inputs can be calculated additionally.
13715 The filter always passes through the first input. The signature of each stream can
13716 be written into a file.
13718 It accepts the following options:
13722 Enable or disable the matching process.
13724 Available values are:
13728 Disable the calculation of a matching (default).
13730 Calculate the matching for the whole video and output whether the whole video
13731 matches or only parts.
13733 Calculate only until a matching is found or the video ends. Should be faster in
13738 Set the number of inputs. The option value must be a non negative integer.
13739 Default value is 1.
13742 Set the path to which the output is written. If there is more than one input,
13743 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
13744 integer), that will be replaced with the input number. If no filename is
13745 specified, no output will be written. This is the default.
13748 Choose the output format.
13750 Available values are:
13754 Use the specified binary representation (default).
13756 Use the specified xml representation.
13760 Set threshold to detect one word as similar. The option value must be an integer
13761 greater than zero. The default value is 9000.
13764 Set threshold to detect all words as similar. The option value must be an integer
13765 greater than zero. The default value is 60000.
13768 Set threshold to detect frames as similar. The option value must be an integer
13769 greater than zero. The default value is 116.
13772 Set the minimum length of a sequence in frames to recognize it as matching
13773 sequence. The option value must be a non negative integer value.
13774 The default value is 0.
13777 Set the minimum relation, that matching frames to all frames must have.
13778 The option value must be a double value between 0 and 1. The default value is 0.5.
13781 @subsection Examples
13785 To calculate the signature of an input video and store it in signature.bin:
13787 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
13791 To detect whether two videos match and store the signatures in XML format in
13792 signature0.xml and signature1.xml:
13794 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 -
13802 Blur the input video without impacting the outlines.
13804 It accepts the following options:
13807 @item luma_radius, lr
13808 Set the luma radius. The option value must be a float number in
13809 the range [0.1,5.0] that specifies the variance of the gaussian filter
13810 used to blur the image (slower if larger). Default value is 1.0.
13812 @item luma_strength, ls
13813 Set the luma strength. The option value must be a float number
13814 in the range [-1.0,1.0] that configures the blurring. A value included
13815 in [0.0,1.0] will blur the image whereas a value included in
13816 [-1.0,0.0] will sharpen the image. Default value is 1.0.
13818 @item luma_threshold, lt
13819 Set the luma threshold used as a coefficient to determine
13820 whether a pixel should be blurred or not. The option value must be an
13821 integer in the range [-30,30]. A value of 0 will filter all the image,
13822 a value included in [0,30] will filter flat areas and a value included
13823 in [-30,0] will filter edges. Default value is 0.
13825 @item chroma_radius, cr
13826 Set the chroma radius. The option value must be a float number in
13827 the range [0.1,5.0] that specifies the variance of the gaussian filter
13828 used to blur the image (slower if larger). Default value is @option{luma_radius}.
13830 @item chroma_strength, cs
13831 Set the chroma strength. The option value must be a float number
13832 in the range [-1.0,1.0] that configures the blurring. A value included
13833 in [0.0,1.0] will blur the image whereas a value included in
13834 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
13836 @item chroma_threshold, ct
13837 Set the chroma threshold used as a coefficient to determine
13838 whether a pixel should be blurred or not. The option value must be an
13839 integer in the range [-30,30]. A value of 0 will filter all the image,
13840 a value included in [0,30] will filter flat areas and a value included
13841 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
13844 If a chroma option is not explicitly set, the corresponding luma value
13849 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
13851 This filter takes in input two input videos, the first input is
13852 considered the "main" source and is passed unchanged to the
13853 output. The second input is used as a "reference" video for computing
13856 Both video inputs must have the same resolution and pixel format for
13857 this filter to work correctly. Also it assumes that both inputs
13858 have the same number of frames, which are compared one by one.
13860 The filter stores the calculated SSIM of each frame.
13862 The description of the accepted parameters follows.
13865 @item stats_file, f
13866 If specified the filter will use the named file to save the SSIM of
13867 each individual frame. When filename equals "-" the data is sent to
13871 The file printed if @var{stats_file} is selected, contains a sequence of
13872 key/value pairs of the form @var{key}:@var{value} for each compared
13875 A description of each shown parameter follows:
13879 sequential number of the input frame, starting from 1
13881 @item Y, U, V, R, G, B
13882 SSIM of the compared frames for the component specified by the suffix.
13885 SSIM of the compared frames for the whole frame.
13888 Same as above but in dB representation.
13891 This filter also supports the @ref{framesync} options.
13895 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13896 [main][ref] ssim="stats_file=stats.log" [out]
13899 On this example the input file being processed is compared with the
13900 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
13901 is stored in @file{stats.log}.
13903 Another example with both psnr and ssim at same time:
13905 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
13910 Convert between different stereoscopic image formats.
13912 The filters accept the following options:
13916 Set stereoscopic image format of input.
13918 Available values for input image formats are:
13921 side by side parallel (left eye left, right eye right)
13924 side by side crosseye (right eye left, left eye right)
13927 side by side parallel with half width resolution
13928 (left eye left, right eye right)
13931 side by side crosseye with half width resolution
13932 (right eye left, left eye right)
13935 above-below (left eye above, right eye below)
13938 above-below (right eye above, left eye below)
13941 above-below with half height resolution
13942 (left eye above, right eye below)
13945 above-below with half height resolution
13946 (right eye above, left eye below)
13949 alternating frames (left eye first, right eye second)
13952 alternating frames (right eye first, left eye second)
13955 interleaved rows (left eye has top row, right eye starts on next row)
13958 interleaved rows (right eye has top row, left eye starts on next row)
13961 interleaved columns, left eye first
13964 interleaved columns, right eye first
13966 Default value is @samp{sbsl}.
13970 Set stereoscopic image format of output.
13974 side by side parallel (left eye left, right eye right)
13977 side by side crosseye (right eye left, left eye right)
13980 side by side parallel with half width resolution
13981 (left eye left, right eye right)
13984 side by side crosseye with half width resolution
13985 (right eye left, left eye right)
13988 above-below (left eye above, right eye below)
13991 above-below (right eye above, left eye below)
13994 above-below with half height resolution
13995 (left eye above, right eye below)
13998 above-below with half height resolution
13999 (right eye above, left eye below)
14002 alternating frames (left eye first, right eye second)
14005 alternating frames (right eye first, left eye second)
14008 interleaved rows (left eye has top row, right eye starts on next row)
14011 interleaved rows (right eye has top row, left eye starts on next row)
14014 anaglyph red/blue gray
14015 (red filter on left eye, blue filter on right eye)
14018 anaglyph red/green gray
14019 (red filter on left eye, green filter on right eye)
14022 anaglyph red/cyan gray
14023 (red filter on left eye, cyan filter on right eye)
14026 anaglyph red/cyan half colored
14027 (red filter on left eye, cyan filter on right eye)
14030 anaglyph red/cyan color
14031 (red filter on left eye, cyan filter on right eye)
14034 anaglyph red/cyan color optimized with the least squares projection of dubois
14035 (red filter on left eye, cyan filter on right eye)
14038 anaglyph green/magenta gray
14039 (green filter on left eye, magenta filter on right eye)
14042 anaglyph green/magenta half colored
14043 (green filter on left eye, magenta filter on right eye)
14046 anaglyph green/magenta colored
14047 (green filter on left eye, magenta filter on right eye)
14050 anaglyph green/magenta color optimized with the least squares projection of dubois
14051 (green filter on left eye, magenta filter on right eye)
14054 anaglyph yellow/blue gray
14055 (yellow filter on left eye, blue filter on right eye)
14058 anaglyph yellow/blue half colored
14059 (yellow filter on left eye, blue filter on right eye)
14062 anaglyph yellow/blue colored
14063 (yellow filter on left eye, blue filter on right eye)
14066 anaglyph yellow/blue color optimized with the least squares projection of dubois
14067 (yellow filter on left eye, blue filter on right eye)
14070 mono output (left eye only)
14073 mono output (right eye only)
14076 checkerboard, left eye first
14079 checkerboard, right eye first
14082 interleaved columns, left eye first
14085 interleaved columns, right eye first
14091 Default value is @samp{arcd}.
14094 @subsection Examples
14098 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14104 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14110 @section streamselect, astreamselect
14111 Select video or audio streams.
14113 The filter accepts the following options:
14117 Set number of inputs. Default is 2.
14120 Set input indexes to remap to outputs.
14123 @subsection Commands
14125 The @code{streamselect} and @code{astreamselect} filter supports the following
14130 Set input indexes to remap to outputs.
14133 @subsection Examples
14137 Select first 5 seconds 1st stream and rest of time 2nd stream:
14139 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14143 Same as above, but for audio:
14145 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14150 Apply sobel operator to input video stream.
14152 The filter accepts the following option:
14156 Set which planes will be processed, unprocessed planes will be copied.
14157 By default value 0xf, all planes will be processed.
14160 Set value which will be multiplied with filtered result.
14163 Set value which will be added to filtered result.
14169 Apply a simple postprocessing filter that compresses and decompresses the image
14170 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14171 and average the results.
14173 The filter accepts the following options:
14177 Set quality. This option defines the number of levels for averaging. It accepts
14178 an integer in the range 0-6. If set to @code{0}, the filter will have no
14179 effect. A value of @code{6} means the higher quality. For each increment of
14180 that value the speed drops by a factor of approximately 2. Default value is
14184 Force a constant quantization parameter. If not set, the filter will use the QP
14185 from the video stream (if available).
14188 Set thresholding mode. Available modes are:
14192 Set hard thresholding (default).
14194 Set soft thresholding (better de-ringing effect, but likely blurrier).
14197 @item use_bframe_qp
14198 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14199 option may cause flicker since the B-Frames have often larger QP. Default is
14200 @code{0} (not enabled).
14206 Draw subtitles on top of input video using the libass library.
14208 To enable compilation of this filter you need to configure FFmpeg with
14209 @code{--enable-libass}. This filter also requires a build with libavcodec and
14210 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14211 Alpha) subtitles format.
14213 The filter accepts the following options:
14217 Set the filename of the subtitle file to read. It must be specified.
14219 @item original_size
14220 Specify the size of the original video, the video for which the ASS file
14221 was composed. For the syntax of this option, check the
14222 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14223 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14224 correctly scale the fonts if the aspect ratio has been changed.
14227 Set a directory path containing fonts that can be used by the filter.
14228 These fonts will be used in addition to whatever the font provider uses.
14231 Process alpha channel, by default alpha channel is untouched.
14234 Set subtitles input character encoding. @code{subtitles} filter only. Only
14235 useful if not UTF-8.
14237 @item stream_index, si
14238 Set subtitles stream index. @code{subtitles} filter only.
14241 Override default style or script info parameters of the subtitles. It accepts a
14242 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14245 If the first key is not specified, it is assumed that the first value
14246 specifies the @option{filename}.
14248 For example, to render the file @file{sub.srt} on top of the input
14249 video, use the command:
14254 which is equivalent to:
14256 subtitles=filename=sub.srt
14259 To render the default subtitles stream from file @file{video.mkv}, use:
14261 subtitles=video.mkv
14264 To render the second subtitles stream from that file, use:
14266 subtitles=video.mkv:si=1
14269 To make the subtitles stream from @file{sub.srt} appear in transparent green
14270 @code{DejaVu Serif}, use:
14272 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14275 @section super2xsai
14277 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14278 Interpolate) pixel art scaling algorithm.
14280 Useful for enlarging pixel art images without reducing sharpness.
14284 Swap two rectangular objects in video.
14286 This filter accepts the following options:
14296 Set 1st rect x coordinate.
14299 Set 1st rect y coordinate.
14302 Set 2nd rect x coordinate.
14305 Set 2nd rect y coordinate.
14307 All expressions are evaluated once for each frame.
14310 The all options are expressions containing the following constants:
14315 The input width and height.
14318 same as @var{w} / @var{h}
14321 input sample aspect ratio
14324 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
14327 The number of the input frame, starting from 0.
14330 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
14333 the position in the file of the input frame, NAN if unknown
14341 Apply telecine process to the video.
14343 This filter accepts the following options:
14352 The default value is @code{top}.
14356 A string of numbers representing the pulldown pattern you wish to apply.
14357 The default value is @code{23}.
14361 Some typical patterns:
14366 24p: 2332 (preferred)
14373 24p: 222222222223 ("Euro pulldown")
14380 Apply threshold effect to video stream.
14382 This filter needs four video streams to perform thresholding.
14383 First stream is stream we are filtering.
14384 Second stream is holding threshold values, third stream is holding min values,
14385 and last, fourth stream is holding max values.
14387 The filter accepts the following option:
14391 Set which planes will be processed, unprocessed planes will be copied.
14392 By default value 0xf, all planes will be processed.
14395 For example if first stream pixel's component value is less then threshold value
14396 of pixel component from 2nd threshold stream, third stream value will picked,
14397 otherwise fourth stream pixel component value will be picked.
14399 Using color source filter one can perform various types of thresholding:
14401 @subsection Examples
14405 Binary threshold, using gray color as threshold:
14407 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
14411 Inverted binary threshold, using gray color as threshold:
14413 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
14417 Truncate binary threshold, using gray color as threshold:
14419 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
14423 Threshold to zero, using gray color as threshold:
14425 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
14429 Inverted threshold to zero, using gray color as threshold:
14431 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
14436 Select the most representative frame in a given sequence of consecutive frames.
14438 The filter accepts the following options:
14442 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
14443 will pick one of them, and then handle the next batch of @var{n} frames until
14444 the end. Default is @code{100}.
14447 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
14448 value will result in a higher memory usage, so a high value is not recommended.
14450 @subsection Examples
14454 Extract one picture each 50 frames:
14460 Complete example of a thumbnail creation with @command{ffmpeg}:
14462 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
14468 Tile several successive frames together.
14470 The filter accepts the following options:
14475 Set the grid size (i.e. the number of lines and columns). For the syntax of
14476 this option, check the
14477 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14480 Set the maximum number of frames to render in the given area. It must be less
14481 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
14482 the area will be used.
14485 Set the outer border margin in pixels.
14488 Set the inner border thickness (i.e. the number of pixels between frames). For
14489 more advanced padding options (such as having different values for the edges),
14490 refer to the pad video filter.
14493 Specify the color of the unused area. For the syntax of this option, check the
14494 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
14498 Set the number of frames to overlap when tiling several successive frames together.
14499 The value must be between @code{0} and @var{nb_frames - 1}.
14502 @subsection Examples
14506 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
14508 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
14510 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
14511 duplicating each output frame to accommodate the originally detected frame
14515 Display @code{5} pictures in an area of @code{3x2} frames,
14516 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
14517 mixed flat and named options:
14519 tile=3x2:nb_frames=5:padding=7:margin=2
14523 @section tinterlace
14525 Perform various types of temporal field interlacing.
14527 Frames are counted starting from 1, so the first input frame is
14530 The filter accepts the following options:
14535 Specify the mode of the interlacing. This option can also be specified
14536 as a value alone. See below for a list of values for this option.
14538 Available values are:
14542 Move odd frames into the upper field, even into the lower field,
14543 generating a double height frame at half frame rate.
14547 Frame 1 Frame 2 Frame 3 Frame 4
14549 11111 22222 33333 44444
14550 11111 22222 33333 44444
14551 11111 22222 33333 44444
14552 11111 22222 33333 44444
14566 Only output odd frames, even frames are dropped, generating a frame with
14567 unchanged height at half frame rate.
14572 Frame 1 Frame 2 Frame 3 Frame 4
14574 11111 22222 33333 44444
14575 11111 22222 33333 44444
14576 11111 22222 33333 44444
14577 11111 22222 33333 44444
14587 Only output even frames, odd frames are dropped, generating a frame with
14588 unchanged height at half frame rate.
14593 Frame 1 Frame 2 Frame 3 Frame 4
14595 11111 22222 33333 44444
14596 11111 22222 33333 44444
14597 11111 22222 33333 44444
14598 11111 22222 33333 44444
14608 Expand each frame to full height, but pad alternate lines with black,
14609 generating a frame with double height at the same input frame rate.
14614 Frame 1 Frame 2 Frame 3 Frame 4
14616 11111 22222 33333 44444
14617 11111 22222 33333 44444
14618 11111 22222 33333 44444
14619 11111 22222 33333 44444
14622 11111 ..... 33333 .....
14623 ..... 22222 ..... 44444
14624 11111 ..... 33333 .....
14625 ..... 22222 ..... 44444
14626 11111 ..... 33333 .....
14627 ..... 22222 ..... 44444
14628 11111 ..... 33333 .....
14629 ..... 22222 ..... 44444
14633 @item interleave_top, 4
14634 Interleave the upper field from odd frames with the lower field from
14635 even frames, generating a frame with unchanged height at half frame rate.
14640 Frame 1 Frame 2 Frame 3 Frame 4
14642 11111<- 22222 33333<- 44444
14643 11111 22222<- 33333 44444<-
14644 11111<- 22222 33333<- 44444
14645 11111 22222<- 33333 44444<-
14655 @item interleave_bottom, 5
14656 Interleave the lower field from odd frames with the upper field from
14657 even frames, generating a frame with unchanged height at half frame rate.
14662 Frame 1 Frame 2 Frame 3 Frame 4
14664 11111 22222<- 33333 44444<-
14665 11111<- 22222 33333<- 44444
14666 11111 22222<- 33333 44444<-
14667 11111<- 22222 33333<- 44444
14677 @item interlacex2, 6
14678 Double frame rate with unchanged height. Frames are inserted each
14679 containing the second temporal field from the previous input frame and
14680 the first temporal field from the next input frame. This mode relies on
14681 the top_field_first flag. Useful for interlaced video displays with no
14682 field synchronisation.
14687 Frame 1 Frame 2 Frame 3 Frame 4
14689 11111 22222 33333 44444
14690 11111 22222 33333 44444
14691 11111 22222 33333 44444
14692 11111 22222 33333 44444
14695 11111 22222 22222 33333 33333 44444 44444
14696 11111 11111 22222 22222 33333 33333 44444
14697 11111 22222 22222 33333 33333 44444 44444
14698 11111 11111 22222 22222 33333 33333 44444
14703 Move odd frames into the upper field, even into the lower field,
14704 generating a double height frame at same frame rate.
14709 Frame 1 Frame 2 Frame 3 Frame 4
14711 11111 22222 33333 44444
14712 11111 22222 33333 44444
14713 11111 22222 33333 44444
14714 11111 22222 33333 44444
14717 11111 33333 33333 55555
14718 22222 22222 44444 44444
14719 11111 33333 33333 55555
14720 22222 22222 44444 44444
14721 11111 33333 33333 55555
14722 22222 22222 44444 44444
14723 11111 33333 33333 55555
14724 22222 22222 44444 44444
14729 Numeric values are deprecated but are accepted for backward
14730 compatibility reasons.
14732 Default mode is @code{merge}.
14735 Specify flags influencing the filter process.
14737 Available value for @var{flags} is:
14740 @item low_pass_filter, vlfp
14741 Enable linear vertical low-pass filtering in the filter.
14742 Vertical low-pass filtering is required when creating an interlaced
14743 destination from a progressive source which contains high-frequency
14744 vertical detail. Filtering will reduce interlace 'twitter' and Moire
14747 @item complex_filter, cvlfp
14748 Enable complex vertical low-pass filtering.
14749 This will slightly less reduce interlace 'twitter' and Moire
14750 patterning but better retain detail and subjective sharpness impression.
14754 Vertical low-pass filtering can only be enabled for @option{mode}
14755 @var{interleave_top} and @var{interleave_bottom}.
14760 Tone map colors from different dynamic ranges.
14762 This filter expects data in single precision floating point, as it needs to
14763 operate on (and can output) out-of-range values. Another filter, such as
14764 @ref{zscale}, is needed to convert the resulting frame to a usable format.
14766 The tonemapping algorithms implemented only work on linear light, so input
14767 data should be linearized beforehand (and possibly correctly tagged).
14770 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
14773 @subsection Options
14774 The filter accepts the following options.
14778 Set the tone map algorithm to use.
14780 Possible values are:
14783 Do not apply any tone map, only desaturate overbright pixels.
14786 Hard-clip any out-of-range values. Use it for perfect color accuracy for
14787 in-range values, while distorting out-of-range values.
14790 Stretch the entire reference gamut to a linear multiple of the display.
14793 Fit a logarithmic transfer between the tone curves.
14796 Preserve overall image brightness with a simple curve, using nonlinear
14797 contrast, which results in flattening details and degrading color accuracy.
14800 Preserve both dark and bright details better than @var{reinhard}, at the cost
14801 of slightly darkening everything. Use it when detail preservation is more
14802 important than color and brightness accuracy.
14805 Smoothly map out-of-range values, while retaining contrast and colors for
14806 in-range material as much as possible. Use it when color accuracy is more
14807 important than detail preservation.
14813 Tune the tone mapping algorithm.
14815 This affects the following algorithms:
14821 Specifies the scale factor to use while stretching.
14825 Specifies the exponent of the function.
14829 Specify an extra linear coefficient to multiply into the signal before clipping.
14833 Specify the local contrast coefficient at the display peak.
14834 Default to 0.5, which means that in-gamut values will be about half as bright
14841 Specify the transition point from linear to mobius transform. Every value
14842 below this point is guaranteed to be mapped 1:1. The higher the value, the
14843 more accurate the result will be, at the cost of losing bright details.
14844 Default to 0.3, which due to the steep initial slope still preserves in-range
14845 colors fairly accurately.
14849 Apply desaturation for highlights that exceed this level of brightness. The
14850 higher the parameter, the more color information will be preserved. This
14851 setting helps prevent unnaturally blown-out colors for super-highlights, by
14852 (smoothly) turning into white instead. This makes images feel more natural,
14853 at the cost of reducing information about out-of-range colors.
14855 The default of 2.0 is somewhat conservative and will mostly just apply to
14856 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
14858 This option works only if the input frame has a supported color tag.
14861 Override signal/nominal/reference peak with this value. Useful when the
14862 embedded peak information in display metadata is not reliable or when tone
14863 mapping from a lower range to a higher range.
14868 Transpose rows with columns in the input video and optionally flip it.
14870 It accepts the following parameters:
14875 Specify the transposition direction.
14877 Can assume the following values:
14879 @item 0, 4, cclock_flip
14880 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
14888 Rotate by 90 degrees clockwise, that is:
14896 Rotate by 90 degrees counterclockwise, that is:
14903 @item 3, 7, clock_flip
14904 Rotate by 90 degrees clockwise and vertically flip, that is:
14912 For values between 4-7, the transposition is only done if the input
14913 video geometry is portrait and not landscape. These values are
14914 deprecated, the @code{passthrough} option should be used instead.
14916 Numerical values are deprecated, and should be dropped in favor of
14917 symbolic constants.
14920 Do not apply the transposition if the input geometry matches the one
14921 specified by the specified value. It accepts the following values:
14924 Always apply transposition.
14926 Preserve portrait geometry (when @var{height} >= @var{width}).
14928 Preserve landscape geometry (when @var{width} >= @var{height}).
14931 Default value is @code{none}.
14934 For example to rotate by 90 degrees clockwise and preserve portrait
14937 transpose=dir=1:passthrough=portrait
14940 The command above can also be specified as:
14942 transpose=1:portrait
14946 Trim the input so that the output contains one continuous subpart of the input.
14948 It accepts the following parameters:
14951 Specify the time of the start of the kept section, i.e. the frame with the
14952 timestamp @var{start} will be the first frame in the output.
14955 Specify the time of the first frame that will be dropped, i.e. the frame
14956 immediately preceding the one with the timestamp @var{end} will be the last
14957 frame in the output.
14960 This is the same as @var{start}, except this option sets the start timestamp
14961 in timebase units instead of seconds.
14964 This is the same as @var{end}, except this option sets the end timestamp
14965 in timebase units instead of seconds.
14968 The maximum duration of the output in seconds.
14971 The number of the first frame that should be passed to the output.
14974 The number of the first frame that should be dropped.
14977 @option{start}, @option{end}, and @option{duration} are expressed as time
14978 duration specifications; see
14979 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14980 for the accepted syntax.
14982 Note that the first two sets of the start/end options and the @option{duration}
14983 option look at the frame timestamp, while the _frame variants simply count the
14984 frames that pass through the filter. Also note that this filter does not modify
14985 the timestamps. If you wish for the output timestamps to start at zero, insert a
14986 setpts filter after the trim filter.
14988 If multiple start or end options are set, this filter tries to be greedy and
14989 keep all the frames that match at least one of the specified constraints. To keep
14990 only the part that matches all the constraints at once, chain multiple trim
14993 The defaults are such that all the input is kept. So it is possible to set e.g.
14994 just the end values to keep everything before the specified time.
14999 Drop everything except the second minute of input:
15001 ffmpeg -i INPUT -vf trim=60:120
15005 Keep only the first second:
15007 ffmpeg -i INPUT -vf trim=duration=1
15012 @section unpremultiply
15013 Apply alpha unpremultiply effect to input video stream using first plane
15014 of second stream as alpha.
15016 Both streams must have same dimensions and same pixel format.
15018 The filter accepts the following option:
15022 Set which planes will be processed, unprocessed planes will be copied.
15023 By default value 0xf, all planes will be processed.
15025 If the format has 1 or 2 components, then luma is bit 0.
15026 If the format has 3 or 4 components:
15027 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15028 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15029 If present, the alpha channel is always the last bit.
15032 Do not require 2nd input for processing, instead use alpha plane from input stream.
15038 Sharpen or blur the input video.
15040 It accepts the following parameters:
15043 @item luma_msize_x, lx
15044 Set the luma matrix horizontal size. It must be an odd integer between
15045 3 and 23. The default value is 5.
15047 @item luma_msize_y, ly
15048 Set the luma matrix vertical size. It must be an odd integer between 3
15049 and 23. The default value is 5.
15051 @item luma_amount, la
15052 Set the luma effect strength. It must be a floating point number, reasonable
15053 values lay between -1.5 and 1.5.
15055 Negative values will blur the input video, while positive values will
15056 sharpen it, a value of zero will disable the effect.
15058 Default value is 1.0.
15060 @item chroma_msize_x, cx
15061 Set the chroma matrix horizontal size. It must be an odd integer
15062 between 3 and 23. The default value is 5.
15064 @item chroma_msize_y, cy
15065 Set the chroma matrix vertical size. It must be an odd integer
15066 between 3 and 23. The default value is 5.
15068 @item chroma_amount, ca
15069 Set the chroma effect strength. It must be a floating point number, reasonable
15070 values lay between -1.5 and 1.5.
15072 Negative values will blur the input video, while positive values will
15073 sharpen it, a value of zero will disable the effect.
15075 Default value is 0.0.
15079 All parameters are optional and default to the equivalent of the
15080 string '5:5:1.0:5:5:0.0'.
15082 @subsection Examples
15086 Apply strong luma sharpen effect:
15088 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15092 Apply a strong blur of both luma and chroma parameters:
15094 unsharp=7:7:-2:7:7:-2
15100 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15101 the image at several (or - in the case of @option{quality} level @code{8} - all)
15102 shifts and average the results.
15104 The way this differs from the behavior of spp is that uspp actually encodes &
15105 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15106 DCT similar to MJPEG.
15108 The filter accepts the following options:
15112 Set quality. This option defines the number of levels for averaging. It accepts
15113 an integer in the range 0-8. If set to @code{0}, the filter will have no
15114 effect. A value of @code{8} means the higher quality. For each increment of
15115 that value the speed drops by a factor of approximately 2. Default value is
15119 Force a constant quantization parameter. If not set, the filter will use the QP
15120 from the video stream (if available).
15123 @section vaguedenoiser
15125 Apply a wavelet based denoiser.
15127 It transforms each frame from the video input into the wavelet domain,
15128 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15129 the obtained coefficients. It does an inverse wavelet transform after.
15130 Due to wavelet properties, it should give a nice smoothed result, and
15131 reduced noise, without blurring picture features.
15133 This filter accepts the following options:
15137 The filtering strength. The higher, the more filtered the video will be.
15138 Hard thresholding can use a higher threshold than soft thresholding
15139 before the video looks overfiltered. Default value is 2.
15142 The filtering method the filter will use.
15144 It accepts the following values:
15147 All values under the threshold will be zeroed.
15150 All values under the threshold will be zeroed. All values above will be
15151 reduced by the threshold.
15154 Scales or nullifies coefficients - intermediary between (more) soft and
15155 (less) hard thresholding.
15158 Default is garrote.
15161 Number of times, the wavelet will decompose the picture. Picture can't
15162 be decomposed beyond a particular point (typically, 8 for a 640x480
15163 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15166 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15169 A list of the planes to process. By default all planes are processed.
15172 @section vectorscope
15174 Display 2 color component values in the two dimensional graph (which is called
15177 This filter accepts the following options:
15181 Set vectorscope mode.
15183 It accepts the following values:
15186 Gray values are displayed on graph, higher brightness means more pixels have
15187 same component color value on location in graph. This is the default mode.
15190 Gray values are displayed on graph. Surrounding pixels values which are not
15191 present in video frame are drawn in gradient of 2 color components which are
15192 set by option @code{x} and @code{y}. The 3rd color component is static.
15195 Actual color components values present in video frame are displayed on graph.
15198 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15199 on graph increases value of another color component, which is luminance by
15200 default values of @code{x} and @code{y}.
15203 Actual colors present in video frame are displayed on graph. If two different
15204 colors map to same position on graph then color with higher value of component
15205 not present in graph is picked.
15208 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15209 component picked from radial gradient.
15213 Set which color component will be represented on X-axis. Default is @code{1}.
15216 Set which color component will be represented on Y-axis. Default is @code{2}.
15219 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15220 of color component which represents frequency of (X, Y) location in graph.
15225 No envelope, this is default.
15228 Instant envelope, even darkest single pixel will be clearly highlighted.
15231 Hold maximum and minimum values presented in graph over time. This way you
15232 can still spot out of range values without constantly looking at vectorscope.
15235 Peak and instant envelope combined together.
15239 Set what kind of graticule to draw.
15247 Set graticule opacity.
15250 Set graticule flags.
15254 Draw graticule for white point.
15257 Draw graticule for black point.
15260 Draw color points short names.
15264 Set background opacity.
15266 @item lthreshold, l
15267 Set low threshold for color component not represented on X or Y axis.
15268 Values lower than this value will be ignored. Default is 0.
15269 Note this value is multiplied with actual max possible value one pixel component
15270 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15273 @item hthreshold, h
15274 Set high threshold for color component not represented on X or Y axis.
15275 Values higher than this value will be ignored. Default is 1.
15276 Note this value is multiplied with actual max possible value one pixel component
15277 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15278 is 0.9 * 255 = 230.
15280 @item colorspace, c
15281 Set what kind of colorspace to use when drawing graticule.
15290 @anchor{vidstabdetect}
15291 @section vidstabdetect
15293 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
15294 @ref{vidstabtransform} for pass 2.
15296 This filter generates a file with relative translation and rotation
15297 transform information about subsequent frames, which is then used by
15298 the @ref{vidstabtransform} filter.
15300 To enable compilation of this filter you need to configure FFmpeg with
15301 @code{--enable-libvidstab}.
15303 This filter accepts the following options:
15307 Set the path to the file used to write the transforms information.
15308 Default value is @file{transforms.trf}.
15311 Set how shaky the video is and how quick the camera is. It accepts an
15312 integer in the range 1-10, a value of 1 means little shakiness, a
15313 value of 10 means strong shakiness. Default value is 5.
15316 Set the accuracy of the detection process. It must be a value in the
15317 range 1-15. A value of 1 means low accuracy, a value of 15 means high
15318 accuracy. Default value is 15.
15321 Set stepsize of the search process. The region around minimum is
15322 scanned with 1 pixel resolution. Default value is 6.
15325 Set minimum contrast. Below this value a local measurement field is
15326 discarded. Must be a floating point value in the range 0-1. Default
15330 Set reference frame number for tripod mode.
15332 If enabled, the motion of the frames is compared to a reference frame
15333 in the filtered stream, identified by the specified number. The idea
15334 is to compensate all movements in a more-or-less static scene and keep
15335 the camera view absolutely still.
15337 If set to 0, it is disabled. The frames are counted starting from 1.
15340 Show fields and transforms in the resulting frames. It accepts an
15341 integer in the range 0-2. Default value is 0, which disables any
15345 @subsection Examples
15349 Use default values:
15355 Analyze strongly shaky movie and put the results in file
15356 @file{mytransforms.trf}:
15358 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
15362 Visualize the result of internal transformations in the resulting
15365 vidstabdetect=show=1
15369 Analyze a video with medium shakiness using @command{ffmpeg}:
15371 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
15375 @anchor{vidstabtransform}
15376 @section vidstabtransform
15378 Video stabilization/deshaking: pass 2 of 2,
15379 see @ref{vidstabdetect} for pass 1.
15381 Read a file with transform information for each frame and
15382 apply/compensate them. Together with the @ref{vidstabdetect}
15383 filter this can be used to deshake videos. See also
15384 @url{http://public.hronopik.de/vid.stab}. It is important to also use
15385 the @ref{unsharp} filter, see below.
15387 To enable compilation of this filter you need to configure FFmpeg with
15388 @code{--enable-libvidstab}.
15390 @subsection Options
15394 Set path to the file used to read the transforms. Default value is
15395 @file{transforms.trf}.
15398 Set the number of frames (value*2 + 1) used for lowpass filtering the
15399 camera movements. Default value is 10.
15401 For example a number of 10 means that 21 frames are used (10 in the
15402 past and 10 in the future) to smoothen the motion in the video. A
15403 larger value leads to a smoother video, but limits the acceleration of
15404 the camera (pan/tilt movements). 0 is a special case where a static
15405 camera is simulated.
15408 Set the camera path optimization algorithm.
15410 Accepted values are:
15413 gaussian kernel low-pass filter on camera motion (default)
15415 averaging on transformations
15419 Set maximal number of pixels to translate frames. Default value is -1,
15423 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
15424 value is -1, meaning no limit.
15427 Specify how to deal with borders that may be visible due to movement
15430 Available values are:
15433 keep image information from previous frame (default)
15435 fill the border black
15439 Invert transforms if set to 1. Default value is 0.
15442 Consider transforms as relative to previous frame if set to 1,
15443 absolute if set to 0. Default value is 0.
15446 Set percentage to zoom. A positive value will result in a zoom-in
15447 effect, a negative value in a zoom-out effect. Default value is 0 (no
15451 Set optimal zooming to avoid borders.
15453 Accepted values are:
15458 optimal static zoom value is determined (only very strong movements
15459 will lead to visible borders) (default)
15461 optimal adaptive zoom value is determined (no borders will be
15462 visible), see @option{zoomspeed}
15465 Note that the value given at zoom is added to the one calculated here.
15468 Set percent to zoom maximally each frame (enabled when
15469 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
15473 Specify type of interpolation.
15475 Available values are:
15480 linear only horizontal
15482 linear in both directions (default)
15484 cubic in both directions (slow)
15488 Enable virtual tripod mode if set to 1, which is equivalent to
15489 @code{relative=0:smoothing=0}. Default value is 0.
15491 Use also @code{tripod} option of @ref{vidstabdetect}.
15494 Increase log verbosity if set to 1. Also the detected global motions
15495 are written to the temporary file @file{global_motions.trf}. Default
15499 @subsection Examples
15503 Use @command{ffmpeg} for a typical stabilization with default values:
15505 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
15508 Note the use of the @ref{unsharp} filter which is always recommended.
15511 Zoom in a bit more and load transform data from a given file:
15513 vidstabtransform=zoom=5:input="mytransforms.trf"
15517 Smoothen the video even more:
15519 vidstabtransform=smoothing=30
15525 Flip the input video vertically.
15527 For example, to vertically flip a video with @command{ffmpeg}:
15529 ffmpeg -i in.avi -vf "vflip" out.avi
15535 Make or reverse a natural vignetting effect.
15537 The filter accepts the following options:
15541 Set lens angle expression as a number of radians.
15543 The value is clipped in the @code{[0,PI/2]} range.
15545 Default value: @code{"PI/5"}
15549 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
15553 Set forward/backward mode.
15555 Available modes are:
15558 The larger the distance from the central point, the darker the image becomes.
15561 The larger the distance from the central point, the brighter the image becomes.
15562 This can be used to reverse a vignette effect, though there is no automatic
15563 detection to extract the lens @option{angle} and other settings (yet). It can
15564 also be used to create a burning effect.
15567 Default value is @samp{forward}.
15570 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
15572 It accepts the following values:
15575 Evaluate expressions only once during the filter initialization.
15578 Evaluate expressions for each incoming frame. This is way slower than the
15579 @samp{init} mode since it requires all the scalers to be re-computed, but it
15580 allows advanced dynamic expressions.
15583 Default value is @samp{init}.
15586 Set dithering to reduce the circular banding effects. Default is @code{1}
15590 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
15591 Setting this value to the SAR of the input will make a rectangular vignetting
15592 following the dimensions of the video.
15594 Default is @code{1/1}.
15597 @subsection Expressions
15599 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
15600 following parameters.
15605 input width and height
15608 the number of input frame, starting from 0
15611 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
15612 @var{TB} units, NAN if undefined
15615 frame rate of the input video, NAN if the input frame rate is unknown
15618 the PTS (Presentation TimeStamp) of the filtered video frame,
15619 expressed in seconds, NAN if undefined
15622 time base of the input video
15626 @subsection Examples
15630 Apply simple strong vignetting effect:
15636 Make a flickering vignetting:
15638 vignette='PI/4+random(1)*PI/50':eval=frame
15643 @section vmafmotion
15645 Obtain the average vmaf motion score of a video.
15646 It is one of the component filters of VMAF.
15648 The obtained average motion score is printed through the logging system.
15650 In the below example the input file @file{ref.mpg} is being processed and score
15654 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
15658 Stack input videos vertically.
15660 All streams must be of same pixel format and of same width.
15662 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
15663 to create same output.
15665 The filter accept the following option:
15669 Set number of input streams. Default is 2.
15672 If set to 1, force the output to terminate when the shortest input
15673 terminates. Default value is 0.
15678 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
15679 Deinterlacing Filter").
15681 Based on the process described by Martin Weston for BBC R&D, and
15682 implemented based on the de-interlace algorithm written by Jim
15683 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
15684 uses filter coefficients calculated by BBC R&D.
15686 There are two sets of filter coefficients, so called "simple":
15687 and "complex". Which set of filter coefficients is used can
15688 be set by passing an optional parameter:
15692 Set the interlacing filter coefficients. Accepts one of the following values:
15696 Simple filter coefficient set.
15698 More-complex filter coefficient set.
15700 Default value is @samp{complex}.
15703 Specify which frames to deinterlace. Accept one of the following values:
15707 Deinterlace all frames,
15709 Only deinterlace frames marked as interlaced.
15712 Default value is @samp{all}.
15716 Video waveform monitor.
15718 The waveform monitor plots color component intensity. By default luminance
15719 only. Each column of the waveform corresponds to a column of pixels in the
15722 It accepts the following options:
15726 Can be either @code{row}, or @code{column}. Default is @code{column}.
15727 In row mode, the graph on the left side represents color component value 0 and
15728 the right side represents value = 255. In column mode, the top side represents
15729 color component value = 0 and bottom side represents value = 255.
15732 Set intensity. Smaller values are useful to find out how many values of the same
15733 luminance are distributed across input rows/columns.
15734 Default value is @code{0.04}. Allowed range is [0, 1].
15737 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
15738 In mirrored mode, higher values will be represented on the left
15739 side for @code{row} mode and at the top for @code{column} mode. Default is
15740 @code{1} (mirrored).
15744 It accepts the following values:
15747 Presents information identical to that in the @code{parade}, except
15748 that the graphs representing color components are superimposed directly
15751 This display mode makes it easier to spot relative differences or similarities
15752 in overlapping areas of the color components that are supposed to be identical,
15753 such as neutral whites, grays, or blacks.
15756 Display separate graph for the color components side by side in
15757 @code{row} mode or one below the other in @code{column} mode.
15760 Display separate graph for the color components side by side in
15761 @code{column} mode or one below the other in @code{row} mode.
15763 Using this display mode makes it easy to spot color casts in the highlights
15764 and shadows of an image, by comparing the contours of the top and the bottom
15765 graphs of each waveform. Since whites, grays, and blacks are characterized
15766 by exactly equal amounts of red, green, and blue, neutral areas of the picture
15767 should display three waveforms of roughly equal width/height. If not, the
15768 correction is easy to perform by making level adjustments the three waveforms.
15770 Default is @code{stack}.
15772 @item components, c
15773 Set which color components to display. Default is 1, which means only luminance
15774 or red color component if input is in RGB colorspace. If is set for example to
15775 7 it will display all 3 (if) available color components.
15780 No envelope, this is default.
15783 Instant envelope, minimum and maximum values presented in graph will be easily
15784 visible even with small @code{step} value.
15787 Hold minimum and maximum values presented in graph across time. This way you
15788 can still spot out of range values without constantly looking at waveforms.
15791 Peak and instant envelope combined together.
15797 No filtering, this is default.
15800 Luma and chroma combined together.
15803 Similar as above, but shows difference between blue and red chroma.
15806 Displays only chroma.
15809 Displays actual color value on waveform.
15812 Similar as above, but with luma showing frequency of chroma values.
15816 Set which graticule to display.
15820 Do not display graticule.
15823 Display green graticule showing legal broadcast ranges.
15827 Set graticule opacity.
15830 Set graticule flags.
15834 Draw numbers above lines. By default enabled.
15837 Draw dots instead of lines.
15841 Set scale used for displaying graticule.
15848 Default is digital.
15851 Set background opacity.
15854 @section weave, doubleweave
15856 The @code{weave} takes a field-based video input and join
15857 each two sequential fields into single frame, producing a new double
15858 height clip with half the frame rate and half the frame count.
15860 The @code{doubleweave} works same as @code{weave} but without
15861 halving frame rate and frame count.
15863 It accepts the following option:
15867 Set first field. Available values are:
15871 Set the frame as top-field-first.
15874 Set the frame as bottom-field-first.
15878 @subsection Examples
15882 Interlace video using @ref{select} and @ref{separatefields} filter:
15884 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
15889 Apply the xBR high-quality magnification filter which is designed for pixel
15890 art. It follows a set of edge-detection rules, see
15891 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
15893 It accepts the following option:
15897 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
15898 @code{3xBR} and @code{4} for @code{4xBR}.
15899 Default is @code{3}.
15905 Deinterlace the input video ("yadif" means "yet another deinterlacing
15908 It accepts the following parameters:
15914 The interlacing mode to adopt. It accepts one of the following values:
15917 @item 0, send_frame
15918 Output one frame for each frame.
15919 @item 1, send_field
15920 Output one frame for each field.
15921 @item 2, send_frame_nospatial
15922 Like @code{send_frame}, but it skips the spatial interlacing check.
15923 @item 3, send_field_nospatial
15924 Like @code{send_field}, but it skips the spatial interlacing check.
15927 The default value is @code{send_frame}.
15930 The picture field parity assumed for the input interlaced video. It accepts one
15931 of the following values:
15935 Assume the top field is first.
15937 Assume the bottom field is first.
15939 Enable automatic detection of field parity.
15942 The default value is @code{auto}.
15943 If the interlacing is unknown or the decoder does not export this information,
15944 top field first will be assumed.
15947 Specify which frames to deinterlace. Accept one of the following
15952 Deinterlace all frames.
15953 @item 1, interlaced
15954 Only deinterlace frames marked as interlaced.
15957 The default value is @code{all}.
15962 Apply Zoom & Pan effect.
15964 This filter accepts the following options:
15968 Set the zoom expression. Default is 1.
15972 Set the x and y expression. Default is 0.
15975 Set the duration expression in number of frames.
15976 This sets for how many number of frames effect will last for
15977 single input image.
15980 Set the output image size, default is 'hd720'.
15983 Set the output frame rate, default is '25'.
15986 Each expression can contain the following constants:
16005 Output frame count.
16009 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16010 for current input frame.
16014 'x' and 'y' of last output frame of previous input frame or 0 when there was
16015 not yet such frame (first input frame).
16018 Last calculated zoom from 'z' expression for current input frame.
16021 Last calculated zoom of last output frame of previous input frame.
16024 Number of output frames for current input frame. Calculated from 'd' expression
16025 for each input frame.
16028 number of output frames created for previous input frame
16031 Rational number: input width / input height
16034 sample aspect ratio
16037 display aspect ratio
16041 @subsection Examples
16045 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16047 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
16051 Zoom-in up to 1.5 and pan always at center of picture:
16053 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16057 Same as above but without pausing:
16059 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16065 Scale (resize) the input video, using the z.lib library:
16066 https://github.com/sekrit-twc/zimg.
16068 The zscale filter forces the output display aspect ratio to be the same
16069 as the input, by changing the output sample aspect ratio.
16071 If the input image format is different from the format requested by
16072 the next filter, the zscale filter will convert the input to the
16075 @subsection Options
16076 The filter accepts the following options.
16081 Set the output video dimension expression. Default value is the input
16084 If the @var{width} or @var{w} value is 0, the input width is used for
16085 the output. If the @var{height} or @var{h} value is 0, the input height
16086 is used for the output.
16088 If one and only one of the values is -n with n >= 1, the zscale filter
16089 will use a value that maintains the aspect ratio of the input image,
16090 calculated from the other specified dimension. After that it will,
16091 however, make sure that the calculated dimension is divisible by n and
16092 adjust the value if necessary.
16094 If both values are -n with n >= 1, the behavior will be identical to
16095 both values being set to 0 as previously detailed.
16097 See below for the list of accepted constants for use in the dimension
16101 Set the video size. For the syntax of this option, check the
16102 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16105 Set the dither type.
16107 Possible values are:
16112 @item error_diffusion
16118 Set the resize filter type.
16120 Possible values are:
16130 Default is bilinear.
16133 Set the color range.
16135 Possible values are:
16142 Default is same as input.
16145 Set the color primaries.
16147 Possible values are:
16157 Default is same as input.
16160 Set the transfer characteristics.
16162 Possible values are:
16176 Default is same as input.
16179 Set the colorspace matrix.
16181 Possible value are:
16192 Default is same as input.
16195 Set the input color range.
16197 Possible values are:
16204 Default is same as input.
16206 @item primariesin, pin
16207 Set the input color primaries.
16209 Possible values are:
16219 Default is same as input.
16221 @item transferin, tin
16222 Set the input transfer characteristics.
16224 Possible values are:
16235 Default is same as input.
16237 @item matrixin, min
16238 Set the input colorspace matrix.
16240 Possible value are:
16252 Set the output chroma location.
16254 Possible values are:
16265 @item chromalin, cin
16266 Set the input chroma location.
16268 Possible values are:
16280 Set the nominal peak luminance.
16283 The values of the @option{w} and @option{h} options are expressions
16284 containing the following constants:
16289 The input width and height
16293 These are the same as @var{in_w} and @var{in_h}.
16297 The output (scaled) width and height
16301 These are the same as @var{out_w} and @var{out_h}
16304 The same as @var{iw} / @var{ih}
16307 input sample aspect ratio
16310 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16314 horizontal and vertical input chroma subsample values. For example for the
16315 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16319 horizontal and vertical output chroma subsample values. For example for the
16320 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16326 @c man end VIDEO FILTERS
16328 @chapter Video Sources
16329 @c man begin VIDEO SOURCES
16331 Below is a description of the currently available video sources.
16335 Buffer video frames, and make them available to the filter chain.
16337 This source is mainly intended for a programmatic use, in particular
16338 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
16340 It accepts the following parameters:
16345 Specify the size (width and height) of the buffered video frames. For the
16346 syntax of this option, check the
16347 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16350 The input video width.
16353 The input video height.
16356 A string representing the pixel format of the buffered video frames.
16357 It may be a number corresponding to a pixel format, or a pixel format
16361 Specify the timebase assumed by the timestamps of the buffered frames.
16364 Specify the frame rate expected for the video stream.
16366 @item pixel_aspect, sar
16367 The sample (pixel) aspect ratio of the input video.
16370 Specify the optional parameters to be used for the scale filter which
16371 is automatically inserted when an input change is detected in the
16372 input size or format.
16374 @item hw_frames_ctx
16375 When using a hardware pixel format, this should be a reference to an
16376 AVHWFramesContext describing input frames.
16381 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
16384 will instruct the source to accept video frames with size 320x240 and
16385 with format "yuv410p", assuming 1/24 as the timestamps timebase and
16386 square pixels (1:1 sample aspect ratio).
16387 Since the pixel format with name "yuv410p" corresponds to the number 6
16388 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
16389 this example corresponds to:
16391 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
16394 Alternatively, the options can be specified as a flat string, but this
16395 syntax is deprecated:
16397 @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}]
16401 Create a pattern generated by an elementary cellular automaton.
16403 The initial state of the cellular automaton can be defined through the
16404 @option{filename} and @option{pattern} options. If such options are
16405 not specified an initial state is created randomly.
16407 At each new frame a new row in the video is filled with the result of
16408 the cellular automaton next generation. The behavior when the whole
16409 frame is filled is defined by the @option{scroll} option.
16411 This source accepts the following options:
16415 Read the initial cellular automaton state, i.e. the starting row, from
16416 the specified file.
16417 In the file, each non-whitespace character is considered an alive
16418 cell, a newline will terminate the row, and further characters in the
16419 file will be ignored.
16422 Read the initial cellular automaton state, i.e. the starting row, from
16423 the specified string.
16425 Each non-whitespace character in the string is considered an alive
16426 cell, a newline will terminate the row, and further characters in the
16427 string will be ignored.
16430 Set the video rate, that is the number of frames generated per second.
16433 @item random_fill_ratio, ratio
16434 Set the random fill ratio for the initial cellular automaton row. It
16435 is a floating point number value ranging from 0 to 1, defaults to
16438 This option is ignored when a file or a pattern is specified.
16440 @item random_seed, seed
16441 Set the seed for filling randomly the initial row, must be an integer
16442 included between 0 and UINT32_MAX. If not specified, or if explicitly
16443 set to -1, the filter will try to use a good random seed on a best
16447 Set the cellular automaton rule, it is a number ranging from 0 to 255.
16448 Default value is 110.
16451 Set the size of the output video. For the syntax of this option, check the
16452 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16454 If @option{filename} or @option{pattern} is specified, the size is set
16455 by default to the width of the specified initial state row, and the
16456 height is set to @var{width} * PHI.
16458 If @option{size} is set, it must contain the width of the specified
16459 pattern string, and the specified pattern will be centered in the
16462 If a filename or a pattern string is not specified, the size value
16463 defaults to "320x518" (used for a randomly generated initial state).
16466 If set to 1, scroll the output upward when all the rows in the output
16467 have been already filled. If set to 0, the new generated row will be
16468 written over the top row just after the bottom row is filled.
16471 @item start_full, full
16472 If set to 1, completely fill the output with generated rows before
16473 outputting the first frame.
16474 This is the default behavior, for disabling set the value to 0.
16477 If set to 1, stitch the left and right row edges together.
16478 This is the default behavior, for disabling set the value to 0.
16481 @subsection Examples
16485 Read the initial state from @file{pattern}, and specify an output of
16488 cellauto=f=pattern:s=200x400
16492 Generate a random initial row with a width of 200 cells, with a fill
16495 cellauto=ratio=2/3:s=200x200
16499 Create a pattern generated by rule 18 starting by a single alive cell
16500 centered on an initial row with width 100:
16502 cellauto=p=@@:s=100x400:full=0:rule=18
16506 Specify a more elaborated initial pattern:
16508 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
16513 @anchor{coreimagesrc}
16514 @section coreimagesrc
16515 Video source generated on GPU using Apple's CoreImage API on OSX.
16517 This video source is a specialized version of the @ref{coreimage} video filter.
16518 Use a core image generator at the beginning of the applied filterchain to
16519 generate the content.
16521 The coreimagesrc video source accepts the following options:
16523 @item list_generators
16524 List all available generators along with all their respective options as well as
16525 possible minimum and maximum values along with the default values.
16527 list_generators=true
16531 Specify the size of the sourced video. For the syntax of this option, check the
16532 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16533 The default value is @code{320x240}.
16536 Specify the frame rate of the sourced video, as the number of frames
16537 generated per second. It has to be a string in the format
16538 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16539 number or a valid video frame rate abbreviation. The default value is
16543 Set the sample aspect ratio of the sourced video.
16546 Set the duration of the sourced video. See
16547 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16548 for the accepted syntax.
16550 If not specified, or the expressed duration is negative, the video is
16551 supposed to be generated forever.
16554 Additionally, all options of the @ref{coreimage} video filter are accepted.
16555 A complete filterchain can be used for further processing of the
16556 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
16557 and examples for details.
16559 @subsection Examples
16564 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
16565 given as complete and escaped command-line for Apple's standard bash shell:
16567 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
16569 This example is equivalent to the QRCode example of @ref{coreimage} without the
16570 need for a nullsrc video source.
16574 @section mandelbrot
16576 Generate a Mandelbrot set fractal, and progressively zoom towards the
16577 point specified with @var{start_x} and @var{start_y}.
16579 This source accepts the following options:
16584 Set the terminal pts value. Default value is 400.
16587 Set the terminal scale value.
16588 Must be a floating point value. Default value is 0.3.
16591 Set the inner coloring mode, that is the algorithm used to draw the
16592 Mandelbrot fractal internal region.
16594 It shall assume one of the following values:
16599 Show time until convergence.
16601 Set color based on point closest to the origin of the iterations.
16606 Default value is @var{mincol}.
16609 Set the bailout value. Default value is 10.0.
16612 Set the maximum of iterations performed by the rendering
16613 algorithm. Default value is 7189.
16616 Set outer coloring mode.
16617 It shall assume one of following values:
16619 @item iteration_count
16620 Set iteration cound mode.
16621 @item normalized_iteration_count
16622 set normalized iteration count mode.
16624 Default value is @var{normalized_iteration_count}.
16627 Set frame rate, expressed as number of frames per second. Default
16631 Set frame size. For the syntax of this option, check the "Video
16632 size" section in the ffmpeg-utils manual. Default value is "640x480".
16635 Set the initial scale value. Default value is 3.0.
16638 Set the initial x position. Must be a floating point value between
16639 -100 and 100. Default value is -0.743643887037158704752191506114774.
16642 Set the initial y position. Must be a floating point value between
16643 -100 and 100. Default value is -0.131825904205311970493132056385139.
16648 Generate various test patterns, as generated by the MPlayer test filter.
16650 The size of the generated video is fixed, and is 256x256.
16651 This source is useful in particular for testing encoding features.
16653 This source accepts the following options:
16658 Specify the frame rate of the sourced video, as the number of frames
16659 generated per second. It has to be a string in the format
16660 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16661 number or a valid video frame rate abbreviation. The default value is
16665 Set the duration of the sourced video. See
16666 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16667 for the accepted syntax.
16669 If not specified, or the expressed duration is negative, the video is
16670 supposed to be generated forever.
16674 Set the number or the name of the test to perform. Supported tests are:
16690 Default value is "all", which will cycle through the list of all tests.
16695 mptestsrc=t=dc_luma
16698 will generate a "dc_luma" test pattern.
16700 @section frei0r_src
16702 Provide a frei0r source.
16704 To enable compilation of this filter you need to install the frei0r
16705 header and configure FFmpeg with @code{--enable-frei0r}.
16707 This source accepts the following parameters:
16712 The size of the video to generate. For the syntax of this option, check the
16713 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16716 The framerate of the generated video. It may be a string of the form
16717 @var{num}/@var{den} or a frame rate abbreviation.
16720 The name to the frei0r source to load. For more information regarding frei0r and
16721 how to set the parameters, read the @ref{frei0r} section in the video filters
16724 @item filter_params
16725 A '|'-separated list of parameters to pass to the frei0r source.
16729 For example, to generate a frei0r partik0l source with size 200x200
16730 and frame rate 10 which is overlaid on the overlay filter main input:
16732 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
16737 Generate a life pattern.
16739 This source is based on a generalization of John Conway's life game.
16741 The sourced input represents a life grid, each pixel represents a cell
16742 which can be in one of two possible states, alive or dead. Every cell
16743 interacts with its eight neighbours, which are the cells that are
16744 horizontally, vertically, or diagonally adjacent.
16746 At each interaction the grid evolves according to the adopted rule,
16747 which specifies the number of neighbor alive cells which will make a
16748 cell stay alive or born. The @option{rule} option allows one to specify
16751 This source accepts the following options:
16755 Set the file from which to read the initial grid state. In the file,
16756 each non-whitespace character is considered an alive cell, and newline
16757 is used to delimit the end of each row.
16759 If this option is not specified, the initial grid is generated
16763 Set the video rate, that is the number of frames generated per second.
16766 @item random_fill_ratio, ratio
16767 Set the random fill ratio for the initial random grid. It is a
16768 floating point number value ranging from 0 to 1, defaults to 1/PHI.
16769 It is ignored when a file is specified.
16771 @item random_seed, seed
16772 Set the seed for filling the initial random grid, must be an integer
16773 included between 0 and UINT32_MAX. If not specified, or if explicitly
16774 set to -1, the filter will try to use a good random seed on a best
16780 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
16781 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
16782 @var{NS} specifies the number of alive neighbor cells which make a
16783 live cell stay alive, and @var{NB} the number of alive neighbor cells
16784 which make a dead cell to become alive (i.e. to "born").
16785 "s" and "b" can be used in place of "S" and "B", respectively.
16787 Alternatively a rule can be specified by an 18-bits integer. The 9
16788 high order bits are used to encode the next cell state if it is alive
16789 for each number of neighbor alive cells, the low order bits specify
16790 the rule for "borning" new cells. Higher order bits encode for an
16791 higher number of neighbor cells.
16792 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
16793 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
16795 Default value is "S23/B3", which is the original Conway's game of life
16796 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
16797 cells, and will born a new cell if there are three alive cells around
16801 Set the size of the output video. For the syntax of this option, check the
16802 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16804 If @option{filename} is specified, the size is set by default to the
16805 same size of the input file. If @option{size} is set, it must contain
16806 the size specified in the input file, and the initial grid defined in
16807 that file is centered in the larger resulting area.
16809 If a filename is not specified, the size value defaults to "320x240"
16810 (used for a randomly generated initial grid).
16813 If set to 1, stitch the left and right grid edges together, and the
16814 top and bottom edges also. Defaults to 1.
16817 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
16818 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
16819 value from 0 to 255.
16822 Set the color of living (or new born) cells.
16825 Set the color of dead cells. If @option{mold} is set, this is the first color
16826 used to represent a dead cell.
16829 Set mold color, for definitely dead and moldy cells.
16831 For the syntax of these 3 color options, check the "Color" section in the
16832 ffmpeg-utils manual.
16835 @subsection Examples
16839 Read a grid from @file{pattern}, and center it on a grid of size
16842 life=f=pattern:s=300x300
16846 Generate a random grid of size 200x200, with a fill ratio of 2/3:
16848 life=ratio=2/3:s=200x200
16852 Specify a custom rule for evolving a randomly generated grid:
16858 Full example with slow death effect (mold) using @command{ffplay}:
16860 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
16867 @anchor{haldclutsrc}
16869 @anchor{rgbtestsrc}
16871 @anchor{smptehdbars}
16874 @anchor{yuvtestsrc}
16875 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
16877 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
16879 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
16881 The @code{color} source provides an uniformly colored input.
16883 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
16884 @ref{haldclut} filter.
16886 The @code{nullsrc} source returns unprocessed video frames. It is
16887 mainly useful to be employed in analysis / debugging tools, or as the
16888 source for filters which ignore the input data.
16890 The @code{rgbtestsrc} source generates an RGB test pattern useful for
16891 detecting RGB vs BGR issues. You should see a red, green and blue
16892 stripe from top to bottom.
16894 The @code{smptebars} source generates a color bars pattern, based on
16895 the SMPTE Engineering Guideline EG 1-1990.
16897 The @code{smptehdbars} source generates a color bars pattern, based on
16898 the SMPTE RP 219-2002.
16900 The @code{testsrc} source generates a test video pattern, showing a
16901 color pattern, a scrolling gradient and a timestamp. This is mainly
16902 intended for testing purposes.
16904 The @code{testsrc2} source is similar to testsrc, but supports more
16905 pixel formats instead of just @code{rgb24}. This allows using it as an
16906 input for other tests without requiring a format conversion.
16908 The @code{yuvtestsrc} source generates an YUV test pattern. You should
16909 see a y, cb and cr stripe from top to bottom.
16911 The sources accept the following parameters:
16916 Specify the alpha (opacity) of the background, only available in the
16917 @code{testsrc2} source. The value must be between 0 (fully transparent) and
16918 255 (fully opaque, the default).
16921 Specify the color of the source, only available in the @code{color}
16922 source. For the syntax of this option, check the "Color" section in the
16923 ffmpeg-utils manual.
16926 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
16927 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
16928 pixels to be used as identity matrix for 3D lookup tables. Each component is
16929 coded on a @code{1/(N*N)} scale.
16932 Specify the size of the sourced video. For the syntax of this option, check the
16933 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16934 The default value is @code{320x240}.
16936 This option is not available with the @code{haldclutsrc} filter.
16939 Specify the frame rate of the sourced video, as the number of frames
16940 generated per second. It has to be a string in the format
16941 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16942 number or a valid video frame rate abbreviation. The default value is
16946 Set the sample aspect ratio of the sourced video.
16949 Set the duration of the sourced video. See
16950 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16951 for the accepted syntax.
16953 If not specified, or the expressed duration is negative, the video is
16954 supposed to be generated forever.
16957 Set the number of decimals to show in the timestamp, only available in the
16958 @code{testsrc} source.
16960 The displayed timestamp value will correspond to the original
16961 timestamp value multiplied by the power of 10 of the specified
16962 value. Default value is 0.
16965 For example the following:
16967 testsrc=duration=5.3:size=qcif:rate=10
16970 will generate a video with a duration of 5.3 seconds, with size
16971 176x144 and a frame rate of 10 frames per second.
16973 The following graph description will generate a red source
16974 with an opacity of 0.2, with size "qcif" and a frame rate of 10
16977 color=c=red@@0.2:s=qcif:r=10
16980 If the input content is to be ignored, @code{nullsrc} can be used. The
16981 following command generates noise in the luminance plane by employing
16982 the @code{geq} filter:
16984 nullsrc=s=256x256, geq=random(1)*255:128:128
16987 @subsection Commands
16989 The @code{color} source supports the following commands:
16993 Set the color of the created image. Accepts the same syntax of the
16994 corresponding @option{color} option.
16997 @c man end VIDEO SOURCES
16999 @chapter Video Sinks
17000 @c man begin VIDEO SINKS
17002 Below is a description of the currently available video sinks.
17004 @section buffersink
17006 Buffer video frames, and make them available to the end of the filter
17009 This sink is mainly intended for programmatic use, in particular
17010 through the interface defined in @file{libavfilter/buffersink.h}
17011 or the options system.
17013 It accepts a pointer to an AVBufferSinkContext structure, which
17014 defines the incoming buffers' formats, to be passed as the opaque
17015 parameter to @code{avfilter_init_filter} for initialization.
17019 Null video sink: do absolutely nothing with the input video. It is
17020 mainly useful as a template and for use in analysis / debugging
17023 @c man end VIDEO SINKS
17025 @chapter Multimedia Filters
17026 @c man begin MULTIMEDIA FILTERS
17028 Below is a description of the currently available multimedia filters.
17032 Convert input audio to a video output, displaying the audio bit scope.
17034 The filter accepts the following options:
17038 Set frame rate, expressed as number of frames per second. Default
17042 Specify the video size for the output. For the syntax of this option, check the
17043 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17044 Default value is @code{1024x256}.
17047 Specify list of colors separated by space or by '|' which will be used to
17048 draw channels. Unrecognized or missing colors will be replaced
17052 @section ahistogram
17054 Convert input audio to a video output, displaying the volume histogram.
17056 The filter accepts the following options:
17060 Specify how histogram is calculated.
17062 It accepts the following values:
17065 Use single histogram for all channels.
17067 Use separate histogram for each channel.
17069 Default is @code{single}.
17072 Set frame rate, expressed as number of frames per second. Default
17076 Specify the video size for the output. For the syntax of this option, check the
17077 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17078 Default value is @code{hd720}.
17083 It accepts the following values:
17094 reverse logarithmic
17096 Default is @code{log}.
17099 Set amplitude scale.
17101 It accepts the following values:
17108 Default is @code{log}.
17111 Set how much frames to accumulate in histogram.
17112 Defauls is 1. Setting this to -1 accumulates all frames.
17115 Set histogram ratio of window height.
17118 Set sonogram sliding.
17120 It accepts the following values:
17123 replace old rows with new ones.
17125 scroll from top to bottom.
17127 Default is @code{replace}.
17130 @section aphasemeter
17132 Convert input audio to a video output, displaying the audio phase.
17134 The filter accepts the following options:
17138 Set the output frame rate. Default value is @code{25}.
17141 Set the video size for the output. For the syntax of this option, check the
17142 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17143 Default value is @code{800x400}.
17148 Specify the red, green, blue contrast. Default values are @code{2},
17149 @code{7} and @code{1}.
17150 Allowed range is @code{[0, 255]}.
17153 Set color which will be used for drawing median phase. If color is
17154 @code{none} which is default, no median phase value will be drawn.
17157 Enable video output. Default is enabled.
17160 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17161 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17162 The @code{-1} means left and right channels are completely out of phase and
17163 @code{1} means channels are in phase.
17165 @section avectorscope
17167 Convert input audio to a video output, representing the audio vector
17170 The filter is used to measure the difference between channels of stereo
17171 audio stream. A monoaural signal, consisting of identical left and right
17172 signal, results in straight vertical line. Any stereo separation is visible
17173 as a deviation from this line, creating a Lissajous figure.
17174 If the straight (or deviation from it) but horizontal line appears this
17175 indicates that the left and right channels are out of phase.
17177 The filter accepts the following options:
17181 Set the vectorscope mode.
17183 Available values are:
17186 Lissajous rotated by 45 degrees.
17189 Same as above but not rotated.
17192 Shape resembling half of circle.
17195 Default value is @samp{lissajous}.
17198 Set the video size for the output. For the syntax of this option, check the
17199 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17200 Default value is @code{400x400}.
17203 Set the output frame rate. Default value is @code{25}.
17209 Specify the red, green, blue and alpha contrast. Default values are @code{40},
17210 @code{160}, @code{80} and @code{255}.
17211 Allowed range is @code{[0, 255]}.
17217 Specify the red, green, blue and alpha fade. Default values are @code{15},
17218 @code{10}, @code{5} and @code{5}.
17219 Allowed range is @code{[0, 255]}.
17222 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
17223 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
17226 Set the vectorscope drawing mode.
17228 Available values are:
17231 Draw dot for each sample.
17234 Draw line between previous and current sample.
17237 Default value is @samp{dot}.
17240 Specify amplitude scale of audio samples.
17242 Available values are:
17258 Swap left channel axis with right channel axis.
17268 Mirror only x axis.
17271 Mirror only y axis.
17279 @subsection Examples
17283 Complete example using @command{ffplay}:
17285 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17286 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
17290 @section bench, abench
17292 Benchmark part of a filtergraph.
17294 The filter accepts the following options:
17298 Start or stop a timer.
17300 Available values are:
17303 Get the current time, set it as frame metadata (using the key
17304 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
17307 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
17308 the input frame metadata to get the time difference. Time difference, average,
17309 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
17310 @code{min}) are then printed. The timestamps are expressed in seconds.
17314 @subsection Examples
17318 Benchmark @ref{selectivecolor} filter:
17320 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
17326 Concatenate audio and video streams, joining them together one after the
17329 The filter works on segments of synchronized video and audio streams. All
17330 segments must have the same number of streams of each type, and that will
17331 also be the number of streams at output.
17333 The filter accepts the following options:
17338 Set the number of segments. Default is 2.
17341 Set the number of output video streams, that is also the number of video
17342 streams in each segment. Default is 1.
17345 Set the number of output audio streams, that is also the number of audio
17346 streams in each segment. Default is 0.
17349 Activate unsafe mode: do not fail if segments have a different format.
17353 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
17354 @var{a} audio outputs.
17356 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
17357 segment, in the same order as the outputs, then the inputs for the second
17360 Related streams do not always have exactly the same duration, for various
17361 reasons including codec frame size or sloppy authoring. For that reason,
17362 related synchronized streams (e.g. a video and its audio track) should be
17363 concatenated at once. The concat filter will use the duration of the longest
17364 stream in each segment (except the last one), and if necessary pad shorter
17365 audio streams with silence.
17367 For this filter to work correctly, all segments must start at timestamp 0.
17369 All corresponding streams must have the same parameters in all segments; the
17370 filtering system will automatically select a common pixel format for video
17371 streams, and a common sample format, sample rate and channel layout for
17372 audio streams, but other settings, such as resolution, must be converted
17373 explicitly by the user.
17375 Different frame rates are acceptable but will result in variable frame rate
17376 at output; be sure to configure the output file to handle it.
17378 @subsection Examples
17382 Concatenate an opening, an episode and an ending, all in bilingual version
17383 (video in stream 0, audio in streams 1 and 2):
17385 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
17386 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
17387 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
17388 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
17392 Concatenate two parts, handling audio and video separately, using the
17393 (a)movie sources, and adjusting the resolution:
17395 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
17396 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
17397 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
17399 Note that a desync will happen at the stitch if the audio and video streams
17400 do not have exactly the same duration in the first file.
17404 @section drawgraph, adrawgraph
17406 Draw a graph using input video or audio metadata.
17408 It accepts the following parameters:
17412 Set 1st frame metadata key from which metadata values will be used to draw a graph.
17415 Set 1st foreground color expression.
17418 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
17421 Set 2nd foreground color expression.
17424 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
17427 Set 3rd foreground color expression.
17430 Set 4th frame metadata key from which metadata values will be used to draw a graph.
17433 Set 4th foreground color expression.
17436 Set minimal value of metadata value.
17439 Set maximal value of metadata value.
17442 Set graph background color. Default is white.
17447 Available values for mode is:
17454 Default is @code{line}.
17459 Available values for slide is:
17462 Draw new frame when right border is reached.
17465 Replace old columns with new ones.
17468 Scroll from right to left.
17471 Scroll from left to right.
17474 Draw single picture.
17477 Default is @code{frame}.
17480 Set size of graph video. For the syntax of this option, check the
17481 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17482 The default value is @code{900x256}.
17484 The foreground color expressions can use the following variables:
17487 Minimal value of metadata value.
17490 Maximal value of metadata value.
17493 Current metadata key value.
17496 The color is defined as 0xAABBGGRR.
17499 Example using metadata from @ref{signalstats} filter:
17501 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
17504 Example using metadata from @ref{ebur128} filter:
17506 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
17512 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
17513 it unchanged. By default, it logs a message at a frequency of 10Hz with the
17514 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
17515 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
17517 The filter also has a video output (see the @var{video} option) with a real
17518 time graph to observe the loudness evolution. The graphic contains the logged
17519 message mentioned above, so it is not printed anymore when this option is set,
17520 unless the verbose logging is set. The main graphing area contains the
17521 short-term loudness (3 seconds of analysis), and the gauge on the right is for
17522 the momentary loudness (400 milliseconds).
17524 More information about the Loudness Recommendation EBU R128 on
17525 @url{http://tech.ebu.ch/loudness}.
17527 The filter accepts the following options:
17532 Activate the video output. The audio stream is passed unchanged whether this
17533 option is set or no. The video stream will be the first output stream if
17534 activated. Default is @code{0}.
17537 Set the video size. This option is for video only. For the syntax of this
17539 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17540 Default and minimum resolution is @code{640x480}.
17543 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
17544 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
17545 other integer value between this range is allowed.
17548 Set metadata injection. If set to @code{1}, the audio input will be segmented
17549 into 100ms output frames, each of them containing various loudness information
17550 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
17552 Default is @code{0}.
17555 Force the frame logging level.
17557 Available values are:
17560 information logging level
17562 verbose logging level
17565 By default, the logging level is set to @var{info}. If the @option{video} or
17566 the @option{metadata} options are set, it switches to @var{verbose}.
17571 Available modes can be cumulated (the option is a @code{flag} type). Possible
17575 Disable any peak mode (default).
17577 Enable sample-peak mode.
17579 Simple peak mode looking for the higher sample value. It logs a message
17580 for sample-peak (identified by @code{SPK}).
17582 Enable true-peak mode.
17584 If enabled, the peak lookup is done on an over-sampled version of the input
17585 stream for better peak accuracy. It logs a message for true-peak.
17586 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
17587 This mode requires a build with @code{libswresample}.
17591 Treat mono input files as "dual mono". If a mono file is intended for playback
17592 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
17593 If set to @code{true}, this option will compensate for this effect.
17594 Multi-channel input files are not affected by this option.
17597 Set a specific pan law to be used for the measurement of dual mono files.
17598 This parameter is optional, and has a default value of -3.01dB.
17601 @subsection Examples
17605 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
17607 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
17611 Run an analysis with @command{ffmpeg}:
17613 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
17617 @section interleave, ainterleave
17619 Temporally interleave frames from several inputs.
17621 @code{interleave} works with video inputs, @code{ainterleave} with audio.
17623 These filters read frames from several inputs and send the oldest
17624 queued frame to the output.
17626 Input streams must have well defined, monotonically increasing frame
17629 In order to submit one frame to output, these filters need to enqueue
17630 at least one frame for each input, so they cannot work in case one
17631 input is not yet terminated and will not receive incoming frames.
17633 For example consider the case when one input is a @code{select} filter
17634 which always drops input frames. The @code{interleave} filter will keep
17635 reading from that input, but it will never be able to send new frames
17636 to output until the input sends an end-of-stream signal.
17638 Also, depending on inputs synchronization, the filters will drop
17639 frames in case one input receives more frames than the other ones, and
17640 the queue is already filled.
17642 These filters accept the following options:
17646 Set the number of different inputs, it is 2 by default.
17649 @subsection Examples
17653 Interleave frames belonging to different streams using @command{ffmpeg}:
17655 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
17659 Add flickering blur effect:
17661 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
17665 @section metadata, ametadata
17667 Manipulate frame metadata.
17669 This filter accepts the following options:
17673 Set mode of operation of the filter.
17675 Can be one of the following:
17679 If both @code{value} and @code{key} is set, select frames
17680 which have such metadata. If only @code{key} is set, select
17681 every frame that has such key in metadata.
17684 Add new metadata @code{key} and @code{value}. If key is already available
17688 Modify value of already present key.
17691 If @code{value} is set, delete only keys that have such value.
17692 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
17696 Print key and its value if metadata was found. If @code{key} is not set print all
17697 metadata values available in frame.
17701 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
17704 Set metadata value which will be used. This option is mandatory for
17705 @code{modify} and @code{add} mode.
17708 Which function to use when comparing metadata value and @code{value}.
17710 Can be one of following:
17714 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
17717 Values are interpreted as strings, returns true if metadata value starts with
17718 the @code{value} option string.
17721 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
17724 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
17727 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
17730 Values are interpreted as floats, returns true if expression from option @code{expr}
17735 Set expression which is used when @code{function} is set to @code{expr}.
17736 The expression is evaluated through the eval API and can contain the following
17741 Float representation of @code{value} from metadata key.
17744 Float representation of @code{value} as supplied by user in @code{value} option.
17748 If specified in @code{print} mode, output is written to the named file. Instead of
17749 plain filename any writable url can be specified. Filename ``-'' is a shorthand
17750 for standard output. If @code{file} option is not set, output is written to the log
17751 with AV_LOG_INFO loglevel.
17755 @subsection Examples
17759 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
17762 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
17765 Print silencedetect output to file @file{metadata.txt}.
17767 silencedetect,ametadata=mode=print:file=metadata.txt
17770 Direct all metadata to a pipe with file descriptor 4.
17772 metadata=mode=print:file='pipe\:4'
17776 @section perms, aperms
17778 Set read/write permissions for the output frames.
17780 These filters are mainly aimed at developers to test direct path in the
17781 following filter in the filtergraph.
17783 The filters accept the following options:
17787 Select the permissions mode.
17789 It accepts the following values:
17792 Do nothing. This is the default.
17794 Set all the output frames read-only.
17796 Set all the output frames directly writable.
17798 Make the frame read-only if writable, and writable if read-only.
17800 Set each output frame read-only or writable randomly.
17804 Set the seed for the @var{random} mode, must be an integer included between
17805 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
17806 @code{-1}, the filter will try to use a good random seed on a best effort
17810 Note: in case of auto-inserted filter between the permission filter and the
17811 following one, the permission might not be received as expected in that
17812 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
17813 perms/aperms filter can avoid this problem.
17815 @section realtime, arealtime
17817 Slow down filtering to match real time approximately.
17819 These filters will pause the filtering for a variable amount of time to
17820 match the output rate with the input timestamps.
17821 They are similar to the @option{re} option to @code{ffmpeg}.
17823 They accept the following options:
17827 Time limit for the pauses. Any pause longer than that will be considered
17828 a timestamp discontinuity and reset the timer. Default is 2 seconds.
17832 @section select, aselect
17834 Select frames to pass in output.
17836 This filter accepts the following options:
17841 Set expression, which is evaluated for each input frame.
17843 If the expression is evaluated to zero, the frame is discarded.
17845 If the evaluation result is negative or NaN, the frame is sent to the
17846 first output; otherwise it is sent to the output with index
17847 @code{ceil(val)-1}, assuming that the input index starts from 0.
17849 For example a value of @code{1.2} corresponds to the output with index
17850 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
17853 Set the number of outputs. The output to which to send the selected
17854 frame is based on the result of the evaluation. Default value is 1.
17857 The expression can contain the following constants:
17861 The (sequential) number of the filtered frame, starting from 0.
17864 The (sequential) number of the selected frame, starting from 0.
17866 @item prev_selected_n
17867 The sequential number of the last selected frame. It's NAN if undefined.
17870 The timebase of the input timestamps.
17873 The PTS (Presentation TimeStamp) of the filtered video frame,
17874 expressed in @var{TB} units. It's NAN if undefined.
17877 The PTS of the filtered video frame,
17878 expressed in seconds. It's NAN if undefined.
17881 The PTS of the previously filtered video frame. It's NAN if undefined.
17883 @item prev_selected_pts
17884 The PTS of the last previously filtered video frame. It's NAN if undefined.
17886 @item prev_selected_t
17887 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
17890 The PTS of the first video frame in the video. It's NAN if undefined.
17893 The time of the first video frame in the video. It's NAN if undefined.
17895 @item pict_type @emph{(video only)}
17896 The type of the filtered frame. It can assume one of the following
17908 @item interlace_type @emph{(video only)}
17909 The frame interlace type. It can assume one of the following values:
17912 The frame is progressive (not interlaced).
17914 The frame is top-field-first.
17916 The frame is bottom-field-first.
17919 @item consumed_sample_n @emph{(audio only)}
17920 the number of selected samples before the current frame
17922 @item samples_n @emph{(audio only)}
17923 the number of samples in the current frame
17925 @item sample_rate @emph{(audio only)}
17926 the input sample rate
17929 This is 1 if the filtered frame is a key-frame, 0 otherwise.
17932 the position in the file of the filtered frame, -1 if the information
17933 is not available (e.g. for synthetic video)
17935 @item scene @emph{(video only)}
17936 value between 0 and 1 to indicate a new scene; a low value reflects a low
17937 probability for the current frame to introduce a new scene, while a higher
17938 value means the current frame is more likely to be one (see the example below)
17940 @item concatdec_select
17941 The concat demuxer can select only part of a concat input file by setting an
17942 inpoint and an outpoint, but the output packets may not be entirely contained
17943 in the selected interval. By using this variable, it is possible to skip frames
17944 generated by the concat demuxer which are not exactly contained in the selected
17947 This works by comparing the frame pts against the @var{lavf.concat.start_time}
17948 and the @var{lavf.concat.duration} packet metadata values which are also
17949 present in the decoded frames.
17951 The @var{concatdec_select} variable is -1 if the frame pts is at least
17952 start_time and either the duration metadata is missing or the frame pts is less
17953 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
17956 That basically means that an input frame is selected if its pts is within the
17957 interval set by the concat demuxer.
17961 The default value of the select expression is "1".
17963 @subsection Examples
17967 Select all frames in input:
17972 The example above is the same as:
17984 Select only I-frames:
17986 select='eq(pict_type\,I)'
17990 Select one frame every 100:
17992 select='not(mod(n\,100))'
17996 Select only frames contained in the 10-20 time interval:
17998 select=between(t\,10\,20)
18002 Select only I-frames contained in the 10-20 time interval:
18004 select=between(t\,10\,20)*eq(pict_type\,I)
18008 Select frames with a minimum distance of 10 seconds:
18010 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18014 Use aselect to select only audio frames with samples number > 100:
18016 aselect='gt(samples_n\,100)'
18020 Create a mosaic of the first scenes:
18022 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18025 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18029 Send even and odd frames to separate outputs, and compose them:
18031 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18035 Select useful frames from an ffconcat file which is using inpoints and
18036 outpoints but where the source files are not intra frame only.
18038 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18042 @section sendcmd, asendcmd
18044 Send commands to filters in the filtergraph.
18046 These filters read commands to be sent to other filters in the
18049 @code{sendcmd} must be inserted between two video filters,
18050 @code{asendcmd} must be inserted between two audio filters, but apart
18051 from that they act the same way.
18053 The specification of commands can be provided in the filter arguments
18054 with the @var{commands} option, or in a file specified by the
18055 @var{filename} option.
18057 These filters accept the following options:
18060 Set the commands to be read and sent to the other filters.
18062 Set the filename of the commands to be read and sent to the other
18066 @subsection Commands syntax
18068 A commands description consists of a sequence of interval
18069 specifications, comprising a list of commands to be executed when a
18070 particular event related to that interval occurs. The occurring event
18071 is typically the current frame time entering or leaving a given time
18074 An interval is specified by the following syntax:
18076 @var{START}[-@var{END}] @var{COMMANDS};
18079 The time interval is specified by the @var{START} and @var{END} times.
18080 @var{END} is optional and defaults to the maximum time.
18082 The current frame time is considered within the specified interval if
18083 it is included in the interval [@var{START}, @var{END}), that is when
18084 the time is greater or equal to @var{START} and is lesser than
18087 @var{COMMANDS} consists of a sequence of one or more command
18088 specifications, separated by ",", relating to that interval. The
18089 syntax of a command specification is given by:
18091 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18094 @var{FLAGS} is optional and specifies the type of events relating to
18095 the time interval which enable sending the specified command, and must
18096 be a non-null sequence of identifier flags separated by "+" or "|" and
18097 enclosed between "[" and "]".
18099 The following flags are recognized:
18102 The command is sent when the current frame timestamp enters the
18103 specified interval. In other words, the command is sent when the
18104 previous frame timestamp was not in the given interval, and the
18108 The command is sent when the current frame timestamp leaves the
18109 specified interval. In other words, the command is sent when the
18110 previous frame timestamp was in the given interval, and the
18114 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18117 @var{TARGET} specifies the target of the command, usually the name of
18118 the filter class or a specific filter instance name.
18120 @var{COMMAND} specifies the name of the command for the target filter.
18122 @var{ARG} is optional and specifies the optional list of argument for
18123 the given @var{COMMAND}.
18125 Between one interval specification and another, whitespaces, or
18126 sequences of characters starting with @code{#} until the end of line,
18127 are ignored and can be used to annotate comments.
18129 A simplified BNF description of the commands specification syntax
18132 @var{COMMAND_FLAG} ::= "enter" | "leave"
18133 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18134 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18135 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18136 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18137 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18140 @subsection Examples
18144 Specify audio tempo change at second 4:
18146 asendcmd=c='4.0 atempo tempo 1.5',atempo
18150 Target a specific filter instance:
18152 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18156 Specify a list of drawtext and hue commands in a file.
18158 # show text in the interval 5-10
18159 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18160 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18162 # desaturate the image in the interval 15-20
18163 15.0-20.0 [enter] hue s 0,
18164 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18166 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18168 # apply an exponential saturation fade-out effect, starting from time 25
18169 25 [enter] hue s exp(25-t)
18172 A filtergraph allowing to read and process the above command list
18173 stored in a file @file{test.cmd}, can be specified with:
18175 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18180 @section setpts, asetpts
18182 Change the PTS (presentation timestamp) of the input frames.
18184 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18186 This filter accepts the following options:
18191 The expression which is evaluated for each frame to construct its timestamp.
18195 The expression is evaluated through the eval API and can contain the following
18200 frame rate, only defined for constant frame-rate video
18203 The presentation timestamp in input
18206 The count of the input frame for video or the number of consumed samples,
18207 not including the current frame for audio, starting from 0.
18209 @item NB_CONSUMED_SAMPLES
18210 The number of consumed samples, not including the current frame (only
18213 @item NB_SAMPLES, S
18214 The number of samples in the current frame (only audio)
18216 @item SAMPLE_RATE, SR
18217 The audio sample rate.
18220 The PTS of the first frame.
18223 the time in seconds of the first frame
18226 State whether the current frame is interlaced.
18229 the time in seconds of the current frame
18232 original position in the file of the frame, or undefined if undefined
18233 for the current frame
18236 The previous input PTS.
18239 previous input time in seconds
18242 The previous output PTS.
18245 previous output time in seconds
18248 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
18252 The wallclock (RTC) time at the start of the movie in microseconds.
18255 The timebase of the input timestamps.
18259 @subsection Examples
18263 Start counting PTS from zero
18265 setpts=PTS-STARTPTS
18269 Apply fast motion effect:
18275 Apply slow motion effect:
18281 Set fixed rate of 25 frames per second:
18287 Set fixed rate 25 fps with some jitter:
18289 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
18293 Apply an offset of 10 seconds to the input PTS:
18299 Generate timestamps from a "live source" and rebase onto the current timebase:
18301 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
18305 Generate timestamps by counting samples:
18312 @section settb, asettb
18314 Set the timebase to use for the output frames timestamps.
18315 It is mainly useful for testing timebase configuration.
18317 It accepts the following parameters:
18322 The expression which is evaluated into the output timebase.
18326 The value for @option{tb} is an arithmetic expression representing a
18327 rational. The expression can contain the constants "AVTB" (the default
18328 timebase), "intb" (the input timebase) and "sr" (the sample rate,
18329 audio only). Default value is "intb".
18331 @subsection Examples
18335 Set the timebase to 1/25:
18341 Set the timebase to 1/10:
18347 Set the timebase to 1001/1000:
18353 Set the timebase to 2*intb:
18359 Set the default timebase value:
18366 Convert input audio to a video output representing frequency spectrum
18367 logarithmically using Brown-Puckette constant Q transform algorithm with
18368 direct frequency domain coefficient calculation (but the transform itself
18369 is not really constant Q, instead the Q factor is actually variable/clamped),
18370 with musical tone scale, from E0 to D#10.
18372 The filter accepts the following options:
18376 Specify the video size for the output. It must be even. For the syntax of this option,
18377 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18378 Default value is @code{1920x1080}.
18381 Set the output frame rate. Default value is @code{25}.
18384 Set the bargraph height. It must be even. Default value is @code{-1} which
18385 computes the bargraph height automatically.
18388 Set the axis height. It must be even. Default value is @code{-1} which computes
18389 the axis height automatically.
18392 Set the sonogram height. It must be even. Default value is @code{-1} which
18393 computes the sonogram height automatically.
18396 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
18397 instead. Default value is @code{1}.
18399 @item sono_v, volume
18400 Specify the sonogram volume expression. It can contain variables:
18403 the @var{bar_v} evaluated expression
18404 @item frequency, freq, f
18405 the frequency where it is evaluated
18406 @item timeclamp, tc
18407 the value of @var{timeclamp} option
18411 @item a_weighting(f)
18412 A-weighting of equal loudness
18413 @item b_weighting(f)
18414 B-weighting of equal loudness
18415 @item c_weighting(f)
18416 C-weighting of equal loudness.
18418 Default value is @code{16}.
18420 @item bar_v, volume2
18421 Specify the bargraph volume expression. It can contain variables:
18424 the @var{sono_v} evaluated expression
18425 @item frequency, freq, f
18426 the frequency where it is evaluated
18427 @item timeclamp, tc
18428 the value of @var{timeclamp} option
18432 @item a_weighting(f)
18433 A-weighting of equal loudness
18434 @item b_weighting(f)
18435 B-weighting of equal loudness
18436 @item c_weighting(f)
18437 C-weighting of equal loudness.
18439 Default value is @code{sono_v}.
18441 @item sono_g, gamma
18442 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
18443 higher gamma makes the spectrum having more range. Default value is @code{3}.
18444 Acceptable range is @code{[1, 7]}.
18446 @item bar_g, gamma2
18447 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
18451 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
18452 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
18454 @item timeclamp, tc
18455 Specify the transform timeclamp. At low frequency, there is trade-off between
18456 accuracy in time domain and frequency domain. If timeclamp is lower,
18457 event in time domain is represented more accurately (such as fast bass drum),
18458 otherwise event in frequency domain is represented more accurately
18459 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
18462 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
18463 limits future samples by applying asymmetric windowing in time domain, useful
18464 when low latency is required. Accepted range is @code{[0, 1]}.
18467 Specify the transform base frequency. Default value is @code{20.01523126408007475},
18468 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
18471 Specify the transform end frequency. Default value is @code{20495.59681441799654},
18472 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
18475 This option is deprecated and ignored.
18478 Specify the transform length in time domain. Use this option to control accuracy
18479 trade-off between time domain and frequency domain at every frequency sample.
18480 It can contain variables:
18482 @item frequency, freq, f
18483 the frequency where it is evaluated
18484 @item timeclamp, tc
18485 the value of @var{timeclamp} option.
18487 Default value is @code{384*tc/(384+tc*f)}.
18490 Specify the transform count for every video frame. Default value is @code{6}.
18491 Acceptable range is @code{[1, 30]}.
18494 Specify the transform count for every single pixel. Default value is @code{0},
18495 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
18498 Specify font file for use with freetype to draw the axis. If not specified,
18499 use embedded font. Note that drawing with font file or embedded font is not
18500 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
18504 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
18505 The : in the pattern may be replaced by | to avoid unnecessary escaping.
18508 Specify font color expression. This is arithmetic expression that should return
18509 integer value 0xRRGGBB. It can contain variables:
18511 @item frequency, freq, f
18512 the frequency where it is evaluated
18513 @item timeclamp, tc
18514 the value of @var{timeclamp} option
18519 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
18520 @item r(x), g(x), b(x)
18521 red, green, and blue value of intensity x.
18523 Default value is @code{st(0, (midi(f)-59.5)/12);
18524 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
18525 r(1-ld(1)) + b(ld(1))}.
18528 Specify image file to draw the axis. This option override @var{fontfile} and
18529 @var{fontcolor} option.
18532 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
18533 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
18534 Default value is @code{1}.
18537 Set colorspace. The accepted values are:
18540 Unspecified (default)
18549 BT.470BG or BT.601-6 625
18552 SMPTE-170M or BT.601-6 525
18558 BT.2020 with non-constant luminance
18563 Set spectrogram color scheme. This is list of floating point values with format
18564 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
18565 The default is @code{1|0.5|0|0|0.5|1}.
18569 @subsection Examples
18573 Playing audio while showing the spectrum:
18575 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
18579 Same as above, but with frame rate 30 fps:
18581 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
18585 Playing at 1280x720:
18587 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
18591 Disable sonogram display:
18597 A1 and its harmonics: A1, A2, (near)E3, A3:
18599 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),
18600 asplit[a][out1]; [a] showcqt [out0]'
18604 Same as above, but with more accuracy in frequency domain:
18606 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),
18607 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
18613 bar_v=10:sono_v=bar_v*a_weighting(f)
18617 Custom gamma, now spectrum is linear to the amplitude.
18623 Custom tlength equation:
18625 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)))'
18629 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
18631 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
18635 Custom font using fontconfig:
18637 font='Courier New,Monospace,mono|bold'
18641 Custom frequency range with custom axis using image file:
18643 axisfile=myaxis.png:basefreq=40:endfreq=10000
18649 Convert input audio to video output representing the audio power spectrum.
18650 Audio amplitude is on Y-axis while frequency is on X-axis.
18652 The filter accepts the following options:
18656 Specify size of video. For the syntax of this option, check the
18657 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18658 Default is @code{1024x512}.
18662 This set how each frequency bin will be represented.
18664 It accepts the following values:
18670 Default is @code{bar}.
18673 Set amplitude scale.
18675 It accepts the following values:
18689 Default is @code{log}.
18692 Set frequency scale.
18694 It accepts the following values:
18703 Reverse logarithmic scale.
18705 Default is @code{lin}.
18710 It accepts the following values:
18726 Default is @code{w2048}
18729 Set windowing function.
18731 It accepts the following values:
18753 Default is @code{hanning}.
18756 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
18757 which means optimal overlap for selected window function will be picked.
18760 Set time averaging. Setting this to 0 will display current maximal peaks.
18761 Default is @code{1}, which means time averaging is disabled.
18764 Specify list of colors separated by space or by '|' which will be used to
18765 draw channel frequencies. Unrecognized or missing colors will be replaced
18769 Set channel display mode.
18771 It accepts the following values:
18776 Default is @code{combined}.
18779 Set minimum amplitude used in @code{log} amplitude scaler.
18783 @anchor{showspectrum}
18784 @section showspectrum
18786 Convert input audio to a video output, representing the audio frequency
18789 The filter accepts the following options:
18793 Specify the video size for the output. For the syntax of this option, check the
18794 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18795 Default value is @code{640x512}.
18798 Specify how the spectrum should slide along the window.
18800 It accepts the following values:
18803 the samples start again on the left when they reach the right
18805 the samples scroll from right to left
18807 frames are only produced when the samples reach the right
18809 the samples scroll from left to right
18812 Default value is @code{replace}.
18815 Specify display mode.
18817 It accepts the following values:
18820 all channels are displayed in the same row
18822 all channels are displayed in separate rows
18825 Default value is @samp{combined}.
18828 Specify display color mode.
18830 It accepts the following values:
18833 each channel is displayed in a separate color
18835 each channel is displayed using the same color scheme
18837 each channel is displayed using the rainbow color scheme
18839 each channel is displayed using the moreland color scheme
18841 each channel is displayed using the nebulae color scheme
18843 each channel is displayed using the fire color scheme
18845 each channel is displayed using the fiery color scheme
18847 each channel is displayed using the fruit color scheme
18849 each channel is displayed using the cool color scheme
18852 Default value is @samp{channel}.
18855 Specify scale used for calculating intensity color values.
18857 It accepts the following values:
18862 square root, default
18873 Default value is @samp{sqrt}.
18876 Set saturation modifier for displayed colors. Negative values provide
18877 alternative color scheme. @code{0} is no saturation at all.
18878 Saturation must be in [-10.0, 10.0] range.
18879 Default value is @code{1}.
18882 Set window function.
18884 It accepts the following values:
18908 Default value is @code{hann}.
18911 Set orientation of time vs frequency axis. Can be @code{vertical} or
18912 @code{horizontal}. Default is @code{vertical}.
18915 Set ratio of overlap window. Default value is @code{0}.
18916 When value is @code{1} overlap is set to recommended size for specific
18917 window function currently used.
18920 Set scale gain for calculating intensity color values.
18921 Default value is @code{1}.
18924 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
18927 Set color rotation, must be in [-1.0, 1.0] range.
18928 Default value is @code{0}.
18931 The usage is very similar to the showwaves filter; see the examples in that
18934 @subsection Examples
18938 Large window with logarithmic color scaling:
18940 showspectrum=s=1280x480:scale=log
18944 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
18946 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18947 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
18951 @section showspectrumpic
18953 Convert input audio to a single video frame, representing the audio frequency
18956 The filter accepts the following options:
18960 Specify the video size for the output. For the syntax of this option, check the
18961 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18962 Default value is @code{4096x2048}.
18965 Specify display mode.
18967 It accepts the following values:
18970 all channels are displayed in the same row
18972 all channels are displayed in separate rows
18974 Default value is @samp{combined}.
18977 Specify display color mode.
18979 It accepts the following values:
18982 each channel is displayed in a separate color
18984 each channel is displayed using the same color scheme
18986 each channel is displayed using the rainbow color scheme
18988 each channel is displayed using the moreland color scheme
18990 each channel is displayed using the nebulae color scheme
18992 each channel is displayed using the fire color scheme
18994 each channel is displayed using the fiery color scheme
18996 each channel is displayed using the fruit color scheme
18998 each channel is displayed using the cool color scheme
19000 Default value is @samp{intensity}.
19003 Specify scale used for calculating intensity color values.
19005 It accepts the following values:
19010 square root, default
19020 Default value is @samp{log}.
19023 Set saturation modifier for displayed colors. Negative values provide
19024 alternative color scheme. @code{0} is no saturation at all.
19025 Saturation must be in [-10.0, 10.0] range.
19026 Default value is @code{1}.
19029 Set window function.
19031 It accepts the following values:
19054 Default value is @code{hann}.
19057 Set orientation of time vs frequency axis. Can be @code{vertical} or
19058 @code{horizontal}. Default is @code{vertical}.
19061 Set scale gain for calculating intensity color values.
19062 Default value is @code{1}.
19065 Draw time and frequency axes and legends. Default is enabled.
19068 Set color rotation, must be in [-1.0, 1.0] range.
19069 Default value is @code{0}.
19072 @subsection Examples
19076 Extract an audio spectrogram of a whole audio track
19077 in a 1024x1024 picture using @command{ffmpeg}:
19079 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19083 @section showvolume
19085 Convert input audio volume to a video output.
19087 The filter accepts the following options:
19094 Set border width, allowed range is [0, 5]. Default is 1.
19097 Set channel width, allowed range is [80, 8192]. Default is 400.
19100 Set channel height, allowed range is [1, 900]. Default is 20.
19103 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19106 Set volume color expression.
19108 The expression can use the following variables:
19112 Current max volume of channel in dB.
19118 Current channel number, starting from 0.
19122 If set, displays channel names. Default is enabled.
19125 If set, displays volume values. Default is enabled.
19128 Set orientation, can be @code{horizontal} or @code{vertical},
19129 default is @code{horizontal}.
19132 Set step size, allowed range s [0, 5]. Default is 0, which means
19138 Convert input audio to a video output, representing the samples waves.
19140 The filter accepts the following options:
19144 Specify the video size for the output. For the syntax of this option, check the
19145 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19146 Default value is @code{600x240}.
19151 Available values are:
19154 Draw a point for each sample.
19157 Draw a vertical line for each sample.
19160 Draw a point for each sample and a line between them.
19163 Draw a centered vertical line for each sample.
19166 Default value is @code{point}.
19169 Set the number of samples which are printed on the same column. A
19170 larger value will decrease the frame rate. Must be a positive
19171 integer. This option can be set only if the value for @var{rate}
19172 is not explicitly specified.
19175 Set the (approximate) output frame rate. This is done by setting the
19176 option @var{n}. Default value is "25".
19178 @item split_channels
19179 Set if channels should be drawn separately or overlap. Default value is 0.
19182 Set colors separated by '|' which are going to be used for drawing of each channel.
19185 Set amplitude scale.
19187 Available values are:
19205 @subsection Examples
19209 Output the input file audio and the corresponding video representation
19212 amovie=a.mp3,asplit[out0],showwaves[out1]
19216 Create a synthetic signal and show it with showwaves, forcing a
19217 frame rate of 30 frames per second:
19219 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
19223 @section showwavespic
19225 Convert input audio to a single video frame, representing the samples waves.
19227 The filter accepts the following options:
19231 Specify the video size for the output. For the syntax of this option, check the
19232 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19233 Default value is @code{600x240}.
19235 @item split_channels
19236 Set if channels should be drawn separately or overlap. Default value is 0.
19239 Set colors separated by '|' which are going to be used for drawing of each channel.
19242 Set amplitude scale.
19244 Available values are:
19262 @subsection Examples
19266 Extract a channel split representation of the wave form of a whole audio track
19267 in a 1024x800 picture using @command{ffmpeg}:
19269 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
19273 @section sidedata, asidedata
19275 Delete frame side data, or select frames based on it.
19277 This filter accepts the following options:
19281 Set mode of operation of the filter.
19283 Can be one of the following:
19287 Select every frame with side data of @code{type}.
19290 Delete side data of @code{type}. If @code{type} is not set, delete all side
19296 Set side data type used with all modes. Must be set for @code{select} mode. For
19297 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
19298 in @file{libavutil/frame.h}. For example, to choose
19299 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
19303 @section spectrumsynth
19305 Sythesize audio from 2 input video spectrums, first input stream represents
19306 magnitude across time and second represents phase across time.
19307 The filter will transform from frequency domain as displayed in videos back
19308 to time domain as presented in audio output.
19310 This filter is primarily created for reversing processed @ref{showspectrum}
19311 filter outputs, but can synthesize sound from other spectrograms too.
19312 But in such case results are going to be poor if the phase data is not
19313 available, because in such cases phase data need to be recreated, usually
19314 its just recreated from random noise.
19315 For best results use gray only output (@code{channel} color mode in
19316 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
19317 @code{lin} scale for phase video. To produce phase, for 2nd video, use
19318 @code{data} option. Inputs videos should generally use @code{fullframe}
19319 slide mode as that saves resources needed for decoding video.
19321 The filter accepts the following options:
19325 Specify sample rate of output audio, the sample rate of audio from which
19326 spectrum was generated may differ.
19329 Set number of channels represented in input video spectrums.
19332 Set scale which was used when generating magnitude input spectrum.
19333 Can be @code{lin} or @code{log}. Default is @code{log}.
19336 Set slide which was used when generating inputs spectrums.
19337 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
19338 Default is @code{fullframe}.
19341 Set window function used for resynthesis.
19344 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19345 which means optimal overlap for selected window function will be picked.
19348 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
19349 Default is @code{vertical}.
19352 @subsection Examples
19356 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
19357 then resynthesize videos back to audio with spectrumsynth:
19359 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
19360 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
19361 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
19365 @section split, asplit
19367 Split input into several identical outputs.
19369 @code{asplit} works with audio input, @code{split} with video.
19371 The filter accepts a single parameter which specifies the number of outputs. If
19372 unspecified, it defaults to 2.
19374 @subsection Examples
19378 Create two separate outputs from the same input:
19380 [in] split [out0][out1]
19384 To create 3 or more outputs, you need to specify the number of
19387 [in] asplit=3 [out0][out1][out2]
19391 Create two separate outputs from the same input, one cropped and
19394 [in] split [splitout1][splitout2];
19395 [splitout1] crop=100:100:0:0 [cropout];
19396 [splitout2] pad=200:200:100:100 [padout];
19400 Create 5 copies of the input audio with @command{ffmpeg}:
19402 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
19408 Receive commands sent through a libzmq client, and forward them to
19409 filters in the filtergraph.
19411 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
19412 must be inserted between two video filters, @code{azmq} between two
19415 To enable these filters you need to install the libzmq library and
19416 headers and configure FFmpeg with @code{--enable-libzmq}.
19418 For more information about libzmq see:
19419 @url{http://www.zeromq.org/}
19421 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
19422 receives messages sent through a network interface defined by the
19423 @option{bind_address} option.
19425 The received message must be in the form:
19427 @var{TARGET} @var{COMMAND} [@var{ARG}]
19430 @var{TARGET} specifies the target of the command, usually the name of
19431 the filter class or a specific filter instance name.
19433 @var{COMMAND} specifies the name of the command for the target filter.
19435 @var{ARG} is optional and specifies the optional argument list for the
19436 given @var{COMMAND}.
19438 Upon reception, the message is processed and the corresponding command
19439 is injected into the filtergraph. Depending on the result, the filter
19440 will send a reply to the client, adopting the format:
19442 @var{ERROR_CODE} @var{ERROR_REASON}
19446 @var{MESSAGE} is optional.
19448 @subsection Examples
19450 Look at @file{tools/zmqsend} for an example of a zmq client which can
19451 be used to send commands processed by these filters.
19453 Consider the following filtergraph generated by @command{ffplay}
19455 ffplay -dumpgraph 1 -f lavfi "
19456 color=s=100x100:c=red [l];
19457 color=s=100x100:c=blue [r];
19458 nullsrc=s=200x100, zmq [bg];
19459 [bg][l] overlay [bg+l];
19460 [bg+l][r] overlay=x=100 "
19463 To change the color of the left side of the video, the following
19464 command can be used:
19466 echo Parsed_color_0 c yellow | tools/zmqsend
19469 To change the right side:
19471 echo Parsed_color_1 c pink | tools/zmqsend
19474 @c man end MULTIMEDIA FILTERS
19476 @chapter Multimedia Sources
19477 @c man begin MULTIMEDIA SOURCES
19479 Below is a description of the currently available multimedia sources.
19483 This is the same as @ref{movie} source, except it selects an audio
19489 Read audio and/or video stream(s) from a movie container.
19491 It accepts the following parameters:
19495 The name of the resource to read (not necessarily a file; it can also be a
19496 device or a stream accessed through some protocol).
19498 @item format_name, f
19499 Specifies the format assumed for the movie to read, and can be either
19500 the name of a container or an input device. If not specified, the
19501 format is guessed from @var{movie_name} or by probing.
19503 @item seek_point, sp
19504 Specifies the seek point in seconds. The frames will be output
19505 starting from this seek point. The parameter is evaluated with
19506 @code{av_strtod}, so the numerical value may be suffixed by an IS
19507 postfix. The default value is "0".
19510 Specifies the streams to read. Several streams can be specified,
19511 separated by "+". The source will then have as many outputs, in the
19512 same order. The syntax is explained in the ``Stream specifiers''
19513 section in the ffmpeg manual. Two special names, "dv" and "da" specify
19514 respectively the default (best suited) video and audio stream. Default
19515 is "dv", or "da" if the filter is called as "amovie".
19517 @item stream_index, si
19518 Specifies the index of the video stream to read. If the value is -1,
19519 the most suitable video stream will be automatically selected. The default
19520 value is "-1". Deprecated. If the filter is called "amovie", it will select
19521 audio instead of video.
19524 Specifies how many times to read the stream in sequence.
19525 If the value is 0, the stream will be looped infinitely.
19526 Default value is "1".
19528 Note that when the movie is looped the source timestamps are not
19529 changed, so it will generate non monotonically increasing timestamps.
19531 @item discontinuity
19532 Specifies the time difference between frames above which the point is
19533 considered a timestamp discontinuity which is removed by adjusting the later
19537 It allows overlaying a second video on top of the main input of
19538 a filtergraph, as shown in this graph:
19540 input -----------> deltapts0 --> overlay --> output
19543 movie --> scale--> deltapts1 -------+
19545 @subsection Examples
19549 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
19550 on top of the input labelled "in":
19552 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
19553 [in] setpts=PTS-STARTPTS [main];
19554 [main][over] overlay=16:16 [out]
19558 Read from a video4linux2 device, and overlay it on top of the input
19561 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
19562 [in] setpts=PTS-STARTPTS [main];
19563 [main][over] overlay=16:16 [out]
19567 Read the first video stream and the audio stream with id 0x81 from
19568 dvd.vob; the video is connected to the pad named "video" and the audio is
19569 connected to the pad named "audio":
19571 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
19575 @subsection Commands
19577 Both movie and amovie support the following commands:
19580 Perform seek using "av_seek_frame".
19581 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
19584 @var{stream_index}: If stream_index is -1, a default
19585 stream is selected, and @var{timestamp} is automatically converted
19586 from AV_TIME_BASE units to the stream specific time_base.
19588 @var{timestamp}: Timestamp in AVStream.time_base units
19589 or, if no stream is specified, in AV_TIME_BASE units.
19591 @var{flags}: Flags which select direction and seeking mode.
19595 Get movie duration in AV_TIME_BASE units.
19599 @c man end MULTIMEDIA SOURCES