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.
434 Copy the input audio source unchanged to the output. This is mainly useful for
439 Apply cross fade from one input audio stream to another input audio stream.
440 The cross fade is applied for specified duration near the end of first stream.
442 The filter accepts the following options:
446 Specify the number of samples for which the cross fade effect has to last.
447 At the end of the cross fade effect the first input audio will be completely
448 silent. Default is 44100.
451 Specify the duration of the cross fade effect. See
452 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
453 for the accepted syntax.
454 By default the duration is determined by @var{nb_samples}.
455 If set this option is used instead of @var{nb_samples}.
458 Should first stream end overlap with second stream start. Default is enabled.
461 Set curve for cross fade transition for first stream.
464 Set curve for cross fade transition for second stream.
466 For description of available curve types see @ref{afade} filter description.
473 Cross fade from one input to another:
475 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
479 Cross fade from one input to another but without overlapping:
481 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
487 Reduce audio bit resolution.
489 This filter is bit crusher with enhanced functionality. A bit crusher
490 is used to audibly reduce number of bits an audio signal is sampled
491 with. This doesn't change the bit depth at all, it just produces the
492 effect. Material reduced in bit depth sounds more harsh and "digital".
493 This filter is able to even round to continuous values instead of discrete
495 Additionally it has a D/C offset which results in different crushing of
496 the lower and the upper half of the signal.
497 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
499 Another feature of this filter is the logarithmic mode.
500 This setting switches from linear distances between bits to logarithmic ones.
501 The result is a much more "natural" sounding crusher which doesn't gate low
502 signals for example. The human ear has a logarithmic perception, too
503 so this kind of crushing is much more pleasant.
504 Logarithmic crushing is also able to get anti-aliased.
506 The filter accepts the following options:
522 Can be linear: @code{lin} or logarithmic: @code{log}.
531 Set sample reduction.
534 Enable LFO. By default disabled.
545 Delay one or more audio channels.
547 Samples in delayed channel are filled with silence.
549 The filter accepts the following option:
553 Set list of delays in milliseconds for each channel separated by '|'.
554 Unused delays will be silently ignored. If number of given delays is
555 smaller than number of channels all remaining channels will not be delayed.
556 If you want to delay exact number of samples, append 'S' to number.
563 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
564 the second channel (and any other channels that may be present) unchanged.
570 Delay second channel by 500 samples, the third channel by 700 samples and leave
571 the first channel (and any other channels that may be present) unchanged.
579 Apply echoing to the input audio.
581 Echoes are reflected sound and can occur naturally amongst mountains
582 (and sometimes large buildings) when talking or shouting; digital echo
583 effects emulate this behaviour and are often used to help fill out the
584 sound of a single instrument or vocal. The time difference between the
585 original signal and the reflection is the @code{delay}, and the
586 loudness of the reflected signal is the @code{decay}.
587 Multiple echoes can have different delays and decays.
589 A description of the accepted parameters follows.
593 Set input gain of reflected signal. Default is @code{0.6}.
596 Set output gain of reflected signal. Default is @code{0.3}.
599 Set list of time intervals in milliseconds between original signal and reflections
600 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
601 Default is @code{1000}.
604 Set list of loudness of reflected signals separated by '|'.
605 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
606 Default is @code{0.5}.
613 Make it sound as if there are twice as many instruments as are actually playing:
615 aecho=0.8:0.88:60:0.4
619 If delay is very short, then it sound like a (metallic) robot playing music:
625 A longer delay will sound like an open air concert in the mountains:
627 aecho=0.8:0.9:1000:0.3
631 Same as above but with one more mountain:
633 aecho=0.8:0.9:1000|1800:0.3|0.25
638 Audio emphasis filter creates or restores material directly taken from LPs or
639 emphased CDs with different filter curves. E.g. to store music on vinyl the
640 signal has to be altered by a filter first to even out the disadvantages of
641 this recording medium.
642 Once the material is played back the inverse filter has to be applied to
643 restore the distortion of the frequency response.
645 The filter accepts the following options:
655 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
656 use @code{production} mode. Default is @code{reproduction} mode.
659 Set filter type. Selects medium. Can be one of the following:
671 select Compact Disc (CD).
677 select 50µs (FM-KF).
679 select 75µs (FM-KF).
685 Modify an audio signal according to the specified expressions.
687 This filter accepts one or more expressions (one for each channel),
688 which are evaluated and used to modify a corresponding audio signal.
690 It accepts the following parameters:
694 Set the '|'-separated expressions list for each separate channel. If
695 the number of input channels is greater than the number of
696 expressions, the last specified expression is used for the remaining
699 @item channel_layout, c
700 Set output channel layout. If not specified, the channel layout is
701 specified by the number of expressions. If set to @samp{same}, it will
702 use by default the same input channel layout.
705 Each expression in @var{exprs} can contain the following constants and functions:
709 channel number of the current expression
712 number of the evaluated sample, starting from 0
718 time of the evaluated sample expressed in seconds
721 @item nb_out_channels
722 input and output number of channels
725 the value of input channel with number @var{CH}
728 Note: this filter is slow. For faster processing you should use a
737 aeval=val(ch)/2:c=same
741 Invert phase of the second channel:
750 Apply fade-in/out effect to input audio.
752 A description of the accepted parameters follows.
756 Specify the effect type, can be either @code{in} for fade-in, or
757 @code{out} for a fade-out effect. Default is @code{in}.
759 @item start_sample, ss
760 Specify the number of the start sample for starting to apply the fade
761 effect. Default is 0.
764 Specify the number of samples for which the fade effect has to last. At
765 the end of the fade-in effect the output audio will have the same
766 volume as the input audio, at the end of the fade-out transition
767 the output audio will be silence. Default is 44100.
770 Specify the start time of the fade effect. Default is 0.
771 The value must be specified as a time duration; see
772 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
773 for the accepted syntax.
774 If set this option is used instead of @var{start_sample}.
777 Specify the duration of the fade effect. See
778 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
779 for the accepted syntax.
780 At the end of the fade-in effect the output audio will have the same
781 volume as the input audio, at the end of the fade-out transition
782 the output audio will be silence.
783 By default the duration is determined by @var{nb_samples}.
784 If set this option is used instead of @var{nb_samples}.
787 Set curve for fade transition.
789 It accepts the following values:
792 select triangular, linear slope (default)
794 select quarter of sine wave
796 select half of sine wave
798 select exponential sine wave
802 select inverted parabola
816 select inverted quarter of sine wave
818 select inverted half of sine wave
820 select double-exponential seat
822 select double-exponential sigmoid
830 Fade in first 15 seconds of audio:
836 Fade out last 25 seconds of a 900 seconds audio:
838 afade=t=out:st=875:d=25
843 Apply arbitrary expressions to samples in frequency domain.
847 Set frequency domain real expression for each separate channel separated
848 by '|'. Default is "1".
849 If the number of input channels is greater than the number of
850 expressions, the last specified expression is used for the remaining
854 Set frequency domain imaginary expression for each separate channel
855 separated by '|'. If not set, @var{real} option is used.
857 Each expression in @var{real} and @var{imag} can contain the following
865 current frequency bin number
868 number of available bins
871 channel number of the current expression
883 It accepts the following values:
899 Default is @code{w4096}
902 Set window function. Default is @code{hann}.
905 Set window overlap. If set to 1, the recommended overlap for selected
906 window function will be picked. Default is @code{0.75}.
913 Leave almost only low frequencies in audio:
915 afftfilt="1-clip((b/nb)*b,0,1)"
921 Apply an arbitrary Frequency Impulse Response filter.
923 This filter is designed for applying long FIR filters,
924 up to 30 seconds long.
926 It can be used as component for digital crossover filters,
927 room equalization, cross talk cancellation, wavefield synthesis,
928 auralization, ambiophonics and ambisonics.
930 This filter uses second stream as FIR coefficients.
931 If second stream holds single channel, it will be used
932 for all input channels in first stream, otherwise
933 number of channels in second stream must be same as
934 number of channels in first stream.
936 It accepts the following parameters:
940 Set dry gain. This sets input gain.
943 Set wet gain. This sets final output gain.
946 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
949 Enable applying gain measured from power of IR.
956 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
958 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
965 Set output format constraints for the input audio. The framework will
966 negotiate the most appropriate format to minimize conversions.
968 It accepts the following parameters:
972 A '|'-separated list of requested sample formats.
975 A '|'-separated list of requested sample rates.
977 @item channel_layouts
978 A '|'-separated list of requested channel layouts.
980 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
981 for the required syntax.
984 If a parameter is omitted, all values are allowed.
986 Force the output to either unsigned 8-bit or signed 16-bit stereo
988 aformat=sample_fmts=u8|s16:channel_layouts=stereo
993 A gate is mainly used to reduce lower parts of a signal. This kind of signal
994 processing reduces disturbing noise between useful signals.
996 Gating is done by detecting the volume below a chosen level @var{threshold}
997 and dividing it by the factor set with @var{ratio}. The bottom of the noise
998 floor is set via @var{range}. Because an exact manipulation of the signal
999 would cause distortion of the waveform the reduction can be levelled over
1000 time. This is done by setting @var{attack} and @var{release}.
1002 @var{attack} determines how long the signal has to fall below the threshold
1003 before any reduction will occur and @var{release} sets the time the signal
1004 has to rise above the threshold to reduce the reduction again.
1005 Shorter signals than the chosen attack time will be left untouched.
1009 Set input level before filtering.
1010 Default is 1. Allowed range is from 0.015625 to 64.
1013 Set the level of gain reduction when the signal is below the threshold.
1014 Default is 0.06125. Allowed range is from 0 to 1.
1017 If a signal rises above this level the gain reduction is released.
1018 Default is 0.125. Allowed range is from 0 to 1.
1021 Set a ratio by which the signal is reduced.
1022 Default is 2. Allowed range is from 1 to 9000.
1025 Amount of milliseconds the signal has to rise above the threshold before gain
1027 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1030 Amount of milliseconds the signal has to fall below the threshold before the
1031 reduction is increased again. Default is 250 milliseconds.
1032 Allowed range is from 0.01 to 9000.
1035 Set amount of amplification of signal after processing.
1036 Default is 1. Allowed range is from 1 to 64.
1039 Curve the sharp knee around the threshold to enter gain reduction more softly.
1040 Default is 2.828427125. Allowed range is from 1 to 8.
1043 Choose if exact signal should be taken for detection or an RMS like one.
1044 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1047 Choose if the average level between all channels or the louder channel affects
1049 Default is @code{average}. Can be @code{average} or @code{maximum}.
1054 The limiter prevents an input signal from rising over a desired threshold.
1055 This limiter uses lookahead technology to prevent your signal from distorting.
1056 It means that there is a small delay after the signal is processed. Keep in mind
1057 that the delay it produces is the attack time you set.
1059 The filter accepts the following options:
1063 Set input gain. Default is 1.
1066 Set output gain. Default is 1.
1069 Don't let signals above this level pass the limiter. Default is 1.
1072 The limiter will reach its attenuation level in this amount of time in
1073 milliseconds. Default is 5 milliseconds.
1076 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1077 Default is 50 milliseconds.
1080 When gain reduction is always needed ASC takes care of releasing to an
1081 average reduction level rather than reaching a reduction of 0 in the release
1085 Select how much the release time is affected by ASC, 0 means nearly no changes
1086 in release time while 1 produces higher release times.
1089 Auto level output signal. Default is enabled.
1090 This normalizes audio back to 0dB if enabled.
1093 Depending on picked setting it is recommended to upsample input 2x or 4x times
1094 with @ref{aresample} before applying this filter.
1098 Apply a two-pole all-pass filter with central frequency (in Hz)
1099 @var{frequency}, and filter-width @var{width}.
1100 An all-pass filter changes the audio's frequency to phase relationship
1101 without changing its frequency to amplitude relationship.
1103 The filter accepts the following options:
1107 Set frequency in Hz.
1110 Set method to specify band-width of filter.
1123 Specify the band-width of a filter in width_type units.
1126 Specify which channels to filter, by default all available are filtered.
1133 The filter accepts the following options:
1137 Set the number of loops.
1140 Set maximal number of samples.
1143 Set first sample of loop.
1149 Merge two or more audio streams into a single multi-channel stream.
1151 The filter accepts the following options:
1156 Set the number of inputs. Default is 2.
1160 If the channel layouts of the inputs are disjoint, and therefore compatible,
1161 the channel layout of the output will be set accordingly and the channels
1162 will be reordered as necessary. If the channel layouts of the inputs are not
1163 disjoint, the output will have all the channels of the first input then all
1164 the channels of the second input, in that order, and the channel layout of
1165 the output will be the default value corresponding to the total number of
1168 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1169 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1170 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1171 first input, b1 is the first channel of the second input).
1173 On the other hand, if both input are in stereo, the output channels will be
1174 in the default order: a1, a2, b1, b2, and the channel layout will be
1175 arbitrarily set to 4.0, which may or may not be the expected value.
1177 All inputs must have the same sample rate, and format.
1179 If inputs do not have the same duration, the output will stop with the
1182 @subsection Examples
1186 Merge two mono files into a stereo stream:
1188 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1192 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1194 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
1200 Mixes multiple audio inputs into a single output.
1202 Note that this filter only supports float samples (the @var{amerge}
1203 and @var{pan} audio filters support many formats). If the @var{amix}
1204 input has integer samples then @ref{aresample} will be automatically
1205 inserted to perform the conversion to float samples.
1209 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1211 will mix 3 input audio streams to a single output with the same duration as the
1212 first input and a dropout transition time of 3 seconds.
1214 It accepts the following parameters:
1218 The number of inputs. If unspecified, it defaults to 2.
1221 How to determine the end-of-stream.
1225 The duration of the longest input. (default)
1228 The duration of the shortest input.
1231 The duration of the first input.
1235 @item dropout_transition
1236 The transition time, in seconds, for volume renormalization when an input
1237 stream ends. The default value is 2 seconds.
1241 @section anequalizer
1243 High-order parametric multiband equalizer for each channel.
1245 It accepts the following parameters:
1249 This option string is in format:
1250 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1251 Each equalizer band is separated by '|'.
1255 Set channel number to which equalization will be applied.
1256 If input doesn't have that channel the entry is ignored.
1259 Set central frequency for band.
1260 If input doesn't have that frequency the entry is ignored.
1263 Set band width in hertz.
1266 Set band gain in dB.
1269 Set filter type for band, optional, can be:
1273 Butterworth, this is default.
1284 With this option activated frequency response of anequalizer is displayed
1288 Set video stream size. Only useful if curves option is activated.
1291 Set max gain that will be displayed. Only useful if curves option is activated.
1292 Setting this to a reasonable value makes it possible to display gain which is derived from
1293 neighbour bands which are too close to each other and thus produce higher gain
1294 when both are activated.
1297 Set frequency scale used to draw frequency response in video output.
1298 Can be linear or logarithmic. Default is logarithmic.
1301 Set color for each channel curve which is going to be displayed in video stream.
1302 This is list of color names separated by space or by '|'.
1303 Unrecognised or missing colors will be replaced by white color.
1306 @subsection Examples
1310 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1311 for first 2 channels using Chebyshev type 1 filter:
1313 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1317 @subsection Commands
1319 This filter supports the following commands:
1322 Alter existing filter parameters.
1323 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1325 @var{fN} is existing filter number, starting from 0, if no such filter is available
1327 @var{freq} set new frequency parameter.
1328 @var{width} set new width parameter in herz.
1329 @var{gain} set new gain parameter in dB.
1331 Full filter invocation with asendcmd may look like this:
1332 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1337 Pass the audio source unchanged to the output.
1341 Pad the end of an audio stream with silence.
1343 This can be used together with @command{ffmpeg} @option{-shortest} to
1344 extend audio streams to the same length as the video stream.
1346 A description of the accepted options follows.
1350 Set silence packet size. Default value is 4096.
1353 Set the number of samples of silence to add to the end. After the
1354 value is reached, the stream is terminated. This option is mutually
1355 exclusive with @option{whole_len}.
1358 Set the minimum total number of samples in the output audio stream. If
1359 the value is longer than the input audio length, silence is added to
1360 the end, until the value is reached. This option is mutually exclusive
1361 with @option{pad_len}.
1364 If neither the @option{pad_len} nor the @option{whole_len} option is
1365 set, the filter will add silence to the end of the input stream
1368 @subsection Examples
1372 Add 1024 samples of silence to the end of the input:
1378 Make sure the audio output will contain at least 10000 samples, pad
1379 the input with silence if required:
1381 apad=whole_len=10000
1385 Use @command{ffmpeg} to pad the audio input with silence, so that the
1386 video stream will always result the shortest and will be converted
1387 until the end in the output file when using the @option{shortest}
1390 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1395 Add a phasing effect to the input audio.
1397 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1398 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1400 A description of the accepted parameters follows.
1404 Set input gain. Default is 0.4.
1407 Set output gain. Default is 0.74
1410 Set delay in milliseconds. Default is 3.0.
1413 Set decay. Default is 0.4.
1416 Set modulation speed in Hz. Default is 0.5.
1419 Set modulation type. Default is triangular.
1421 It accepts the following values:
1430 Audio pulsator is something between an autopanner and a tremolo.
1431 But it can produce funny stereo effects as well. Pulsator changes the volume
1432 of the left and right channel based on a LFO (low frequency oscillator) with
1433 different waveforms and shifted phases.
1434 This filter have the ability to define an offset between left and right
1435 channel. An offset of 0 means that both LFO shapes match each other.
1436 The left and right channel are altered equally - a conventional tremolo.
1437 An offset of 50% means that the shape of the right channel is exactly shifted
1438 in phase (or moved backwards about half of the frequency) - pulsator acts as
1439 an autopanner. At 1 both curves match again. Every setting in between moves the
1440 phase shift gapless between all stages and produces some "bypassing" sounds with
1441 sine and triangle waveforms. The more you set the offset near 1 (starting from
1442 the 0.5) the faster the signal passes from the left to the right speaker.
1444 The filter accepts the following options:
1448 Set input gain. By default it is 1. Range is [0.015625 - 64].
1451 Set output gain. By default it is 1. Range is [0.015625 - 64].
1454 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1455 sawup or sawdown. Default is sine.
1458 Set modulation. Define how much of original signal is affected by the LFO.
1461 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1464 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1467 Set pulse width. Default is 1. Allowed range is [0 - 2].
1470 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1473 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1477 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1481 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1482 if timing is set to hz.
1488 Resample the input audio to the specified parameters, using the
1489 libswresample library. If none are specified then the filter will
1490 automatically convert between its input and output.
1492 This filter is also able to stretch/squeeze the audio data to make it match
1493 the timestamps or to inject silence / cut out audio to make it match the
1494 timestamps, do a combination of both or do neither.
1496 The filter accepts the syntax
1497 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1498 expresses a sample rate and @var{resampler_options} is a list of
1499 @var{key}=@var{value} pairs, separated by ":". See the
1500 @ref{Resampler Options,,the "Resampler Options" section in the
1501 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1502 for the complete list of supported options.
1504 @subsection Examples
1508 Resample the input audio to 44100Hz:
1514 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1515 samples per second compensation:
1517 aresample=async=1000
1523 Reverse an audio clip.
1525 Warning: This filter requires memory to buffer the entire clip, so trimming
1528 @subsection Examples
1532 Take the first 5 seconds of a clip, and reverse it.
1534 atrim=end=5,areverse
1538 @section asetnsamples
1540 Set the number of samples per each output audio frame.
1542 The last output packet may contain a different number of samples, as
1543 the filter will flush all the remaining samples when the input audio
1546 The filter accepts the following options:
1550 @item nb_out_samples, n
1551 Set the number of frames per each output audio frame. The number is
1552 intended as the number of samples @emph{per each channel}.
1553 Default value is 1024.
1556 If set to 1, the filter will pad the last audio frame with zeroes, so
1557 that the last frame will contain the same number of samples as the
1558 previous ones. Default value is 1.
1561 For example, to set the number of per-frame samples to 1234 and
1562 disable padding for the last frame, use:
1564 asetnsamples=n=1234:p=0
1569 Set the sample rate without altering the PCM data.
1570 This will result in a change of speed and pitch.
1572 The filter accepts the following options:
1575 @item sample_rate, r
1576 Set the output sample rate. Default is 44100 Hz.
1581 Show a line containing various information for each input audio frame.
1582 The input audio is not modified.
1584 The shown line contains a sequence of key/value pairs of the form
1585 @var{key}:@var{value}.
1587 The following values are shown in the output:
1591 The (sequential) number of the input frame, starting from 0.
1594 The presentation timestamp of the input frame, in time base units; the time base
1595 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1598 The presentation timestamp of the input frame in seconds.
1601 position of the frame in the input stream, -1 if this information in
1602 unavailable and/or meaningless (for example in case of synthetic audio)
1611 The sample rate for the audio frame.
1614 The number of samples (per channel) in the frame.
1617 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1618 audio, the data is treated as if all the planes were concatenated.
1620 @item plane_checksums
1621 A list of Adler-32 checksums for each data plane.
1627 Display time domain statistical information about the audio channels.
1628 Statistics are calculated and displayed for each audio channel and,
1629 where applicable, an overall figure is also given.
1631 It accepts the following option:
1634 Short window length in seconds, used for peak and trough RMS measurement.
1635 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1639 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1640 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1643 Available keys for each channel are:
1677 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1678 this @code{lavfi.astats.Overall.Peak_count}.
1680 For description what each key means read below.
1683 Set number of frame after which stats are going to be recalculated.
1684 Default is disabled.
1687 A description of each shown parameter follows:
1691 Mean amplitude displacement from zero.
1694 Minimal sample level.
1697 Maximal sample level.
1699 @item Min difference
1700 Minimal difference between two consecutive samples.
1702 @item Max difference
1703 Maximal difference between two consecutive samples.
1705 @item Mean difference
1706 Mean difference between two consecutive samples.
1707 The average of each difference between two consecutive samples.
1709 @item RMS difference
1710 Root Mean Square difference between two consecutive samples.
1714 Standard peak and RMS level measured in dBFS.
1718 Peak and trough values for RMS level measured over a short window.
1721 Standard ratio of peak to RMS level (note: not in dB).
1724 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1725 (i.e. either @var{Min level} or @var{Max level}).
1728 Number of occasions (not the number of samples) that the signal attained either
1729 @var{Min level} or @var{Max level}.
1732 Overall bit depth of audio. Number of bits used for each sample.
1735 Measured dynamic range of audio in dB.
1742 The filter accepts exactly one parameter, the audio tempo. If not
1743 specified then the filter will assume nominal 1.0 tempo. Tempo must
1744 be in the [0.5, 2.0] range.
1746 @subsection Examples
1750 Slow down audio to 80% tempo:
1756 To speed up audio to 125% tempo:
1764 Trim the input so that the output contains one continuous subpart of the input.
1766 It accepts the following parameters:
1769 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1770 sample with the timestamp @var{start} will be the first sample in the output.
1773 Specify time of the first audio sample that will be dropped, i.e. the
1774 audio sample immediately preceding the one with the timestamp @var{end} will be
1775 the last sample in the output.
1778 Same as @var{start}, except this option sets the start timestamp in samples
1782 Same as @var{end}, except this option sets the end timestamp in samples instead
1786 The maximum duration of the output in seconds.
1789 The number of the first sample that should be output.
1792 The number of the first sample that should be dropped.
1795 @option{start}, @option{end}, and @option{duration} are expressed as time
1796 duration specifications; see
1797 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1799 Note that the first two sets of the start/end options and the @option{duration}
1800 option look at the frame timestamp, while the _sample options simply count the
1801 samples that pass through the filter. So start/end_pts and start/end_sample will
1802 give different results when the timestamps are wrong, inexact or do not start at
1803 zero. Also note that this filter does not modify the timestamps. If you wish
1804 to have the output timestamps start at zero, insert the asetpts filter after the
1807 If multiple start or end options are set, this filter tries to be greedy and
1808 keep all samples that match at least one of the specified constraints. To keep
1809 only the part that matches all the constraints at once, chain multiple atrim
1812 The defaults are such that all the input is kept. So it is possible to set e.g.
1813 just the end values to keep everything before the specified time.
1818 Drop everything except the second minute of input:
1820 ffmpeg -i INPUT -af atrim=60:120
1824 Keep only the first 1000 samples:
1826 ffmpeg -i INPUT -af atrim=end_sample=1000
1833 Apply a two-pole Butterworth band-pass filter with central
1834 frequency @var{frequency}, and (3dB-point) band-width width.
1835 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1836 instead of the default: constant 0dB peak gain.
1837 The filter roll off at 6dB per octave (20dB per decade).
1839 The filter accepts the following options:
1843 Set the filter's central frequency. Default is @code{3000}.
1846 Constant skirt gain if set to 1. Defaults to 0.
1849 Set method to specify band-width of filter.
1862 Specify the band-width of a filter in width_type units.
1865 Specify which channels to filter, by default all available are filtered.
1870 Apply a two-pole Butterworth band-reject filter with central
1871 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1872 The filter roll off at 6dB per octave (20dB per decade).
1874 The filter accepts the following options:
1878 Set the filter's central frequency. Default is @code{3000}.
1881 Set method to specify band-width of filter.
1894 Specify the band-width of a filter in width_type units.
1897 Specify which channels to filter, by default all available are filtered.
1902 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1903 shelving filter with a response similar to that of a standard
1904 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1906 The filter accepts the following options:
1910 Give the gain at 0 Hz. Its useful range is about -20
1911 (for a large cut) to +20 (for a large boost).
1912 Beware of clipping when using a positive gain.
1915 Set the filter's central frequency and so can be used
1916 to extend or reduce the frequency range to be boosted or cut.
1917 The default value is @code{100} Hz.
1920 Set method to specify band-width of filter.
1933 Determine how steep is the filter's shelf transition.
1936 Specify which channels to filter, by default all available are filtered.
1941 Apply a biquad IIR filter with the given coefficients.
1942 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1943 are the numerator and denominator coefficients respectively.
1944 and @var{channels}, @var{c} specify which channels to filter, by default all
1945 available are filtered.
1948 Bauer stereo to binaural transformation, which improves headphone listening of
1949 stereo audio records.
1951 To enable compilation of this filter you need to configure FFmpeg with
1952 @code{--enable-libbs2b}.
1954 It accepts the following parameters:
1958 Pre-defined crossfeed level.
1962 Default level (fcut=700, feed=50).
1965 Chu Moy circuit (fcut=700, feed=60).
1968 Jan Meier circuit (fcut=650, feed=95).
1973 Cut frequency (in Hz).
1982 Remap input channels to new locations.
1984 It accepts the following parameters:
1987 Map channels from input to output. The argument is a '|'-separated list of
1988 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1989 @var{in_channel} form. @var{in_channel} can be either the name of the input
1990 channel (e.g. FL for front left) or its index in the input channel layout.
1991 @var{out_channel} is the name of the output channel or its index in the output
1992 channel layout. If @var{out_channel} is not given then it is implicitly an
1993 index, starting with zero and increasing by one for each mapping.
1995 @item channel_layout
1996 The channel layout of the output stream.
1999 If no mapping is present, the filter will implicitly map input channels to
2000 output channels, preserving indices.
2002 For example, assuming a 5.1+downmix input MOV file,
2004 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2006 will create an output WAV file tagged as stereo from the downmix channels of
2009 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2011 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2014 @section channelsplit
2016 Split each channel from an input audio stream into a separate output stream.
2018 It accepts the following parameters:
2020 @item channel_layout
2021 The channel layout of the input stream. The default is "stereo".
2024 For example, assuming a stereo input MP3 file,
2026 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2028 will create an output Matroska file with two audio streams, one containing only
2029 the left channel and the other the right channel.
2031 Split a 5.1 WAV file into per-channel files:
2033 ffmpeg -i in.wav -filter_complex
2034 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2035 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2036 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2041 Add a chorus effect to the audio.
2043 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2045 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2046 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2047 The modulation depth defines the range the modulated delay is played before or after
2048 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2049 sound tuned around the original one, like in a chorus where some vocals are slightly
2052 It accepts the following parameters:
2055 Set input gain. Default is 0.4.
2058 Set output gain. Default is 0.4.
2061 Set delays. A typical delay is around 40ms to 60ms.
2073 @subsection Examples
2079 chorus=0.7:0.9:55:0.4:0.25:2
2085 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2089 Fuller sounding chorus with three delays:
2091 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
2096 Compress or expand the audio's dynamic range.
2098 It accepts the following parameters:
2104 A list of times in seconds for each channel over which the instantaneous level
2105 of the input signal is averaged to determine its volume. @var{attacks} refers to
2106 increase of volume and @var{decays} refers to decrease of volume. For most
2107 situations, the attack time (response to the audio getting louder) should be
2108 shorter than the decay time, because the human ear is more sensitive to sudden
2109 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2110 a typical value for decay is 0.8 seconds.
2111 If specified number of attacks & decays is lower than number of channels, the last
2112 set attack/decay will be used for all remaining channels.
2115 A list of points for the transfer function, specified in dB relative to the
2116 maximum possible signal amplitude. Each key points list must be defined using
2117 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2118 @code{x0/y0 x1/y1 x2/y2 ....}
2120 The input values must be in strictly increasing order but the transfer function
2121 does not have to be monotonically rising. The point @code{0/0} is assumed but
2122 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2123 function are @code{-70/-70|-60/-20|1/0}.
2126 Set the curve radius in dB for all joints. It defaults to 0.01.
2129 Set the additional gain in dB to be applied at all points on the transfer
2130 function. This allows for easy adjustment of the overall gain.
2134 Set an initial volume, in dB, to be assumed for each channel when filtering
2135 starts. This permits the user to supply a nominal level initially, so that, for
2136 example, a very large gain is not applied to initial signal levels before the
2137 companding has begun to operate. A typical value for audio which is initially
2138 quiet is -90 dB. It defaults to 0.
2141 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2142 delayed before being fed to the volume adjuster. Specifying a delay
2143 approximately equal to the attack/decay times allows the filter to effectively
2144 operate in predictive rather than reactive mode. It defaults to 0.
2148 @subsection Examples
2152 Make music with both quiet and loud passages suitable for listening to in a
2155 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2158 Another example for audio with whisper and explosion parts:
2160 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2164 A noise gate for when the noise is at a lower level than the signal:
2166 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2170 Here is another noise gate, this time for when the noise is at a higher level
2171 than the signal (making it, in some ways, similar to squelch):
2173 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2177 2:1 compression starting at -6dB:
2179 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2183 2:1 compression starting at -9dB:
2185 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2189 2:1 compression starting at -12dB:
2191 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2195 2:1 compression starting at -18dB:
2197 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2201 3:1 compression starting at -15dB:
2203 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2209 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2215 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
2219 Hard limiter at -6dB:
2221 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2225 Hard limiter at -12dB:
2227 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2231 Hard noise gate at -35 dB:
2233 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2239 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2243 @section compensationdelay
2245 Compensation Delay Line is a metric based delay to compensate differing
2246 positions of microphones or speakers.
2248 For example, you have recorded guitar with two microphones placed in
2249 different location. Because the front of sound wave has fixed speed in
2250 normal conditions, the phasing of microphones can vary and depends on
2251 their location and interposition. The best sound mix can be achieved when
2252 these microphones are in phase (synchronized). Note that distance of
2253 ~30 cm between microphones makes one microphone to capture signal in
2254 antiphase to another microphone. That makes the final mix sounding moody.
2255 This filter helps to solve phasing problems by adding different delays
2256 to each microphone track and make them synchronized.
2258 The best result can be reached when you take one track as base and
2259 synchronize other tracks one by one with it.
2260 Remember that synchronization/delay tolerance depends on sample rate, too.
2261 Higher sample rates will give more tolerance.
2263 It accepts the following parameters:
2267 Set millimeters distance. This is compensation distance for fine tuning.
2271 Set cm distance. This is compensation distance for tightening distance setup.
2275 Set meters distance. This is compensation distance for hard distance setup.
2279 Set dry amount. Amount of unprocessed (dry) signal.
2283 Set wet amount. Amount of processed (wet) signal.
2287 Set temperature degree in Celsius. This is the temperature of the environment.
2292 Apply headphone crossfeed filter.
2294 Crossfeed is the process of blending the left and right channels of stereo
2296 It is mainly used to reduce extreme stereo separation of low frequencies.
2298 The intent is to produce more speaker like sound to the listener.
2300 The filter accepts the following options:
2304 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2305 This sets gain of low shelf filter for side part of stereo image.
2306 Default is -6dB. Max allowed is -30db when strength is set to 1.
2309 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2310 This sets cut off frequency of low shelf filter. Default is cut off near
2311 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2314 Set input gain. Default is 0.9.
2317 Set output gain. Default is 1.
2320 @section crystalizer
2321 Simple algorithm to expand audio dynamic range.
2323 The filter accepts the following options:
2327 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2328 (unchanged sound) to 10.0 (maximum effect).
2331 Enable clipping. By default is enabled.
2335 Apply a DC shift to the audio.
2337 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2338 in the recording chain) from the audio. The effect of a DC offset is reduced
2339 headroom and hence volume. The @ref{astats} filter can be used to determine if
2340 a signal has a DC offset.
2344 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2348 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2349 used to prevent clipping.
2353 Dynamic Audio Normalizer.
2355 This filter applies a certain amount of gain to the input audio in order
2356 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2357 contrast to more "simple" normalization algorithms, the Dynamic Audio
2358 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2359 This allows for applying extra gain to the "quiet" sections of the audio
2360 while avoiding distortions or clipping the "loud" sections. In other words:
2361 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2362 sections, in the sense that the volume of each section is brought to the
2363 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2364 this goal *without* applying "dynamic range compressing". It will retain 100%
2365 of the dynamic range *within* each section of the audio file.
2369 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2370 Default is 500 milliseconds.
2371 The Dynamic Audio Normalizer processes the input audio in small chunks,
2372 referred to as frames. This is required, because a peak magnitude has no
2373 meaning for just a single sample value. Instead, we need to determine the
2374 peak magnitude for a contiguous sequence of sample values. While a "standard"
2375 normalizer would simply use the peak magnitude of the complete file, the
2376 Dynamic Audio Normalizer determines the peak magnitude individually for each
2377 frame. The length of a frame is specified in milliseconds. By default, the
2378 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2379 been found to give good results with most files.
2380 Note that the exact frame length, in number of samples, will be determined
2381 automatically, based on the sampling rate of the individual input audio file.
2384 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2385 number. Default is 31.
2386 Probably the most important parameter of the Dynamic Audio Normalizer is the
2387 @code{window size} of the Gaussian smoothing filter. The filter's window size
2388 is specified in frames, centered around the current frame. For the sake of
2389 simplicity, this must be an odd number. Consequently, the default value of 31
2390 takes into account the current frame, as well as the 15 preceding frames and
2391 the 15 subsequent frames. Using a larger window results in a stronger
2392 smoothing effect and thus in less gain variation, i.e. slower gain
2393 adaptation. Conversely, using a smaller window results in a weaker smoothing
2394 effect and thus in more gain variation, i.e. faster gain adaptation.
2395 In other words, the more you increase this value, the more the Dynamic Audio
2396 Normalizer will behave like a "traditional" normalization filter. On the
2397 contrary, the more you decrease this value, the more the Dynamic Audio
2398 Normalizer will behave like a dynamic range compressor.
2401 Set the target peak value. This specifies the highest permissible magnitude
2402 level for the normalized audio input. This filter will try to approach the
2403 target peak magnitude as closely as possible, but at the same time it also
2404 makes sure that the normalized signal will never exceed the peak magnitude.
2405 A frame's maximum local gain factor is imposed directly by the target peak
2406 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2407 It is not recommended to go above this value.
2410 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2411 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2412 factor for each input frame, i.e. the maximum gain factor that does not
2413 result in clipping or distortion. The maximum gain factor is determined by
2414 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2415 additionally bounds the frame's maximum gain factor by a predetermined
2416 (global) maximum gain factor. This is done in order to avoid excessive gain
2417 factors in "silent" or almost silent frames. By default, the maximum gain
2418 factor is 10.0, For most inputs the default value should be sufficient and
2419 it usually is not recommended to increase this value. Though, for input
2420 with an extremely low overall volume level, it may be necessary to allow even
2421 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2422 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2423 Instead, a "sigmoid" threshold function will be applied. This way, the
2424 gain factors will smoothly approach the threshold value, but never exceed that
2428 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2429 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2430 This means that the maximum local gain factor for each frame is defined
2431 (only) by the frame's highest magnitude sample. This way, the samples can
2432 be amplified as much as possible without exceeding the maximum signal
2433 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2434 Normalizer can also take into account the frame's root mean square,
2435 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2436 determine the power of a time-varying signal. It is therefore considered
2437 that the RMS is a better approximation of the "perceived loudness" than
2438 just looking at the signal's peak magnitude. Consequently, by adjusting all
2439 frames to a constant RMS value, a uniform "perceived loudness" can be
2440 established. If a target RMS value has been specified, a frame's local gain
2441 factor is defined as the factor that would result in exactly that RMS value.
2442 Note, however, that the maximum local gain factor is still restricted by the
2443 frame's highest magnitude sample, in order to prevent clipping.
2446 Enable channels coupling. By default is enabled.
2447 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2448 amount. This means the same gain factor will be applied to all channels, i.e.
2449 the maximum possible gain factor is determined by the "loudest" channel.
2450 However, in some recordings, it may happen that the volume of the different
2451 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2452 In this case, this option can be used to disable the channel coupling. This way,
2453 the gain factor will be determined independently for each channel, depending
2454 only on the individual channel's highest magnitude sample. This allows for
2455 harmonizing the volume of the different channels.
2458 Enable DC bias correction. By default is disabled.
2459 An audio signal (in the time domain) is a sequence of sample values.
2460 In the Dynamic Audio Normalizer these sample values are represented in the
2461 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2462 audio signal, or "waveform", should be centered around the zero point.
2463 That means if we calculate the mean value of all samples in a file, or in a
2464 single frame, then the result should be 0.0 or at least very close to that
2465 value. If, however, there is a significant deviation of the mean value from
2466 0.0, in either positive or negative direction, this is referred to as a
2467 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2468 Audio Normalizer provides optional DC bias correction.
2469 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2470 the mean value, or "DC correction" offset, of each input frame and subtract
2471 that value from all of the frame's sample values which ensures those samples
2472 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2473 boundaries, the DC correction offset values will be interpolated smoothly
2474 between neighbouring frames.
2477 Enable alternative boundary mode. By default is disabled.
2478 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2479 around each frame. This includes the preceding frames as well as the
2480 subsequent frames. However, for the "boundary" frames, located at the very
2481 beginning and at the very end of the audio file, not all neighbouring
2482 frames are available. In particular, for the first few frames in the audio
2483 file, the preceding frames are not known. And, similarly, for the last few
2484 frames in the audio file, the subsequent frames are not known. Thus, the
2485 question arises which gain factors should be assumed for the missing frames
2486 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2487 to deal with this situation. The default boundary mode assumes a gain factor
2488 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2489 "fade out" at the beginning and at the end of the input, respectively.
2492 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2493 By default, the Dynamic Audio Normalizer does not apply "traditional"
2494 compression. This means that signal peaks will not be pruned and thus the
2495 full dynamic range will be retained within each local neighbourhood. However,
2496 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2497 normalization algorithm with a more "traditional" compression.
2498 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2499 (thresholding) function. If (and only if) the compression feature is enabled,
2500 all input frames will be processed by a soft knee thresholding function prior
2501 to the actual normalization process. Put simply, the thresholding function is
2502 going to prune all samples whose magnitude exceeds a certain threshold value.
2503 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2504 value. Instead, the threshold value will be adjusted for each individual
2506 In general, smaller parameters result in stronger compression, and vice versa.
2507 Values below 3.0 are not recommended, because audible distortion may appear.
2512 Make audio easier to listen to on headphones.
2514 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2515 so that when listened to on headphones the stereo image is moved from
2516 inside your head (standard for headphones) to outside and in front of
2517 the listener (standard for speakers).
2523 Apply a two-pole peaking equalisation (EQ) filter. With this
2524 filter, the signal-level at and around a selected frequency can
2525 be increased or decreased, whilst (unlike bandpass and bandreject
2526 filters) that at all other frequencies is unchanged.
2528 In order to produce complex equalisation curves, this filter can
2529 be given several times, each with a different central frequency.
2531 The filter accepts the following options:
2535 Set the filter's central frequency in Hz.
2538 Set method to specify band-width of filter.
2551 Specify the band-width of a filter in width_type units.
2554 Set the required gain or attenuation in dB.
2555 Beware of clipping when using a positive gain.
2558 Specify which channels to filter, by default all available are filtered.
2561 @subsection Examples
2564 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2566 equalizer=f=1000:t=h:width=200:g=-10
2570 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2572 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2576 @section extrastereo
2578 Linearly increases the difference between left and right channels which
2579 adds some sort of "live" effect to playback.
2581 The filter accepts the following options:
2585 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2586 (average of both channels), with 1.0 sound will be unchanged, with
2587 -1.0 left and right channels will be swapped.
2590 Enable clipping. By default is enabled.
2593 @section firequalizer
2594 Apply FIR Equalization using arbitrary frequency response.
2596 The filter accepts the following option:
2600 Set gain curve equation (in dB). The expression can contain variables:
2603 the evaluated frequency
2607 channel number, set to 0 when multichannels evaluation is disabled
2609 channel id, see libavutil/channel_layout.h, set to the first channel id when
2610 multichannels evaluation is disabled
2614 channel_layout, see libavutil/channel_layout.h
2619 @item gain_interpolate(f)
2620 interpolate gain on frequency f based on gain_entry
2621 @item cubic_interpolate(f)
2622 same as gain_interpolate, but smoother
2624 This option is also available as command. Default is @code{gain_interpolate(f)}.
2627 Set gain entry for gain_interpolate function. The expression can
2631 store gain entry at frequency f with value g
2633 This option is also available as command.
2636 Set filter delay in seconds. Higher value means more accurate.
2637 Default is @code{0.01}.
2640 Set filter accuracy in Hz. Lower value means more accurate.
2641 Default is @code{5}.
2644 Set window function. Acceptable values are:
2647 rectangular window, useful when gain curve is already smooth
2649 hann window (default)
2655 3-terms continuous 1st derivative nuttall window
2657 minimum 3-terms discontinuous nuttall window
2659 4-terms continuous 1st derivative nuttall window
2661 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2663 blackman-harris window
2669 If enabled, use fixed number of audio samples. This improves speed when
2670 filtering with large delay. Default is disabled.
2673 Enable multichannels evaluation on gain. Default is disabled.
2676 Enable zero phase mode by subtracting timestamp to compensate delay.
2677 Default is disabled.
2680 Set scale used by gain. Acceptable values are:
2683 linear frequency, linear gain
2685 linear frequency, logarithmic (in dB) gain (default)
2687 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2689 logarithmic frequency, logarithmic gain
2693 Set file for dumping, suitable for gnuplot.
2696 Set scale for dumpfile. Acceptable values are same with scale option.
2700 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2701 Default is disabled.
2704 Enable minimum phase impulse response. Default is disabled.
2707 @subsection Examples
2712 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2715 lowpass at 1000 Hz with gain_entry:
2717 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2720 custom equalization:
2722 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2725 higher delay with zero phase to compensate delay:
2727 firequalizer=delay=0.1:fixed=on:zero_phase=on
2730 lowpass on left channel, highpass on right channel:
2732 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2733 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2738 Apply a flanging effect to the audio.
2740 The filter accepts the following options:
2744 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2747 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
2750 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2754 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2755 Default value is 71.
2758 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2761 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2762 Default value is @var{sinusoidal}.
2765 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2766 Default value is 25.
2769 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2770 Default is @var{linear}.
2774 Apply Haas effect to audio.
2776 Note that this makes most sense to apply on mono signals.
2777 With this filter applied to mono signals it give some directionality and
2778 stretches its stereo image.
2780 The filter accepts the following options:
2784 Set input level. By default is @var{1}, or 0dB
2787 Set output level. By default is @var{1}, or 0dB.
2790 Set gain applied to side part of signal. By default is @var{1}.
2793 Set kind of middle source. Can be one of the following:
2803 Pick middle part signal of stereo image.
2806 Pick side part signal of stereo image.
2810 Change middle phase. By default is disabled.
2813 Set left channel delay. By default is @var{2.05} milliseconds.
2816 Set left channel balance. By default is @var{-1}.
2819 Set left channel gain. By default is @var{1}.
2822 Change left phase. By default is disabled.
2825 Set right channel delay. By defaults is @var{2.12} milliseconds.
2828 Set right channel balance. By default is @var{1}.
2831 Set right channel gain. By default is @var{1}.
2834 Change right phase. By default is enabled.
2839 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2840 embedded HDCD codes is expanded into a 20-bit PCM stream.
2842 The filter supports the Peak Extend and Low-level Gain Adjustment features
2843 of HDCD, and detects the Transient Filter flag.
2846 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2849 When using the filter with wav, note the default encoding for wav is 16-bit,
2850 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2851 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2853 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2854 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2857 The filter accepts the following options:
2860 @item disable_autoconvert
2861 Disable any automatic format conversion or resampling in the filter graph.
2863 @item process_stereo
2864 Process the stereo channels together. If target_gain does not match between
2865 channels, consider it invalid and use the last valid target_gain.
2868 Set the code detect timer period in ms.
2871 Always extend peaks above -3dBFS even if PE isn't signaled.
2874 Replace audio with a solid tone and adjust the amplitude to signal some
2875 specific aspect of the decoding process. The output file can be loaded in
2876 an audio editor alongside the original to aid analysis.
2878 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2885 Gain adjustment level at each sample
2887 Samples where peak extend occurs
2889 Samples where the code detect timer is active
2891 Samples where the target gain does not match between channels
2897 Apply head-related transfer functions (HRTFs) to create virtual
2898 loudspeakers around the user for binaural listening via headphones.
2899 The HRIRs are provided via additional streams, for each channel
2900 one stereo input stream is needed.
2902 The filter accepts the following options:
2906 Set mapping of input streams for convolution.
2907 The argument is a '|'-separated list of channel names in order as they
2908 are given as additional stream inputs for filter.
2909 This also specify number of input streams. Number of input streams
2910 must be not less than number of channels in first stream plus one.
2913 Set gain applied to audio. Value is in dB. Default is 0.
2916 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
2917 processing audio in time domain which is slow.
2918 @var{freq} is processing audio in frequency domain which is fast.
2919 Default is @var{freq}.
2922 Set custom gain for LFE channels. Value is in dB. Default is 0.
2925 @subsection Examples
2929 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
2930 each amovie filter use stereo file with IR coefficients as input.
2931 The files give coefficients for each position of virtual loudspeaker:
2933 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"
2940 Apply a high-pass filter with 3dB point frequency.
2941 The filter can be either single-pole, or double-pole (the default).
2942 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2944 The filter accepts the following options:
2948 Set frequency in Hz. Default is 3000.
2951 Set number of poles. Default is 2.
2954 Set method to specify band-width of filter.
2967 Specify the band-width of a filter in width_type units.
2968 Applies only to double-pole filter.
2969 The default is 0.707q and gives a Butterworth response.
2972 Specify which channels to filter, by default all available are filtered.
2977 Join multiple input streams into one multi-channel stream.
2979 It accepts the following parameters:
2983 The number of input streams. It defaults to 2.
2985 @item channel_layout
2986 The desired output channel layout. It defaults to stereo.
2989 Map channels from inputs to output. The argument is a '|'-separated list of
2990 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2991 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2992 can be either the name of the input channel (e.g. FL for front left) or its
2993 index in the specified input stream. @var{out_channel} is the name of the output
2997 The filter will attempt to guess the mappings when they are not specified
2998 explicitly. It does so by first trying to find an unused matching input channel
2999 and if that fails it picks the first unused input channel.
3001 Join 3 inputs (with properly set channel layouts):
3003 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3006 Build a 5.1 output from 6 single-channel streams:
3008 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3009 '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'
3015 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3017 To enable compilation of this filter you need to configure FFmpeg with
3018 @code{--enable-ladspa}.
3022 Specifies the name of LADSPA plugin library to load. If the environment
3023 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3024 each one of the directories specified by the colon separated list in
3025 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3026 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3027 @file{/usr/lib/ladspa/}.
3030 Specifies the plugin within the library. Some libraries contain only
3031 one plugin, but others contain many of them. If this is not set filter
3032 will list all available plugins within the specified library.
3035 Set the '|' separated list of controls which are zero or more floating point
3036 values that determine the behavior of the loaded plugin (for example delay,
3038 Controls need to be defined using the following syntax:
3039 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3040 @var{valuei} is the value set on the @var{i}-th control.
3041 Alternatively they can be also defined using the following syntax:
3042 @var{value0}|@var{value1}|@var{value2}|..., where
3043 @var{valuei} is the value set on the @var{i}-th control.
3044 If @option{controls} is set to @code{help}, all available controls and
3045 their valid ranges are printed.
3047 @item sample_rate, s
3048 Specify the sample rate, default to 44100. Only used if plugin have
3052 Set the number of samples per channel per each output frame, default
3053 is 1024. Only used if plugin have zero inputs.
3056 Set the minimum duration of the sourced audio. See
3057 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3058 for the accepted syntax.
3059 Note that the resulting duration may be greater than the specified duration,
3060 as the generated audio is always cut at the end of a complete frame.
3061 If not specified, or the expressed duration is negative, the audio is
3062 supposed to be generated forever.
3063 Only used if plugin have zero inputs.
3067 @subsection Examples
3071 List all available plugins within amp (LADSPA example plugin) library:
3077 List all available controls and their valid ranges for @code{vcf_notch}
3078 plugin from @code{VCF} library:
3080 ladspa=f=vcf:p=vcf_notch:c=help
3084 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3087 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3091 Add reverberation to the audio using TAP-plugins
3092 (Tom's Audio Processing plugins):
3094 ladspa=file=tap_reverb:tap_reverb
3098 Generate white noise, with 0.2 amplitude:
3100 ladspa=file=cmt:noise_source_white:c=c0=.2
3104 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3105 @code{C* Audio Plugin Suite} (CAPS) library:
3107 ladspa=file=caps:Click:c=c1=20'
3111 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3113 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3117 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3118 @code{SWH Plugins} collection:
3120 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3124 Attenuate low frequencies using Multiband EQ from Steve Harris
3125 @code{SWH Plugins} collection:
3127 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3131 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3134 ladspa=caps:Narrower
3138 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3140 ladspa=caps:White:.2
3144 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3146 ladspa=caps:Fractal:c=c1=1
3150 Dynamic volume normalization using @code{VLevel} plugin:
3152 ladspa=vlevel-ladspa:vlevel_mono
3156 @subsection Commands
3158 This filter supports the following commands:
3161 Modify the @var{N}-th control value.
3163 If the specified value is not valid, it is ignored and prior one is kept.
3168 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3169 Support for both single pass (livestreams, files) and double pass (files) modes.
3170 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3171 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3172 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3174 The filter accepts the following options:
3178 Set integrated loudness target.
3179 Range is -70.0 - -5.0. Default value is -24.0.
3182 Set loudness range target.
3183 Range is 1.0 - 20.0. Default value is 7.0.
3186 Set maximum true peak.
3187 Range is -9.0 - +0.0. Default value is -2.0.
3189 @item measured_I, measured_i
3190 Measured IL of input file.
3191 Range is -99.0 - +0.0.
3193 @item measured_LRA, measured_lra
3194 Measured LRA of input file.
3195 Range is 0.0 - 99.0.
3197 @item measured_TP, measured_tp
3198 Measured true peak of input file.
3199 Range is -99.0 - +99.0.
3201 @item measured_thresh
3202 Measured threshold of input file.
3203 Range is -99.0 - +0.0.
3206 Set offset gain. Gain is applied before the true-peak limiter.
3207 Range is -99.0 - +99.0. Default is +0.0.
3210 Normalize linearly if possible.
3211 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3212 to be specified in order to use this mode.
3213 Options are true or false. Default is true.
3216 Treat mono input files as "dual-mono". If a mono file is intended for playback
3217 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3218 If set to @code{true}, this option will compensate for this effect.
3219 Multi-channel input files are not affected by this option.
3220 Options are true or false. Default is false.
3223 Set print format for stats. Options are summary, json, or none.
3224 Default value is none.
3229 Apply a low-pass filter with 3dB point frequency.
3230 The filter can be either single-pole or double-pole (the default).
3231 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3233 The filter accepts the following options:
3237 Set frequency in Hz. Default is 500.
3240 Set number of poles. Default is 2.
3243 Set method to specify band-width of filter.
3256 Specify the band-width of a filter in width_type units.
3257 Applies only to double-pole filter.
3258 The default is 0.707q and gives a Butterworth response.
3261 Specify which channels to filter, by default all available are filtered.
3264 @subsection Examples
3267 Lowpass only LFE channel, it LFE is not present it does nothing:
3276 Mix channels with specific gain levels. The filter accepts the output
3277 channel layout followed by a set of channels definitions.
3279 This filter is also designed to efficiently remap the channels of an audio
3282 The filter accepts parameters of the form:
3283 "@var{l}|@var{outdef}|@var{outdef}|..."
3287 output channel layout or number of channels
3290 output channel specification, of the form:
3291 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3294 output channel to define, either a channel name (FL, FR, etc.) or a channel
3295 number (c0, c1, etc.)
3298 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3301 input channel to use, see out_name for details; it is not possible to mix
3302 named and numbered input channels
3305 If the `=' in a channel specification is replaced by `<', then the gains for
3306 that specification will be renormalized so that the total is 1, thus
3307 avoiding clipping noise.
3309 @subsection Mixing examples
3311 For example, if you want to down-mix from stereo to mono, but with a bigger
3312 factor for the left channel:
3314 pan=1c|c0=0.9*c0+0.1*c1
3317 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3318 7-channels surround:
3320 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3323 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3324 that should be preferred (see "-ac" option) unless you have very specific
3327 @subsection Remapping examples
3329 The channel remapping will be effective if, and only if:
3332 @item gain coefficients are zeroes or ones,
3333 @item only one input per channel output,
3336 If all these conditions are satisfied, the filter will notify the user ("Pure
3337 channel mapping detected"), and use an optimized and lossless method to do the
3340 For example, if you have a 5.1 source and want a stereo audio stream by
3341 dropping the extra channels:
3343 pan="stereo| c0=FL | c1=FR"
3346 Given the same source, you can also switch front left and front right channels
3347 and keep the input channel layout:
3349 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3352 If the input is a stereo audio stream, you can mute the front left channel (and
3353 still keep the stereo channel layout) with:
3358 Still with a stereo audio stream input, you can copy the right channel in both
3359 front left and right:
3361 pan="stereo| c0=FR | c1=FR"
3366 ReplayGain scanner filter. This filter takes an audio stream as an input and
3367 outputs it unchanged.
3368 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3372 Convert the audio sample format, sample rate and channel layout. It is
3373 not meant to be used directly.
3376 Apply time-stretching and pitch-shifting with librubberband.
3378 The filter accepts the following options:
3382 Set tempo scale factor.
3385 Set pitch scale factor.
3388 Set transients detector.
3389 Possible values are:
3398 Possible values are:
3407 Possible values are:
3414 Set processing window size.
3415 Possible values are:
3424 Possible values are:
3431 Enable formant preservation when shift pitching.
3432 Possible values are:
3440 Possible values are:
3449 Possible values are:
3456 @section sidechaincompress
3458 This filter acts like normal compressor but has the ability to compress
3459 detected signal using second input signal.
3460 It needs two input streams and returns one output stream.
3461 First input stream will be processed depending on second stream signal.
3462 The filtered signal then can be filtered with other filters in later stages of
3463 processing. See @ref{pan} and @ref{amerge} filter.
3465 The filter accepts the following options:
3469 Set input gain. Default is 1. Range is between 0.015625 and 64.
3472 If a signal of second stream raises above this level it will affect the gain
3473 reduction of first stream.
3474 By default is 0.125. Range is between 0.00097563 and 1.
3477 Set a ratio about which the signal is reduced. 1:2 means that if the level
3478 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3479 Default is 2. Range is between 1 and 20.
3482 Amount of milliseconds the signal has to rise above the threshold before gain
3483 reduction starts. Default is 20. Range is between 0.01 and 2000.
3486 Amount of milliseconds the signal has to fall below the threshold before
3487 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3490 Set the amount by how much signal will be amplified after processing.
3491 Default is 1. Range is from 1 to 64.
3494 Curve the sharp knee around the threshold to enter gain reduction more softly.
3495 Default is 2.82843. Range is between 1 and 8.
3498 Choose if the @code{average} level between all channels of side-chain stream
3499 or the louder(@code{maximum}) channel of side-chain stream affects the
3500 reduction. Default is @code{average}.
3503 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3504 of @code{rms}. Default is @code{rms} which is mainly smoother.
3507 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3510 How much to use compressed signal in output. Default is 1.
3511 Range is between 0 and 1.
3514 @subsection Examples
3518 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3519 depending on the signal of 2nd input and later compressed signal to be
3520 merged with 2nd input:
3522 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3526 @section sidechaingate
3528 A sidechain gate acts like a normal (wideband) gate but has the ability to
3529 filter the detected signal before sending it to the gain reduction stage.
3530 Normally a gate uses the full range signal to detect a level above the
3532 For example: If you cut all lower frequencies from your sidechain signal
3533 the gate will decrease the volume of your track only if not enough highs
3534 appear. With this technique you are able to reduce the resonation of a
3535 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3537 It needs two input streams and returns one output stream.
3538 First input stream will be processed depending on second stream signal.
3540 The filter accepts the following options:
3544 Set input level before filtering.
3545 Default is 1. Allowed range is from 0.015625 to 64.
3548 Set the level of gain reduction when the signal is below the threshold.
3549 Default is 0.06125. Allowed range is from 0 to 1.
3552 If a signal rises above this level the gain reduction is released.
3553 Default is 0.125. Allowed range is from 0 to 1.
3556 Set a ratio about which the signal is reduced.
3557 Default is 2. Allowed range is from 1 to 9000.
3560 Amount of milliseconds the signal has to rise above the threshold before gain
3562 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3565 Amount of milliseconds the signal has to fall below the threshold before the
3566 reduction is increased again. Default is 250 milliseconds.
3567 Allowed range is from 0.01 to 9000.
3570 Set amount of amplification of signal after processing.
3571 Default is 1. Allowed range is from 1 to 64.
3574 Curve the sharp knee around the threshold to enter gain reduction more softly.
3575 Default is 2.828427125. Allowed range is from 1 to 8.
3578 Choose if exact signal should be taken for detection or an RMS like one.
3579 Default is rms. Can be peak or rms.
3582 Choose if the average level between all channels or the louder channel affects
3584 Default is average. Can be average or maximum.
3587 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3590 @section silencedetect
3592 Detect silence in an audio stream.
3594 This filter logs a message when it detects that the input audio volume is less
3595 or equal to a noise tolerance value for a duration greater or equal to the
3596 minimum detected noise duration.
3598 The printed times and duration are expressed in seconds.
3600 The filter accepts the following options:
3604 Set silence duration until notification (default is 2 seconds).
3607 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3608 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3611 @subsection Examples
3615 Detect 5 seconds of silence with -50dB noise tolerance:
3617 silencedetect=n=-50dB:d=5
3621 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3622 tolerance in @file{silence.mp3}:
3624 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3628 @section silenceremove
3630 Remove silence from the beginning, middle or end of the audio.
3632 The filter accepts the following options:
3636 This value is used to indicate if audio should be trimmed at beginning of
3637 the audio. A value of zero indicates no silence should be trimmed from the
3638 beginning. When specifying a non-zero value, it trims audio up until it
3639 finds non-silence. Normally, when trimming silence from beginning of audio
3640 the @var{start_periods} will be @code{1} but it can be increased to higher
3641 values to trim all audio up to specific count of non-silence periods.
3642 Default value is @code{0}.
3644 @item start_duration
3645 Specify the amount of time that non-silence must be detected before it stops
3646 trimming audio. By increasing the duration, bursts of noises can be treated
3647 as silence and trimmed off. Default value is @code{0}.
3649 @item start_threshold
3650 This indicates what sample value should be treated as silence. For digital
3651 audio, a value of @code{0} may be fine but for audio recorded from analog,
3652 you may wish to increase the value to account for background noise.
3653 Can be specified in dB (in case "dB" is appended to the specified value)
3654 or amplitude ratio. Default value is @code{0}.
3657 Set the count for trimming silence from the end of audio.
3658 To remove silence from the middle of a file, specify a @var{stop_periods}
3659 that is negative. This value is then treated as a positive value and is
3660 used to indicate the effect should restart processing as specified by
3661 @var{start_periods}, making it suitable for removing periods of silence
3662 in the middle of the audio.
3663 Default value is @code{0}.
3666 Specify a duration of silence that must exist before audio is not copied any
3667 more. By specifying a higher duration, silence that is wanted can be left in
3669 Default value is @code{0}.
3671 @item stop_threshold
3672 This is the same as @option{start_threshold} but for trimming silence from
3674 Can be specified in dB (in case "dB" is appended to the specified value)
3675 or amplitude ratio. Default value is @code{0}.
3678 This indicates that @var{stop_duration} length of audio should be left intact
3679 at the beginning of each period of silence.
3680 For example, if you want to remove long pauses between words but do not want
3681 to remove the pauses completely. Default value is @code{0}.
3684 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3685 and works better with digital silence which is exactly 0.
3686 Default value is @code{rms}.
3689 Set ratio used to calculate size of window for detecting silence.
3690 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3693 @subsection Examples
3697 The following example shows how this filter can be used to start a recording
3698 that does not contain the delay at the start which usually occurs between
3699 pressing the record button and the start of the performance:
3701 silenceremove=1:5:0.02
3705 Trim all silence encountered from beginning to end where there is more than 1
3706 second of silence in audio:
3708 silenceremove=0:0:0:-1:1:-90dB
3714 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3715 loudspeakers around the user for binaural listening via headphones (audio
3716 formats up to 9 channels supported).
3717 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3718 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3719 Austrian Academy of Sciences.
3721 To enable compilation of this filter you need to configure FFmpeg with
3722 @code{--enable-libmysofa}.
3724 The filter accepts the following options:
3728 Set the SOFA file used for rendering.
3731 Set gain applied to audio. Value is in dB. Default is 0.
3734 Set rotation of virtual loudspeakers in deg. Default is 0.
3737 Set elevation of virtual speakers in deg. Default is 0.
3740 Set distance in meters between loudspeakers and the listener with near-field
3741 HRTFs. Default is 1.
3744 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3745 processing audio in time domain which is slow.
3746 @var{freq} is processing audio in frequency domain which is fast.
3747 Default is @var{freq}.
3750 Set custom positions of virtual loudspeakers. Syntax for this option is:
3751 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3752 Each virtual loudspeaker is described with short channel name following with
3753 azimuth and elevation in degrees.
3754 Each virtual loudspeaker description is separated by '|'.
3755 For example to override front left and front right channel positions use:
3756 'speakers=FL 45 15|FR 345 15'.
3757 Descriptions with unrecognised channel names are ignored.
3760 Set custom gain for LFE channels. Value is in dB. Default is 0.
3763 @subsection Examples
3767 Using ClubFritz6 sofa file:
3769 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3773 Using ClubFritz12 sofa file and bigger radius with small rotation:
3775 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3779 Similar as above but with custom speaker positions for front left, front right, back left and back right
3780 and also with custom gain:
3782 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
3786 @section stereotools
3788 This filter has some handy utilities to manage stereo signals, for converting
3789 M/S stereo recordings to L/R signal while having control over the parameters
3790 or spreading the stereo image of master track.
3792 The filter accepts the following options:
3796 Set input level before filtering for both channels. Defaults is 1.
3797 Allowed range is from 0.015625 to 64.
3800 Set output level after filtering for both channels. Defaults is 1.
3801 Allowed range is from 0.015625 to 64.
3804 Set input balance between both channels. Default is 0.
3805 Allowed range is from -1 to 1.
3808 Set output balance between both channels. Default is 0.
3809 Allowed range is from -1 to 1.
3812 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3813 clipping. Disabled by default.
3816 Mute the left channel. Disabled by default.
3819 Mute the right channel. Disabled by default.
3822 Change the phase of the left channel. Disabled by default.
3825 Change the phase of the right channel. Disabled by default.
3828 Set stereo mode. Available values are:
3832 Left/Right to Left/Right, this is default.
3835 Left/Right to Mid/Side.
3838 Mid/Side to Left/Right.
3841 Left/Right to Left/Left.
3844 Left/Right to Right/Right.
3847 Left/Right to Left + Right.
3850 Left/Right to Right/Left.
3853 Mid/Side to Left/Left.
3856 Mid/Side to Right/Right.
3860 Set level of side signal. Default is 1.
3861 Allowed range is from 0.015625 to 64.
3864 Set balance of side signal. Default is 0.
3865 Allowed range is from -1 to 1.
3868 Set level of the middle signal. Default is 1.
3869 Allowed range is from 0.015625 to 64.
3872 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3875 Set stereo base between mono and inversed channels. Default is 0.
3876 Allowed range is from -1 to 1.
3879 Set delay in milliseconds how much to delay left from right channel and
3880 vice versa. Default is 0. Allowed range is from -20 to 20.
3883 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3886 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3888 @item bmode_in, bmode_out
3889 Set balance mode for balance_in/balance_out option.
3891 Can be one of the following:
3895 Classic balance mode. Attenuate one channel at time.
3896 Gain is raised up to 1.
3899 Similar as classic mode above but gain is raised up to 2.
3902 Equal power distribution, from -6dB to +6dB range.
3906 @subsection Examples
3910 Apply karaoke like effect:
3912 stereotools=mlev=0.015625
3916 Convert M/S signal to L/R:
3918 "stereotools=mode=ms>lr"
3922 @section stereowiden
3924 This filter enhance the stereo effect by suppressing signal common to both
3925 channels and by delaying the signal of left into right and vice versa,
3926 thereby widening the stereo effect.
3928 The filter accepts the following options:
3932 Time in milliseconds of the delay of left signal into right and vice versa.
3933 Default is 20 milliseconds.
3936 Amount of gain in delayed signal into right and vice versa. Gives a delay
3937 effect of left signal in right output and vice versa which gives widening
3938 effect. Default is 0.3.
3941 Cross feed of left into right with inverted phase. This helps in suppressing
3942 the mono. If the value is 1 it will cancel all the signal common to both
3943 channels. Default is 0.3.
3946 Set level of input signal of original channel. Default is 0.8.
3949 @section superequalizer
3950 Apply 18 band equalizer.
3952 The filter accepts the following options:
3959 Set 131Hz band gain.
3961 Set 185Hz band gain.
3963 Set 262Hz band gain.
3965 Set 370Hz band gain.
3967 Set 523Hz band gain.
3969 Set 740Hz band gain.
3971 Set 1047Hz band gain.
3973 Set 1480Hz band gain.
3975 Set 2093Hz band gain.
3977 Set 2960Hz band gain.
3979 Set 4186Hz band gain.
3981 Set 5920Hz band gain.
3983 Set 8372Hz band gain.
3985 Set 11840Hz band gain.
3987 Set 16744Hz band gain.
3989 Set 20000Hz band gain.
3993 Apply audio surround upmix filter.
3995 This filter allows to produce multichannel output from audio stream.
3997 The filter accepts the following options:
4001 Set output channel layout. By default, this is @var{5.1}.
4003 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4004 for the required syntax.
4007 Set input channel layout. By default, this is @var{stereo}.
4009 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4010 for the required syntax.
4013 Set input volume level. By default, this is @var{1}.
4016 Set output volume level. By default, this is @var{1}.
4019 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4022 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4025 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4028 Set front center input volume. By default, this is @var{1}.
4031 Set front center output volume. By default, this is @var{1}.
4034 Set LFE input volume. By default, this is @var{1}.
4037 Set LFE output volume. By default, this is @var{1}.
4042 Boost or cut treble (upper) frequencies of the audio using a two-pole
4043 shelving filter with a response similar to that of a standard
4044 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4046 The filter accepts the following options:
4050 Give the gain at whichever is the lower of ~22 kHz and the
4051 Nyquist frequency. Its useful range is about -20 (for a large cut)
4052 to +20 (for a large boost). Beware of clipping when using a positive gain.
4055 Set the filter's central frequency and so can be used
4056 to extend or reduce the frequency range to be boosted or cut.
4057 The default value is @code{3000} Hz.
4060 Set method to specify band-width of filter.
4073 Determine how steep is the filter's shelf transition.
4076 Specify which channels to filter, by default all available are filtered.
4081 Sinusoidal amplitude modulation.
4083 The filter accepts the following options:
4087 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4088 (20 Hz or lower) will result in a tremolo effect.
4089 This filter may also be used as a ring modulator by specifying
4090 a modulation frequency higher than 20 Hz.
4091 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4094 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4095 Default value is 0.5.
4100 Sinusoidal phase modulation.
4102 The filter accepts the following options:
4106 Modulation frequency in Hertz.
4107 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4110 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4111 Default value is 0.5.
4116 Adjust the input audio volume.
4118 It accepts the following parameters:
4122 Set audio volume expression.
4124 Output values are clipped to the maximum value.
4126 The output audio volume is given by the relation:
4128 @var{output_volume} = @var{volume} * @var{input_volume}
4131 The default value for @var{volume} is "1.0".
4134 This parameter represents the mathematical precision.
4136 It determines which input sample formats will be allowed, which affects the
4137 precision of the volume scaling.
4141 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4143 32-bit floating-point; this limits input sample format to FLT. (default)
4145 64-bit floating-point; this limits input sample format to DBL.
4149 Choose the behaviour on encountering ReplayGain side data in input frames.
4153 Remove ReplayGain side data, ignoring its contents (the default).
4156 Ignore ReplayGain side data, but leave it in the frame.
4159 Prefer the track gain, if present.
4162 Prefer the album gain, if present.
4165 @item replaygain_preamp
4166 Pre-amplification gain in dB to apply to the selected replaygain gain.
4168 Default value for @var{replaygain_preamp} is 0.0.
4171 Set when the volume expression is evaluated.
4173 It accepts the following values:
4176 only evaluate expression once during the filter initialization, or
4177 when the @samp{volume} command is sent
4180 evaluate expression for each incoming frame
4183 Default value is @samp{once}.
4186 The volume expression can contain the following parameters.
4190 frame number (starting at zero)
4193 @item nb_consumed_samples
4194 number of samples consumed by the filter
4196 number of samples in the current frame
4198 original frame position in the file
4204 PTS at start of stream
4206 time at start of stream
4212 last set volume value
4215 Note that when @option{eval} is set to @samp{once} only the
4216 @var{sample_rate} and @var{tb} variables are available, all other
4217 variables will evaluate to NAN.
4219 @subsection Commands
4221 This filter supports the following commands:
4224 Modify the volume expression.
4225 The command accepts the same syntax of the corresponding option.
4227 If the specified expression is not valid, it is kept at its current
4229 @item replaygain_noclip
4230 Prevent clipping by limiting the gain applied.
4232 Default value for @var{replaygain_noclip} is 1.
4236 @subsection Examples
4240 Halve the input audio volume:
4244 volume=volume=-6.0206dB
4247 In all the above example the named key for @option{volume} can be
4248 omitted, for example like in:
4254 Increase input audio power by 6 decibels using fixed-point precision:
4256 volume=volume=6dB:precision=fixed
4260 Fade volume after time 10 with an annihilation period of 5 seconds:
4262 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4266 @section volumedetect
4268 Detect the volume of the input video.
4270 The filter has no parameters. The input is not modified. Statistics about
4271 the volume will be printed in the log when the input stream end is reached.
4273 In particular it will show the mean volume (root mean square), maximum
4274 volume (on a per-sample basis), and the beginning of a histogram of the
4275 registered volume values (from the maximum value to a cumulated 1/1000 of
4278 All volumes are in decibels relative to the maximum PCM value.
4280 @subsection Examples
4282 Here is an excerpt of the output:
4284 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4285 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4286 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4287 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4288 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4289 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4290 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4291 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4292 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4298 The mean square energy is approximately -27 dB, or 10^-2.7.
4300 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4302 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4305 In other words, raising the volume by +4 dB does not cause any clipping,
4306 raising it by +5 dB causes clipping for 6 samples, etc.
4308 @c man end AUDIO FILTERS
4310 @chapter Audio Sources
4311 @c man begin AUDIO SOURCES
4313 Below is a description of the currently available audio sources.
4317 Buffer audio frames, and make them available to the filter chain.
4319 This source is mainly intended for a programmatic use, in particular
4320 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4322 It accepts the following parameters:
4326 The timebase which will be used for timestamps of submitted frames. It must be
4327 either a floating-point number or in @var{numerator}/@var{denominator} form.
4330 The sample rate of the incoming audio buffers.
4333 The sample format of the incoming audio buffers.
4334 Either a sample format name or its corresponding integer representation from
4335 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4337 @item channel_layout
4338 The channel layout of the incoming audio buffers.
4339 Either a channel layout name from channel_layout_map in
4340 @file{libavutil/channel_layout.c} or its corresponding integer representation
4341 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4344 The number of channels of the incoming audio buffers.
4345 If both @var{channels} and @var{channel_layout} are specified, then they
4350 @subsection Examples
4353 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4356 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4357 Since the sample format with name "s16p" corresponds to the number
4358 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4361 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4366 Generate an audio signal specified by an expression.
4368 This source accepts in input one or more expressions (one for each
4369 channel), which are evaluated and used to generate a corresponding
4372 This source accepts the following options:
4376 Set the '|'-separated expressions list for each separate channel. In case the
4377 @option{channel_layout} option is not specified, the selected channel layout
4378 depends on the number of provided expressions. Otherwise the last
4379 specified expression is applied to the remaining output channels.
4381 @item channel_layout, c
4382 Set the channel layout. The number of channels in the specified layout
4383 must be equal to the number of specified expressions.
4386 Set the minimum duration of the sourced audio. See
4387 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4388 for the accepted syntax.
4389 Note that the resulting duration may be greater than the specified
4390 duration, as the generated audio is always cut at the end of a
4393 If not specified, or the expressed duration is negative, the audio is
4394 supposed to be generated forever.
4397 Set the number of samples per channel per each output frame,
4400 @item sample_rate, s
4401 Specify the sample rate, default to 44100.
4404 Each expression in @var{exprs} can contain the following constants:
4408 number of the evaluated sample, starting from 0
4411 time of the evaluated sample expressed in seconds, starting from 0
4418 @subsection Examples
4428 Generate a sin signal with frequency of 440 Hz, set sample rate to
4431 aevalsrc="sin(440*2*PI*t):s=8000"
4435 Generate a two channels signal, specify the channel layout (Front
4436 Center + Back Center) explicitly:
4438 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4442 Generate white noise:
4444 aevalsrc="-2+random(0)"
4448 Generate an amplitude modulated signal:
4450 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4454 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4456 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4463 The null audio source, return unprocessed audio frames. It is mainly useful
4464 as a template and to be employed in analysis / debugging tools, or as
4465 the source for filters which ignore the input data (for example the sox
4468 This source accepts the following options:
4472 @item channel_layout, cl
4474 Specifies the channel layout, and can be either an integer or a string
4475 representing a channel layout. The default value of @var{channel_layout}
4478 Check the channel_layout_map definition in
4479 @file{libavutil/channel_layout.c} for the mapping between strings and
4480 channel layout values.
4482 @item sample_rate, r
4483 Specifies the sample rate, and defaults to 44100.
4486 Set the number of samples per requested frames.
4490 @subsection Examples
4494 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4496 anullsrc=r=48000:cl=4
4500 Do the same operation with a more obvious syntax:
4502 anullsrc=r=48000:cl=mono
4506 All the parameters need to be explicitly defined.
4510 Synthesize a voice utterance using the libflite library.
4512 To enable compilation of this filter you need to configure FFmpeg with
4513 @code{--enable-libflite}.
4515 Note that the flite library is not thread-safe.
4517 The filter accepts the following options:
4522 If set to 1, list the names of the available voices and exit
4523 immediately. Default value is 0.
4526 Set the maximum number of samples per frame. Default value is 512.
4529 Set the filename containing the text to speak.
4532 Set the text to speak.
4535 Set the voice to use for the speech synthesis. Default value is
4536 @code{kal}. See also the @var{list_voices} option.
4539 @subsection Examples
4543 Read from file @file{speech.txt}, and synthesize the text using the
4544 standard flite voice:
4546 flite=textfile=speech.txt
4550 Read the specified text selecting the @code{slt} voice:
4552 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4556 Input text to ffmpeg:
4558 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4562 Make @file{ffplay} speak the specified text, using @code{flite} and
4563 the @code{lavfi} device:
4565 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4569 For more information about libflite, check:
4570 @url{http://www.speech.cs.cmu.edu/flite/}
4574 Generate a noise audio signal.
4576 The filter accepts the following options:
4579 @item sample_rate, r
4580 Specify the sample rate. Default value is 48000 Hz.
4583 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4587 Specify the duration of the generated audio stream. Not specifying this option
4588 results in noise with an infinite length.
4590 @item color, colour, c
4591 Specify the color of noise. Available noise colors are white, pink, brown,
4592 blue and violet. Default color is white.
4595 Specify a value used to seed the PRNG.
4598 Set the number of samples per each output frame, default is 1024.
4601 @subsection Examples
4606 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4608 anoisesrc=d=60:c=pink:r=44100:a=0.5
4614 Generate an audio signal made of a sine wave with amplitude 1/8.
4616 The audio signal is bit-exact.
4618 The filter accepts the following options:
4623 Set the carrier frequency. Default is 440 Hz.
4625 @item beep_factor, b
4626 Enable a periodic beep every second with frequency @var{beep_factor} times
4627 the carrier frequency. Default is 0, meaning the beep is disabled.
4629 @item sample_rate, r
4630 Specify the sample rate, default is 44100.
4633 Specify the duration of the generated audio stream.
4635 @item samples_per_frame
4636 Set the number of samples per output frame.
4638 The expression can contain the following constants:
4642 The (sequential) number of the output audio frame, starting from 0.
4645 The PTS (Presentation TimeStamp) of the output audio frame,
4646 expressed in @var{TB} units.
4649 The PTS of the output audio frame, expressed in seconds.
4652 The timebase of the output audio frames.
4655 Default is @code{1024}.
4658 @subsection Examples
4663 Generate a simple 440 Hz sine wave:
4669 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4673 sine=frequency=220:beep_factor=4:duration=5
4677 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4680 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4684 @c man end AUDIO SOURCES
4686 @chapter Audio Sinks
4687 @c man begin AUDIO SINKS
4689 Below is a description of the currently available audio sinks.
4691 @section abuffersink
4693 Buffer audio frames, and make them available to the end of filter chain.
4695 This sink is mainly intended for programmatic use, in particular
4696 through the interface defined in @file{libavfilter/buffersink.h}
4697 or the options system.
4699 It accepts a pointer to an AVABufferSinkContext structure, which
4700 defines the incoming buffers' formats, to be passed as the opaque
4701 parameter to @code{avfilter_init_filter} for initialization.
4704 Null audio sink; do absolutely nothing with the input audio. It is
4705 mainly useful as a template and for use in analysis / debugging
4708 @c man end AUDIO SINKS
4710 @chapter Video Filters
4711 @c man begin VIDEO FILTERS
4713 When you configure your FFmpeg build, you can disable any of the
4714 existing filters using @code{--disable-filters}.
4715 The configure output will show the video filters included in your
4718 Below is a description of the currently available video filters.
4720 @section alphaextract
4722 Extract the alpha component from the input as a grayscale video. This
4723 is especially useful with the @var{alphamerge} filter.
4727 Add or replace the alpha component of the primary input with the
4728 grayscale value of a second input. This is intended for use with
4729 @var{alphaextract} to allow the transmission or storage of frame
4730 sequences that have alpha in a format that doesn't support an alpha
4733 For example, to reconstruct full frames from a normal YUV-encoded video
4734 and a separate video created with @var{alphaextract}, you might use:
4736 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4739 Since this filter is designed for reconstruction, it operates on frame
4740 sequences without considering timestamps, and terminates when either
4741 input reaches end of stream. This will cause problems if your encoding
4742 pipeline drops frames. If you're trying to apply an image as an
4743 overlay to a video stream, consider the @var{overlay} filter instead.
4747 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4748 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4749 Substation Alpha) subtitles files.
4751 This filter accepts the following option in addition to the common options from
4752 the @ref{subtitles} filter:
4756 Set the shaping engine
4758 Available values are:
4761 The default libass shaping engine, which is the best available.
4763 Fast, font-agnostic shaper that can do only substitutions
4765 Slower shaper using OpenType for substitutions and positioning
4768 The default is @code{auto}.
4772 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4774 The filter accepts the following options:
4778 Set threshold A for 1st plane. Default is 0.02.
4779 Valid range is 0 to 0.3.
4782 Set threshold B for 1st plane. Default is 0.04.
4783 Valid range is 0 to 5.
4786 Set threshold A for 2nd plane. Default is 0.02.
4787 Valid range is 0 to 0.3.
4790 Set threshold B for 2nd plane. Default is 0.04.
4791 Valid range is 0 to 5.
4794 Set threshold A for 3rd plane. Default is 0.02.
4795 Valid range is 0 to 0.3.
4798 Set threshold B for 3rd plane. Default is 0.04.
4799 Valid range is 0 to 5.
4801 Threshold A is designed to react on abrupt changes in the input signal and
4802 threshold B is designed to react on continuous changes in the input signal.
4805 Set number of frames filter will use for averaging. Default is 33. Must be odd
4806 number in range [5, 129].
4809 Set what planes of frame filter will use for averaging. Default is all.
4814 Apply average blur filter.
4816 The filter accepts the following options:
4820 Set horizontal kernel size.
4823 Set which planes to filter. By default all planes are filtered.
4826 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4827 Default is @code{0}.
4832 Compute the bounding box for the non-black pixels in the input frame
4835 This filter computes the bounding box containing all the pixels with a
4836 luminance value greater than the minimum allowed value.
4837 The parameters describing the bounding box are printed on the filter
4840 The filter accepts the following option:
4844 Set the minimal luminance value. Default is @code{16}.
4847 @section bitplanenoise
4849 Show and measure bit plane noise.
4851 The filter accepts the following options:
4855 Set which plane to analyze. Default is @code{1}.
4858 Filter out noisy pixels from @code{bitplane} set above.
4859 Default is disabled.
4862 @section blackdetect
4864 Detect video intervals that are (almost) completely black. Can be
4865 useful to detect chapter transitions, commercials, or invalid
4866 recordings. Output lines contains the time for the start, end and
4867 duration of the detected black interval expressed in seconds.
4869 In order to display the output lines, you need to set the loglevel at
4870 least to the AV_LOG_INFO value.
4872 The filter accepts the following options:
4875 @item black_min_duration, d
4876 Set the minimum detected black duration expressed in seconds. It must
4877 be a non-negative floating point number.
4879 Default value is 2.0.
4881 @item picture_black_ratio_th, pic_th
4882 Set the threshold for considering a picture "black".
4883 Express the minimum value for the ratio:
4885 @var{nb_black_pixels} / @var{nb_pixels}
4888 for which a picture is considered black.
4889 Default value is 0.98.
4891 @item pixel_black_th, pix_th
4892 Set the threshold for considering a pixel "black".
4894 The threshold expresses the maximum pixel luminance value for which a
4895 pixel is considered "black". The provided value is scaled according to
4896 the following equation:
4898 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4901 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4902 the input video format, the range is [0-255] for YUV full-range
4903 formats and [16-235] for YUV non full-range formats.
4905 Default value is 0.10.
4908 The following example sets the maximum pixel threshold to the minimum
4909 value, and detects only black intervals of 2 or more seconds:
4911 blackdetect=d=2:pix_th=0.00
4916 Detect frames that are (almost) completely black. Can be useful to
4917 detect chapter transitions or commercials. Output lines consist of
4918 the frame number of the detected frame, the percentage of blackness,
4919 the position in the file if known or -1 and the timestamp in seconds.
4921 In order to display the output lines, you need to set the loglevel at
4922 least to the AV_LOG_INFO value.
4924 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
4925 The value represents the percentage of pixels in the picture that
4926 are below the threshold value.
4928 It accepts the following parameters:
4933 The percentage of the pixels that have to be below the threshold; it defaults to
4936 @item threshold, thresh
4937 The threshold below which a pixel value is considered black; it defaults to
4942 @section blend, tblend
4944 Blend two video frames into each other.
4946 The @code{blend} filter takes two input streams and outputs one
4947 stream, the first input is the "top" layer and second input is
4948 "bottom" layer. By default, the output terminates when the longest input terminates.
4950 The @code{tblend} (time blend) filter takes two consecutive frames
4951 from one single stream, and outputs the result obtained by blending
4952 the new frame on top of the old frame.
4954 A description of the accepted options follows.
4962 Set blend mode for specific pixel component or all pixel components in case
4963 of @var{all_mode}. Default value is @code{normal}.
4965 Available values for component modes are:
5007 Set blend opacity for specific pixel component or all pixel components in case
5008 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5015 Set blend expression for specific pixel component or all pixel components in case
5016 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5018 The expressions can use the following variables:
5022 The sequential number of the filtered frame, starting from @code{0}.
5026 the coordinates of the current sample
5030 the width and height of currently filtered plane
5034 Width and height scale depending on the currently filtered plane. It is the
5035 ratio between the corresponding luma plane number of pixels and the current
5036 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5037 @code{0.5,0.5} for chroma planes.
5040 Time of the current frame, expressed in seconds.
5043 Value of pixel component at current location for first video frame (top layer).
5046 Value of pixel component at current location for second video frame (bottom layer).
5050 The @code{blend} filter also supports the @ref{framesync} options.
5052 @subsection Examples
5056 Apply transition from bottom layer to top layer in first 10 seconds:
5058 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5062 Apply linear horizontal transition from top layer to bottom layer:
5064 blend=all_expr='A*(X/W)+B*(1-X/W)'
5068 Apply 1x1 checkerboard effect:
5070 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5074 Apply uncover left effect:
5076 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5080 Apply uncover down effect:
5082 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5086 Apply uncover up-left effect:
5088 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5092 Split diagonally video and shows top and bottom layer on each side:
5094 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5098 Display differences between the current and the previous frame:
5100 tblend=all_mode=grainextract
5106 Apply a boxblur algorithm to the input video.
5108 It accepts the following parameters:
5112 @item luma_radius, lr
5113 @item luma_power, lp
5114 @item chroma_radius, cr
5115 @item chroma_power, cp
5116 @item alpha_radius, ar
5117 @item alpha_power, ap
5121 A description of the accepted options follows.
5124 @item luma_radius, lr
5125 @item chroma_radius, cr
5126 @item alpha_radius, ar
5127 Set an expression for the box radius in pixels used for blurring the
5128 corresponding input plane.
5130 The radius value must be a non-negative number, and must not be
5131 greater than the value of the expression @code{min(w,h)/2} for the
5132 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5135 Default value for @option{luma_radius} is "2". If not specified,
5136 @option{chroma_radius} and @option{alpha_radius} default to the
5137 corresponding value set for @option{luma_radius}.
5139 The expressions can contain the following constants:
5143 The input width and height in pixels.
5147 The input chroma image width and height in pixels.
5151 The horizontal and vertical chroma subsample values. For example, for the
5152 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5155 @item luma_power, lp
5156 @item chroma_power, cp
5157 @item alpha_power, ap
5158 Specify how many times the boxblur filter is applied to the
5159 corresponding plane.
5161 Default value for @option{luma_power} is 2. If not specified,
5162 @option{chroma_power} and @option{alpha_power} default to the
5163 corresponding value set for @option{luma_power}.
5165 A value of 0 will disable the effect.
5168 @subsection Examples
5172 Apply a boxblur filter with the luma, chroma, and alpha radii
5175 boxblur=luma_radius=2:luma_power=1
5180 Set the luma radius to 2, and alpha and chroma radius to 0:
5182 boxblur=2:1:cr=0:ar=0
5186 Set the luma and chroma radii to a fraction of the video dimension:
5188 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5194 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5195 Deinterlacing Filter").
5197 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5198 interpolation algorithms.
5199 It accepts the following parameters:
5203 The interlacing mode to adopt. It accepts one of the following values:
5207 Output one frame for each frame.
5209 Output one frame for each field.
5212 The default value is @code{send_field}.
5215 The picture field parity assumed for the input interlaced video. It accepts one
5216 of the following values:
5220 Assume the top field is first.
5222 Assume the bottom field is first.
5224 Enable automatic detection of field parity.
5227 The default value is @code{auto}.
5228 If the interlacing is unknown or the decoder does not export this information,
5229 top field first will be assumed.
5232 Specify which frames to deinterlace. Accept one of the following
5237 Deinterlace all frames.
5239 Only deinterlace frames marked as interlaced.
5242 The default value is @code{all}.
5246 YUV colorspace color/chroma keying.
5248 The filter accepts the following options:
5252 The color which will be replaced with transparency.
5255 Similarity percentage with the key color.
5257 0.01 matches only the exact key color, while 1.0 matches everything.
5262 0.0 makes pixels either fully transparent, or not transparent at all.
5264 Higher values result in semi-transparent pixels, with a higher transparency
5265 the more similar the pixels color is to the key color.
5268 Signals that the color passed is already in YUV instead of RGB.
5270 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5271 This can be used to pass exact YUV values as hexadecimal numbers.
5274 @subsection Examples
5278 Make every green pixel in the input image transparent:
5280 ffmpeg -i input.png -vf chromakey=green out.png
5284 Overlay a greenscreen-video on top of a static black background.
5286 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
5292 Display CIE color diagram with pixels overlaid onto it.
5294 The filter accepts the following options:
5309 @item uhdtv, rec2020
5322 Set what gamuts to draw.
5324 See @code{system} option for available values.
5327 Set ciescope size, by default set to 512.
5330 Set intensity used to map input pixel values to CIE diagram.
5333 Set contrast used to draw tongue colors that are out of active color system gamut.
5336 Correct gamma displayed on scope, by default enabled.
5339 Show white point on CIE diagram, by default disabled.
5342 Set input gamma. Used only with XYZ input color space.
5347 Visualize information exported by some codecs.
5349 Some codecs can export information through frames using side-data or other
5350 means. For example, some MPEG based codecs export motion vectors through the
5351 @var{export_mvs} flag in the codec @option{flags2} option.
5353 The filter accepts the following option:
5357 Set motion vectors to visualize.
5359 Available flags for @var{mv} are:
5363 forward predicted MVs of P-frames
5365 forward predicted MVs of B-frames
5367 backward predicted MVs of B-frames
5371 Display quantization parameters using the chroma planes.
5374 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5376 Available flags for @var{mv_type} are:
5380 forward predicted MVs
5382 backward predicted MVs
5385 @item frame_type, ft
5386 Set frame type to visualize motion vectors of.
5388 Available flags for @var{frame_type} are:
5392 intra-coded frames (I-frames)
5394 predicted frames (P-frames)
5396 bi-directionally predicted frames (B-frames)
5400 @subsection Examples
5404 Visualize forward predicted MVs of all frames using @command{ffplay}:
5406 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5410 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5412 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5416 @section colorbalance
5417 Modify intensity of primary colors (red, green and blue) of input frames.
5419 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5420 regions for the red-cyan, green-magenta or blue-yellow balance.
5422 A positive adjustment value shifts the balance towards the primary color, a negative
5423 value towards the complementary color.
5425 The filter accepts the following options:
5431 Adjust red, green and blue shadows (darkest pixels).
5436 Adjust red, green and blue midtones (medium pixels).
5441 Adjust red, green and blue highlights (brightest pixels).
5443 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5446 @subsection Examples
5450 Add red color cast to shadows:
5457 RGB colorspace color keying.
5459 The filter accepts the following options:
5463 The color which will be replaced with transparency.
5466 Similarity percentage with the key color.
5468 0.01 matches only the exact key color, while 1.0 matches everything.
5473 0.0 makes pixels either fully transparent, or not transparent at all.
5475 Higher values result in semi-transparent pixels, with a higher transparency
5476 the more similar the pixels color is to the key color.
5479 @subsection Examples
5483 Make every green pixel in the input image transparent:
5485 ffmpeg -i input.png -vf colorkey=green out.png
5489 Overlay a greenscreen-video on top of a static background image.
5491 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
5495 @section colorlevels
5497 Adjust video input frames using levels.
5499 The filter accepts the following options:
5506 Adjust red, green, blue and alpha input black point.
5507 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5513 Adjust red, green, blue and alpha input white point.
5514 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5516 Input levels are used to lighten highlights (bright tones), darken shadows
5517 (dark tones), change the balance of bright and dark tones.
5523 Adjust red, green, blue and alpha output black point.
5524 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5530 Adjust red, green, blue and alpha output white point.
5531 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5533 Output levels allows manual selection of a constrained output level range.
5536 @subsection Examples
5540 Make video output darker:
5542 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5548 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5552 Make video output lighter:
5554 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5558 Increase brightness:
5560 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5564 @section colorchannelmixer
5566 Adjust video input frames by re-mixing color channels.
5568 This filter modifies a color channel by adding the values associated to
5569 the other channels of the same pixels. For example if the value to
5570 modify is red, the output value will be:
5572 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5575 The filter accepts the following options:
5582 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5583 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5589 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5590 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5596 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5597 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5603 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5604 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5606 Allowed ranges for options are @code{[-2.0, 2.0]}.
5609 @subsection Examples
5613 Convert source to grayscale:
5615 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5618 Simulate sepia tones:
5620 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5624 @section colormatrix
5626 Convert color matrix.
5628 The filter accepts the following options:
5633 Specify the source and destination color matrix. Both values must be
5636 The accepted values are:
5664 For example to convert from BT.601 to SMPTE-240M, use the command:
5666 colormatrix=bt601:smpte240m
5671 Convert colorspace, transfer characteristics or color primaries.
5672 Input video needs to have an even size.
5674 The filter accepts the following options:
5679 Specify all color properties at once.
5681 The accepted values are:
5711 Specify output colorspace.
5713 The accepted values are:
5722 BT.470BG or BT.601-6 625
5725 SMPTE-170M or BT.601-6 525
5734 BT.2020 with non-constant luminance
5740 Specify output transfer characteristics.
5742 The accepted values are:
5754 Constant gamma of 2.2
5757 Constant gamma of 2.8
5760 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5778 BT.2020 for 10-bits content
5781 BT.2020 for 12-bits content
5787 Specify output color primaries.
5789 The accepted values are:
5798 BT.470BG or BT.601-6 625
5801 SMPTE-170M or BT.601-6 525
5825 Specify output color range.
5827 The accepted values are:
5830 TV (restricted) range
5833 MPEG (restricted) range
5844 Specify output color format.
5846 The accepted values are:
5849 YUV 4:2:0 planar 8-bits
5852 YUV 4:2:0 planar 10-bits
5855 YUV 4:2:0 planar 12-bits
5858 YUV 4:2:2 planar 8-bits
5861 YUV 4:2:2 planar 10-bits
5864 YUV 4:2:2 planar 12-bits
5867 YUV 4:4:4 planar 8-bits
5870 YUV 4:4:4 planar 10-bits
5873 YUV 4:4:4 planar 12-bits
5878 Do a fast conversion, which skips gamma/primary correction. This will take
5879 significantly less CPU, but will be mathematically incorrect. To get output
5880 compatible with that produced by the colormatrix filter, use fast=1.
5883 Specify dithering mode.
5885 The accepted values are:
5891 Floyd-Steinberg dithering
5895 Whitepoint adaptation mode.
5897 The accepted values are:
5900 Bradford whitepoint adaptation
5903 von Kries whitepoint adaptation
5906 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5910 Override all input properties at once. Same accepted values as @ref{all}.
5913 Override input colorspace. Same accepted values as @ref{space}.
5916 Override input color primaries. Same accepted values as @ref{primaries}.
5919 Override input transfer characteristics. Same accepted values as @ref{trc}.
5922 Override input color range. Same accepted values as @ref{range}.
5926 The filter converts the transfer characteristics, color space and color
5927 primaries to the specified user values. The output value, if not specified,
5928 is set to a default value based on the "all" property. If that property is
5929 also not specified, the filter will log an error. The output color range and
5930 format default to the same value as the input color range and format. The
5931 input transfer characteristics, color space, color primaries and color range
5932 should be set on the input data. If any of these are missing, the filter will
5933 log an error and no conversion will take place.
5935 For example to convert the input to SMPTE-240M, use the command:
5937 colorspace=smpte240m
5940 @section convolution
5942 Apply convolution 3x3 or 5x5 filter.
5944 The filter accepts the following options:
5951 Set matrix for each plane.
5952 Matrix is sequence of 9 or 25 signed integers.
5958 Set multiplier for calculated value for each plane.
5964 Set bias for each plane. This value is added to the result of the multiplication.
5965 Useful for making the overall image brighter or darker. Default is 0.0.
5968 @subsection Examples
5974 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"
5980 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"
5986 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"
5992 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"
5996 Apply laplacian edge detector which includes diagonals:
5998 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"
6004 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"
6010 Apply 2D convolution of video stream in frequency domain using second stream
6013 The filter accepts the following options:
6017 Set which planes to process.
6020 Set which impulse video frames will be processed, can be @var{first}
6021 or @var{all}. Default is @var{all}.
6024 The @code{convolve} filter also supports the @ref{framesync} options.
6028 Copy the input video source unchanged to the output. This is mainly useful for
6033 Video filtering on GPU using Apple's CoreImage API on OSX.
6035 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6036 processed by video hardware. However, software-based OpenGL implementations
6037 exist which means there is no guarantee for hardware processing. It depends on
6040 There are many filters and image generators provided by Apple that come with a
6041 large variety of options. The filter has to be referenced by its name along
6044 The coreimage filter accepts the following options:
6047 List all available filters and generators along with all their respective
6048 options as well as possible minimum and maximum values along with the default
6055 Specify all filters by their respective name and options.
6056 Use @var{list_filters} to determine all valid filter names and options.
6057 Numerical options are specified by a float value and are automatically clamped
6058 to their respective value range. Vector and color options have to be specified
6059 by a list of space separated float values. Character escaping has to be done.
6060 A special option name @code{default} is available to use default options for a
6063 It is required to specify either @code{default} or at least one of the filter options.
6064 All omitted options are used with their default values.
6065 The syntax of the filter string is as follows:
6067 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6071 Specify a rectangle where the output of the filter chain is copied into the
6072 input image. It is given by a list of space separated float values:
6074 output_rect=x\ y\ width\ height
6076 If not given, the output rectangle equals the dimensions of the input image.
6077 The output rectangle is automatically cropped at the borders of the input
6078 image. Negative values are valid for each component.
6080 output_rect=25\ 25\ 100\ 100
6084 Several filters can be chained for successive processing without GPU-HOST
6085 transfers allowing for fast processing of complex filter chains.
6086 Currently, only filters with zero (generators) or exactly one (filters) input
6087 image and one output image are supported. Also, transition filters are not yet
6090 Some filters generate output images with additional padding depending on the
6091 respective filter kernel. The padding is automatically removed to ensure the
6092 filter output has the same size as the input image.
6094 For image generators, the size of the output image is determined by the
6095 previous output image of the filter chain or the input image of the whole
6096 filterchain, respectively. The generators do not use the pixel information of
6097 this image to generate their output. However, the generated output is
6098 blended onto this image, resulting in partial or complete coverage of the
6101 The @ref{coreimagesrc} video source can be used for generating input images
6102 which are directly fed into the filter chain. By using it, providing input
6103 images by another video source or an input video is not required.
6105 @subsection Examples
6110 List all filters available:
6112 coreimage=list_filters=true
6116 Use the CIBoxBlur filter with default options to blur an image:
6118 coreimage=filter=CIBoxBlur@@default
6122 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6123 its center at 100x100 and a radius of 50 pixels:
6125 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6129 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6130 given as complete and escaped command-line for Apple's standard bash shell:
6132 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6138 Crop the input video to given dimensions.
6140 It accepts the following parameters:
6144 The width of the output video. It defaults to @code{iw}.
6145 This expression is evaluated only once during the filter
6146 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6149 The height of the output video. It defaults to @code{ih}.
6150 This expression is evaluated only once during the filter
6151 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6154 The horizontal position, in the input video, of the left edge of the output
6155 video. It defaults to @code{(in_w-out_w)/2}.
6156 This expression is evaluated per-frame.
6159 The vertical position, in the input video, of the top edge of the output video.
6160 It defaults to @code{(in_h-out_h)/2}.
6161 This expression is evaluated per-frame.
6164 If set to 1 will force the output display aspect ratio
6165 to be the same of the input, by changing the output sample aspect
6166 ratio. It defaults to 0.
6169 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6170 width/height/x/y as specified and will not be rounded to nearest smaller value.
6174 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6175 expressions containing the following constants:
6180 The computed values for @var{x} and @var{y}. They are evaluated for
6185 The input width and height.
6189 These are the same as @var{in_w} and @var{in_h}.
6193 The output (cropped) width and height.
6197 These are the same as @var{out_w} and @var{out_h}.
6200 same as @var{iw} / @var{ih}
6203 input sample aspect ratio
6206 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6210 horizontal and vertical chroma subsample values. For example for the
6211 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6214 The number of the input frame, starting from 0.
6217 the position in the file of the input frame, NAN if unknown
6220 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6224 The expression for @var{out_w} may depend on the value of @var{out_h},
6225 and the expression for @var{out_h} may depend on @var{out_w}, but they
6226 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6227 evaluated after @var{out_w} and @var{out_h}.
6229 The @var{x} and @var{y} parameters specify the expressions for the
6230 position of the top-left corner of the output (non-cropped) area. They
6231 are evaluated for each frame. If the evaluated value is not valid, it
6232 is approximated to the nearest valid value.
6234 The expression for @var{x} may depend on @var{y}, and the expression
6235 for @var{y} may depend on @var{x}.
6237 @subsection Examples
6241 Crop area with size 100x100 at position (12,34).
6246 Using named options, the example above becomes:
6248 crop=w=100:h=100:x=12:y=34
6252 Crop the central input area with size 100x100:
6258 Crop the central input area with size 2/3 of the input video:
6260 crop=2/3*in_w:2/3*in_h
6264 Crop the input video central square:
6271 Delimit the rectangle with the top-left corner placed at position
6272 100:100 and the right-bottom corner corresponding to the right-bottom
6273 corner of the input image.
6275 crop=in_w-100:in_h-100:100:100
6279 Crop 10 pixels from the left and right borders, and 20 pixels from
6280 the top and bottom borders
6282 crop=in_w-2*10:in_h-2*20
6286 Keep only the bottom right quarter of the input image:
6288 crop=in_w/2:in_h/2:in_w/2:in_h/2
6292 Crop height for getting Greek harmony:
6294 crop=in_w:1/PHI*in_w
6298 Apply trembling effect:
6300 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)
6304 Apply erratic camera effect depending on timestamp:
6306 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)"
6310 Set x depending on the value of y:
6312 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6316 @subsection Commands
6318 This filter supports the following commands:
6324 Set width/height of the output video and the horizontal/vertical position
6326 The command accepts the same syntax of the corresponding option.
6328 If the specified expression is not valid, it is kept at its current
6334 Auto-detect the crop size.
6336 It calculates the necessary cropping parameters and prints the
6337 recommended parameters via the logging system. The detected dimensions
6338 correspond to the non-black area of the input video.
6340 It accepts the following parameters:
6345 Set higher black value threshold, which can be optionally specified
6346 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6347 value greater to the set value is considered non-black. It defaults to 24.
6348 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6349 on the bitdepth of the pixel format.
6352 The value which the width/height should be divisible by. It defaults to
6353 16. The offset is automatically adjusted to center the video. Use 2 to
6354 get only even dimensions (needed for 4:2:2 video). 16 is best when
6355 encoding to most video codecs.
6357 @item reset_count, reset
6358 Set the counter that determines after how many frames cropdetect will
6359 reset the previously detected largest video area and start over to
6360 detect the current optimal crop area. Default value is 0.
6362 This can be useful when channel logos distort the video area. 0
6363 indicates 'never reset', and returns the largest area encountered during
6370 Apply color adjustments using curves.
6372 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6373 component (red, green and blue) has its values defined by @var{N} key points
6374 tied from each other using a smooth curve. The x-axis represents the pixel
6375 values from the input frame, and the y-axis the new pixel values to be set for
6378 By default, a component curve is defined by the two points @var{(0;0)} and
6379 @var{(1;1)}. This creates a straight line where each original pixel value is
6380 "adjusted" to its own value, which means no change to the image.
6382 The filter allows you to redefine these two points and add some more. A new
6383 curve (using a natural cubic spline interpolation) will be define to pass
6384 smoothly through all these new coordinates. The new defined points needs to be
6385 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6386 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6387 the vector spaces, the values will be clipped accordingly.
6389 The filter accepts the following options:
6393 Select one of the available color presets. This option can be used in addition
6394 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6395 options takes priority on the preset values.
6396 Available presets are:
6399 @item color_negative
6402 @item increase_contrast
6404 @item linear_contrast
6405 @item medium_contrast
6407 @item strong_contrast
6410 Default is @code{none}.
6412 Set the master key points. These points will define a second pass mapping. It
6413 is sometimes called a "luminance" or "value" mapping. It can be used with
6414 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6415 post-processing LUT.
6417 Set the key points for the red component.
6419 Set the key points for the green component.
6421 Set the key points for the blue component.
6423 Set the key points for all components (not including master).
6424 Can be used in addition to the other key points component
6425 options. In this case, the unset component(s) will fallback on this
6426 @option{all} setting.
6428 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6430 Save Gnuplot script of the curves in specified file.
6433 To avoid some filtergraph syntax conflicts, each key points list need to be
6434 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6436 @subsection Examples
6440 Increase slightly the middle level of blue:
6442 curves=blue='0/0 0.5/0.58 1/1'
6448 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'
6450 Here we obtain the following coordinates for each components:
6453 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6455 @code{(0;0) (0.50;0.48) (1;1)}
6457 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6461 The previous example can also be achieved with the associated built-in preset:
6463 curves=preset=vintage
6473 Use a Photoshop preset and redefine the points of the green component:
6475 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6479 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6480 and @command{gnuplot}:
6482 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6483 gnuplot -p /tmp/curves.plt
6489 Video data analysis filter.
6491 This filter shows hexadecimal pixel values of part of video.
6493 The filter accepts the following options:
6497 Set output video size.
6500 Set x offset from where to pick pixels.
6503 Set y offset from where to pick pixels.
6506 Set scope mode, can be one of the following:
6509 Draw hexadecimal pixel values with white color on black background.
6512 Draw hexadecimal pixel values with input video pixel color on black
6516 Draw hexadecimal pixel values on color background picked from input video,
6517 the text color is picked in such way so its always visible.
6521 Draw rows and columns numbers on left and top of video.
6524 Set background opacity.
6529 Denoise frames using 2D DCT (frequency domain filtering).
6531 This filter is not designed for real time.
6533 The filter accepts the following options:
6537 Set the noise sigma constant.
6539 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6540 coefficient (absolute value) below this threshold with be dropped.
6542 If you need a more advanced filtering, see @option{expr}.
6544 Default is @code{0}.
6547 Set number overlapping pixels for each block. Since the filter can be slow, you
6548 may want to reduce this value, at the cost of a less effective filter and the
6549 risk of various artefacts.
6551 If the overlapping value doesn't permit processing the whole input width or
6552 height, a warning will be displayed and according borders won't be denoised.
6554 Default value is @var{blocksize}-1, which is the best possible setting.
6557 Set the coefficient factor expression.
6559 For each coefficient of a DCT block, this expression will be evaluated as a
6560 multiplier value for the coefficient.
6562 If this is option is set, the @option{sigma} option will be ignored.
6564 The absolute value of the coefficient can be accessed through the @var{c}
6568 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6569 @var{blocksize}, which is the width and height of the processed blocks.
6571 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6572 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6573 on the speed processing. Also, a larger block size does not necessarily means a
6577 @subsection Examples
6579 Apply a denoise with a @option{sigma} of @code{4.5}:
6584 The same operation can be achieved using the expression system:
6586 dctdnoiz=e='gte(c, 4.5*3)'
6589 Violent denoise using a block size of @code{16x16}:
6596 Remove banding artifacts from input video.
6597 It works by replacing banded pixels with average value of referenced pixels.
6599 The filter accepts the following options:
6606 Set banding detection threshold for each plane. Default is 0.02.
6607 Valid range is 0.00003 to 0.5.
6608 If difference between current pixel and reference pixel is less than threshold,
6609 it will be considered as banded.
6612 Banding detection range in pixels. Default is 16. If positive, random number
6613 in range 0 to set value will be used. If negative, exact absolute value
6615 The range defines square of four pixels around current pixel.
6618 Set direction in radians from which four pixel will be compared. If positive,
6619 random direction from 0 to set direction will be picked. If negative, exact of
6620 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6621 will pick only pixels on same row and -PI/2 will pick only pixels on same
6625 If enabled, current pixel is compared with average value of all four
6626 surrounding pixels. The default is enabled. If disabled current pixel is
6627 compared with all four surrounding pixels. The pixel is considered banded
6628 if only all four differences with surrounding pixels are less than threshold.
6631 If enabled, current pixel is changed if and only if all pixel components are banded,
6632 e.g. banding detection threshold is triggered for all color components.
6633 The default is disabled.
6639 Drop duplicated frames at regular intervals.
6641 The filter accepts the following options:
6645 Set the number of frames from which one will be dropped. Setting this to
6646 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6647 Default is @code{5}.
6650 Set the threshold for duplicate detection. If the difference metric for a frame
6651 is less than or equal to this value, then it is declared as duplicate. Default
6655 Set scene change threshold. Default is @code{15}.
6659 Set the size of the x and y-axis blocks used during metric calculations.
6660 Larger blocks give better noise suppression, but also give worse detection of
6661 small movements. Must be a power of two. Default is @code{32}.
6664 Mark main input as a pre-processed input and activate clean source input
6665 stream. This allows the input to be pre-processed with various filters to help
6666 the metrics calculation while keeping the frame selection lossless. When set to
6667 @code{1}, the first stream is for the pre-processed input, and the second
6668 stream is the clean source from where the kept frames are chosen. Default is
6672 Set whether or not chroma is considered in the metric calculations. Default is
6678 Apply deflate effect to the video.
6680 This filter replaces the pixel by the local(3x3) average by taking into account
6681 only values lower than the pixel.
6683 It accepts the following options:
6690 Limit the maximum change for each plane, default is 65535.
6691 If 0, plane will remain unchanged.
6696 Remove temporal frame luminance variations.
6698 It accepts the following options:
6702 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
6705 Set averaging mode to smooth temporal luminance variations.
6707 Available values are:
6732 Do not actually modify frame. Useful when one only wants metadata.
6737 Remove judder produced by partially interlaced telecined content.
6739 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6740 source was partially telecined content then the output of @code{pullup,dejudder}
6741 will have a variable frame rate. May change the recorded frame rate of the
6742 container. Aside from that change, this filter will not affect constant frame
6745 The option available in this filter is:
6749 Specify the length of the window over which the judder repeats.
6751 Accepts any integer greater than 1. Useful values are:
6755 If the original was telecined from 24 to 30 fps (Film to NTSC).
6758 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6761 If a mixture of the two.
6764 The default is @samp{4}.
6769 Suppress a TV station logo by a simple interpolation of the surrounding
6770 pixels. Just set a rectangle covering the logo and watch it disappear
6771 (and sometimes something even uglier appear - your mileage may vary).
6773 It accepts the following parameters:
6778 Specify the top left corner coordinates of the logo. They must be
6783 Specify the width and height of the logo to clear. They must be
6787 Specify the thickness of the fuzzy edge of the rectangle (added to
6788 @var{w} and @var{h}). The default value is 1. This option is
6789 deprecated, setting higher values should no longer be necessary and
6793 When set to 1, a green rectangle is drawn on the screen to simplify
6794 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6795 The default value is 0.
6797 The rectangle is drawn on the outermost pixels which will be (partly)
6798 replaced with interpolated values. The values of the next pixels
6799 immediately outside this rectangle in each direction will be used to
6800 compute the interpolated pixel values inside the rectangle.
6804 @subsection Examples
6808 Set a rectangle covering the area with top left corner coordinates 0,0
6809 and size 100x77, and a band of size 10:
6811 delogo=x=0:y=0:w=100:h=77:band=10
6818 Attempt to fix small changes in horizontal and/or vertical shift. This
6819 filter helps remove camera shake from hand-holding a camera, bumping a
6820 tripod, moving on a vehicle, etc.
6822 The filter accepts the following options:
6830 Specify a rectangular area where to limit the search for motion
6832 If desired the search for motion vectors can be limited to a
6833 rectangular area of the frame defined by its top left corner, width
6834 and height. These parameters have the same meaning as the drawbox
6835 filter which can be used to visualise the position of the bounding
6838 This is useful when simultaneous movement of subjects within the frame
6839 might be confused for camera motion by the motion vector search.
6841 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6842 then the full frame is used. This allows later options to be set
6843 without specifying the bounding box for the motion vector search.
6845 Default - search the whole frame.
6849 Specify the maximum extent of movement in x and y directions in the
6850 range 0-64 pixels. Default 16.
6853 Specify how to generate pixels to fill blanks at the edge of the
6854 frame. Available values are:
6857 Fill zeroes at blank locations
6859 Original image at blank locations
6861 Extruded edge value at blank locations
6863 Mirrored edge at blank locations
6865 Default value is @samp{mirror}.
6868 Specify the blocksize to use for motion search. Range 4-128 pixels,
6872 Specify the contrast threshold for blocks. Only blocks with more than
6873 the specified contrast (difference between darkest and lightest
6874 pixels) will be considered. Range 1-255, default 125.
6877 Specify the search strategy. Available values are:
6880 Set exhaustive search
6882 Set less exhaustive search.
6884 Default value is @samp{exhaustive}.
6887 If set then a detailed log of the motion search is written to the
6891 If set to 1, specify using OpenCL capabilities, only available if
6892 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6898 Remove unwanted contamination of foreground colors, caused by reflected color of
6899 greenscreen or bluescreen.
6901 This filter accepts the following options:
6905 Set what type of despill to use.
6908 Set how spillmap will be generated.
6911 Set how much to get rid of still remaining spill.
6914 Controls amount of red in spill area.
6917 Controls amount of green in spill area.
6918 Should be -1 for greenscreen.
6921 Controls amount of blue in spill area.
6922 Should be -1 for bluescreen.
6925 Controls brightness of spill area, preserving colors.
6928 Modify alpha from generated spillmap.
6933 Apply an exact inverse of the telecine operation. It requires a predefined
6934 pattern specified using the pattern option which must be the same as that passed
6935 to the telecine filter.
6937 This filter accepts the following options:
6946 The default value is @code{top}.
6950 A string of numbers representing the pulldown pattern you wish to apply.
6951 The default value is @code{23}.
6954 A number representing position of the first frame with respect to the telecine
6955 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6960 Apply dilation effect to the video.
6962 This filter replaces the pixel by the local(3x3) maximum.
6964 It accepts the following options:
6971 Limit the maximum change for each plane, default is 65535.
6972 If 0, plane will remain unchanged.
6975 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6978 Flags to local 3x3 coordinates maps like this:
6987 Displace pixels as indicated by second and third input stream.
6989 It takes three input streams and outputs one stream, the first input is the
6990 source, and second and third input are displacement maps.
6992 The second input specifies how much to displace pixels along the
6993 x-axis, while the third input specifies how much to displace pixels
6995 If one of displacement map streams terminates, last frame from that
6996 displacement map will be used.
6998 Note that once generated, displacements maps can be reused over and over again.
7000 A description of the accepted options follows.
7004 Set displace behavior for pixels that are out of range.
7006 Available values are:
7009 Missing pixels are replaced by black pixels.
7012 Adjacent pixels will spread out to replace missing pixels.
7015 Out of range pixels are wrapped so they point to pixels of other side.
7018 Out of range pixels will be replaced with mirrored pixels.
7020 Default is @samp{smear}.
7024 @subsection Examples
7028 Add ripple effect to rgb input of video size hd720:
7030 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
7034 Add wave effect to rgb input of video size hd720:
7036 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
7042 Draw a colored box on the input image.
7044 It accepts the following parameters:
7049 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7053 The expressions which specify the width and height of the box; if 0 they are interpreted as
7054 the input width and height. It defaults to 0.
7057 Specify the color of the box to write. For the general syntax of this option,
7058 check the "Color" section in the ffmpeg-utils manual. If the special
7059 value @code{invert} is used, the box edge color is the same as the
7060 video with inverted luma.
7063 The expression which sets the thickness of the box edge. Default value is @code{3}.
7065 See below for the list of accepted constants.
7068 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7069 following constants:
7073 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7077 horizontal and vertical chroma subsample values. For example for the
7078 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7082 The input width and height.
7085 The input sample aspect ratio.
7089 The x and y offset coordinates where the box is drawn.
7093 The width and height of the drawn box.
7096 The thickness of the drawn box.
7098 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7099 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7103 @subsection Examples
7107 Draw a black box around the edge of the input image:
7113 Draw a box with color red and an opacity of 50%:
7115 drawbox=10:20:200:60:red@@0.5
7118 The previous example can be specified as:
7120 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7124 Fill the box with pink color:
7126 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
7130 Draw a 2-pixel red 2.40:1 mask:
7132 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
7138 Draw a grid on the input image.
7140 It accepts the following parameters:
7145 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7149 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7150 input width and height, respectively, minus @code{thickness}, so image gets
7151 framed. Default to 0.
7154 Specify the color of the grid. For the general syntax of this option,
7155 check the "Color" section in the ffmpeg-utils manual. If the special
7156 value @code{invert} is used, the grid color is the same as the
7157 video with inverted luma.
7160 The expression which sets the thickness of the grid line. Default value is @code{1}.
7162 See below for the list of accepted constants.
7165 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7166 following constants:
7170 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7174 horizontal and vertical chroma subsample values. For example for the
7175 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7179 The input grid cell width and height.
7182 The input sample aspect ratio.
7186 The x and y coordinates of some point of grid intersection (meant to configure offset).
7190 The width and height of the drawn cell.
7193 The thickness of the drawn cell.
7195 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7196 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7200 @subsection Examples
7204 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7206 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7210 Draw a white 3x3 grid with an opacity of 50%:
7212 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7219 Draw a text string or text from a specified file on top of a video, using the
7220 libfreetype library.
7222 To enable compilation of this filter, you need to configure FFmpeg with
7223 @code{--enable-libfreetype}.
7224 To enable default font fallback and the @var{font} option you need to
7225 configure FFmpeg with @code{--enable-libfontconfig}.
7226 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7227 @code{--enable-libfribidi}.
7231 It accepts the following parameters:
7236 Used to draw a box around text using the background color.
7237 The value must be either 1 (enable) or 0 (disable).
7238 The default value of @var{box} is 0.
7241 Set the width of the border to be drawn around the box using @var{boxcolor}.
7242 The default value of @var{boxborderw} is 0.
7245 The color to be used for drawing box around text. For the syntax of this
7246 option, check the "Color" section in the ffmpeg-utils manual.
7248 The default value of @var{boxcolor} is "white".
7251 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7252 The default value of @var{line_spacing} is 0.
7255 Set the width of the border to be drawn around the text using @var{bordercolor}.
7256 The default value of @var{borderw} is 0.
7259 Set the color to be used for drawing border around text. For the syntax of this
7260 option, check the "Color" section in the ffmpeg-utils manual.
7262 The default value of @var{bordercolor} is "black".
7265 Select how the @var{text} is expanded. Can be either @code{none},
7266 @code{strftime} (deprecated) or
7267 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7271 Set a start time for the count. Value is in microseconds. Only applied
7272 in the deprecated strftime expansion mode. To emulate in normal expansion
7273 mode use the @code{pts} function, supplying the start time (in seconds)
7274 as the second argument.
7277 If true, check and fix text coords to avoid clipping.
7280 The color to be used for drawing fonts. For the syntax of this option, check
7281 the "Color" section in the ffmpeg-utils manual.
7283 The default value of @var{fontcolor} is "black".
7285 @item fontcolor_expr
7286 String which is expanded the same way as @var{text} to obtain dynamic
7287 @var{fontcolor} value. By default this option has empty value and is not
7288 processed. When this option is set, it overrides @var{fontcolor} option.
7291 The font family to be used for drawing text. By default Sans.
7294 The font file to be used for drawing text. The path must be included.
7295 This parameter is mandatory if the fontconfig support is disabled.
7298 Draw the text applying alpha blending. The value can
7299 be a number between 0.0 and 1.0.
7300 The expression accepts the same variables @var{x, y} as well.
7301 The default value is 1.
7302 Please see @var{fontcolor_expr}.
7305 The font size to be used for drawing text.
7306 The default value of @var{fontsize} is 16.
7309 If set to 1, attempt to shape the text (for example, reverse the order of
7310 right-to-left text and join Arabic characters) before drawing it.
7311 Otherwise, just draw the text exactly as given.
7312 By default 1 (if supported).
7315 The flags to be used for loading the fonts.
7317 The flags map the corresponding flags supported by libfreetype, and are
7318 a combination of the following values:
7325 @item vertical_layout
7326 @item force_autohint
7329 @item ignore_global_advance_width
7331 @item ignore_transform
7337 Default value is "default".
7339 For more information consult the documentation for the FT_LOAD_*
7343 The color to be used for drawing a shadow behind the drawn text. For the
7344 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
7346 The default value of @var{shadowcolor} is "black".
7350 The x and y offsets for the text shadow position with respect to the
7351 position of the text. They can be either positive or negative
7352 values. The default value for both is "0".
7355 The starting frame number for the n/frame_num variable. The default value
7359 The size in number of spaces to use for rendering the tab.
7363 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7364 format. It can be used with or without text parameter. @var{timecode_rate}
7365 option must be specified.
7367 @item timecode_rate, rate, r
7368 Set the timecode frame rate (timecode only).
7371 If set to 1, the output of the timecode option will wrap around at 24 hours.
7372 Default is 0 (disabled).
7375 The text string to be drawn. The text must be a sequence of UTF-8
7377 This parameter is mandatory if no file is specified with the parameter
7381 A text file containing text to be drawn. The text must be a sequence
7382 of UTF-8 encoded characters.
7384 This parameter is mandatory if no text string is specified with the
7385 parameter @var{text}.
7387 If both @var{text} and @var{textfile} are specified, an error is thrown.
7390 If set to 1, the @var{textfile} will be reloaded before each frame.
7391 Be sure to update it atomically, or it may be read partially, or even fail.
7395 The expressions which specify the offsets where text will be drawn
7396 within the video frame. They are relative to the top/left border of the
7399 The default value of @var{x} and @var{y} is "0".
7401 See below for the list of accepted constants and functions.
7404 The parameters for @var{x} and @var{y} are expressions containing the
7405 following constants and functions:
7409 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7413 horizontal and vertical chroma subsample values. For example for the
7414 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7417 the height of each text line
7425 @item max_glyph_a, ascent
7426 the maximum distance from the baseline to the highest/upper grid
7427 coordinate used to place a glyph outline point, for all the rendered
7429 It is a positive value, due to the grid's orientation with the Y axis
7432 @item max_glyph_d, descent
7433 the maximum distance from the baseline to the lowest grid coordinate
7434 used to place a glyph outline point, for all the rendered glyphs.
7435 This is a negative value, due to the grid's orientation, with the Y axis
7439 maximum glyph height, that is the maximum height for all the glyphs
7440 contained in the rendered text, it is equivalent to @var{ascent} -
7444 maximum glyph width, that is the maximum width for all the glyphs
7445 contained in the rendered text
7448 the number of input frame, starting from 0
7450 @item rand(min, max)
7451 return a random number included between @var{min} and @var{max}
7454 The input sample aspect ratio.
7457 timestamp expressed in seconds, NAN if the input timestamp is unknown
7460 the height of the rendered text
7463 the width of the rendered text
7467 the x and y offset coordinates where the text is drawn.
7469 These parameters allow the @var{x} and @var{y} expressions to refer
7470 each other, so you can for example specify @code{y=x/dar}.
7473 @anchor{drawtext_expansion}
7474 @subsection Text expansion
7476 If @option{expansion} is set to @code{strftime},
7477 the filter recognizes strftime() sequences in the provided text and
7478 expands them accordingly. Check the documentation of strftime(). This
7479 feature is deprecated.
7481 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7483 If @option{expansion} is set to @code{normal} (which is the default),
7484 the following expansion mechanism is used.
7486 The backslash character @samp{\}, followed by any character, always expands to
7487 the second character.
7489 Sequences of the form @code{%@{...@}} are expanded. The text between the
7490 braces is a function name, possibly followed by arguments separated by ':'.
7491 If the arguments contain special characters or delimiters (':' or '@}'),
7492 they should be escaped.
7494 Note that they probably must also be escaped as the value for the
7495 @option{text} option in the filter argument string and as the filter
7496 argument in the filtergraph description, and possibly also for the shell,
7497 that makes up to four levels of escaping; using a text file avoids these
7500 The following functions are available:
7505 The expression evaluation result.
7507 It must take one argument specifying the expression to be evaluated,
7508 which accepts the same constants and functions as the @var{x} and
7509 @var{y} values. Note that not all constants should be used, for
7510 example the text size is not known when evaluating the expression, so
7511 the constants @var{text_w} and @var{text_h} will have an undefined
7514 @item expr_int_format, eif
7515 Evaluate the expression's value and output as formatted integer.
7517 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7518 The second argument specifies the output format. Allowed values are @samp{x},
7519 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7520 @code{printf} function.
7521 The third parameter is optional and sets the number of positions taken by the output.
7522 It can be used to add padding with zeros from the left.
7525 The time at which the filter is running, expressed in UTC.
7526 It can accept an argument: a strftime() format string.
7529 The time at which the filter is running, expressed in the local time zone.
7530 It can accept an argument: a strftime() format string.
7533 Frame metadata. Takes one or two arguments.
7535 The first argument is mandatory and specifies the metadata key.
7537 The second argument is optional and specifies a default value, used when the
7538 metadata key is not found or empty.
7541 The frame number, starting from 0.
7544 A 1 character description of the current picture type.
7547 The timestamp of the current frame.
7548 It can take up to three arguments.
7550 The first argument is the format of the timestamp; it defaults to @code{flt}
7551 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7552 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7553 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7554 @code{localtime} stands for the timestamp of the frame formatted as
7555 local time zone time.
7557 The second argument is an offset added to the timestamp.
7559 If the format is set to @code{localtime} or @code{gmtime},
7560 a third argument may be supplied: a strftime() format string.
7561 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7564 @subsection Examples
7568 Draw "Test Text" with font FreeSerif, using the default values for the
7569 optional parameters.
7572 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7576 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7577 and y=50 (counting from the top-left corner of the screen), text is
7578 yellow with a red box around it. Both the text and the box have an
7582 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7583 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7586 Note that the double quotes are not necessary if spaces are not used
7587 within the parameter list.
7590 Show the text at the center of the video frame:
7592 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7596 Show the text at a random position, switching to a new position every 30 seconds:
7598 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)"
7602 Show a text line sliding from right to left in the last row of the video
7603 frame. The file @file{LONG_LINE} is assumed to contain a single line
7606 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7610 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7612 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7616 Draw a single green letter "g", at the center of the input video.
7617 The glyph baseline is placed at half screen height.
7619 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7623 Show text for 1 second every 3 seconds:
7625 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7629 Use fontconfig to set the font. Note that the colons need to be escaped.
7631 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7635 Print the date of a real-time encoding (see strftime(3)):
7637 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7641 Show text fading in and out (appearing/disappearing):
7644 DS=1.0 # display start
7645 DE=10.0 # display end
7646 FID=1.5 # fade in duration
7647 FOD=5 # fade out duration
7648 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 @}"
7652 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
7653 and the @option{fontsize} value are included in the @option{y} offset.
7655 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
7656 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
7661 For more information about libfreetype, check:
7662 @url{http://www.freetype.org/}.
7664 For more information about fontconfig, check:
7665 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7667 For more information about libfribidi, check:
7668 @url{http://fribidi.org/}.
7672 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7674 The filter accepts the following options:
7679 Set low and high threshold values used by the Canny thresholding
7682 The high threshold selects the "strong" edge pixels, which are then
7683 connected through 8-connectivity with the "weak" edge pixels selected
7684 by the low threshold.
7686 @var{low} and @var{high} threshold values must be chosen in the range
7687 [0,1], and @var{low} should be lesser or equal to @var{high}.
7689 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7693 Define the drawing mode.
7697 Draw white/gray wires on black background.
7700 Mix the colors to create a paint/cartoon effect.
7703 Default value is @var{wires}.
7706 @subsection Examples
7710 Standard edge detection with custom values for the hysteresis thresholding:
7712 edgedetect=low=0.1:high=0.4
7716 Painting effect without thresholding:
7718 edgedetect=mode=colormix:high=0
7723 Set brightness, contrast, saturation and approximate gamma adjustment.
7725 The filter accepts the following options:
7729 Set the contrast expression. The value must be a float value in range
7730 @code{-2.0} to @code{2.0}. The default value is "1".
7733 Set the brightness expression. The value must be a float value in
7734 range @code{-1.0} to @code{1.0}. The default value is "0".
7737 Set the saturation expression. The value must be a float in
7738 range @code{0.0} to @code{3.0}. The default value is "1".
7741 Set the gamma expression. The value must be a float in range
7742 @code{0.1} to @code{10.0}. The default value is "1".
7745 Set the gamma expression for red. The value must be a float in
7746 range @code{0.1} to @code{10.0}. The default value is "1".
7749 Set the gamma expression for green. The value must be a float in range
7750 @code{0.1} to @code{10.0}. The default value is "1".
7753 Set the gamma expression for blue. The value must be a float in range
7754 @code{0.1} to @code{10.0}. The default value is "1".
7757 Set the gamma weight expression. It can be used to reduce the effect
7758 of a high gamma value on bright image areas, e.g. keep them from
7759 getting overamplified and just plain white. The value must be a float
7760 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7761 gamma correction all the way down while @code{1.0} leaves it at its
7762 full strength. Default is "1".
7765 Set when the expressions for brightness, contrast, saturation and
7766 gamma expressions are evaluated.
7768 It accepts the following values:
7771 only evaluate expressions once during the filter initialization or
7772 when a command is processed
7775 evaluate expressions for each incoming frame
7778 Default value is @samp{init}.
7781 The expressions accept the following parameters:
7784 frame count of the input frame starting from 0
7787 byte position of the corresponding packet in the input file, NAN if
7791 frame rate of the input video, NAN if the input frame rate is unknown
7794 timestamp expressed in seconds, NAN if the input timestamp is unknown
7797 @subsection Commands
7798 The filter supports the following commands:
7802 Set the contrast expression.
7805 Set the brightness expression.
7808 Set the saturation expression.
7811 Set the gamma expression.
7814 Set the gamma_r expression.
7817 Set gamma_g expression.
7820 Set gamma_b expression.
7823 Set gamma_weight expression.
7825 The command accepts the same syntax of the corresponding option.
7827 If the specified expression is not valid, it is kept at its current
7834 Apply erosion effect to the video.
7836 This filter replaces the pixel by the local(3x3) minimum.
7838 It accepts the following options:
7845 Limit the maximum change for each plane, default is 65535.
7846 If 0, plane will remain unchanged.
7849 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7852 Flags to local 3x3 coordinates maps like this:
7859 @section extractplanes
7861 Extract color channel components from input video stream into
7862 separate grayscale video streams.
7864 The filter accepts the following option:
7868 Set plane(s) to extract.
7870 Available values for planes are:
7881 Choosing planes not available in the input will result in an error.
7882 That means you cannot select @code{r}, @code{g}, @code{b} planes
7883 with @code{y}, @code{u}, @code{v} planes at same time.
7886 @subsection Examples
7890 Extract luma, u and v color channel component from input video frame
7891 into 3 grayscale outputs:
7893 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
7899 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7901 For each input image, the filter will compute the optimal mapping from
7902 the input to the output given the codebook length, that is the number
7903 of distinct output colors.
7905 This filter accepts the following options.
7908 @item codebook_length, l
7909 Set codebook length. The value must be a positive integer, and
7910 represents the number of distinct output colors. Default value is 256.
7913 Set the maximum number of iterations to apply for computing the optimal
7914 mapping. The higher the value the better the result and the higher the
7915 computation time. Default value is 1.
7918 Set a random seed, must be an integer included between 0 and
7919 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7920 will try to use a good random seed on a best effort basis.
7923 Set pal8 output pixel format. This option does not work with codebook
7924 length greater than 256.
7929 Apply a fade-in/out effect to the input video.
7931 It accepts the following parameters:
7935 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7937 Default is @code{in}.
7939 @item start_frame, s
7940 Specify the number of the frame to start applying the fade
7941 effect at. Default is 0.
7944 The number of frames that the fade effect lasts. At the end of the
7945 fade-in effect, the output video will have the same intensity as the input video.
7946 At the end of the fade-out transition, the output video will be filled with the
7947 selected @option{color}.
7951 If set to 1, fade only alpha channel, if one exists on the input.
7954 @item start_time, st
7955 Specify the timestamp (in seconds) of the frame to start to apply the fade
7956 effect. If both start_frame and start_time are specified, the fade will start at
7957 whichever comes last. Default is 0.
7960 The number of seconds for which the fade effect has to last. At the end of the
7961 fade-in effect the output video will have the same intensity as the input video,
7962 at the end of the fade-out transition the output video will be filled with the
7963 selected @option{color}.
7964 If both duration and nb_frames are specified, duration is used. Default is 0
7965 (nb_frames is used by default).
7968 Specify the color of the fade. Default is "black".
7971 @subsection Examples
7975 Fade in the first 30 frames of video:
7980 The command above is equivalent to:
7986 Fade out the last 45 frames of a 200-frame video:
7989 fade=type=out:start_frame=155:nb_frames=45
7993 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7995 fade=in:0:25, fade=out:975:25
7999 Make the first 5 frames yellow, then fade in from frame 5-24:
8001 fade=in:5:20:color=yellow
8005 Fade in alpha over first 25 frames of video:
8007 fade=in:0:25:alpha=1
8011 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8013 fade=t=in:st=5.5:d=0.5
8019 Apply arbitrary expressions to samples in frequency domain
8023 Adjust the dc value (gain) of the luma plane of the image. The filter
8024 accepts an integer value in range @code{0} to @code{1000}. The default
8025 value is set to @code{0}.
8028 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8029 filter accepts an integer value in range @code{0} to @code{1000}. The
8030 default value is set to @code{0}.
8033 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8034 filter accepts an integer value in range @code{0} to @code{1000}. The
8035 default value is set to @code{0}.
8038 Set the frequency domain weight expression for the luma plane.
8041 Set the frequency domain weight expression for the 1st chroma plane.
8044 Set the frequency domain weight expression for the 2nd chroma plane.
8047 Set when the expressions are evaluated.
8049 It accepts the following values:
8052 Only evaluate expressions once during the filter initialization.
8055 Evaluate expressions for each incoming frame.
8058 Default value is @samp{init}.
8060 The filter accepts the following variables:
8063 The coordinates of the current sample.
8067 The width and height of the image.
8070 The number of input frame, starting from 0.
8073 @subsection Examples
8079 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8085 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8091 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8097 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8104 Extract a single field from an interlaced image using stride
8105 arithmetic to avoid wasting CPU time. The output frames are marked as
8108 The filter accepts the following options:
8112 Specify whether to extract the top (if the value is @code{0} or
8113 @code{top}) or the bottom field (if the value is @code{1} or
8119 Create new frames by copying the top and bottom fields from surrounding frames
8120 supplied as numbers by the hint file.
8124 Set file containing hints: absolute/relative frame numbers.
8126 There must be one line for each frame in a clip. Each line must contain two
8127 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8128 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8129 is current frame number for @code{absolute} mode or out of [-1, 1] range
8130 for @code{relative} mode. First number tells from which frame to pick up top
8131 field and second number tells from which frame to pick up bottom field.
8133 If optionally followed by @code{+} output frame will be marked as interlaced,
8134 else if followed by @code{-} output frame will be marked as progressive, else
8135 it will be marked same as input frame.
8136 If line starts with @code{#} or @code{;} that line is skipped.
8139 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8142 Example of first several lines of @code{hint} file for @code{relative} mode:
8145 1,0 - # second frame, use third's frame top field and second's frame bottom field
8146 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8163 Field matching filter for inverse telecine. It is meant to reconstruct the
8164 progressive frames from a telecined stream. The filter does not drop duplicated
8165 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8166 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8168 The separation of the field matching and the decimation is notably motivated by
8169 the possibility of inserting a de-interlacing filter fallback between the two.
8170 If the source has mixed telecined and real interlaced content,
8171 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8172 But these remaining combed frames will be marked as interlaced, and thus can be
8173 de-interlaced by a later filter such as @ref{yadif} before decimation.
8175 In addition to the various configuration options, @code{fieldmatch} can take an
8176 optional second stream, activated through the @option{ppsrc} option. If
8177 enabled, the frames reconstruction will be based on the fields and frames from
8178 this second stream. This allows the first input to be pre-processed in order to
8179 help the various algorithms of the filter, while keeping the output lossless
8180 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8181 or brightness/contrast adjustments can help.
8183 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8184 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8185 which @code{fieldmatch} is based on. While the semantic and usage are very
8186 close, some behaviour and options names can differ.
8188 The @ref{decimate} filter currently only works for constant frame rate input.
8189 If your input has mixed telecined (30fps) and progressive content with a lower
8190 framerate like 24fps use the following filterchain to produce the necessary cfr
8191 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8193 The filter accepts the following options:
8197 Specify the assumed field order of the input stream. Available values are:
8201 Auto detect parity (use FFmpeg's internal parity value).
8203 Assume bottom field first.
8205 Assume top field first.
8208 Note that it is sometimes recommended not to trust the parity announced by the
8211 Default value is @var{auto}.
8214 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8215 sense that it won't risk creating jerkiness due to duplicate frames when
8216 possible, but if there are bad edits or blended fields it will end up
8217 outputting combed frames when a good match might actually exist. On the other
8218 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8219 but will almost always find a good frame if there is one. The other values are
8220 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8221 jerkiness and creating duplicate frames versus finding good matches in sections
8222 with bad edits, orphaned fields, blended fields, etc.
8224 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8226 Available values are:
8230 2-way matching (p/c)
8232 2-way matching, and trying 3rd match if still combed (p/c + n)
8234 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8236 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8237 still combed (p/c + n + u/b)
8239 3-way matching (p/c/n)
8241 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8242 detected as combed (p/c/n + u/b)
8245 The parenthesis at the end indicate the matches that would be used for that
8246 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8249 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8252 Default value is @var{pc_n}.
8255 Mark the main input stream as a pre-processed input, and enable the secondary
8256 input stream as the clean source to pick the fields from. See the filter
8257 introduction for more details. It is similar to the @option{clip2} feature from
8260 Default value is @code{0} (disabled).
8263 Set the field to match from. It is recommended to set this to the same value as
8264 @option{order} unless you experience matching failures with that setting. In
8265 certain circumstances changing the field that is used to match from can have a
8266 large impact on matching performance. Available values are:
8270 Automatic (same value as @option{order}).
8272 Match from the bottom field.
8274 Match from the top field.
8277 Default value is @var{auto}.
8280 Set whether or not chroma is included during the match comparisons. In most
8281 cases it is recommended to leave this enabled. You should set this to @code{0}
8282 only if your clip has bad chroma problems such as heavy rainbowing or other
8283 artifacts. Setting this to @code{0} could also be used to speed things up at
8284 the cost of some accuracy.
8286 Default value is @code{1}.
8290 These define an exclusion band which excludes the lines between @option{y0} and
8291 @option{y1} from being included in the field matching decision. An exclusion
8292 band can be used to ignore subtitles, a logo, or other things that may
8293 interfere with the matching. @option{y0} sets the starting scan line and
8294 @option{y1} sets the ending line; all lines in between @option{y0} and
8295 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8296 @option{y0} and @option{y1} to the same value will disable the feature.
8297 @option{y0} and @option{y1} defaults to @code{0}.
8300 Set the scene change detection threshold as a percentage of maximum change on
8301 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8302 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8303 @option{scthresh} is @code{[0.0, 100.0]}.
8305 Default value is @code{12.0}.
8308 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8309 account the combed scores of matches when deciding what match to use as the
8310 final match. Available values are:
8314 No final matching based on combed scores.
8316 Combed scores are only used when a scene change is detected.
8318 Use combed scores all the time.
8321 Default is @var{sc}.
8324 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8325 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8326 Available values are:
8330 No forced calculation.
8332 Force p/c/n calculations.
8334 Force p/c/n/u/b calculations.
8337 Default value is @var{none}.
8340 This is the area combing threshold used for combed frame detection. This
8341 essentially controls how "strong" or "visible" combing must be to be detected.
8342 Larger values mean combing must be more visible and smaller values mean combing
8343 can be less visible or strong and still be detected. Valid settings are from
8344 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8345 be detected as combed). This is basically a pixel difference value. A good
8346 range is @code{[8, 12]}.
8348 Default value is @code{9}.
8351 Sets whether or not chroma is considered in the combed frame decision. Only
8352 disable this if your source has chroma problems (rainbowing, etc.) that are
8353 causing problems for the combed frame detection with chroma enabled. Actually,
8354 using @option{chroma}=@var{0} is usually more reliable, except for the case
8355 where there is chroma only combing in the source.
8357 Default value is @code{0}.
8361 Respectively set the x-axis and y-axis size of the window used during combed
8362 frame detection. This has to do with the size of the area in which
8363 @option{combpel} pixels are required to be detected as combed for a frame to be
8364 declared combed. See the @option{combpel} parameter description for more info.
8365 Possible values are any number that is a power of 2 starting at 4 and going up
8368 Default value is @code{16}.
8371 The number of combed pixels inside any of the @option{blocky} by
8372 @option{blockx} size blocks on the frame for the frame to be detected as
8373 combed. While @option{cthresh} controls how "visible" the combing must be, this
8374 setting controls "how much" combing there must be in any localized area (a
8375 window defined by the @option{blockx} and @option{blocky} settings) on the
8376 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8377 which point no frames will ever be detected as combed). This setting is known
8378 as @option{MI} in TFM/VFM vocabulary.
8380 Default value is @code{80}.
8383 @anchor{p/c/n/u/b meaning}
8384 @subsection p/c/n/u/b meaning
8386 @subsubsection p/c/n
8388 We assume the following telecined stream:
8391 Top fields: 1 2 2 3 4
8392 Bottom fields: 1 2 3 4 4
8395 The numbers correspond to the progressive frame the fields relate to. Here, the
8396 first two frames are progressive, the 3rd and 4th are combed, and so on.
8398 When @code{fieldmatch} is configured to run a matching from bottom
8399 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8404 B 1 2 3 4 4 <-- matching reference
8413 As a result of the field matching, we can see that some frames get duplicated.
8414 To perform a complete inverse telecine, you need to rely on a decimation filter
8415 after this operation. See for instance the @ref{decimate} filter.
8417 The same operation now matching from top fields (@option{field}=@var{top})
8422 T 1 2 2 3 4 <-- matching reference
8432 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8433 basically, they refer to the frame and field of the opposite parity:
8436 @item @var{p} matches the field of the opposite parity in the previous frame
8437 @item @var{c} matches the field of the opposite parity in the current frame
8438 @item @var{n} matches the field of the opposite parity in the next frame
8443 The @var{u} and @var{b} matching are a bit special in the sense that they match
8444 from the opposite parity flag. In the following examples, we assume that we are
8445 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8446 'x' is placed above and below each matched fields.
8448 With bottom matching (@option{field}=@var{bottom}):
8453 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8454 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8462 With top matching (@option{field}=@var{top}):
8467 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8468 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8476 @subsection Examples
8478 Simple IVTC of a top field first telecined stream:
8480 fieldmatch=order=tff:combmatch=none, decimate
8483 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8485 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8490 Transform the field order of the input video.
8492 It accepts the following parameters:
8497 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8498 for bottom field first.
8501 The default value is @samp{tff}.
8503 The transformation is done by shifting the picture content up or down
8504 by one line, and filling the remaining line with appropriate picture content.
8505 This method is consistent with most broadcast field order converters.
8507 If the input video is not flagged as being interlaced, or it is already
8508 flagged as being of the required output field order, then this filter does
8509 not alter the incoming video.
8511 It is very useful when converting to or from PAL DV material,
8512 which is bottom field first.
8516 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8519 @section fifo, afifo
8521 Buffer input images and send them when they are requested.
8523 It is mainly useful when auto-inserted by the libavfilter
8526 It does not take parameters.
8530 Find a rectangular object
8532 It accepts the following options:
8536 Filepath of the object image, needs to be in gray8.
8539 Detection threshold, default is 0.5.
8542 Number of mipmaps, default is 3.
8544 @item xmin, ymin, xmax, ymax
8545 Specifies the rectangle in which to search.
8548 @subsection Examples
8552 Generate a representative palette of a given video using @command{ffmpeg}:
8554 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8560 Cover a rectangular object
8562 It accepts the following options:
8566 Filepath of the optional cover image, needs to be in yuv420.
8571 It accepts the following values:
8574 cover it by the supplied image
8576 cover it by interpolating the surrounding pixels
8579 Default value is @var{blur}.
8582 @subsection Examples
8586 Generate a representative palette of a given video using @command{ffmpeg}:
8588 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8594 Flood area with values of same pixel components with another values.
8596 It accepts the following options:
8599 Set pixel x coordinate.
8602 Set pixel y coordinate.
8605 Set source #0 component value.
8608 Set source #1 component value.
8611 Set source #2 component value.
8614 Set source #3 component value.
8617 Set destination #0 component value.
8620 Set destination #1 component value.
8623 Set destination #2 component value.
8626 Set destination #3 component value.
8632 Convert the input video to one of the specified pixel formats.
8633 Libavfilter will try to pick one that is suitable as input to
8636 It accepts the following parameters:
8640 A '|'-separated list of pixel format names, such as
8641 "pix_fmts=yuv420p|monow|rgb24".
8645 @subsection Examples
8649 Convert the input video to the @var{yuv420p} format
8651 format=pix_fmts=yuv420p
8654 Convert the input video to any of the formats in the list
8656 format=pix_fmts=yuv420p|yuv444p|yuv410p
8663 Convert the video to specified constant frame rate by duplicating or dropping
8664 frames as necessary.
8666 It accepts the following parameters:
8670 The desired output frame rate. The default is @code{25}.
8675 Possible values are:
8678 zero round towards 0
8682 round towards -infinity
8684 round towards +infinity
8688 The default is @code{near}.
8691 Assume the first PTS should be the given value, in seconds. This allows for
8692 padding/trimming at the start of stream. By default, no assumption is made
8693 about the first frame's expected PTS, so no padding or trimming is done.
8694 For example, this could be set to 0 to pad the beginning with duplicates of
8695 the first frame if a video stream starts after the audio stream or to trim any
8696 frames with a negative PTS.
8700 Alternatively, the options can be specified as a flat string:
8701 @var{fps}[:@var{round}].
8703 See also the @ref{setpts} filter.
8705 @subsection Examples
8709 A typical usage in order to set the fps to 25:
8715 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8717 fps=fps=film:round=near
8723 Pack two different video streams into a stereoscopic video, setting proper
8724 metadata on supported codecs. The two views should have the same size and
8725 framerate and processing will stop when the shorter video ends. Please note
8726 that you may conveniently adjust view properties with the @ref{scale} and
8729 It accepts the following parameters:
8733 The desired packing format. Supported values are:
8738 The views are next to each other (default).
8741 The views are on top of each other.
8744 The views are packed by line.
8747 The views are packed by column.
8750 The views are temporally interleaved.
8759 # Convert left and right views into a frame-sequential video
8760 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8762 # Convert views into a side-by-side video with the same output resolution as the input
8763 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
8768 Change the frame rate by interpolating new video output frames from the source
8771 This filter is not designed to function correctly with interlaced media. If
8772 you wish to change the frame rate of interlaced media then you are required
8773 to deinterlace before this filter and re-interlace after this filter.
8775 A description of the accepted options follows.
8779 Specify the output frames per second. This option can also be specified
8780 as a value alone. The default is @code{50}.
8783 Specify the start of a range where the output frame will be created as a
8784 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8785 the default is @code{15}.
8788 Specify the end of a range where the output frame will be created as a
8789 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8790 the default is @code{240}.
8793 Specify the level at which a scene change is detected as a value between
8794 0 and 100 to indicate a new scene; a low value reflects a low
8795 probability for the current frame to introduce a new scene, while a higher
8796 value means the current frame is more likely to be one.
8797 The default is @code{7}.
8800 Specify flags influencing the filter process.
8802 Available value for @var{flags} is:
8805 @item scene_change_detect, scd
8806 Enable scene change detection using the value of the option @var{scene}.
8807 This flag is enabled by default.
8813 Select one frame every N-th frame.
8815 This filter accepts the following option:
8818 Select frame after every @code{step} frames.
8819 Allowed values are positive integers higher than 0. Default value is @code{1}.
8825 Apply a frei0r effect to the input video.
8827 To enable the compilation of this filter, you need to install the frei0r
8828 header and configure FFmpeg with @code{--enable-frei0r}.
8830 It accepts the following parameters:
8835 The name of the frei0r effect to load. If the environment variable
8836 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8837 directories specified by the colon-separated list in @env{FREI0R_PATH}.
8838 Otherwise, the standard frei0r paths are searched, in this order:
8839 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8840 @file{/usr/lib/frei0r-1/}.
8843 A '|'-separated list of parameters to pass to the frei0r effect.
8847 A frei0r effect parameter can be a boolean (its value is either
8848 "y" or "n"), a double, a color (specified as
8849 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8850 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8851 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8852 @var{X} and @var{Y} are floating point numbers) and/or a string.
8854 The number and types of parameters depend on the loaded effect. If an
8855 effect parameter is not specified, the default value is set.
8857 @subsection Examples
8861 Apply the distort0r effect, setting the first two double parameters:
8863 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8867 Apply the colordistance effect, taking a color as the first parameter:
8869 frei0r=colordistance:0.2/0.3/0.4
8870 frei0r=colordistance:violet
8871 frei0r=colordistance:0x112233
8875 Apply the perspective effect, specifying the top left and top right image
8878 frei0r=perspective:0.2/0.2|0.8/0.2
8882 For more information, see
8883 @url{http://frei0r.dyne.org}
8887 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8889 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8890 processing filter, one of them is performed once per block, not per pixel.
8891 This allows for much higher speed.
8893 The filter accepts the following options:
8897 Set quality. This option defines the number of levels for averaging. It accepts
8898 an integer in the range 4-5. Default value is @code{4}.
8901 Force a constant quantization parameter. It accepts an integer in range 0-63.
8902 If not set, the filter will use the QP from the video stream (if available).
8905 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8906 more details but also more artifacts, while higher values make the image smoother
8907 but also blurrier. Default value is @code{0} − PSNR optimal.
8910 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8911 option may cause flicker since the B-Frames have often larger QP. Default is
8912 @code{0} (not enabled).
8918 Apply Gaussian blur filter.
8920 The filter accepts the following options:
8924 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8927 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8930 Set which planes to filter. By default all planes are filtered.
8933 Set vertical sigma, if negative it will be same as @code{sigma}.
8934 Default is @code{-1}.
8939 The filter accepts the following options:
8943 Set the luminance expression.
8945 Set the chrominance blue expression.
8947 Set the chrominance red expression.
8949 Set the alpha expression.
8951 Set the red expression.
8953 Set the green expression.
8955 Set the blue expression.
8958 The colorspace is selected according to the specified options. If one
8959 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8960 options is specified, the filter will automatically select a YCbCr
8961 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8962 @option{blue_expr} options is specified, it will select an RGB
8965 If one of the chrominance expression is not defined, it falls back on the other
8966 one. If no alpha expression is specified it will evaluate to opaque value.
8967 If none of chrominance expressions are specified, they will evaluate
8968 to the luminance expression.
8970 The expressions can use the following variables and functions:
8974 The sequential number of the filtered frame, starting from @code{0}.
8978 The coordinates of the current sample.
8982 The width and height of the image.
8986 Width and height scale depending on the currently filtered plane. It is the
8987 ratio between the corresponding luma plane number of pixels and the current
8988 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8989 @code{0.5,0.5} for chroma planes.
8992 Time of the current frame, expressed in seconds.
8995 Return the value of the pixel at location (@var{x},@var{y}) of the current
8999 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9003 Return the value of the pixel at location (@var{x},@var{y}) of the
9004 blue-difference chroma plane. Return 0 if there is no such plane.
9007 Return the value of the pixel at location (@var{x},@var{y}) of the
9008 red-difference chroma plane. Return 0 if there is no such plane.
9013 Return the value of the pixel at location (@var{x},@var{y}) of the
9014 red/green/blue component. Return 0 if there is no such component.
9017 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9018 plane. Return 0 if there is no such plane.
9021 For functions, if @var{x} and @var{y} are outside the area, the value will be
9022 automatically clipped to the closer edge.
9024 @subsection Examples
9028 Flip the image horizontally:
9034 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9035 wavelength of 100 pixels:
9037 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9041 Generate a fancy enigmatic moving light:
9043 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
9047 Generate a quick emboss effect:
9049 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9053 Modify RGB components depending on pixel position:
9055 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9059 Create a radial gradient that is the same size as the input (also see
9060 the @ref{vignette} filter):
9062 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9068 Fix the banding artifacts that are sometimes introduced into nearly flat
9069 regions by truncation to 8-bit color depth.
9070 Interpolate the gradients that should go where the bands are, and
9073 It is designed for playback only. Do not use it prior to
9074 lossy compression, because compression tends to lose the dither and
9075 bring back the bands.
9077 It accepts the following parameters:
9082 The maximum amount by which the filter will change any one pixel. This is also
9083 the threshold for detecting nearly flat regions. Acceptable values range from
9084 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9088 The neighborhood to fit the gradient to. A larger radius makes for smoother
9089 gradients, but also prevents the filter from modifying the pixels near detailed
9090 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9091 values will be clipped to the valid range.
9095 Alternatively, the options can be specified as a flat string:
9096 @var{strength}[:@var{radius}]
9098 @subsection Examples
9102 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9108 Specify radius, omitting the strength (which will fall-back to the default
9119 Apply a Hald CLUT to a video stream.
9121 First input is the video stream to process, and second one is the Hald CLUT.
9122 The Hald CLUT input can be a simple picture or a complete video stream.
9124 The filter accepts the following options:
9128 Force termination when the shortest input terminates. Default is @code{0}.
9130 Continue applying the last CLUT after the end of the stream. A value of
9131 @code{0} disable the filter after the last frame of the CLUT is reached.
9132 Default is @code{1}.
9135 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9136 filters share the same internals).
9138 More information about the Hald CLUT can be found on Eskil Steenberg's website
9139 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9141 @subsection Workflow examples
9143 @subsubsection Hald CLUT video stream
9145 Generate an identity Hald CLUT stream altered with various effects:
9147 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
9150 Note: make sure you use a lossless codec.
9152 Then use it with @code{haldclut} to apply it on some random stream:
9154 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9157 The Hald CLUT will be applied to the 10 first seconds (duration of
9158 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9159 to the remaining frames of the @code{mandelbrot} stream.
9161 @subsubsection Hald CLUT with preview
9163 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9164 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9165 biggest possible square starting at the top left of the picture. The remaining
9166 padding pixels (bottom or right) will be ignored. This area can be used to add
9167 a preview of the Hald CLUT.
9169 Typically, the following generated Hald CLUT will be supported by the
9170 @code{haldclut} filter:
9173 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9174 pad=iw+320 [padded_clut];
9175 smptebars=s=320x256, split [a][b];
9176 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9177 [main][b] overlay=W-320" -frames:v 1 clut.png
9180 It contains the original and a preview of the effect of the CLUT: SMPTE color
9181 bars are displayed on the right-top, and below the same color bars processed by
9184 Then, the effect of this Hald CLUT can be visualized with:
9186 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9191 Flip the input video horizontally.
9193 For example, to horizontally flip the input video with @command{ffmpeg}:
9195 ffmpeg -i in.avi -vf "hflip" out.avi
9199 This filter applies a global color histogram equalization on a
9202 It can be used to correct video that has a compressed range of pixel
9203 intensities. The filter redistributes the pixel intensities to
9204 equalize their distribution across the intensity range. It may be
9205 viewed as an "automatically adjusting contrast filter". This filter is
9206 useful only for correcting degraded or poorly captured source
9209 The filter accepts the following options:
9213 Determine the amount of equalization to be applied. As the strength
9214 is reduced, the distribution of pixel intensities more-and-more
9215 approaches that of the input frame. The value must be a float number
9216 in the range [0,1] and defaults to 0.200.
9219 Set the maximum intensity that can generated and scale the output
9220 values appropriately. The strength should be set as desired and then
9221 the intensity can be limited if needed to avoid washing-out. The value
9222 must be a float number in the range [0,1] and defaults to 0.210.
9225 Set the antibanding level. If enabled the filter will randomly vary
9226 the luminance of output pixels by a small amount to avoid banding of
9227 the histogram. Possible values are @code{none}, @code{weak} or
9228 @code{strong}. It defaults to @code{none}.
9233 Compute and draw a color distribution histogram for the input video.
9235 The computed histogram is a representation of the color component
9236 distribution in an image.
9238 Standard histogram displays the color components distribution in an image.
9239 Displays color graph for each color component. Shows distribution of
9240 the Y, U, V, A or R, G, B components, depending on input format, in the
9241 current frame. Below each graph a color component scale meter is shown.
9243 The filter accepts the following options:
9247 Set height of level. Default value is @code{200}.
9248 Allowed range is [50, 2048].
9251 Set height of color scale. Default value is @code{12}.
9252 Allowed range is [0, 40].
9256 It accepts the following values:
9259 Per color component graphs are placed below each other.
9262 Per color component graphs are placed side by side.
9265 Presents information identical to that in the @code{parade}, except
9266 that the graphs representing color components are superimposed directly
9269 Default is @code{stack}.
9272 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9273 Default is @code{linear}.
9276 Set what color components to display.
9277 Default is @code{7}.
9280 Set foreground opacity. Default is @code{0.7}.
9283 Set background opacity. Default is @code{0.5}.
9286 @subsection Examples
9291 Calculate and draw histogram:
9293 ffplay -i input -vf histogram
9301 This is a high precision/quality 3d denoise filter. It aims to reduce
9302 image noise, producing smooth images and making still images really
9303 still. It should enhance compressibility.
9305 It accepts the following optional parameters:
9309 A non-negative floating point number which specifies spatial luma strength.
9312 @item chroma_spatial
9313 A non-negative floating point number which specifies spatial chroma strength.
9314 It defaults to 3.0*@var{luma_spatial}/4.0.
9317 A floating point number which specifies luma temporal strength. It defaults to
9318 6.0*@var{luma_spatial}/4.0.
9321 A floating point number which specifies chroma temporal strength. It defaults to
9322 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9327 Download hardware frames to system memory.
9329 The input must be in hardware frames, and the output a non-hardware format.
9330 Not all formats will be supported on the output - it may be necessary to insert
9331 an additional @option{format} filter immediately following in the graph to get
9332 the output in a supported format.
9336 Map hardware frames to system memory or to another device.
9338 This filter has several different modes of operation; which one is used depends
9339 on the input and output formats:
9342 Hardware frame input, normal frame output
9344 Map the input frames to system memory and pass them to the output. If the
9345 original hardware frame is later required (for example, after overlaying
9346 something else on part of it), the @option{hwmap} filter can be used again
9347 in the next mode to retrieve it.
9349 Normal frame input, hardware frame output
9351 If the input is actually a software-mapped hardware frame, then unmap it -
9352 that is, return the original hardware frame.
9354 Otherwise, a device must be provided. Create new hardware surfaces on that
9355 device for the output, then map them back to the software format at the input
9356 and give those frames to the preceding filter. This will then act like the
9357 @option{hwupload} filter, but may be able to avoid an additional copy when
9358 the input is already in a compatible format.
9360 Hardware frame input and output
9362 A device must be supplied for the output, either directly or with the
9363 @option{derive_device} option. The input and output devices must be of
9364 different types and compatible - the exact meaning of this is
9365 system-dependent, but typically it means that they must refer to the same
9366 underlying hardware context (for example, refer to the same graphics card).
9368 If the input frames were originally created on the output device, then unmap
9369 to retrieve the original frames.
9371 Otherwise, map the frames to the output device - create new hardware frames
9372 on the output corresponding to the frames on the input.
9375 The following additional parameters are accepted:
9379 Set the frame mapping mode. Some combination of:
9382 The mapped frame should be readable.
9384 The mapped frame should be writeable.
9386 The mapping will always overwrite the entire frame.
9388 This may improve performance in some cases, as the original contents of the
9389 frame need not be loaded.
9391 The mapping must not involve any copying.
9393 Indirect mappings to copies of frames are created in some cases where either
9394 direct mapping is not possible or it would have unexpected properties.
9395 Setting this flag ensures that the mapping is direct and will fail if that is
9398 Defaults to @var{read+write} if not specified.
9400 @item derive_device @var{type}
9401 Rather than using the device supplied at initialisation, instead derive a new
9402 device of type @var{type} from the device the input frames exist on.
9405 In a hardware to hardware mapping, map in reverse - create frames in the sink
9406 and map them back to the source. This may be necessary in some cases where
9407 a mapping in one direction is required but only the opposite direction is
9408 supported by the devices being used.
9410 This option is dangerous - it may break the preceding filter in undefined
9411 ways if there are any additional constraints on that filter's output.
9412 Do not use it without fully understanding the implications of its use.
9417 Upload system memory frames to hardware surfaces.
9419 The device to upload to must be supplied when the filter is initialised. If
9420 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9423 @anchor{hwupload_cuda}
9424 @section hwupload_cuda
9426 Upload system memory frames to a CUDA device.
9428 It accepts the following optional parameters:
9432 The number of the CUDA device to use
9437 Apply a high-quality magnification filter designed for pixel art. This filter
9438 was originally created by Maxim Stepin.
9440 It accepts the following option:
9444 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9445 @code{hq3x} and @code{4} for @code{hq4x}.
9446 Default is @code{3}.
9450 Stack input videos horizontally.
9452 All streams must be of same pixel format and of same height.
9454 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9455 to create same output.
9457 The filter accept the following option:
9461 Set number of input streams. Default is 2.
9464 If set to 1, force the output to terminate when the shortest input
9465 terminates. Default value is 0.
9470 Modify the hue and/or the saturation of the input.
9472 It accepts the following parameters:
9476 Specify the hue angle as a number of degrees. It accepts an expression,
9477 and defaults to "0".
9480 Specify the saturation in the [-10,10] range. It accepts an expression and
9484 Specify the hue angle as a number of radians. It accepts an
9485 expression, and defaults to "0".
9488 Specify the brightness in the [-10,10] range. It accepts an expression and
9492 @option{h} and @option{H} are mutually exclusive, and can't be
9493 specified at the same time.
9495 The @option{b}, @option{h}, @option{H} and @option{s} option values are
9496 expressions containing the following constants:
9500 frame count of the input frame starting from 0
9503 presentation timestamp of the input frame expressed in time base units
9506 frame rate of the input video, NAN if the input frame rate is unknown
9509 timestamp expressed in seconds, NAN if the input timestamp is unknown
9512 time base of the input video
9515 @subsection Examples
9519 Set the hue to 90 degrees and the saturation to 1.0:
9525 Same command but expressing the hue in radians:
9531 Rotate hue and make the saturation swing between 0
9532 and 2 over a period of 1 second:
9534 hue="H=2*PI*t: s=sin(2*PI*t)+1"
9538 Apply a 3 seconds saturation fade-in effect starting at 0:
9543 The general fade-in expression can be written as:
9545 hue="s=min(0\, max((t-START)/DURATION\, 1))"
9549 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
9551 hue="s=max(0\, min(1\, (8-t)/3))"
9554 The general fade-out expression can be written as:
9556 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
9561 @subsection Commands
9563 This filter supports the following commands:
9569 Modify the hue and/or the saturation and/or brightness of the input video.
9570 The command accepts the same syntax of the corresponding option.
9572 If the specified expression is not valid, it is kept at its current
9578 Grow first stream into second stream by connecting components.
9579 This makes it possible to build more robust edge masks.
9581 This filter accepts the following options:
9585 Set which planes will be processed as bitmap, unprocessed planes will be
9586 copied from first stream.
9587 By default value 0xf, all planes will be processed.
9590 Set threshold which is used in filtering. If pixel component value is higher than
9591 this value filter algorithm for connecting components is activated.
9592 By default value is 0.
9597 Detect video interlacing type.
9599 This filter tries to detect if the input frames are interlaced, progressive,
9600 top or bottom field first. It will also try to detect fields that are
9601 repeated between adjacent frames (a sign of telecine).
9603 Single frame detection considers only immediately adjacent frames when classifying each frame.
9604 Multiple frame detection incorporates the classification history of previous frames.
9606 The filter will log these metadata values:
9609 @item single.current_frame
9610 Detected type of current frame using single-frame detection. One of:
9611 ``tff'' (top field first), ``bff'' (bottom field first),
9612 ``progressive'', or ``undetermined''
9615 Cumulative number of frames detected as top field first using single-frame detection.
9618 Cumulative number of frames detected as top field first using multiple-frame detection.
9621 Cumulative number of frames detected as bottom field first using single-frame detection.
9623 @item multiple.current_frame
9624 Detected type of current frame using multiple-frame detection. One of:
9625 ``tff'' (top field first), ``bff'' (bottom field first),
9626 ``progressive'', or ``undetermined''
9629 Cumulative number of frames detected as bottom field first using multiple-frame detection.
9631 @item single.progressive
9632 Cumulative number of frames detected as progressive using single-frame detection.
9634 @item multiple.progressive
9635 Cumulative number of frames detected as progressive using multiple-frame detection.
9637 @item single.undetermined
9638 Cumulative number of frames that could not be classified using single-frame detection.
9640 @item multiple.undetermined
9641 Cumulative number of frames that could not be classified using multiple-frame detection.
9643 @item repeated.current_frame
9644 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
9646 @item repeated.neither
9647 Cumulative number of frames with no repeated field.
9650 Cumulative number of frames with the top field repeated from the previous frame's top field.
9652 @item repeated.bottom
9653 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
9656 The filter accepts the following options:
9660 Set interlacing threshold.
9662 Set progressive threshold.
9664 Threshold for repeated field detection.
9666 Number of frames after which a given frame's contribution to the
9667 statistics is halved (i.e., it contributes only 0.5 to its
9668 classification). The default of 0 means that all frames seen are given
9669 full weight of 1.0 forever.
9670 @item analyze_interlaced_flag
9671 When this is not 0 then idet will use the specified number of frames to determine
9672 if the interlaced flag is accurate, it will not count undetermined frames.
9673 If the flag is found to be accurate it will be used without any further
9674 computations, if it is found to be inaccurate it will be cleared without any
9675 further computations. This allows inserting the idet filter as a low computational
9676 method to clean up the interlaced flag
9681 Deinterleave or interleave fields.
9683 This filter allows one to process interlaced images fields without
9684 deinterlacing them. Deinterleaving splits the input frame into 2
9685 fields (so called half pictures). Odd lines are moved to the top
9686 half of the output image, even lines to the bottom half.
9687 You can process (filter) them independently and then re-interleave them.
9689 The filter accepts the following options:
9693 @item chroma_mode, c
9695 Available values for @var{luma_mode}, @var{chroma_mode} and
9696 @var{alpha_mode} are:
9702 @item deinterleave, d
9703 Deinterleave fields, placing one above the other.
9706 Interleave fields. Reverse the effect of deinterleaving.
9708 Default value is @code{none}.
9711 @item chroma_swap, cs
9712 @item alpha_swap, as
9713 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9718 Apply inflate effect to the video.
9720 This filter replaces the pixel by the local(3x3) average by taking into account
9721 only values higher than the pixel.
9723 It accepts the following options:
9730 Limit the maximum change for each plane, default is 65535.
9731 If 0, plane will remain unchanged.
9736 Simple interlacing filter from progressive contents. This interleaves upper (or
9737 lower) lines from odd frames with lower (or upper) lines from even frames,
9738 halving the frame rate and preserving image height.
9741 Original Original New Frame
9742 Frame 'j' Frame 'j+1' (tff)
9743 ========== =========== ==================
9744 Line 0 --------------------> Frame 'j' Line 0
9745 Line 1 Line 1 ----> Frame 'j+1' Line 1
9746 Line 2 ---------------------> Frame 'j' Line 2
9747 Line 3 Line 3 ----> Frame 'j+1' Line 3
9749 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9752 It accepts the following optional parameters:
9756 This determines whether the interlaced frame is taken from the even
9757 (tff - default) or odd (bff) lines of the progressive frame.
9760 Vertical lowpass filter to avoid twitter interlacing and
9761 reduce moire patterns.
9765 Disable vertical lowpass filter
9768 Enable linear filter (default)
9771 Enable complex filter. This will slightly less reduce twitter and moire
9772 but better retain detail and subjective sharpness impression.
9779 Deinterlace input video by applying Donald Graft's adaptive kernel
9780 deinterling. Work on interlaced parts of a video to produce
9783 The description of the accepted parameters follows.
9787 Set the threshold which affects the filter's tolerance when
9788 determining if a pixel line must be processed. It must be an integer
9789 in the range [0,255] and defaults to 10. A value of 0 will result in
9790 applying the process on every pixels.
9793 Paint pixels exceeding the threshold value to white if set to 1.
9797 Set the fields order. Swap fields if set to 1, leave fields alone if
9801 Enable additional sharpening if set to 1. Default is 0.
9804 Enable twoway sharpening if set to 1. Default is 0.
9807 @subsection Examples
9811 Apply default values:
9813 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9817 Enable additional sharpening:
9823 Paint processed pixels in white:
9829 @section lenscorrection
9831 Correct radial lens distortion
9833 This filter can be used to correct for radial distortion as can result from the use
9834 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9835 one can use tools available for example as part of opencv or simply trial-and-error.
9836 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9837 and extract the k1 and k2 coefficients from the resulting matrix.
9839 Note that effectively the same filter is available in the open-source tools Krita and
9840 Digikam from the KDE project.
9842 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9843 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9844 brightness distribution, so you may want to use both filters together in certain
9845 cases, though you will have to take care of ordering, i.e. whether vignetting should
9846 be applied before or after lens correction.
9850 The filter accepts the following options:
9854 Relative x-coordinate of the focal point of the image, and thereby the center of the
9855 distortion. This value has a range [0,1] and is expressed as fractions of the image
9858 Relative y-coordinate of the focal point of the image, and thereby the center of the
9859 distortion. This value has a range [0,1] and is expressed as fractions of the image
9862 Coefficient of the quadratic correction term. 0.5 means no correction.
9864 Coefficient of the double quadratic correction term. 0.5 means no correction.
9867 The formula that generates the correction is:
9869 @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)
9871 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9872 distances from the focal point in the source and target images, respectively.
9876 Obtain the average VMAF (Video Multi-Method Assessment Fusion)
9877 score between two input videos.
9879 This filter takes two input videos.
9881 Both video inputs must have the same resolution and pixel format for
9882 this filter to work correctly. Also it assumes that both inputs
9883 have the same number of frames, which are compared one by one.
9885 The obtained average VMAF score is printed through the logging system.
9887 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
9888 After installing the library it can be enabled using:
9889 @code{./configure --enable-libvmaf}.
9890 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
9892 On the below examples the input file @file{main.mpg} being processed is
9893 compared with the reference file @file{ref.mpg}.
9895 The filter has following options:
9899 Set the model path which is to be used for SVM.
9900 Default value: @code{"vmaf_v0.6.1.pkl"}
9903 Set the file path to be used to store logs.
9906 Set the format of the log file (xml or json).
9908 @item enable_transform
9909 Enables transform for computing vmaf.
9912 Invokes the phone model which will generate VMAF scores higher than in the
9913 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
9916 Enables computing psnr along with vmaf.
9919 Enables computing ssim along with vmaf.
9922 Enables computing ms_ssim along with vmaf.
9925 Set the pool method to be used for computing vmaf.
9928 This filter also supports the @ref{framesync} options.
9932 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
9935 Example with options:
9937 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
9942 Limits the pixel components values to the specified range [min, max].
9944 The filter accepts the following options:
9948 Lower bound. Defaults to the lowest allowed value for the input.
9951 Upper bound. Defaults to the highest allowed value for the input.
9954 Specify which planes will be processed. Defaults to all available.
9961 The filter accepts the following options:
9965 Set the number of loops.
9968 Set maximal size in number of frames.
9971 Set first frame of loop.
9977 Apply a 3D LUT to an input video.
9979 The filter accepts the following options:
9983 Set the 3D LUT file name.
9985 Currently supported formats:
9997 Select interpolation mode.
9999 Available values are:
10003 Use values from the nearest defined point.
10005 Interpolate values using the 8 points defining a cube.
10007 Interpolate values using a tetrahedron.
10011 This filter also supports the @ref{framesync} options.
10015 Turn certain luma values into transparency.
10017 The filter accepts the following options:
10021 Set the luma which will be used as base for transparency.
10022 Default value is @code{0}.
10025 Set the range of luma values to be keyed out.
10026 Default value is @code{0}.
10029 Set the range of softness. Default value is @code{0}.
10030 Use this to control gradual transition from zero to full transparency.
10033 @section lut, lutrgb, lutyuv
10035 Compute a look-up table for binding each pixel component input value
10036 to an output value, and apply it to the input video.
10038 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10039 to an RGB input video.
10041 These filters accept the following parameters:
10044 set first pixel component expression
10046 set second pixel component expression
10048 set third pixel component expression
10050 set fourth pixel component expression, corresponds to the alpha component
10053 set red component expression
10055 set green component expression
10057 set blue component expression
10059 alpha component expression
10062 set Y/luminance component expression
10064 set U/Cb component expression
10066 set V/Cr component expression
10069 Each of them specifies the expression to use for computing the lookup table for
10070 the corresponding pixel component values.
10072 The exact component associated to each of the @var{c*} options depends on the
10075 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10076 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10078 The expressions can contain the following constants and functions:
10083 The input width and height.
10086 The input value for the pixel component.
10089 The input value, clipped to the @var{minval}-@var{maxval} range.
10092 The maximum value for the pixel component.
10095 The minimum value for the pixel component.
10098 The negated value for the pixel component value, clipped to the
10099 @var{minval}-@var{maxval} range; it corresponds to the expression
10100 "maxval-clipval+minval".
10103 The computed value in @var{val}, clipped to the
10104 @var{minval}-@var{maxval} range.
10106 @item gammaval(gamma)
10107 The computed gamma correction value of the pixel component value,
10108 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10110 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10114 All expressions default to "val".
10116 @subsection Examples
10120 Negate input video:
10122 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10123 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10126 The above is the same as:
10128 lutrgb="r=negval:g=negval:b=negval"
10129 lutyuv="y=negval:u=negval:v=negval"
10139 Remove chroma components, turning the video into a graytone image:
10141 lutyuv="u=128:v=128"
10145 Apply a luma burning effect:
10151 Remove green and blue components:
10157 Set a constant alpha channel value on input:
10159 format=rgba,lutrgb=a="maxval-minval/2"
10163 Correct luminance gamma by a factor of 0.5:
10165 lutyuv=y=gammaval(0.5)
10169 Discard least significant bits of luma:
10171 lutyuv=y='bitand(val, 128+64+32)'
10175 Technicolor like effect:
10177 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10181 @section lut2, tlut2
10183 The @code{lut2} filter takes two input streams and outputs one
10186 The @code{tlut2} (time lut2) filter takes two consecutive frames
10187 from one single stream.
10189 This filter accepts the following parameters:
10192 set first pixel component expression
10194 set second pixel component expression
10196 set third pixel component expression
10198 set fourth pixel component expression, corresponds to the alpha component
10201 Each of them specifies the expression to use for computing the lookup table for
10202 the corresponding pixel component values.
10204 The exact component associated to each of the @var{c*} options depends on the
10207 The expressions can contain the following constants:
10212 The input width and height.
10215 The first input value for the pixel component.
10218 The second input value for the pixel component.
10221 The first input video bit depth.
10224 The second input video bit depth.
10227 All expressions default to "x".
10229 @subsection Examples
10233 Highlight differences between two RGB video streams:
10235 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)'
10239 Highlight differences between two YUV video streams:
10241 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)'
10245 Show max difference between two video streams:
10247 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)))'
10251 @section maskedclamp
10253 Clamp the first input stream with the second input and third input stream.
10255 Returns the value of first stream to be between second input
10256 stream - @code{undershoot} and third input stream + @code{overshoot}.
10258 This filter accepts the following options:
10261 Default value is @code{0}.
10264 Default value is @code{0}.
10267 Set which planes will be processed as bitmap, unprocessed planes will be
10268 copied from first stream.
10269 By default value 0xf, all planes will be processed.
10272 @section maskedmerge
10274 Merge the first input stream with the second input stream using per pixel
10275 weights in the third input stream.
10277 A value of 0 in the third stream pixel component means that pixel component
10278 from first stream is returned unchanged, while maximum value (eg. 255 for
10279 8-bit videos) means that pixel component from second stream is returned
10280 unchanged. Intermediate values define the amount of merging between both
10281 input stream's pixel components.
10283 This filter accepts the following options:
10286 Set which planes will be processed as bitmap, unprocessed planes will be
10287 copied from first stream.
10288 By default value 0xf, all planes will be processed.
10293 Apply motion-compensation deinterlacing.
10295 It needs one field per frame as input and must thus be used together
10296 with yadif=1/3 or equivalent.
10298 This filter accepts the following options:
10301 Set the deinterlacing mode.
10303 It accepts one of the following values:
10308 use iterative motion estimation
10310 like @samp{slow}, but use multiple reference frames.
10312 Default value is @samp{fast}.
10315 Set the picture field parity assumed for the input video. It must be
10316 one of the following values:
10320 assume top field first
10322 assume bottom field first
10325 Default value is @samp{bff}.
10328 Set per-block quantization parameter (QP) used by the internal
10331 Higher values should result in a smoother motion vector field but less
10332 optimal individual vectors. Default value is 1.
10335 @section mergeplanes
10337 Merge color channel components from several video streams.
10339 The filter accepts up to 4 input streams, and merge selected input
10340 planes to the output video.
10342 This filter accepts the following options:
10345 Set input to output plane mapping. Default is @code{0}.
10347 The mappings is specified as a bitmap. It should be specified as a
10348 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10349 mapping for the first plane of the output stream. 'A' sets the number of
10350 the input stream to use (from 0 to 3), and 'a' the plane number of the
10351 corresponding input to use (from 0 to 3). The rest of the mappings is
10352 similar, 'Bb' describes the mapping for the output stream second
10353 plane, 'Cc' describes the mapping for the output stream third plane and
10354 'Dd' describes the mapping for the output stream fourth plane.
10357 Set output pixel format. Default is @code{yuva444p}.
10360 @subsection Examples
10364 Merge three gray video streams of same width and height into single video stream:
10366 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10370 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10372 [a0][a1]mergeplanes=0x00010210:yuva444p
10376 Swap Y and A plane in yuva444p stream:
10378 format=yuva444p,mergeplanes=0x03010200:yuva444p
10382 Swap U and V plane in yuv420p stream:
10384 format=yuv420p,mergeplanes=0x000201:yuv420p
10388 Cast a rgb24 clip to yuv444p:
10390 format=rgb24,mergeplanes=0x000102:yuv444p
10396 Estimate and export motion vectors using block matching algorithms.
10397 Motion vectors are stored in frame side data to be used by other filters.
10399 This filter accepts the following options:
10402 Specify the motion estimation method. Accepts one of the following values:
10406 Exhaustive search algorithm.
10408 Three step search algorithm.
10410 Two dimensional logarithmic search algorithm.
10412 New three step search algorithm.
10414 Four step search algorithm.
10416 Diamond search algorithm.
10418 Hexagon-based search algorithm.
10420 Enhanced predictive zonal search algorithm.
10422 Uneven multi-hexagon search algorithm.
10424 Default value is @samp{esa}.
10427 Macroblock size. Default @code{16}.
10430 Search parameter. Default @code{7}.
10433 @section midequalizer
10435 Apply Midway Image Equalization effect using two video streams.
10437 Midway Image Equalization adjusts a pair of images to have the same
10438 histogram, while maintaining their dynamics as much as possible. It's
10439 useful for e.g. matching exposures from a pair of stereo cameras.
10441 This filter has two inputs and one output, which must be of same pixel format, but
10442 may be of different sizes. The output of filter is first input adjusted with
10443 midway histogram of both inputs.
10445 This filter accepts the following option:
10449 Set which planes to process. Default is @code{15}, which is all available planes.
10452 @section minterpolate
10454 Convert the video to specified frame rate using motion interpolation.
10456 This filter accepts the following options:
10459 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}.
10462 Motion interpolation mode. Following values are accepted:
10465 Duplicate previous or next frame for interpolating new ones.
10467 Blend source frames. Interpolated frame is mean of previous and next frames.
10469 Motion compensated interpolation. Following options are effective when this mode is selected:
10473 Motion compensation mode. Following values are accepted:
10476 Overlapped block motion compensation.
10478 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10480 Default mode is @samp{obmc}.
10483 Motion estimation mode. Following values are accepted:
10486 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
10488 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
10490 Default mode is @samp{bilat}.
10493 The algorithm to be used for motion estimation. Following values are accepted:
10496 Exhaustive search algorithm.
10498 Three step search algorithm.
10500 Two dimensional logarithmic search algorithm.
10502 New three step search algorithm.
10504 Four step search algorithm.
10506 Diamond search algorithm.
10508 Hexagon-based search algorithm.
10510 Enhanced predictive zonal search algorithm.
10512 Uneven multi-hexagon search algorithm.
10514 Default algorithm is @samp{epzs}.
10517 Macroblock size. Default @code{16}.
10520 Motion estimation search parameter. Default @code{32}.
10523 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).
10528 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:
10531 Disable scene change detection.
10533 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
10535 Default method is @samp{fdiff}.
10537 @item scd_threshold
10538 Scene change detection threshold. Default is @code{5.0}.
10541 @section mpdecimate
10543 Drop frames that do not differ greatly from the previous frame in
10544 order to reduce frame rate.
10546 The main use of this filter is for very-low-bitrate encoding
10547 (e.g. streaming over dialup modem), but it could in theory be used for
10548 fixing movies that were inverse-telecined incorrectly.
10550 A description of the accepted options follows.
10554 Set the maximum number of consecutive frames which can be dropped (if
10555 positive), or the minimum interval between dropped frames (if
10556 negative). If the value is 0, the frame is dropped disregarding the
10557 number of previous sequentially dropped frames.
10559 Default value is 0.
10564 Set the dropping threshold values.
10566 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
10567 represent actual pixel value differences, so a threshold of 64
10568 corresponds to 1 unit of difference for each pixel, or the same spread
10569 out differently over the block.
10571 A frame is a candidate for dropping if no 8x8 blocks differ by more
10572 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
10573 meaning the whole image) differ by more than a threshold of @option{lo}.
10575 Default value for @option{hi} is 64*12, default value for @option{lo} is
10576 64*5, and default value for @option{frac} is 0.33.
10582 Negate input video.
10584 It accepts an integer in input; if non-zero it negates the
10585 alpha component (if available). The default value in input is 0.
10589 Denoise frames using Non-Local Means algorithm.
10591 Each pixel is adjusted by looking for other pixels with similar contexts. This
10592 context similarity is defined by comparing their surrounding patches of size
10593 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
10596 Note that the research area defines centers for patches, which means some
10597 patches will be made of pixels outside that research area.
10599 The filter accepts the following options.
10603 Set denoising strength.
10609 Same as @option{p} but for chroma planes.
10611 The default value is @var{0} and means automatic.
10617 Same as @option{r} but for chroma planes.
10619 The default value is @var{0} and means automatic.
10624 Deinterlace video using neural network edge directed interpolation.
10626 This filter accepts the following options:
10630 Mandatory option, without binary file filter can not work.
10631 Currently file can be found here:
10632 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
10635 Set which frames to deinterlace, by default it is @code{all}.
10636 Can be @code{all} or @code{interlaced}.
10639 Set mode of operation.
10641 Can be one of the following:
10645 Use frame flags, both fields.
10647 Use frame flags, single field.
10649 Use top field only.
10651 Use bottom field only.
10653 Use both fields, top first.
10655 Use both fields, bottom first.
10659 Set which planes to process, by default filter process all frames.
10662 Set size of local neighborhood around each pixel, used by the predictor neural
10665 Can be one of the following:
10678 Set the number of neurons in predictor neural network.
10679 Can be one of the following:
10690 Controls the number of different neural network predictions that are blended
10691 together to compute the final output value. Can be @code{fast}, default or
10695 Set which set of weights to use in the predictor.
10696 Can be one of the following:
10700 weights trained to minimize absolute error
10702 weights trained to minimize squared error
10706 Controls whether or not the prescreener neural network is used to decide
10707 which pixels should be processed by the predictor neural network and which
10708 can be handled by simple cubic interpolation.
10709 The prescreener is trained to know whether cubic interpolation will be
10710 sufficient for a pixel or whether it should be predicted by the predictor nn.
10711 The computational complexity of the prescreener nn is much less than that of
10712 the predictor nn. Since most pixels can be handled by cubic interpolation,
10713 using the prescreener generally results in much faster processing.
10714 The prescreener is pretty accurate, so the difference between using it and not
10715 using it is almost always unnoticeable.
10717 Can be one of the following:
10725 Default is @code{new}.
10728 Set various debugging flags.
10733 Force libavfilter not to use any of the specified pixel formats for the
10734 input to the next filter.
10736 It accepts the following parameters:
10740 A '|'-separated list of pixel format names, such as
10741 pix_fmts=yuv420p|monow|rgb24".
10745 @subsection Examples
10749 Force libavfilter to use a format different from @var{yuv420p} for the
10750 input to the vflip filter:
10752 noformat=pix_fmts=yuv420p,vflip
10756 Convert the input video to any of the formats not contained in the list:
10758 noformat=yuv420p|yuv444p|yuv410p
10764 Add noise on video input frame.
10766 The filter accepts the following options:
10774 Set noise seed for specific pixel component or all pixel components in case
10775 of @var{all_seed}. Default value is @code{123457}.
10777 @item all_strength, alls
10778 @item c0_strength, c0s
10779 @item c1_strength, c1s
10780 @item c2_strength, c2s
10781 @item c3_strength, c3s
10782 Set noise strength for specific pixel component or all pixel components in case
10783 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
10785 @item all_flags, allf
10786 @item c0_flags, c0f
10787 @item c1_flags, c1f
10788 @item c2_flags, c2f
10789 @item c3_flags, c3f
10790 Set pixel component flags or set flags for all components if @var{all_flags}.
10791 Available values for component flags are:
10794 averaged temporal noise (smoother)
10796 mix random noise with a (semi)regular pattern
10798 temporal noise (noise pattern changes between frames)
10800 uniform noise (gaussian otherwise)
10804 @subsection Examples
10806 Add temporal and uniform noise to input video:
10808 noise=alls=20:allf=t+u
10813 Pass the video source unchanged to the output.
10816 Optical Character Recognition
10818 This filter uses Tesseract for optical character recognition.
10820 It accepts the following options:
10824 Set datapath to tesseract data. Default is to use whatever was
10825 set at installation.
10828 Set language, default is "eng".
10831 Set character whitelist.
10834 Set character blacklist.
10837 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
10841 Apply a video transform using libopencv.
10843 To enable this filter, install the libopencv library and headers and
10844 configure FFmpeg with @code{--enable-libopencv}.
10846 It accepts the following parameters:
10851 The name of the libopencv filter to apply.
10853 @item filter_params
10854 The parameters to pass to the libopencv filter. If not specified, the default
10855 values are assumed.
10859 Refer to the official libopencv documentation for more precise
10861 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
10863 Several libopencv filters are supported; see the following subsections.
10868 Dilate an image by using a specific structuring element.
10869 It corresponds to the libopencv function @code{cvDilate}.
10871 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10873 @var{struct_el} represents a structuring element, and has the syntax:
10874 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10876 @var{cols} and @var{rows} represent the number of columns and rows of
10877 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10878 point, and @var{shape} the shape for the structuring element. @var{shape}
10879 must be "rect", "cross", "ellipse", or "custom".
10881 If the value for @var{shape} is "custom", it must be followed by a
10882 string of the form "=@var{filename}". The file with name
10883 @var{filename} is assumed to represent a binary image, with each
10884 printable character corresponding to a bright pixel. When a custom
10885 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10886 or columns and rows of the read file are assumed instead.
10888 The default value for @var{struct_el} is "3x3+0x0/rect".
10890 @var{nb_iterations} specifies the number of times the transform is
10891 applied to the image, and defaults to 1.
10895 # Use the default values
10898 # Dilate using a structuring element with a 5x5 cross, iterating two times
10899 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10901 # Read the shape from the file diamond.shape, iterating two times.
10902 # The file diamond.shape may contain a pattern of characters like this
10908 # The specified columns and rows are ignored
10909 # but the anchor point coordinates are not
10910 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10915 Erode an image by using a specific structuring element.
10916 It corresponds to the libopencv function @code{cvErode}.
10918 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10919 with the same syntax and semantics as the @ref{dilate} filter.
10923 Smooth the input video.
10925 The filter takes the following parameters:
10926 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10928 @var{type} is the type of smooth filter to apply, and must be one of
10929 the following values: "blur", "blur_no_scale", "median", "gaussian",
10930 or "bilateral". The default value is "gaussian".
10932 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10933 depend on the smooth type. @var{param1} and
10934 @var{param2} accept integer positive values or 0. @var{param3} and
10935 @var{param4} accept floating point values.
10937 The default value for @var{param1} is 3. The default value for the
10938 other parameters is 0.
10940 These parameters correspond to the parameters assigned to the
10941 libopencv function @code{cvSmooth}.
10943 @section oscilloscope
10945 2D Video Oscilloscope.
10947 Useful to measure spatial impulse, step responses, chroma delays, etc.
10949 It accepts the following parameters:
10953 Set scope center x position.
10956 Set scope center y position.
10959 Set scope size, relative to frame diagonal.
10962 Set scope tilt/rotation.
10968 Set trace center x position.
10971 Set trace center y position.
10974 Set trace width, relative to width of frame.
10977 Set trace height, relative to height of frame.
10980 Set which components to trace. By default it traces first three components.
10983 Draw trace grid. By default is enabled.
10986 Draw some statistics. By default is enabled.
10989 Draw scope. By default is enabled.
10992 @subsection Examples
10996 Inspect full first row of video frame.
10998 oscilloscope=x=0.5:y=0:s=1
11002 Inspect full last row of video frame.
11004 oscilloscope=x=0.5:y=1:s=1
11008 Inspect full 5th line of video frame of height 1080.
11010 oscilloscope=x=0.5:y=5/1080:s=1
11014 Inspect full last column of video frame.
11016 oscilloscope=x=1:y=0.5:s=1:t=1
11024 Overlay one video on top of another.
11026 It takes two inputs and has one output. The first input is the "main"
11027 video on which the second input is overlaid.
11029 It accepts the following parameters:
11031 A description of the accepted options follows.
11036 Set the expression for the x and y coordinates of the overlaid video
11037 on the main video. Default value is "0" for both expressions. In case
11038 the expression is invalid, it is set to a huge value (meaning that the
11039 overlay will not be displayed within the output visible area).
11042 See @ref{framesync}.
11045 Set when the expressions for @option{x}, and @option{y} are evaluated.
11047 It accepts the following values:
11050 only evaluate expressions once during the filter initialization or
11051 when a command is processed
11054 evaluate expressions for each incoming frame
11057 Default value is @samp{frame}.
11060 See @ref{framesync}.
11063 Set the format for the output video.
11065 It accepts the following values:
11068 force YUV420 output
11071 force YUV422 output
11074 force YUV444 output
11077 force packed RGB output
11080 force planar RGB output
11083 automatically pick format
11086 Default value is @samp{yuv420}.
11089 See @ref{framesync}.
11092 The @option{x}, and @option{y} expressions can contain the following
11098 The main input width and height.
11102 The overlay input width and height.
11106 The computed values for @var{x} and @var{y}. They are evaluated for
11111 horizontal and vertical chroma subsample values of the output
11112 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11116 the number of input frame, starting from 0
11119 the position in the file of the input frame, NAN if unknown
11122 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11126 This filter also supports the @ref{framesync} options.
11128 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11129 when evaluation is done @emph{per frame}, and will evaluate to NAN
11130 when @option{eval} is set to @samp{init}.
11132 Be aware that frames are taken from each input video in timestamp
11133 order, hence, if their initial timestamps differ, it is a good idea
11134 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11135 have them begin in the same zero timestamp, as the example for
11136 the @var{movie} filter does.
11138 You can chain together more overlays but you should test the
11139 efficiency of such approach.
11141 @subsection Commands
11143 This filter supports the following commands:
11147 Modify the x and y of the overlay input.
11148 The command accepts the same syntax of the corresponding option.
11150 If the specified expression is not valid, it is kept at its current
11154 @subsection Examples
11158 Draw the overlay at 10 pixels from the bottom right corner of the main
11161 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11164 Using named options the example above becomes:
11166 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11170 Insert a transparent PNG logo in the bottom left corner of the input,
11171 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11173 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11177 Insert 2 different transparent PNG logos (second logo on bottom
11178 right corner) using the @command{ffmpeg} tool:
11180 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
11184 Add a transparent color layer on top of the main video; @code{WxH}
11185 must specify the size of the main input to the overlay filter:
11187 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11191 Play an original video and a filtered version (here with the deshake
11192 filter) side by side using the @command{ffplay} tool:
11194 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11197 The above command is the same as:
11199 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11203 Make a sliding overlay appearing from the left to the right top part of the
11204 screen starting since time 2:
11206 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11210 Compose output by putting two input videos side to side:
11212 ffmpeg -i left.avi -i right.avi -filter_complex "
11213 nullsrc=size=200x100 [background];
11214 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11215 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11216 [background][left] overlay=shortest=1 [background+left];
11217 [background+left][right] overlay=shortest=1:x=100 [left+right]
11222 Mask 10-20 seconds of a video by applying the delogo filter to a section
11224 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11225 -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]'
11230 Chain several overlays in cascade:
11232 nullsrc=s=200x200 [bg];
11233 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11234 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11235 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11236 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11237 [in3] null, [mid2] overlay=100:100 [out0]
11244 Apply Overcomplete Wavelet denoiser.
11246 The filter accepts the following options:
11252 Larger depth values will denoise lower frequency components more, but
11253 slow down filtering.
11255 Must be an int in the range 8-16, default is @code{8}.
11257 @item luma_strength, ls
11260 Must be a double value in the range 0-1000, default is @code{1.0}.
11262 @item chroma_strength, cs
11263 Set chroma strength.
11265 Must be a double value in the range 0-1000, default is @code{1.0}.
11271 Add paddings to the input image, and place the original input at the
11272 provided @var{x}, @var{y} coordinates.
11274 It accepts the following parameters:
11279 Specify an expression for the size of the output image with the
11280 paddings added. If the value for @var{width} or @var{height} is 0, the
11281 corresponding input size is used for the output.
11283 The @var{width} expression can reference the value set by the
11284 @var{height} expression, and vice versa.
11286 The default value of @var{width} and @var{height} is 0.
11290 Specify the offsets to place the input image at within the padded area,
11291 with respect to the top/left border of the output image.
11293 The @var{x} expression can reference the value set by the @var{y}
11294 expression, and vice versa.
11296 The default value of @var{x} and @var{y} is 0.
11298 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11299 so the input image is centered on the padded area.
11302 Specify the color of the padded area. For the syntax of this option,
11303 check the "Color" section in the ffmpeg-utils manual.
11305 The default value of @var{color} is "black".
11308 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11310 It accepts the following values:
11314 Only evaluate expressions once during the filter initialization or when
11315 a command is processed.
11318 Evaluate expressions for each incoming frame.
11322 Default value is @samp{init}.
11325 Pad to aspect instead to a resolution.
11329 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11330 options are expressions containing the following constants:
11335 The input video width and height.
11339 These are the same as @var{in_w} and @var{in_h}.
11343 The output width and height (the size of the padded area), as
11344 specified by the @var{width} and @var{height} expressions.
11348 These are the same as @var{out_w} and @var{out_h}.
11352 The x and y offsets as specified by the @var{x} and @var{y}
11353 expressions, or NAN if not yet specified.
11356 same as @var{iw} / @var{ih}
11359 input sample aspect ratio
11362 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11366 The horizontal and vertical chroma subsample values. For example for the
11367 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11370 @subsection Examples
11374 Add paddings with the color "violet" to the input video. The output video
11375 size is 640x480, and the top-left corner of the input video is placed at
11378 pad=640:480:0:40:violet
11381 The example above is equivalent to the following command:
11383 pad=width=640:height=480:x=0:y=40:color=violet
11387 Pad the input to get an output with dimensions increased by 3/2,
11388 and put the input video at the center of the padded area:
11390 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
11394 Pad the input to get a squared output with size equal to the maximum
11395 value between the input width and height, and put the input video at
11396 the center of the padded area:
11398 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
11402 Pad the input to get a final w/h ratio of 16:9:
11404 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
11408 In case of anamorphic video, in order to set the output display aspect
11409 correctly, it is necessary to use @var{sar} in the expression,
11410 according to the relation:
11412 (ih * X / ih) * sar = output_dar
11413 X = output_dar / sar
11416 Thus the previous example needs to be modified to:
11418 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
11422 Double the output size and put the input video in the bottom-right
11423 corner of the output padded area:
11425 pad="2*iw:2*ih:ow-iw:oh-ih"
11429 @anchor{palettegen}
11430 @section palettegen
11432 Generate one palette for a whole video stream.
11434 It accepts the following options:
11438 Set the maximum number of colors to quantize in the palette.
11439 Note: the palette will still contain 256 colors; the unused palette entries
11442 @item reserve_transparent
11443 Create a palette of 255 colors maximum and reserve the last one for
11444 transparency. Reserving the transparency color is useful for GIF optimization.
11445 If not set, the maximum of colors in the palette will be 256. You probably want
11446 to disable this option for a standalone image.
11450 Set statistics mode.
11452 It accepts the following values:
11455 Compute full frame histograms.
11457 Compute histograms only for the part that differs from previous frame. This
11458 might be relevant to give more importance to the moving part of your input if
11459 the background is static.
11461 Compute new histogram for each frame.
11464 Default value is @var{full}.
11467 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
11468 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
11469 color quantization of the palette. This information is also visible at
11470 @var{info} logging level.
11472 @subsection Examples
11476 Generate a representative palette of a given video using @command{ffmpeg}:
11478 ffmpeg -i input.mkv -vf palettegen palette.png
11482 @section paletteuse
11484 Use a palette to downsample an input video stream.
11486 The filter takes two inputs: one video stream and a palette. The palette must
11487 be a 256 pixels image.
11489 It accepts the following options:
11493 Select dithering mode. Available algorithms are:
11496 Ordered 8x8 bayer dithering (deterministic)
11498 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
11499 Note: this dithering is sometimes considered "wrong" and is included as a
11501 @item floyd_steinberg
11502 Floyd and Steingberg dithering (error diffusion)
11504 Frankie Sierra dithering v2 (error diffusion)
11506 Frankie Sierra dithering v2 "Lite" (error diffusion)
11509 Default is @var{sierra2_4a}.
11512 When @var{bayer} dithering is selected, this option defines the scale of the
11513 pattern (how much the crosshatch pattern is visible). A low value means more
11514 visible pattern for less banding, and higher value means less visible pattern
11515 at the cost of more banding.
11517 The option must be an integer value in the range [0,5]. Default is @var{2}.
11520 If set, define the zone to process
11524 Only the changing rectangle will be reprocessed. This is similar to GIF
11525 cropping/offsetting compression mechanism. This option can be useful for speed
11526 if only a part of the image is changing, and has use cases such as limiting the
11527 scope of the error diffusal @option{dither} to the rectangle that bounds the
11528 moving scene (it leads to more deterministic output if the scene doesn't change
11529 much, and as a result less moving noise and better GIF compression).
11532 Default is @var{none}.
11535 Take new palette for each output frame.
11538 @subsection Examples
11542 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
11543 using @command{ffmpeg}:
11545 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
11549 @section perspective
11551 Correct perspective of video not recorded perpendicular to the screen.
11553 A description of the accepted parameters follows.
11564 Set coordinates expression for top left, top right, bottom left and bottom right corners.
11565 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
11566 If the @code{sense} option is set to @code{source}, then the specified points will be sent
11567 to the corners of the destination. If the @code{sense} option is set to @code{destination},
11568 then the corners of the source will be sent to the specified coordinates.
11570 The expressions can use the following variables:
11575 the width and height of video frame.
11579 Output frame count.
11582 @item interpolation
11583 Set interpolation for perspective correction.
11585 It accepts the following values:
11591 Default value is @samp{linear}.
11594 Set interpretation of coordinate options.
11596 It accepts the following values:
11600 Send point in the source specified by the given coordinates to
11601 the corners of the destination.
11603 @item 1, destination
11605 Send the corners of the source to the point in the destination specified
11606 by the given coordinates.
11608 Default value is @samp{source}.
11612 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
11614 It accepts the following values:
11617 only evaluate expressions once during the filter initialization or
11618 when a command is processed
11621 evaluate expressions for each incoming frame
11624 Default value is @samp{init}.
11629 Delay interlaced video by one field time so that the field order changes.
11631 The intended use is to fix PAL movies that have been captured with the
11632 opposite field order to the film-to-video transfer.
11634 A description of the accepted parameters follows.
11640 It accepts the following values:
11643 Capture field order top-first, transfer bottom-first.
11644 Filter will delay the bottom field.
11647 Capture field order bottom-first, transfer top-first.
11648 Filter will delay the top field.
11651 Capture and transfer with the same field order. This mode only exists
11652 for the documentation of the other options to refer to, but if you
11653 actually select it, the filter will faithfully do nothing.
11656 Capture field order determined automatically by field flags, transfer
11658 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
11659 basis using field flags. If no field information is available,
11660 then this works just like @samp{u}.
11663 Capture unknown or varying, transfer opposite.
11664 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
11665 analyzing the images and selecting the alternative that produces best
11666 match between the fields.
11669 Capture top-first, transfer unknown or varying.
11670 Filter selects among @samp{t} and @samp{p} using image analysis.
11673 Capture bottom-first, transfer unknown or varying.
11674 Filter selects among @samp{b} and @samp{p} using image analysis.
11677 Capture determined by field flags, transfer unknown or varying.
11678 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
11679 image analysis. If no field information is available, then this works just
11680 like @samp{U}. This is the default mode.
11683 Both capture and transfer unknown or varying.
11684 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
11688 @section pixdesctest
11690 Pixel format descriptor test filter, mainly useful for internal
11691 testing. The output video should be equal to the input video.
11695 format=monow, pixdesctest
11698 can be used to test the monowhite pixel format descriptor definition.
11702 Display sample values of color channels. Mainly useful for checking color and levels.
11704 The filters accept the following options:
11708 Set scope X position, relative offset on X axis.
11711 Set scope Y position, relative offset on Y axis.
11720 Set window opacity. This window also holds statistics about pixel area.
11723 Set window X position, relative offset on X axis.
11726 Set window Y position, relative offset on Y axis.
11731 Enable the specified chain of postprocessing subfilters using libpostproc. This
11732 library should be automatically selected with a GPL build (@code{--enable-gpl}).
11733 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
11734 Each subfilter and some options have a short and a long name that can be used
11735 interchangeably, i.e. dr/dering are the same.
11737 The filters accept the following options:
11741 Set postprocessing subfilters string.
11744 All subfilters share common options to determine their scope:
11748 Honor the quality commands for this subfilter.
11751 Do chrominance filtering, too (default).
11754 Do luminance filtering only (no chrominance).
11757 Do chrominance filtering only (no luminance).
11760 These options can be appended after the subfilter name, separated by a '|'.
11762 Available subfilters are:
11765 @item hb/hdeblock[|difference[|flatness]]
11766 Horizontal deblocking filter
11769 Difference factor where higher values mean more deblocking (default: @code{32}).
11771 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11774 @item vb/vdeblock[|difference[|flatness]]
11775 Vertical deblocking filter
11778 Difference factor where higher values mean more deblocking (default: @code{32}).
11780 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11783 @item ha/hadeblock[|difference[|flatness]]
11784 Accurate horizontal deblocking filter
11787 Difference factor where higher values mean more deblocking (default: @code{32}).
11789 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11792 @item va/vadeblock[|difference[|flatness]]
11793 Accurate vertical deblocking filter
11796 Difference factor where higher values mean more deblocking (default: @code{32}).
11798 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11802 The horizontal and vertical deblocking filters share the difference and
11803 flatness values so you cannot set different horizontal and vertical
11807 @item h1/x1hdeblock
11808 Experimental horizontal deblocking filter
11810 @item v1/x1vdeblock
11811 Experimental vertical deblocking filter
11816 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
11819 larger -> stronger filtering
11821 larger -> stronger filtering
11823 larger -> stronger filtering
11826 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
11829 Stretch luminance to @code{0-255}.
11832 @item lb/linblenddeint
11833 Linear blend deinterlacing filter that deinterlaces the given block by
11834 filtering all lines with a @code{(1 2 1)} filter.
11836 @item li/linipoldeint
11837 Linear interpolating deinterlacing filter that deinterlaces the given block by
11838 linearly interpolating every second line.
11840 @item ci/cubicipoldeint
11841 Cubic interpolating deinterlacing filter deinterlaces the given block by
11842 cubically interpolating every second line.
11844 @item md/mediandeint
11845 Median deinterlacing filter that deinterlaces the given block by applying a
11846 median filter to every second line.
11848 @item fd/ffmpegdeint
11849 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
11850 second line with a @code{(-1 4 2 4 -1)} filter.
11853 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
11854 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
11856 @item fq/forceQuant[|quantizer]
11857 Overrides the quantizer table from the input with the constant quantizer you
11865 Default pp filter combination (@code{hb|a,vb|a,dr|a})
11868 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
11871 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
11874 @subsection Examples
11878 Apply horizontal and vertical deblocking, deringing and automatic
11879 brightness/contrast:
11885 Apply default filters without brightness/contrast correction:
11891 Apply default filters and temporal denoiser:
11893 pp=default/tmpnoise|1|2|3
11897 Apply deblocking on luminance only, and switch vertical deblocking on or off
11898 automatically depending on available CPU time:
11905 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
11906 similar to spp = 6 with 7 point DCT, where only the center sample is
11909 The filter accepts the following options:
11913 Force a constant quantization parameter. It accepts an integer in range
11914 0 to 63. If not set, the filter will use the QP from the video stream
11918 Set thresholding mode. Available modes are:
11922 Set hard thresholding.
11924 Set soft thresholding (better de-ringing effect, but likely blurrier).
11926 Set medium thresholding (good results, default).
11930 @section premultiply
11931 Apply alpha premultiply effect to input video stream using first plane
11932 of second stream as alpha.
11934 Both streams must have same dimensions and same pixel format.
11936 The filter accepts the following option:
11940 Set which planes will be processed, unprocessed planes will be copied.
11941 By default value 0xf, all planes will be processed.
11944 Do not require 2nd input for processing, instead use alpha plane from input stream.
11948 Apply prewitt operator to input video stream.
11950 The filter accepts the following option:
11954 Set which planes will be processed, unprocessed planes will be copied.
11955 By default value 0xf, all planes will be processed.
11958 Set value which will be multiplied with filtered result.
11961 Set value which will be added to filtered result.
11964 @section pseudocolor
11966 Alter frame colors in video with pseudocolors.
11968 This filter accept the following options:
11972 set pixel first component expression
11975 set pixel second component expression
11978 set pixel third component expression
11981 set pixel fourth component expression, corresponds to the alpha component
11984 set component to use as base for altering colors
11987 Each of them specifies the expression to use for computing the lookup table for
11988 the corresponding pixel component values.
11990 The expressions can contain the following constants and functions:
11995 The input width and height.
11998 The input value for the pixel component.
12000 @item ymin, umin, vmin, amin
12001 The minimum allowed component value.
12003 @item ymax, umax, vmax, amax
12004 The maximum allowed component value.
12007 All expressions default to "val".
12009 @subsection Examples
12013 Change too high luma values to gradient:
12015 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'"
12021 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12022 Ratio) between two input videos.
12024 This filter takes in input two input videos, the first input is
12025 considered the "main" source and is passed unchanged to the
12026 output. The second input is used as a "reference" video for computing
12029 Both video inputs must have the same resolution and pixel format for
12030 this filter to work correctly. Also it assumes that both inputs
12031 have the same number of frames, which are compared one by one.
12033 The obtained average PSNR is printed through the logging system.
12035 The filter stores the accumulated MSE (mean squared error) of each
12036 frame, and at the end of the processing it is averaged across all frames
12037 equally, and the following formula is applied to obtain the PSNR:
12040 PSNR = 10*log10(MAX^2/MSE)
12043 Where MAX is the average of the maximum values of each component of the
12046 The description of the accepted parameters follows.
12049 @item stats_file, f
12050 If specified the filter will use the named file to save the PSNR of
12051 each individual frame. When filename equals "-" the data is sent to
12054 @item stats_version
12055 Specifies which version of the stats file format to use. Details of
12056 each format are written below.
12057 Default value is 1.
12059 @item stats_add_max
12060 Determines whether the max value is output to the stats log.
12061 Default value is 0.
12062 Requires stats_version >= 2. If this is set and stats_version < 2,
12063 the filter will return an error.
12066 This filter also supports the @ref{framesync} options.
12068 The file printed if @var{stats_file} is selected, contains a sequence of
12069 key/value pairs of the form @var{key}:@var{value} for each compared
12072 If a @var{stats_version} greater than 1 is specified, a header line precedes
12073 the list of per-frame-pair stats, with key value pairs following the frame
12074 format with the following parameters:
12077 @item psnr_log_version
12078 The version of the log file format. Will match @var{stats_version}.
12081 A comma separated list of the per-frame-pair parameters included in
12085 A description of each shown per-frame-pair parameter follows:
12089 sequential number of the input frame, starting from 1
12092 Mean Square Error pixel-by-pixel average difference of the compared
12093 frames, averaged over all the image components.
12095 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
12096 Mean Square Error pixel-by-pixel average difference of the compared
12097 frames for the component specified by the suffix.
12099 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12100 Peak Signal to Noise ratio of the compared frames for the component
12101 specified by the suffix.
12103 @item max_avg, max_y, max_u, max_v
12104 Maximum allowed value for each channel, and average over all
12110 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12111 [main][ref] psnr="stats_file=stats.log" [out]
12114 On this example the input file being processed is compared with the
12115 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12116 is stored in @file{stats.log}.
12121 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12122 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12125 The pullup filter is designed to take advantage of future context in making
12126 its decisions. This filter is stateless in the sense that it does not lock
12127 onto a pattern to follow, but it instead looks forward to the following
12128 fields in order to identify matches and rebuild progressive frames.
12130 To produce content with an even framerate, insert the fps filter after
12131 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12132 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12134 The filter accepts the following options:
12141 These options set the amount of "junk" to ignore at the left, right, top, and
12142 bottom of the image, respectively. Left and right are in units of 8 pixels,
12143 while top and bottom are in units of 2 lines.
12144 The default is 8 pixels on each side.
12147 Set the strict breaks. Setting this option to 1 will reduce the chances of
12148 filter generating an occasional mismatched frame, but it may also cause an
12149 excessive number of frames to be dropped during high motion sequences.
12150 Conversely, setting it to -1 will make filter match fields more easily.
12151 This may help processing of video where there is slight blurring between
12152 the fields, but may also cause there to be interlaced frames in the output.
12153 Default value is @code{0}.
12156 Set the metric plane to use. It accepts the following values:
12162 Use chroma blue plane.
12165 Use chroma red plane.
12168 This option may be set to use chroma plane instead of the default luma plane
12169 for doing filter's computations. This may improve accuracy on very clean
12170 source material, but more likely will decrease accuracy, especially if there
12171 is chroma noise (rainbow effect) or any grayscale video.
12172 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12173 load and make pullup usable in realtime on slow machines.
12176 For best results (without duplicated frames in the output file) it is
12177 necessary to change the output frame rate. For example, to inverse
12178 telecine NTSC input:
12180 ffmpeg -i input -vf pullup -r 24000/1001 ...
12185 Change video quantization parameters (QP).
12187 The filter accepts the following option:
12191 Set expression for quantization parameter.
12194 The expression is evaluated through the eval API and can contain, among others,
12195 the following constants:
12199 1 if index is not 129, 0 otherwise.
12202 Sequential index starting from -129 to 128.
12205 @subsection Examples
12209 Some equation like:
12217 Flush video frames from internal cache of frames into a random order.
12218 No frame is discarded.
12219 Inspired by @ref{frei0r} nervous filter.
12223 Set size in number of frames of internal cache, in range from @code{2} to
12224 @code{512}. Default is @code{30}.
12227 Set seed for random number generator, must be an integer included between
12228 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12229 less than @code{0}, the filter will try to use a good random seed on a
12233 @section readeia608
12235 Read closed captioning (EIA-608) information from the top lines of a video frame.
12237 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
12238 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
12239 with EIA-608 data (starting from 0). A description of each metadata value follows:
12242 @item lavfi.readeia608.X.cc
12243 The two bytes stored as EIA-608 data (printed in hexadecimal).
12245 @item lavfi.readeia608.X.line
12246 The number of the line on which the EIA-608 data was identified and read.
12249 This filter accepts the following options:
12253 Set the line to start scanning for EIA-608 data. Default is @code{0}.
12256 Set the line to end scanning for EIA-608 data. Default is @code{29}.
12259 Set minimal acceptable amplitude change for sync codes detection.
12260 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
12263 Set the ratio of width reserved for sync code detection.
12264 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
12267 Set the max peaks height difference for sync code detection.
12268 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12271 Set max peaks period difference for sync code detection.
12272 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12275 Set the first two max start code bits differences.
12276 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
12279 Set the minimum ratio of bits height compared to 3rd start code bit.
12280 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
12283 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
12286 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
12289 Enable checking the parity bit. In the event of a parity error, the filter will output
12290 @code{0x00} for that character. Default is false.
12293 @subsection Examples
12297 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
12299 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
12305 Read vertical interval timecode (VITC) information from the top lines of a
12308 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
12309 timecode value, if a valid timecode has been detected. Further metadata key
12310 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
12311 timecode data has been found or not.
12313 This filter accepts the following options:
12317 Set the maximum number of lines to scan for VITC data. If the value is set to
12318 @code{-1} the full video frame is scanned. Default is @code{45}.
12321 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
12322 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
12325 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
12326 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
12329 @subsection Examples
12333 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
12334 draw @code{--:--:--:--} as a placeholder:
12336 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
12342 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
12344 Destination pixel at position (X, Y) will be picked from source (x, y) position
12345 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
12346 value for pixel will be used for destination pixel.
12348 Xmap and Ymap input video streams must be of same dimensions. Output video stream
12349 will have Xmap/Ymap video stream dimensions.
12350 Xmap and Ymap input video streams are 16bit depth, single channel.
12352 @section removegrain
12354 The removegrain filter is a spatial denoiser for progressive video.
12358 Set mode for the first plane.
12361 Set mode for the second plane.
12364 Set mode for the third plane.
12367 Set mode for the fourth plane.
12370 Range of mode is from 0 to 24. Description of each mode follows:
12374 Leave input plane unchanged. Default.
12377 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
12380 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
12383 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
12386 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
12387 This is equivalent to a median filter.
12390 Line-sensitive clipping giving the minimal change.
12393 Line-sensitive clipping, intermediate.
12396 Line-sensitive clipping, intermediate.
12399 Line-sensitive clipping, intermediate.
12402 Line-sensitive clipping on a line where the neighbours pixels are the closest.
12405 Replaces the target pixel with the closest neighbour.
12408 [1 2 1] horizontal and vertical kernel blur.
12414 Bob mode, interpolates top field from the line where the neighbours
12415 pixels are the closest.
12418 Bob mode, interpolates bottom field from the line where the neighbours
12419 pixels are the closest.
12422 Bob mode, interpolates top field. Same as 13 but with a more complicated
12423 interpolation formula.
12426 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
12427 interpolation formula.
12430 Clips the pixel with the minimum and maximum of respectively the maximum and
12431 minimum of each pair of opposite neighbour pixels.
12434 Line-sensitive clipping using opposite neighbours whose greatest distance from
12435 the current pixel is minimal.
12438 Replaces the pixel with the average of its 8 neighbours.
12441 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
12444 Clips pixels using the averages of opposite neighbour.
12447 Same as mode 21 but simpler and faster.
12450 Small edge and halo removal, but reputed useless.
12456 @section removelogo
12458 Suppress a TV station logo, using an image file to determine which
12459 pixels comprise the logo. It works by filling in the pixels that
12460 comprise the logo with neighboring pixels.
12462 The filter accepts the following options:
12466 Set the filter bitmap file, which can be any image format supported by
12467 libavformat. The width and height of the image file must match those of the
12468 video stream being processed.
12471 Pixels in the provided bitmap image with a value of zero are not
12472 considered part of the logo, non-zero pixels are considered part of
12473 the logo. If you use white (255) for the logo and black (0) for the
12474 rest, you will be safe. For making the filter bitmap, it is
12475 recommended to take a screen capture of a black frame with the logo
12476 visible, and then using a threshold filter followed by the erode
12477 filter once or twice.
12479 If needed, little splotches can be fixed manually. Remember that if
12480 logo pixels are not covered, the filter quality will be much
12481 reduced. Marking too many pixels as part of the logo does not hurt as
12482 much, but it will increase the amount of blurring needed to cover over
12483 the image and will destroy more information than necessary, and extra
12484 pixels will slow things down on a large logo.
12486 @section repeatfields
12488 This filter uses the repeat_field flag from the Video ES headers and hard repeats
12489 fields based on its value.
12493 Reverse a video clip.
12495 Warning: This filter requires memory to buffer the entire clip, so trimming
12498 @subsection Examples
12502 Take the first 5 seconds of a clip, and reverse it.
12509 Apply roberts cross operator to input video stream.
12511 The filter accepts the following option:
12515 Set which planes will be processed, unprocessed planes will be copied.
12516 By default value 0xf, all planes will be processed.
12519 Set value which will be multiplied with filtered result.
12522 Set value which will be added to filtered result.
12527 Rotate video by an arbitrary angle expressed in radians.
12529 The filter accepts the following options:
12531 A description of the optional parameters follows.
12534 Set an expression for the angle by which to rotate the input video
12535 clockwise, expressed as a number of radians. A negative value will
12536 result in a counter-clockwise rotation. By default it is set to "0".
12538 This expression is evaluated for each frame.
12541 Set the output width expression, default value is "iw".
12542 This expression is evaluated just once during configuration.
12545 Set the output height expression, default value is "ih".
12546 This expression is evaluated just once during configuration.
12549 Enable bilinear interpolation if set to 1, a value of 0 disables
12550 it. Default value is 1.
12553 Set the color used to fill the output area not covered by the rotated
12554 image. For the general syntax of this option, check the "Color" section in the
12555 ffmpeg-utils manual. If the special value "none" is selected then no
12556 background is printed (useful for example if the background is never shown).
12558 Default value is "black".
12561 The expressions for the angle and the output size can contain the
12562 following constants and functions:
12566 sequential number of the input frame, starting from 0. It is always NAN
12567 before the first frame is filtered.
12570 time in seconds of the input frame, it is set to 0 when the filter is
12571 configured. It is always NAN before the first frame is filtered.
12575 horizontal and vertical chroma subsample values. For example for the
12576 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12580 the input video width and height
12584 the output width and height, that is the size of the padded area as
12585 specified by the @var{width} and @var{height} expressions
12589 the minimal width/height required for completely containing the input
12590 video rotated by @var{a} radians.
12592 These are only available when computing the @option{out_w} and
12593 @option{out_h} expressions.
12596 @subsection Examples
12600 Rotate the input by PI/6 radians clockwise:
12606 Rotate the input by PI/6 radians counter-clockwise:
12612 Rotate the input by 45 degrees clockwise:
12618 Apply a constant rotation with period T, starting from an angle of PI/3:
12620 rotate=PI/3+2*PI*t/T
12624 Make the input video rotation oscillating with a period of T
12625 seconds and an amplitude of A radians:
12627 rotate=A*sin(2*PI/T*t)
12631 Rotate the video, output size is chosen so that the whole rotating
12632 input video is always completely contained in the output:
12634 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
12638 Rotate the video, reduce the output size so that no background is ever
12641 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
12645 @subsection Commands
12647 The filter supports the following commands:
12651 Set the angle expression.
12652 The command accepts the same syntax of the corresponding option.
12654 If the specified expression is not valid, it is kept at its current
12660 Apply Shape Adaptive Blur.
12662 The filter accepts the following options:
12665 @item luma_radius, lr
12666 Set luma blur filter strength, must be a value in range 0.1-4.0, default
12667 value is 1.0. A greater value will result in a more blurred image, and
12668 in slower processing.
12670 @item luma_pre_filter_radius, lpfr
12671 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
12674 @item luma_strength, ls
12675 Set luma maximum difference between pixels to still be considered, must
12676 be a value in the 0.1-100.0 range, default value is 1.0.
12678 @item chroma_radius, cr
12679 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
12680 greater value will result in a more blurred image, and in slower
12683 @item chroma_pre_filter_radius, cpfr
12684 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
12686 @item chroma_strength, cs
12687 Set chroma maximum difference between pixels to still be considered,
12688 must be a value in the -0.9-100.0 range.
12691 Each chroma option value, if not explicitly specified, is set to the
12692 corresponding luma option value.
12697 Scale (resize) the input video, using the libswscale library.
12699 The scale filter forces the output display aspect ratio to be the same
12700 of the input, by changing the output sample aspect ratio.
12702 If the input image format is different from the format requested by
12703 the next filter, the scale filter will convert the input to the
12706 @subsection Options
12707 The filter accepts the following options, or any of the options
12708 supported by the libswscale scaler.
12710 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
12711 the complete list of scaler options.
12716 Set the output video dimension expression. Default value is the input
12719 If the @var{width} or @var{w} value is 0, the input width is used for
12720 the output. If the @var{height} or @var{h} value is 0, the input height
12721 is used for the output.
12723 If one and only one of the values is -n with n >= 1, the scale filter
12724 will use a value that maintains the aspect ratio of the input image,
12725 calculated from the other specified dimension. After that it will,
12726 however, make sure that the calculated dimension is divisible by n and
12727 adjust the value if necessary.
12729 If both values are -n with n >= 1, the behavior will be identical to
12730 both values being set to 0 as previously detailed.
12732 See below for the list of accepted constants for use in the dimension
12736 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
12740 Only evaluate expressions once during the filter initialization or when a command is processed.
12743 Evaluate expressions for each incoming frame.
12747 Default value is @samp{init}.
12751 Set the interlacing mode. It accepts the following values:
12755 Force interlaced aware scaling.
12758 Do not apply interlaced scaling.
12761 Select interlaced aware scaling depending on whether the source frames
12762 are flagged as interlaced or not.
12765 Default value is @samp{0}.
12768 Set libswscale scaling flags. See
12769 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12770 complete list of values. If not explicitly specified the filter applies
12774 @item param0, param1
12775 Set libswscale input parameters for scaling algorithms that need them. See
12776 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12777 complete documentation. If not explicitly specified the filter applies
12783 Set the video size. For the syntax of this option, check the
12784 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12786 @item in_color_matrix
12787 @item out_color_matrix
12788 Set in/output YCbCr color space type.
12790 This allows the autodetected value to be overridden as well as allows forcing
12791 a specific value used for the output and encoder.
12793 If not specified, the color space type depends on the pixel format.
12799 Choose automatically.
12802 Format conforming to International Telecommunication Union (ITU)
12803 Recommendation BT.709.
12806 Set color space conforming to the United States Federal Communications
12807 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
12810 Set color space conforming to:
12814 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
12817 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
12820 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
12825 Set color space conforming to SMPTE ST 240:1999.
12830 Set in/output YCbCr sample range.
12832 This allows the autodetected value to be overridden as well as allows forcing
12833 a specific value used for the output and encoder. If not specified, the
12834 range depends on the pixel format. Possible values:
12838 Choose automatically.
12841 Set full range (0-255 in case of 8-bit luma).
12844 Set "MPEG" range (16-235 in case of 8-bit luma).
12847 @item force_original_aspect_ratio
12848 Enable decreasing or increasing output video width or height if necessary to
12849 keep the original aspect ratio. Possible values:
12853 Scale the video as specified and disable this feature.
12856 The output video dimensions will automatically be decreased if needed.
12859 The output video dimensions will automatically be increased if needed.
12863 One useful instance of this option is that when you know a specific device's
12864 maximum allowed resolution, you can use this to limit the output video to
12865 that, while retaining the aspect ratio. For example, device A allows
12866 1280x720 playback, and your video is 1920x800. Using this option (set it to
12867 decrease) and specifying 1280x720 to the command line makes the output
12870 Please note that this is a different thing than specifying -1 for @option{w}
12871 or @option{h}, you still need to specify the output resolution for this option
12876 The values of the @option{w} and @option{h} options are expressions
12877 containing the following constants:
12882 The input width and height
12886 These are the same as @var{in_w} and @var{in_h}.
12890 The output (scaled) width and height
12894 These are the same as @var{out_w} and @var{out_h}
12897 The same as @var{iw} / @var{ih}
12900 input sample aspect ratio
12903 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
12907 horizontal and vertical input chroma subsample values. For example for the
12908 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12912 horizontal and vertical output chroma subsample values. For example for the
12913 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12916 @subsection Examples
12920 Scale the input video to a size of 200x100
12925 This is equivalent to:
12936 Specify a size abbreviation for the output size:
12941 which can also be written as:
12947 Scale the input to 2x:
12949 scale=w=2*iw:h=2*ih
12953 The above is the same as:
12955 scale=2*in_w:2*in_h
12959 Scale the input to 2x with forced interlaced scaling:
12961 scale=2*iw:2*ih:interl=1
12965 Scale the input to half size:
12967 scale=w=iw/2:h=ih/2
12971 Increase the width, and set the height to the same size:
12977 Seek Greek harmony:
12984 Increase the height, and set the width to 3/2 of the height:
12986 scale=w=3/2*oh:h=3/5*ih
12990 Increase the size, making the size a multiple of the chroma
12993 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
12997 Increase the width to a maximum of 500 pixels,
12998 keeping the same aspect ratio as the input:
13000 scale=w='min(500\, iw*3/2):h=-1'
13004 @subsection Commands
13006 This filter supports the following commands:
13010 Set the output video dimension expression.
13011 The command accepts the same syntax of the corresponding option.
13013 If the specified expression is not valid, it is kept at its current
13019 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13020 format conversion on CUDA video frames. Setting the output width and height
13021 works in the same way as for the @var{scale} filter.
13023 The following additional options are accepted:
13026 The pixel format of the output CUDA frames. If set to the string "same" (the
13027 default), the input format will be kept. Note that automatic format negotiation
13028 and conversion is not yet supported for hardware frames
13031 The interpolation algorithm used for resizing. One of the following:
13038 @item cubic2p_bspline
13039 2-parameter cubic (B=1, C=0)
13041 @item cubic2p_catmullrom
13042 2-parameter cubic (B=0, C=1/2)
13044 @item cubic2p_b05c03
13045 2-parameter cubic (B=1/2, C=3/10)
13057 Scale (resize) the input video, based on a reference video.
13059 See the scale filter for available options, scale2ref supports the same but
13060 uses the reference video instead of the main input as basis. scale2ref also
13061 supports the following additional constants for the @option{w} and
13062 @option{h} options:
13067 The main input video's width and height
13070 The same as @var{main_w} / @var{main_h}
13073 The main input video's sample aspect ratio
13075 @item main_dar, mdar
13076 The main input video's display aspect ratio. Calculated from
13077 @code{(main_w / main_h) * main_sar}.
13081 The main input video's horizontal and vertical chroma subsample values.
13082 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13086 @subsection Examples
13090 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13092 'scale2ref[b][a];[a][b]overlay'
13096 @anchor{selectivecolor}
13097 @section selectivecolor
13099 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13100 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13101 by the "purity" of the color (that is, how saturated it already is).
13103 This filter is similar to the Adobe Photoshop Selective Color tool.
13105 The filter accepts the following options:
13108 @item correction_method
13109 Select color correction method.
13111 Available values are:
13114 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13117 Specified adjustments are relative to the original component value.
13119 Default is @code{absolute}.
13121 Adjustments for red pixels (pixels where the red component is the maximum)
13123 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13125 Adjustments for green pixels (pixels where the green component is the maximum)
13127 Adjustments for cyan pixels (pixels where the red component is the minimum)
13129 Adjustments for blue pixels (pixels where the blue component is the maximum)
13131 Adjustments for magenta pixels (pixels where the green component is the minimum)
13133 Adjustments for white pixels (pixels where all components are greater than 128)
13135 Adjustments for all pixels except pure black and pure white
13137 Adjustments for black pixels (pixels where all components are lesser than 128)
13139 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13142 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13143 4 space separated floating point adjustment values in the [-1,1] range,
13144 respectively to adjust the amount of cyan, magenta, yellow and black for the
13145 pixels of its range.
13147 @subsection Examples
13151 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13152 increase magenta by 27% in blue areas:
13154 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13158 Use a Photoshop selective color preset:
13160 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13164 @anchor{separatefields}
13165 @section separatefields
13167 The @code{separatefields} takes a frame-based video input and splits
13168 each frame into its components fields, producing a new half height clip
13169 with twice the frame rate and twice the frame count.
13171 This filter use field-dominance information in frame to decide which
13172 of each pair of fields to place first in the output.
13173 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13175 @section setdar, setsar
13177 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13180 This is done by changing the specified Sample (aka Pixel) Aspect
13181 Ratio, according to the following equation:
13183 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13186 Keep in mind that the @code{setdar} filter does not modify the pixel
13187 dimensions of the video frame. Also, the display aspect ratio set by
13188 this filter may be changed by later filters in the filterchain,
13189 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13192 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13193 the filter output video.
13195 Note that as a consequence of the application of this filter, the
13196 output display aspect ratio will change according to the equation
13199 Keep in mind that the sample aspect ratio set by the @code{setsar}
13200 filter may be changed by later filters in the filterchain, e.g. if
13201 another "setsar" or a "setdar" filter is applied.
13203 It accepts the following parameters:
13206 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13207 Set the aspect ratio used by the filter.
13209 The parameter can be a floating point number string, an expression, or
13210 a string of the form @var{num}:@var{den}, where @var{num} and
13211 @var{den} are the numerator and denominator of the aspect ratio. If
13212 the parameter is not specified, it is assumed the value "0".
13213 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13217 Set the maximum integer value to use for expressing numerator and
13218 denominator when reducing the expressed aspect ratio to a rational.
13219 Default value is @code{100}.
13223 The parameter @var{sar} is an expression containing
13224 the following constants:
13228 These are approximated values for the mathematical constants e
13229 (Euler's number), pi (Greek pi), and phi (the golden ratio).
13232 The input width and height.
13235 These are the same as @var{w} / @var{h}.
13238 The input sample aspect ratio.
13241 The input display aspect ratio. It is the same as
13242 (@var{w} / @var{h}) * @var{sar}.
13245 Horizontal and vertical chroma subsample values. For example, for the
13246 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13249 @subsection Examples
13254 To change the display aspect ratio to 16:9, specify one of the following:
13261 To change the sample aspect ratio to 10:11, specify:
13267 To set a display aspect ratio of 16:9, and specify a maximum integer value of
13268 1000 in the aspect ratio reduction, use the command:
13270 setdar=ratio=16/9:max=1000
13278 Force field for the output video frame.
13280 The @code{setfield} filter marks the interlace type field for the
13281 output frames. It does not change the input frame, but only sets the
13282 corresponding property, which affects how the frame is treated by
13283 following filters (e.g. @code{fieldorder} or @code{yadif}).
13285 The filter accepts the following options:
13290 Available values are:
13294 Keep the same field property.
13297 Mark the frame as bottom-field-first.
13300 Mark the frame as top-field-first.
13303 Mark the frame as progressive.
13309 Show a line containing various information for each input video frame.
13310 The input video is not modified.
13312 The shown line contains a sequence of key/value pairs of the form
13313 @var{key}:@var{value}.
13315 The following values are shown in the output:
13319 The (sequential) number of the input frame, starting from 0.
13322 The Presentation TimeStamp of the input frame, expressed as a number of
13323 time base units. The time base unit depends on the filter input pad.
13326 The Presentation TimeStamp of the input frame, expressed as a number of
13330 The position of the frame in the input stream, or -1 if this information is
13331 unavailable and/or meaningless (for example in case of synthetic video).
13334 The pixel format name.
13337 The sample aspect ratio of the input frame, expressed in the form
13338 @var{num}/@var{den}.
13341 The size of the input frame. For the syntax of this option, check the
13342 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13345 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
13346 for bottom field first).
13349 This is 1 if the frame is a key frame, 0 otherwise.
13352 The picture type of the input frame ("I" for an I-frame, "P" for a
13353 P-frame, "B" for a B-frame, or "?" for an unknown type).
13354 Also refer to the documentation of the @code{AVPictureType} enum and of
13355 the @code{av_get_picture_type_char} function defined in
13356 @file{libavutil/avutil.h}.
13359 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
13361 @item plane_checksum
13362 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
13363 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
13366 @section showpalette
13368 Displays the 256 colors palette of each frame. This filter is only relevant for
13369 @var{pal8} pixel format frames.
13371 It accepts the following option:
13375 Set the size of the box used to represent one palette color entry. Default is
13376 @code{30} (for a @code{30x30} pixel box).
13379 @section shuffleframes
13381 Reorder and/or duplicate and/or drop video frames.
13383 It accepts the following parameters:
13387 Set the destination indexes of input frames.
13388 This is space or '|' separated list of indexes that maps input frames to output
13389 frames. Number of indexes also sets maximal value that each index may have.
13390 '-1' index have special meaning and that is to drop frame.
13393 The first frame has the index 0. The default is to keep the input unchanged.
13395 @subsection Examples
13399 Swap second and third frame of every three frames of the input:
13401 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
13405 Swap 10th and 1st frame of every ten frames of the input:
13407 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
13411 @section shuffleplanes
13413 Reorder and/or duplicate video planes.
13415 It accepts the following parameters:
13420 The index of the input plane to be used as the first output plane.
13423 The index of the input plane to be used as the second output plane.
13426 The index of the input plane to be used as the third output plane.
13429 The index of the input plane to be used as the fourth output plane.
13433 The first plane has the index 0. The default is to keep the input unchanged.
13435 @subsection Examples
13439 Swap the second and third planes of the input:
13441 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
13445 @anchor{signalstats}
13446 @section signalstats
13447 Evaluate various visual metrics that assist in determining issues associated
13448 with the digitization of analog video media.
13450 By default the filter will log these metadata values:
13454 Display the minimal Y value contained within the input frame. Expressed in
13458 Display the Y value at the 10% percentile within the input frame. Expressed in
13462 Display the average Y value within the input frame. Expressed in range of
13466 Display the Y value at the 90% percentile within the input frame. Expressed in
13470 Display the maximum Y value contained within the input frame. Expressed in
13474 Display the minimal U value contained within the input frame. Expressed in
13478 Display the U value at the 10% percentile within the input frame. Expressed in
13482 Display the average U value within the input frame. Expressed in range of
13486 Display the U value at the 90% percentile within the input frame. Expressed in
13490 Display the maximum U value contained within the input frame. Expressed in
13494 Display the minimal V value contained within the input frame. Expressed in
13498 Display the V value at the 10% percentile within the input frame. Expressed in
13502 Display the average V value within the input frame. Expressed in range of
13506 Display the V value at the 90% percentile within the input frame. Expressed in
13510 Display the maximum V value contained within the input frame. Expressed in
13514 Display the minimal saturation value contained within the input frame.
13515 Expressed in range of [0-~181.02].
13518 Display the saturation value at the 10% percentile within the input frame.
13519 Expressed in range of [0-~181.02].
13522 Display the average saturation value within the input frame. Expressed in range
13526 Display the saturation value at the 90% percentile within the input frame.
13527 Expressed in range of [0-~181.02].
13530 Display the maximum saturation value contained within the input frame.
13531 Expressed in range of [0-~181.02].
13534 Display the median value for hue within the input frame. Expressed in range of
13538 Display the average value for hue within the input frame. Expressed in range of
13542 Display the average of sample value difference between all values of the Y
13543 plane in the current frame and corresponding values of the previous input frame.
13544 Expressed in range of [0-255].
13547 Display the average of sample value difference between all values of the U
13548 plane in the current frame and corresponding values of the previous input frame.
13549 Expressed in range of [0-255].
13552 Display the average of sample value difference between all values of the V
13553 plane in the current frame and corresponding values of the previous input frame.
13554 Expressed in range of [0-255].
13557 Display bit depth of Y plane in current frame.
13558 Expressed in range of [0-16].
13561 Display bit depth of U plane in current frame.
13562 Expressed in range of [0-16].
13565 Display bit depth of V plane in current frame.
13566 Expressed in range of [0-16].
13569 The filter accepts the following options:
13575 @option{stat} specify an additional form of image analysis.
13576 @option{out} output video with the specified type of pixel highlighted.
13578 Both options accept the following values:
13582 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
13583 unlike the neighboring pixels of the same field. Examples of temporal outliers
13584 include the results of video dropouts, head clogs, or tape tracking issues.
13587 Identify @var{vertical line repetition}. Vertical line repetition includes
13588 similar rows of pixels within a frame. In born-digital video vertical line
13589 repetition is common, but this pattern is uncommon in video digitized from an
13590 analog source. When it occurs in video that results from the digitization of an
13591 analog source it can indicate concealment from a dropout compensator.
13594 Identify pixels that fall outside of legal broadcast range.
13598 Set the highlight color for the @option{out} option. The default color is
13602 @subsection Examples
13606 Output data of various video metrics:
13608 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
13612 Output specific data about the minimum and maximum values of the Y plane per frame:
13614 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
13618 Playback video while highlighting pixels that are outside of broadcast range in red.
13620 ffplay example.mov -vf signalstats="out=brng:color=red"
13624 Playback video with signalstats metadata drawn over the frame.
13626 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
13629 The contents of signalstat_drawtext.txt used in the command are:
13632 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
13633 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
13634 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
13635 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
13643 Calculates the MPEG-7 Video Signature. The filter can handle more than one
13644 input. In this case the matching between the inputs can be calculated additionally.
13645 The filter always passes through the first input. The signature of each stream can
13646 be written into a file.
13648 It accepts the following options:
13652 Enable or disable the matching process.
13654 Available values are:
13658 Disable the calculation of a matching (default).
13660 Calculate the matching for the whole video and output whether the whole video
13661 matches or only parts.
13663 Calculate only until a matching is found or the video ends. Should be faster in
13668 Set the number of inputs. The option value must be a non negative integer.
13669 Default value is 1.
13672 Set the path to which the output is written. If there is more than one input,
13673 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
13674 integer), that will be replaced with the input number. If no filename is
13675 specified, no output will be written. This is the default.
13678 Choose the output format.
13680 Available values are:
13684 Use the specified binary representation (default).
13686 Use the specified xml representation.
13690 Set threshold to detect one word as similar. The option value must be an integer
13691 greater than zero. The default value is 9000.
13694 Set threshold to detect all words as similar. The option value must be an integer
13695 greater than zero. The default value is 60000.
13698 Set threshold to detect frames as similar. The option value must be an integer
13699 greater than zero. The default value is 116.
13702 Set the minimum length of a sequence in frames to recognize it as matching
13703 sequence. The option value must be a non negative integer value.
13704 The default value is 0.
13707 Set the minimum relation, that matching frames to all frames must have.
13708 The option value must be a double value between 0 and 1. The default value is 0.5.
13711 @subsection Examples
13715 To calculate the signature of an input video and store it in signature.bin:
13717 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
13721 To detect whether two videos match and store the signatures in XML format in
13722 signature0.xml and signature1.xml:
13724 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 -
13732 Blur the input video without impacting the outlines.
13734 It accepts the following options:
13737 @item luma_radius, lr
13738 Set the luma radius. The option value must be a float number in
13739 the range [0.1,5.0] that specifies the variance of the gaussian filter
13740 used to blur the image (slower if larger). Default value is 1.0.
13742 @item luma_strength, ls
13743 Set the luma strength. The option value must be a float number
13744 in the range [-1.0,1.0] that configures the blurring. A value included
13745 in [0.0,1.0] will blur the image whereas a value included in
13746 [-1.0,0.0] will sharpen the image. Default value is 1.0.
13748 @item luma_threshold, lt
13749 Set the luma threshold used as a coefficient to determine
13750 whether a pixel should be blurred or not. The option value must be an
13751 integer in the range [-30,30]. A value of 0 will filter all the image,
13752 a value included in [0,30] will filter flat areas and a value included
13753 in [-30,0] will filter edges. Default value is 0.
13755 @item chroma_radius, cr
13756 Set the chroma radius. The option value must be a float number in
13757 the range [0.1,5.0] that specifies the variance of the gaussian filter
13758 used to blur the image (slower if larger). Default value is @option{luma_radius}.
13760 @item chroma_strength, cs
13761 Set the chroma strength. The option value must be a float number
13762 in the range [-1.0,1.0] that configures the blurring. A value included
13763 in [0.0,1.0] will blur the image whereas a value included in
13764 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
13766 @item chroma_threshold, ct
13767 Set the chroma threshold used as a coefficient to determine
13768 whether a pixel should be blurred or not. The option value must be an
13769 integer in the range [-30,30]. A value of 0 will filter all the image,
13770 a value included in [0,30] will filter flat areas and a value included
13771 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
13774 If a chroma option is not explicitly set, the corresponding luma value
13779 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
13781 This filter takes in input two input videos, the first input is
13782 considered the "main" source and is passed unchanged to the
13783 output. The second input is used as a "reference" video for computing
13786 Both video inputs must have the same resolution and pixel format for
13787 this filter to work correctly. Also it assumes that both inputs
13788 have the same number of frames, which are compared one by one.
13790 The filter stores the calculated SSIM of each frame.
13792 The description of the accepted parameters follows.
13795 @item stats_file, f
13796 If specified the filter will use the named file to save the SSIM of
13797 each individual frame. When filename equals "-" the data is sent to
13801 The file printed if @var{stats_file} is selected, contains a sequence of
13802 key/value pairs of the form @var{key}:@var{value} for each compared
13805 A description of each shown parameter follows:
13809 sequential number of the input frame, starting from 1
13811 @item Y, U, V, R, G, B
13812 SSIM of the compared frames for the component specified by the suffix.
13815 SSIM of the compared frames for the whole frame.
13818 Same as above but in dB representation.
13821 This filter also supports the @ref{framesync} options.
13825 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13826 [main][ref] ssim="stats_file=stats.log" [out]
13829 On this example the input file being processed is compared with the
13830 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
13831 is stored in @file{stats.log}.
13833 Another example with both psnr and ssim at same time:
13835 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
13840 Convert between different stereoscopic image formats.
13842 The filters accept the following options:
13846 Set stereoscopic image format of input.
13848 Available values for input image formats are:
13851 side by side parallel (left eye left, right eye right)
13854 side by side crosseye (right eye left, left eye right)
13857 side by side parallel with half width resolution
13858 (left eye left, right eye right)
13861 side by side crosseye with half width resolution
13862 (right eye left, left eye right)
13865 above-below (left eye above, right eye below)
13868 above-below (right eye above, left eye below)
13871 above-below with half height resolution
13872 (left eye above, right eye below)
13875 above-below with half height resolution
13876 (right eye above, left eye below)
13879 alternating frames (left eye first, right eye second)
13882 alternating frames (right eye first, left eye second)
13885 interleaved rows (left eye has top row, right eye starts on next row)
13888 interleaved rows (right eye has top row, left eye starts on next row)
13891 interleaved columns, left eye first
13894 interleaved columns, right eye first
13896 Default value is @samp{sbsl}.
13900 Set stereoscopic image format of output.
13904 side by side parallel (left eye left, right eye right)
13907 side by side crosseye (right eye left, left eye right)
13910 side by side parallel with half width resolution
13911 (left eye left, right eye right)
13914 side by side crosseye with half width resolution
13915 (right eye left, left eye right)
13918 above-below (left eye above, right eye below)
13921 above-below (right eye above, left eye below)
13924 above-below with half height resolution
13925 (left eye above, right eye below)
13928 above-below with half height resolution
13929 (right eye above, left eye below)
13932 alternating frames (left eye first, right eye second)
13935 alternating frames (right eye first, left eye second)
13938 interleaved rows (left eye has top row, right eye starts on next row)
13941 interleaved rows (right eye has top row, left eye starts on next row)
13944 anaglyph red/blue gray
13945 (red filter on left eye, blue filter on right eye)
13948 anaglyph red/green gray
13949 (red filter on left eye, green filter on right eye)
13952 anaglyph red/cyan gray
13953 (red filter on left eye, cyan filter on right eye)
13956 anaglyph red/cyan half colored
13957 (red filter on left eye, cyan filter on right eye)
13960 anaglyph red/cyan color
13961 (red filter on left eye, cyan filter on right eye)
13964 anaglyph red/cyan color optimized with the least squares projection of dubois
13965 (red filter on left eye, cyan filter on right eye)
13968 anaglyph green/magenta gray
13969 (green filter on left eye, magenta filter on right eye)
13972 anaglyph green/magenta half colored
13973 (green filter on left eye, magenta filter on right eye)
13976 anaglyph green/magenta colored
13977 (green filter on left eye, magenta filter on right eye)
13980 anaglyph green/magenta color optimized with the least squares projection of dubois
13981 (green filter on left eye, magenta filter on right eye)
13984 anaglyph yellow/blue gray
13985 (yellow filter on left eye, blue filter on right eye)
13988 anaglyph yellow/blue half colored
13989 (yellow filter on left eye, blue filter on right eye)
13992 anaglyph yellow/blue colored
13993 (yellow filter on left eye, blue filter on right eye)
13996 anaglyph yellow/blue color optimized with the least squares projection of dubois
13997 (yellow filter on left eye, blue filter on right eye)
14000 mono output (left eye only)
14003 mono output (right eye only)
14006 checkerboard, left eye first
14009 checkerboard, right eye first
14012 interleaved columns, left eye first
14015 interleaved columns, right eye first
14021 Default value is @samp{arcd}.
14024 @subsection Examples
14028 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14034 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14040 @section streamselect, astreamselect
14041 Select video or audio streams.
14043 The filter accepts the following options:
14047 Set number of inputs. Default is 2.
14050 Set input indexes to remap to outputs.
14053 @subsection Commands
14055 The @code{streamselect} and @code{astreamselect} filter supports the following
14060 Set input indexes to remap to outputs.
14063 @subsection Examples
14067 Select first 5 seconds 1st stream and rest of time 2nd stream:
14069 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14073 Same as above, but for audio:
14075 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14080 Apply sobel operator to input video stream.
14082 The filter accepts the following option:
14086 Set which planes will be processed, unprocessed planes will be copied.
14087 By default value 0xf, all planes will be processed.
14090 Set value which will be multiplied with filtered result.
14093 Set value which will be added to filtered result.
14099 Apply a simple postprocessing filter that compresses and decompresses the image
14100 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14101 and average the results.
14103 The filter accepts the following options:
14107 Set quality. This option defines the number of levels for averaging. It accepts
14108 an integer in the range 0-6. If set to @code{0}, the filter will have no
14109 effect. A value of @code{6} means the higher quality. For each increment of
14110 that value the speed drops by a factor of approximately 2. Default value is
14114 Force a constant quantization parameter. If not set, the filter will use the QP
14115 from the video stream (if available).
14118 Set thresholding mode. Available modes are:
14122 Set hard thresholding (default).
14124 Set soft thresholding (better de-ringing effect, but likely blurrier).
14127 @item use_bframe_qp
14128 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14129 option may cause flicker since the B-Frames have often larger QP. Default is
14130 @code{0} (not enabled).
14136 Draw subtitles on top of input video using the libass library.
14138 To enable compilation of this filter you need to configure FFmpeg with
14139 @code{--enable-libass}. This filter also requires a build with libavcodec and
14140 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14141 Alpha) subtitles format.
14143 The filter accepts the following options:
14147 Set the filename of the subtitle file to read. It must be specified.
14149 @item original_size
14150 Specify the size of the original video, the video for which the ASS file
14151 was composed. For the syntax of this option, check the
14152 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14153 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14154 correctly scale the fonts if the aspect ratio has been changed.
14157 Set a directory path containing fonts that can be used by the filter.
14158 These fonts will be used in addition to whatever the font provider uses.
14161 Process alpha channel, by default alpha channel is untouched.
14164 Set subtitles input character encoding. @code{subtitles} filter only. Only
14165 useful if not UTF-8.
14167 @item stream_index, si
14168 Set subtitles stream index. @code{subtitles} filter only.
14171 Override default style or script info parameters of the subtitles. It accepts a
14172 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14175 If the first key is not specified, it is assumed that the first value
14176 specifies the @option{filename}.
14178 For example, to render the file @file{sub.srt} on top of the input
14179 video, use the command:
14184 which is equivalent to:
14186 subtitles=filename=sub.srt
14189 To render the default subtitles stream from file @file{video.mkv}, use:
14191 subtitles=video.mkv
14194 To render the second subtitles stream from that file, use:
14196 subtitles=video.mkv:si=1
14199 To make the subtitles stream from @file{sub.srt} appear in transparent green
14200 @code{DejaVu Serif}, use:
14202 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14205 @section super2xsai
14207 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14208 Interpolate) pixel art scaling algorithm.
14210 Useful for enlarging pixel art images without reducing sharpness.
14214 Swap two rectangular objects in video.
14216 This filter accepts the following options:
14226 Set 1st rect x coordinate.
14229 Set 1st rect y coordinate.
14232 Set 2nd rect x coordinate.
14235 Set 2nd rect y coordinate.
14237 All expressions are evaluated once for each frame.
14240 The all options are expressions containing the following constants:
14245 The input width and height.
14248 same as @var{w} / @var{h}
14251 input sample aspect ratio
14254 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
14257 The number of the input frame, starting from 0.
14260 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
14263 the position in the file of the input frame, NAN if unknown
14271 Apply telecine process to the video.
14273 This filter accepts the following options:
14282 The default value is @code{top}.
14286 A string of numbers representing the pulldown pattern you wish to apply.
14287 The default value is @code{23}.
14291 Some typical patterns:
14296 24p: 2332 (preferred)
14303 24p: 222222222223 ("Euro pulldown")
14310 Apply threshold effect to video stream.
14312 This filter needs four video streams to perform thresholding.
14313 First stream is stream we are filtering.
14314 Second stream is holding threshold values, third stream is holding min values,
14315 and last, fourth stream is holding max values.
14317 The filter accepts the following option:
14321 Set which planes will be processed, unprocessed planes will be copied.
14322 By default value 0xf, all planes will be processed.
14325 For example if first stream pixel's component value is less then threshold value
14326 of pixel component from 2nd threshold stream, third stream value will picked,
14327 otherwise fourth stream pixel component value will be picked.
14329 Using color source filter one can perform various types of thresholding:
14331 @subsection Examples
14335 Binary threshold, using gray color as threshold:
14337 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
14341 Inverted binary threshold, using gray color as threshold:
14343 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
14347 Truncate binary threshold, using gray color as threshold:
14349 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
14353 Threshold to zero, using gray color as threshold:
14355 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
14359 Inverted threshold to zero, using gray color as threshold:
14361 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
14366 Select the most representative frame in a given sequence of consecutive frames.
14368 The filter accepts the following options:
14372 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
14373 will pick one of them, and then handle the next batch of @var{n} frames until
14374 the end. Default is @code{100}.
14377 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
14378 value will result in a higher memory usage, so a high value is not recommended.
14380 @subsection Examples
14384 Extract one picture each 50 frames:
14390 Complete example of a thumbnail creation with @command{ffmpeg}:
14392 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
14398 Tile several successive frames together.
14400 The filter accepts the following options:
14405 Set the grid size (i.e. the number of lines and columns). For the syntax of
14406 this option, check the
14407 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14410 Set the maximum number of frames to render in the given area. It must be less
14411 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
14412 the area will be used.
14415 Set the outer border margin in pixels.
14418 Set the inner border thickness (i.e. the number of pixels between frames). For
14419 more advanced padding options (such as having different values for the edges),
14420 refer to the pad video filter.
14423 Specify the color of the unused area. For the syntax of this option, check the
14424 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
14428 @subsection Examples
14432 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
14434 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
14436 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
14437 duplicating each output frame to accommodate the originally detected frame
14441 Display @code{5} pictures in an area of @code{3x2} frames,
14442 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
14443 mixed flat and named options:
14445 tile=3x2:nb_frames=5:padding=7:margin=2
14449 @section tinterlace
14451 Perform various types of temporal field interlacing.
14453 Frames are counted starting from 1, so the first input frame is
14456 The filter accepts the following options:
14461 Specify the mode of the interlacing. This option can also be specified
14462 as a value alone. See below for a list of values for this option.
14464 Available values are:
14468 Move odd frames into the upper field, even into the lower field,
14469 generating a double height frame at half frame rate.
14473 Frame 1 Frame 2 Frame 3 Frame 4
14475 11111 22222 33333 44444
14476 11111 22222 33333 44444
14477 11111 22222 33333 44444
14478 11111 22222 33333 44444
14492 Only output odd frames, even frames are dropped, generating a frame with
14493 unchanged height at half frame rate.
14498 Frame 1 Frame 2 Frame 3 Frame 4
14500 11111 22222 33333 44444
14501 11111 22222 33333 44444
14502 11111 22222 33333 44444
14503 11111 22222 33333 44444
14513 Only output even frames, odd frames are dropped, generating a frame with
14514 unchanged height at half frame rate.
14519 Frame 1 Frame 2 Frame 3 Frame 4
14521 11111 22222 33333 44444
14522 11111 22222 33333 44444
14523 11111 22222 33333 44444
14524 11111 22222 33333 44444
14534 Expand each frame to full height, but pad alternate lines with black,
14535 generating a frame with double height at the same input frame rate.
14540 Frame 1 Frame 2 Frame 3 Frame 4
14542 11111 22222 33333 44444
14543 11111 22222 33333 44444
14544 11111 22222 33333 44444
14545 11111 22222 33333 44444
14548 11111 ..... 33333 .....
14549 ..... 22222 ..... 44444
14550 11111 ..... 33333 .....
14551 ..... 22222 ..... 44444
14552 11111 ..... 33333 .....
14553 ..... 22222 ..... 44444
14554 11111 ..... 33333 .....
14555 ..... 22222 ..... 44444
14559 @item interleave_top, 4
14560 Interleave the upper field from odd frames with the lower field from
14561 even frames, generating a frame with unchanged height at half frame rate.
14566 Frame 1 Frame 2 Frame 3 Frame 4
14568 11111<- 22222 33333<- 44444
14569 11111 22222<- 33333 44444<-
14570 11111<- 22222 33333<- 44444
14571 11111 22222<- 33333 44444<-
14581 @item interleave_bottom, 5
14582 Interleave the lower field from odd frames with the upper field from
14583 even frames, generating a frame with unchanged height at half frame rate.
14588 Frame 1 Frame 2 Frame 3 Frame 4
14590 11111 22222<- 33333 44444<-
14591 11111<- 22222 33333<- 44444
14592 11111 22222<- 33333 44444<-
14593 11111<- 22222 33333<- 44444
14603 @item interlacex2, 6
14604 Double frame rate with unchanged height. Frames are inserted each
14605 containing the second temporal field from the previous input frame and
14606 the first temporal field from the next input frame. This mode relies on
14607 the top_field_first flag. Useful for interlaced video displays with no
14608 field synchronisation.
14613 Frame 1 Frame 2 Frame 3 Frame 4
14615 11111 22222 33333 44444
14616 11111 22222 33333 44444
14617 11111 22222 33333 44444
14618 11111 22222 33333 44444
14621 11111 22222 22222 33333 33333 44444 44444
14622 11111 11111 22222 22222 33333 33333 44444
14623 11111 22222 22222 33333 33333 44444 44444
14624 11111 11111 22222 22222 33333 33333 44444
14629 Move odd frames into the upper field, even into the lower field,
14630 generating a double height frame at same frame rate.
14635 Frame 1 Frame 2 Frame 3 Frame 4
14637 11111 22222 33333 44444
14638 11111 22222 33333 44444
14639 11111 22222 33333 44444
14640 11111 22222 33333 44444
14643 11111 33333 33333 55555
14644 22222 22222 44444 44444
14645 11111 33333 33333 55555
14646 22222 22222 44444 44444
14647 11111 33333 33333 55555
14648 22222 22222 44444 44444
14649 11111 33333 33333 55555
14650 22222 22222 44444 44444
14655 Numeric values are deprecated but are accepted for backward
14656 compatibility reasons.
14658 Default mode is @code{merge}.
14661 Specify flags influencing the filter process.
14663 Available value for @var{flags} is:
14666 @item low_pass_filter, vlfp
14667 Enable linear vertical low-pass filtering in the filter.
14668 Vertical low-pass filtering is required when creating an interlaced
14669 destination from a progressive source which contains high-frequency
14670 vertical detail. Filtering will reduce interlace 'twitter' and Moire
14673 @item complex_filter, cvlfp
14674 Enable complex vertical low-pass filtering.
14675 This will slightly less reduce interlace 'twitter' and Moire
14676 patterning but better retain detail and subjective sharpness impression.
14680 Vertical low-pass filtering can only be enabled for @option{mode}
14681 @var{interleave_top} and @var{interleave_bottom}.
14686 Tone map colors from different dynamic ranges.
14688 This filter expects data in single precision floating point, as it needs to
14689 operate on (and can output) out-of-range values. Another filter, such as
14690 @ref{zscale}, is needed to convert the resulting frame to a usable format.
14692 The tonemapping algorithms implemented only work on linear light, so input
14693 data should be linearized beforehand (and possibly correctly tagged).
14696 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
14699 @subsection Options
14700 The filter accepts the following options.
14704 Set the tone map algorithm to use.
14706 Possible values are:
14709 Do not apply any tone map, only desaturate overbright pixels.
14712 Hard-clip any out-of-range values. Use it for perfect color accuracy for
14713 in-range values, while distorting out-of-range values.
14716 Stretch the entire reference gamut to a linear multiple of the display.
14719 Fit a logarithmic transfer between the tone curves.
14722 Preserve overall image brightness with a simple curve, using nonlinear
14723 contrast, which results in flattening details and degrading color accuracy.
14726 Preserve both dark and bright details better than @var{reinhard}, at the cost
14727 of slightly darkening everything. Use it when detail preservation is more
14728 important than color and brightness accuracy.
14731 Smoothly map out-of-range values, while retaining contrast and colors for
14732 in-range material as much as possible. Use it when color accuracy is more
14733 important than detail preservation.
14739 Tune the tone mapping algorithm.
14741 This affects the following algorithms:
14747 Specifies the scale factor to use while stretching.
14751 Specifies the exponent of the function.
14755 Specify an extra linear coefficient to multiply into the signal before clipping.
14759 Specify the local contrast coefficient at the display peak.
14760 Default to 0.5, which means that in-gamut values will be about half as bright
14767 Specify the transition point from linear to mobius transform. Every value
14768 below this point is guaranteed to be mapped 1:1. The higher the value, the
14769 more accurate the result will be, at the cost of losing bright details.
14770 Default to 0.3, which due to the steep initial slope still preserves in-range
14771 colors fairly accurately.
14775 Apply desaturation for highlights that exceed this level of brightness. The
14776 higher the parameter, the more color information will be preserved. This
14777 setting helps prevent unnaturally blown-out colors for super-highlights, by
14778 (smoothly) turning into white instead. This makes images feel more natural,
14779 at the cost of reducing information about out-of-range colors.
14781 The default of 2.0 is somewhat conservative and will mostly just apply to
14782 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
14784 This option works only if the input frame has a supported color tag.
14787 Override signal/nominal/reference peak with this value. Useful when the
14788 embedded peak information in display metadata is not reliable or when tone
14789 mapping from a lower range to a higher range.
14794 Transpose rows with columns in the input video and optionally flip it.
14796 It accepts the following parameters:
14801 Specify the transposition direction.
14803 Can assume the following values:
14805 @item 0, 4, cclock_flip
14806 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
14814 Rotate by 90 degrees clockwise, that is:
14822 Rotate by 90 degrees counterclockwise, that is:
14829 @item 3, 7, clock_flip
14830 Rotate by 90 degrees clockwise and vertically flip, that is:
14838 For values between 4-7, the transposition is only done if the input
14839 video geometry is portrait and not landscape. These values are
14840 deprecated, the @code{passthrough} option should be used instead.
14842 Numerical values are deprecated, and should be dropped in favor of
14843 symbolic constants.
14846 Do not apply the transposition if the input geometry matches the one
14847 specified by the specified value. It accepts the following values:
14850 Always apply transposition.
14852 Preserve portrait geometry (when @var{height} >= @var{width}).
14854 Preserve landscape geometry (when @var{width} >= @var{height}).
14857 Default value is @code{none}.
14860 For example to rotate by 90 degrees clockwise and preserve portrait
14863 transpose=dir=1:passthrough=portrait
14866 The command above can also be specified as:
14868 transpose=1:portrait
14872 Trim the input so that the output contains one continuous subpart of the input.
14874 It accepts the following parameters:
14877 Specify the time of the start of the kept section, i.e. the frame with the
14878 timestamp @var{start} will be the first frame in the output.
14881 Specify the time of the first frame that will be dropped, i.e. the frame
14882 immediately preceding the one with the timestamp @var{end} will be the last
14883 frame in the output.
14886 This is the same as @var{start}, except this option sets the start timestamp
14887 in timebase units instead of seconds.
14890 This is the same as @var{end}, except this option sets the end timestamp
14891 in timebase units instead of seconds.
14894 The maximum duration of the output in seconds.
14897 The number of the first frame that should be passed to the output.
14900 The number of the first frame that should be dropped.
14903 @option{start}, @option{end}, and @option{duration} are expressed as time
14904 duration specifications; see
14905 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14906 for the accepted syntax.
14908 Note that the first two sets of the start/end options and the @option{duration}
14909 option look at the frame timestamp, while the _frame variants simply count the
14910 frames that pass through the filter. Also note that this filter does not modify
14911 the timestamps. If you wish for the output timestamps to start at zero, insert a
14912 setpts filter after the trim filter.
14914 If multiple start or end options are set, this filter tries to be greedy and
14915 keep all the frames that match at least one of the specified constraints. To keep
14916 only the part that matches all the constraints at once, chain multiple trim
14919 The defaults are such that all the input is kept. So it is possible to set e.g.
14920 just the end values to keep everything before the specified time.
14925 Drop everything except the second minute of input:
14927 ffmpeg -i INPUT -vf trim=60:120
14931 Keep only the first second:
14933 ffmpeg -i INPUT -vf trim=duration=1
14938 @section unpremultiply
14939 Apply alpha unpremultiply effect to input video stream using first plane
14940 of second stream as alpha.
14942 Both streams must have same dimensions and same pixel format.
14944 The filter accepts the following option:
14948 Set which planes will be processed, unprocessed planes will be copied.
14949 By default value 0xf, all planes will be processed.
14951 If the format has 1 or 2 components, then luma is bit 0.
14952 If the format has 3 or 4 components:
14953 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
14954 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
14955 If present, the alpha channel is always the last bit.
14958 Do not require 2nd input for processing, instead use alpha plane from input stream.
14964 Sharpen or blur the input video.
14966 It accepts the following parameters:
14969 @item luma_msize_x, lx
14970 Set the luma matrix horizontal size. It must be an odd integer between
14971 3 and 23. The default value is 5.
14973 @item luma_msize_y, ly
14974 Set the luma matrix vertical size. It must be an odd integer between 3
14975 and 23. The default value is 5.
14977 @item luma_amount, la
14978 Set the luma effect strength. It must be a floating point number, reasonable
14979 values lay between -1.5 and 1.5.
14981 Negative values will blur the input video, while positive values will
14982 sharpen it, a value of zero will disable the effect.
14984 Default value is 1.0.
14986 @item chroma_msize_x, cx
14987 Set the chroma matrix horizontal size. It must be an odd integer
14988 between 3 and 23. The default value is 5.
14990 @item chroma_msize_y, cy
14991 Set the chroma matrix vertical size. It must be an odd integer
14992 between 3 and 23. The default value is 5.
14994 @item chroma_amount, ca
14995 Set the chroma effect strength. It must be a floating point number, reasonable
14996 values lay between -1.5 and 1.5.
14998 Negative values will blur the input video, while positive values will
14999 sharpen it, a value of zero will disable the effect.
15001 Default value is 0.0.
15004 If set to 1, specify using OpenCL capabilities, only available if
15005 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
15009 All parameters are optional and default to the equivalent of the
15010 string '5:5:1.0:5:5:0.0'.
15012 @subsection Examples
15016 Apply strong luma sharpen effect:
15018 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15022 Apply a strong blur of both luma and chroma parameters:
15024 unsharp=7:7:-2:7:7:-2
15030 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15031 the image at several (or - in the case of @option{quality} level @code{8} - all)
15032 shifts and average the results.
15034 The way this differs from the behavior of spp is that uspp actually encodes &
15035 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15036 DCT similar to MJPEG.
15038 The filter accepts the following options:
15042 Set quality. This option defines the number of levels for averaging. It accepts
15043 an integer in the range 0-8. If set to @code{0}, the filter will have no
15044 effect. A value of @code{8} means the higher quality. For each increment of
15045 that value the speed drops by a factor of approximately 2. Default value is
15049 Force a constant quantization parameter. If not set, the filter will use the QP
15050 from the video stream (if available).
15053 @section vaguedenoiser
15055 Apply a wavelet based denoiser.
15057 It transforms each frame from the video input into the wavelet domain,
15058 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15059 the obtained coefficients. It does an inverse wavelet transform after.
15060 Due to wavelet properties, it should give a nice smoothed result, and
15061 reduced noise, without blurring picture features.
15063 This filter accepts the following options:
15067 The filtering strength. The higher, the more filtered the video will be.
15068 Hard thresholding can use a higher threshold than soft thresholding
15069 before the video looks overfiltered. Default value is 2.
15072 The filtering method the filter will use.
15074 It accepts the following values:
15077 All values under the threshold will be zeroed.
15080 All values under the threshold will be zeroed. All values above will be
15081 reduced by the threshold.
15084 Scales or nullifies coefficients - intermediary between (more) soft and
15085 (less) hard thresholding.
15088 Default is garrote.
15091 Number of times, the wavelet will decompose the picture. Picture can't
15092 be decomposed beyond a particular point (typically, 8 for a 640x480
15093 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15096 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15099 A list of the planes to process. By default all planes are processed.
15102 @section vectorscope
15104 Display 2 color component values in the two dimensional graph (which is called
15107 This filter accepts the following options:
15111 Set vectorscope mode.
15113 It accepts the following values:
15116 Gray values are displayed on graph, higher brightness means more pixels have
15117 same component color value on location in graph. This is the default mode.
15120 Gray values are displayed on graph. Surrounding pixels values which are not
15121 present in video frame are drawn in gradient of 2 color components which are
15122 set by option @code{x} and @code{y}. The 3rd color component is static.
15125 Actual color components values present in video frame are displayed on graph.
15128 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15129 on graph increases value of another color component, which is luminance by
15130 default values of @code{x} and @code{y}.
15133 Actual colors present in video frame are displayed on graph. If two different
15134 colors map to same position on graph then color with higher value of component
15135 not present in graph is picked.
15138 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15139 component picked from radial gradient.
15143 Set which color component will be represented on X-axis. Default is @code{1}.
15146 Set which color component will be represented on Y-axis. Default is @code{2}.
15149 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15150 of color component which represents frequency of (X, Y) location in graph.
15155 No envelope, this is default.
15158 Instant envelope, even darkest single pixel will be clearly highlighted.
15161 Hold maximum and minimum values presented in graph over time. This way you
15162 can still spot out of range values without constantly looking at vectorscope.
15165 Peak and instant envelope combined together.
15169 Set what kind of graticule to draw.
15177 Set graticule opacity.
15180 Set graticule flags.
15184 Draw graticule for white point.
15187 Draw graticule for black point.
15190 Draw color points short names.
15194 Set background opacity.
15196 @item lthreshold, l
15197 Set low threshold for color component not represented on X or Y axis.
15198 Values lower than this value will be ignored. Default is 0.
15199 Note this value is multiplied with actual max possible value one pixel component
15200 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15203 @item hthreshold, h
15204 Set high threshold for color component not represented on X or Y axis.
15205 Values higher than this value will be ignored. Default is 1.
15206 Note this value is multiplied with actual max possible value one pixel component
15207 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15208 is 0.9 * 255 = 230.
15210 @item colorspace, c
15211 Set what kind of colorspace to use when drawing graticule.
15220 @anchor{vidstabdetect}
15221 @section vidstabdetect
15223 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
15224 @ref{vidstabtransform} for pass 2.
15226 This filter generates a file with relative translation and rotation
15227 transform information about subsequent frames, which is then used by
15228 the @ref{vidstabtransform} filter.
15230 To enable compilation of this filter you need to configure FFmpeg with
15231 @code{--enable-libvidstab}.
15233 This filter accepts the following options:
15237 Set the path to the file used to write the transforms information.
15238 Default value is @file{transforms.trf}.
15241 Set how shaky the video is and how quick the camera is. It accepts an
15242 integer in the range 1-10, a value of 1 means little shakiness, a
15243 value of 10 means strong shakiness. Default value is 5.
15246 Set the accuracy of the detection process. It must be a value in the
15247 range 1-15. A value of 1 means low accuracy, a value of 15 means high
15248 accuracy. Default value is 15.
15251 Set stepsize of the search process. The region around minimum is
15252 scanned with 1 pixel resolution. Default value is 6.
15255 Set minimum contrast. Below this value a local measurement field is
15256 discarded. Must be a floating point value in the range 0-1. Default
15260 Set reference frame number for tripod mode.
15262 If enabled, the motion of the frames is compared to a reference frame
15263 in the filtered stream, identified by the specified number. The idea
15264 is to compensate all movements in a more-or-less static scene and keep
15265 the camera view absolutely still.
15267 If set to 0, it is disabled. The frames are counted starting from 1.
15270 Show fields and transforms in the resulting frames. It accepts an
15271 integer in the range 0-2. Default value is 0, which disables any
15275 @subsection Examples
15279 Use default values:
15285 Analyze strongly shaky movie and put the results in file
15286 @file{mytransforms.trf}:
15288 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
15292 Visualize the result of internal transformations in the resulting
15295 vidstabdetect=show=1
15299 Analyze a video with medium shakiness using @command{ffmpeg}:
15301 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
15305 @anchor{vidstabtransform}
15306 @section vidstabtransform
15308 Video stabilization/deshaking: pass 2 of 2,
15309 see @ref{vidstabdetect} for pass 1.
15311 Read a file with transform information for each frame and
15312 apply/compensate them. Together with the @ref{vidstabdetect}
15313 filter this can be used to deshake videos. See also
15314 @url{http://public.hronopik.de/vid.stab}. It is important to also use
15315 the @ref{unsharp} filter, see below.
15317 To enable compilation of this filter you need to configure FFmpeg with
15318 @code{--enable-libvidstab}.
15320 @subsection Options
15324 Set path to the file used to read the transforms. Default value is
15325 @file{transforms.trf}.
15328 Set the number of frames (value*2 + 1) used for lowpass filtering the
15329 camera movements. Default value is 10.
15331 For example a number of 10 means that 21 frames are used (10 in the
15332 past and 10 in the future) to smoothen the motion in the video. A
15333 larger value leads to a smoother video, but limits the acceleration of
15334 the camera (pan/tilt movements). 0 is a special case where a static
15335 camera is simulated.
15338 Set the camera path optimization algorithm.
15340 Accepted values are:
15343 gaussian kernel low-pass filter on camera motion (default)
15345 averaging on transformations
15349 Set maximal number of pixels to translate frames. Default value is -1,
15353 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
15354 value is -1, meaning no limit.
15357 Specify how to deal with borders that may be visible due to movement
15360 Available values are:
15363 keep image information from previous frame (default)
15365 fill the border black
15369 Invert transforms if set to 1. Default value is 0.
15372 Consider transforms as relative to previous frame if set to 1,
15373 absolute if set to 0. Default value is 0.
15376 Set percentage to zoom. A positive value will result in a zoom-in
15377 effect, a negative value in a zoom-out effect. Default value is 0 (no
15381 Set optimal zooming to avoid borders.
15383 Accepted values are:
15388 optimal static zoom value is determined (only very strong movements
15389 will lead to visible borders) (default)
15391 optimal adaptive zoom value is determined (no borders will be
15392 visible), see @option{zoomspeed}
15395 Note that the value given at zoom is added to the one calculated here.
15398 Set percent to zoom maximally each frame (enabled when
15399 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
15403 Specify type of interpolation.
15405 Available values are:
15410 linear only horizontal
15412 linear in both directions (default)
15414 cubic in both directions (slow)
15418 Enable virtual tripod mode if set to 1, which is equivalent to
15419 @code{relative=0:smoothing=0}. Default value is 0.
15421 Use also @code{tripod} option of @ref{vidstabdetect}.
15424 Increase log verbosity if set to 1. Also the detected global motions
15425 are written to the temporary file @file{global_motions.trf}. Default
15429 @subsection Examples
15433 Use @command{ffmpeg} for a typical stabilization with default values:
15435 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
15438 Note the use of the @ref{unsharp} filter which is always recommended.
15441 Zoom in a bit more and load transform data from a given file:
15443 vidstabtransform=zoom=5:input="mytransforms.trf"
15447 Smoothen the video even more:
15449 vidstabtransform=smoothing=30
15455 Flip the input video vertically.
15457 For example, to vertically flip a video with @command{ffmpeg}:
15459 ffmpeg -i in.avi -vf "vflip" out.avi
15465 Make or reverse a natural vignetting effect.
15467 The filter accepts the following options:
15471 Set lens angle expression as a number of radians.
15473 The value is clipped in the @code{[0,PI/2]} range.
15475 Default value: @code{"PI/5"}
15479 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
15483 Set forward/backward mode.
15485 Available modes are:
15488 The larger the distance from the central point, the darker the image becomes.
15491 The larger the distance from the central point, the brighter the image becomes.
15492 This can be used to reverse a vignette effect, though there is no automatic
15493 detection to extract the lens @option{angle} and other settings (yet). It can
15494 also be used to create a burning effect.
15497 Default value is @samp{forward}.
15500 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
15502 It accepts the following values:
15505 Evaluate expressions only once during the filter initialization.
15508 Evaluate expressions for each incoming frame. This is way slower than the
15509 @samp{init} mode since it requires all the scalers to be re-computed, but it
15510 allows advanced dynamic expressions.
15513 Default value is @samp{init}.
15516 Set dithering to reduce the circular banding effects. Default is @code{1}
15520 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
15521 Setting this value to the SAR of the input will make a rectangular vignetting
15522 following the dimensions of the video.
15524 Default is @code{1/1}.
15527 @subsection Expressions
15529 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
15530 following parameters.
15535 input width and height
15538 the number of input frame, starting from 0
15541 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
15542 @var{TB} units, NAN if undefined
15545 frame rate of the input video, NAN if the input frame rate is unknown
15548 the PTS (Presentation TimeStamp) of the filtered video frame,
15549 expressed in seconds, NAN if undefined
15552 time base of the input video
15556 @subsection Examples
15560 Apply simple strong vignetting effect:
15566 Make a flickering vignetting:
15568 vignette='PI/4+random(1)*PI/50':eval=frame
15574 Stack input videos vertically.
15576 All streams must be of same pixel format and of same width.
15578 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
15579 to create same output.
15581 The filter accept the following option:
15585 Set number of input streams. Default is 2.
15588 If set to 1, force the output to terminate when the shortest input
15589 terminates. Default value is 0.
15594 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
15595 Deinterlacing Filter").
15597 Based on the process described by Martin Weston for BBC R&D, and
15598 implemented based on the de-interlace algorithm written by Jim
15599 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
15600 uses filter coefficients calculated by BBC R&D.
15602 There are two sets of filter coefficients, so called "simple":
15603 and "complex". Which set of filter coefficients is used can
15604 be set by passing an optional parameter:
15608 Set the interlacing filter coefficients. Accepts one of the following values:
15612 Simple filter coefficient set.
15614 More-complex filter coefficient set.
15616 Default value is @samp{complex}.
15619 Specify which frames to deinterlace. Accept one of the following values:
15623 Deinterlace all frames,
15625 Only deinterlace frames marked as interlaced.
15628 Default value is @samp{all}.
15632 Video waveform monitor.
15634 The waveform monitor plots color component intensity. By default luminance
15635 only. Each column of the waveform corresponds to a column of pixels in the
15638 It accepts the following options:
15642 Can be either @code{row}, or @code{column}. Default is @code{column}.
15643 In row mode, the graph on the left side represents color component value 0 and
15644 the right side represents value = 255. In column mode, the top side represents
15645 color component value = 0 and bottom side represents value = 255.
15648 Set intensity. Smaller values are useful to find out how many values of the same
15649 luminance are distributed across input rows/columns.
15650 Default value is @code{0.04}. Allowed range is [0, 1].
15653 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
15654 In mirrored mode, higher values will be represented on the left
15655 side for @code{row} mode and at the top for @code{column} mode. Default is
15656 @code{1} (mirrored).
15660 It accepts the following values:
15663 Presents information identical to that in the @code{parade}, except
15664 that the graphs representing color components are superimposed directly
15667 This display mode makes it easier to spot relative differences or similarities
15668 in overlapping areas of the color components that are supposed to be identical,
15669 such as neutral whites, grays, or blacks.
15672 Display separate graph for the color components side by side in
15673 @code{row} mode or one below the other in @code{column} mode.
15676 Display separate graph for the color components side by side in
15677 @code{column} mode or one below the other in @code{row} mode.
15679 Using this display mode makes it easy to spot color casts in the highlights
15680 and shadows of an image, by comparing the contours of the top and the bottom
15681 graphs of each waveform. Since whites, grays, and blacks are characterized
15682 by exactly equal amounts of red, green, and blue, neutral areas of the picture
15683 should display three waveforms of roughly equal width/height. If not, the
15684 correction is easy to perform by making level adjustments the three waveforms.
15686 Default is @code{stack}.
15688 @item components, c
15689 Set which color components to display. Default is 1, which means only luminance
15690 or red color component if input is in RGB colorspace. If is set for example to
15691 7 it will display all 3 (if) available color components.
15696 No envelope, this is default.
15699 Instant envelope, minimum and maximum values presented in graph will be easily
15700 visible even with small @code{step} value.
15703 Hold minimum and maximum values presented in graph across time. This way you
15704 can still spot out of range values without constantly looking at waveforms.
15707 Peak and instant envelope combined together.
15713 No filtering, this is default.
15716 Luma and chroma combined together.
15719 Similar as above, but shows difference between blue and red chroma.
15722 Displays only chroma.
15725 Displays actual color value on waveform.
15728 Similar as above, but with luma showing frequency of chroma values.
15732 Set which graticule to display.
15736 Do not display graticule.
15739 Display green graticule showing legal broadcast ranges.
15743 Set graticule opacity.
15746 Set graticule flags.
15750 Draw numbers above lines. By default enabled.
15753 Draw dots instead of lines.
15757 Set scale used for displaying graticule.
15764 Default is digital.
15767 Set background opacity.
15770 @section weave, doubleweave
15772 The @code{weave} takes a field-based video input and join
15773 each two sequential fields into single frame, producing a new double
15774 height clip with half the frame rate and half the frame count.
15776 The @code{doubleweave} works same as @code{weave} but without
15777 halving frame rate and frame count.
15779 It accepts the following option:
15783 Set first field. Available values are:
15787 Set the frame as top-field-first.
15790 Set the frame as bottom-field-first.
15794 @subsection Examples
15798 Interlace video using @ref{select} and @ref{separatefields} filter:
15800 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
15805 Apply the xBR high-quality magnification filter which is designed for pixel
15806 art. It follows a set of edge-detection rules, see
15807 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
15809 It accepts the following option:
15813 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
15814 @code{3xBR} and @code{4} for @code{4xBR}.
15815 Default is @code{3}.
15821 Deinterlace the input video ("yadif" means "yet another deinterlacing
15824 It accepts the following parameters:
15830 The interlacing mode to adopt. It accepts one of the following values:
15833 @item 0, send_frame
15834 Output one frame for each frame.
15835 @item 1, send_field
15836 Output one frame for each field.
15837 @item 2, send_frame_nospatial
15838 Like @code{send_frame}, but it skips the spatial interlacing check.
15839 @item 3, send_field_nospatial
15840 Like @code{send_field}, but it skips the spatial interlacing check.
15843 The default value is @code{send_frame}.
15846 The picture field parity assumed for the input interlaced video. It accepts one
15847 of the following values:
15851 Assume the top field is first.
15853 Assume the bottom field is first.
15855 Enable automatic detection of field parity.
15858 The default value is @code{auto}.
15859 If the interlacing is unknown or the decoder does not export this information,
15860 top field first will be assumed.
15863 Specify which frames to deinterlace. Accept one of the following
15868 Deinterlace all frames.
15869 @item 1, interlaced
15870 Only deinterlace frames marked as interlaced.
15873 The default value is @code{all}.
15878 Apply Zoom & Pan effect.
15880 This filter accepts the following options:
15884 Set the zoom expression. Default is 1.
15888 Set the x and y expression. Default is 0.
15891 Set the duration expression in number of frames.
15892 This sets for how many number of frames effect will last for
15893 single input image.
15896 Set the output image size, default is 'hd720'.
15899 Set the output frame rate, default is '25'.
15902 Each expression can contain the following constants:
15921 Output frame count.
15925 Last calculated 'x' and 'y' position from 'x' and 'y' expression
15926 for current input frame.
15930 'x' and 'y' of last output frame of previous input frame or 0 when there was
15931 not yet such frame (first input frame).
15934 Last calculated zoom from 'z' expression for current input frame.
15937 Last calculated zoom of last output frame of previous input frame.
15940 Number of output frames for current input frame. Calculated from 'd' expression
15941 for each input frame.
15944 number of output frames created for previous input frame
15947 Rational number: input width / input height
15950 sample aspect ratio
15953 display aspect ratio
15957 @subsection Examples
15961 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
15963 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
15967 Zoom-in up to 1.5 and pan always at center of picture:
15969 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
15973 Same as above but without pausing:
15975 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
15981 Scale (resize) the input video, using the z.lib library:
15982 https://github.com/sekrit-twc/zimg.
15984 The zscale filter forces the output display aspect ratio to be the same
15985 as the input, by changing the output sample aspect ratio.
15987 If the input image format is different from the format requested by
15988 the next filter, the zscale filter will convert the input to the
15991 @subsection Options
15992 The filter accepts the following options.
15997 Set the output video dimension expression. Default value is the input
16000 If the @var{width} or @var{w} value is 0, the input width is used for
16001 the output. If the @var{height} or @var{h} value is 0, the input height
16002 is used for the output.
16004 If one and only one of the values is -n with n >= 1, the zscale filter
16005 will use a value that maintains the aspect ratio of the input image,
16006 calculated from the other specified dimension. After that it will,
16007 however, make sure that the calculated dimension is divisible by n and
16008 adjust the value if necessary.
16010 If both values are -n with n >= 1, the behavior will be identical to
16011 both values being set to 0 as previously detailed.
16013 See below for the list of accepted constants for use in the dimension
16017 Set the video size. For the syntax of this option, check the
16018 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16021 Set the dither type.
16023 Possible values are:
16028 @item error_diffusion
16034 Set the resize filter type.
16036 Possible values are:
16046 Default is bilinear.
16049 Set the color range.
16051 Possible values are:
16058 Default is same as input.
16061 Set the color primaries.
16063 Possible values are:
16073 Default is same as input.
16076 Set the transfer characteristics.
16078 Possible values are:
16092 Default is same as input.
16095 Set the colorspace matrix.
16097 Possible value are:
16108 Default is same as input.
16111 Set the input color range.
16113 Possible values are:
16120 Default is same as input.
16122 @item primariesin, pin
16123 Set the input color primaries.
16125 Possible values are:
16135 Default is same as input.
16137 @item transferin, tin
16138 Set the input transfer characteristics.
16140 Possible values are:
16151 Default is same as input.
16153 @item matrixin, min
16154 Set the input colorspace matrix.
16156 Possible value are:
16168 Set the output chroma location.
16170 Possible values are:
16181 @item chromalin, cin
16182 Set the input chroma location.
16184 Possible values are:
16196 Set the nominal peak luminance.
16199 The values of the @option{w} and @option{h} options are expressions
16200 containing the following constants:
16205 The input width and height
16209 These are the same as @var{in_w} and @var{in_h}.
16213 The output (scaled) width and height
16217 These are the same as @var{out_w} and @var{out_h}
16220 The same as @var{iw} / @var{ih}
16223 input sample aspect ratio
16226 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16230 horizontal and vertical input chroma subsample values. For example for the
16231 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16235 horizontal and vertical output chroma subsample values. For example for the
16236 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16242 @c man end VIDEO FILTERS
16244 @chapter Video Sources
16245 @c man begin VIDEO SOURCES
16247 Below is a description of the currently available video sources.
16251 Buffer video frames, and make them available to the filter chain.
16253 This source is mainly intended for a programmatic use, in particular
16254 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
16256 It accepts the following parameters:
16261 Specify the size (width and height) of the buffered video frames. For the
16262 syntax of this option, check the
16263 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16266 The input video width.
16269 The input video height.
16272 A string representing the pixel format of the buffered video frames.
16273 It may be a number corresponding to a pixel format, or a pixel format
16277 Specify the timebase assumed by the timestamps of the buffered frames.
16280 Specify the frame rate expected for the video stream.
16282 @item pixel_aspect, sar
16283 The sample (pixel) aspect ratio of the input video.
16286 Specify the optional parameters to be used for the scale filter which
16287 is automatically inserted when an input change is detected in the
16288 input size or format.
16290 @item hw_frames_ctx
16291 When using a hardware pixel format, this should be a reference to an
16292 AVHWFramesContext describing input frames.
16297 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
16300 will instruct the source to accept video frames with size 320x240 and
16301 with format "yuv410p", assuming 1/24 as the timestamps timebase and
16302 square pixels (1:1 sample aspect ratio).
16303 Since the pixel format with name "yuv410p" corresponds to the number 6
16304 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
16305 this example corresponds to:
16307 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
16310 Alternatively, the options can be specified as a flat string, but this
16311 syntax is deprecated:
16313 @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}]
16317 Create a pattern generated by an elementary cellular automaton.
16319 The initial state of the cellular automaton can be defined through the
16320 @option{filename} and @option{pattern} options. If such options are
16321 not specified an initial state is created randomly.
16323 At each new frame a new row in the video is filled with the result of
16324 the cellular automaton next generation. The behavior when the whole
16325 frame is filled is defined by the @option{scroll} option.
16327 This source accepts the following options:
16331 Read the initial cellular automaton state, i.e. the starting row, from
16332 the specified file.
16333 In the file, each non-whitespace character is considered an alive
16334 cell, a newline will terminate the row, and further characters in the
16335 file will be ignored.
16338 Read the initial cellular automaton state, i.e. the starting row, from
16339 the specified string.
16341 Each non-whitespace character in the string is considered an alive
16342 cell, a newline will terminate the row, and further characters in the
16343 string will be ignored.
16346 Set the video rate, that is the number of frames generated per second.
16349 @item random_fill_ratio, ratio
16350 Set the random fill ratio for the initial cellular automaton row. It
16351 is a floating point number value ranging from 0 to 1, defaults to
16354 This option is ignored when a file or a pattern is specified.
16356 @item random_seed, seed
16357 Set the seed for filling randomly the initial row, must be an integer
16358 included between 0 and UINT32_MAX. If not specified, or if explicitly
16359 set to -1, the filter will try to use a good random seed on a best
16363 Set the cellular automaton rule, it is a number ranging from 0 to 255.
16364 Default value is 110.
16367 Set the size of the output video. For the syntax of this option, check the
16368 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16370 If @option{filename} or @option{pattern} is specified, the size is set
16371 by default to the width of the specified initial state row, and the
16372 height is set to @var{width} * PHI.
16374 If @option{size} is set, it must contain the width of the specified
16375 pattern string, and the specified pattern will be centered in the
16378 If a filename or a pattern string is not specified, the size value
16379 defaults to "320x518" (used for a randomly generated initial state).
16382 If set to 1, scroll the output upward when all the rows in the output
16383 have been already filled. If set to 0, the new generated row will be
16384 written over the top row just after the bottom row is filled.
16387 @item start_full, full
16388 If set to 1, completely fill the output with generated rows before
16389 outputting the first frame.
16390 This is the default behavior, for disabling set the value to 0.
16393 If set to 1, stitch the left and right row edges together.
16394 This is the default behavior, for disabling set the value to 0.
16397 @subsection Examples
16401 Read the initial state from @file{pattern}, and specify an output of
16404 cellauto=f=pattern:s=200x400
16408 Generate a random initial row with a width of 200 cells, with a fill
16411 cellauto=ratio=2/3:s=200x200
16415 Create a pattern generated by rule 18 starting by a single alive cell
16416 centered on an initial row with width 100:
16418 cellauto=p=@@:s=100x400:full=0:rule=18
16422 Specify a more elaborated initial pattern:
16424 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
16429 @anchor{coreimagesrc}
16430 @section coreimagesrc
16431 Video source generated on GPU using Apple's CoreImage API on OSX.
16433 This video source is a specialized version of the @ref{coreimage} video filter.
16434 Use a core image generator at the beginning of the applied filterchain to
16435 generate the content.
16437 The coreimagesrc video source accepts the following options:
16439 @item list_generators
16440 List all available generators along with all their respective options as well as
16441 possible minimum and maximum values along with the default values.
16443 list_generators=true
16447 Specify the size of the sourced video. For the syntax of this option, check the
16448 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16449 The default value is @code{320x240}.
16452 Specify the frame rate of the sourced video, as the number of frames
16453 generated per second. It has to be a string in the format
16454 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16455 number or a valid video frame rate abbreviation. The default value is
16459 Set the sample aspect ratio of the sourced video.
16462 Set the duration of the sourced video. See
16463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16464 for the accepted syntax.
16466 If not specified, or the expressed duration is negative, the video is
16467 supposed to be generated forever.
16470 Additionally, all options of the @ref{coreimage} video filter are accepted.
16471 A complete filterchain can be used for further processing of the
16472 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
16473 and examples for details.
16475 @subsection Examples
16480 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
16481 given as complete and escaped command-line for Apple's standard bash shell:
16483 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
16485 This example is equivalent to the QRCode example of @ref{coreimage} without the
16486 need for a nullsrc video source.
16490 @section mandelbrot
16492 Generate a Mandelbrot set fractal, and progressively zoom towards the
16493 point specified with @var{start_x} and @var{start_y}.
16495 This source accepts the following options:
16500 Set the terminal pts value. Default value is 400.
16503 Set the terminal scale value.
16504 Must be a floating point value. Default value is 0.3.
16507 Set the inner coloring mode, that is the algorithm used to draw the
16508 Mandelbrot fractal internal region.
16510 It shall assume one of the following values:
16515 Show time until convergence.
16517 Set color based on point closest to the origin of the iterations.
16522 Default value is @var{mincol}.
16525 Set the bailout value. Default value is 10.0.
16528 Set the maximum of iterations performed by the rendering
16529 algorithm. Default value is 7189.
16532 Set outer coloring mode.
16533 It shall assume one of following values:
16535 @item iteration_count
16536 Set iteration cound mode.
16537 @item normalized_iteration_count
16538 set normalized iteration count mode.
16540 Default value is @var{normalized_iteration_count}.
16543 Set frame rate, expressed as number of frames per second. Default
16547 Set frame size. For the syntax of this option, check the "Video
16548 size" section in the ffmpeg-utils manual. Default value is "640x480".
16551 Set the initial scale value. Default value is 3.0.
16554 Set the initial x position. Must be a floating point value between
16555 -100 and 100. Default value is -0.743643887037158704752191506114774.
16558 Set the initial y position. Must be a floating point value between
16559 -100 and 100. Default value is -0.131825904205311970493132056385139.
16564 Generate various test patterns, as generated by the MPlayer test filter.
16566 The size of the generated video is fixed, and is 256x256.
16567 This source is useful in particular for testing encoding features.
16569 This source accepts the following options:
16574 Specify the frame rate of the sourced video, as the number of frames
16575 generated per second. It has to be a string in the format
16576 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16577 number or a valid video frame rate abbreviation. The default value is
16581 Set the duration of the sourced video. See
16582 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16583 for the accepted syntax.
16585 If not specified, or the expressed duration is negative, the video is
16586 supposed to be generated forever.
16590 Set the number or the name of the test to perform. Supported tests are:
16606 Default value is "all", which will cycle through the list of all tests.
16611 mptestsrc=t=dc_luma
16614 will generate a "dc_luma" test pattern.
16616 @section frei0r_src
16618 Provide a frei0r source.
16620 To enable compilation of this filter you need to install the frei0r
16621 header and configure FFmpeg with @code{--enable-frei0r}.
16623 This source accepts the following parameters:
16628 The size of the video to generate. For the syntax of this option, check the
16629 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16632 The framerate of the generated video. It may be a string of the form
16633 @var{num}/@var{den} or a frame rate abbreviation.
16636 The name to the frei0r source to load. For more information regarding frei0r and
16637 how to set the parameters, read the @ref{frei0r} section in the video filters
16640 @item filter_params
16641 A '|'-separated list of parameters to pass to the frei0r source.
16645 For example, to generate a frei0r partik0l source with size 200x200
16646 and frame rate 10 which is overlaid on the overlay filter main input:
16648 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
16653 Generate a life pattern.
16655 This source is based on a generalization of John Conway's life game.
16657 The sourced input represents a life grid, each pixel represents a cell
16658 which can be in one of two possible states, alive or dead. Every cell
16659 interacts with its eight neighbours, which are the cells that are
16660 horizontally, vertically, or diagonally adjacent.
16662 At each interaction the grid evolves according to the adopted rule,
16663 which specifies the number of neighbor alive cells which will make a
16664 cell stay alive or born. The @option{rule} option allows one to specify
16667 This source accepts the following options:
16671 Set the file from which to read the initial grid state. In the file,
16672 each non-whitespace character is considered an alive cell, and newline
16673 is used to delimit the end of each row.
16675 If this option is not specified, the initial grid is generated
16679 Set the video rate, that is the number of frames generated per second.
16682 @item random_fill_ratio, ratio
16683 Set the random fill ratio for the initial random grid. It is a
16684 floating point number value ranging from 0 to 1, defaults to 1/PHI.
16685 It is ignored when a file is specified.
16687 @item random_seed, seed
16688 Set the seed for filling the initial random grid, must be an integer
16689 included between 0 and UINT32_MAX. If not specified, or if explicitly
16690 set to -1, the filter will try to use a good random seed on a best
16696 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
16697 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
16698 @var{NS} specifies the number of alive neighbor cells which make a
16699 live cell stay alive, and @var{NB} the number of alive neighbor cells
16700 which make a dead cell to become alive (i.e. to "born").
16701 "s" and "b" can be used in place of "S" and "B", respectively.
16703 Alternatively a rule can be specified by an 18-bits integer. The 9
16704 high order bits are used to encode the next cell state if it is alive
16705 for each number of neighbor alive cells, the low order bits specify
16706 the rule for "borning" new cells. Higher order bits encode for an
16707 higher number of neighbor cells.
16708 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
16709 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
16711 Default value is "S23/B3", which is the original Conway's game of life
16712 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
16713 cells, and will born a new cell if there are three alive cells around
16717 Set the size of the output video. For the syntax of this option, check the
16718 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16720 If @option{filename} is specified, the size is set by default to the
16721 same size of the input file. If @option{size} is set, it must contain
16722 the size specified in the input file, and the initial grid defined in
16723 that file is centered in the larger resulting area.
16725 If a filename is not specified, the size value defaults to "320x240"
16726 (used for a randomly generated initial grid).
16729 If set to 1, stitch the left and right grid edges together, and the
16730 top and bottom edges also. Defaults to 1.
16733 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
16734 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
16735 value from 0 to 255.
16738 Set the color of living (or new born) cells.
16741 Set the color of dead cells. If @option{mold} is set, this is the first color
16742 used to represent a dead cell.
16745 Set mold color, for definitely dead and moldy cells.
16747 For the syntax of these 3 color options, check the "Color" section in the
16748 ffmpeg-utils manual.
16751 @subsection Examples
16755 Read a grid from @file{pattern}, and center it on a grid of size
16758 life=f=pattern:s=300x300
16762 Generate a random grid of size 200x200, with a fill ratio of 2/3:
16764 life=ratio=2/3:s=200x200
16768 Specify a custom rule for evolving a randomly generated grid:
16774 Full example with slow death effect (mold) using @command{ffplay}:
16776 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
16783 @anchor{haldclutsrc}
16785 @anchor{rgbtestsrc}
16787 @anchor{smptehdbars}
16790 @anchor{yuvtestsrc}
16791 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
16793 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
16795 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
16797 The @code{color} source provides an uniformly colored input.
16799 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
16800 @ref{haldclut} filter.
16802 The @code{nullsrc} source returns unprocessed video frames. It is
16803 mainly useful to be employed in analysis / debugging tools, or as the
16804 source for filters which ignore the input data.
16806 The @code{rgbtestsrc} source generates an RGB test pattern useful for
16807 detecting RGB vs BGR issues. You should see a red, green and blue
16808 stripe from top to bottom.
16810 The @code{smptebars} source generates a color bars pattern, based on
16811 the SMPTE Engineering Guideline EG 1-1990.
16813 The @code{smptehdbars} source generates a color bars pattern, based on
16814 the SMPTE RP 219-2002.
16816 The @code{testsrc} source generates a test video pattern, showing a
16817 color pattern, a scrolling gradient and a timestamp. This is mainly
16818 intended for testing purposes.
16820 The @code{testsrc2} source is similar to testsrc, but supports more
16821 pixel formats instead of just @code{rgb24}. This allows using it as an
16822 input for other tests without requiring a format conversion.
16824 The @code{yuvtestsrc} source generates an YUV test pattern. You should
16825 see a y, cb and cr stripe from top to bottom.
16827 The sources accept the following parameters:
16832 Specify the alpha (opacity) of the background, only available in the
16833 @code{testsrc2} source. The value must be between 0 (fully transparent) and
16834 255 (fully opaque, the default).
16837 Specify the color of the source, only available in the @code{color}
16838 source. For the syntax of this option, check the "Color" section in the
16839 ffmpeg-utils manual.
16842 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
16843 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
16844 pixels to be used as identity matrix for 3D lookup tables. Each component is
16845 coded on a @code{1/(N*N)} scale.
16848 Specify the size of the sourced video. For the syntax of this option, check the
16849 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16850 The default value is @code{320x240}.
16852 This option is not available with the @code{haldclutsrc} filter.
16855 Specify the frame rate of the sourced video, as the number of frames
16856 generated per second. It has to be a string in the format
16857 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16858 number or a valid video frame rate abbreviation. The default value is
16862 Set the sample aspect ratio of the sourced video.
16865 Set the duration of the sourced video. See
16866 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16867 for the accepted syntax.
16869 If not specified, or the expressed duration is negative, the video is
16870 supposed to be generated forever.
16873 Set the number of decimals to show in the timestamp, only available in the
16874 @code{testsrc} source.
16876 The displayed timestamp value will correspond to the original
16877 timestamp value multiplied by the power of 10 of the specified
16878 value. Default value is 0.
16881 For example the following:
16883 testsrc=duration=5.3:size=qcif:rate=10
16886 will generate a video with a duration of 5.3 seconds, with size
16887 176x144 and a frame rate of 10 frames per second.
16889 The following graph description will generate a red source
16890 with an opacity of 0.2, with size "qcif" and a frame rate of 10
16893 color=c=red@@0.2:s=qcif:r=10
16896 If the input content is to be ignored, @code{nullsrc} can be used. The
16897 following command generates noise in the luminance plane by employing
16898 the @code{geq} filter:
16900 nullsrc=s=256x256, geq=random(1)*255:128:128
16903 @subsection Commands
16905 The @code{color} source supports the following commands:
16909 Set the color of the created image. Accepts the same syntax of the
16910 corresponding @option{color} option.
16913 @c man end VIDEO SOURCES
16915 @chapter Video Sinks
16916 @c man begin VIDEO SINKS
16918 Below is a description of the currently available video sinks.
16920 @section buffersink
16922 Buffer video frames, and make them available to the end of the filter
16925 This sink is mainly intended for programmatic use, in particular
16926 through the interface defined in @file{libavfilter/buffersink.h}
16927 or the options system.
16929 It accepts a pointer to an AVBufferSinkContext structure, which
16930 defines the incoming buffers' formats, to be passed as the opaque
16931 parameter to @code{avfilter_init_filter} for initialization.
16935 Null video sink: do absolutely nothing with the input video. It is
16936 mainly useful as a template and for use in analysis / debugging
16939 @c man end VIDEO SINKS
16941 @chapter Multimedia Filters
16942 @c man begin MULTIMEDIA FILTERS
16944 Below is a description of the currently available multimedia filters.
16948 Convert input audio to a video output, displaying the audio bit scope.
16950 The filter accepts the following options:
16954 Set frame rate, expressed as number of frames per second. Default
16958 Specify the video size for the output. For the syntax of this option, check the
16959 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16960 Default value is @code{1024x256}.
16963 Specify list of colors separated by space or by '|' which will be used to
16964 draw channels. Unrecognized or missing colors will be replaced
16968 @section ahistogram
16970 Convert input audio to a video output, displaying the volume histogram.
16972 The filter accepts the following options:
16976 Specify how histogram is calculated.
16978 It accepts the following values:
16981 Use single histogram for all channels.
16983 Use separate histogram for each channel.
16985 Default is @code{single}.
16988 Set frame rate, expressed as number of frames per second. Default
16992 Specify the video size for the output. For the syntax of this option, check the
16993 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16994 Default value is @code{hd720}.
16999 It accepts the following values:
17010 reverse logarithmic
17012 Default is @code{log}.
17015 Set amplitude scale.
17017 It accepts the following values:
17024 Default is @code{log}.
17027 Set how much frames to accumulate in histogram.
17028 Defauls is 1. Setting this to -1 accumulates all frames.
17031 Set histogram ratio of window height.
17034 Set sonogram sliding.
17036 It accepts the following values:
17039 replace old rows with new ones.
17041 scroll from top to bottom.
17043 Default is @code{replace}.
17046 @section aphasemeter
17048 Convert input audio to a video output, displaying the audio phase.
17050 The filter accepts the following options:
17054 Set the output frame rate. Default value is @code{25}.
17057 Set the video size for the output. For the syntax of this option, check the
17058 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17059 Default value is @code{800x400}.
17064 Specify the red, green, blue contrast. Default values are @code{2},
17065 @code{7} and @code{1}.
17066 Allowed range is @code{[0, 255]}.
17069 Set color which will be used for drawing median phase. If color is
17070 @code{none} which is default, no median phase value will be drawn.
17073 Enable video output. Default is enabled.
17076 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17077 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17078 The @code{-1} means left and right channels are completely out of phase and
17079 @code{1} means channels are in phase.
17081 @section avectorscope
17083 Convert input audio to a video output, representing the audio vector
17086 The filter is used to measure the difference between channels of stereo
17087 audio stream. A monoaural signal, consisting of identical left and right
17088 signal, results in straight vertical line. Any stereo separation is visible
17089 as a deviation from this line, creating a Lissajous figure.
17090 If the straight (or deviation from it) but horizontal line appears this
17091 indicates that the left and right channels are out of phase.
17093 The filter accepts the following options:
17097 Set the vectorscope mode.
17099 Available values are:
17102 Lissajous rotated by 45 degrees.
17105 Same as above but not rotated.
17108 Shape resembling half of circle.
17111 Default value is @samp{lissajous}.
17114 Set the video size for the output. For the syntax of this option, check the
17115 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17116 Default value is @code{400x400}.
17119 Set the output frame rate. Default value is @code{25}.
17125 Specify the red, green, blue and alpha contrast. Default values are @code{40},
17126 @code{160}, @code{80} and @code{255}.
17127 Allowed range is @code{[0, 255]}.
17133 Specify the red, green, blue and alpha fade. Default values are @code{15},
17134 @code{10}, @code{5} and @code{5}.
17135 Allowed range is @code{[0, 255]}.
17138 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
17139 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
17142 Set the vectorscope drawing mode.
17144 Available values are:
17147 Draw dot for each sample.
17150 Draw line between previous and current sample.
17153 Default value is @samp{dot}.
17156 Specify amplitude scale of audio samples.
17158 Available values are:
17175 @subsection Examples
17179 Complete example using @command{ffplay}:
17181 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17182 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
17186 @section bench, abench
17188 Benchmark part of a filtergraph.
17190 The filter accepts the following options:
17194 Start or stop a timer.
17196 Available values are:
17199 Get the current time, set it as frame metadata (using the key
17200 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
17203 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
17204 the input frame metadata to get the time difference. Time difference, average,
17205 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
17206 @code{min}) are then printed. The timestamps are expressed in seconds.
17210 @subsection Examples
17214 Benchmark @ref{selectivecolor} filter:
17216 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
17222 Concatenate audio and video streams, joining them together one after the
17225 The filter works on segments of synchronized video and audio streams. All
17226 segments must have the same number of streams of each type, and that will
17227 also be the number of streams at output.
17229 The filter accepts the following options:
17234 Set the number of segments. Default is 2.
17237 Set the number of output video streams, that is also the number of video
17238 streams in each segment. Default is 1.
17241 Set the number of output audio streams, that is also the number of audio
17242 streams in each segment. Default is 0.
17245 Activate unsafe mode: do not fail if segments have a different format.
17249 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
17250 @var{a} audio outputs.
17252 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
17253 segment, in the same order as the outputs, then the inputs for the second
17256 Related streams do not always have exactly the same duration, for various
17257 reasons including codec frame size or sloppy authoring. For that reason,
17258 related synchronized streams (e.g. a video and its audio track) should be
17259 concatenated at once. The concat filter will use the duration of the longest
17260 stream in each segment (except the last one), and if necessary pad shorter
17261 audio streams with silence.
17263 For this filter to work correctly, all segments must start at timestamp 0.
17265 All corresponding streams must have the same parameters in all segments; the
17266 filtering system will automatically select a common pixel format for video
17267 streams, and a common sample format, sample rate and channel layout for
17268 audio streams, but other settings, such as resolution, must be converted
17269 explicitly by the user.
17271 Different frame rates are acceptable but will result in variable frame rate
17272 at output; be sure to configure the output file to handle it.
17274 @subsection Examples
17278 Concatenate an opening, an episode and an ending, all in bilingual version
17279 (video in stream 0, audio in streams 1 and 2):
17281 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
17282 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
17283 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
17284 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
17288 Concatenate two parts, handling audio and video separately, using the
17289 (a)movie sources, and adjusting the resolution:
17291 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
17292 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
17293 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
17295 Note that a desync will happen at the stitch if the audio and video streams
17296 do not have exactly the same duration in the first file.
17300 @section drawgraph, adrawgraph
17302 Draw a graph using input video or audio metadata.
17304 It accepts the following parameters:
17308 Set 1st frame metadata key from which metadata values will be used to draw a graph.
17311 Set 1st foreground color expression.
17314 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
17317 Set 2nd foreground color expression.
17320 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
17323 Set 3rd foreground color expression.
17326 Set 4th frame metadata key from which metadata values will be used to draw a graph.
17329 Set 4th foreground color expression.
17332 Set minimal value of metadata value.
17335 Set maximal value of metadata value.
17338 Set graph background color. Default is white.
17343 Available values for mode is:
17350 Default is @code{line}.
17355 Available values for slide is:
17358 Draw new frame when right border is reached.
17361 Replace old columns with new ones.
17364 Scroll from right to left.
17367 Scroll from left to right.
17370 Draw single picture.
17373 Default is @code{frame}.
17376 Set size of graph video. For the syntax of this option, check the
17377 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17378 The default value is @code{900x256}.
17380 The foreground color expressions can use the following variables:
17383 Minimal value of metadata value.
17386 Maximal value of metadata value.
17389 Current metadata key value.
17392 The color is defined as 0xAABBGGRR.
17395 Example using metadata from @ref{signalstats} filter:
17397 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
17400 Example using metadata from @ref{ebur128} filter:
17402 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
17408 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
17409 it unchanged. By default, it logs a message at a frequency of 10Hz with the
17410 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
17411 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
17413 The filter also has a video output (see the @var{video} option) with a real
17414 time graph to observe the loudness evolution. The graphic contains the logged
17415 message mentioned above, so it is not printed anymore when this option is set,
17416 unless the verbose logging is set. The main graphing area contains the
17417 short-term loudness (3 seconds of analysis), and the gauge on the right is for
17418 the momentary loudness (400 milliseconds).
17420 More information about the Loudness Recommendation EBU R128 on
17421 @url{http://tech.ebu.ch/loudness}.
17423 The filter accepts the following options:
17428 Activate the video output. The audio stream is passed unchanged whether this
17429 option is set or no. The video stream will be the first output stream if
17430 activated. Default is @code{0}.
17433 Set the video size. This option is for video only. For the syntax of this
17435 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17436 Default and minimum resolution is @code{640x480}.
17439 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
17440 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
17441 other integer value between this range is allowed.
17444 Set metadata injection. If set to @code{1}, the audio input will be segmented
17445 into 100ms output frames, each of them containing various loudness information
17446 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
17448 Default is @code{0}.
17451 Force the frame logging level.
17453 Available values are:
17456 information logging level
17458 verbose logging level
17461 By default, the logging level is set to @var{info}. If the @option{video} or
17462 the @option{metadata} options are set, it switches to @var{verbose}.
17467 Available modes can be cumulated (the option is a @code{flag} type). Possible
17471 Disable any peak mode (default).
17473 Enable sample-peak mode.
17475 Simple peak mode looking for the higher sample value. It logs a message
17476 for sample-peak (identified by @code{SPK}).
17478 Enable true-peak mode.
17480 If enabled, the peak lookup is done on an over-sampled version of the input
17481 stream for better peak accuracy. It logs a message for true-peak.
17482 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
17483 This mode requires a build with @code{libswresample}.
17487 Treat mono input files as "dual mono". If a mono file is intended for playback
17488 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
17489 If set to @code{true}, this option will compensate for this effect.
17490 Multi-channel input files are not affected by this option.
17493 Set a specific pan law to be used for the measurement of dual mono files.
17494 This parameter is optional, and has a default value of -3.01dB.
17497 @subsection Examples
17501 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
17503 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
17507 Run an analysis with @command{ffmpeg}:
17509 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
17513 @section interleave, ainterleave
17515 Temporally interleave frames from several inputs.
17517 @code{interleave} works with video inputs, @code{ainterleave} with audio.
17519 These filters read frames from several inputs and send the oldest
17520 queued frame to the output.
17522 Input streams must have well defined, monotonically increasing frame
17525 In order to submit one frame to output, these filters need to enqueue
17526 at least one frame for each input, so they cannot work in case one
17527 input is not yet terminated and will not receive incoming frames.
17529 For example consider the case when one input is a @code{select} filter
17530 which always drops input frames. The @code{interleave} filter will keep
17531 reading from that input, but it will never be able to send new frames
17532 to output until the input sends an end-of-stream signal.
17534 Also, depending on inputs synchronization, the filters will drop
17535 frames in case one input receives more frames than the other ones, and
17536 the queue is already filled.
17538 These filters accept the following options:
17542 Set the number of different inputs, it is 2 by default.
17545 @subsection Examples
17549 Interleave frames belonging to different streams using @command{ffmpeg}:
17551 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
17555 Add flickering blur effect:
17557 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
17561 @section metadata, ametadata
17563 Manipulate frame metadata.
17565 This filter accepts the following options:
17569 Set mode of operation of the filter.
17571 Can be one of the following:
17575 If both @code{value} and @code{key} is set, select frames
17576 which have such metadata. If only @code{key} is set, select
17577 every frame that has such key in metadata.
17580 Add new metadata @code{key} and @code{value}. If key is already available
17584 Modify value of already present key.
17587 If @code{value} is set, delete only keys that have such value.
17588 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
17592 Print key and its value if metadata was found. If @code{key} is not set print all
17593 metadata values available in frame.
17597 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
17600 Set metadata value which will be used. This option is mandatory for
17601 @code{modify} and @code{add} mode.
17604 Which function to use when comparing metadata value and @code{value}.
17606 Can be one of following:
17610 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
17613 Values are interpreted as strings, returns true if metadata value starts with
17614 the @code{value} option string.
17617 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
17620 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
17623 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
17626 Values are interpreted as floats, returns true if expression from option @code{expr}
17631 Set expression which is used when @code{function} is set to @code{expr}.
17632 The expression is evaluated through the eval API and can contain the following
17637 Float representation of @code{value} from metadata key.
17640 Float representation of @code{value} as supplied by user in @code{value} option.
17644 If specified in @code{print} mode, output is written to the named file. Instead of
17645 plain filename any writable url can be specified. Filename ``-'' is a shorthand
17646 for standard output. If @code{file} option is not set, output is written to the log
17647 with AV_LOG_INFO loglevel.
17651 @subsection Examples
17655 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
17658 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
17661 Print silencedetect output to file @file{metadata.txt}.
17663 silencedetect,ametadata=mode=print:file=metadata.txt
17666 Direct all metadata to a pipe with file descriptor 4.
17668 metadata=mode=print:file='pipe\:4'
17672 @section perms, aperms
17674 Set read/write permissions for the output frames.
17676 These filters are mainly aimed at developers to test direct path in the
17677 following filter in the filtergraph.
17679 The filters accept the following options:
17683 Select the permissions mode.
17685 It accepts the following values:
17688 Do nothing. This is the default.
17690 Set all the output frames read-only.
17692 Set all the output frames directly writable.
17694 Make the frame read-only if writable, and writable if read-only.
17696 Set each output frame read-only or writable randomly.
17700 Set the seed for the @var{random} mode, must be an integer included between
17701 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
17702 @code{-1}, the filter will try to use a good random seed on a best effort
17706 Note: in case of auto-inserted filter between the permission filter and the
17707 following one, the permission might not be received as expected in that
17708 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
17709 perms/aperms filter can avoid this problem.
17711 @section realtime, arealtime
17713 Slow down filtering to match real time approximately.
17715 These filters will pause the filtering for a variable amount of time to
17716 match the output rate with the input timestamps.
17717 They are similar to the @option{re} option to @code{ffmpeg}.
17719 They accept the following options:
17723 Time limit for the pauses. Any pause longer than that will be considered
17724 a timestamp discontinuity and reset the timer. Default is 2 seconds.
17728 @section select, aselect
17730 Select frames to pass in output.
17732 This filter accepts the following options:
17737 Set expression, which is evaluated for each input frame.
17739 If the expression is evaluated to zero, the frame is discarded.
17741 If the evaluation result is negative or NaN, the frame is sent to the
17742 first output; otherwise it is sent to the output with index
17743 @code{ceil(val)-1}, assuming that the input index starts from 0.
17745 For example a value of @code{1.2} corresponds to the output with index
17746 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
17749 Set the number of outputs. The output to which to send the selected
17750 frame is based on the result of the evaluation. Default value is 1.
17753 The expression can contain the following constants:
17757 The (sequential) number of the filtered frame, starting from 0.
17760 The (sequential) number of the selected frame, starting from 0.
17762 @item prev_selected_n
17763 The sequential number of the last selected frame. It's NAN if undefined.
17766 The timebase of the input timestamps.
17769 The PTS (Presentation TimeStamp) of the filtered video frame,
17770 expressed in @var{TB} units. It's NAN if undefined.
17773 The PTS of the filtered video frame,
17774 expressed in seconds. It's NAN if undefined.
17777 The PTS of the previously filtered video frame. It's NAN if undefined.
17779 @item prev_selected_pts
17780 The PTS of the last previously filtered video frame. It's NAN if undefined.
17782 @item prev_selected_t
17783 The PTS of the last previously selected video frame. It's NAN if undefined.
17786 The PTS of the first video frame in the video. It's NAN if undefined.
17789 The time of the first video frame in the video. It's NAN if undefined.
17791 @item pict_type @emph{(video only)}
17792 The type of the filtered frame. It can assume one of the following
17804 @item interlace_type @emph{(video only)}
17805 The frame interlace type. It can assume one of the following values:
17808 The frame is progressive (not interlaced).
17810 The frame is top-field-first.
17812 The frame is bottom-field-first.
17815 @item consumed_sample_n @emph{(audio only)}
17816 the number of selected samples before the current frame
17818 @item samples_n @emph{(audio only)}
17819 the number of samples in the current frame
17821 @item sample_rate @emph{(audio only)}
17822 the input sample rate
17825 This is 1 if the filtered frame is a key-frame, 0 otherwise.
17828 the position in the file of the filtered frame, -1 if the information
17829 is not available (e.g. for synthetic video)
17831 @item scene @emph{(video only)}
17832 value between 0 and 1 to indicate a new scene; a low value reflects a low
17833 probability for the current frame to introduce a new scene, while a higher
17834 value means the current frame is more likely to be one (see the example below)
17836 @item concatdec_select
17837 The concat demuxer can select only part of a concat input file by setting an
17838 inpoint and an outpoint, but the output packets may not be entirely contained
17839 in the selected interval. By using this variable, it is possible to skip frames
17840 generated by the concat demuxer which are not exactly contained in the selected
17843 This works by comparing the frame pts against the @var{lavf.concat.start_time}
17844 and the @var{lavf.concat.duration} packet metadata values which are also
17845 present in the decoded frames.
17847 The @var{concatdec_select} variable is -1 if the frame pts is at least
17848 start_time and either the duration metadata is missing or the frame pts is less
17849 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
17852 That basically means that an input frame is selected if its pts is within the
17853 interval set by the concat demuxer.
17857 The default value of the select expression is "1".
17859 @subsection Examples
17863 Select all frames in input:
17868 The example above is the same as:
17880 Select only I-frames:
17882 select='eq(pict_type\,I)'
17886 Select one frame every 100:
17888 select='not(mod(n\,100))'
17892 Select only frames contained in the 10-20 time interval:
17894 select=between(t\,10\,20)
17898 Select only I-frames contained in the 10-20 time interval:
17900 select=between(t\,10\,20)*eq(pict_type\,I)
17904 Select frames with a minimum distance of 10 seconds:
17906 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
17910 Use aselect to select only audio frames with samples number > 100:
17912 aselect='gt(samples_n\,100)'
17916 Create a mosaic of the first scenes:
17918 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
17921 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
17925 Send even and odd frames to separate outputs, and compose them:
17927 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
17931 Select useful frames from an ffconcat file which is using inpoints and
17932 outpoints but where the source files are not intra frame only.
17934 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
17938 @section sendcmd, asendcmd
17940 Send commands to filters in the filtergraph.
17942 These filters read commands to be sent to other filters in the
17945 @code{sendcmd} must be inserted between two video filters,
17946 @code{asendcmd} must be inserted between two audio filters, but apart
17947 from that they act the same way.
17949 The specification of commands can be provided in the filter arguments
17950 with the @var{commands} option, or in a file specified by the
17951 @var{filename} option.
17953 These filters accept the following options:
17956 Set the commands to be read and sent to the other filters.
17958 Set the filename of the commands to be read and sent to the other
17962 @subsection Commands syntax
17964 A commands description consists of a sequence of interval
17965 specifications, comprising a list of commands to be executed when a
17966 particular event related to that interval occurs. The occurring event
17967 is typically the current frame time entering or leaving a given time
17970 An interval is specified by the following syntax:
17972 @var{START}[-@var{END}] @var{COMMANDS};
17975 The time interval is specified by the @var{START} and @var{END} times.
17976 @var{END} is optional and defaults to the maximum time.
17978 The current frame time is considered within the specified interval if
17979 it is included in the interval [@var{START}, @var{END}), that is when
17980 the time is greater or equal to @var{START} and is lesser than
17983 @var{COMMANDS} consists of a sequence of one or more command
17984 specifications, separated by ",", relating to that interval. The
17985 syntax of a command specification is given by:
17987 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
17990 @var{FLAGS} is optional and specifies the type of events relating to
17991 the time interval which enable sending the specified command, and must
17992 be a non-null sequence of identifier flags separated by "+" or "|" and
17993 enclosed between "[" and "]".
17995 The following flags are recognized:
17998 The command is sent when the current frame timestamp enters the
17999 specified interval. In other words, the command is sent when the
18000 previous frame timestamp was not in the given interval, and the
18004 The command is sent when the current frame timestamp leaves the
18005 specified interval. In other words, the command is sent when the
18006 previous frame timestamp was in the given interval, and the
18010 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18013 @var{TARGET} specifies the target of the command, usually the name of
18014 the filter class or a specific filter instance name.
18016 @var{COMMAND} specifies the name of the command for the target filter.
18018 @var{ARG} is optional and specifies the optional list of argument for
18019 the given @var{COMMAND}.
18021 Between one interval specification and another, whitespaces, or
18022 sequences of characters starting with @code{#} until the end of line,
18023 are ignored and can be used to annotate comments.
18025 A simplified BNF description of the commands specification syntax
18028 @var{COMMAND_FLAG} ::= "enter" | "leave"
18029 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18030 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18031 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18032 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18033 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18036 @subsection Examples
18040 Specify audio tempo change at second 4:
18042 asendcmd=c='4.0 atempo tempo 1.5',atempo
18046 Target a specific filter instance:
18048 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18052 Specify a list of drawtext and hue commands in a file.
18054 # show text in the interval 5-10
18055 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18056 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18058 # desaturate the image in the interval 15-20
18059 15.0-20.0 [enter] hue s 0,
18060 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18062 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18064 # apply an exponential saturation fade-out effect, starting from time 25
18065 25 [enter] hue s exp(25-t)
18068 A filtergraph allowing to read and process the above command list
18069 stored in a file @file{test.cmd}, can be specified with:
18071 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18076 @section setpts, asetpts
18078 Change the PTS (presentation timestamp) of the input frames.
18080 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18082 This filter accepts the following options:
18087 The expression which is evaluated for each frame to construct its timestamp.
18091 The expression is evaluated through the eval API and can contain the following
18096 frame rate, only defined for constant frame-rate video
18099 The presentation timestamp in input
18102 The count of the input frame for video or the number of consumed samples,
18103 not including the current frame for audio, starting from 0.
18105 @item NB_CONSUMED_SAMPLES
18106 The number of consumed samples, not including the current frame (only
18109 @item NB_SAMPLES, S
18110 The number of samples in the current frame (only audio)
18112 @item SAMPLE_RATE, SR
18113 The audio sample rate.
18116 The PTS of the first frame.
18119 the time in seconds of the first frame
18122 State whether the current frame is interlaced.
18125 the time in seconds of the current frame
18128 original position in the file of the frame, or undefined if undefined
18129 for the current frame
18132 The previous input PTS.
18135 previous input time in seconds
18138 The previous output PTS.
18141 previous output time in seconds
18144 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
18148 The wallclock (RTC) time at the start of the movie in microseconds.
18151 The timebase of the input timestamps.
18155 @subsection Examples
18159 Start counting PTS from zero
18161 setpts=PTS-STARTPTS
18165 Apply fast motion effect:
18171 Apply slow motion effect:
18177 Set fixed rate of 25 frames per second:
18183 Set fixed rate 25 fps with some jitter:
18185 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
18189 Apply an offset of 10 seconds to the input PTS:
18195 Generate timestamps from a "live source" and rebase onto the current timebase:
18197 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
18201 Generate timestamps by counting samples:
18208 @section settb, asettb
18210 Set the timebase to use for the output frames timestamps.
18211 It is mainly useful for testing timebase configuration.
18213 It accepts the following parameters:
18218 The expression which is evaluated into the output timebase.
18222 The value for @option{tb} is an arithmetic expression representing a
18223 rational. The expression can contain the constants "AVTB" (the default
18224 timebase), "intb" (the input timebase) and "sr" (the sample rate,
18225 audio only). Default value is "intb".
18227 @subsection Examples
18231 Set the timebase to 1/25:
18237 Set the timebase to 1/10:
18243 Set the timebase to 1001/1000:
18249 Set the timebase to 2*intb:
18255 Set the default timebase value:
18262 Convert input audio to a video output representing frequency spectrum
18263 logarithmically using Brown-Puckette constant Q transform algorithm with
18264 direct frequency domain coefficient calculation (but the transform itself
18265 is not really constant Q, instead the Q factor is actually variable/clamped),
18266 with musical tone scale, from E0 to D#10.
18268 The filter accepts the following options:
18272 Specify the video size for the output. It must be even. For the syntax of this option,
18273 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18274 Default value is @code{1920x1080}.
18277 Set the output frame rate. Default value is @code{25}.
18280 Set the bargraph height. It must be even. Default value is @code{-1} which
18281 computes the bargraph height automatically.
18284 Set the axis height. It must be even. Default value is @code{-1} which computes
18285 the axis height automatically.
18288 Set the sonogram height. It must be even. Default value is @code{-1} which
18289 computes the sonogram height automatically.
18292 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
18293 instead. Default value is @code{1}.
18295 @item sono_v, volume
18296 Specify the sonogram volume expression. It can contain variables:
18299 the @var{bar_v} evaluated expression
18300 @item frequency, freq, f
18301 the frequency where it is evaluated
18302 @item timeclamp, tc
18303 the value of @var{timeclamp} option
18307 @item a_weighting(f)
18308 A-weighting of equal loudness
18309 @item b_weighting(f)
18310 B-weighting of equal loudness
18311 @item c_weighting(f)
18312 C-weighting of equal loudness.
18314 Default value is @code{16}.
18316 @item bar_v, volume2
18317 Specify the bargraph volume expression. It can contain variables:
18320 the @var{sono_v} evaluated expression
18321 @item frequency, freq, f
18322 the frequency where it is evaluated
18323 @item timeclamp, tc
18324 the value of @var{timeclamp} option
18328 @item a_weighting(f)
18329 A-weighting of equal loudness
18330 @item b_weighting(f)
18331 B-weighting of equal loudness
18332 @item c_weighting(f)
18333 C-weighting of equal loudness.
18335 Default value is @code{sono_v}.
18337 @item sono_g, gamma
18338 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
18339 higher gamma makes the spectrum having more range. Default value is @code{3}.
18340 Acceptable range is @code{[1, 7]}.
18342 @item bar_g, gamma2
18343 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
18347 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
18348 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
18350 @item timeclamp, tc
18351 Specify the transform timeclamp. At low frequency, there is trade-off between
18352 accuracy in time domain and frequency domain. If timeclamp is lower,
18353 event in time domain is represented more accurately (such as fast bass drum),
18354 otherwise event in frequency domain is represented more accurately
18355 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
18358 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
18359 limits future samples by applying asymmetric windowing in time domain, useful
18360 when low latency is required. Accepted range is @code{[0, 1]}.
18363 Specify the transform base frequency. Default value is @code{20.01523126408007475},
18364 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
18367 Specify the transform end frequency. Default value is @code{20495.59681441799654},
18368 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
18371 This option is deprecated and ignored.
18374 Specify the transform length in time domain. Use this option to control accuracy
18375 trade-off between time domain and frequency domain at every frequency sample.
18376 It can contain variables:
18378 @item frequency, freq, f
18379 the frequency where it is evaluated
18380 @item timeclamp, tc
18381 the value of @var{timeclamp} option.
18383 Default value is @code{384*tc/(384+tc*f)}.
18386 Specify the transform count for every video frame. Default value is @code{6}.
18387 Acceptable range is @code{[1, 30]}.
18390 Specify the transform count for every single pixel. Default value is @code{0},
18391 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
18394 Specify font file for use with freetype to draw the axis. If not specified,
18395 use embedded font. Note that drawing with font file or embedded font is not
18396 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
18400 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
18401 The : in the pattern may be replaced by | to avoid unnecessary escaping.
18404 Specify font color expression. This is arithmetic expression that should return
18405 integer value 0xRRGGBB. It can contain variables:
18407 @item frequency, freq, f
18408 the frequency where it is evaluated
18409 @item timeclamp, tc
18410 the value of @var{timeclamp} option
18415 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
18416 @item r(x), g(x), b(x)
18417 red, green, and blue value of intensity x.
18419 Default value is @code{st(0, (midi(f)-59.5)/12);
18420 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
18421 r(1-ld(1)) + b(ld(1))}.
18424 Specify image file to draw the axis. This option override @var{fontfile} and
18425 @var{fontcolor} option.
18428 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
18429 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
18430 Default value is @code{1}.
18433 Set colorspace. The accepted values are:
18436 Unspecified (default)
18445 BT.470BG or BT.601-6 625
18448 SMPTE-170M or BT.601-6 525
18454 BT.2020 with non-constant luminance
18459 Set spectrogram color scheme. This is list of floating point values with format
18460 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
18461 The default is @code{1|0.5|0|0|0.5|1}.
18465 @subsection Examples
18469 Playing audio while showing the spectrum:
18471 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
18475 Same as above, but with frame rate 30 fps:
18477 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
18481 Playing at 1280x720:
18483 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
18487 Disable sonogram display:
18493 A1 and its harmonics: A1, A2, (near)E3, A3:
18495 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),
18496 asplit[a][out1]; [a] showcqt [out0]'
18500 Same as above, but with more accuracy in frequency domain:
18502 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),
18503 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
18509 bar_v=10:sono_v=bar_v*a_weighting(f)
18513 Custom gamma, now spectrum is linear to the amplitude.
18519 Custom tlength equation:
18521 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)))'
18525 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
18527 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
18531 Custom font using fontconfig:
18533 font='Courier New,Monospace,mono|bold'
18537 Custom frequency range with custom axis using image file:
18539 axisfile=myaxis.png:basefreq=40:endfreq=10000
18545 Convert input audio to video output representing the audio power spectrum.
18546 Audio amplitude is on Y-axis while frequency is on X-axis.
18548 The filter accepts the following options:
18552 Specify size of video. For the syntax of this option, check the
18553 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18554 Default is @code{1024x512}.
18558 This set how each frequency bin will be represented.
18560 It accepts the following values:
18566 Default is @code{bar}.
18569 Set amplitude scale.
18571 It accepts the following values:
18585 Default is @code{log}.
18588 Set frequency scale.
18590 It accepts the following values:
18599 Reverse logarithmic scale.
18601 Default is @code{lin}.
18606 It accepts the following values:
18622 Default is @code{w2048}
18625 Set windowing function.
18627 It accepts the following values:
18649 Default is @code{hanning}.
18652 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
18653 which means optimal overlap for selected window function will be picked.
18656 Set time averaging. Setting this to 0 will display current maximal peaks.
18657 Default is @code{1}, which means time averaging is disabled.
18660 Specify list of colors separated by space or by '|' which will be used to
18661 draw channel frequencies. Unrecognized or missing colors will be replaced
18665 Set channel display mode.
18667 It accepts the following values:
18672 Default is @code{combined}.
18675 Set minimum amplitude used in @code{log} amplitude scaler.
18679 @anchor{showspectrum}
18680 @section showspectrum
18682 Convert input audio to a video output, representing the audio frequency
18685 The filter accepts the following options:
18689 Specify the video size for the output. For the syntax of this option, check the
18690 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18691 Default value is @code{640x512}.
18694 Specify how the spectrum should slide along the window.
18696 It accepts the following values:
18699 the samples start again on the left when they reach the right
18701 the samples scroll from right to left
18703 frames are only produced when the samples reach the right
18705 the samples scroll from left to right
18708 Default value is @code{replace}.
18711 Specify display mode.
18713 It accepts the following values:
18716 all channels are displayed in the same row
18718 all channels are displayed in separate rows
18721 Default value is @samp{combined}.
18724 Specify display color mode.
18726 It accepts the following values:
18729 each channel is displayed in a separate color
18731 each channel is displayed using the same color scheme
18733 each channel is displayed using the rainbow color scheme
18735 each channel is displayed using the moreland color scheme
18737 each channel is displayed using the nebulae color scheme
18739 each channel is displayed using the fire color scheme
18741 each channel is displayed using the fiery color scheme
18743 each channel is displayed using the fruit color scheme
18745 each channel is displayed using the cool color scheme
18748 Default value is @samp{channel}.
18751 Specify scale used for calculating intensity color values.
18753 It accepts the following values:
18758 square root, default
18769 Default value is @samp{sqrt}.
18772 Set saturation modifier for displayed colors. Negative values provide
18773 alternative color scheme. @code{0} is no saturation at all.
18774 Saturation must be in [-10.0, 10.0] range.
18775 Default value is @code{1}.
18778 Set window function.
18780 It accepts the following values:
18804 Default value is @code{hann}.
18807 Set orientation of time vs frequency axis. Can be @code{vertical} or
18808 @code{horizontal}. Default is @code{vertical}.
18811 Set ratio of overlap window. Default value is @code{0}.
18812 When value is @code{1} overlap is set to recommended size for specific
18813 window function currently used.
18816 Set scale gain for calculating intensity color values.
18817 Default value is @code{1}.
18820 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
18823 Set color rotation, must be in [-1.0, 1.0] range.
18824 Default value is @code{0}.
18827 The usage is very similar to the showwaves filter; see the examples in that
18830 @subsection Examples
18834 Large window with logarithmic color scaling:
18836 showspectrum=s=1280x480:scale=log
18840 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
18842 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18843 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
18847 @section showspectrumpic
18849 Convert input audio to a single video frame, representing the audio frequency
18852 The filter accepts the following options:
18856 Specify the video size for the output. For the syntax of this option, check the
18857 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18858 Default value is @code{4096x2048}.
18861 Specify display mode.
18863 It accepts the following values:
18866 all channels are displayed in the same row
18868 all channels are displayed in separate rows
18870 Default value is @samp{combined}.
18873 Specify display color mode.
18875 It accepts the following values:
18878 each channel is displayed in a separate color
18880 each channel is displayed using the same color scheme
18882 each channel is displayed using the rainbow color scheme
18884 each channel is displayed using the moreland color scheme
18886 each channel is displayed using the nebulae color scheme
18888 each channel is displayed using the fire color scheme
18890 each channel is displayed using the fiery color scheme
18892 each channel is displayed using the fruit color scheme
18894 each channel is displayed using the cool color scheme
18896 Default value is @samp{intensity}.
18899 Specify scale used for calculating intensity color values.
18901 It accepts the following values:
18906 square root, default
18916 Default value is @samp{log}.
18919 Set saturation modifier for displayed colors. Negative values provide
18920 alternative color scheme. @code{0} is no saturation at all.
18921 Saturation must be in [-10.0, 10.0] range.
18922 Default value is @code{1}.
18925 Set window function.
18927 It accepts the following values:
18950 Default value is @code{hann}.
18953 Set orientation of time vs frequency axis. Can be @code{vertical} or
18954 @code{horizontal}. Default is @code{vertical}.
18957 Set scale gain for calculating intensity color values.
18958 Default value is @code{1}.
18961 Draw time and frequency axes and legends. Default is enabled.
18964 Set color rotation, must be in [-1.0, 1.0] range.
18965 Default value is @code{0}.
18968 @subsection Examples
18972 Extract an audio spectrogram of a whole audio track
18973 in a 1024x1024 picture using @command{ffmpeg}:
18975 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
18979 @section showvolume
18981 Convert input audio volume to a video output.
18983 The filter accepts the following options:
18990 Set border width, allowed range is [0, 5]. Default is 1.
18993 Set channel width, allowed range is [80, 8192]. Default is 400.
18996 Set channel height, allowed range is [1, 900]. Default is 20.
18999 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19002 Set volume color expression.
19004 The expression can use the following variables:
19008 Current max volume of channel in dB.
19014 Current channel number, starting from 0.
19018 If set, displays channel names. Default is enabled.
19021 If set, displays volume values. Default is enabled.
19024 Set orientation, can be @code{horizontal} or @code{vertical},
19025 default is @code{horizontal}.
19028 Set step size, allowed range s [0, 5]. Default is 0, which means
19034 Convert input audio to a video output, representing the samples waves.
19036 The filter accepts the following options:
19040 Specify the video size for the output. For the syntax of this option, check the
19041 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19042 Default value is @code{600x240}.
19047 Available values are:
19050 Draw a point for each sample.
19053 Draw a vertical line for each sample.
19056 Draw a point for each sample and a line between them.
19059 Draw a centered vertical line for each sample.
19062 Default value is @code{point}.
19065 Set the number of samples which are printed on the same column. A
19066 larger value will decrease the frame rate. Must be a positive
19067 integer. This option can be set only if the value for @var{rate}
19068 is not explicitly specified.
19071 Set the (approximate) output frame rate. This is done by setting the
19072 option @var{n}. Default value is "25".
19074 @item split_channels
19075 Set if channels should be drawn separately or overlap. Default value is 0.
19078 Set colors separated by '|' which are going to be used for drawing of each channel.
19081 Set amplitude scale.
19083 Available values are:
19101 @subsection Examples
19105 Output the input file audio and the corresponding video representation
19108 amovie=a.mp3,asplit[out0],showwaves[out1]
19112 Create a synthetic signal and show it with showwaves, forcing a
19113 frame rate of 30 frames per second:
19115 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
19119 @section showwavespic
19121 Convert input audio to a single video frame, representing the samples waves.
19123 The filter accepts the following options:
19127 Specify the video size for the output. For the syntax of this option, check the
19128 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19129 Default value is @code{600x240}.
19131 @item split_channels
19132 Set if channels should be drawn separately or overlap. Default value is 0.
19135 Set colors separated by '|' which are going to be used for drawing of each channel.
19138 Set amplitude scale.
19140 Available values are:
19158 @subsection Examples
19162 Extract a channel split representation of the wave form of a whole audio track
19163 in a 1024x800 picture using @command{ffmpeg}:
19165 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
19169 @section sidedata, asidedata
19171 Delete frame side data, or select frames based on it.
19173 This filter accepts the following options:
19177 Set mode of operation of the filter.
19179 Can be one of the following:
19183 Select every frame with side data of @code{type}.
19186 Delete side data of @code{type}. If @code{type} is not set, delete all side
19192 Set side data type used with all modes. Must be set for @code{select} mode. For
19193 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
19194 in @file{libavutil/frame.h}. For example, to choose
19195 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
19199 @section spectrumsynth
19201 Sythesize audio from 2 input video spectrums, first input stream represents
19202 magnitude across time and second represents phase across time.
19203 The filter will transform from frequency domain as displayed in videos back
19204 to time domain as presented in audio output.
19206 This filter is primarily created for reversing processed @ref{showspectrum}
19207 filter outputs, but can synthesize sound from other spectrograms too.
19208 But in such case results are going to be poor if the phase data is not
19209 available, because in such cases phase data need to be recreated, usually
19210 its just recreated from random noise.
19211 For best results use gray only output (@code{channel} color mode in
19212 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
19213 @code{lin} scale for phase video. To produce phase, for 2nd video, use
19214 @code{data} option. Inputs videos should generally use @code{fullframe}
19215 slide mode as that saves resources needed for decoding video.
19217 The filter accepts the following options:
19221 Specify sample rate of output audio, the sample rate of audio from which
19222 spectrum was generated may differ.
19225 Set number of channels represented in input video spectrums.
19228 Set scale which was used when generating magnitude input spectrum.
19229 Can be @code{lin} or @code{log}. Default is @code{log}.
19232 Set slide which was used when generating inputs spectrums.
19233 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
19234 Default is @code{fullframe}.
19237 Set window function used for resynthesis.
19240 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19241 which means optimal overlap for selected window function will be picked.
19244 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
19245 Default is @code{vertical}.
19248 @subsection Examples
19252 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
19253 then resynthesize videos back to audio with spectrumsynth:
19255 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
19256 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
19257 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
19261 @section split, asplit
19263 Split input into several identical outputs.
19265 @code{asplit} works with audio input, @code{split} with video.
19267 The filter accepts a single parameter which specifies the number of outputs. If
19268 unspecified, it defaults to 2.
19270 @subsection Examples
19274 Create two separate outputs from the same input:
19276 [in] split [out0][out1]
19280 To create 3 or more outputs, you need to specify the number of
19283 [in] asplit=3 [out0][out1][out2]
19287 Create two separate outputs from the same input, one cropped and
19290 [in] split [splitout1][splitout2];
19291 [splitout1] crop=100:100:0:0 [cropout];
19292 [splitout2] pad=200:200:100:100 [padout];
19296 Create 5 copies of the input audio with @command{ffmpeg}:
19298 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
19304 Receive commands sent through a libzmq client, and forward them to
19305 filters in the filtergraph.
19307 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
19308 must be inserted between two video filters, @code{azmq} between two
19311 To enable these filters you need to install the libzmq library and
19312 headers and configure FFmpeg with @code{--enable-libzmq}.
19314 For more information about libzmq see:
19315 @url{http://www.zeromq.org/}
19317 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
19318 receives messages sent through a network interface defined by the
19319 @option{bind_address} option.
19321 The received message must be in the form:
19323 @var{TARGET} @var{COMMAND} [@var{ARG}]
19326 @var{TARGET} specifies the target of the command, usually the name of
19327 the filter class or a specific filter instance name.
19329 @var{COMMAND} specifies the name of the command for the target filter.
19331 @var{ARG} is optional and specifies the optional argument list for the
19332 given @var{COMMAND}.
19334 Upon reception, the message is processed and the corresponding command
19335 is injected into the filtergraph. Depending on the result, the filter
19336 will send a reply to the client, adopting the format:
19338 @var{ERROR_CODE} @var{ERROR_REASON}
19342 @var{MESSAGE} is optional.
19344 @subsection Examples
19346 Look at @file{tools/zmqsend} for an example of a zmq client which can
19347 be used to send commands processed by these filters.
19349 Consider the following filtergraph generated by @command{ffplay}
19351 ffplay -dumpgraph 1 -f lavfi "
19352 color=s=100x100:c=red [l];
19353 color=s=100x100:c=blue [r];
19354 nullsrc=s=200x100, zmq [bg];
19355 [bg][l] overlay [bg+l];
19356 [bg+l][r] overlay=x=100 "
19359 To change the color of the left side of the video, the following
19360 command can be used:
19362 echo Parsed_color_0 c yellow | tools/zmqsend
19365 To change the right side:
19367 echo Parsed_color_1 c pink | tools/zmqsend
19370 @c man end MULTIMEDIA FILTERS
19372 @chapter Multimedia Sources
19373 @c man begin MULTIMEDIA SOURCES
19375 Below is a description of the currently available multimedia sources.
19379 This is the same as @ref{movie} source, except it selects an audio
19385 Read audio and/or video stream(s) from a movie container.
19387 It accepts the following parameters:
19391 The name of the resource to read (not necessarily a file; it can also be a
19392 device or a stream accessed through some protocol).
19394 @item format_name, f
19395 Specifies the format assumed for the movie to read, and can be either
19396 the name of a container or an input device. If not specified, the
19397 format is guessed from @var{movie_name} or by probing.
19399 @item seek_point, sp
19400 Specifies the seek point in seconds. The frames will be output
19401 starting from this seek point. The parameter is evaluated with
19402 @code{av_strtod}, so the numerical value may be suffixed by an IS
19403 postfix. The default value is "0".
19406 Specifies the streams to read. Several streams can be specified,
19407 separated by "+". The source will then have as many outputs, in the
19408 same order. The syntax is explained in the ``Stream specifiers''
19409 section in the ffmpeg manual. Two special names, "dv" and "da" specify
19410 respectively the default (best suited) video and audio stream. Default
19411 is "dv", or "da" if the filter is called as "amovie".
19413 @item stream_index, si
19414 Specifies the index of the video stream to read. If the value is -1,
19415 the most suitable video stream will be automatically selected. The default
19416 value is "-1". Deprecated. If the filter is called "amovie", it will select
19417 audio instead of video.
19420 Specifies how many times to read the stream in sequence.
19421 If the value is 0, the stream will be looped infinitely.
19422 Default value is "1".
19424 Note that when the movie is looped the source timestamps are not
19425 changed, so it will generate non monotonically increasing timestamps.
19427 @item discontinuity
19428 Specifies the time difference between frames above which the point is
19429 considered a timestamp discontinuity which is removed by adjusting the later
19433 It allows overlaying a second video on top of the main input of
19434 a filtergraph, as shown in this graph:
19436 input -----------> deltapts0 --> overlay --> output
19439 movie --> scale--> deltapts1 -------+
19441 @subsection Examples
19445 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
19446 on top of the input labelled "in":
19448 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
19449 [in] setpts=PTS-STARTPTS [main];
19450 [main][over] overlay=16:16 [out]
19454 Read from a video4linux2 device, and overlay it on top of the input
19457 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
19458 [in] setpts=PTS-STARTPTS [main];
19459 [main][over] overlay=16:16 [out]
19463 Read the first video stream and the audio stream with id 0x81 from
19464 dvd.vob; the video is connected to the pad named "video" and the audio is
19465 connected to the pad named "audio":
19467 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
19471 @subsection Commands
19473 Both movie and amovie support the following commands:
19476 Perform seek using "av_seek_frame".
19477 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
19480 @var{stream_index}: If stream_index is -1, a default
19481 stream is selected, and @var{timestamp} is automatically converted
19482 from AV_TIME_BASE units to the stream specific time_base.
19484 @var{timestamp}: Timestamp in AVStream.time_base units
19485 or, if no stream is specified, in AV_TIME_BASE units.
19487 @var{flags}: Flags which select direction and seeking mode.
19491 Get movie duration in AV_TIME_BASE units.
19495 @c man end MULTIMEDIA SOURCES