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 -c:a 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 versions of the flite library prior to 2.0 are 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.festvox.org/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). Value will be rounded to nearest
7369 integer. Minimum value is "1".
7370 Drop-frame timecode is supported for frame rates 30 & 60.
7373 If set to 1, the output of the timecode option will wrap around at 24 hours.
7374 Default is 0 (disabled).
7377 The text string to be drawn. The text must be a sequence of UTF-8
7379 This parameter is mandatory if no file is specified with the parameter
7383 A text file containing text to be drawn. The text must be a sequence
7384 of UTF-8 encoded characters.
7386 This parameter is mandatory if no text string is specified with the
7387 parameter @var{text}.
7389 If both @var{text} and @var{textfile} are specified, an error is thrown.
7392 If set to 1, the @var{textfile} will be reloaded before each frame.
7393 Be sure to update it atomically, or it may be read partially, or even fail.
7397 The expressions which specify the offsets where text will be drawn
7398 within the video frame. They are relative to the top/left border of the
7401 The default value of @var{x} and @var{y} is "0".
7403 See below for the list of accepted constants and functions.
7406 The parameters for @var{x} and @var{y} are expressions containing the
7407 following constants and functions:
7411 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7415 horizontal and vertical chroma subsample values. For example for the
7416 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7419 the height of each text line
7427 @item max_glyph_a, ascent
7428 the maximum distance from the baseline to the highest/upper grid
7429 coordinate used to place a glyph outline point, for all the rendered
7431 It is a positive value, due to the grid's orientation with the Y axis
7434 @item max_glyph_d, descent
7435 the maximum distance from the baseline to the lowest grid coordinate
7436 used to place a glyph outline point, for all the rendered glyphs.
7437 This is a negative value, due to the grid's orientation, with the Y axis
7441 maximum glyph height, that is the maximum height for all the glyphs
7442 contained in the rendered text, it is equivalent to @var{ascent} -
7446 maximum glyph width, that is the maximum width for all the glyphs
7447 contained in the rendered text
7450 the number of input frame, starting from 0
7452 @item rand(min, max)
7453 return a random number included between @var{min} and @var{max}
7456 The input sample aspect ratio.
7459 timestamp expressed in seconds, NAN if the input timestamp is unknown
7462 the height of the rendered text
7465 the width of the rendered text
7469 the x and y offset coordinates where the text is drawn.
7471 These parameters allow the @var{x} and @var{y} expressions to refer
7472 each other, so you can for example specify @code{y=x/dar}.
7475 @anchor{drawtext_expansion}
7476 @subsection Text expansion
7478 If @option{expansion} is set to @code{strftime},
7479 the filter recognizes strftime() sequences in the provided text and
7480 expands them accordingly. Check the documentation of strftime(). This
7481 feature is deprecated.
7483 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7485 If @option{expansion} is set to @code{normal} (which is the default),
7486 the following expansion mechanism is used.
7488 The backslash character @samp{\}, followed by any character, always expands to
7489 the second character.
7491 Sequences of the form @code{%@{...@}} are expanded. The text between the
7492 braces is a function name, possibly followed by arguments separated by ':'.
7493 If the arguments contain special characters or delimiters (':' or '@}'),
7494 they should be escaped.
7496 Note that they probably must also be escaped as the value for the
7497 @option{text} option in the filter argument string and as the filter
7498 argument in the filtergraph description, and possibly also for the shell,
7499 that makes up to four levels of escaping; using a text file avoids these
7502 The following functions are available:
7507 The expression evaluation result.
7509 It must take one argument specifying the expression to be evaluated,
7510 which accepts the same constants and functions as the @var{x} and
7511 @var{y} values. Note that not all constants should be used, for
7512 example the text size is not known when evaluating the expression, so
7513 the constants @var{text_w} and @var{text_h} will have an undefined
7516 @item expr_int_format, eif
7517 Evaluate the expression's value and output as formatted integer.
7519 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7520 The second argument specifies the output format. Allowed values are @samp{x},
7521 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7522 @code{printf} function.
7523 The third parameter is optional and sets the number of positions taken by the output.
7524 It can be used to add padding with zeros from the left.
7527 The time at which the filter is running, expressed in UTC.
7528 It can accept an argument: a strftime() format string.
7531 The time at which the filter is running, expressed in the local time zone.
7532 It can accept an argument: a strftime() format string.
7535 Frame metadata. Takes one or two arguments.
7537 The first argument is mandatory and specifies the metadata key.
7539 The second argument is optional and specifies a default value, used when the
7540 metadata key is not found or empty.
7543 The frame number, starting from 0.
7546 A 1 character description of the current picture type.
7549 The timestamp of the current frame.
7550 It can take up to three arguments.
7552 The first argument is the format of the timestamp; it defaults to @code{flt}
7553 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7554 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7555 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7556 @code{localtime} stands for the timestamp of the frame formatted as
7557 local time zone time.
7559 The second argument is an offset added to the timestamp.
7561 If the format is set to @code{localtime} or @code{gmtime},
7562 a third argument may be supplied: a strftime() format string.
7563 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7566 @subsection Examples
7570 Draw "Test Text" with font FreeSerif, using the default values for the
7571 optional parameters.
7574 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7578 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7579 and y=50 (counting from the top-left corner of the screen), text is
7580 yellow with a red box around it. Both the text and the box have an
7584 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7585 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7588 Note that the double quotes are not necessary if spaces are not used
7589 within the parameter list.
7592 Show the text at the center of the video frame:
7594 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7598 Show the text at a random position, switching to a new position every 30 seconds:
7600 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)"
7604 Show a text line sliding from right to left in the last row of the video
7605 frame. The file @file{LONG_LINE} is assumed to contain a single line
7608 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7612 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7614 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7618 Draw a single green letter "g", at the center of the input video.
7619 The glyph baseline is placed at half screen height.
7621 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7625 Show text for 1 second every 3 seconds:
7627 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7631 Use fontconfig to set the font. Note that the colons need to be escaped.
7633 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7637 Print the date of a real-time encoding (see strftime(3)):
7639 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7643 Show text fading in and out (appearing/disappearing):
7646 DS=1.0 # display start
7647 DE=10.0 # display end
7648 FID=1.5 # fade in duration
7649 FOD=5 # fade out duration
7650 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 @}"
7654 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
7655 and the @option{fontsize} value are included in the @option{y} offset.
7657 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
7658 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
7663 For more information about libfreetype, check:
7664 @url{http://www.freetype.org/}.
7666 For more information about fontconfig, check:
7667 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7669 For more information about libfribidi, check:
7670 @url{http://fribidi.org/}.
7674 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7676 The filter accepts the following options:
7681 Set low and high threshold values used by the Canny thresholding
7684 The high threshold selects the "strong" edge pixels, which are then
7685 connected through 8-connectivity with the "weak" edge pixels selected
7686 by the low threshold.
7688 @var{low} and @var{high} threshold values must be chosen in the range
7689 [0,1], and @var{low} should be lesser or equal to @var{high}.
7691 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7695 Define the drawing mode.
7699 Draw white/gray wires on black background.
7702 Mix the colors to create a paint/cartoon effect.
7705 Default value is @var{wires}.
7708 @subsection Examples
7712 Standard edge detection with custom values for the hysteresis thresholding:
7714 edgedetect=low=0.1:high=0.4
7718 Painting effect without thresholding:
7720 edgedetect=mode=colormix:high=0
7725 Set brightness, contrast, saturation and approximate gamma adjustment.
7727 The filter accepts the following options:
7731 Set the contrast expression. The value must be a float value in range
7732 @code{-2.0} to @code{2.0}. The default value is "1".
7735 Set the brightness expression. The value must be a float value in
7736 range @code{-1.0} to @code{1.0}. The default value is "0".
7739 Set the saturation expression. The value must be a float in
7740 range @code{0.0} to @code{3.0}. The default value is "1".
7743 Set the gamma expression. The value must be a float in range
7744 @code{0.1} to @code{10.0}. The default value is "1".
7747 Set the gamma expression for red. The value must be a float in
7748 range @code{0.1} to @code{10.0}. The default value is "1".
7751 Set the gamma expression for green. The value must be a float in range
7752 @code{0.1} to @code{10.0}. The default value is "1".
7755 Set the gamma expression for blue. The value must be a float in range
7756 @code{0.1} to @code{10.0}. The default value is "1".
7759 Set the gamma weight expression. It can be used to reduce the effect
7760 of a high gamma value on bright image areas, e.g. keep them from
7761 getting overamplified and just plain white. The value must be a float
7762 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7763 gamma correction all the way down while @code{1.0} leaves it at its
7764 full strength. Default is "1".
7767 Set when the expressions for brightness, contrast, saturation and
7768 gamma expressions are evaluated.
7770 It accepts the following values:
7773 only evaluate expressions once during the filter initialization or
7774 when a command is processed
7777 evaluate expressions for each incoming frame
7780 Default value is @samp{init}.
7783 The expressions accept the following parameters:
7786 frame count of the input frame starting from 0
7789 byte position of the corresponding packet in the input file, NAN if
7793 frame rate of the input video, NAN if the input frame rate is unknown
7796 timestamp expressed in seconds, NAN if the input timestamp is unknown
7799 @subsection Commands
7800 The filter supports the following commands:
7804 Set the contrast expression.
7807 Set the brightness expression.
7810 Set the saturation expression.
7813 Set the gamma expression.
7816 Set the gamma_r expression.
7819 Set gamma_g expression.
7822 Set gamma_b expression.
7825 Set gamma_weight expression.
7827 The command accepts the same syntax of the corresponding option.
7829 If the specified expression is not valid, it is kept at its current
7836 Apply erosion effect to the video.
7838 This filter replaces the pixel by the local(3x3) minimum.
7840 It accepts the following options:
7847 Limit the maximum change for each plane, default is 65535.
7848 If 0, plane will remain unchanged.
7851 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7854 Flags to local 3x3 coordinates maps like this:
7861 @section extractplanes
7863 Extract color channel components from input video stream into
7864 separate grayscale video streams.
7866 The filter accepts the following option:
7870 Set plane(s) to extract.
7872 Available values for planes are:
7883 Choosing planes not available in the input will result in an error.
7884 That means you cannot select @code{r}, @code{g}, @code{b} planes
7885 with @code{y}, @code{u}, @code{v} planes at same time.
7888 @subsection Examples
7892 Extract luma, u and v color channel component from input video frame
7893 into 3 grayscale outputs:
7895 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
7901 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7903 For each input image, the filter will compute the optimal mapping from
7904 the input to the output given the codebook length, that is the number
7905 of distinct output colors.
7907 This filter accepts the following options.
7910 @item codebook_length, l
7911 Set codebook length. The value must be a positive integer, and
7912 represents the number of distinct output colors. Default value is 256.
7915 Set the maximum number of iterations to apply for computing the optimal
7916 mapping. The higher the value the better the result and the higher the
7917 computation time. Default value is 1.
7920 Set a random seed, must be an integer included between 0 and
7921 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7922 will try to use a good random seed on a best effort basis.
7925 Set pal8 output pixel format. This option does not work with codebook
7926 length greater than 256.
7931 Apply a fade-in/out effect to the input video.
7933 It accepts the following parameters:
7937 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7939 Default is @code{in}.
7941 @item start_frame, s
7942 Specify the number of the frame to start applying the fade
7943 effect at. Default is 0.
7946 The number of frames that the fade effect lasts. At the end of the
7947 fade-in effect, the output video will have the same intensity as the input video.
7948 At the end of the fade-out transition, the output video will be filled with the
7949 selected @option{color}.
7953 If set to 1, fade only alpha channel, if one exists on the input.
7956 @item start_time, st
7957 Specify the timestamp (in seconds) of the frame to start to apply the fade
7958 effect. If both start_frame and start_time are specified, the fade will start at
7959 whichever comes last. Default is 0.
7962 The number of seconds for which the fade effect has to last. At the end of the
7963 fade-in effect the output video will have the same intensity as the input video,
7964 at the end of the fade-out transition the output video will be filled with the
7965 selected @option{color}.
7966 If both duration and nb_frames are specified, duration is used. Default is 0
7967 (nb_frames is used by default).
7970 Specify the color of the fade. Default is "black".
7973 @subsection Examples
7977 Fade in the first 30 frames of video:
7982 The command above is equivalent to:
7988 Fade out the last 45 frames of a 200-frame video:
7991 fade=type=out:start_frame=155:nb_frames=45
7995 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7997 fade=in:0:25, fade=out:975:25
8001 Make the first 5 frames yellow, then fade in from frame 5-24:
8003 fade=in:5:20:color=yellow
8007 Fade in alpha over first 25 frames of video:
8009 fade=in:0:25:alpha=1
8013 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8015 fade=t=in:st=5.5:d=0.5
8021 Apply arbitrary expressions to samples in frequency domain
8025 Adjust the dc value (gain) of the luma plane of the image. The filter
8026 accepts an integer value in range @code{0} to @code{1000}. The default
8027 value is set to @code{0}.
8030 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8031 filter accepts an integer value in range @code{0} to @code{1000}. The
8032 default value is set to @code{0}.
8035 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8036 filter accepts an integer value in range @code{0} to @code{1000}. The
8037 default value is set to @code{0}.
8040 Set the frequency domain weight expression for the luma plane.
8043 Set the frequency domain weight expression for the 1st chroma plane.
8046 Set the frequency domain weight expression for the 2nd chroma plane.
8049 Set when the expressions are evaluated.
8051 It accepts the following values:
8054 Only evaluate expressions once during the filter initialization.
8057 Evaluate expressions for each incoming frame.
8060 Default value is @samp{init}.
8062 The filter accepts the following variables:
8065 The coordinates of the current sample.
8069 The width and height of the image.
8072 The number of input frame, starting from 0.
8075 @subsection Examples
8081 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8087 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8093 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8099 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8106 Extract a single field from an interlaced image using stride
8107 arithmetic to avoid wasting CPU time. The output frames are marked as
8110 The filter accepts the following options:
8114 Specify whether to extract the top (if the value is @code{0} or
8115 @code{top}) or the bottom field (if the value is @code{1} or
8121 Create new frames by copying the top and bottom fields from surrounding frames
8122 supplied as numbers by the hint file.
8126 Set file containing hints: absolute/relative frame numbers.
8128 There must be one line for each frame in a clip. Each line must contain two
8129 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8130 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8131 is current frame number for @code{absolute} mode or out of [-1, 1] range
8132 for @code{relative} mode. First number tells from which frame to pick up top
8133 field and second number tells from which frame to pick up bottom field.
8135 If optionally followed by @code{+} output frame will be marked as interlaced,
8136 else if followed by @code{-} output frame will be marked as progressive, else
8137 it will be marked same as input frame.
8138 If line starts with @code{#} or @code{;} that line is skipped.
8141 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8144 Example of first several lines of @code{hint} file for @code{relative} mode:
8147 1,0 - # second frame, use third's frame top field and second's frame bottom field
8148 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8165 Field matching filter for inverse telecine. It is meant to reconstruct the
8166 progressive frames from a telecined stream. The filter does not drop duplicated
8167 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8168 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8170 The separation of the field matching and the decimation is notably motivated by
8171 the possibility of inserting a de-interlacing filter fallback between the two.
8172 If the source has mixed telecined and real interlaced content,
8173 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8174 But these remaining combed frames will be marked as interlaced, and thus can be
8175 de-interlaced by a later filter such as @ref{yadif} before decimation.
8177 In addition to the various configuration options, @code{fieldmatch} can take an
8178 optional second stream, activated through the @option{ppsrc} option. If
8179 enabled, the frames reconstruction will be based on the fields and frames from
8180 this second stream. This allows the first input to be pre-processed in order to
8181 help the various algorithms of the filter, while keeping the output lossless
8182 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8183 or brightness/contrast adjustments can help.
8185 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8186 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8187 which @code{fieldmatch} is based on. While the semantic and usage are very
8188 close, some behaviour and options names can differ.
8190 The @ref{decimate} filter currently only works for constant frame rate input.
8191 If your input has mixed telecined (30fps) and progressive content with a lower
8192 framerate like 24fps use the following filterchain to produce the necessary cfr
8193 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8195 The filter accepts the following options:
8199 Specify the assumed field order of the input stream. Available values are:
8203 Auto detect parity (use FFmpeg's internal parity value).
8205 Assume bottom field first.
8207 Assume top field first.
8210 Note that it is sometimes recommended not to trust the parity announced by the
8213 Default value is @var{auto}.
8216 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8217 sense that it won't risk creating jerkiness due to duplicate frames when
8218 possible, but if there are bad edits or blended fields it will end up
8219 outputting combed frames when a good match might actually exist. On the other
8220 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8221 but will almost always find a good frame if there is one. The other values are
8222 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8223 jerkiness and creating duplicate frames versus finding good matches in sections
8224 with bad edits, orphaned fields, blended fields, etc.
8226 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8228 Available values are:
8232 2-way matching (p/c)
8234 2-way matching, and trying 3rd match if still combed (p/c + n)
8236 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8238 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8239 still combed (p/c + n + u/b)
8241 3-way matching (p/c/n)
8243 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8244 detected as combed (p/c/n + u/b)
8247 The parenthesis at the end indicate the matches that would be used for that
8248 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8251 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8254 Default value is @var{pc_n}.
8257 Mark the main input stream as a pre-processed input, and enable the secondary
8258 input stream as the clean source to pick the fields from. See the filter
8259 introduction for more details. It is similar to the @option{clip2} feature from
8262 Default value is @code{0} (disabled).
8265 Set the field to match from. It is recommended to set this to the same value as
8266 @option{order} unless you experience matching failures with that setting. In
8267 certain circumstances changing the field that is used to match from can have a
8268 large impact on matching performance. Available values are:
8272 Automatic (same value as @option{order}).
8274 Match from the bottom field.
8276 Match from the top field.
8279 Default value is @var{auto}.
8282 Set whether or not chroma is included during the match comparisons. In most
8283 cases it is recommended to leave this enabled. You should set this to @code{0}
8284 only if your clip has bad chroma problems such as heavy rainbowing or other
8285 artifacts. Setting this to @code{0} could also be used to speed things up at
8286 the cost of some accuracy.
8288 Default value is @code{1}.
8292 These define an exclusion band which excludes the lines between @option{y0} and
8293 @option{y1} from being included in the field matching decision. An exclusion
8294 band can be used to ignore subtitles, a logo, or other things that may
8295 interfere with the matching. @option{y0} sets the starting scan line and
8296 @option{y1} sets the ending line; all lines in between @option{y0} and
8297 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8298 @option{y0} and @option{y1} to the same value will disable the feature.
8299 @option{y0} and @option{y1} defaults to @code{0}.
8302 Set the scene change detection threshold as a percentage of maximum change on
8303 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8304 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8305 @option{scthresh} is @code{[0.0, 100.0]}.
8307 Default value is @code{12.0}.
8310 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8311 account the combed scores of matches when deciding what match to use as the
8312 final match. Available values are:
8316 No final matching based on combed scores.
8318 Combed scores are only used when a scene change is detected.
8320 Use combed scores all the time.
8323 Default is @var{sc}.
8326 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8327 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8328 Available values are:
8332 No forced calculation.
8334 Force p/c/n calculations.
8336 Force p/c/n/u/b calculations.
8339 Default value is @var{none}.
8342 This is the area combing threshold used for combed frame detection. This
8343 essentially controls how "strong" or "visible" combing must be to be detected.
8344 Larger values mean combing must be more visible and smaller values mean combing
8345 can be less visible or strong and still be detected. Valid settings are from
8346 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8347 be detected as combed). This is basically a pixel difference value. A good
8348 range is @code{[8, 12]}.
8350 Default value is @code{9}.
8353 Sets whether or not chroma is considered in the combed frame decision. Only
8354 disable this if your source has chroma problems (rainbowing, etc.) that are
8355 causing problems for the combed frame detection with chroma enabled. Actually,
8356 using @option{chroma}=@var{0} is usually more reliable, except for the case
8357 where there is chroma only combing in the source.
8359 Default value is @code{0}.
8363 Respectively set the x-axis and y-axis size of the window used during combed
8364 frame detection. This has to do with the size of the area in which
8365 @option{combpel} pixels are required to be detected as combed for a frame to be
8366 declared combed. See the @option{combpel} parameter description for more info.
8367 Possible values are any number that is a power of 2 starting at 4 and going up
8370 Default value is @code{16}.
8373 The number of combed pixels inside any of the @option{blocky} by
8374 @option{blockx} size blocks on the frame for the frame to be detected as
8375 combed. While @option{cthresh} controls how "visible" the combing must be, this
8376 setting controls "how much" combing there must be in any localized area (a
8377 window defined by the @option{blockx} and @option{blocky} settings) on the
8378 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8379 which point no frames will ever be detected as combed). This setting is known
8380 as @option{MI} in TFM/VFM vocabulary.
8382 Default value is @code{80}.
8385 @anchor{p/c/n/u/b meaning}
8386 @subsection p/c/n/u/b meaning
8388 @subsubsection p/c/n
8390 We assume the following telecined stream:
8393 Top fields: 1 2 2 3 4
8394 Bottom fields: 1 2 3 4 4
8397 The numbers correspond to the progressive frame the fields relate to. Here, the
8398 first two frames are progressive, the 3rd and 4th are combed, and so on.
8400 When @code{fieldmatch} is configured to run a matching from bottom
8401 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8406 B 1 2 3 4 4 <-- matching reference
8415 As a result of the field matching, we can see that some frames get duplicated.
8416 To perform a complete inverse telecine, you need to rely on a decimation filter
8417 after this operation. See for instance the @ref{decimate} filter.
8419 The same operation now matching from top fields (@option{field}=@var{top})
8424 T 1 2 2 3 4 <-- matching reference
8434 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8435 basically, they refer to the frame and field of the opposite parity:
8438 @item @var{p} matches the field of the opposite parity in the previous frame
8439 @item @var{c} matches the field of the opposite parity in the current frame
8440 @item @var{n} matches the field of the opposite parity in the next frame
8445 The @var{u} and @var{b} matching are a bit special in the sense that they match
8446 from the opposite parity flag. In the following examples, we assume that we are
8447 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8448 'x' is placed above and below each matched fields.
8450 With bottom matching (@option{field}=@var{bottom}):
8455 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8456 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8464 With top matching (@option{field}=@var{top}):
8469 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8470 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8478 @subsection Examples
8480 Simple IVTC of a top field first telecined stream:
8482 fieldmatch=order=tff:combmatch=none, decimate
8485 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8487 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8492 Transform the field order of the input video.
8494 It accepts the following parameters:
8499 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8500 for bottom field first.
8503 The default value is @samp{tff}.
8505 The transformation is done by shifting the picture content up or down
8506 by one line, and filling the remaining line with appropriate picture content.
8507 This method is consistent with most broadcast field order converters.
8509 If the input video is not flagged as being interlaced, or it is already
8510 flagged as being of the required output field order, then this filter does
8511 not alter the incoming video.
8513 It is very useful when converting to or from PAL DV material,
8514 which is bottom field first.
8518 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8521 @section fifo, afifo
8523 Buffer input images and send them when they are requested.
8525 It is mainly useful when auto-inserted by the libavfilter
8528 It does not take parameters.
8532 Find a rectangular object
8534 It accepts the following options:
8538 Filepath of the object image, needs to be in gray8.
8541 Detection threshold, default is 0.5.
8544 Number of mipmaps, default is 3.
8546 @item xmin, ymin, xmax, ymax
8547 Specifies the rectangle in which to search.
8550 @subsection Examples
8554 Generate a representative palette of a given video using @command{ffmpeg}:
8556 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8562 Cover a rectangular object
8564 It accepts the following options:
8568 Filepath of the optional cover image, needs to be in yuv420.
8573 It accepts the following values:
8576 cover it by the supplied image
8578 cover it by interpolating the surrounding pixels
8581 Default value is @var{blur}.
8584 @subsection Examples
8588 Generate a representative palette of a given video using @command{ffmpeg}:
8590 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8596 Flood area with values of same pixel components with another values.
8598 It accepts the following options:
8601 Set pixel x coordinate.
8604 Set pixel y coordinate.
8607 Set source #0 component value.
8610 Set source #1 component value.
8613 Set source #2 component value.
8616 Set source #3 component value.
8619 Set destination #0 component value.
8622 Set destination #1 component value.
8625 Set destination #2 component value.
8628 Set destination #3 component value.
8634 Convert the input video to one of the specified pixel formats.
8635 Libavfilter will try to pick one that is suitable as input to
8638 It accepts the following parameters:
8642 A '|'-separated list of pixel format names, such as
8643 "pix_fmts=yuv420p|monow|rgb24".
8647 @subsection Examples
8651 Convert the input video to the @var{yuv420p} format
8653 format=pix_fmts=yuv420p
8656 Convert the input video to any of the formats in the list
8658 format=pix_fmts=yuv420p|yuv444p|yuv410p
8665 Convert the video to specified constant frame rate by duplicating or dropping
8666 frames as necessary.
8668 It accepts the following parameters:
8672 The desired output frame rate. The default is @code{25}.
8675 Assume the first PTS should be the given value, in seconds. This allows for
8676 padding/trimming at the start of stream. By default, no assumption is made
8677 about the first frame's expected PTS, so no padding or trimming is done.
8678 For example, this could be set to 0 to pad the beginning with duplicates of
8679 the first frame if a video stream starts after the audio stream or to trim any
8680 frames with a negative PTS.
8683 Timestamp (PTS) rounding method.
8685 Possible values are:
8692 round towards -infinity
8694 round towards +infinity
8698 The default is @code{near}.
8701 Action performed when reading the last frame.
8703 Possible values are:
8706 Use same timestamp rounding method as used for other frames.
8708 Pass through last frame if input duration has not been reached yet.
8710 The default is @code{round}.
8714 Alternatively, the options can be specified as a flat string:
8715 @var{fps}[:@var{start_time}[:@var{round}]].
8717 See also the @ref{setpts} filter.
8719 @subsection Examples
8723 A typical usage in order to set the fps to 25:
8729 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8731 fps=fps=film:round=near
8737 Pack two different video streams into a stereoscopic video, setting proper
8738 metadata on supported codecs. The two views should have the same size and
8739 framerate and processing will stop when the shorter video ends. Please note
8740 that you may conveniently adjust view properties with the @ref{scale} and
8743 It accepts the following parameters:
8747 The desired packing format. Supported values are:
8752 The views are next to each other (default).
8755 The views are on top of each other.
8758 The views are packed by line.
8761 The views are packed by column.
8764 The views are temporally interleaved.
8773 # Convert left and right views into a frame-sequential video
8774 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8776 # Convert views into a side-by-side video with the same output resolution as the input
8777 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
8782 Change the frame rate by interpolating new video output frames from the source
8785 This filter is not designed to function correctly with interlaced media. If
8786 you wish to change the frame rate of interlaced media then you are required
8787 to deinterlace before this filter and re-interlace after this filter.
8789 A description of the accepted options follows.
8793 Specify the output frames per second. This option can also be specified
8794 as a value alone. The default is @code{50}.
8797 Specify the start of a range where the output frame will be created as a
8798 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8799 the default is @code{15}.
8802 Specify the end of a range where the output frame will be created as a
8803 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8804 the default is @code{240}.
8807 Specify the level at which a scene change is detected as a value between
8808 0 and 100 to indicate a new scene; a low value reflects a low
8809 probability for the current frame to introduce a new scene, while a higher
8810 value means the current frame is more likely to be one.
8811 The default is @code{7}.
8814 Specify flags influencing the filter process.
8816 Available value for @var{flags} is:
8819 @item scene_change_detect, scd
8820 Enable scene change detection using the value of the option @var{scene}.
8821 This flag is enabled by default.
8827 Select one frame every N-th frame.
8829 This filter accepts the following option:
8832 Select frame after every @code{step} frames.
8833 Allowed values are positive integers higher than 0. Default value is @code{1}.
8839 Apply a frei0r effect to the input video.
8841 To enable the compilation of this filter, you need to install the frei0r
8842 header and configure FFmpeg with @code{--enable-frei0r}.
8844 It accepts the following parameters:
8849 The name of the frei0r effect to load. If the environment variable
8850 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8851 directories specified by the colon-separated list in @env{FREI0R_PATH}.
8852 Otherwise, the standard frei0r paths are searched, in this order:
8853 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8854 @file{/usr/lib/frei0r-1/}.
8857 A '|'-separated list of parameters to pass to the frei0r effect.
8861 A frei0r effect parameter can be a boolean (its value is either
8862 "y" or "n"), a double, a color (specified as
8863 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8864 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8865 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8866 @var{X} and @var{Y} are floating point numbers) and/or a string.
8868 The number and types of parameters depend on the loaded effect. If an
8869 effect parameter is not specified, the default value is set.
8871 @subsection Examples
8875 Apply the distort0r effect, setting the first two double parameters:
8877 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8881 Apply the colordistance effect, taking a color as the first parameter:
8883 frei0r=colordistance:0.2/0.3/0.4
8884 frei0r=colordistance:violet
8885 frei0r=colordistance:0x112233
8889 Apply the perspective effect, specifying the top left and top right image
8892 frei0r=perspective:0.2/0.2|0.8/0.2
8896 For more information, see
8897 @url{http://frei0r.dyne.org}
8901 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8903 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8904 processing filter, one of them is performed once per block, not per pixel.
8905 This allows for much higher speed.
8907 The filter accepts the following options:
8911 Set quality. This option defines the number of levels for averaging. It accepts
8912 an integer in the range 4-5. Default value is @code{4}.
8915 Force a constant quantization parameter. It accepts an integer in range 0-63.
8916 If not set, the filter will use the QP from the video stream (if available).
8919 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8920 more details but also more artifacts, while higher values make the image smoother
8921 but also blurrier. Default value is @code{0} − PSNR optimal.
8924 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8925 option may cause flicker since the B-Frames have often larger QP. Default is
8926 @code{0} (not enabled).
8932 Apply Gaussian blur filter.
8934 The filter accepts the following options:
8938 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8941 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8944 Set which planes to filter. By default all planes are filtered.
8947 Set vertical sigma, if negative it will be same as @code{sigma}.
8948 Default is @code{-1}.
8953 The filter accepts the following options:
8957 Set the luminance expression.
8959 Set the chrominance blue expression.
8961 Set the chrominance red expression.
8963 Set the alpha expression.
8965 Set the red expression.
8967 Set the green expression.
8969 Set the blue expression.
8972 The colorspace is selected according to the specified options. If one
8973 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8974 options is specified, the filter will automatically select a YCbCr
8975 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8976 @option{blue_expr} options is specified, it will select an RGB
8979 If one of the chrominance expression is not defined, it falls back on the other
8980 one. If no alpha expression is specified it will evaluate to opaque value.
8981 If none of chrominance expressions are specified, they will evaluate
8982 to the luminance expression.
8984 The expressions can use the following variables and functions:
8988 The sequential number of the filtered frame, starting from @code{0}.
8992 The coordinates of the current sample.
8996 The width and height of the image.
9000 Width and height scale depending on the currently filtered plane. It is the
9001 ratio between the corresponding luma plane number of pixels and the current
9002 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9003 @code{0.5,0.5} for chroma planes.
9006 Time of the current frame, expressed in seconds.
9009 Return the value of the pixel at location (@var{x},@var{y}) of the current
9013 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9017 Return the value of the pixel at location (@var{x},@var{y}) of the
9018 blue-difference chroma plane. Return 0 if there is no such plane.
9021 Return the value of the pixel at location (@var{x},@var{y}) of the
9022 red-difference chroma plane. Return 0 if there is no such plane.
9027 Return the value of the pixel at location (@var{x},@var{y}) of the
9028 red/green/blue component. Return 0 if there is no such component.
9031 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9032 plane. Return 0 if there is no such plane.
9035 For functions, if @var{x} and @var{y} are outside the area, the value will be
9036 automatically clipped to the closer edge.
9038 @subsection Examples
9042 Flip the image horizontally:
9048 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9049 wavelength of 100 pixels:
9051 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9055 Generate a fancy enigmatic moving light:
9057 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
9061 Generate a quick emboss effect:
9063 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9067 Modify RGB components depending on pixel position:
9069 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9073 Create a radial gradient that is the same size as the input (also see
9074 the @ref{vignette} filter):
9076 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9082 Fix the banding artifacts that are sometimes introduced into nearly flat
9083 regions by truncation to 8-bit color depth.
9084 Interpolate the gradients that should go where the bands are, and
9087 It is designed for playback only. Do not use it prior to
9088 lossy compression, because compression tends to lose the dither and
9089 bring back the bands.
9091 It accepts the following parameters:
9096 The maximum amount by which the filter will change any one pixel. This is also
9097 the threshold for detecting nearly flat regions. Acceptable values range from
9098 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9102 The neighborhood to fit the gradient to. A larger radius makes for smoother
9103 gradients, but also prevents the filter from modifying the pixels near detailed
9104 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9105 values will be clipped to the valid range.
9109 Alternatively, the options can be specified as a flat string:
9110 @var{strength}[:@var{radius}]
9112 @subsection Examples
9116 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9122 Specify radius, omitting the strength (which will fall-back to the default
9133 Apply a Hald CLUT to a video stream.
9135 First input is the video stream to process, and second one is the Hald CLUT.
9136 The Hald CLUT input can be a simple picture or a complete video stream.
9138 The filter accepts the following options:
9142 Force termination when the shortest input terminates. Default is @code{0}.
9144 Continue applying the last CLUT after the end of the stream. A value of
9145 @code{0} disable the filter after the last frame of the CLUT is reached.
9146 Default is @code{1}.
9149 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9150 filters share the same internals).
9152 More information about the Hald CLUT can be found on Eskil Steenberg's website
9153 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9155 @subsection Workflow examples
9157 @subsubsection Hald CLUT video stream
9159 Generate an identity Hald CLUT stream altered with various effects:
9161 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
9164 Note: make sure you use a lossless codec.
9166 Then use it with @code{haldclut} to apply it on some random stream:
9168 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9171 The Hald CLUT will be applied to the 10 first seconds (duration of
9172 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9173 to the remaining frames of the @code{mandelbrot} stream.
9175 @subsubsection Hald CLUT with preview
9177 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9178 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9179 biggest possible square starting at the top left of the picture. The remaining
9180 padding pixels (bottom or right) will be ignored. This area can be used to add
9181 a preview of the Hald CLUT.
9183 Typically, the following generated Hald CLUT will be supported by the
9184 @code{haldclut} filter:
9187 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9188 pad=iw+320 [padded_clut];
9189 smptebars=s=320x256, split [a][b];
9190 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9191 [main][b] overlay=W-320" -frames:v 1 clut.png
9194 It contains the original and a preview of the effect of the CLUT: SMPTE color
9195 bars are displayed on the right-top, and below the same color bars processed by
9198 Then, the effect of this Hald CLUT can be visualized with:
9200 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9205 Flip the input video horizontally.
9207 For example, to horizontally flip the input video with @command{ffmpeg}:
9209 ffmpeg -i in.avi -vf "hflip" out.avi
9213 This filter applies a global color histogram equalization on a
9216 It can be used to correct video that has a compressed range of pixel
9217 intensities. The filter redistributes the pixel intensities to
9218 equalize their distribution across the intensity range. It may be
9219 viewed as an "automatically adjusting contrast filter". This filter is
9220 useful only for correcting degraded or poorly captured source
9223 The filter accepts the following options:
9227 Determine the amount of equalization to be applied. As the strength
9228 is reduced, the distribution of pixel intensities more-and-more
9229 approaches that of the input frame. The value must be a float number
9230 in the range [0,1] and defaults to 0.200.
9233 Set the maximum intensity that can generated and scale the output
9234 values appropriately. The strength should be set as desired and then
9235 the intensity can be limited if needed to avoid washing-out. The value
9236 must be a float number in the range [0,1] and defaults to 0.210.
9239 Set the antibanding level. If enabled the filter will randomly vary
9240 the luminance of output pixels by a small amount to avoid banding of
9241 the histogram. Possible values are @code{none}, @code{weak} or
9242 @code{strong}. It defaults to @code{none}.
9247 Compute and draw a color distribution histogram for the input video.
9249 The computed histogram is a representation of the color component
9250 distribution in an image.
9252 Standard histogram displays the color components distribution in an image.
9253 Displays color graph for each color component. Shows distribution of
9254 the Y, U, V, A or R, G, B components, depending on input format, in the
9255 current frame. Below each graph a color component scale meter is shown.
9257 The filter accepts the following options:
9261 Set height of level. Default value is @code{200}.
9262 Allowed range is [50, 2048].
9265 Set height of color scale. Default value is @code{12}.
9266 Allowed range is [0, 40].
9270 It accepts the following values:
9273 Per color component graphs are placed below each other.
9276 Per color component graphs are placed side by side.
9279 Presents information identical to that in the @code{parade}, except
9280 that the graphs representing color components are superimposed directly
9283 Default is @code{stack}.
9286 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9287 Default is @code{linear}.
9290 Set what color components to display.
9291 Default is @code{7}.
9294 Set foreground opacity. Default is @code{0.7}.
9297 Set background opacity. Default is @code{0.5}.
9300 @subsection Examples
9305 Calculate and draw histogram:
9307 ffplay -i input -vf histogram
9315 This is a high precision/quality 3d denoise filter. It aims to reduce
9316 image noise, producing smooth images and making still images really
9317 still. It should enhance compressibility.
9319 It accepts the following optional parameters:
9323 A non-negative floating point number which specifies spatial luma strength.
9326 @item chroma_spatial
9327 A non-negative floating point number which specifies spatial chroma strength.
9328 It defaults to 3.0*@var{luma_spatial}/4.0.
9331 A floating point number which specifies luma temporal strength. It defaults to
9332 6.0*@var{luma_spatial}/4.0.
9335 A floating point number which specifies chroma temporal strength. It defaults to
9336 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9341 Download hardware frames to system memory.
9343 The input must be in hardware frames, and the output a non-hardware format.
9344 Not all formats will be supported on the output - it may be necessary to insert
9345 an additional @option{format} filter immediately following in the graph to get
9346 the output in a supported format.
9350 Map hardware frames to system memory or to another device.
9352 This filter has several different modes of operation; which one is used depends
9353 on the input and output formats:
9356 Hardware frame input, normal frame output
9358 Map the input frames to system memory and pass them to the output. If the
9359 original hardware frame is later required (for example, after overlaying
9360 something else on part of it), the @option{hwmap} filter can be used again
9361 in the next mode to retrieve it.
9363 Normal frame input, hardware frame output
9365 If the input is actually a software-mapped hardware frame, then unmap it -
9366 that is, return the original hardware frame.
9368 Otherwise, a device must be provided. Create new hardware surfaces on that
9369 device for the output, then map them back to the software format at the input
9370 and give those frames to the preceding filter. This will then act like the
9371 @option{hwupload} filter, but may be able to avoid an additional copy when
9372 the input is already in a compatible format.
9374 Hardware frame input and output
9376 A device must be supplied for the output, either directly or with the
9377 @option{derive_device} option. The input and output devices must be of
9378 different types and compatible - the exact meaning of this is
9379 system-dependent, but typically it means that they must refer to the same
9380 underlying hardware context (for example, refer to the same graphics card).
9382 If the input frames were originally created on the output device, then unmap
9383 to retrieve the original frames.
9385 Otherwise, map the frames to the output device - create new hardware frames
9386 on the output corresponding to the frames on the input.
9389 The following additional parameters are accepted:
9393 Set the frame mapping mode. Some combination of:
9396 The mapped frame should be readable.
9398 The mapped frame should be writeable.
9400 The mapping will always overwrite the entire frame.
9402 This may improve performance in some cases, as the original contents of the
9403 frame need not be loaded.
9405 The mapping must not involve any copying.
9407 Indirect mappings to copies of frames are created in some cases where either
9408 direct mapping is not possible or it would have unexpected properties.
9409 Setting this flag ensures that the mapping is direct and will fail if that is
9412 Defaults to @var{read+write} if not specified.
9414 @item derive_device @var{type}
9415 Rather than using the device supplied at initialisation, instead derive a new
9416 device of type @var{type} from the device the input frames exist on.
9419 In a hardware to hardware mapping, map in reverse - create frames in the sink
9420 and map them back to the source. This may be necessary in some cases where
9421 a mapping in one direction is required but only the opposite direction is
9422 supported by the devices being used.
9424 This option is dangerous - it may break the preceding filter in undefined
9425 ways if there are any additional constraints on that filter's output.
9426 Do not use it without fully understanding the implications of its use.
9431 Upload system memory frames to hardware surfaces.
9433 The device to upload to must be supplied when the filter is initialised. If
9434 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9437 @anchor{hwupload_cuda}
9438 @section hwupload_cuda
9440 Upload system memory frames to a CUDA device.
9442 It accepts the following optional parameters:
9446 The number of the CUDA device to use
9451 Apply a high-quality magnification filter designed for pixel art. This filter
9452 was originally created by Maxim Stepin.
9454 It accepts the following option:
9458 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9459 @code{hq3x} and @code{4} for @code{hq4x}.
9460 Default is @code{3}.
9464 Stack input videos horizontally.
9466 All streams must be of same pixel format and of same height.
9468 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9469 to create same output.
9471 The filter accept the following option:
9475 Set number of input streams. Default is 2.
9478 If set to 1, force the output to terminate when the shortest input
9479 terminates. Default value is 0.
9484 Modify the hue and/or the saturation of the input.
9486 It accepts the following parameters:
9490 Specify the hue angle as a number of degrees. It accepts an expression,
9491 and defaults to "0".
9494 Specify the saturation in the [-10,10] range. It accepts an expression and
9498 Specify the hue angle as a number of radians. It accepts an
9499 expression, and defaults to "0".
9502 Specify the brightness in the [-10,10] range. It accepts an expression and
9506 @option{h} and @option{H} are mutually exclusive, and can't be
9507 specified at the same time.
9509 The @option{b}, @option{h}, @option{H} and @option{s} option values are
9510 expressions containing the following constants:
9514 frame count of the input frame starting from 0
9517 presentation timestamp of the input frame expressed in time base units
9520 frame rate of the input video, NAN if the input frame rate is unknown
9523 timestamp expressed in seconds, NAN if the input timestamp is unknown
9526 time base of the input video
9529 @subsection Examples
9533 Set the hue to 90 degrees and the saturation to 1.0:
9539 Same command but expressing the hue in radians:
9545 Rotate hue and make the saturation swing between 0
9546 and 2 over a period of 1 second:
9548 hue="H=2*PI*t: s=sin(2*PI*t)+1"
9552 Apply a 3 seconds saturation fade-in effect starting at 0:
9557 The general fade-in expression can be written as:
9559 hue="s=min(0\, max((t-START)/DURATION\, 1))"
9563 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
9565 hue="s=max(0\, min(1\, (8-t)/3))"
9568 The general fade-out expression can be written as:
9570 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
9575 @subsection Commands
9577 This filter supports the following commands:
9583 Modify the hue and/or the saturation and/or brightness of the input video.
9584 The command accepts the same syntax of the corresponding option.
9586 If the specified expression is not valid, it is kept at its current
9592 Grow first stream into second stream by connecting components.
9593 This makes it possible to build more robust edge masks.
9595 This filter accepts the following options:
9599 Set which planes will be processed as bitmap, unprocessed planes will be
9600 copied from first stream.
9601 By default value 0xf, all planes will be processed.
9604 Set threshold which is used in filtering. If pixel component value is higher than
9605 this value filter algorithm for connecting components is activated.
9606 By default value is 0.
9611 Detect video interlacing type.
9613 This filter tries to detect if the input frames are interlaced, progressive,
9614 top or bottom field first. It will also try to detect fields that are
9615 repeated between adjacent frames (a sign of telecine).
9617 Single frame detection considers only immediately adjacent frames when classifying each frame.
9618 Multiple frame detection incorporates the classification history of previous frames.
9620 The filter will log these metadata values:
9623 @item single.current_frame
9624 Detected type of current frame using single-frame detection. One of:
9625 ``tff'' (top field first), ``bff'' (bottom field first),
9626 ``progressive'', or ``undetermined''
9629 Cumulative number of frames detected as top field first using single-frame detection.
9632 Cumulative number of frames detected as top field first using multiple-frame detection.
9635 Cumulative number of frames detected as bottom field first using single-frame detection.
9637 @item multiple.current_frame
9638 Detected type of current frame using multiple-frame detection. One of:
9639 ``tff'' (top field first), ``bff'' (bottom field first),
9640 ``progressive'', or ``undetermined''
9643 Cumulative number of frames detected as bottom field first using multiple-frame detection.
9645 @item single.progressive
9646 Cumulative number of frames detected as progressive using single-frame detection.
9648 @item multiple.progressive
9649 Cumulative number of frames detected as progressive using multiple-frame detection.
9651 @item single.undetermined
9652 Cumulative number of frames that could not be classified using single-frame detection.
9654 @item multiple.undetermined
9655 Cumulative number of frames that could not be classified using multiple-frame detection.
9657 @item repeated.current_frame
9658 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
9660 @item repeated.neither
9661 Cumulative number of frames with no repeated field.
9664 Cumulative number of frames with the top field repeated from the previous frame's top field.
9666 @item repeated.bottom
9667 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
9670 The filter accepts the following options:
9674 Set interlacing threshold.
9676 Set progressive threshold.
9678 Threshold for repeated field detection.
9680 Number of frames after which a given frame's contribution to the
9681 statistics is halved (i.e., it contributes only 0.5 to its
9682 classification). The default of 0 means that all frames seen are given
9683 full weight of 1.0 forever.
9684 @item analyze_interlaced_flag
9685 When this is not 0 then idet will use the specified number of frames to determine
9686 if the interlaced flag is accurate, it will not count undetermined frames.
9687 If the flag is found to be accurate it will be used without any further
9688 computations, if it is found to be inaccurate it will be cleared without any
9689 further computations. This allows inserting the idet filter as a low computational
9690 method to clean up the interlaced flag
9695 Deinterleave or interleave fields.
9697 This filter allows one to process interlaced images fields without
9698 deinterlacing them. Deinterleaving splits the input frame into 2
9699 fields (so called half pictures). Odd lines are moved to the top
9700 half of the output image, even lines to the bottom half.
9701 You can process (filter) them independently and then re-interleave them.
9703 The filter accepts the following options:
9707 @item chroma_mode, c
9709 Available values for @var{luma_mode}, @var{chroma_mode} and
9710 @var{alpha_mode} are:
9716 @item deinterleave, d
9717 Deinterleave fields, placing one above the other.
9720 Interleave fields. Reverse the effect of deinterleaving.
9722 Default value is @code{none}.
9725 @item chroma_swap, cs
9726 @item alpha_swap, as
9727 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9732 Apply inflate effect to the video.
9734 This filter replaces the pixel by the local(3x3) average by taking into account
9735 only values higher than the pixel.
9737 It accepts the following options:
9744 Limit the maximum change for each plane, default is 65535.
9745 If 0, plane will remain unchanged.
9750 Simple interlacing filter from progressive contents. This interleaves upper (or
9751 lower) lines from odd frames with lower (or upper) lines from even frames,
9752 halving the frame rate and preserving image height.
9755 Original Original New Frame
9756 Frame 'j' Frame 'j+1' (tff)
9757 ========== =========== ==================
9758 Line 0 --------------------> Frame 'j' Line 0
9759 Line 1 Line 1 ----> Frame 'j+1' Line 1
9760 Line 2 ---------------------> Frame 'j' Line 2
9761 Line 3 Line 3 ----> Frame 'j+1' Line 3
9763 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9766 It accepts the following optional parameters:
9770 This determines whether the interlaced frame is taken from the even
9771 (tff - default) or odd (bff) lines of the progressive frame.
9774 Vertical lowpass filter to avoid twitter interlacing and
9775 reduce moire patterns.
9779 Disable vertical lowpass filter
9782 Enable linear filter (default)
9785 Enable complex filter. This will slightly less reduce twitter and moire
9786 but better retain detail and subjective sharpness impression.
9793 Deinterlace input video by applying Donald Graft's adaptive kernel
9794 deinterling. Work on interlaced parts of a video to produce
9797 The description of the accepted parameters follows.
9801 Set the threshold which affects the filter's tolerance when
9802 determining if a pixel line must be processed. It must be an integer
9803 in the range [0,255] and defaults to 10. A value of 0 will result in
9804 applying the process on every pixels.
9807 Paint pixels exceeding the threshold value to white if set to 1.
9811 Set the fields order. Swap fields if set to 1, leave fields alone if
9815 Enable additional sharpening if set to 1. Default is 0.
9818 Enable twoway sharpening if set to 1. Default is 0.
9821 @subsection Examples
9825 Apply default values:
9827 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9831 Enable additional sharpening:
9837 Paint processed pixels in white:
9843 @section lenscorrection
9845 Correct radial lens distortion
9847 This filter can be used to correct for radial distortion as can result from the use
9848 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9849 one can use tools available for example as part of opencv or simply trial-and-error.
9850 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9851 and extract the k1 and k2 coefficients from the resulting matrix.
9853 Note that effectively the same filter is available in the open-source tools Krita and
9854 Digikam from the KDE project.
9856 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9857 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9858 brightness distribution, so you may want to use both filters together in certain
9859 cases, though you will have to take care of ordering, i.e. whether vignetting should
9860 be applied before or after lens correction.
9864 The filter accepts the following options:
9868 Relative x-coordinate of the focal point of the image, and thereby the center of the
9869 distortion. This value has a range [0,1] and is expressed as fractions of the image
9872 Relative y-coordinate of the focal point of the image, and thereby the center of the
9873 distortion. This value has a range [0,1] and is expressed as fractions of the image
9876 Coefficient of the quadratic correction term. 0.5 means no correction.
9878 Coefficient of the double quadratic correction term. 0.5 means no correction.
9881 The formula that generates the correction is:
9883 @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)
9885 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9886 distances from the focal point in the source and target images, respectively.
9890 Obtain the VMAF (Video Multi-Method Assessment Fusion)
9891 score between two input videos.
9893 The obtained VMAF score is printed through the logging system.
9895 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
9896 After installing the library it can be enabled using:
9897 @code{./configure --enable-libvmaf}.
9898 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
9900 The filter has following options:
9904 Set the model path which is to be used for SVM.
9905 Default value: @code{"vmaf_v0.6.1.pkl"}
9908 Set the file path to be used to store logs.
9911 Set the format of the log file (xml or json).
9913 @item enable_transform
9914 Enables transform for computing vmaf.
9917 Invokes the phone model which will generate VMAF scores higher than in the
9918 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
9921 Enables computing psnr along with vmaf.
9924 Enables computing ssim along with vmaf.
9927 Enables computing ms_ssim along with vmaf.
9930 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
9933 This filter also supports the @ref{framesync} options.
9935 On the below examples the input file @file{main.mpg} being processed is
9936 compared with the reference file @file{ref.mpg}.
9939 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
9942 Example with options:
9944 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
9949 Limits the pixel components values to the specified range [min, max].
9951 The filter accepts the following options:
9955 Lower bound. Defaults to the lowest allowed value for the input.
9958 Upper bound. Defaults to the highest allowed value for the input.
9961 Specify which planes will be processed. Defaults to all available.
9968 The filter accepts the following options:
9972 Set the number of loops.
9975 Set maximal size in number of frames.
9978 Set first frame of loop.
9984 Apply a 3D LUT to an input video.
9986 The filter accepts the following options:
9990 Set the 3D LUT file name.
9992 Currently supported formats:
10004 Select interpolation mode.
10006 Available values are:
10010 Use values from the nearest defined point.
10012 Interpolate values using the 8 points defining a cube.
10014 Interpolate values using a tetrahedron.
10018 This filter also supports the @ref{framesync} options.
10022 Turn certain luma values into transparency.
10024 The filter accepts the following options:
10028 Set the luma which will be used as base for transparency.
10029 Default value is @code{0}.
10032 Set the range of luma values to be keyed out.
10033 Default value is @code{0}.
10036 Set the range of softness. Default value is @code{0}.
10037 Use this to control gradual transition from zero to full transparency.
10040 @section lut, lutrgb, lutyuv
10042 Compute a look-up table for binding each pixel component input value
10043 to an output value, and apply it to the input video.
10045 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10046 to an RGB input video.
10048 These filters accept the following parameters:
10051 set first pixel component expression
10053 set second pixel component expression
10055 set third pixel component expression
10057 set fourth pixel component expression, corresponds to the alpha component
10060 set red component expression
10062 set green component expression
10064 set blue component expression
10066 alpha component expression
10069 set Y/luminance component expression
10071 set U/Cb component expression
10073 set V/Cr component expression
10076 Each of them specifies the expression to use for computing the lookup table for
10077 the corresponding pixel component values.
10079 The exact component associated to each of the @var{c*} options depends on the
10082 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10083 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10085 The expressions can contain the following constants and functions:
10090 The input width and height.
10093 The input value for the pixel component.
10096 The input value, clipped to the @var{minval}-@var{maxval} range.
10099 The maximum value for the pixel component.
10102 The minimum value for the pixel component.
10105 The negated value for the pixel component value, clipped to the
10106 @var{minval}-@var{maxval} range; it corresponds to the expression
10107 "maxval-clipval+minval".
10110 The computed value in @var{val}, clipped to the
10111 @var{minval}-@var{maxval} range.
10113 @item gammaval(gamma)
10114 The computed gamma correction value of the pixel component value,
10115 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10117 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10121 All expressions default to "val".
10123 @subsection Examples
10127 Negate input video:
10129 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10130 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10133 The above is the same as:
10135 lutrgb="r=negval:g=negval:b=negval"
10136 lutyuv="y=negval:u=negval:v=negval"
10146 Remove chroma components, turning the video into a graytone image:
10148 lutyuv="u=128:v=128"
10152 Apply a luma burning effect:
10158 Remove green and blue components:
10164 Set a constant alpha channel value on input:
10166 format=rgba,lutrgb=a="maxval-minval/2"
10170 Correct luminance gamma by a factor of 0.5:
10172 lutyuv=y=gammaval(0.5)
10176 Discard least significant bits of luma:
10178 lutyuv=y='bitand(val, 128+64+32)'
10182 Technicolor like effect:
10184 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10188 @section lut2, tlut2
10190 The @code{lut2} filter takes two input streams and outputs one
10193 The @code{tlut2} (time lut2) filter takes two consecutive frames
10194 from one single stream.
10196 This filter accepts the following parameters:
10199 set first pixel component expression
10201 set second pixel component expression
10203 set third pixel component expression
10205 set fourth pixel component expression, corresponds to the alpha component
10208 Each of them specifies the expression to use for computing the lookup table for
10209 the corresponding pixel component values.
10211 The exact component associated to each of the @var{c*} options depends on the
10214 The expressions can contain the following constants:
10219 The input width and height.
10222 The first input value for the pixel component.
10225 The second input value for the pixel component.
10228 The first input video bit depth.
10231 The second input video bit depth.
10234 All expressions default to "x".
10236 @subsection Examples
10240 Highlight differences between two RGB video streams:
10242 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)'
10246 Highlight differences between two YUV video streams:
10248 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)'
10252 Show max difference between two video streams:
10254 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)))'
10258 @section maskedclamp
10260 Clamp the first input stream with the second input and third input stream.
10262 Returns the value of first stream to be between second input
10263 stream - @code{undershoot} and third input stream + @code{overshoot}.
10265 This filter accepts the following options:
10268 Default value is @code{0}.
10271 Default value is @code{0}.
10274 Set which planes will be processed as bitmap, unprocessed planes will be
10275 copied from first stream.
10276 By default value 0xf, all planes will be processed.
10279 @section maskedmerge
10281 Merge the first input stream with the second input stream using per pixel
10282 weights in the third input stream.
10284 A value of 0 in the third stream pixel component means that pixel component
10285 from first stream is returned unchanged, while maximum value (eg. 255 for
10286 8-bit videos) means that pixel component from second stream is returned
10287 unchanged. Intermediate values define the amount of merging between both
10288 input stream's pixel components.
10290 This filter accepts the following options:
10293 Set which planes will be processed as bitmap, unprocessed planes will be
10294 copied from first stream.
10295 By default value 0xf, all planes will be processed.
10300 Apply motion-compensation deinterlacing.
10302 It needs one field per frame as input and must thus be used together
10303 with yadif=1/3 or equivalent.
10305 This filter accepts the following options:
10308 Set the deinterlacing mode.
10310 It accepts one of the following values:
10315 use iterative motion estimation
10317 like @samp{slow}, but use multiple reference frames.
10319 Default value is @samp{fast}.
10322 Set the picture field parity assumed for the input video. It must be
10323 one of the following values:
10327 assume top field first
10329 assume bottom field first
10332 Default value is @samp{bff}.
10335 Set per-block quantization parameter (QP) used by the internal
10338 Higher values should result in a smoother motion vector field but less
10339 optimal individual vectors. Default value is 1.
10342 @section mergeplanes
10344 Merge color channel components from several video streams.
10346 The filter accepts up to 4 input streams, and merge selected input
10347 planes to the output video.
10349 This filter accepts the following options:
10352 Set input to output plane mapping. Default is @code{0}.
10354 The mappings is specified as a bitmap. It should be specified as a
10355 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10356 mapping for the first plane of the output stream. 'A' sets the number of
10357 the input stream to use (from 0 to 3), and 'a' the plane number of the
10358 corresponding input to use (from 0 to 3). The rest of the mappings is
10359 similar, 'Bb' describes the mapping for the output stream second
10360 plane, 'Cc' describes the mapping for the output stream third plane and
10361 'Dd' describes the mapping for the output stream fourth plane.
10364 Set output pixel format. Default is @code{yuva444p}.
10367 @subsection Examples
10371 Merge three gray video streams of same width and height into single video stream:
10373 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10377 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10379 [a0][a1]mergeplanes=0x00010210:yuva444p
10383 Swap Y and A plane in yuva444p stream:
10385 format=yuva444p,mergeplanes=0x03010200:yuva444p
10389 Swap U and V plane in yuv420p stream:
10391 format=yuv420p,mergeplanes=0x000201:yuv420p
10395 Cast a rgb24 clip to yuv444p:
10397 format=rgb24,mergeplanes=0x000102:yuv444p
10403 Estimate and export motion vectors using block matching algorithms.
10404 Motion vectors are stored in frame side data to be used by other filters.
10406 This filter accepts the following options:
10409 Specify the motion estimation method. Accepts one of the following values:
10413 Exhaustive search algorithm.
10415 Three step search algorithm.
10417 Two dimensional logarithmic search algorithm.
10419 New three step search algorithm.
10421 Four step search algorithm.
10423 Diamond search algorithm.
10425 Hexagon-based search algorithm.
10427 Enhanced predictive zonal search algorithm.
10429 Uneven multi-hexagon search algorithm.
10431 Default value is @samp{esa}.
10434 Macroblock size. Default @code{16}.
10437 Search parameter. Default @code{7}.
10440 @section midequalizer
10442 Apply Midway Image Equalization effect using two video streams.
10444 Midway Image Equalization adjusts a pair of images to have the same
10445 histogram, while maintaining their dynamics as much as possible. It's
10446 useful for e.g. matching exposures from a pair of stereo cameras.
10448 This filter has two inputs and one output, which must be of same pixel format, but
10449 may be of different sizes. The output of filter is first input adjusted with
10450 midway histogram of both inputs.
10452 This filter accepts the following option:
10456 Set which planes to process. Default is @code{15}, which is all available planes.
10459 @section minterpolate
10461 Convert the video to specified frame rate using motion interpolation.
10463 This filter accepts the following options:
10466 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}.
10469 Motion interpolation mode. Following values are accepted:
10472 Duplicate previous or next frame for interpolating new ones.
10474 Blend source frames. Interpolated frame is mean of previous and next frames.
10476 Motion compensated interpolation. Following options are effective when this mode is selected:
10480 Motion compensation mode. Following values are accepted:
10483 Overlapped block motion compensation.
10485 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10487 Default mode is @samp{obmc}.
10490 Motion estimation mode. Following values are accepted:
10493 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
10495 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
10497 Default mode is @samp{bilat}.
10500 The algorithm to be used for motion estimation. Following values are accepted:
10503 Exhaustive search algorithm.
10505 Three step search algorithm.
10507 Two dimensional logarithmic search algorithm.
10509 New three step search algorithm.
10511 Four step search algorithm.
10513 Diamond search algorithm.
10515 Hexagon-based search algorithm.
10517 Enhanced predictive zonal search algorithm.
10519 Uneven multi-hexagon search algorithm.
10521 Default algorithm is @samp{epzs}.
10524 Macroblock size. Default @code{16}.
10527 Motion estimation search parameter. Default @code{32}.
10530 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).
10535 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:
10538 Disable scene change detection.
10540 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
10542 Default method is @samp{fdiff}.
10544 @item scd_threshold
10545 Scene change detection threshold. Default is @code{5.0}.
10548 @section mpdecimate
10550 Drop frames that do not differ greatly from the previous frame in
10551 order to reduce frame rate.
10553 The main use of this filter is for very-low-bitrate encoding
10554 (e.g. streaming over dialup modem), but it could in theory be used for
10555 fixing movies that were inverse-telecined incorrectly.
10557 A description of the accepted options follows.
10561 Set the maximum number of consecutive frames which can be dropped (if
10562 positive), or the minimum interval between dropped frames (if
10563 negative). If the value is 0, the frame is dropped disregarding the
10564 number of previous sequentially dropped frames.
10566 Default value is 0.
10571 Set the dropping threshold values.
10573 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
10574 represent actual pixel value differences, so a threshold of 64
10575 corresponds to 1 unit of difference for each pixel, or the same spread
10576 out differently over the block.
10578 A frame is a candidate for dropping if no 8x8 blocks differ by more
10579 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
10580 meaning the whole image) differ by more than a threshold of @option{lo}.
10582 Default value for @option{hi} is 64*12, default value for @option{lo} is
10583 64*5, and default value for @option{frac} is 0.33.
10589 Negate input video.
10591 It accepts an integer in input; if non-zero it negates the
10592 alpha component (if available). The default value in input is 0.
10596 Denoise frames using Non-Local Means algorithm.
10598 Each pixel is adjusted by looking for other pixels with similar contexts. This
10599 context similarity is defined by comparing their surrounding patches of size
10600 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
10603 Note that the research area defines centers for patches, which means some
10604 patches will be made of pixels outside that research area.
10606 The filter accepts the following options.
10610 Set denoising strength.
10616 Same as @option{p} but for chroma planes.
10618 The default value is @var{0} and means automatic.
10624 Same as @option{r} but for chroma planes.
10626 The default value is @var{0} and means automatic.
10631 Deinterlace video using neural network edge directed interpolation.
10633 This filter accepts the following options:
10637 Mandatory option, without binary file filter can not work.
10638 Currently file can be found here:
10639 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
10642 Set which frames to deinterlace, by default it is @code{all}.
10643 Can be @code{all} or @code{interlaced}.
10646 Set mode of operation.
10648 Can be one of the following:
10652 Use frame flags, both fields.
10654 Use frame flags, single field.
10656 Use top field only.
10658 Use bottom field only.
10660 Use both fields, top first.
10662 Use both fields, bottom first.
10666 Set which planes to process, by default filter process all frames.
10669 Set size of local neighborhood around each pixel, used by the predictor neural
10672 Can be one of the following:
10685 Set the number of neurons in predictor neural network.
10686 Can be one of the following:
10697 Controls the number of different neural network predictions that are blended
10698 together to compute the final output value. Can be @code{fast}, default or
10702 Set which set of weights to use in the predictor.
10703 Can be one of the following:
10707 weights trained to minimize absolute error
10709 weights trained to minimize squared error
10713 Controls whether or not the prescreener neural network is used to decide
10714 which pixels should be processed by the predictor neural network and which
10715 can be handled by simple cubic interpolation.
10716 The prescreener is trained to know whether cubic interpolation will be
10717 sufficient for a pixel or whether it should be predicted by the predictor nn.
10718 The computational complexity of the prescreener nn is much less than that of
10719 the predictor nn. Since most pixels can be handled by cubic interpolation,
10720 using the prescreener generally results in much faster processing.
10721 The prescreener is pretty accurate, so the difference between using it and not
10722 using it is almost always unnoticeable.
10724 Can be one of the following:
10732 Default is @code{new}.
10735 Set various debugging flags.
10740 Force libavfilter not to use any of the specified pixel formats for the
10741 input to the next filter.
10743 It accepts the following parameters:
10747 A '|'-separated list of pixel format names, such as
10748 pix_fmts=yuv420p|monow|rgb24".
10752 @subsection Examples
10756 Force libavfilter to use a format different from @var{yuv420p} for the
10757 input to the vflip filter:
10759 noformat=pix_fmts=yuv420p,vflip
10763 Convert the input video to any of the formats not contained in the list:
10765 noformat=yuv420p|yuv444p|yuv410p
10771 Add noise on video input frame.
10773 The filter accepts the following options:
10781 Set noise seed for specific pixel component or all pixel components in case
10782 of @var{all_seed}. Default value is @code{123457}.
10784 @item all_strength, alls
10785 @item c0_strength, c0s
10786 @item c1_strength, c1s
10787 @item c2_strength, c2s
10788 @item c3_strength, c3s
10789 Set noise strength for specific pixel component or all pixel components in case
10790 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
10792 @item all_flags, allf
10793 @item c0_flags, c0f
10794 @item c1_flags, c1f
10795 @item c2_flags, c2f
10796 @item c3_flags, c3f
10797 Set pixel component flags or set flags for all components if @var{all_flags}.
10798 Available values for component flags are:
10801 averaged temporal noise (smoother)
10803 mix random noise with a (semi)regular pattern
10805 temporal noise (noise pattern changes between frames)
10807 uniform noise (gaussian otherwise)
10811 @subsection Examples
10813 Add temporal and uniform noise to input video:
10815 noise=alls=20:allf=t+u
10820 Pass the video source unchanged to the output.
10823 Optical Character Recognition
10825 This filter uses Tesseract for optical character recognition.
10827 It accepts the following options:
10831 Set datapath to tesseract data. Default is to use whatever was
10832 set at installation.
10835 Set language, default is "eng".
10838 Set character whitelist.
10841 Set character blacklist.
10844 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
10848 Apply a video transform using libopencv.
10850 To enable this filter, install the libopencv library and headers and
10851 configure FFmpeg with @code{--enable-libopencv}.
10853 It accepts the following parameters:
10858 The name of the libopencv filter to apply.
10860 @item filter_params
10861 The parameters to pass to the libopencv filter. If not specified, the default
10862 values are assumed.
10866 Refer to the official libopencv documentation for more precise
10868 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
10870 Several libopencv filters are supported; see the following subsections.
10875 Dilate an image by using a specific structuring element.
10876 It corresponds to the libopencv function @code{cvDilate}.
10878 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10880 @var{struct_el} represents a structuring element, and has the syntax:
10881 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10883 @var{cols} and @var{rows} represent the number of columns and rows of
10884 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10885 point, and @var{shape} the shape for the structuring element. @var{shape}
10886 must be "rect", "cross", "ellipse", or "custom".
10888 If the value for @var{shape} is "custom", it must be followed by a
10889 string of the form "=@var{filename}". The file with name
10890 @var{filename} is assumed to represent a binary image, with each
10891 printable character corresponding to a bright pixel. When a custom
10892 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10893 or columns and rows of the read file are assumed instead.
10895 The default value for @var{struct_el} is "3x3+0x0/rect".
10897 @var{nb_iterations} specifies the number of times the transform is
10898 applied to the image, and defaults to 1.
10902 # Use the default values
10905 # Dilate using a structuring element with a 5x5 cross, iterating two times
10906 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10908 # Read the shape from the file diamond.shape, iterating two times.
10909 # The file diamond.shape may contain a pattern of characters like this
10915 # The specified columns and rows are ignored
10916 # but the anchor point coordinates are not
10917 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10922 Erode an image by using a specific structuring element.
10923 It corresponds to the libopencv function @code{cvErode}.
10925 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10926 with the same syntax and semantics as the @ref{dilate} filter.
10930 Smooth the input video.
10932 The filter takes the following parameters:
10933 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10935 @var{type} is the type of smooth filter to apply, and must be one of
10936 the following values: "blur", "blur_no_scale", "median", "gaussian",
10937 or "bilateral". The default value is "gaussian".
10939 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10940 depend on the smooth type. @var{param1} and
10941 @var{param2} accept integer positive values or 0. @var{param3} and
10942 @var{param4} accept floating point values.
10944 The default value for @var{param1} is 3. The default value for the
10945 other parameters is 0.
10947 These parameters correspond to the parameters assigned to the
10948 libopencv function @code{cvSmooth}.
10950 @section oscilloscope
10952 2D Video Oscilloscope.
10954 Useful to measure spatial impulse, step responses, chroma delays, etc.
10956 It accepts the following parameters:
10960 Set scope center x position.
10963 Set scope center y position.
10966 Set scope size, relative to frame diagonal.
10969 Set scope tilt/rotation.
10975 Set trace center x position.
10978 Set trace center y position.
10981 Set trace width, relative to width of frame.
10984 Set trace height, relative to height of frame.
10987 Set which components to trace. By default it traces first three components.
10990 Draw trace grid. By default is enabled.
10993 Draw some statistics. By default is enabled.
10996 Draw scope. By default is enabled.
10999 @subsection Examples
11003 Inspect full first row of video frame.
11005 oscilloscope=x=0.5:y=0:s=1
11009 Inspect full last row of video frame.
11011 oscilloscope=x=0.5:y=1:s=1
11015 Inspect full 5th line of video frame of height 1080.
11017 oscilloscope=x=0.5:y=5/1080:s=1
11021 Inspect full last column of video frame.
11023 oscilloscope=x=1:y=0.5:s=1:t=1
11031 Overlay one video on top of another.
11033 It takes two inputs and has one output. The first input is the "main"
11034 video on which the second input is overlaid.
11036 It accepts the following parameters:
11038 A description of the accepted options follows.
11043 Set the expression for the x and y coordinates of the overlaid video
11044 on the main video. Default value is "0" for both expressions. In case
11045 the expression is invalid, it is set to a huge value (meaning that the
11046 overlay will not be displayed within the output visible area).
11049 See @ref{framesync}.
11052 Set when the expressions for @option{x}, and @option{y} are evaluated.
11054 It accepts the following values:
11057 only evaluate expressions once during the filter initialization or
11058 when a command is processed
11061 evaluate expressions for each incoming frame
11064 Default value is @samp{frame}.
11067 See @ref{framesync}.
11070 Set the format for the output video.
11072 It accepts the following values:
11075 force YUV420 output
11078 force YUV422 output
11081 force YUV444 output
11084 force packed RGB output
11087 force planar RGB output
11090 automatically pick format
11093 Default value is @samp{yuv420}.
11096 See @ref{framesync}.
11099 The @option{x}, and @option{y} expressions can contain the following
11105 The main input width and height.
11109 The overlay input width and height.
11113 The computed values for @var{x} and @var{y}. They are evaluated for
11118 horizontal and vertical chroma subsample values of the output
11119 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11123 the number of input frame, starting from 0
11126 the position in the file of the input frame, NAN if unknown
11129 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11133 This filter also supports the @ref{framesync} options.
11135 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11136 when evaluation is done @emph{per frame}, and will evaluate to NAN
11137 when @option{eval} is set to @samp{init}.
11139 Be aware that frames are taken from each input video in timestamp
11140 order, hence, if their initial timestamps differ, it is a good idea
11141 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11142 have them begin in the same zero timestamp, as the example for
11143 the @var{movie} filter does.
11145 You can chain together more overlays but you should test the
11146 efficiency of such approach.
11148 @subsection Commands
11150 This filter supports the following commands:
11154 Modify the x and y of the overlay input.
11155 The command accepts the same syntax of the corresponding option.
11157 If the specified expression is not valid, it is kept at its current
11161 @subsection Examples
11165 Draw the overlay at 10 pixels from the bottom right corner of the main
11168 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11171 Using named options the example above becomes:
11173 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11177 Insert a transparent PNG logo in the bottom left corner of the input,
11178 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11180 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11184 Insert 2 different transparent PNG logos (second logo on bottom
11185 right corner) using the @command{ffmpeg} tool:
11187 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
11191 Add a transparent color layer on top of the main video; @code{WxH}
11192 must specify the size of the main input to the overlay filter:
11194 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11198 Play an original video and a filtered version (here with the deshake
11199 filter) side by side using the @command{ffplay} tool:
11201 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11204 The above command is the same as:
11206 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11210 Make a sliding overlay appearing from the left to the right top part of the
11211 screen starting since time 2:
11213 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11217 Compose output by putting two input videos side to side:
11219 ffmpeg -i left.avi -i right.avi -filter_complex "
11220 nullsrc=size=200x100 [background];
11221 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11222 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11223 [background][left] overlay=shortest=1 [background+left];
11224 [background+left][right] overlay=shortest=1:x=100 [left+right]
11229 Mask 10-20 seconds of a video by applying the delogo filter to a section
11231 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11232 -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]'
11237 Chain several overlays in cascade:
11239 nullsrc=s=200x200 [bg];
11240 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11241 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11242 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11243 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11244 [in3] null, [mid2] overlay=100:100 [out0]
11251 Apply Overcomplete Wavelet denoiser.
11253 The filter accepts the following options:
11259 Larger depth values will denoise lower frequency components more, but
11260 slow down filtering.
11262 Must be an int in the range 8-16, default is @code{8}.
11264 @item luma_strength, ls
11267 Must be a double value in the range 0-1000, default is @code{1.0}.
11269 @item chroma_strength, cs
11270 Set chroma strength.
11272 Must be a double value in the range 0-1000, default is @code{1.0}.
11278 Add paddings to the input image, and place the original input at the
11279 provided @var{x}, @var{y} coordinates.
11281 It accepts the following parameters:
11286 Specify an expression for the size of the output image with the
11287 paddings added. If the value for @var{width} or @var{height} is 0, the
11288 corresponding input size is used for the output.
11290 The @var{width} expression can reference the value set by the
11291 @var{height} expression, and vice versa.
11293 The default value of @var{width} and @var{height} is 0.
11297 Specify the offsets to place the input image at within the padded area,
11298 with respect to the top/left border of the output image.
11300 The @var{x} expression can reference the value set by the @var{y}
11301 expression, and vice versa.
11303 The default value of @var{x} and @var{y} is 0.
11305 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11306 so the input image is centered on the padded area.
11309 Specify the color of the padded area. For the syntax of this option,
11310 check the "Color" section in the ffmpeg-utils manual.
11312 The default value of @var{color} is "black".
11315 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11317 It accepts the following values:
11321 Only evaluate expressions once during the filter initialization or when
11322 a command is processed.
11325 Evaluate expressions for each incoming frame.
11329 Default value is @samp{init}.
11332 Pad to aspect instead to a resolution.
11336 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11337 options are expressions containing the following constants:
11342 The input video width and height.
11346 These are the same as @var{in_w} and @var{in_h}.
11350 The output width and height (the size of the padded area), as
11351 specified by the @var{width} and @var{height} expressions.
11355 These are the same as @var{out_w} and @var{out_h}.
11359 The x and y offsets as specified by the @var{x} and @var{y}
11360 expressions, or NAN if not yet specified.
11363 same as @var{iw} / @var{ih}
11366 input sample aspect ratio
11369 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11373 The horizontal and vertical chroma subsample values. For example for the
11374 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11377 @subsection Examples
11381 Add paddings with the color "violet" to the input video. The output video
11382 size is 640x480, and the top-left corner of the input video is placed at
11385 pad=640:480:0:40:violet
11388 The example above is equivalent to the following command:
11390 pad=width=640:height=480:x=0:y=40:color=violet
11394 Pad the input to get an output with dimensions increased by 3/2,
11395 and put the input video at the center of the padded area:
11397 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
11401 Pad the input to get a squared output with size equal to the maximum
11402 value between the input width and height, and put the input video at
11403 the center of the padded area:
11405 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
11409 Pad the input to get a final w/h ratio of 16:9:
11411 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
11415 In case of anamorphic video, in order to set the output display aspect
11416 correctly, it is necessary to use @var{sar} in the expression,
11417 according to the relation:
11419 (ih * X / ih) * sar = output_dar
11420 X = output_dar / sar
11423 Thus the previous example needs to be modified to:
11425 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
11429 Double the output size and put the input video in the bottom-right
11430 corner of the output padded area:
11432 pad="2*iw:2*ih:ow-iw:oh-ih"
11436 @anchor{palettegen}
11437 @section palettegen
11439 Generate one palette for a whole video stream.
11441 It accepts the following options:
11445 Set the maximum number of colors to quantize in the palette.
11446 Note: the palette will still contain 256 colors; the unused palette entries
11449 @item reserve_transparent
11450 Create a palette of 255 colors maximum and reserve the last one for
11451 transparency. Reserving the transparency color is useful for GIF optimization.
11452 If not set, the maximum of colors in the palette will be 256. You probably want
11453 to disable this option for a standalone image.
11456 @item transparency_color
11457 Set the color that will be used as background for transparency.
11460 Set statistics mode.
11462 It accepts the following values:
11465 Compute full frame histograms.
11467 Compute histograms only for the part that differs from previous frame. This
11468 might be relevant to give more importance to the moving part of your input if
11469 the background is static.
11471 Compute new histogram for each frame.
11474 Default value is @var{full}.
11477 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
11478 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
11479 color quantization of the palette. This information is also visible at
11480 @var{info} logging level.
11482 @subsection Examples
11486 Generate a representative palette of a given video using @command{ffmpeg}:
11488 ffmpeg -i input.mkv -vf palettegen palette.png
11492 @section paletteuse
11494 Use a palette to downsample an input video stream.
11496 The filter takes two inputs: one video stream and a palette. The palette must
11497 be a 256 pixels image.
11499 It accepts the following options:
11503 Select dithering mode. Available algorithms are:
11506 Ordered 8x8 bayer dithering (deterministic)
11508 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
11509 Note: this dithering is sometimes considered "wrong" and is included as a
11511 @item floyd_steinberg
11512 Floyd and Steingberg dithering (error diffusion)
11514 Frankie Sierra dithering v2 (error diffusion)
11516 Frankie Sierra dithering v2 "Lite" (error diffusion)
11519 Default is @var{sierra2_4a}.
11522 When @var{bayer} dithering is selected, this option defines the scale of the
11523 pattern (how much the crosshatch pattern is visible). A low value means more
11524 visible pattern for less banding, and higher value means less visible pattern
11525 at the cost of more banding.
11527 The option must be an integer value in the range [0,5]. Default is @var{2}.
11530 If set, define the zone to process
11534 Only the changing rectangle will be reprocessed. This is similar to GIF
11535 cropping/offsetting compression mechanism. This option can be useful for speed
11536 if only a part of the image is changing, and has use cases such as limiting the
11537 scope of the error diffusal @option{dither} to the rectangle that bounds the
11538 moving scene (it leads to more deterministic output if the scene doesn't change
11539 much, and as a result less moving noise and better GIF compression).
11542 Default is @var{none}.
11545 Take new palette for each output frame.
11547 @item alpha_threshold
11548 Sets the alpha threshold for transparency. Alpha values above this threshold
11549 will be treated as completely opaque, and values below this threshold will be
11550 treated as completely transparent.
11552 The option must be an integer value in the range [0,255]. Default is @var{128}.
11555 @subsection Examples
11559 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
11560 using @command{ffmpeg}:
11562 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
11566 @section perspective
11568 Correct perspective of video not recorded perpendicular to the screen.
11570 A description of the accepted parameters follows.
11581 Set coordinates expression for top left, top right, bottom left and bottom right corners.
11582 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
11583 If the @code{sense} option is set to @code{source}, then the specified points will be sent
11584 to the corners of the destination. If the @code{sense} option is set to @code{destination},
11585 then the corners of the source will be sent to the specified coordinates.
11587 The expressions can use the following variables:
11592 the width and height of video frame.
11596 Output frame count.
11599 @item interpolation
11600 Set interpolation for perspective correction.
11602 It accepts the following values:
11608 Default value is @samp{linear}.
11611 Set interpretation of coordinate options.
11613 It accepts the following values:
11617 Send point in the source specified by the given coordinates to
11618 the corners of the destination.
11620 @item 1, destination
11622 Send the corners of the source to the point in the destination specified
11623 by the given coordinates.
11625 Default value is @samp{source}.
11629 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
11631 It accepts the following values:
11634 only evaluate expressions once during the filter initialization or
11635 when a command is processed
11638 evaluate expressions for each incoming frame
11641 Default value is @samp{init}.
11646 Delay interlaced video by one field time so that the field order changes.
11648 The intended use is to fix PAL movies that have been captured with the
11649 opposite field order to the film-to-video transfer.
11651 A description of the accepted parameters follows.
11657 It accepts the following values:
11660 Capture field order top-first, transfer bottom-first.
11661 Filter will delay the bottom field.
11664 Capture field order bottom-first, transfer top-first.
11665 Filter will delay the top field.
11668 Capture and transfer with the same field order. This mode only exists
11669 for the documentation of the other options to refer to, but if you
11670 actually select it, the filter will faithfully do nothing.
11673 Capture field order determined automatically by field flags, transfer
11675 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
11676 basis using field flags. If no field information is available,
11677 then this works just like @samp{u}.
11680 Capture unknown or varying, transfer opposite.
11681 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
11682 analyzing the images and selecting the alternative that produces best
11683 match between the fields.
11686 Capture top-first, transfer unknown or varying.
11687 Filter selects among @samp{t} and @samp{p} using image analysis.
11690 Capture bottom-first, transfer unknown or varying.
11691 Filter selects among @samp{b} and @samp{p} using image analysis.
11694 Capture determined by field flags, transfer unknown or varying.
11695 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
11696 image analysis. If no field information is available, then this works just
11697 like @samp{U}. This is the default mode.
11700 Both capture and transfer unknown or varying.
11701 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
11705 @section pixdesctest
11707 Pixel format descriptor test filter, mainly useful for internal
11708 testing. The output video should be equal to the input video.
11712 format=monow, pixdesctest
11715 can be used to test the monowhite pixel format descriptor definition.
11719 Display sample values of color channels. Mainly useful for checking color
11720 and levels. Minimum supported resolution is 640x480.
11722 The filters accept the following options:
11726 Set scope X position, relative offset on X axis.
11729 Set scope Y position, relative offset on Y axis.
11738 Set window opacity. This window also holds statistics about pixel area.
11741 Set window X position, relative offset on X axis.
11744 Set window Y position, relative offset on Y axis.
11749 Enable the specified chain of postprocessing subfilters using libpostproc. This
11750 library should be automatically selected with a GPL build (@code{--enable-gpl}).
11751 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
11752 Each subfilter and some options have a short and a long name that can be used
11753 interchangeably, i.e. dr/dering are the same.
11755 The filters accept the following options:
11759 Set postprocessing subfilters string.
11762 All subfilters share common options to determine their scope:
11766 Honor the quality commands for this subfilter.
11769 Do chrominance filtering, too (default).
11772 Do luminance filtering only (no chrominance).
11775 Do chrominance filtering only (no luminance).
11778 These options can be appended after the subfilter name, separated by a '|'.
11780 Available subfilters are:
11783 @item hb/hdeblock[|difference[|flatness]]
11784 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 vb/vdeblock[|difference[|flatness]]
11793 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}).
11801 @item ha/hadeblock[|difference[|flatness]]
11802 Accurate horizontal deblocking filter
11805 Difference factor where higher values mean more deblocking (default: @code{32}).
11807 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11810 @item va/vadeblock[|difference[|flatness]]
11811 Accurate vertical deblocking filter
11814 Difference factor where higher values mean more deblocking (default: @code{32}).
11816 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11820 The horizontal and vertical deblocking filters share the difference and
11821 flatness values so you cannot set different horizontal and vertical
11825 @item h1/x1hdeblock
11826 Experimental horizontal deblocking filter
11828 @item v1/x1vdeblock
11829 Experimental vertical deblocking filter
11834 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
11837 larger -> stronger filtering
11839 larger -> stronger filtering
11841 larger -> stronger filtering
11844 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
11847 Stretch luminance to @code{0-255}.
11850 @item lb/linblenddeint
11851 Linear blend deinterlacing filter that deinterlaces the given block by
11852 filtering all lines with a @code{(1 2 1)} filter.
11854 @item li/linipoldeint
11855 Linear interpolating deinterlacing filter that deinterlaces the given block by
11856 linearly interpolating every second line.
11858 @item ci/cubicipoldeint
11859 Cubic interpolating deinterlacing filter deinterlaces the given block by
11860 cubically interpolating every second line.
11862 @item md/mediandeint
11863 Median deinterlacing filter that deinterlaces the given block by applying a
11864 median filter to every second line.
11866 @item fd/ffmpegdeint
11867 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
11868 second line with a @code{(-1 4 2 4 -1)} filter.
11871 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
11872 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
11874 @item fq/forceQuant[|quantizer]
11875 Overrides the quantizer table from the input with the constant quantizer you
11883 Default pp filter combination (@code{hb|a,vb|a,dr|a})
11886 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
11889 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
11892 @subsection Examples
11896 Apply horizontal and vertical deblocking, deringing and automatic
11897 brightness/contrast:
11903 Apply default filters without brightness/contrast correction:
11909 Apply default filters and temporal denoiser:
11911 pp=default/tmpnoise|1|2|3
11915 Apply deblocking on luminance only, and switch vertical deblocking on or off
11916 automatically depending on available CPU time:
11923 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
11924 similar to spp = 6 with 7 point DCT, where only the center sample is
11927 The filter accepts the following options:
11931 Force a constant quantization parameter. It accepts an integer in range
11932 0 to 63. If not set, the filter will use the QP from the video stream
11936 Set thresholding mode. Available modes are:
11940 Set hard thresholding.
11942 Set soft thresholding (better de-ringing effect, but likely blurrier).
11944 Set medium thresholding (good results, default).
11948 @section premultiply
11949 Apply alpha premultiply effect to input video stream using first plane
11950 of second stream as alpha.
11952 Both streams must have same dimensions and same pixel format.
11954 The filter accepts the following option:
11958 Set which planes will be processed, unprocessed planes will be copied.
11959 By default value 0xf, all planes will be processed.
11962 Do not require 2nd input for processing, instead use alpha plane from input stream.
11966 Apply prewitt operator to input video stream.
11968 The filter accepts the following option:
11972 Set which planes will be processed, unprocessed planes will be copied.
11973 By default value 0xf, all planes will be processed.
11976 Set value which will be multiplied with filtered result.
11979 Set value which will be added to filtered result.
11982 @section pseudocolor
11984 Alter frame colors in video with pseudocolors.
11986 This filter accept the following options:
11990 set pixel first component expression
11993 set pixel second component expression
11996 set pixel third component expression
11999 set pixel fourth component expression, corresponds to the alpha component
12002 set component to use as base for altering colors
12005 Each of them specifies the expression to use for computing the lookup table for
12006 the corresponding pixel component values.
12008 The expressions can contain the following constants and functions:
12013 The input width and height.
12016 The input value for the pixel component.
12018 @item ymin, umin, vmin, amin
12019 The minimum allowed component value.
12021 @item ymax, umax, vmax, amax
12022 The maximum allowed component value.
12025 All expressions default to "val".
12027 @subsection Examples
12031 Change too high luma values to gradient:
12033 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'"
12039 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12040 Ratio) between two input videos.
12042 This filter takes in input two input videos, the first input is
12043 considered the "main" source and is passed unchanged to the
12044 output. The second input is used as a "reference" video for computing
12047 Both video inputs must have the same resolution and pixel format for
12048 this filter to work correctly. Also it assumes that both inputs
12049 have the same number of frames, which are compared one by one.
12051 The obtained average PSNR is printed through the logging system.
12053 The filter stores the accumulated MSE (mean squared error) of each
12054 frame, and at the end of the processing it is averaged across all frames
12055 equally, and the following formula is applied to obtain the PSNR:
12058 PSNR = 10*log10(MAX^2/MSE)
12061 Where MAX is the average of the maximum values of each component of the
12064 The description of the accepted parameters follows.
12067 @item stats_file, f
12068 If specified the filter will use the named file to save the PSNR of
12069 each individual frame. When filename equals "-" the data is sent to
12072 @item stats_version
12073 Specifies which version of the stats file format to use. Details of
12074 each format are written below.
12075 Default value is 1.
12077 @item stats_add_max
12078 Determines whether the max value is output to the stats log.
12079 Default value is 0.
12080 Requires stats_version >= 2. If this is set and stats_version < 2,
12081 the filter will return an error.
12084 This filter also supports the @ref{framesync} options.
12086 The file printed if @var{stats_file} is selected, contains a sequence of
12087 key/value pairs of the form @var{key}:@var{value} for each compared
12090 If a @var{stats_version} greater than 1 is specified, a header line precedes
12091 the list of per-frame-pair stats, with key value pairs following the frame
12092 format with the following parameters:
12095 @item psnr_log_version
12096 The version of the log file format. Will match @var{stats_version}.
12099 A comma separated list of the per-frame-pair parameters included in
12103 A description of each shown per-frame-pair parameter follows:
12107 sequential number of the input frame, starting from 1
12110 Mean Square Error pixel-by-pixel average difference of the compared
12111 frames, averaged over all the image components.
12113 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
12114 Mean Square Error pixel-by-pixel average difference of the compared
12115 frames for the component specified by the suffix.
12117 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12118 Peak Signal to Noise ratio of the compared frames for the component
12119 specified by the suffix.
12121 @item max_avg, max_y, max_u, max_v
12122 Maximum allowed value for each channel, and average over all
12128 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12129 [main][ref] psnr="stats_file=stats.log" [out]
12132 On this example the input file being processed is compared with the
12133 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12134 is stored in @file{stats.log}.
12139 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12140 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12143 The pullup filter is designed to take advantage of future context in making
12144 its decisions. This filter is stateless in the sense that it does not lock
12145 onto a pattern to follow, but it instead looks forward to the following
12146 fields in order to identify matches and rebuild progressive frames.
12148 To produce content with an even framerate, insert the fps filter after
12149 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12150 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12152 The filter accepts the following options:
12159 These options set the amount of "junk" to ignore at the left, right, top, and
12160 bottom of the image, respectively. Left and right are in units of 8 pixels,
12161 while top and bottom are in units of 2 lines.
12162 The default is 8 pixels on each side.
12165 Set the strict breaks. Setting this option to 1 will reduce the chances of
12166 filter generating an occasional mismatched frame, but it may also cause an
12167 excessive number of frames to be dropped during high motion sequences.
12168 Conversely, setting it to -1 will make filter match fields more easily.
12169 This may help processing of video where there is slight blurring between
12170 the fields, but may also cause there to be interlaced frames in the output.
12171 Default value is @code{0}.
12174 Set the metric plane to use. It accepts the following values:
12180 Use chroma blue plane.
12183 Use chroma red plane.
12186 This option may be set to use chroma plane instead of the default luma plane
12187 for doing filter's computations. This may improve accuracy on very clean
12188 source material, but more likely will decrease accuracy, especially if there
12189 is chroma noise (rainbow effect) or any grayscale video.
12190 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12191 load and make pullup usable in realtime on slow machines.
12194 For best results (without duplicated frames in the output file) it is
12195 necessary to change the output frame rate. For example, to inverse
12196 telecine NTSC input:
12198 ffmpeg -i input -vf pullup -r 24000/1001 ...
12203 Change video quantization parameters (QP).
12205 The filter accepts the following option:
12209 Set expression for quantization parameter.
12212 The expression is evaluated through the eval API and can contain, among others,
12213 the following constants:
12217 1 if index is not 129, 0 otherwise.
12220 Sequential index starting from -129 to 128.
12223 @subsection Examples
12227 Some equation like:
12235 Flush video frames from internal cache of frames into a random order.
12236 No frame is discarded.
12237 Inspired by @ref{frei0r} nervous filter.
12241 Set size in number of frames of internal cache, in range from @code{2} to
12242 @code{512}. Default is @code{30}.
12245 Set seed for random number generator, must be an integer included between
12246 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12247 less than @code{0}, the filter will try to use a good random seed on a
12251 @section readeia608
12253 Read closed captioning (EIA-608) information from the top lines of a video frame.
12255 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
12256 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
12257 with EIA-608 data (starting from 0). A description of each metadata value follows:
12260 @item lavfi.readeia608.X.cc
12261 The two bytes stored as EIA-608 data (printed in hexadecimal).
12263 @item lavfi.readeia608.X.line
12264 The number of the line on which the EIA-608 data was identified and read.
12267 This filter accepts the following options:
12271 Set the line to start scanning for EIA-608 data. Default is @code{0}.
12274 Set the line to end scanning for EIA-608 data. Default is @code{29}.
12277 Set minimal acceptable amplitude change for sync codes detection.
12278 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
12281 Set the ratio of width reserved for sync code detection.
12282 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
12285 Set the max peaks height difference for sync code detection.
12286 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12289 Set max peaks period difference for sync code detection.
12290 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
12293 Set the first two max start code bits differences.
12294 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
12297 Set the minimum ratio of bits height compared to 3rd start code bit.
12298 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
12301 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
12304 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
12307 Enable checking the parity bit. In the event of a parity error, the filter will output
12308 @code{0x00} for that character. Default is false.
12311 @subsection Examples
12315 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
12317 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
12323 Read vertical interval timecode (VITC) information from the top lines of a
12326 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
12327 timecode value, if a valid timecode has been detected. Further metadata key
12328 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
12329 timecode data has been found or not.
12331 This filter accepts the following options:
12335 Set the maximum number of lines to scan for VITC data. If the value is set to
12336 @code{-1} the full video frame is scanned. Default is @code{45}.
12339 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
12340 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
12343 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
12344 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
12347 @subsection Examples
12351 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
12352 draw @code{--:--:--:--} as a placeholder:
12354 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
12360 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
12362 Destination pixel at position (X, Y) will be picked from source (x, y) position
12363 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
12364 value for pixel will be used for destination pixel.
12366 Xmap and Ymap input video streams must be of same dimensions. Output video stream
12367 will have Xmap/Ymap video stream dimensions.
12368 Xmap and Ymap input video streams are 16bit depth, single channel.
12370 @section removegrain
12372 The removegrain filter is a spatial denoiser for progressive video.
12376 Set mode for the first plane.
12379 Set mode for the second plane.
12382 Set mode for the third plane.
12385 Set mode for the fourth plane.
12388 Range of mode is from 0 to 24. Description of each mode follows:
12392 Leave input plane unchanged. Default.
12395 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
12398 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
12401 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
12404 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
12405 This is equivalent to a median filter.
12408 Line-sensitive clipping giving the minimal change.
12411 Line-sensitive clipping, intermediate.
12414 Line-sensitive clipping, intermediate.
12417 Line-sensitive clipping, intermediate.
12420 Line-sensitive clipping on a line where the neighbours pixels are the closest.
12423 Replaces the target pixel with the closest neighbour.
12426 [1 2 1] horizontal and vertical kernel blur.
12432 Bob mode, interpolates top field from the line where the neighbours
12433 pixels are the closest.
12436 Bob mode, interpolates bottom field from the line where the neighbours
12437 pixels are the closest.
12440 Bob mode, interpolates top field. Same as 13 but with a more complicated
12441 interpolation formula.
12444 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
12445 interpolation formula.
12448 Clips the pixel with the minimum and maximum of respectively the maximum and
12449 minimum of each pair of opposite neighbour pixels.
12452 Line-sensitive clipping using opposite neighbours whose greatest distance from
12453 the current pixel is minimal.
12456 Replaces the pixel with the average of its 8 neighbours.
12459 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
12462 Clips pixels using the averages of opposite neighbour.
12465 Same as mode 21 but simpler and faster.
12468 Small edge and halo removal, but reputed useless.
12474 @section removelogo
12476 Suppress a TV station logo, using an image file to determine which
12477 pixels comprise the logo. It works by filling in the pixels that
12478 comprise the logo with neighboring pixels.
12480 The filter accepts the following options:
12484 Set the filter bitmap file, which can be any image format supported by
12485 libavformat. The width and height of the image file must match those of the
12486 video stream being processed.
12489 Pixels in the provided bitmap image with a value of zero are not
12490 considered part of the logo, non-zero pixels are considered part of
12491 the logo. If you use white (255) for the logo and black (0) for the
12492 rest, you will be safe. For making the filter bitmap, it is
12493 recommended to take a screen capture of a black frame with the logo
12494 visible, and then using a threshold filter followed by the erode
12495 filter once or twice.
12497 If needed, little splotches can be fixed manually. Remember that if
12498 logo pixels are not covered, the filter quality will be much
12499 reduced. Marking too many pixels as part of the logo does not hurt as
12500 much, but it will increase the amount of blurring needed to cover over
12501 the image and will destroy more information than necessary, and extra
12502 pixels will slow things down on a large logo.
12504 @section repeatfields
12506 This filter uses the repeat_field flag from the Video ES headers and hard repeats
12507 fields based on its value.
12511 Reverse a video clip.
12513 Warning: This filter requires memory to buffer the entire clip, so trimming
12516 @subsection Examples
12520 Take the first 5 seconds of a clip, and reverse it.
12527 Apply roberts cross operator to input video stream.
12529 The filter accepts the following option:
12533 Set which planes will be processed, unprocessed planes will be copied.
12534 By default value 0xf, all planes will be processed.
12537 Set value which will be multiplied with filtered result.
12540 Set value which will be added to filtered result.
12545 Rotate video by an arbitrary angle expressed in radians.
12547 The filter accepts the following options:
12549 A description of the optional parameters follows.
12552 Set an expression for the angle by which to rotate the input video
12553 clockwise, expressed as a number of radians. A negative value will
12554 result in a counter-clockwise rotation. By default it is set to "0".
12556 This expression is evaluated for each frame.
12559 Set the output width expression, default value is "iw".
12560 This expression is evaluated just once during configuration.
12563 Set the output height expression, default value is "ih".
12564 This expression is evaluated just once during configuration.
12567 Enable bilinear interpolation if set to 1, a value of 0 disables
12568 it. Default value is 1.
12571 Set the color used to fill the output area not covered by the rotated
12572 image. For the general syntax of this option, check the "Color" section in the
12573 ffmpeg-utils manual. If the special value "none" is selected then no
12574 background is printed (useful for example if the background is never shown).
12576 Default value is "black".
12579 The expressions for the angle and the output size can contain the
12580 following constants and functions:
12584 sequential number of the input frame, starting from 0. It is always NAN
12585 before the first frame is filtered.
12588 time in seconds of the input frame, it is set to 0 when the filter is
12589 configured. It is always NAN before the first frame is filtered.
12593 horizontal and vertical chroma subsample values. For example for the
12594 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12598 the input video width and height
12602 the output width and height, that is the size of the padded area as
12603 specified by the @var{width} and @var{height} expressions
12607 the minimal width/height required for completely containing the input
12608 video rotated by @var{a} radians.
12610 These are only available when computing the @option{out_w} and
12611 @option{out_h} expressions.
12614 @subsection Examples
12618 Rotate the input by PI/6 radians clockwise:
12624 Rotate the input by PI/6 radians counter-clockwise:
12630 Rotate the input by 45 degrees clockwise:
12636 Apply a constant rotation with period T, starting from an angle of PI/3:
12638 rotate=PI/3+2*PI*t/T
12642 Make the input video rotation oscillating with a period of T
12643 seconds and an amplitude of A radians:
12645 rotate=A*sin(2*PI/T*t)
12649 Rotate the video, output size is chosen so that the whole rotating
12650 input video is always completely contained in the output:
12652 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
12656 Rotate the video, reduce the output size so that no background is ever
12659 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
12663 @subsection Commands
12665 The filter supports the following commands:
12669 Set the angle expression.
12670 The command accepts the same syntax of the corresponding option.
12672 If the specified expression is not valid, it is kept at its current
12678 Apply Shape Adaptive Blur.
12680 The filter accepts the following options:
12683 @item luma_radius, lr
12684 Set luma blur filter strength, must be a value in range 0.1-4.0, default
12685 value is 1.0. A greater value will result in a more blurred image, and
12686 in slower processing.
12688 @item luma_pre_filter_radius, lpfr
12689 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
12692 @item luma_strength, ls
12693 Set luma maximum difference between pixels to still be considered, must
12694 be a value in the 0.1-100.0 range, default value is 1.0.
12696 @item chroma_radius, cr
12697 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
12698 greater value will result in a more blurred image, and in slower
12701 @item chroma_pre_filter_radius, cpfr
12702 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
12704 @item chroma_strength, cs
12705 Set chroma maximum difference between pixels to still be considered,
12706 must be a value in the -0.9-100.0 range.
12709 Each chroma option value, if not explicitly specified, is set to the
12710 corresponding luma option value.
12715 Scale (resize) the input video, using the libswscale library.
12717 The scale filter forces the output display aspect ratio to be the same
12718 of the input, by changing the output sample aspect ratio.
12720 If the input image format is different from the format requested by
12721 the next filter, the scale filter will convert the input to the
12724 @subsection Options
12725 The filter accepts the following options, or any of the options
12726 supported by the libswscale scaler.
12728 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
12729 the complete list of scaler options.
12734 Set the output video dimension expression. Default value is the input
12737 If the @var{width} or @var{w} value is 0, the input width is used for
12738 the output. If the @var{height} or @var{h} value is 0, the input height
12739 is used for the output.
12741 If one and only one of the values is -n with n >= 1, the scale filter
12742 will use a value that maintains the aspect ratio of the input image,
12743 calculated from the other specified dimension. After that it will,
12744 however, make sure that the calculated dimension is divisible by n and
12745 adjust the value if necessary.
12747 If both values are -n with n >= 1, the behavior will be identical to
12748 both values being set to 0 as previously detailed.
12750 See below for the list of accepted constants for use in the dimension
12754 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
12758 Only evaluate expressions once during the filter initialization or when a command is processed.
12761 Evaluate expressions for each incoming frame.
12765 Default value is @samp{init}.
12769 Set the interlacing mode. It accepts the following values:
12773 Force interlaced aware scaling.
12776 Do not apply interlaced scaling.
12779 Select interlaced aware scaling depending on whether the source frames
12780 are flagged as interlaced or not.
12783 Default value is @samp{0}.
12786 Set libswscale scaling flags. See
12787 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12788 complete list of values. If not explicitly specified the filter applies
12792 @item param0, param1
12793 Set libswscale input parameters for scaling algorithms that need them. See
12794 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12795 complete documentation. If not explicitly specified the filter applies
12801 Set the video size. For the syntax of this option, check the
12802 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12804 @item in_color_matrix
12805 @item out_color_matrix
12806 Set in/output YCbCr color space type.
12808 This allows the autodetected value to be overridden as well as allows forcing
12809 a specific value used for the output and encoder.
12811 If not specified, the color space type depends on the pixel format.
12817 Choose automatically.
12820 Format conforming to International Telecommunication Union (ITU)
12821 Recommendation BT.709.
12824 Set color space conforming to the United States Federal Communications
12825 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
12828 Set color space conforming to:
12832 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
12835 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
12838 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
12843 Set color space conforming to SMPTE ST 240:1999.
12848 Set in/output YCbCr sample range.
12850 This allows the autodetected value to be overridden as well as allows forcing
12851 a specific value used for the output and encoder. If not specified, the
12852 range depends on the pixel format. Possible values:
12856 Choose automatically.
12859 Set full range (0-255 in case of 8-bit luma).
12862 Set "MPEG" range (16-235 in case of 8-bit luma).
12865 @item force_original_aspect_ratio
12866 Enable decreasing or increasing output video width or height if necessary to
12867 keep the original aspect ratio. Possible values:
12871 Scale the video as specified and disable this feature.
12874 The output video dimensions will automatically be decreased if needed.
12877 The output video dimensions will automatically be increased if needed.
12881 One useful instance of this option is that when you know a specific device's
12882 maximum allowed resolution, you can use this to limit the output video to
12883 that, while retaining the aspect ratio. For example, device A allows
12884 1280x720 playback, and your video is 1920x800. Using this option (set it to
12885 decrease) and specifying 1280x720 to the command line makes the output
12888 Please note that this is a different thing than specifying -1 for @option{w}
12889 or @option{h}, you still need to specify the output resolution for this option
12894 The values of the @option{w} and @option{h} options are expressions
12895 containing the following constants:
12900 The input width and height
12904 These are the same as @var{in_w} and @var{in_h}.
12908 The output (scaled) width and height
12912 These are the same as @var{out_w} and @var{out_h}
12915 The same as @var{iw} / @var{ih}
12918 input sample aspect ratio
12921 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
12925 horizontal and vertical input chroma subsample values. For example for the
12926 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12930 horizontal and vertical output chroma subsample values. For example for the
12931 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12934 @subsection Examples
12938 Scale the input video to a size of 200x100
12943 This is equivalent to:
12954 Specify a size abbreviation for the output size:
12959 which can also be written as:
12965 Scale the input to 2x:
12967 scale=w=2*iw:h=2*ih
12971 The above is the same as:
12973 scale=2*in_w:2*in_h
12977 Scale the input to 2x with forced interlaced scaling:
12979 scale=2*iw:2*ih:interl=1
12983 Scale the input to half size:
12985 scale=w=iw/2:h=ih/2
12989 Increase the width, and set the height to the same size:
12995 Seek Greek harmony:
13002 Increase the height, and set the width to 3/2 of the height:
13004 scale=w=3/2*oh:h=3/5*ih
13008 Increase the size, making the size a multiple of the chroma
13011 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13015 Increase the width to a maximum of 500 pixels,
13016 keeping the same aspect ratio as the input:
13018 scale=w='min(500\, iw*3/2):h=-1'
13022 @subsection Commands
13024 This filter supports the following commands:
13028 Set the output video dimension expression.
13029 The command accepts the same syntax of the corresponding option.
13031 If the specified expression is not valid, it is kept at its current
13037 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13038 format conversion on CUDA video frames. Setting the output width and height
13039 works in the same way as for the @var{scale} filter.
13041 The following additional options are accepted:
13044 The pixel format of the output CUDA frames. If set to the string "same" (the
13045 default), the input format will be kept. Note that automatic format negotiation
13046 and conversion is not yet supported for hardware frames
13049 The interpolation algorithm used for resizing. One of the following:
13056 @item cubic2p_bspline
13057 2-parameter cubic (B=1, C=0)
13059 @item cubic2p_catmullrom
13060 2-parameter cubic (B=0, C=1/2)
13062 @item cubic2p_b05c03
13063 2-parameter cubic (B=1/2, C=3/10)
13075 Scale (resize) the input video, based on a reference video.
13077 See the scale filter for available options, scale2ref supports the same but
13078 uses the reference video instead of the main input as basis. scale2ref also
13079 supports the following additional constants for the @option{w} and
13080 @option{h} options:
13085 The main input video's width and height
13088 The same as @var{main_w} / @var{main_h}
13091 The main input video's sample aspect ratio
13093 @item main_dar, mdar
13094 The main input video's display aspect ratio. Calculated from
13095 @code{(main_w / main_h) * main_sar}.
13099 The main input video's horizontal and vertical chroma subsample values.
13100 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13104 @subsection Examples
13108 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13110 'scale2ref[b][a];[a][b]overlay'
13114 @anchor{selectivecolor}
13115 @section selectivecolor
13117 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13118 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13119 by the "purity" of the color (that is, how saturated it already is).
13121 This filter is similar to the Adobe Photoshop Selective Color tool.
13123 The filter accepts the following options:
13126 @item correction_method
13127 Select color correction method.
13129 Available values are:
13132 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13135 Specified adjustments are relative to the original component value.
13137 Default is @code{absolute}.
13139 Adjustments for red pixels (pixels where the red component is the maximum)
13141 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13143 Adjustments for green pixels (pixels where the green component is the maximum)
13145 Adjustments for cyan pixels (pixels where the red component is the minimum)
13147 Adjustments for blue pixels (pixels where the blue component is the maximum)
13149 Adjustments for magenta pixels (pixels where the green component is the minimum)
13151 Adjustments for white pixels (pixels where all components are greater than 128)
13153 Adjustments for all pixels except pure black and pure white
13155 Adjustments for black pixels (pixels where all components are lesser than 128)
13157 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13160 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13161 4 space separated floating point adjustment values in the [-1,1] range,
13162 respectively to adjust the amount of cyan, magenta, yellow and black for the
13163 pixels of its range.
13165 @subsection Examples
13169 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13170 increase magenta by 27% in blue areas:
13172 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13176 Use a Photoshop selective color preset:
13178 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13182 @anchor{separatefields}
13183 @section separatefields
13185 The @code{separatefields} takes a frame-based video input and splits
13186 each frame into its components fields, producing a new half height clip
13187 with twice the frame rate and twice the frame count.
13189 This filter use field-dominance information in frame to decide which
13190 of each pair of fields to place first in the output.
13191 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13193 @section setdar, setsar
13195 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13198 This is done by changing the specified Sample (aka Pixel) Aspect
13199 Ratio, according to the following equation:
13201 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13204 Keep in mind that the @code{setdar} filter does not modify the pixel
13205 dimensions of the video frame. Also, the display aspect ratio set by
13206 this filter may be changed by later filters in the filterchain,
13207 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13210 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13211 the filter output video.
13213 Note that as a consequence of the application of this filter, the
13214 output display aspect ratio will change according to the equation
13217 Keep in mind that the sample aspect ratio set by the @code{setsar}
13218 filter may be changed by later filters in the filterchain, e.g. if
13219 another "setsar" or a "setdar" filter is applied.
13221 It accepts the following parameters:
13224 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13225 Set the aspect ratio used by the filter.
13227 The parameter can be a floating point number string, an expression, or
13228 a string of the form @var{num}:@var{den}, where @var{num} and
13229 @var{den} are the numerator and denominator of the aspect ratio. If
13230 the parameter is not specified, it is assumed the value "0".
13231 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13235 Set the maximum integer value to use for expressing numerator and
13236 denominator when reducing the expressed aspect ratio to a rational.
13237 Default value is @code{100}.
13241 The parameter @var{sar} is an expression containing
13242 the following constants:
13246 These are approximated values for the mathematical constants e
13247 (Euler's number), pi (Greek pi), and phi (the golden ratio).
13250 The input width and height.
13253 These are the same as @var{w} / @var{h}.
13256 The input sample aspect ratio.
13259 The input display aspect ratio. It is the same as
13260 (@var{w} / @var{h}) * @var{sar}.
13263 Horizontal and vertical chroma subsample values. For example, for the
13264 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13267 @subsection Examples
13272 To change the display aspect ratio to 16:9, specify one of the following:
13279 To change the sample aspect ratio to 10:11, specify:
13285 To set a display aspect ratio of 16:9, and specify a maximum integer value of
13286 1000 in the aspect ratio reduction, use the command:
13288 setdar=ratio=16/9:max=1000
13296 Force field for the output video frame.
13298 The @code{setfield} filter marks the interlace type field for the
13299 output frames. It does not change the input frame, but only sets the
13300 corresponding property, which affects how the frame is treated by
13301 following filters (e.g. @code{fieldorder} or @code{yadif}).
13303 The filter accepts the following options:
13308 Available values are:
13312 Keep the same field property.
13315 Mark the frame as bottom-field-first.
13318 Mark the frame as top-field-first.
13321 Mark the frame as progressive.
13327 Show a line containing various information for each input video frame.
13328 The input video is not modified.
13330 The shown line contains a sequence of key/value pairs of the form
13331 @var{key}:@var{value}.
13333 The following values are shown in the output:
13337 The (sequential) number of the input frame, starting from 0.
13340 The Presentation TimeStamp of the input frame, expressed as a number of
13341 time base units. The time base unit depends on the filter input pad.
13344 The Presentation TimeStamp of the input frame, expressed as a number of
13348 The position of the frame in the input stream, or -1 if this information is
13349 unavailable and/or meaningless (for example in case of synthetic video).
13352 The pixel format name.
13355 The sample aspect ratio of the input frame, expressed in the form
13356 @var{num}/@var{den}.
13359 The size of the input frame. For the syntax of this option, check the
13360 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13363 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
13364 for bottom field first).
13367 This is 1 if the frame is a key frame, 0 otherwise.
13370 The picture type of the input frame ("I" for an I-frame, "P" for a
13371 P-frame, "B" for a B-frame, or "?" for an unknown type).
13372 Also refer to the documentation of the @code{AVPictureType} enum and of
13373 the @code{av_get_picture_type_char} function defined in
13374 @file{libavutil/avutil.h}.
13377 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
13379 @item plane_checksum
13380 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
13381 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
13384 @section showpalette
13386 Displays the 256 colors palette of each frame. This filter is only relevant for
13387 @var{pal8} pixel format frames.
13389 It accepts the following option:
13393 Set the size of the box used to represent one palette color entry. Default is
13394 @code{30} (for a @code{30x30} pixel box).
13397 @section shuffleframes
13399 Reorder and/or duplicate and/or drop video frames.
13401 It accepts the following parameters:
13405 Set the destination indexes of input frames.
13406 This is space or '|' separated list of indexes that maps input frames to output
13407 frames. Number of indexes also sets maximal value that each index may have.
13408 '-1' index have special meaning and that is to drop frame.
13411 The first frame has the index 0. The default is to keep the input unchanged.
13413 @subsection Examples
13417 Swap second and third frame of every three frames of the input:
13419 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
13423 Swap 10th and 1st frame of every ten frames of the input:
13425 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
13429 @section shuffleplanes
13431 Reorder and/or duplicate video planes.
13433 It accepts the following parameters:
13438 The index of the input plane to be used as the first output plane.
13441 The index of the input plane to be used as the second output plane.
13444 The index of the input plane to be used as the third output plane.
13447 The index of the input plane to be used as the fourth output plane.
13451 The first plane has the index 0. The default is to keep the input unchanged.
13453 @subsection Examples
13457 Swap the second and third planes of the input:
13459 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
13463 @anchor{signalstats}
13464 @section signalstats
13465 Evaluate various visual metrics that assist in determining issues associated
13466 with the digitization of analog video media.
13468 By default the filter will log these metadata values:
13472 Display the minimal Y value contained within the input frame. Expressed in
13476 Display the Y value at the 10% percentile within the input frame. Expressed in
13480 Display the average Y value within the input frame. Expressed in range of
13484 Display the Y value at the 90% percentile within the input frame. Expressed in
13488 Display the maximum Y value contained within the input frame. Expressed in
13492 Display the minimal U value contained within the input frame. Expressed in
13496 Display the U value at the 10% percentile within the input frame. Expressed in
13500 Display the average U value within the input frame. Expressed in range of
13504 Display the U value at the 90% percentile within the input frame. Expressed in
13508 Display the maximum U value contained within the input frame. Expressed in
13512 Display the minimal V value contained within the input frame. Expressed in
13516 Display the V value at the 10% percentile within the input frame. Expressed in
13520 Display the average V value within the input frame. Expressed in range of
13524 Display the V value at the 90% percentile within the input frame. Expressed in
13528 Display the maximum V value contained within the input frame. Expressed in
13532 Display the minimal saturation value contained within the input frame.
13533 Expressed in range of [0-~181.02].
13536 Display the saturation value at the 10% percentile within the input frame.
13537 Expressed in range of [0-~181.02].
13540 Display the average saturation value within the input frame. Expressed in range
13544 Display the saturation value at the 90% percentile within the input frame.
13545 Expressed in range of [0-~181.02].
13548 Display the maximum saturation value contained within the input frame.
13549 Expressed in range of [0-~181.02].
13552 Display the median value for hue within the input frame. Expressed in range of
13556 Display the average value for hue within the input frame. Expressed in range of
13560 Display the average of sample value difference between all values of the Y
13561 plane in the current frame and corresponding values of the previous input frame.
13562 Expressed in range of [0-255].
13565 Display the average of sample value difference between all values of the U
13566 plane in the current frame and corresponding values of the previous input frame.
13567 Expressed in range of [0-255].
13570 Display the average of sample value difference between all values of the V
13571 plane in the current frame and corresponding values of the previous input frame.
13572 Expressed in range of [0-255].
13575 Display bit depth of Y plane in current frame.
13576 Expressed in range of [0-16].
13579 Display bit depth of U plane in current frame.
13580 Expressed in range of [0-16].
13583 Display bit depth of V plane in current frame.
13584 Expressed in range of [0-16].
13587 The filter accepts the following options:
13593 @option{stat} specify an additional form of image analysis.
13594 @option{out} output video with the specified type of pixel highlighted.
13596 Both options accept the following values:
13600 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
13601 unlike the neighboring pixels of the same field. Examples of temporal outliers
13602 include the results of video dropouts, head clogs, or tape tracking issues.
13605 Identify @var{vertical line repetition}. Vertical line repetition includes
13606 similar rows of pixels within a frame. In born-digital video vertical line
13607 repetition is common, but this pattern is uncommon in video digitized from an
13608 analog source. When it occurs in video that results from the digitization of an
13609 analog source it can indicate concealment from a dropout compensator.
13612 Identify pixels that fall outside of legal broadcast range.
13616 Set the highlight color for the @option{out} option. The default color is
13620 @subsection Examples
13624 Output data of various video metrics:
13626 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
13630 Output specific data about the minimum and maximum values of the Y plane per frame:
13632 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
13636 Playback video while highlighting pixels that are outside of broadcast range in red.
13638 ffplay example.mov -vf signalstats="out=brng:color=red"
13642 Playback video with signalstats metadata drawn over the frame.
13644 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
13647 The contents of signalstat_drawtext.txt used in the command are:
13650 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
13651 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
13652 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
13653 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
13661 Calculates the MPEG-7 Video Signature. The filter can handle more than one
13662 input. In this case the matching between the inputs can be calculated additionally.
13663 The filter always passes through the first input. The signature of each stream can
13664 be written into a file.
13666 It accepts the following options:
13670 Enable or disable the matching process.
13672 Available values are:
13676 Disable the calculation of a matching (default).
13678 Calculate the matching for the whole video and output whether the whole video
13679 matches or only parts.
13681 Calculate only until a matching is found or the video ends. Should be faster in
13686 Set the number of inputs. The option value must be a non negative integer.
13687 Default value is 1.
13690 Set the path to which the output is written. If there is more than one input,
13691 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
13692 integer), that will be replaced with the input number. If no filename is
13693 specified, no output will be written. This is the default.
13696 Choose the output format.
13698 Available values are:
13702 Use the specified binary representation (default).
13704 Use the specified xml representation.
13708 Set threshold to detect one word as similar. The option value must be an integer
13709 greater than zero. The default value is 9000.
13712 Set threshold to detect all words as similar. The option value must be an integer
13713 greater than zero. The default value is 60000.
13716 Set threshold to detect frames as similar. The option value must be an integer
13717 greater than zero. The default value is 116.
13720 Set the minimum length of a sequence in frames to recognize it as matching
13721 sequence. The option value must be a non negative integer value.
13722 The default value is 0.
13725 Set the minimum relation, that matching frames to all frames must have.
13726 The option value must be a double value between 0 and 1. The default value is 0.5.
13729 @subsection Examples
13733 To calculate the signature of an input video and store it in signature.bin:
13735 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
13739 To detect whether two videos match and store the signatures in XML format in
13740 signature0.xml and signature1.xml:
13742 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 -
13750 Blur the input video without impacting the outlines.
13752 It accepts the following options:
13755 @item luma_radius, lr
13756 Set the luma 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 1.0.
13760 @item luma_strength, ls
13761 Set the luma 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 1.0.
13766 @item luma_threshold, lt
13767 Set the luma 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 0.
13773 @item chroma_radius, cr
13774 Set the chroma radius. The option value must be a float number in
13775 the range [0.1,5.0] that specifies the variance of the gaussian filter
13776 used to blur the image (slower if larger). Default value is @option{luma_radius}.
13778 @item chroma_strength, cs
13779 Set the chroma strength. The option value must be a float number
13780 in the range [-1.0,1.0] that configures the blurring. A value included
13781 in [0.0,1.0] will blur the image whereas a value included in
13782 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
13784 @item chroma_threshold, ct
13785 Set the chroma threshold used as a coefficient to determine
13786 whether a pixel should be blurred or not. The option value must be an
13787 integer in the range [-30,30]. A value of 0 will filter all the image,
13788 a value included in [0,30] will filter flat areas and a value included
13789 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
13792 If a chroma option is not explicitly set, the corresponding luma value
13797 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
13799 This filter takes in input two input videos, the first input is
13800 considered the "main" source and is passed unchanged to the
13801 output. The second input is used as a "reference" video for computing
13804 Both video inputs must have the same resolution and pixel format for
13805 this filter to work correctly. Also it assumes that both inputs
13806 have the same number of frames, which are compared one by one.
13808 The filter stores the calculated SSIM of each frame.
13810 The description of the accepted parameters follows.
13813 @item stats_file, f
13814 If specified the filter will use the named file to save the SSIM of
13815 each individual frame. When filename equals "-" the data is sent to
13819 The file printed if @var{stats_file} is selected, contains a sequence of
13820 key/value pairs of the form @var{key}:@var{value} for each compared
13823 A description of each shown parameter follows:
13827 sequential number of the input frame, starting from 1
13829 @item Y, U, V, R, G, B
13830 SSIM of the compared frames for the component specified by the suffix.
13833 SSIM of the compared frames for the whole frame.
13836 Same as above but in dB representation.
13839 This filter also supports the @ref{framesync} options.
13843 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13844 [main][ref] ssim="stats_file=stats.log" [out]
13847 On this example the input file being processed is compared with the
13848 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
13849 is stored in @file{stats.log}.
13851 Another example with both psnr and ssim at same time:
13853 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
13858 Convert between different stereoscopic image formats.
13860 The filters accept the following options:
13864 Set stereoscopic image format of input.
13866 Available values for input image formats are:
13869 side by side parallel (left eye left, right eye right)
13872 side by side crosseye (right eye left, left eye right)
13875 side by side parallel with half width resolution
13876 (left eye left, right eye right)
13879 side by side crosseye with half width resolution
13880 (right eye left, left eye right)
13883 above-below (left eye above, right eye below)
13886 above-below (right eye above, left eye below)
13889 above-below with half height resolution
13890 (left eye above, right eye below)
13893 above-below with half height resolution
13894 (right eye above, left eye below)
13897 alternating frames (left eye first, right eye second)
13900 alternating frames (right eye first, left eye second)
13903 interleaved rows (left eye has top row, right eye starts on next row)
13906 interleaved rows (right eye has top row, left eye starts on next row)
13909 interleaved columns, left eye first
13912 interleaved columns, right eye first
13914 Default value is @samp{sbsl}.
13918 Set stereoscopic image format of output.
13922 side by side parallel (left eye left, right eye right)
13925 side by side crosseye (right eye left, left eye right)
13928 side by side parallel with half width resolution
13929 (left eye left, right eye right)
13932 side by side crosseye with half width resolution
13933 (right eye left, left eye right)
13936 above-below (left eye above, right eye below)
13939 above-below (right eye above, left eye below)
13942 above-below with half height resolution
13943 (left eye above, right eye below)
13946 above-below with half height resolution
13947 (right eye above, left eye below)
13950 alternating frames (left eye first, right eye second)
13953 alternating frames (right eye first, left eye second)
13956 interleaved rows (left eye has top row, right eye starts on next row)
13959 interleaved rows (right eye has top row, left eye starts on next row)
13962 anaglyph red/blue gray
13963 (red filter on left eye, blue filter on right eye)
13966 anaglyph red/green gray
13967 (red filter on left eye, green filter on right eye)
13970 anaglyph red/cyan gray
13971 (red filter on left eye, cyan filter on right eye)
13974 anaglyph red/cyan half colored
13975 (red filter on left eye, cyan filter on right eye)
13978 anaglyph red/cyan color
13979 (red filter on left eye, cyan filter on right eye)
13982 anaglyph red/cyan color optimized with the least squares projection of dubois
13983 (red filter on left eye, cyan filter on right eye)
13986 anaglyph green/magenta gray
13987 (green filter on left eye, magenta filter on right eye)
13990 anaglyph green/magenta half colored
13991 (green filter on left eye, magenta filter on right eye)
13994 anaglyph green/magenta colored
13995 (green filter on left eye, magenta filter on right eye)
13998 anaglyph green/magenta color optimized with the least squares projection of dubois
13999 (green filter on left eye, magenta filter on right eye)
14002 anaglyph yellow/blue gray
14003 (yellow filter on left eye, blue filter on right eye)
14006 anaglyph yellow/blue half colored
14007 (yellow filter on left eye, blue filter on right eye)
14010 anaglyph yellow/blue colored
14011 (yellow filter on left eye, blue filter on right eye)
14014 anaglyph yellow/blue color optimized with the least squares projection of dubois
14015 (yellow filter on left eye, blue filter on right eye)
14018 mono output (left eye only)
14021 mono output (right eye only)
14024 checkerboard, left eye first
14027 checkerboard, right eye first
14030 interleaved columns, left eye first
14033 interleaved columns, right eye first
14039 Default value is @samp{arcd}.
14042 @subsection Examples
14046 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14052 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14058 @section streamselect, astreamselect
14059 Select video or audio streams.
14061 The filter accepts the following options:
14065 Set number of inputs. Default is 2.
14068 Set input indexes to remap to outputs.
14071 @subsection Commands
14073 The @code{streamselect} and @code{astreamselect} filter supports the following
14078 Set input indexes to remap to outputs.
14081 @subsection Examples
14085 Select first 5 seconds 1st stream and rest of time 2nd stream:
14087 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14091 Same as above, but for audio:
14093 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14098 Apply sobel operator to input video stream.
14100 The filter accepts the following option:
14104 Set which planes will be processed, unprocessed planes will be copied.
14105 By default value 0xf, all planes will be processed.
14108 Set value which will be multiplied with filtered result.
14111 Set value which will be added to filtered result.
14117 Apply a simple postprocessing filter that compresses and decompresses the image
14118 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14119 and average the results.
14121 The filter accepts the following options:
14125 Set quality. This option defines the number of levels for averaging. It accepts
14126 an integer in the range 0-6. If set to @code{0}, the filter will have no
14127 effect. A value of @code{6} means the higher quality. For each increment of
14128 that value the speed drops by a factor of approximately 2. Default value is
14132 Force a constant quantization parameter. If not set, the filter will use the QP
14133 from the video stream (if available).
14136 Set thresholding mode. Available modes are:
14140 Set hard thresholding (default).
14142 Set soft thresholding (better de-ringing effect, but likely blurrier).
14145 @item use_bframe_qp
14146 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14147 option may cause flicker since the B-Frames have often larger QP. Default is
14148 @code{0} (not enabled).
14154 Draw subtitles on top of input video using the libass library.
14156 To enable compilation of this filter you need to configure FFmpeg with
14157 @code{--enable-libass}. This filter also requires a build with libavcodec and
14158 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14159 Alpha) subtitles format.
14161 The filter accepts the following options:
14165 Set the filename of the subtitle file to read. It must be specified.
14167 @item original_size
14168 Specify the size of the original video, the video for which the ASS file
14169 was composed. For the syntax of this option, check the
14170 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14171 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14172 correctly scale the fonts if the aspect ratio has been changed.
14175 Set a directory path containing fonts that can be used by the filter.
14176 These fonts will be used in addition to whatever the font provider uses.
14179 Process alpha channel, by default alpha channel is untouched.
14182 Set subtitles input character encoding. @code{subtitles} filter only. Only
14183 useful if not UTF-8.
14185 @item stream_index, si
14186 Set subtitles stream index. @code{subtitles} filter only.
14189 Override default style or script info parameters of the subtitles. It accepts a
14190 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14193 If the first key is not specified, it is assumed that the first value
14194 specifies the @option{filename}.
14196 For example, to render the file @file{sub.srt} on top of the input
14197 video, use the command:
14202 which is equivalent to:
14204 subtitles=filename=sub.srt
14207 To render the default subtitles stream from file @file{video.mkv}, use:
14209 subtitles=video.mkv
14212 To render the second subtitles stream from that file, use:
14214 subtitles=video.mkv:si=1
14217 To make the subtitles stream from @file{sub.srt} appear in transparent green
14218 @code{DejaVu Serif}, use:
14220 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14223 @section super2xsai
14225 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14226 Interpolate) pixel art scaling algorithm.
14228 Useful for enlarging pixel art images without reducing sharpness.
14232 Swap two rectangular objects in video.
14234 This filter accepts the following options:
14244 Set 1st rect x coordinate.
14247 Set 1st rect y coordinate.
14250 Set 2nd rect x coordinate.
14253 Set 2nd rect y coordinate.
14255 All expressions are evaluated once for each frame.
14258 The all options are expressions containing the following constants:
14263 The input width and height.
14266 same as @var{w} / @var{h}
14269 input sample aspect ratio
14272 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
14275 The number of the input frame, starting from 0.
14278 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
14281 the position in the file of the input frame, NAN if unknown
14289 Apply telecine process to the video.
14291 This filter accepts the following options:
14300 The default value is @code{top}.
14304 A string of numbers representing the pulldown pattern you wish to apply.
14305 The default value is @code{23}.
14309 Some typical patterns:
14314 24p: 2332 (preferred)
14321 24p: 222222222223 ("Euro pulldown")
14328 Apply threshold effect to video stream.
14330 This filter needs four video streams to perform thresholding.
14331 First stream is stream we are filtering.
14332 Second stream is holding threshold values, third stream is holding min values,
14333 and last, fourth stream is holding max values.
14335 The filter accepts the following option:
14339 Set which planes will be processed, unprocessed planes will be copied.
14340 By default value 0xf, all planes will be processed.
14343 For example if first stream pixel's component value is less then threshold value
14344 of pixel component from 2nd threshold stream, third stream value will picked,
14345 otherwise fourth stream pixel component value will be picked.
14347 Using color source filter one can perform various types of thresholding:
14349 @subsection Examples
14353 Binary threshold, using gray color as threshold:
14355 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
14359 Inverted binary threshold, using gray color as threshold:
14361 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
14365 Truncate binary threshold, using gray color as threshold:
14367 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
14371 Threshold to zero, using gray color as threshold:
14373 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
14377 Inverted threshold to zero, using gray color as threshold:
14379 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
14384 Select the most representative frame in a given sequence of consecutive frames.
14386 The filter accepts the following options:
14390 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
14391 will pick one of them, and then handle the next batch of @var{n} frames until
14392 the end. Default is @code{100}.
14395 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
14396 value will result in a higher memory usage, so a high value is not recommended.
14398 @subsection Examples
14402 Extract one picture each 50 frames:
14408 Complete example of a thumbnail creation with @command{ffmpeg}:
14410 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
14416 Tile several successive frames together.
14418 The filter accepts the following options:
14423 Set the grid size (i.e. the number of lines and columns). For the syntax of
14424 this option, check the
14425 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14428 Set the maximum number of frames to render in the given area. It must be less
14429 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
14430 the area will be used.
14433 Set the outer border margin in pixels.
14436 Set the inner border thickness (i.e. the number of pixels between frames). For
14437 more advanced padding options (such as having different values for the edges),
14438 refer to the pad video filter.
14441 Specify the color of the unused area. For the syntax of this option, check the
14442 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
14446 @subsection Examples
14450 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
14452 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
14454 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
14455 duplicating each output frame to accommodate the originally detected frame
14459 Display @code{5} pictures in an area of @code{3x2} frames,
14460 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
14461 mixed flat and named options:
14463 tile=3x2:nb_frames=5:padding=7:margin=2
14467 @section tinterlace
14469 Perform various types of temporal field interlacing.
14471 Frames are counted starting from 1, so the first input frame is
14474 The filter accepts the following options:
14479 Specify the mode of the interlacing. This option can also be specified
14480 as a value alone. See below for a list of values for this option.
14482 Available values are:
14486 Move odd frames into the upper field, even into the lower field,
14487 generating a double height frame at half frame rate.
14491 Frame 1 Frame 2 Frame 3 Frame 4
14493 11111 22222 33333 44444
14494 11111 22222 33333 44444
14495 11111 22222 33333 44444
14496 11111 22222 33333 44444
14510 Only output odd frames, even frames are dropped, generating a frame with
14511 unchanged height at half frame rate.
14516 Frame 1 Frame 2 Frame 3 Frame 4
14518 11111 22222 33333 44444
14519 11111 22222 33333 44444
14520 11111 22222 33333 44444
14521 11111 22222 33333 44444
14531 Only output even frames, odd frames are dropped, generating a frame with
14532 unchanged height at half frame rate.
14537 Frame 1 Frame 2 Frame 3 Frame 4
14539 11111 22222 33333 44444
14540 11111 22222 33333 44444
14541 11111 22222 33333 44444
14542 11111 22222 33333 44444
14552 Expand each frame to full height, but pad alternate lines with black,
14553 generating a frame with double height at the same input frame rate.
14558 Frame 1 Frame 2 Frame 3 Frame 4
14560 11111 22222 33333 44444
14561 11111 22222 33333 44444
14562 11111 22222 33333 44444
14563 11111 22222 33333 44444
14566 11111 ..... 33333 .....
14567 ..... 22222 ..... 44444
14568 11111 ..... 33333 .....
14569 ..... 22222 ..... 44444
14570 11111 ..... 33333 .....
14571 ..... 22222 ..... 44444
14572 11111 ..... 33333 .....
14573 ..... 22222 ..... 44444
14577 @item interleave_top, 4
14578 Interleave the upper field from odd frames with the lower field from
14579 even frames, generating a frame with unchanged height at half frame rate.
14584 Frame 1 Frame 2 Frame 3 Frame 4
14586 11111<- 22222 33333<- 44444
14587 11111 22222<- 33333 44444<-
14588 11111<- 22222 33333<- 44444
14589 11111 22222<- 33333 44444<-
14599 @item interleave_bottom, 5
14600 Interleave the lower field from odd frames with the upper field from
14601 even frames, generating a frame with unchanged height at half frame rate.
14606 Frame 1 Frame 2 Frame 3 Frame 4
14608 11111 22222<- 33333 44444<-
14609 11111<- 22222 33333<- 44444
14610 11111 22222<- 33333 44444<-
14611 11111<- 22222 33333<- 44444
14621 @item interlacex2, 6
14622 Double frame rate with unchanged height. Frames are inserted each
14623 containing the second temporal field from the previous input frame and
14624 the first temporal field from the next input frame. This mode relies on
14625 the top_field_first flag. Useful for interlaced video displays with no
14626 field synchronisation.
14631 Frame 1 Frame 2 Frame 3 Frame 4
14633 11111 22222 33333 44444
14634 11111 22222 33333 44444
14635 11111 22222 33333 44444
14636 11111 22222 33333 44444
14639 11111 22222 22222 33333 33333 44444 44444
14640 11111 11111 22222 22222 33333 33333 44444
14641 11111 22222 22222 33333 33333 44444 44444
14642 11111 11111 22222 22222 33333 33333 44444
14647 Move odd frames into the upper field, even into the lower field,
14648 generating a double height frame at same frame rate.
14653 Frame 1 Frame 2 Frame 3 Frame 4
14655 11111 22222 33333 44444
14656 11111 22222 33333 44444
14657 11111 22222 33333 44444
14658 11111 22222 33333 44444
14661 11111 33333 33333 55555
14662 22222 22222 44444 44444
14663 11111 33333 33333 55555
14664 22222 22222 44444 44444
14665 11111 33333 33333 55555
14666 22222 22222 44444 44444
14667 11111 33333 33333 55555
14668 22222 22222 44444 44444
14673 Numeric values are deprecated but are accepted for backward
14674 compatibility reasons.
14676 Default mode is @code{merge}.
14679 Specify flags influencing the filter process.
14681 Available value for @var{flags} is:
14684 @item low_pass_filter, vlfp
14685 Enable linear vertical low-pass filtering in the filter.
14686 Vertical low-pass filtering is required when creating an interlaced
14687 destination from a progressive source which contains high-frequency
14688 vertical detail. Filtering will reduce interlace 'twitter' and Moire
14691 @item complex_filter, cvlfp
14692 Enable complex vertical low-pass filtering.
14693 This will slightly less reduce interlace 'twitter' and Moire
14694 patterning but better retain detail and subjective sharpness impression.
14698 Vertical low-pass filtering can only be enabled for @option{mode}
14699 @var{interleave_top} and @var{interleave_bottom}.
14704 Tone map colors from different dynamic ranges.
14706 This filter expects data in single precision floating point, as it needs to
14707 operate on (and can output) out-of-range values. Another filter, such as
14708 @ref{zscale}, is needed to convert the resulting frame to a usable format.
14710 The tonemapping algorithms implemented only work on linear light, so input
14711 data should be linearized beforehand (and possibly correctly tagged).
14714 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
14717 @subsection Options
14718 The filter accepts the following options.
14722 Set the tone map algorithm to use.
14724 Possible values are:
14727 Do not apply any tone map, only desaturate overbright pixels.
14730 Hard-clip any out-of-range values. Use it for perfect color accuracy for
14731 in-range values, while distorting out-of-range values.
14734 Stretch the entire reference gamut to a linear multiple of the display.
14737 Fit a logarithmic transfer between the tone curves.
14740 Preserve overall image brightness with a simple curve, using nonlinear
14741 contrast, which results in flattening details and degrading color accuracy.
14744 Preserve both dark and bright details better than @var{reinhard}, at the cost
14745 of slightly darkening everything. Use it when detail preservation is more
14746 important than color and brightness accuracy.
14749 Smoothly map out-of-range values, while retaining contrast and colors for
14750 in-range material as much as possible. Use it when color accuracy is more
14751 important than detail preservation.
14757 Tune the tone mapping algorithm.
14759 This affects the following algorithms:
14765 Specifies the scale factor to use while stretching.
14769 Specifies the exponent of the function.
14773 Specify an extra linear coefficient to multiply into the signal before clipping.
14777 Specify the local contrast coefficient at the display peak.
14778 Default to 0.5, which means that in-gamut values will be about half as bright
14785 Specify the transition point from linear to mobius transform. Every value
14786 below this point is guaranteed to be mapped 1:1. The higher the value, the
14787 more accurate the result will be, at the cost of losing bright details.
14788 Default to 0.3, which due to the steep initial slope still preserves in-range
14789 colors fairly accurately.
14793 Apply desaturation for highlights that exceed this level of brightness. The
14794 higher the parameter, the more color information will be preserved. This
14795 setting helps prevent unnaturally blown-out colors for super-highlights, by
14796 (smoothly) turning into white instead. This makes images feel more natural,
14797 at the cost of reducing information about out-of-range colors.
14799 The default of 2.0 is somewhat conservative and will mostly just apply to
14800 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
14802 This option works only if the input frame has a supported color tag.
14805 Override signal/nominal/reference peak with this value. Useful when the
14806 embedded peak information in display metadata is not reliable or when tone
14807 mapping from a lower range to a higher range.
14812 Transpose rows with columns in the input video and optionally flip it.
14814 It accepts the following parameters:
14819 Specify the transposition direction.
14821 Can assume the following values:
14823 @item 0, 4, cclock_flip
14824 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
14832 Rotate by 90 degrees clockwise, that is:
14840 Rotate by 90 degrees counterclockwise, that is:
14847 @item 3, 7, clock_flip
14848 Rotate by 90 degrees clockwise and vertically flip, that is:
14856 For values between 4-7, the transposition is only done if the input
14857 video geometry is portrait and not landscape. These values are
14858 deprecated, the @code{passthrough} option should be used instead.
14860 Numerical values are deprecated, and should be dropped in favor of
14861 symbolic constants.
14864 Do not apply the transposition if the input geometry matches the one
14865 specified by the specified value. It accepts the following values:
14868 Always apply transposition.
14870 Preserve portrait geometry (when @var{height} >= @var{width}).
14872 Preserve landscape geometry (when @var{width} >= @var{height}).
14875 Default value is @code{none}.
14878 For example to rotate by 90 degrees clockwise and preserve portrait
14881 transpose=dir=1:passthrough=portrait
14884 The command above can also be specified as:
14886 transpose=1:portrait
14890 Trim the input so that the output contains one continuous subpart of the input.
14892 It accepts the following parameters:
14895 Specify the time of the start of the kept section, i.e. the frame with the
14896 timestamp @var{start} will be the first frame in the output.
14899 Specify the time of the first frame that will be dropped, i.e. the frame
14900 immediately preceding the one with the timestamp @var{end} will be the last
14901 frame in the output.
14904 This is the same as @var{start}, except this option sets the start timestamp
14905 in timebase units instead of seconds.
14908 This is the same as @var{end}, except this option sets the end timestamp
14909 in timebase units instead of seconds.
14912 The maximum duration of the output in seconds.
14915 The number of the first frame that should be passed to the output.
14918 The number of the first frame that should be dropped.
14921 @option{start}, @option{end}, and @option{duration} are expressed as time
14922 duration specifications; see
14923 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14924 for the accepted syntax.
14926 Note that the first two sets of the start/end options and the @option{duration}
14927 option look at the frame timestamp, while the _frame variants simply count the
14928 frames that pass through the filter. Also note that this filter does not modify
14929 the timestamps. If you wish for the output timestamps to start at zero, insert a
14930 setpts filter after the trim filter.
14932 If multiple start or end options are set, this filter tries to be greedy and
14933 keep all the frames that match at least one of the specified constraints. To keep
14934 only the part that matches all the constraints at once, chain multiple trim
14937 The defaults are such that all the input is kept. So it is possible to set e.g.
14938 just the end values to keep everything before the specified time.
14943 Drop everything except the second minute of input:
14945 ffmpeg -i INPUT -vf trim=60:120
14949 Keep only the first second:
14951 ffmpeg -i INPUT -vf trim=duration=1
14956 @section unpremultiply
14957 Apply alpha unpremultiply effect to input video stream using first plane
14958 of second stream as alpha.
14960 Both streams must have same dimensions and same pixel format.
14962 The filter accepts the following option:
14966 Set which planes will be processed, unprocessed planes will be copied.
14967 By default value 0xf, all planes will be processed.
14969 If the format has 1 or 2 components, then luma is bit 0.
14970 If the format has 3 or 4 components:
14971 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
14972 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
14973 If present, the alpha channel is always the last bit.
14976 Do not require 2nd input for processing, instead use alpha plane from input stream.
14982 Sharpen or blur the input video.
14984 It accepts the following parameters:
14987 @item luma_msize_x, lx
14988 Set the luma matrix horizontal size. It must be an odd integer between
14989 3 and 23. The default value is 5.
14991 @item luma_msize_y, ly
14992 Set the luma matrix vertical size. It must be an odd integer between 3
14993 and 23. The default value is 5.
14995 @item luma_amount, la
14996 Set the luma effect strength. It must be a floating point number, reasonable
14997 values lay between -1.5 and 1.5.
14999 Negative values will blur the input video, while positive values will
15000 sharpen it, a value of zero will disable the effect.
15002 Default value is 1.0.
15004 @item chroma_msize_x, cx
15005 Set the chroma matrix horizontal size. It must be an odd integer
15006 between 3 and 23. The default value is 5.
15008 @item chroma_msize_y, cy
15009 Set the chroma matrix vertical size. It must be an odd integer
15010 between 3 and 23. The default value is 5.
15012 @item chroma_amount, ca
15013 Set the chroma effect strength. It must be a floating point number, reasonable
15014 values lay between -1.5 and 1.5.
15016 Negative values will blur the input video, while positive values will
15017 sharpen it, a value of zero will disable the effect.
15019 Default value is 0.0.
15022 If set to 1, specify using OpenCL capabilities, only available if
15023 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
15027 All parameters are optional and default to the equivalent of the
15028 string '5:5:1.0:5:5:0.0'.
15030 @subsection Examples
15034 Apply strong luma sharpen effect:
15036 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15040 Apply a strong blur of both luma and chroma parameters:
15042 unsharp=7:7:-2:7:7:-2
15048 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15049 the image at several (or - in the case of @option{quality} level @code{8} - all)
15050 shifts and average the results.
15052 The way this differs from the behavior of spp is that uspp actually encodes &
15053 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15054 DCT similar to MJPEG.
15056 The filter accepts the following options:
15060 Set quality. This option defines the number of levels for averaging. It accepts
15061 an integer in the range 0-8. If set to @code{0}, the filter will have no
15062 effect. A value of @code{8} means the higher quality. For each increment of
15063 that value the speed drops by a factor of approximately 2. Default value is
15067 Force a constant quantization parameter. If not set, the filter will use the QP
15068 from the video stream (if available).
15071 @section vaguedenoiser
15073 Apply a wavelet based denoiser.
15075 It transforms each frame from the video input into the wavelet domain,
15076 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15077 the obtained coefficients. It does an inverse wavelet transform after.
15078 Due to wavelet properties, it should give a nice smoothed result, and
15079 reduced noise, without blurring picture features.
15081 This filter accepts the following options:
15085 The filtering strength. The higher, the more filtered the video will be.
15086 Hard thresholding can use a higher threshold than soft thresholding
15087 before the video looks overfiltered. Default value is 2.
15090 The filtering method the filter will use.
15092 It accepts the following values:
15095 All values under the threshold will be zeroed.
15098 All values under the threshold will be zeroed. All values above will be
15099 reduced by the threshold.
15102 Scales or nullifies coefficients - intermediary between (more) soft and
15103 (less) hard thresholding.
15106 Default is garrote.
15109 Number of times, the wavelet will decompose the picture. Picture can't
15110 be decomposed beyond a particular point (typically, 8 for a 640x480
15111 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15114 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15117 A list of the planes to process. By default all planes are processed.
15120 @section vectorscope
15122 Display 2 color component values in the two dimensional graph (which is called
15125 This filter accepts the following options:
15129 Set vectorscope mode.
15131 It accepts the following values:
15134 Gray values are displayed on graph, higher brightness means more pixels have
15135 same component color value on location in graph. This is the default mode.
15138 Gray values are displayed on graph. Surrounding pixels values which are not
15139 present in video frame are drawn in gradient of 2 color components which are
15140 set by option @code{x} and @code{y}. The 3rd color component is static.
15143 Actual color components values present in video frame are displayed on graph.
15146 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15147 on graph increases value of another color component, which is luminance by
15148 default values of @code{x} and @code{y}.
15151 Actual colors present in video frame are displayed on graph. If two different
15152 colors map to same position on graph then color with higher value of component
15153 not present in graph is picked.
15156 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15157 component picked from radial gradient.
15161 Set which color component will be represented on X-axis. Default is @code{1}.
15164 Set which color component will be represented on Y-axis. Default is @code{2}.
15167 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15168 of color component which represents frequency of (X, Y) location in graph.
15173 No envelope, this is default.
15176 Instant envelope, even darkest single pixel will be clearly highlighted.
15179 Hold maximum and minimum values presented in graph over time. This way you
15180 can still spot out of range values without constantly looking at vectorscope.
15183 Peak and instant envelope combined together.
15187 Set what kind of graticule to draw.
15195 Set graticule opacity.
15198 Set graticule flags.
15202 Draw graticule for white point.
15205 Draw graticule for black point.
15208 Draw color points short names.
15212 Set background opacity.
15214 @item lthreshold, l
15215 Set low threshold for color component not represented on X or Y axis.
15216 Values lower than this value will be ignored. Default is 0.
15217 Note this value is multiplied with actual max possible value one pixel component
15218 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15221 @item hthreshold, h
15222 Set high threshold for color component not represented on X or Y axis.
15223 Values higher than this value will be ignored. Default is 1.
15224 Note this value is multiplied with actual max possible value one pixel component
15225 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15226 is 0.9 * 255 = 230.
15228 @item colorspace, c
15229 Set what kind of colorspace to use when drawing graticule.
15238 @anchor{vidstabdetect}
15239 @section vidstabdetect
15241 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
15242 @ref{vidstabtransform} for pass 2.
15244 This filter generates a file with relative translation and rotation
15245 transform information about subsequent frames, which is then used by
15246 the @ref{vidstabtransform} filter.
15248 To enable compilation of this filter you need to configure FFmpeg with
15249 @code{--enable-libvidstab}.
15251 This filter accepts the following options:
15255 Set the path to the file used to write the transforms information.
15256 Default value is @file{transforms.trf}.
15259 Set how shaky the video is and how quick the camera is. It accepts an
15260 integer in the range 1-10, a value of 1 means little shakiness, a
15261 value of 10 means strong shakiness. Default value is 5.
15264 Set the accuracy of the detection process. It must be a value in the
15265 range 1-15. A value of 1 means low accuracy, a value of 15 means high
15266 accuracy. Default value is 15.
15269 Set stepsize of the search process. The region around minimum is
15270 scanned with 1 pixel resolution. Default value is 6.
15273 Set minimum contrast. Below this value a local measurement field is
15274 discarded. Must be a floating point value in the range 0-1. Default
15278 Set reference frame number for tripod mode.
15280 If enabled, the motion of the frames is compared to a reference frame
15281 in the filtered stream, identified by the specified number. The idea
15282 is to compensate all movements in a more-or-less static scene and keep
15283 the camera view absolutely still.
15285 If set to 0, it is disabled. The frames are counted starting from 1.
15288 Show fields and transforms in the resulting frames. It accepts an
15289 integer in the range 0-2. Default value is 0, which disables any
15293 @subsection Examples
15297 Use default values:
15303 Analyze strongly shaky movie and put the results in file
15304 @file{mytransforms.trf}:
15306 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
15310 Visualize the result of internal transformations in the resulting
15313 vidstabdetect=show=1
15317 Analyze a video with medium shakiness using @command{ffmpeg}:
15319 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
15323 @anchor{vidstabtransform}
15324 @section vidstabtransform
15326 Video stabilization/deshaking: pass 2 of 2,
15327 see @ref{vidstabdetect} for pass 1.
15329 Read a file with transform information for each frame and
15330 apply/compensate them. Together with the @ref{vidstabdetect}
15331 filter this can be used to deshake videos. See also
15332 @url{http://public.hronopik.de/vid.stab}. It is important to also use
15333 the @ref{unsharp} filter, see below.
15335 To enable compilation of this filter you need to configure FFmpeg with
15336 @code{--enable-libvidstab}.
15338 @subsection Options
15342 Set path to the file used to read the transforms. Default value is
15343 @file{transforms.trf}.
15346 Set the number of frames (value*2 + 1) used for lowpass filtering the
15347 camera movements. Default value is 10.
15349 For example a number of 10 means that 21 frames are used (10 in the
15350 past and 10 in the future) to smoothen the motion in the video. A
15351 larger value leads to a smoother video, but limits the acceleration of
15352 the camera (pan/tilt movements). 0 is a special case where a static
15353 camera is simulated.
15356 Set the camera path optimization algorithm.
15358 Accepted values are:
15361 gaussian kernel low-pass filter on camera motion (default)
15363 averaging on transformations
15367 Set maximal number of pixels to translate frames. Default value is -1,
15371 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
15372 value is -1, meaning no limit.
15375 Specify how to deal with borders that may be visible due to movement
15378 Available values are:
15381 keep image information from previous frame (default)
15383 fill the border black
15387 Invert transforms if set to 1. Default value is 0.
15390 Consider transforms as relative to previous frame if set to 1,
15391 absolute if set to 0. Default value is 0.
15394 Set percentage to zoom. A positive value will result in a zoom-in
15395 effect, a negative value in a zoom-out effect. Default value is 0 (no
15399 Set optimal zooming to avoid borders.
15401 Accepted values are:
15406 optimal static zoom value is determined (only very strong movements
15407 will lead to visible borders) (default)
15409 optimal adaptive zoom value is determined (no borders will be
15410 visible), see @option{zoomspeed}
15413 Note that the value given at zoom is added to the one calculated here.
15416 Set percent to zoom maximally each frame (enabled when
15417 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
15421 Specify type of interpolation.
15423 Available values are:
15428 linear only horizontal
15430 linear in both directions (default)
15432 cubic in both directions (slow)
15436 Enable virtual tripod mode if set to 1, which is equivalent to
15437 @code{relative=0:smoothing=0}. Default value is 0.
15439 Use also @code{tripod} option of @ref{vidstabdetect}.
15442 Increase log verbosity if set to 1. Also the detected global motions
15443 are written to the temporary file @file{global_motions.trf}. Default
15447 @subsection Examples
15451 Use @command{ffmpeg} for a typical stabilization with default values:
15453 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
15456 Note the use of the @ref{unsharp} filter which is always recommended.
15459 Zoom in a bit more and load transform data from a given file:
15461 vidstabtransform=zoom=5:input="mytransforms.trf"
15465 Smoothen the video even more:
15467 vidstabtransform=smoothing=30
15473 Flip the input video vertically.
15475 For example, to vertically flip a video with @command{ffmpeg}:
15477 ffmpeg -i in.avi -vf "vflip" out.avi
15483 Make or reverse a natural vignetting effect.
15485 The filter accepts the following options:
15489 Set lens angle expression as a number of radians.
15491 The value is clipped in the @code{[0,PI/2]} range.
15493 Default value: @code{"PI/5"}
15497 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
15501 Set forward/backward mode.
15503 Available modes are:
15506 The larger the distance from the central point, the darker the image becomes.
15509 The larger the distance from the central point, the brighter the image becomes.
15510 This can be used to reverse a vignette effect, though there is no automatic
15511 detection to extract the lens @option{angle} and other settings (yet). It can
15512 also be used to create a burning effect.
15515 Default value is @samp{forward}.
15518 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
15520 It accepts the following values:
15523 Evaluate expressions only once during the filter initialization.
15526 Evaluate expressions for each incoming frame. This is way slower than the
15527 @samp{init} mode since it requires all the scalers to be re-computed, but it
15528 allows advanced dynamic expressions.
15531 Default value is @samp{init}.
15534 Set dithering to reduce the circular banding effects. Default is @code{1}
15538 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
15539 Setting this value to the SAR of the input will make a rectangular vignetting
15540 following the dimensions of the video.
15542 Default is @code{1/1}.
15545 @subsection Expressions
15547 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
15548 following parameters.
15553 input width and height
15556 the number of input frame, starting from 0
15559 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
15560 @var{TB} units, NAN if undefined
15563 frame rate of the input video, NAN if the input frame rate is unknown
15566 the PTS (Presentation TimeStamp) of the filtered video frame,
15567 expressed in seconds, NAN if undefined
15570 time base of the input video
15574 @subsection Examples
15578 Apply simple strong vignetting effect:
15584 Make a flickering vignetting:
15586 vignette='PI/4+random(1)*PI/50':eval=frame
15591 @section vmafmotion
15593 Obtain the average vmaf motion score of a video.
15594 It is one of the component filters of VMAF.
15596 The obtained average motion score is printed through the logging system.
15598 In the below example the input file @file{ref.mpg} is being processed and score
15602 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
15606 Stack input videos vertically.
15608 All streams must be of same pixel format and of same width.
15610 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
15611 to create same output.
15613 The filter accept the following option:
15617 Set number of input streams. Default is 2.
15620 If set to 1, force the output to terminate when the shortest input
15621 terminates. Default value is 0.
15626 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
15627 Deinterlacing Filter").
15629 Based on the process described by Martin Weston for BBC R&D, and
15630 implemented based on the de-interlace algorithm written by Jim
15631 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
15632 uses filter coefficients calculated by BBC R&D.
15634 There are two sets of filter coefficients, so called "simple":
15635 and "complex". Which set of filter coefficients is used can
15636 be set by passing an optional parameter:
15640 Set the interlacing filter coefficients. Accepts one of the following values:
15644 Simple filter coefficient set.
15646 More-complex filter coefficient set.
15648 Default value is @samp{complex}.
15651 Specify which frames to deinterlace. Accept one of the following values:
15655 Deinterlace all frames,
15657 Only deinterlace frames marked as interlaced.
15660 Default value is @samp{all}.
15664 Video waveform monitor.
15666 The waveform monitor plots color component intensity. By default luminance
15667 only. Each column of the waveform corresponds to a column of pixels in the
15670 It accepts the following options:
15674 Can be either @code{row}, or @code{column}. Default is @code{column}.
15675 In row mode, the graph on the left side represents color component value 0 and
15676 the right side represents value = 255. In column mode, the top side represents
15677 color component value = 0 and bottom side represents value = 255.
15680 Set intensity. Smaller values are useful to find out how many values of the same
15681 luminance are distributed across input rows/columns.
15682 Default value is @code{0.04}. Allowed range is [0, 1].
15685 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
15686 In mirrored mode, higher values will be represented on the left
15687 side for @code{row} mode and at the top for @code{column} mode. Default is
15688 @code{1} (mirrored).
15692 It accepts the following values:
15695 Presents information identical to that in the @code{parade}, except
15696 that the graphs representing color components are superimposed directly
15699 This display mode makes it easier to spot relative differences or similarities
15700 in overlapping areas of the color components that are supposed to be identical,
15701 such as neutral whites, grays, or blacks.
15704 Display separate graph for the color components side by side in
15705 @code{row} mode or one below the other in @code{column} mode.
15708 Display separate graph for the color components side by side in
15709 @code{column} mode or one below the other in @code{row} mode.
15711 Using this display mode makes it easy to spot color casts in the highlights
15712 and shadows of an image, by comparing the contours of the top and the bottom
15713 graphs of each waveform. Since whites, grays, and blacks are characterized
15714 by exactly equal amounts of red, green, and blue, neutral areas of the picture
15715 should display three waveforms of roughly equal width/height. If not, the
15716 correction is easy to perform by making level adjustments the three waveforms.
15718 Default is @code{stack}.
15720 @item components, c
15721 Set which color components to display. Default is 1, which means only luminance
15722 or red color component if input is in RGB colorspace. If is set for example to
15723 7 it will display all 3 (if) available color components.
15728 No envelope, this is default.
15731 Instant envelope, minimum and maximum values presented in graph will be easily
15732 visible even with small @code{step} value.
15735 Hold minimum and maximum values presented in graph across time. This way you
15736 can still spot out of range values without constantly looking at waveforms.
15739 Peak and instant envelope combined together.
15745 No filtering, this is default.
15748 Luma and chroma combined together.
15751 Similar as above, but shows difference between blue and red chroma.
15754 Displays only chroma.
15757 Displays actual color value on waveform.
15760 Similar as above, but with luma showing frequency of chroma values.
15764 Set which graticule to display.
15768 Do not display graticule.
15771 Display green graticule showing legal broadcast ranges.
15775 Set graticule opacity.
15778 Set graticule flags.
15782 Draw numbers above lines. By default enabled.
15785 Draw dots instead of lines.
15789 Set scale used for displaying graticule.
15796 Default is digital.
15799 Set background opacity.
15802 @section weave, doubleweave
15804 The @code{weave} takes a field-based video input and join
15805 each two sequential fields into single frame, producing a new double
15806 height clip with half the frame rate and half the frame count.
15808 The @code{doubleweave} works same as @code{weave} but without
15809 halving frame rate and frame count.
15811 It accepts the following option:
15815 Set first field. Available values are:
15819 Set the frame as top-field-first.
15822 Set the frame as bottom-field-first.
15826 @subsection Examples
15830 Interlace video using @ref{select} and @ref{separatefields} filter:
15832 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
15837 Apply the xBR high-quality magnification filter which is designed for pixel
15838 art. It follows a set of edge-detection rules, see
15839 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
15841 It accepts the following option:
15845 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
15846 @code{3xBR} and @code{4} for @code{4xBR}.
15847 Default is @code{3}.
15853 Deinterlace the input video ("yadif" means "yet another deinterlacing
15856 It accepts the following parameters:
15862 The interlacing mode to adopt. It accepts one of the following values:
15865 @item 0, send_frame
15866 Output one frame for each frame.
15867 @item 1, send_field
15868 Output one frame for each field.
15869 @item 2, send_frame_nospatial
15870 Like @code{send_frame}, but it skips the spatial interlacing check.
15871 @item 3, send_field_nospatial
15872 Like @code{send_field}, but it skips the spatial interlacing check.
15875 The default value is @code{send_frame}.
15878 The picture field parity assumed for the input interlaced video. It accepts one
15879 of the following values:
15883 Assume the top field is first.
15885 Assume the bottom field is first.
15887 Enable automatic detection of field parity.
15890 The default value is @code{auto}.
15891 If the interlacing is unknown or the decoder does not export this information,
15892 top field first will be assumed.
15895 Specify which frames to deinterlace. Accept one of the following
15900 Deinterlace all frames.
15901 @item 1, interlaced
15902 Only deinterlace frames marked as interlaced.
15905 The default value is @code{all}.
15910 Apply Zoom & Pan effect.
15912 This filter accepts the following options:
15916 Set the zoom expression. Default is 1.
15920 Set the x and y expression. Default is 0.
15923 Set the duration expression in number of frames.
15924 This sets for how many number of frames effect will last for
15925 single input image.
15928 Set the output image size, default is 'hd720'.
15931 Set the output frame rate, default is '25'.
15934 Each expression can contain the following constants:
15953 Output frame count.
15957 Last calculated 'x' and 'y' position from 'x' and 'y' expression
15958 for current input frame.
15962 'x' and 'y' of last output frame of previous input frame or 0 when there was
15963 not yet such frame (first input frame).
15966 Last calculated zoom from 'z' expression for current input frame.
15969 Last calculated zoom of last output frame of previous input frame.
15972 Number of output frames for current input frame. Calculated from 'd' expression
15973 for each input frame.
15976 number of output frames created for previous input frame
15979 Rational number: input width / input height
15982 sample aspect ratio
15985 display aspect ratio
15989 @subsection Examples
15993 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
15995 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
15999 Zoom-in up to 1.5 and pan always at center of picture:
16001 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16005 Same as above but without pausing:
16007 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16013 Scale (resize) the input video, using the z.lib library:
16014 https://github.com/sekrit-twc/zimg.
16016 The zscale filter forces the output display aspect ratio to be the same
16017 as the input, by changing the output sample aspect ratio.
16019 If the input image format is different from the format requested by
16020 the next filter, the zscale filter will convert the input to the
16023 @subsection Options
16024 The filter accepts the following options.
16029 Set the output video dimension expression. Default value is the input
16032 If the @var{width} or @var{w} value is 0, the input width is used for
16033 the output. If the @var{height} or @var{h} value is 0, the input height
16034 is used for the output.
16036 If one and only one of the values is -n with n >= 1, the zscale filter
16037 will use a value that maintains the aspect ratio of the input image,
16038 calculated from the other specified dimension. After that it will,
16039 however, make sure that the calculated dimension is divisible by n and
16040 adjust the value if necessary.
16042 If both values are -n with n >= 1, the behavior will be identical to
16043 both values being set to 0 as previously detailed.
16045 See below for the list of accepted constants for use in the dimension
16049 Set the video size. For the syntax of this option, check the
16050 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16053 Set the dither type.
16055 Possible values are:
16060 @item error_diffusion
16066 Set the resize filter type.
16068 Possible values are:
16078 Default is bilinear.
16081 Set the color range.
16083 Possible values are:
16090 Default is same as input.
16093 Set the color primaries.
16095 Possible values are:
16105 Default is same as input.
16108 Set the transfer characteristics.
16110 Possible values are:
16124 Default is same as input.
16127 Set the colorspace matrix.
16129 Possible value are:
16140 Default is same as input.
16143 Set the input color range.
16145 Possible values are:
16152 Default is same as input.
16154 @item primariesin, pin
16155 Set the input color primaries.
16157 Possible values are:
16167 Default is same as input.
16169 @item transferin, tin
16170 Set the input transfer characteristics.
16172 Possible values are:
16183 Default is same as input.
16185 @item matrixin, min
16186 Set the input colorspace matrix.
16188 Possible value are:
16200 Set the output chroma location.
16202 Possible values are:
16213 @item chromalin, cin
16214 Set the input chroma location.
16216 Possible values are:
16228 Set the nominal peak luminance.
16231 The values of the @option{w} and @option{h} options are expressions
16232 containing the following constants:
16237 The input width and height
16241 These are the same as @var{in_w} and @var{in_h}.
16245 The output (scaled) width and height
16249 These are the same as @var{out_w} and @var{out_h}
16252 The same as @var{iw} / @var{ih}
16255 input sample aspect ratio
16258 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
16262 horizontal and vertical input chroma subsample values. For example for the
16263 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16267 horizontal and vertical output chroma subsample values. For example for the
16268 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16274 @c man end VIDEO FILTERS
16276 @chapter Video Sources
16277 @c man begin VIDEO SOURCES
16279 Below is a description of the currently available video sources.
16283 Buffer video frames, and make them available to the filter chain.
16285 This source is mainly intended for a programmatic use, in particular
16286 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
16288 It accepts the following parameters:
16293 Specify the size (width and height) of the buffered video frames. For the
16294 syntax of this option, check the
16295 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16298 The input video width.
16301 The input video height.
16304 A string representing the pixel format of the buffered video frames.
16305 It may be a number corresponding to a pixel format, or a pixel format
16309 Specify the timebase assumed by the timestamps of the buffered frames.
16312 Specify the frame rate expected for the video stream.
16314 @item pixel_aspect, sar
16315 The sample (pixel) aspect ratio of the input video.
16318 Specify the optional parameters to be used for the scale filter which
16319 is automatically inserted when an input change is detected in the
16320 input size or format.
16322 @item hw_frames_ctx
16323 When using a hardware pixel format, this should be a reference to an
16324 AVHWFramesContext describing input frames.
16329 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
16332 will instruct the source to accept video frames with size 320x240 and
16333 with format "yuv410p", assuming 1/24 as the timestamps timebase and
16334 square pixels (1:1 sample aspect ratio).
16335 Since the pixel format with name "yuv410p" corresponds to the number 6
16336 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
16337 this example corresponds to:
16339 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
16342 Alternatively, the options can be specified as a flat string, but this
16343 syntax is deprecated:
16345 @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}]
16349 Create a pattern generated by an elementary cellular automaton.
16351 The initial state of the cellular automaton can be defined through the
16352 @option{filename} and @option{pattern} options. If such options are
16353 not specified an initial state is created randomly.
16355 At each new frame a new row in the video is filled with the result of
16356 the cellular automaton next generation. The behavior when the whole
16357 frame is filled is defined by the @option{scroll} option.
16359 This source accepts the following options:
16363 Read the initial cellular automaton state, i.e. the starting row, from
16364 the specified file.
16365 In the file, each non-whitespace character is considered an alive
16366 cell, a newline will terminate the row, and further characters in the
16367 file will be ignored.
16370 Read the initial cellular automaton state, i.e. the starting row, from
16371 the specified string.
16373 Each non-whitespace character in the string is considered an alive
16374 cell, a newline will terminate the row, and further characters in the
16375 string will be ignored.
16378 Set the video rate, that is the number of frames generated per second.
16381 @item random_fill_ratio, ratio
16382 Set the random fill ratio for the initial cellular automaton row. It
16383 is a floating point number value ranging from 0 to 1, defaults to
16386 This option is ignored when a file or a pattern is specified.
16388 @item random_seed, seed
16389 Set the seed for filling randomly the initial row, must be an integer
16390 included between 0 and UINT32_MAX. If not specified, or if explicitly
16391 set to -1, the filter will try to use a good random seed on a best
16395 Set the cellular automaton rule, it is a number ranging from 0 to 255.
16396 Default value is 110.
16399 Set the size of the output video. For the syntax of this option, check the
16400 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16402 If @option{filename} or @option{pattern} is specified, the size is set
16403 by default to the width of the specified initial state row, and the
16404 height is set to @var{width} * PHI.
16406 If @option{size} is set, it must contain the width of the specified
16407 pattern string, and the specified pattern will be centered in the
16410 If a filename or a pattern string is not specified, the size value
16411 defaults to "320x518" (used for a randomly generated initial state).
16414 If set to 1, scroll the output upward when all the rows in the output
16415 have been already filled. If set to 0, the new generated row will be
16416 written over the top row just after the bottom row is filled.
16419 @item start_full, full
16420 If set to 1, completely fill the output with generated rows before
16421 outputting the first frame.
16422 This is the default behavior, for disabling set the value to 0.
16425 If set to 1, stitch the left and right row edges together.
16426 This is the default behavior, for disabling set the value to 0.
16429 @subsection Examples
16433 Read the initial state from @file{pattern}, and specify an output of
16436 cellauto=f=pattern:s=200x400
16440 Generate a random initial row with a width of 200 cells, with a fill
16443 cellauto=ratio=2/3:s=200x200
16447 Create a pattern generated by rule 18 starting by a single alive cell
16448 centered on an initial row with width 100:
16450 cellauto=p=@@:s=100x400:full=0:rule=18
16454 Specify a more elaborated initial pattern:
16456 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
16461 @anchor{coreimagesrc}
16462 @section coreimagesrc
16463 Video source generated on GPU using Apple's CoreImage API on OSX.
16465 This video source is a specialized version of the @ref{coreimage} video filter.
16466 Use a core image generator at the beginning of the applied filterchain to
16467 generate the content.
16469 The coreimagesrc video source accepts the following options:
16471 @item list_generators
16472 List all available generators along with all their respective options as well as
16473 possible minimum and maximum values along with the default values.
16475 list_generators=true
16479 Specify the size of the sourced video. For the syntax of this option, check the
16480 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16481 The default value is @code{320x240}.
16484 Specify the frame rate of the sourced video, as the number of frames
16485 generated per second. It has to be a string in the format
16486 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16487 number or a valid video frame rate abbreviation. The default value is
16491 Set the sample aspect ratio of the sourced video.
16494 Set the duration of the sourced video. See
16495 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16496 for the accepted syntax.
16498 If not specified, or the expressed duration is negative, the video is
16499 supposed to be generated forever.
16502 Additionally, all options of the @ref{coreimage} video filter are accepted.
16503 A complete filterchain can be used for further processing of the
16504 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
16505 and examples for details.
16507 @subsection Examples
16512 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
16513 given as complete and escaped command-line for Apple's standard bash shell:
16515 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
16517 This example is equivalent to the QRCode example of @ref{coreimage} without the
16518 need for a nullsrc video source.
16522 @section mandelbrot
16524 Generate a Mandelbrot set fractal, and progressively zoom towards the
16525 point specified with @var{start_x} and @var{start_y}.
16527 This source accepts the following options:
16532 Set the terminal pts value. Default value is 400.
16535 Set the terminal scale value.
16536 Must be a floating point value. Default value is 0.3.
16539 Set the inner coloring mode, that is the algorithm used to draw the
16540 Mandelbrot fractal internal region.
16542 It shall assume one of the following values:
16547 Show time until convergence.
16549 Set color based on point closest to the origin of the iterations.
16554 Default value is @var{mincol}.
16557 Set the bailout value. Default value is 10.0.
16560 Set the maximum of iterations performed by the rendering
16561 algorithm. Default value is 7189.
16564 Set outer coloring mode.
16565 It shall assume one of following values:
16567 @item iteration_count
16568 Set iteration cound mode.
16569 @item normalized_iteration_count
16570 set normalized iteration count mode.
16572 Default value is @var{normalized_iteration_count}.
16575 Set frame rate, expressed as number of frames per second. Default
16579 Set frame size. For the syntax of this option, check the "Video
16580 size" section in the ffmpeg-utils manual. Default value is "640x480".
16583 Set the initial scale value. Default value is 3.0.
16586 Set the initial x position. Must be a floating point value between
16587 -100 and 100. Default value is -0.743643887037158704752191506114774.
16590 Set the initial y position. Must be a floating point value between
16591 -100 and 100. Default value is -0.131825904205311970493132056385139.
16596 Generate various test patterns, as generated by the MPlayer test filter.
16598 The size of the generated video is fixed, and is 256x256.
16599 This source is useful in particular for testing encoding features.
16601 This source accepts the following options:
16606 Specify the frame rate of the sourced video, as the number of frames
16607 generated per second. It has to be a string in the format
16608 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16609 number or a valid video frame rate abbreviation. The default value is
16613 Set the duration of the sourced video. See
16614 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16615 for the accepted syntax.
16617 If not specified, or the expressed duration is negative, the video is
16618 supposed to be generated forever.
16622 Set the number or the name of the test to perform. Supported tests are:
16638 Default value is "all", which will cycle through the list of all tests.
16643 mptestsrc=t=dc_luma
16646 will generate a "dc_luma" test pattern.
16648 @section frei0r_src
16650 Provide a frei0r source.
16652 To enable compilation of this filter you need to install the frei0r
16653 header and configure FFmpeg with @code{--enable-frei0r}.
16655 This source accepts the following parameters:
16660 The size of the video to generate. For the syntax of this option, check the
16661 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16664 The framerate of the generated video. It may be a string of the form
16665 @var{num}/@var{den} or a frame rate abbreviation.
16668 The name to the frei0r source to load. For more information regarding frei0r and
16669 how to set the parameters, read the @ref{frei0r} section in the video filters
16672 @item filter_params
16673 A '|'-separated list of parameters to pass to the frei0r source.
16677 For example, to generate a frei0r partik0l source with size 200x200
16678 and frame rate 10 which is overlaid on the overlay filter main input:
16680 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
16685 Generate a life pattern.
16687 This source is based on a generalization of John Conway's life game.
16689 The sourced input represents a life grid, each pixel represents a cell
16690 which can be in one of two possible states, alive or dead. Every cell
16691 interacts with its eight neighbours, which are the cells that are
16692 horizontally, vertically, or diagonally adjacent.
16694 At each interaction the grid evolves according to the adopted rule,
16695 which specifies the number of neighbor alive cells which will make a
16696 cell stay alive or born. The @option{rule} option allows one to specify
16699 This source accepts the following options:
16703 Set the file from which to read the initial grid state. In the file,
16704 each non-whitespace character is considered an alive cell, and newline
16705 is used to delimit the end of each row.
16707 If this option is not specified, the initial grid is generated
16711 Set the video rate, that is the number of frames generated per second.
16714 @item random_fill_ratio, ratio
16715 Set the random fill ratio for the initial random grid. It is a
16716 floating point number value ranging from 0 to 1, defaults to 1/PHI.
16717 It is ignored when a file is specified.
16719 @item random_seed, seed
16720 Set the seed for filling the initial random grid, must be an integer
16721 included between 0 and UINT32_MAX. If not specified, or if explicitly
16722 set to -1, the filter will try to use a good random seed on a best
16728 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
16729 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
16730 @var{NS} specifies the number of alive neighbor cells which make a
16731 live cell stay alive, and @var{NB} the number of alive neighbor cells
16732 which make a dead cell to become alive (i.e. to "born").
16733 "s" and "b" can be used in place of "S" and "B", respectively.
16735 Alternatively a rule can be specified by an 18-bits integer. The 9
16736 high order bits are used to encode the next cell state if it is alive
16737 for each number of neighbor alive cells, the low order bits specify
16738 the rule for "borning" new cells. Higher order bits encode for an
16739 higher number of neighbor cells.
16740 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
16741 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
16743 Default value is "S23/B3", which is the original Conway's game of life
16744 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
16745 cells, and will born a new cell if there are three alive cells around
16749 Set the size of the output video. For the syntax of this option, check the
16750 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16752 If @option{filename} is specified, the size is set by default to the
16753 same size of the input file. If @option{size} is set, it must contain
16754 the size specified in the input file, and the initial grid defined in
16755 that file is centered in the larger resulting area.
16757 If a filename is not specified, the size value defaults to "320x240"
16758 (used for a randomly generated initial grid).
16761 If set to 1, stitch the left and right grid edges together, and the
16762 top and bottom edges also. Defaults to 1.
16765 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
16766 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
16767 value from 0 to 255.
16770 Set the color of living (or new born) cells.
16773 Set the color of dead cells. If @option{mold} is set, this is the first color
16774 used to represent a dead cell.
16777 Set mold color, for definitely dead and moldy cells.
16779 For the syntax of these 3 color options, check the "Color" section in the
16780 ffmpeg-utils manual.
16783 @subsection Examples
16787 Read a grid from @file{pattern}, and center it on a grid of size
16790 life=f=pattern:s=300x300
16794 Generate a random grid of size 200x200, with a fill ratio of 2/3:
16796 life=ratio=2/3:s=200x200
16800 Specify a custom rule for evolving a randomly generated grid:
16806 Full example with slow death effect (mold) using @command{ffplay}:
16808 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
16815 @anchor{haldclutsrc}
16817 @anchor{rgbtestsrc}
16819 @anchor{smptehdbars}
16822 @anchor{yuvtestsrc}
16823 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
16825 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
16827 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
16829 The @code{color} source provides an uniformly colored input.
16831 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
16832 @ref{haldclut} filter.
16834 The @code{nullsrc} source returns unprocessed video frames. It is
16835 mainly useful to be employed in analysis / debugging tools, or as the
16836 source for filters which ignore the input data.
16838 The @code{rgbtestsrc} source generates an RGB test pattern useful for
16839 detecting RGB vs BGR issues. You should see a red, green and blue
16840 stripe from top to bottom.
16842 The @code{smptebars} source generates a color bars pattern, based on
16843 the SMPTE Engineering Guideline EG 1-1990.
16845 The @code{smptehdbars} source generates a color bars pattern, based on
16846 the SMPTE RP 219-2002.
16848 The @code{testsrc} source generates a test video pattern, showing a
16849 color pattern, a scrolling gradient and a timestamp. This is mainly
16850 intended for testing purposes.
16852 The @code{testsrc2} source is similar to testsrc, but supports more
16853 pixel formats instead of just @code{rgb24}. This allows using it as an
16854 input for other tests without requiring a format conversion.
16856 The @code{yuvtestsrc} source generates an YUV test pattern. You should
16857 see a y, cb and cr stripe from top to bottom.
16859 The sources accept the following parameters:
16864 Specify the alpha (opacity) of the background, only available in the
16865 @code{testsrc2} source. The value must be between 0 (fully transparent) and
16866 255 (fully opaque, the default).
16869 Specify the color of the source, only available in the @code{color}
16870 source. For the syntax of this option, check the "Color" section in the
16871 ffmpeg-utils manual.
16874 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
16875 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
16876 pixels to be used as identity matrix for 3D lookup tables. Each component is
16877 coded on a @code{1/(N*N)} scale.
16880 Specify the size of the sourced video. For the syntax of this option, check the
16881 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16882 The default value is @code{320x240}.
16884 This option is not available with the @code{haldclutsrc} filter.
16887 Specify the frame rate of the sourced video, as the number of frames
16888 generated per second. It has to be a string in the format
16889 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
16890 number or a valid video frame rate abbreviation. The default value is
16894 Set the sample aspect ratio of the sourced video.
16897 Set the duration of the sourced video. See
16898 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16899 for the accepted syntax.
16901 If not specified, or the expressed duration is negative, the video is
16902 supposed to be generated forever.
16905 Set the number of decimals to show in the timestamp, only available in the
16906 @code{testsrc} source.
16908 The displayed timestamp value will correspond to the original
16909 timestamp value multiplied by the power of 10 of the specified
16910 value. Default value is 0.
16913 For example the following:
16915 testsrc=duration=5.3:size=qcif:rate=10
16918 will generate a video with a duration of 5.3 seconds, with size
16919 176x144 and a frame rate of 10 frames per second.
16921 The following graph description will generate a red source
16922 with an opacity of 0.2, with size "qcif" and a frame rate of 10
16925 color=c=red@@0.2:s=qcif:r=10
16928 If the input content is to be ignored, @code{nullsrc} can be used. The
16929 following command generates noise in the luminance plane by employing
16930 the @code{geq} filter:
16932 nullsrc=s=256x256, geq=random(1)*255:128:128
16935 @subsection Commands
16937 The @code{color} source supports the following commands:
16941 Set the color of the created image. Accepts the same syntax of the
16942 corresponding @option{color} option.
16945 @c man end VIDEO SOURCES
16947 @chapter Video Sinks
16948 @c man begin VIDEO SINKS
16950 Below is a description of the currently available video sinks.
16952 @section buffersink
16954 Buffer video frames, and make them available to the end of the filter
16957 This sink is mainly intended for programmatic use, in particular
16958 through the interface defined in @file{libavfilter/buffersink.h}
16959 or the options system.
16961 It accepts a pointer to an AVBufferSinkContext structure, which
16962 defines the incoming buffers' formats, to be passed as the opaque
16963 parameter to @code{avfilter_init_filter} for initialization.
16967 Null video sink: do absolutely nothing with the input video. It is
16968 mainly useful as a template and for use in analysis / debugging
16971 @c man end VIDEO SINKS
16973 @chapter Multimedia Filters
16974 @c man begin MULTIMEDIA FILTERS
16976 Below is a description of the currently available multimedia filters.
16980 Convert input audio to a video output, displaying the audio bit scope.
16982 The filter accepts the following options:
16986 Set frame rate, expressed as number of frames per second. Default
16990 Specify the video size for the output. For the syntax of this option, check the
16991 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16992 Default value is @code{1024x256}.
16995 Specify list of colors separated by space or by '|' which will be used to
16996 draw channels. Unrecognized or missing colors will be replaced
17000 @section ahistogram
17002 Convert input audio to a video output, displaying the volume histogram.
17004 The filter accepts the following options:
17008 Specify how histogram is calculated.
17010 It accepts the following values:
17013 Use single histogram for all channels.
17015 Use separate histogram for each channel.
17017 Default is @code{single}.
17020 Set frame rate, expressed as number of frames per second. Default
17024 Specify the video size for the output. For the syntax of this option, check the
17025 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17026 Default value is @code{hd720}.
17031 It accepts the following values:
17042 reverse logarithmic
17044 Default is @code{log}.
17047 Set amplitude scale.
17049 It accepts the following values:
17056 Default is @code{log}.
17059 Set how much frames to accumulate in histogram.
17060 Defauls is 1. Setting this to -1 accumulates all frames.
17063 Set histogram ratio of window height.
17066 Set sonogram sliding.
17068 It accepts the following values:
17071 replace old rows with new ones.
17073 scroll from top to bottom.
17075 Default is @code{replace}.
17078 @section aphasemeter
17080 Convert input audio to a video output, displaying the audio phase.
17082 The filter accepts the following options:
17086 Set the output frame rate. Default value is @code{25}.
17089 Set the video size for the output. For the syntax of this option, check the
17090 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17091 Default value is @code{800x400}.
17096 Specify the red, green, blue contrast. Default values are @code{2},
17097 @code{7} and @code{1}.
17098 Allowed range is @code{[0, 255]}.
17101 Set color which will be used for drawing median phase. If color is
17102 @code{none} which is default, no median phase value will be drawn.
17105 Enable video output. Default is enabled.
17108 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17109 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17110 The @code{-1} means left and right channels are completely out of phase and
17111 @code{1} means channels are in phase.
17113 @section avectorscope
17115 Convert input audio to a video output, representing the audio vector
17118 The filter is used to measure the difference between channels of stereo
17119 audio stream. A monoaural signal, consisting of identical left and right
17120 signal, results in straight vertical line. Any stereo separation is visible
17121 as a deviation from this line, creating a Lissajous figure.
17122 If the straight (or deviation from it) but horizontal line appears this
17123 indicates that the left and right channels are out of phase.
17125 The filter accepts the following options:
17129 Set the vectorscope mode.
17131 Available values are:
17134 Lissajous rotated by 45 degrees.
17137 Same as above but not rotated.
17140 Shape resembling half of circle.
17143 Default value is @samp{lissajous}.
17146 Set the video size for the output. For the syntax of this option, check the
17147 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17148 Default value is @code{400x400}.
17151 Set the output frame rate. Default value is @code{25}.
17157 Specify the red, green, blue and alpha contrast. Default values are @code{40},
17158 @code{160}, @code{80} and @code{255}.
17159 Allowed range is @code{[0, 255]}.
17165 Specify the red, green, blue and alpha fade. Default values are @code{15},
17166 @code{10}, @code{5} and @code{5}.
17167 Allowed range is @code{[0, 255]}.
17170 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
17171 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
17174 Set the vectorscope drawing mode.
17176 Available values are:
17179 Draw dot for each sample.
17182 Draw line between previous and current sample.
17185 Default value is @samp{dot}.
17188 Specify amplitude scale of audio samples.
17190 Available values are:
17207 @subsection Examples
17211 Complete example using @command{ffplay}:
17213 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17214 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
17218 @section bench, abench
17220 Benchmark part of a filtergraph.
17222 The filter accepts the following options:
17226 Start or stop a timer.
17228 Available values are:
17231 Get the current time, set it as frame metadata (using the key
17232 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
17235 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
17236 the input frame metadata to get the time difference. Time difference, average,
17237 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
17238 @code{min}) are then printed. The timestamps are expressed in seconds.
17242 @subsection Examples
17246 Benchmark @ref{selectivecolor} filter:
17248 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
17254 Concatenate audio and video streams, joining them together one after the
17257 The filter works on segments of synchronized video and audio streams. All
17258 segments must have the same number of streams of each type, and that will
17259 also be the number of streams at output.
17261 The filter accepts the following options:
17266 Set the number of segments. Default is 2.
17269 Set the number of output video streams, that is also the number of video
17270 streams in each segment. Default is 1.
17273 Set the number of output audio streams, that is also the number of audio
17274 streams in each segment. Default is 0.
17277 Activate unsafe mode: do not fail if segments have a different format.
17281 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
17282 @var{a} audio outputs.
17284 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
17285 segment, in the same order as the outputs, then the inputs for the second
17288 Related streams do not always have exactly the same duration, for various
17289 reasons including codec frame size or sloppy authoring. For that reason,
17290 related synchronized streams (e.g. a video and its audio track) should be
17291 concatenated at once. The concat filter will use the duration of the longest
17292 stream in each segment (except the last one), and if necessary pad shorter
17293 audio streams with silence.
17295 For this filter to work correctly, all segments must start at timestamp 0.
17297 All corresponding streams must have the same parameters in all segments; the
17298 filtering system will automatically select a common pixel format for video
17299 streams, and a common sample format, sample rate and channel layout for
17300 audio streams, but other settings, such as resolution, must be converted
17301 explicitly by the user.
17303 Different frame rates are acceptable but will result in variable frame rate
17304 at output; be sure to configure the output file to handle it.
17306 @subsection Examples
17310 Concatenate an opening, an episode and an ending, all in bilingual version
17311 (video in stream 0, audio in streams 1 and 2):
17313 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
17314 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
17315 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
17316 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
17320 Concatenate two parts, handling audio and video separately, using the
17321 (a)movie sources, and adjusting the resolution:
17323 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
17324 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
17325 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
17327 Note that a desync will happen at the stitch if the audio and video streams
17328 do not have exactly the same duration in the first file.
17332 @section drawgraph, adrawgraph
17334 Draw a graph using input video or audio metadata.
17336 It accepts the following parameters:
17340 Set 1st frame metadata key from which metadata values will be used to draw a graph.
17343 Set 1st foreground color expression.
17346 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
17349 Set 2nd foreground color expression.
17352 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
17355 Set 3rd foreground color expression.
17358 Set 4th frame metadata key from which metadata values will be used to draw a graph.
17361 Set 4th foreground color expression.
17364 Set minimal value of metadata value.
17367 Set maximal value of metadata value.
17370 Set graph background color. Default is white.
17375 Available values for mode is:
17382 Default is @code{line}.
17387 Available values for slide is:
17390 Draw new frame when right border is reached.
17393 Replace old columns with new ones.
17396 Scroll from right to left.
17399 Scroll from left to right.
17402 Draw single picture.
17405 Default is @code{frame}.
17408 Set size of graph video. For the syntax of this option, check the
17409 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17410 The default value is @code{900x256}.
17412 The foreground color expressions can use the following variables:
17415 Minimal value of metadata value.
17418 Maximal value of metadata value.
17421 Current metadata key value.
17424 The color is defined as 0xAABBGGRR.
17427 Example using metadata from @ref{signalstats} filter:
17429 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
17432 Example using metadata from @ref{ebur128} filter:
17434 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
17440 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
17441 it unchanged. By default, it logs a message at a frequency of 10Hz with the
17442 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
17443 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
17445 The filter also has a video output (see the @var{video} option) with a real
17446 time graph to observe the loudness evolution. The graphic contains the logged
17447 message mentioned above, so it is not printed anymore when this option is set,
17448 unless the verbose logging is set. The main graphing area contains the
17449 short-term loudness (3 seconds of analysis), and the gauge on the right is for
17450 the momentary loudness (400 milliseconds).
17452 More information about the Loudness Recommendation EBU R128 on
17453 @url{http://tech.ebu.ch/loudness}.
17455 The filter accepts the following options:
17460 Activate the video output. The audio stream is passed unchanged whether this
17461 option is set or no. The video stream will be the first output stream if
17462 activated. Default is @code{0}.
17465 Set the video size. This option is for video only. For the syntax of this
17467 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17468 Default and minimum resolution is @code{640x480}.
17471 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
17472 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
17473 other integer value between this range is allowed.
17476 Set metadata injection. If set to @code{1}, the audio input will be segmented
17477 into 100ms output frames, each of them containing various loudness information
17478 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
17480 Default is @code{0}.
17483 Force the frame logging level.
17485 Available values are:
17488 information logging level
17490 verbose logging level
17493 By default, the logging level is set to @var{info}. If the @option{video} or
17494 the @option{metadata} options are set, it switches to @var{verbose}.
17499 Available modes can be cumulated (the option is a @code{flag} type). Possible
17503 Disable any peak mode (default).
17505 Enable sample-peak mode.
17507 Simple peak mode looking for the higher sample value. It logs a message
17508 for sample-peak (identified by @code{SPK}).
17510 Enable true-peak mode.
17512 If enabled, the peak lookup is done on an over-sampled version of the input
17513 stream for better peak accuracy. It logs a message for true-peak.
17514 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
17515 This mode requires a build with @code{libswresample}.
17519 Treat mono input files as "dual mono". If a mono file is intended for playback
17520 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
17521 If set to @code{true}, this option will compensate for this effect.
17522 Multi-channel input files are not affected by this option.
17525 Set a specific pan law to be used for the measurement of dual mono files.
17526 This parameter is optional, and has a default value of -3.01dB.
17529 @subsection Examples
17533 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
17535 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
17539 Run an analysis with @command{ffmpeg}:
17541 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
17545 @section interleave, ainterleave
17547 Temporally interleave frames from several inputs.
17549 @code{interleave} works with video inputs, @code{ainterleave} with audio.
17551 These filters read frames from several inputs and send the oldest
17552 queued frame to the output.
17554 Input streams must have well defined, monotonically increasing frame
17557 In order to submit one frame to output, these filters need to enqueue
17558 at least one frame for each input, so they cannot work in case one
17559 input is not yet terminated and will not receive incoming frames.
17561 For example consider the case when one input is a @code{select} filter
17562 which always drops input frames. The @code{interleave} filter will keep
17563 reading from that input, but it will never be able to send new frames
17564 to output until the input sends an end-of-stream signal.
17566 Also, depending on inputs synchronization, the filters will drop
17567 frames in case one input receives more frames than the other ones, and
17568 the queue is already filled.
17570 These filters accept the following options:
17574 Set the number of different inputs, it is 2 by default.
17577 @subsection Examples
17581 Interleave frames belonging to different streams using @command{ffmpeg}:
17583 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
17587 Add flickering blur effect:
17589 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
17593 @section metadata, ametadata
17595 Manipulate frame metadata.
17597 This filter accepts the following options:
17601 Set mode of operation of the filter.
17603 Can be one of the following:
17607 If both @code{value} and @code{key} is set, select frames
17608 which have such metadata. If only @code{key} is set, select
17609 every frame that has such key in metadata.
17612 Add new metadata @code{key} and @code{value}. If key is already available
17616 Modify value of already present key.
17619 If @code{value} is set, delete only keys that have such value.
17620 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
17624 Print key and its value if metadata was found. If @code{key} is not set print all
17625 metadata values available in frame.
17629 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
17632 Set metadata value which will be used. This option is mandatory for
17633 @code{modify} and @code{add} mode.
17636 Which function to use when comparing metadata value and @code{value}.
17638 Can be one of following:
17642 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
17645 Values are interpreted as strings, returns true if metadata value starts with
17646 the @code{value} option string.
17649 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
17652 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
17655 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
17658 Values are interpreted as floats, returns true if expression from option @code{expr}
17663 Set expression which is used when @code{function} is set to @code{expr}.
17664 The expression is evaluated through the eval API and can contain the following
17669 Float representation of @code{value} from metadata key.
17672 Float representation of @code{value} as supplied by user in @code{value} option.
17676 If specified in @code{print} mode, output is written to the named file. Instead of
17677 plain filename any writable url can be specified. Filename ``-'' is a shorthand
17678 for standard output. If @code{file} option is not set, output is written to the log
17679 with AV_LOG_INFO loglevel.
17683 @subsection Examples
17687 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
17690 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
17693 Print silencedetect output to file @file{metadata.txt}.
17695 silencedetect,ametadata=mode=print:file=metadata.txt
17698 Direct all metadata to a pipe with file descriptor 4.
17700 metadata=mode=print:file='pipe\:4'
17704 @section perms, aperms
17706 Set read/write permissions for the output frames.
17708 These filters are mainly aimed at developers to test direct path in the
17709 following filter in the filtergraph.
17711 The filters accept the following options:
17715 Select the permissions mode.
17717 It accepts the following values:
17720 Do nothing. This is the default.
17722 Set all the output frames read-only.
17724 Set all the output frames directly writable.
17726 Make the frame read-only if writable, and writable if read-only.
17728 Set each output frame read-only or writable randomly.
17732 Set the seed for the @var{random} mode, must be an integer included between
17733 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
17734 @code{-1}, the filter will try to use a good random seed on a best effort
17738 Note: in case of auto-inserted filter between the permission filter and the
17739 following one, the permission might not be received as expected in that
17740 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
17741 perms/aperms filter can avoid this problem.
17743 @section realtime, arealtime
17745 Slow down filtering to match real time approximately.
17747 These filters will pause the filtering for a variable amount of time to
17748 match the output rate with the input timestamps.
17749 They are similar to the @option{re} option to @code{ffmpeg}.
17751 They accept the following options:
17755 Time limit for the pauses. Any pause longer than that will be considered
17756 a timestamp discontinuity and reset the timer. Default is 2 seconds.
17760 @section select, aselect
17762 Select frames to pass in output.
17764 This filter accepts the following options:
17769 Set expression, which is evaluated for each input frame.
17771 If the expression is evaluated to zero, the frame is discarded.
17773 If the evaluation result is negative or NaN, the frame is sent to the
17774 first output; otherwise it is sent to the output with index
17775 @code{ceil(val)-1}, assuming that the input index starts from 0.
17777 For example a value of @code{1.2} corresponds to the output with index
17778 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
17781 Set the number of outputs. The output to which to send the selected
17782 frame is based on the result of the evaluation. Default value is 1.
17785 The expression can contain the following constants:
17789 The (sequential) number of the filtered frame, starting from 0.
17792 The (sequential) number of the selected frame, starting from 0.
17794 @item prev_selected_n
17795 The sequential number of the last selected frame. It's NAN if undefined.
17798 The timebase of the input timestamps.
17801 The PTS (Presentation TimeStamp) of the filtered video frame,
17802 expressed in @var{TB} units. It's NAN if undefined.
17805 The PTS of the filtered video frame,
17806 expressed in seconds. It's NAN if undefined.
17809 The PTS of the previously filtered video frame. It's NAN if undefined.
17811 @item prev_selected_pts
17812 The PTS of the last previously filtered video frame. It's NAN if undefined.
17814 @item prev_selected_t
17815 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
17818 The PTS of the first video frame in the video. It's NAN if undefined.
17821 The time of the first video frame in the video. It's NAN if undefined.
17823 @item pict_type @emph{(video only)}
17824 The type of the filtered frame. It can assume one of the following
17836 @item interlace_type @emph{(video only)}
17837 The frame interlace type. It can assume one of the following values:
17840 The frame is progressive (not interlaced).
17842 The frame is top-field-first.
17844 The frame is bottom-field-first.
17847 @item consumed_sample_n @emph{(audio only)}
17848 the number of selected samples before the current frame
17850 @item samples_n @emph{(audio only)}
17851 the number of samples in the current frame
17853 @item sample_rate @emph{(audio only)}
17854 the input sample rate
17857 This is 1 if the filtered frame is a key-frame, 0 otherwise.
17860 the position in the file of the filtered frame, -1 if the information
17861 is not available (e.g. for synthetic video)
17863 @item scene @emph{(video only)}
17864 value between 0 and 1 to indicate a new scene; a low value reflects a low
17865 probability for the current frame to introduce a new scene, while a higher
17866 value means the current frame is more likely to be one (see the example below)
17868 @item concatdec_select
17869 The concat demuxer can select only part of a concat input file by setting an
17870 inpoint and an outpoint, but the output packets may not be entirely contained
17871 in the selected interval. By using this variable, it is possible to skip frames
17872 generated by the concat demuxer which are not exactly contained in the selected
17875 This works by comparing the frame pts against the @var{lavf.concat.start_time}
17876 and the @var{lavf.concat.duration} packet metadata values which are also
17877 present in the decoded frames.
17879 The @var{concatdec_select} variable is -1 if the frame pts is at least
17880 start_time and either the duration metadata is missing or the frame pts is less
17881 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
17884 That basically means that an input frame is selected if its pts is within the
17885 interval set by the concat demuxer.
17889 The default value of the select expression is "1".
17891 @subsection Examples
17895 Select all frames in input:
17900 The example above is the same as:
17912 Select only I-frames:
17914 select='eq(pict_type\,I)'
17918 Select one frame every 100:
17920 select='not(mod(n\,100))'
17924 Select only frames contained in the 10-20 time interval:
17926 select=between(t\,10\,20)
17930 Select only I-frames contained in the 10-20 time interval:
17932 select=between(t\,10\,20)*eq(pict_type\,I)
17936 Select frames with a minimum distance of 10 seconds:
17938 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
17942 Use aselect to select only audio frames with samples number > 100:
17944 aselect='gt(samples_n\,100)'
17948 Create a mosaic of the first scenes:
17950 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
17953 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
17957 Send even and odd frames to separate outputs, and compose them:
17959 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
17963 Select useful frames from an ffconcat file which is using inpoints and
17964 outpoints but where the source files are not intra frame only.
17966 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
17970 @section sendcmd, asendcmd
17972 Send commands to filters in the filtergraph.
17974 These filters read commands to be sent to other filters in the
17977 @code{sendcmd} must be inserted between two video filters,
17978 @code{asendcmd} must be inserted between two audio filters, but apart
17979 from that they act the same way.
17981 The specification of commands can be provided in the filter arguments
17982 with the @var{commands} option, or in a file specified by the
17983 @var{filename} option.
17985 These filters accept the following options:
17988 Set the commands to be read and sent to the other filters.
17990 Set the filename of the commands to be read and sent to the other
17994 @subsection Commands syntax
17996 A commands description consists of a sequence of interval
17997 specifications, comprising a list of commands to be executed when a
17998 particular event related to that interval occurs. The occurring event
17999 is typically the current frame time entering or leaving a given time
18002 An interval is specified by the following syntax:
18004 @var{START}[-@var{END}] @var{COMMANDS};
18007 The time interval is specified by the @var{START} and @var{END} times.
18008 @var{END} is optional and defaults to the maximum time.
18010 The current frame time is considered within the specified interval if
18011 it is included in the interval [@var{START}, @var{END}), that is when
18012 the time is greater or equal to @var{START} and is lesser than
18015 @var{COMMANDS} consists of a sequence of one or more command
18016 specifications, separated by ",", relating to that interval. The
18017 syntax of a command specification is given by:
18019 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18022 @var{FLAGS} is optional and specifies the type of events relating to
18023 the time interval which enable sending the specified command, and must
18024 be a non-null sequence of identifier flags separated by "+" or "|" and
18025 enclosed between "[" and "]".
18027 The following flags are recognized:
18030 The command is sent when the current frame timestamp enters the
18031 specified interval. In other words, the command is sent when the
18032 previous frame timestamp was not in the given interval, and the
18036 The command is sent when the current frame timestamp leaves the
18037 specified interval. In other words, the command is sent when the
18038 previous frame timestamp was in the given interval, and the
18042 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18045 @var{TARGET} specifies the target of the command, usually the name of
18046 the filter class or a specific filter instance name.
18048 @var{COMMAND} specifies the name of the command for the target filter.
18050 @var{ARG} is optional and specifies the optional list of argument for
18051 the given @var{COMMAND}.
18053 Between one interval specification and another, whitespaces, or
18054 sequences of characters starting with @code{#} until the end of line,
18055 are ignored and can be used to annotate comments.
18057 A simplified BNF description of the commands specification syntax
18060 @var{COMMAND_FLAG} ::= "enter" | "leave"
18061 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18062 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18063 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18064 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18065 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18068 @subsection Examples
18072 Specify audio tempo change at second 4:
18074 asendcmd=c='4.0 atempo tempo 1.5',atempo
18078 Target a specific filter instance:
18080 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18084 Specify a list of drawtext and hue commands in a file.
18086 # show text in the interval 5-10
18087 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18088 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18090 # desaturate the image in the interval 15-20
18091 15.0-20.0 [enter] hue s 0,
18092 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18094 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18096 # apply an exponential saturation fade-out effect, starting from time 25
18097 25 [enter] hue s exp(25-t)
18100 A filtergraph allowing to read and process the above command list
18101 stored in a file @file{test.cmd}, can be specified with:
18103 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18108 @section setpts, asetpts
18110 Change the PTS (presentation timestamp) of the input frames.
18112 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18114 This filter accepts the following options:
18119 The expression which is evaluated for each frame to construct its timestamp.
18123 The expression is evaluated through the eval API and can contain the following
18128 frame rate, only defined for constant frame-rate video
18131 The presentation timestamp in input
18134 The count of the input frame for video or the number of consumed samples,
18135 not including the current frame for audio, starting from 0.
18137 @item NB_CONSUMED_SAMPLES
18138 The number of consumed samples, not including the current frame (only
18141 @item NB_SAMPLES, S
18142 The number of samples in the current frame (only audio)
18144 @item SAMPLE_RATE, SR
18145 The audio sample rate.
18148 The PTS of the first frame.
18151 the time in seconds of the first frame
18154 State whether the current frame is interlaced.
18157 the time in seconds of the current frame
18160 original position in the file of the frame, or undefined if undefined
18161 for the current frame
18164 The previous input PTS.
18167 previous input time in seconds
18170 The previous output PTS.
18173 previous output time in seconds
18176 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
18180 The wallclock (RTC) time at the start of the movie in microseconds.
18183 The timebase of the input timestamps.
18187 @subsection Examples
18191 Start counting PTS from zero
18193 setpts=PTS-STARTPTS
18197 Apply fast motion effect:
18203 Apply slow motion effect:
18209 Set fixed rate of 25 frames per second:
18215 Set fixed rate 25 fps with some jitter:
18217 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
18221 Apply an offset of 10 seconds to the input PTS:
18227 Generate timestamps from a "live source" and rebase onto the current timebase:
18229 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
18233 Generate timestamps by counting samples:
18240 @section settb, asettb
18242 Set the timebase to use for the output frames timestamps.
18243 It is mainly useful for testing timebase configuration.
18245 It accepts the following parameters:
18250 The expression which is evaluated into the output timebase.
18254 The value for @option{tb} is an arithmetic expression representing a
18255 rational. The expression can contain the constants "AVTB" (the default
18256 timebase), "intb" (the input timebase) and "sr" (the sample rate,
18257 audio only). Default value is "intb".
18259 @subsection Examples
18263 Set the timebase to 1/25:
18269 Set the timebase to 1/10:
18275 Set the timebase to 1001/1000:
18281 Set the timebase to 2*intb:
18287 Set the default timebase value:
18294 Convert input audio to a video output representing frequency spectrum
18295 logarithmically using Brown-Puckette constant Q transform algorithm with
18296 direct frequency domain coefficient calculation (but the transform itself
18297 is not really constant Q, instead the Q factor is actually variable/clamped),
18298 with musical tone scale, from E0 to D#10.
18300 The filter accepts the following options:
18304 Specify the video size for the output. It must be even. For the syntax of this option,
18305 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18306 Default value is @code{1920x1080}.
18309 Set the output frame rate. Default value is @code{25}.
18312 Set the bargraph height. It must be even. Default value is @code{-1} which
18313 computes the bargraph height automatically.
18316 Set the axis height. It must be even. Default value is @code{-1} which computes
18317 the axis height automatically.
18320 Set the sonogram height. It must be even. Default value is @code{-1} which
18321 computes the sonogram height automatically.
18324 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
18325 instead. Default value is @code{1}.
18327 @item sono_v, volume
18328 Specify the sonogram volume expression. It can contain variables:
18331 the @var{bar_v} evaluated expression
18332 @item frequency, freq, f
18333 the frequency where it is evaluated
18334 @item timeclamp, tc
18335 the value of @var{timeclamp} option
18339 @item a_weighting(f)
18340 A-weighting of equal loudness
18341 @item b_weighting(f)
18342 B-weighting of equal loudness
18343 @item c_weighting(f)
18344 C-weighting of equal loudness.
18346 Default value is @code{16}.
18348 @item bar_v, volume2
18349 Specify the bargraph volume expression. It can contain variables:
18352 the @var{sono_v} evaluated expression
18353 @item frequency, freq, f
18354 the frequency where it is evaluated
18355 @item timeclamp, tc
18356 the value of @var{timeclamp} option
18360 @item a_weighting(f)
18361 A-weighting of equal loudness
18362 @item b_weighting(f)
18363 B-weighting of equal loudness
18364 @item c_weighting(f)
18365 C-weighting of equal loudness.
18367 Default value is @code{sono_v}.
18369 @item sono_g, gamma
18370 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
18371 higher gamma makes the spectrum having more range. Default value is @code{3}.
18372 Acceptable range is @code{[1, 7]}.
18374 @item bar_g, gamma2
18375 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
18379 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
18380 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
18382 @item timeclamp, tc
18383 Specify the transform timeclamp. At low frequency, there is trade-off between
18384 accuracy in time domain and frequency domain. If timeclamp is lower,
18385 event in time domain is represented more accurately (such as fast bass drum),
18386 otherwise event in frequency domain is represented more accurately
18387 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
18390 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
18391 limits future samples by applying asymmetric windowing in time domain, useful
18392 when low latency is required. Accepted range is @code{[0, 1]}.
18395 Specify the transform base frequency. Default value is @code{20.01523126408007475},
18396 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
18399 Specify the transform end frequency. Default value is @code{20495.59681441799654},
18400 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
18403 This option is deprecated and ignored.
18406 Specify the transform length in time domain. Use this option to control accuracy
18407 trade-off between time domain and frequency domain at every frequency sample.
18408 It can contain variables:
18410 @item frequency, freq, f
18411 the frequency where it is evaluated
18412 @item timeclamp, tc
18413 the value of @var{timeclamp} option.
18415 Default value is @code{384*tc/(384+tc*f)}.
18418 Specify the transform count for every video frame. Default value is @code{6}.
18419 Acceptable range is @code{[1, 30]}.
18422 Specify the transform count for every single pixel. Default value is @code{0},
18423 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
18426 Specify font file for use with freetype to draw the axis. If not specified,
18427 use embedded font. Note that drawing with font file or embedded font is not
18428 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
18432 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
18433 The : in the pattern may be replaced by | to avoid unnecessary escaping.
18436 Specify font color expression. This is arithmetic expression that should return
18437 integer value 0xRRGGBB. It can contain variables:
18439 @item frequency, freq, f
18440 the frequency where it is evaluated
18441 @item timeclamp, tc
18442 the value of @var{timeclamp} option
18447 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
18448 @item r(x), g(x), b(x)
18449 red, green, and blue value of intensity x.
18451 Default value is @code{st(0, (midi(f)-59.5)/12);
18452 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
18453 r(1-ld(1)) + b(ld(1))}.
18456 Specify image file to draw the axis. This option override @var{fontfile} and
18457 @var{fontcolor} option.
18460 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
18461 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
18462 Default value is @code{1}.
18465 Set colorspace. The accepted values are:
18468 Unspecified (default)
18477 BT.470BG or BT.601-6 625
18480 SMPTE-170M or BT.601-6 525
18486 BT.2020 with non-constant luminance
18491 Set spectrogram color scheme. This is list of floating point values with format
18492 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
18493 The default is @code{1|0.5|0|0|0.5|1}.
18497 @subsection Examples
18501 Playing audio while showing the spectrum:
18503 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
18507 Same as above, but with frame rate 30 fps:
18509 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
18513 Playing at 1280x720:
18515 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
18519 Disable sonogram display:
18525 A1 and its harmonics: A1, A2, (near)E3, A3:
18527 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),
18528 asplit[a][out1]; [a] showcqt [out0]'
18532 Same as above, but with more accuracy in frequency domain:
18534 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),
18535 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
18541 bar_v=10:sono_v=bar_v*a_weighting(f)
18545 Custom gamma, now spectrum is linear to the amplitude.
18551 Custom tlength equation:
18553 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)))'
18557 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
18559 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
18563 Custom font using fontconfig:
18565 font='Courier New,Monospace,mono|bold'
18569 Custom frequency range with custom axis using image file:
18571 axisfile=myaxis.png:basefreq=40:endfreq=10000
18577 Convert input audio to video output representing the audio power spectrum.
18578 Audio amplitude is on Y-axis while frequency is on X-axis.
18580 The filter accepts the following options:
18584 Specify size of video. For the syntax of this option, check the
18585 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18586 Default is @code{1024x512}.
18590 This set how each frequency bin will be represented.
18592 It accepts the following values:
18598 Default is @code{bar}.
18601 Set amplitude scale.
18603 It accepts the following values:
18617 Default is @code{log}.
18620 Set frequency scale.
18622 It accepts the following values:
18631 Reverse logarithmic scale.
18633 Default is @code{lin}.
18638 It accepts the following values:
18654 Default is @code{w2048}
18657 Set windowing function.
18659 It accepts the following values:
18681 Default is @code{hanning}.
18684 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
18685 which means optimal overlap for selected window function will be picked.
18688 Set time averaging. Setting this to 0 will display current maximal peaks.
18689 Default is @code{1}, which means time averaging is disabled.
18692 Specify list of colors separated by space or by '|' which will be used to
18693 draw channel frequencies. Unrecognized or missing colors will be replaced
18697 Set channel display mode.
18699 It accepts the following values:
18704 Default is @code{combined}.
18707 Set minimum amplitude used in @code{log} amplitude scaler.
18711 @anchor{showspectrum}
18712 @section showspectrum
18714 Convert input audio to a video output, representing the audio frequency
18717 The filter accepts the following options:
18721 Specify the video size for the output. For the syntax of this option, check the
18722 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18723 Default value is @code{640x512}.
18726 Specify how the spectrum should slide along the window.
18728 It accepts the following values:
18731 the samples start again on the left when they reach the right
18733 the samples scroll from right to left
18735 frames are only produced when the samples reach the right
18737 the samples scroll from left to right
18740 Default value is @code{replace}.
18743 Specify display mode.
18745 It accepts the following values:
18748 all channels are displayed in the same row
18750 all channels are displayed in separate rows
18753 Default value is @samp{combined}.
18756 Specify display color mode.
18758 It accepts the following values:
18761 each channel is displayed in a separate color
18763 each channel is displayed using the same color scheme
18765 each channel is displayed using the rainbow color scheme
18767 each channel is displayed using the moreland color scheme
18769 each channel is displayed using the nebulae color scheme
18771 each channel is displayed using the fire color scheme
18773 each channel is displayed using the fiery color scheme
18775 each channel is displayed using the fruit color scheme
18777 each channel is displayed using the cool color scheme
18780 Default value is @samp{channel}.
18783 Specify scale used for calculating intensity color values.
18785 It accepts the following values:
18790 square root, default
18801 Default value is @samp{sqrt}.
18804 Set saturation modifier for displayed colors. Negative values provide
18805 alternative color scheme. @code{0} is no saturation at all.
18806 Saturation must be in [-10.0, 10.0] range.
18807 Default value is @code{1}.
18810 Set window function.
18812 It accepts the following values:
18836 Default value is @code{hann}.
18839 Set orientation of time vs frequency axis. Can be @code{vertical} or
18840 @code{horizontal}. Default is @code{vertical}.
18843 Set ratio of overlap window. Default value is @code{0}.
18844 When value is @code{1} overlap is set to recommended size for specific
18845 window function currently used.
18848 Set scale gain for calculating intensity color values.
18849 Default value is @code{1}.
18852 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
18855 Set color rotation, must be in [-1.0, 1.0] range.
18856 Default value is @code{0}.
18859 The usage is very similar to the showwaves filter; see the examples in that
18862 @subsection Examples
18866 Large window with logarithmic color scaling:
18868 showspectrum=s=1280x480:scale=log
18872 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
18874 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18875 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
18879 @section showspectrumpic
18881 Convert input audio to a single video frame, representing the audio frequency
18884 The filter accepts the following options:
18888 Specify the video size for the output. For the syntax of this option, check the
18889 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18890 Default value is @code{4096x2048}.
18893 Specify display mode.
18895 It accepts the following values:
18898 all channels are displayed in the same row
18900 all channels are displayed in separate rows
18902 Default value is @samp{combined}.
18905 Specify display color mode.
18907 It accepts the following values:
18910 each channel is displayed in a separate color
18912 each channel is displayed using the same color scheme
18914 each channel is displayed using the rainbow color scheme
18916 each channel is displayed using the moreland color scheme
18918 each channel is displayed using the nebulae color scheme
18920 each channel is displayed using the fire color scheme
18922 each channel is displayed using the fiery color scheme
18924 each channel is displayed using the fruit color scheme
18926 each channel is displayed using the cool color scheme
18928 Default value is @samp{intensity}.
18931 Specify scale used for calculating intensity color values.
18933 It accepts the following values:
18938 square root, default
18948 Default value is @samp{log}.
18951 Set saturation modifier for displayed colors. Negative values provide
18952 alternative color scheme. @code{0} is no saturation at all.
18953 Saturation must be in [-10.0, 10.0] range.
18954 Default value is @code{1}.
18957 Set window function.
18959 It accepts the following values:
18982 Default value is @code{hann}.
18985 Set orientation of time vs frequency axis. Can be @code{vertical} or
18986 @code{horizontal}. Default is @code{vertical}.
18989 Set scale gain for calculating intensity color values.
18990 Default value is @code{1}.
18993 Draw time and frequency axes and legends. Default is enabled.
18996 Set color rotation, must be in [-1.0, 1.0] range.
18997 Default value is @code{0}.
19000 @subsection Examples
19004 Extract an audio spectrogram of a whole audio track
19005 in a 1024x1024 picture using @command{ffmpeg}:
19007 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19011 @section showvolume
19013 Convert input audio volume to a video output.
19015 The filter accepts the following options:
19022 Set border width, allowed range is [0, 5]. Default is 1.
19025 Set channel width, allowed range is [80, 8192]. Default is 400.
19028 Set channel height, allowed range is [1, 900]. Default is 20.
19031 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19034 Set volume color expression.
19036 The expression can use the following variables:
19040 Current max volume of channel in dB.
19046 Current channel number, starting from 0.
19050 If set, displays channel names. Default is enabled.
19053 If set, displays volume values. Default is enabled.
19056 Set orientation, can be @code{horizontal} or @code{vertical},
19057 default is @code{horizontal}.
19060 Set step size, allowed range s [0, 5]. Default is 0, which means
19066 Convert input audio to a video output, representing the samples waves.
19068 The filter accepts the following options:
19072 Specify the video size for the output. For the syntax of this option, check the
19073 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19074 Default value is @code{600x240}.
19079 Available values are:
19082 Draw a point for each sample.
19085 Draw a vertical line for each sample.
19088 Draw a point for each sample and a line between them.
19091 Draw a centered vertical line for each sample.
19094 Default value is @code{point}.
19097 Set the number of samples which are printed on the same column. A
19098 larger value will decrease the frame rate. Must be a positive
19099 integer. This option can be set only if the value for @var{rate}
19100 is not explicitly specified.
19103 Set the (approximate) output frame rate. This is done by setting the
19104 option @var{n}. Default value is "25".
19106 @item split_channels
19107 Set if channels should be drawn separately or overlap. Default value is 0.
19110 Set colors separated by '|' which are going to be used for drawing of each channel.
19113 Set amplitude scale.
19115 Available values are:
19133 @subsection Examples
19137 Output the input file audio and the corresponding video representation
19140 amovie=a.mp3,asplit[out0],showwaves[out1]
19144 Create a synthetic signal and show it with showwaves, forcing a
19145 frame rate of 30 frames per second:
19147 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
19151 @section showwavespic
19153 Convert input audio to a single video frame, representing the samples waves.
19155 The filter accepts the following options:
19159 Specify the video size for the output. For the syntax of this option, check the
19160 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19161 Default value is @code{600x240}.
19163 @item split_channels
19164 Set if channels should be drawn separately or overlap. Default value is 0.
19167 Set colors separated by '|' which are going to be used for drawing of each channel.
19170 Set amplitude scale.
19172 Available values are:
19190 @subsection Examples
19194 Extract a channel split representation of the wave form of a whole audio track
19195 in a 1024x800 picture using @command{ffmpeg}:
19197 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
19201 @section sidedata, asidedata
19203 Delete frame side data, or select frames based on it.
19205 This filter accepts the following options:
19209 Set mode of operation of the filter.
19211 Can be one of the following:
19215 Select every frame with side data of @code{type}.
19218 Delete side data of @code{type}. If @code{type} is not set, delete all side
19224 Set side data type used with all modes. Must be set for @code{select} mode. For
19225 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
19226 in @file{libavutil/frame.h}. For example, to choose
19227 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
19231 @section spectrumsynth
19233 Sythesize audio from 2 input video spectrums, first input stream represents
19234 magnitude across time and second represents phase across time.
19235 The filter will transform from frequency domain as displayed in videos back
19236 to time domain as presented in audio output.
19238 This filter is primarily created for reversing processed @ref{showspectrum}
19239 filter outputs, but can synthesize sound from other spectrograms too.
19240 But in such case results are going to be poor if the phase data is not
19241 available, because in such cases phase data need to be recreated, usually
19242 its just recreated from random noise.
19243 For best results use gray only output (@code{channel} color mode in
19244 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
19245 @code{lin} scale for phase video. To produce phase, for 2nd video, use
19246 @code{data} option. Inputs videos should generally use @code{fullframe}
19247 slide mode as that saves resources needed for decoding video.
19249 The filter accepts the following options:
19253 Specify sample rate of output audio, the sample rate of audio from which
19254 spectrum was generated may differ.
19257 Set number of channels represented in input video spectrums.
19260 Set scale which was used when generating magnitude input spectrum.
19261 Can be @code{lin} or @code{log}. Default is @code{log}.
19264 Set slide which was used when generating inputs spectrums.
19265 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
19266 Default is @code{fullframe}.
19269 Set window function used for resynthesis.
19272 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19273 which means optimal overlap for selected window function will be picked.
19276 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
19277 Default is @code{vertical}.
19280 @subsection Examples
19284 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
19285 then resynthesize videos back to audio with spectrumsynth:
19287 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
19288 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
19289 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
19293 @section split, asplit
19295 Split input into several identical outputs.
19297 @code{asplit} works with audio input, @code{split} with video.
19299 The filter accepts a single parameter which specifies the number of outputs. If
19300 unspecified, it defaults to 2.
19302 @subsection Examples
19306 Create two separate outputs from the same input:
19308 [in] split [out0][out1]
19312 To create 3 or more outputs, you need to specify the number of
19315 [in] asplit=3 [out0][out1][out2]
19319 Create two separate outputs from the same input, one cropped and
19322 [in] split [splitout1][splitout2];
19323 [splitout1] crop=100:100:0:0 [cropout];
19324 [splitout2] pad=200:200:100:100 [padout];
19328 Create 5 copies of the input audio with @command{ffmpeg}:
19330 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
19336 Receive commands sent through a libzmq client, and forward them to
19337 filters in the filtergraph.
19339 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
19340 must be inserted between two video filters, @code{azmq} between two
19343 To enable these filters you need to install the libzmq library and
19344 headers and configure FFmpeg with @code{--enable-libzmq}.
19346 For more information about libzmq see:
19347 @url{http://www.zeromq.org/}
19349 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
19350 receives messages sent through a network interface defined by the
19351 @option{bind_address} option.
19353 The received message must be in the form:
19355 @var{TARGET} @var{COMMAND} [@var{ARG}]
19358 @var{TARGET} specifies the target of the command, usually the name of
19359 the filter class or a specific filter instance name.
19361 @var{COMMAND} specifies the name of the command for the target filter.
19363 @var{ARG} is optional and specifies the optional argument list for the
19364 given @var{COMMAND}.
19366 Upon reception, the message is processed and the corresponding command
19367 is injected into the filtergraph. Depending on the result, the filter
19368 will send a reply to the client, adopting the format:
19370 @var{ERROR_CODE} @var{ERROR_REASON}
19374 @var{MESSAGE} is optional.
19376 @subsection Examples
19378 Look at @file{tools/zmqsend} for an example of a zmq client which can
19379 be used to send commands processed by these filters.
19381 Consider the following filtergraph generated by @command{ffplay}
19383 ffplay -dumpgraph 1 -f lavfi "
19384 color=s=100x100:c=red [l];
19385 color=s=100x100:c=blue [r];
19386 nullsrc=s=200x100, zmq [bg];
19387 [bg][l] overlay [bg+l];
19388 [bg+l][r] overlay=x=100 "
19391 To change the color of the left side of the video, the following
19392 command can be used:
19394 echo Parsed_color_0 c yellow | tools/zmqsend
19397 To change the right side:
19399 echo Parsed_color_1 c pink | tools/zmqsend
19402 @c man end MULTIMEDIA FILTERS
19404 @chapter Multimedia Sources
19405 @c man begin MULTIMEDIA SOURCES
19407 Below is a description of the currently available multimedia sources.
19411 This is the same as @ref{movie} source, except it selects an audio
19417 Read audio and/or video stream(s) from a movie container.
19419 It accepts the following parameters:
19423 The name of the resource to read (not necessarily a file; it can also be a
19424 device or a stream accessed through some protocol).
19426 @item format_name, f
19427 Specifies the format assumed for the movie to read, and can be either
19428 the name of a container or an input device. If not specified, the
19429 format is guessed from @var{movie_name} or by probing.
19431 @item seek_point, sp
19432 Specifies the seek point in seconds. The frames will be output
19433 starting from this seek point. The parameter is evaluated with
19434 @code{av_strtod}, so the numerical value may be suffixed by an IS
19435 postfix. The default value is "0".
19438 Specifies the streams to read. Several streams can be specified,
19439 separated by "+". The source will then have as many outputs, in the
19440 same order. The syntax is explained in the ``Stream specifiers''
19441 section in the ffmpeg manual. Two special names, "dv" and "da" specify
19442 respectively the default (best suited) video and audio stream. Default
19443 is "dv", or "da" if the filter is called as "amovie".
19445 @item stream_index, si
19446 Specifies the index of the video stream to read. If the value is -1,
19447 the most suitable video stream will be automatically selected. The default
19448 value is "-1". Deprecated. If the filter is called "amovie", it will select
19449 audio instead of video.
19452 Specifies how many times to read the stream in sequence.
19453 If the value is 0, the stream will be looped infinitely.
19454 Default value is "1".
19456 Note that when the movie is looped the source timestamps are not
19457 changed, so it will generate non monotonically increasing timestamps.
19459 @item discontinuity
19460 Specifies the time difference between frames above which the point is
19461 considered a timestamp discontinuity which is removed by adjusting the later
19465 It allows overlaying a second video on top of the main input of
19466 a filtergraph, as shown in this graph:
19468 input -----------> deltapts0 --> overlay --> output
19471 movie --> scale--> deltapts1 -------+
19473 @subsection Examples
19477 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
19478 on top of the input labelled "in":
19480 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
19481 [in] setpts=PTS-STARTPTS [main];
19482 [main][over] overlay=16:16 [out]
19486 Read from a video4linux2 device, and overlay it on top of the input
19489 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
19490 [in] setpts=PTS-STARTPTS [main];
19491 [main][over] overlay=16:16 [out]
19495 Read the first video stream and the audio stream with id 0x81 from
19496 dvd.vob; the video is connected to the pad named "video" and the audio is
19497 connected to the pad named "audio":
19499 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
19503 @subsection Commands
19505 Both movie and amovie support the following commands:
19508 Perform seek using "av_seek_frame".
19509 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
19512 @var{stream_index}: If stream_index is -1, a default
19513 stream is selected, and @var{timestamp} is automatically converted
19514 from AV_TIME_BASE units to the stream specific time_base.
19516 @var{timestamp}: Timestamp in AVStream.time_base units
19517 or, if no stream is specified, in AV_TIME_BASE units.
19519 @var{flags}: Flags which select direction and seeking mode.
19523 Get movie duration in AV_TIME_BASE units.
19527 @c man end MULTIMEDIA SOURCES