1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @section Notes on filtergraph escaping
226 Filtergraph description composition entails several levels of
227 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
228 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
229 information about the employed escaping procedure.
231 A first level escaping affects the content of each filter option
232 value, which may contain the special character @code{:} used to
233 separate values, or one of the escaping characters @code{\'}.
235 A second level escaping affects the whole filter description, which
236 may contain the escaping characters @code{\'} or the special
237 characters @code{[],;} used by the filtergraph description.
239 Finally, when you specify a filtergraph on a shell commandline, you
240 need to perform a third level escaping for the shell special
241 characters contained within it.
243 For example, consider the following string to be embedded in
244 the @ref{drawtext} filter description @option{text} value:
246 this is a 'string': may contain one, or more, special characters
249 This string contains the @code{'} special escaping character, and the
250 @code{:} special character, so it needs to be escaped in this way:
252 text=this is a \'string\'\: may contain one, or more, special characters
255 A second level of escaping is required when embedding the filter
256 description in a filtergraph description, in order to escape all the
257 filtergraph special characters. Thus the example above becomes:
259 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
261 (note that in addition to the @code{\'} escaping special characters,
262 also @code{,} needs to be escaped).
264 Finally an additional level of escaping is needed when writing the
265 filtergraph description in a shell command, which depends on the
266 escaping rules of the adopted shell. For example, assuming that
267 @code{\} is special and needs to be escaped with another @code{\}, the
268 previous string will finally result in:
270 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
273 @chapter Timeline editing
275 Some filters support a generic @option{enable} option. For the filters
276 supporting timeline editing, this option can be set to an expression which is
277 evaluated before sending a frame to the filter. If the evaluation is non-zero,
278 the filter will be enabled, otherwise the frame will be sent unchanged to the
279 next filter in the filtergraph.
281 The expression accepts the following values:
284 timestamp expressed in seconds, NAN if the input timestamp is unknown
287 sequential number of the input frame, starting from 0
290 the position in the file of the input frame, NAN if unknown
294 width and height of the input frame if video
297 Additionally, these filters support an @option{enable} command that can be used
298 to re-define the expression.
300 Like any other filtering option, the @option{enable} option follows the same
303 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
304 minutes, and a @ref{curves} filter starting at 3 seconds:
306 smartblur = enable='between(t,10,3*60)',
307 curves = enable='gte(t,3)' : preset=cross_process
310 See @code{ffmpeg -filters} to view which filters have timeline support.
312 @c man end FILTERGRAPH DESCRIPTION
315 @chapter Options for filters with several inputs (framesync)
316 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
318 Some filters with several inputs support a common set of options.
319 These options can only be set by name, not with the short notation.
323 The action to take when EOF is encountered on the secondary input; it accepts
324 one of the following values:
328 Repeat the last frame (the default).
332 Pass the main input through.
336 If set to 1, force the output to terminate when the shortest input
337 terminates. Default value is 0.
340 If set to 1, force the filter to extend the last frame of secondary streams
341 until the end of the primary stream. A value of 0 disables this behavior.
345 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
347 @chapter Audio Filters
348 @c man begin AUDIO FILTERS
350 When you configure your FFmpeg build, you can disable any of the
351 existing filters using @code{--disable-filters}.
352 The configure output will show the audio filters included in your
355 Below is a description of the currently available audio filters.
359 A compressor is mainly used to reduce the dynamic range of a signal.
360 Especially modern music is mostly compressed at a high ratio to
361 improve the overall loudness. It's done to get the highest attention
362 of a listener, "fatten" the sound and bring more "power" to the track.
363 If a signal is compressed too much it may sound dull or "dead"
364 afterwards or it may start to "pump" (which could be a powerful effect
365 but can also destroy a track completely).
366 The right compression is the key to reach a professional sound and is
367 the high art of mixing and mastering. Because of its complex settings
368 it may take a long time to get the right feeling for this kind of effect.
370 Compression is done by detecting the volume above a chosen level
371 @code{threshold} and dividing it by the factor set with @code{ratio}.
372 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
373 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
374 the signal would cause distortion of the waveform the reduction can be
375 levelled over the time. This is done by setting "Attack" and "Release".
376 @code{attack} determines how long the signal has to rise above the threshold
377 before any reduction will occur and @code{release} sets the time the signal
378 has to fall below the threshold to reduce the reduction again. Shorter signals
379 than the chosen attack time will be left untouched.
380 The overall reduction of the signal can be made up afterwards with the
381 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
382 raising the makeup to this level results in a signal twice as loud than the
383 source. To gain a softer entry in the compression the @code{knee} flattens the
384 hard edge at the threshold in the range of the chosen decibels.
386 The filter accepts the following options:
390 Set input gain. Default is 1. Range is between 0.015625 and 64.
393 If a signal of stream rises above this level it will affect the gain
395 By default it is 0.125. Range is between 0.00097563 and 1.
398 Set a ratio by which the signal is reduced. 1:2 means that if the level
399 rose 4dB above the threshold, it will be only 2dB above after the reduction.
400 Default is 2. Range is between 1 and 20.
403 Amount of milliseconds the signal has to rise above the threshold before gain
404 reduction starts. Default is 20. Range is between 0.01 and 2000.
407 Amount of milliseconds the signal has to fall below the threshold before
408 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
411 Set the amount by how much signal will be amplified after processing.
412 Default is 1. Range is from 1 to 64.
415 Curve the sharp knee around the threshold to enter gain reduction more softly.
416 Default is 2.82843. Range is between 1 and 8.
419 Choose if the @code{average} level between all channels of input stream
420 or the louder(@code{maximum}) channel of input stream affects the
421 reduction. Default is @code{average}.
424 Should the exact signal be taken in case of @code{peak} or an RMS one in case
425 of @code{rms}. Default is @code{rms} which is mostly smoother.
428 How much to use compressed signal in output. Default is 1.
429 Range is between 0 and 1.
433 Simple audio dynamic range commpression/expansion filter.
435 The filter accepts the following options:
439 Set contrast. Default is 33. Allowed range is between 0 and 100.
444 Copy the input audio source unchanged to the output. This is mainly useful for
449 Apply cross fade from one input audio stream to another input audio stream.
450 The cross fade is applied for specified duration near the end of first stream.
452 The filter accepts the following options:
456 Specify the number of samples for which the cross fade effect has to last.
457 At the end of the cross fade effect the first input audio will be completely
458 silent. Default is 44100.
461 Specify the duration of the cross fade effect. See
462 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
463 for the accepted syntax.
464 By default the duration is determined by @var{nb_samples}.
465 If set this option is used instead of @var{nb_samples}.
468 Should first stream end overlap with second stream start. Default is enabled.
471 Set curve for cross fade transition for first stream.
474 Set curve for cross fade transition for second stream.
476 For description of available curve types see @ref{afade} filter description.
483 Cross fade from one input to another:
485 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
489 Cross fade from one input to another but without overlapping:
491 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Reduce audio bit resolution.
499 This filter is bit crusher with enhanced functionality. A bit crusher
500 is used to audibly reduce number of bits an audio signal is sampled
501 with. This doesn't change the bit depth at all, it just produces the
502 effect. Material reduced in bit depth sounds more harsh and "digital".
503 This filter is able to even round to continuous values instead of discrete
505 Additionally it has a D/C offset which results in different crushing of
506 the lower and the upper half of the signal.
507 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
509 Another feature of this filter is the logarithmic mode.
510 This setting switches from linear distances between bits to logarithmic ones.
511 The result is a much more "natural" sounding crusher which doesn't gate low
512 signals for example. The human ear has a logarithmic perception,
513 so this kind of crushing is much more pleasant.
514 Logarithmic crushing is also able to get anti-aliased.
516 The filter accepts the following options:
532 Can be linear: @code{lin} or logarithmic: @code{log}.
541 Set sample reduction.
544 Enable LFO. By default disabled.
555 Delay one or more audio channels.
557 Samples in delayed channel are filled with silence.
559 The filter accepts the following option:
563 Set list of delays in milliseconds for each channel separated by '|'.
564 Unused delays will be silently ignored. If number of given delays is
565 smaller than number of channels all remaining channels will not be delayed.
566 If you want to delay exact number of samples, append 'S' to number.
573 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
574 the second channel (and any other channels that may be present) unchanged.
580 Delay second channel by 500 samples, the third channel by 700 samples and leave
581 the first channel (and any other channels that may be present) unchanged.
589 Apply echoing to the input audio.
591 Echoes are reflected sound and can occur naturally amongst mountains
592 (and sometimes large buildings) when talking or shouting; digital echo
593 effects emulate this behaviour and are often used to help fill out the
594 sound of a single instrument or vocal. The time difference between the
595 original signal and the reflection is the @code{delay}, and the
596 loudness of the reflected signal is the @code{decay}.
597 Multiple echoes can have different delays and decays.
599 A description of the accepted parameters follows.
603 Set input gain of reflected signal. Default is @code{0.6}.
606 Set output gain of reflected signal. Default is @code{0.3}.
609 Set list of time intervals in milliseconds between original signal and reflections
610 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
611 Default is @code{1000}.
614 Set list of loudness of reflected signals separated by '|'.
615 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
616 Default is @code{0.5}.
623 Make it sound as if there are twice as many instruments as are actually playing:
625 aecho=0.8:0.88:60:0.4
629 If delay is very short, then it sound like a (metallic) robot playing music:
635 A longer delay will sound like an open air concert in the mountains:
637 aecho=0.8:0.9:1000:0.3
641 Same as above but with one more mountain:
643 aecho=0.8:0.9:1000|1800:0.3|0.25
648 Audio emphasis filter creates or restores material directly taken from LPs or
649 emphased CDs with different filter curves. E.g. to store music on vinyl the
650 signal has to be altered by a filter first to even out the disadvantages of
651 this recording medium.
652 Once the material is played back the inverse filter has to be applied to
653 restore the distortion of the frequency response.
655 The filter accepts the following options:
665 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
666 use @code{production} mode. Default is @code{reproduction} mode.
669 Set filter type. Selects medium. Can be one of the following:
681 select Compact Disc (CD).
687 select 50µs (FM-KF).
689 select 75µs (FM-KF).
695 Modify an audio signal according to the specified expressions.
697 This filter accepts one or more expressions (one for each channel),
698 which are evaluated and used to modify a corresponding audio signal.
700 It accepts the following parameters:
704 Set the '|'-separated expressions list for each separate channel. If
705 the number of input channels is greater than the number of
706 expressions, the last specified expression is used for the remaining
709 @item channel_layout, c
710 Set output channel layout. If not specified, the channel layout is
711 specified by the number of expressions. If set to @samp{same}, it will
712 use by default the same input channel layout.
715 Each expression in @var{exprs} can contain the following constants and functions:
719 channel number of the current expression
722 number of the evaluated sample, starting from 0
728 time of the evaluated sample expressed in seconds
731 @item nb_out_channels
732 input and output number of channels
735 the value of input channel with number @var{CH}
738 Note: this filter is slow. For faster processing you should use a
747 aeval=val(ch)/2:c=same
751 Invert phase of the second channel:
760 Apply fade-in/out effect to input audio.
762 A description of the accepted parameters follows.
766 Specify the effect type, can be either @code{in} for fade-in, or
767 @code{out} for a fade-out effect. Default is @code{in}.
769 @item start_sample, ss
770 Specify the number of the start sample for starting to apply the fade
771 effect. Default is 0.
774 Specify the number of samples for which the fade effect has to last. At
775 the end of the fade-in effect the output audio will have the same
776 volume as the input audio, at the end of the fade-out transition
777 the output audio will be silence. Default is 44100.
780 Specify the start time of the fade effect. Default is 0.
781 The value must be specified as a time duration; see
782 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
783 for the accepted syntax.
784 If set this option is used instead of @var{start_sample}.
787 Specify the duration of the fade effect. See
788 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
789 for the accepted syntax.
790 At the end of the fade-in effect the output audio will have the same
791 volume as the input audio, at the end of the fade-out transition
792 the output audio will be silence.
793 By default the duration is determined by @var{nb_samples}.
794 If set this option is used instead of @var{nb_samples}.
797 Set curve for fade transition.
799 It accepts the following values:
802 select triangular, linear slope (default)
804 select quarter of sine wave
806 select half of sine wave
808 select exponential sine wave
812 select inverted parabola
826 select inverted quarter of sine wave
828 select inverted half of sine wave
830 select double-exponential seat
832 select double-exponential sigmoid
840 Fade in first 15 seconds of audio:
846 Fade out last 25 seconds of a 900 seconds audio:
848 afade=t=out:st=875:d=25
853 Apply arbitrary expressions to samples in frequency domain.
857 Set frequency domain real expression for each separate channel separated
858 by '|'. Default is "1".
859 If the number of input channels is greater than the number of
860 expressions, the last specified expression is used for the remaining
864 Set frequency domain imaginary expression for each separate channel
865 separated by '|'. If not set, @var{real} option is used.
867 Each expression in @var{real} and @var{imag} can contain the following
875 current frequency bin number
878 number of available bins
881 channel number of the current expression
893 It accepts the following values:
909 Default is @code{w4096}
912 Set window function. Default is @code{hann}.
915 Set window overlap. If set to 1, the recommended overlap for selected
916 window function will be picked. Default is @code{0.75}.
923 Leave almost only low frequencies in audio:
925 afftfilt="1-clip((b/nb)*b,0,1)"
932 Apply an arbitrary Frequency Impulse Response filter.
934 This filter is designed for applying long FIR filters,
935 up to 30 seconds long.
937 It can be used as component for digital crossover filters,
938 room equalization, cross talk cancellation, wavefield synthesis,
939 auralization, ambiophonics and ambisonics.
941 This filter uses second stream as FIR coefficients.
942 If second stream holds single channel, it will be used
943 for all input channels in first stream, otherwise
944 number of channels in second stream must be same as
945 number of channels in first stream.
947 It accepts the following parameters:
951 Set dry gain. This sets input gain.
954 Set wet gain. This sets final output gain.
957 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
960 Enable applying gain measured from power of IR.
967 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
969 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
976 Set output format constraints for the input audio. The framework will
977 negotiate the most appropriate format to minimize conversions.
979 It accepts the following parameters:
983 A '|'-separated list of requested sample formats.
986 A '|'-separated list of requested sample rates.
988 @item channel_layouts
989 A '|'-separated list of requested channel layouts.
991 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
992 for the required syntax.
995 If a parameter is omitted, all values are allowed.
997 Force the output to either unsigned 8-bit or signed 16-bit stereo
999 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1004 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1005 processing reduces disturbing noise between useful signals.
1007 Gating is done by detecting the volume below a chosen level @var{threshold}
1008 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1009 floor is set via @var{range}. Because an exact manipulation of the signal
1010 would cause distortion of the waveform the reduction can be levelled over
1011 time. This is done by setting @var{attack} and @var{release}.
1013 @var{attack} determines how long the signal has to fall below the threshold
1014 before any reduction will occur and @var{release} sets the time the signal
1015 has to rise above the threshold to reduce the reduction again.
1016 Shorter signals than the chosen attack time will be left untouched.
1020 Set input level before filtering.
1021 Default is 1. Allowed range is from 0.015625 to 64.
1024 Set the level of gain reduction when the signal is below the threshold.
1025 Default is 0.06125. Allowed range is from 0 to 1.
1028 If a signal rises above this level the gain reduction is released.
1029 Default is 0.125. Allowed range is from 0 to 1.
1032 Set a ratio by which the signal is reduced.
1033 Default is 2. Allowed range is from 1 to 9000.
1036 Amount of milliseconds the signal has to rise above the threshold before gain
1038 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1041 Amount of milliseconds the signal has to fall below the threshold before the
1042 reduction is increased again. Default is 250 milliseconds.
1043 Allowed range is from 0.01 to 9000.
1046 Set amount of amplification of signal after processing.
1047 Default is 1. Allowed range is from 1 to 64.
1050 Curve the sharp knee around the threshold to enter gain reduction more softly.
1051 Default is 2.828427125. Allowed range is from 1 to 8.
1054 Choose if exact signal should be taken for detection or an RMS like one.
1055 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1058 Choose if the average level between all channels or the louder channel affects
1060 Default is @code{average}. Can be @code{average} or @code{maximum}.
1065 Apply an arbitrary Infinite Impulse Response filter.
1067 It accepts the following parameters:
1071 Set numerator/zeros coefficients.
1074 Set denominator/poles coefficients.
1086 Set coefficients format.
1092 Z-plane zeros/poles, cartesian (default)
1094 Z-plane zeros/poles, polar radians
1096 Z-plane zeros/poles, polar degrees
1100 Set kind of processing.
1101 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1104 Set filtering precision.
1108 double-precision floating-point (default)
1110 single-precision floating-point
1119 Coefficients in @code{tf} format are separated by spaces and are in ascending
1122 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1123 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1126 Different coefficients and gains can be provided for every channel, in such case
1127 use '|' to separate coefficients or gains. Last provided coefficients will be
1128 used for all remaining channels.
1130 @subsection Examples
1134 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1136 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1140 Same as above but in @code{zp} format:
1142 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1148 The limiter prevents an input signal from rising over a desired threshold.
1149 This limiter uses lookahead technology to prevent your signal from distorting.
1150 It means that there is a small delay after the signal is processed. Keep in mind
1151 that the delay it produces is the attack time you set.
1153 The filter accepts the following options:
1157 Set input gain. Default is 1.
1160 Set output gain. Default is 1.
1163 Don't let signals above this level pass the limiter. Default is 1.
1166 The limiter will reach its attenuation level in this amount of time in
1167 milliseconds. Default is 5 milliseconds.
1170 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1171 Default is 50 milliseconds.
1174 When gain reduction is always needed ASC takes care of releasing to an
1175 average reduction level rather than reaching a reduction of 0 in the release
1179 Select how much the release time is affected by ASC, 0 means nearly no changes
1180 in release time while 1 produces higher release times.
1183 Auto level output signal. Default is enabled.
1184 This normalizes audio back to 0dB if enabled.
1187 Depending on picked setting it is recommended to upsample input 2x or 4x times
1188 with @ref{aresample} before applying this filter.
1192 Apply a two-pole all-pass filter with central frequency (in Hz)
1193 @var{frequency}, and filter-width @var{width}.
1194 An all-pass filter changes the audio's frequency to phase relationship
1195 without changing its frequency to amplitude relationship.
1197 The filter accepts the following options:
1201 Set frequency in Hz.
1204 Set method to specify band-width of filter.
1219 Specify the band-width of a filter in width_type units.
1222 Specify which channels to filter, by default all available are filtered.
1225 @subsection Commands
1227 This filter supports the following commands:
1230 Change allpass frequency.
1231 Syntax for the command is : "@var{frequency}"
1234 Change allpass width_type.
1235 Syntax for the command is : "@var{width_type}"
1238 Change allpass width.
1239 Syntax for the command is : "@var{width}"
1246 The filter accepts the following options:
1250 Set the number of loops. Setting this value to -1 will result in infinite loops.
1254 Set maximal number of samples. Default is 0.
1257 Set first sample of loop. Default is 0.
1263 Merge two or more audio streams into a single multi-channel stream.
1265 The filter accepts the following options:
1270 Set the number of inputs. Default is 2.
1274 If the channel layouts of the inputs are disjoint, and therefore compatible,
1275 the channel layout of the output will be set accordingly and the channels
1276 will be reordered as necessary. If the channel layouts of the inputs are not
1277 disjoint, the output will have all the channels of the first input then all
1278 the channels of the second input, in that order, and the channel layout of
1279 the output will be the default value corresponding to the total number of
1282 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1283 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1284 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1285 first input, b1 is the first channel of the second input).
1287 On the other hand, if both input are in stereo, the output channels will be
1288 in the default order: a1, a2, b1, b2, and the channel layout will be
1289 arbitrarily set to 4.0, which may or may not be the expected value.
1291 All inputs must have the same sample rate, and format.
1293 If inputs do not have the same duration, the output will stop with the
1296 @subsection Examples
1300 Merge two mono files into a stereo stream:
1302 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1306 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1308 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
1314 Mixes multiple audio inputs into a single output.
1316 Note that this filter only supports float samples (the @var{amerge}
1317 and @var{pan} audio filters support many formats). If the @var{amix}
1318 input has integer samples then @ref{aresample} will be automatically
1319 inserted to perform the conversion to float samples.
1323 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1325 will mix 3 input audio streams to a single output with the same duration as the
1326 first input and a dropout transition time of 3 seconds.
1328 It accepts the following parameters:
1332 The number of inputs. If unspecified, it defaults to 2.
1335 How to determine the end-of-stream.
1339 The duration of the longest input. (default)
1342 The duration of the shortest input.
1345 The duration of the first input.
1349 @item dropout_transition
1350 The transition time, in seconds, for volume renormalization when an input
1351 stream ends. The default value is 2 seconds.
1355 @section anequalizer
1357 High-order parametric multiband equalizer for each channel.
1359 It accepts the following parameters:
1363 This option string is in format:
1364 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1365 Each equalizer band is separated by '|'.
1369 Set channel number to which equalization will be applied.
1370 If input doesn't have that channel the entry is ignored.
1373 Set central frequency for band.
1374 If input doesn't have that frequency the entry is ignored.
1377 Set band width in hertz.
1380 Set band gain in dB.
1383 Set filter type for band, optional, can be:
1387 Butterworth, this is default.
1398 With this option activated frequency response of anequalizer is displayed
1402 Set video stream size. Only useful if curves option is activated.
1405 Set max gain that will be displayed. Only useful if curves option is activated.
1406 Setting this to a reasonable value makes it possible to display gain which is derived from
1407 neighbour bands which are too close to each other and thus produce higher gain
1408 when both are activated.
1411 Set frequency scale used to draw frequency response in video output.
1412 Can be linear or logarithmic. Default is logarithmic.
1415 Set color for each channel curve which is going to be displayed in video stream.
1416 This is list of color names separated by space or by '|'.
1417 Unrecognised or missing colors will be replaced by white color.
1420 @subsection Examples
1424 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1425 for first 2 channels using Chebyshev type 1 filter:
1427 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1431 @subsection Commands
1433 This filter supports the following commands:
1436 Alter existing filter parameters.
1437 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1439 @var{fN} is existing filter number, starting from 0, if no such filter is available
1441 @var{freq} set new frequency parameter.
1442 @var{width} set new width parameter in herz.
1443 @var{gain} set new gain parameter in dB.
1445 Full filter invocation with asendcmd may look like this:
1446 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1451 Pass the audio source unchanged to the output.
1455 Pad the end of an audio stream with silence.
1457 This can be used together with @command{ffmpeg} @option{-shortest} to
1458 extend audio streams to the same length as the video stream.
1460 A description of the accepted options follows.
1464 Set silence packet size. Default value is 4096.
1467 Set the number of samples of silence to add to the end. After the
1468 value is reached, the stream is terminated. This option is mutually
1469 exclusive with @option{whole_len}.
1472 Set the minimum total number of samples in the output audio stream. If
1473 the value is longer than the input audio length, silence is added to
1474 the end, until the value is reached. This option is mutually exclusive
1475 with @option{pad_len}.
1478 If neither the @option{pad_len} nor the @option{whole_len} option is
1479 set, the filter will add silence to the end of the input stream
1482 @subsection Examples
1486 Add 1024 samples of silence to the end of the input:
1492 Make sure the audio output will contain at least 10000 samples, pad
1493 the input with silence if required:
1495 apad=whole_len=10000
1499 Use @command{ffmpeg} to pad the audio input with silence, so that the
1500 video stream will always result the shortest and will be converted
1501 until the end in the output file when using the @option{shortest}
1504 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1509 Add a phasing effect to the input audio.
1511 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1512 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1514 A description of the accepted parameters follows.
1518 Set input gain. Default is 0.4.
1521 Set output gain. Default is 0.74
1524 Set delay in milliseconds. Default is 3.0.
1527 Set decay. Default is 0.4.
1530 Set modulation speed in Hz. Default is 0.5.
1533 Set modulation type. Default is triangular.
1535 It accepts the following values:
1544 Audio pulsator is something between an autopanner and a tremolo.
1545 But it can produce funny stereo effects as well. Pulsator changes the volume
1546 of the left and right channel based on a LFO (low frequency oscillator) with
1547 different waveforms and shifted phases.
1548 This filter have the ability to define an offset between left and right
1549 channel. An offset of 0 means that both LFO shapes match each other.
1550 The left and right channel are altered equally - a conventional tremolo.
1551 An offset of 50% means that the shape of the right channel is exactly shifted
1552 in phase (or moved backwards about half of the frequency) - pulsator acts as
1553 an autopanner. At 1 both curves match again. Every setting in between moves the
1554 phase shift gapless between all stages and produces some "bypassing" sounds with
1555 sine and triangle waveforms. The more you set the offset near 1 (starting from
1556 the 0.5) the faster the signal passes from the left to the right speaker.
1558 The filter accepts the following options:
1562 Set input gain. By default it is 1. Range is [0.015625 - 64].
1565 Set output gain. By default it is 1. Range is [0.015625 - 64].
1568 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1569 sawup or sawdown. Default is sine.
1572 Set modulation. Define how much of original signal is affected by the LFO.
1575 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1578 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1581 Set pulse width. Default is 1. Allowed range is [0 - 2].
1584 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1587 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1591 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1595 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1596 if timing is set to hz.
1602 Resample the input audio to the specified parameters, using the
1603 libswresample library. If none are specified then the filter will
1604 automatically convert between its input and output.
1606 This filter is also able to stretch/squeeze the audio data to make it match
1607 the timestamps or to inject silence / cut out audio to make it match the
1608 timestamps, do a combination of both or do neither.
1610 The filter accepts the syntax
1611 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1612 expresses a sample rate and @var{resampler_options} is a list of
1613 @var{key}=@var{value} pairs, separated by ":". See the
1614 @ref{Resampler Options,,"Resampler Options" section in the
1615 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1616 for the complete list of supported options.
1618 @subsection Examples
1622 Resample the input audio to 44100Hz:
1628 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1629 samples per second compensation:
1631 aresample=async=1000
1637 Reverse an audio clip.
1639 Warning: This filter requires memory to buffer the entire clip, so trimming
1642 @subsection Examples
1646 Take the first 5 seconds of a clip, and reverse it.
1648 atrim=end=5,areverse
1652 @section asetnsamples
1654 Set the number of samples per each output audio frame.
1656 The last output packet may contain a different number of samples, as
1657 the filter will flush all the remaining samples when the input audio
1660 The filter accepts the following options:
1664 @item nb_out_samples, n
1665 Set the number of frames per each output audio frame. The number is
1666 intended as the number of samples @emph{per each channel}.
1667 Default value is 1024.
1670 If set to 1, the filter will pad the last audio frame with zeroes, so
1671 that the last frame will contain the same number of samples as the
1672 previous ones. Default value is 1.
1675 For example, to set the number of per-frame samples to 1234 and
1676 disable padding for the last frame, use:
1678 asetnsamples=n=1234:p=0
1683 Set the sample rate without altering the PCM data.
1684 This will result in a change of speed and pitch.
1686 The filter accepts the following options:
1689 @item sample_rate, r
1690 Set the output sample rate. Default is 44100 Hz.
1695 Show a line containing various information for each input audio frame.
1696 The input audio is not modified.
1698 The shown line contains a sequence of key/value pairs of the form
1699 @var{key}:@var{value}.
1701 The following values are shown in the output:
1705 The (sequential) number of the input frame, starting from 0.
1708 The presentation timestamp of the input frame, in time base units; the time base
1709 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1712 The presentation timestamp of the input frame in seconds.
1715 position of the frame in the input stream, -1 if this information in
1716 unavailable and/or meaningless (for example in case of synthetic audio)
1725 The sample rate for the audio frame.
1728 The number of samples (per channel) in the frame.
1731 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1732 audio, the data is treated as if all the planes were concatenated.
1734 @item plane_checksums
1735 A list of Adler-32 checksums for each data plane.
1741 Display time domain statistical information about the audio channels.
1742 Statistics are calculated and displayed for each audio channel and,
1743 where applicable, an overall figure is also given.
1745 It accepts the following option:
1748 Short window length in seconds, used for peak and trough RMS measurement.
1749 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1753 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1754 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1757 Available keys for each channel are:
1791 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1792 this @code{lavfi.astats.Overall.Peak_count}.
1794 For description what each key means read below.
1797 Set number of frame after which stats are going to be recalculated.
1798 Default is disabled.
1801 A description of each shown parameter follows:
1805 Mean amplitude displacement from zero.
1808 Minimal sample level.
1811 Maximal sample level.
1813 @item Min difference
1814 Minimal difference between two consecutive samples.
1816 @item Max difference
1817 Maximal difference between two consecutive samples.
1819 @item Mean difference
1820 Mean difference between two consecutive samples.
1821 The average of each difference between two consecutive samples.
1823 @item RMS difference
1824 Root Mean Square difference between two consecutive samples.
1828 Standard peak and RMS level measured in dBFS.
1832 Peak and trough values for RMS level measured over a short window.
1835 Standard ratio of peak to RMS level (note: not in dB).
1838 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1839 (i.e. either @var{Min level} or @var{Max level}).
1842 Number of occasions (not the number of samples) that the signal attained either
1843 @var{Min level} or @var{Max level}.
1846 Overall bit depth of audio. Number of bits used for each sample.
1849 Measured dynamic range of audio in dB.
1856 The filter accepts exactly one parameter, the audio tempo. If not
1857 specified then the filter will assume nominal 1.0 tempo. Tempo must
1858 be in the [0.5, 2.0] range.
1860 @subsection Examples
1864 Slow down audio to 80% tempo:
1870 To speed up audio to 125% tempo:
1878 Trim the input so that the output contains one continuous subpart of the input.
1880 It accepts the following parameters:
1883 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1884 sample with the timestamp @var{start} will be the first sample in the output.
1887 Specify time of the first audio sample that will be dropped, i.e. the
1888 audio sample immediately preceding the one with the timestamp @var{end} will be
1889 the last sample in the output.
1892 Same as @var{start}, except this option sets the start timestamp in samples
1896 Same as @var{end}, except this option sets the end timestamp in samples instead
1900 The maximum duration of the output in seconds.
1903 The number of the first sample that should be output.
1906 The number of the first sample that should be dropped.
1909 @option{start}, @option{end}, and @option{duration} are expressed as time
1910 duration specifications; see
1911 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1913 Note that the first two sets of the start/end options and the @option{duration}
1914 option look at the frame timestamp, while the _sample options simply count the
1915 samples that pass through the filter. So start/end_pts and start/end_sample will
1916 give different results when the timestamps are wrong, inexact or do not start at
1917 zero. Also note that this filter does not modify the timestamps. If you wish
1918 to have the output timestamps start at zero, insert the asetpts filter after the
1921 If multiple start or end options are set, this filter tries to be greedy and
1922 keep all samples that match at least one of the specified constraints. To keep
1923 only the part that matches all the constraints at once, chain multiple atrim
1926 The defaults are such that all the input is kept. So it is possible to set e.g.
1927 just the end values to keep everything before the specified time.
1932 Drop everything except the second minute of input:
1934 ffmpeg -i INPUT -af atrim=60:120
1938 Keep only the first 1000 samples:
1940 ffmpeg -i INPUT -af atrim=end_sample=1000
1947 Apply a two-pole Butterworth band-pass filter with central
1948 frequency @var{frequency}, and (3dB-point) band-width width.
1949 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1950 instead of the default: constant 0dB peak gain.
1951 The filter roll off at 6dB per octave (20dB per decade).
1953 The filter accepts the following options:
1957 Set the filter's central frequency. Default is @code{3000}.
1960 Constant skirt gain if set to 1. Defaults to 0.
1963 Set method to specify band-width of filter.
1978 Specify the band-width of a filter in width_type units.
1981 Specify which channels to filter, by default all available are filtered.
1984 @subsection Commands
1986 This filter supports the following commands:
1989 Change bandpass frequency.
1990 Syntax for the command is : "@var{frequency}"
1993 Change bandpass width_type.
1994 Syntax for the command is : "@var{width_type}"
1997 Change bandpass width.
1998 Syntax for the command is : "@var{width}"
2003 Apply a two-pole Butterworth band-reject filter with central
2004 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2005 The filter roll off at 6dB per octave (20dB per decade).
2007 The filter accepts the following options:
2011 Set the filter's central frequency. Default is @code{3000}.
2014 Set method to specify band-width of filter.
2029 Specify the band-width of a filter in width_type units.
2032 Specify which channels to filter, by default all available are filtered.
2035 @subsection Commands
2037 This filter supports the following commands:
2040 Change bandreject frequency.
2041 Syntax for the command is : "@var{frequency}"
2044 Change bandreject width_type.
2045 Syntax for the command is : "@var{width_type}"
2048 Change bandreject width.
2049 Syntax for the command is : "@var{width}"
2054 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2055 shelving filter with a response similar to that of a standard
2056 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2058 The filter accepts the following options:
2062 Give the gain at 0 Hz. Its useful range is about -20
2063 (for a large cut) to +20 (for a large boost).
2064 Beware of clipping when using a positive gain.
2067 Set the filter's central frequency and so can be used
2068 to extend or reduce the frequency range to be boosted or cut.
2069 The default value is @code{100} Hz.
2072 Set method to specify band-width of filter.
2087 Determine how steep is the filter's shelf transition.
2090 Specify which channels to filter, by default all available are filtered.
2093 @subsection Commands
2095 This filter supports the following commands:
2098 Change bass frequency.
2099 Syntax for the command is : "@var{frequency}"
2102 Change bass width_type.
2103 Syntax for the command is : "@var{width_type}"
2107 Syntax for the command is : "@var{width}"
2111 Syntax for the command is : "@var{gain}"
2116 Apply a biquad IIR filter with the given coefficients.
2117 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2118 are the numerator and denominator coefficients respectively.
2119 and @var{channels}, @var{c} specify which channels to filter, by default all
2120 available are filtered.
2122 @subsection Commands
2124 This filter supports the following commands:
2132 Change biquad parameter.
2133 Syntax for the command is : "@var{value}"
2137 Bauer stereo to binaural transformation, which improves headphone listening of
2138 stereo audio records.
2140 To enable compilation of this filter you need to configure FFmpeg with
2141 @code{--enable-libbs2b}.
2143 It accepts the following parameters:
2147 Pre-defined crossfeed level.
2151 Default level (fcut=700, feed=50).
2154 Chu Moy circuit (fcut=700, feed=60).
2157 Jan Meier circuit (fcut=650, feed=95).
2162 Cut frequency (in Hz).
2171 Remap input channels to new locations.
2173 It accepts the following parameters:
2176 Map channels from input to output. The argument is a '|'-separated list of
2177 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2178 @var{in_channel} form. @var{in_channel} can be either the name of the input
2179 channel (e.g. FL for front left) or its index in the input channel layout.
2180 @var{out_channel} is the name of the output channel or its index in the output
2181 channel layout. If @var{out_channel} is not given then it is implicitly an
2182 index, starting with zero and increasing by one for each mapping.
2184 @item channel_layout
2185 The channel layout of the output stream.
2188 If no mapping is present, the filter will implicitly map input channels to
2189 output channels, preserving indices.
2191 @subsection Examples
2195 For example, assuming a 5.1+downmix input MOV file,
2197 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2199 will create an output WAV file tagged as stereo from the downmix channels of
2203 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2205 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2209 @section channelsplit
2211 Split each channel from an input audio stream into a separate output stream.
2213 It accepts the following parameters:
2215 @item channel_layout
2216 The channel layout of the input stream. The default is "stereo".
2218 A channel layout describing the channels to be extracted as separate output streams
2219 or "all" to extract each input channel as a separate stream. The default is "all".
2221 Choosing channels not present in channel layout in the input will result in an error.
2224 @subsection Examples
2228 For example, assuming a stereo input MP3 file,
2230 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2232 will create an output Matroska file with two audio streams, one containing only
2233 the left channel and the other the right channel.
2236 Split a 5.1 WAV file into per-channel files:
2238 ffmpeg -i in.wav -filter_complex
2239 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2240 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2241 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2246 Extract only LFE from a 5.1 WAV file:
2248 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2249 -map '[LFE]' lfe.wav
2254 Add a chorus effect to the audio.
2256 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2258 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2259 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2260 The modulation depth defines the range the modulated delay is played before or after
2261 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2262 sound tuned around the original one, like in a chorus where some vocals are slightly
2265 It accepts the following parameters:
2268 Set input gain. Default is 0.4.
2271 Set output gain. Default is 0.4.
2274 Set delays. A typical delay is around 40ms to 60ms.
2286 @subsection Examples
2292 chorus=0.7:0.9:55:0.4:0.25:2
2298 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2302 Fuller sounding chorus with three delays:
2304 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
2309 Compress or expand the audio's dynamic range.
2311 It accepts the following parameters:
2317 A list of times in seconds for each channel over which the instantaneous level
2318 of the input signal is averaged to determine its volume. @var{attacks} refers to
2319 increase of volume and @var{decays} refers to decrease of volume. For most
2320 situations, the attack time (response to the audio getting louder) should be
2321 shorter than the decay time, because the human ear is more sensitive to sudden
2322 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2323 a typical value for decay is 0.8 seconds.
2324 If specified number of attacks & decays is lower than number of channels, the last
2325 set attack/decay will be used for all remaining channels.
2328 A list of points for the transfer function, specified in dB relative to the
2329 maximum possible signal amplitude. Each key points list must be defined using
2330 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2331 @code{x0/y0 x1/y1 x2/y2 ....}
2333 The input values must be in strictly increasing order but the transfer function
2334 does not have to be monotonically rising. The point @code{0/0} is assumed but
2335 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2336 function are @code{-70/-70|-60/-20|1/0}.
2339 Set the curve radius in dB for all joints. It defaults to 0.01.
2342 Set the additional gain in dB to be applied at all points on the transfer
2343 function. This allows for easy adjustment of the overall gain.
2347 Set an initial volume, in dB, to be assumed for each channel when filtering
2348 starts. This permits the user to supply a nominal level initially, so that, for
2349 example, a very large gain is not applied to initial signal levels before the
2350 companding has begun to operate. A typical value for audio which is initially
2351 quiet is -90 dB. It defaults to 0.
2354 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2355 delayed before being fed to the volume adjuster. Specifying a delay
2356 approximately equal to the attack/decay times allows the filter to effectively
2357 operate in predictive rather than reactive mode. It defaults to 0.
2361 @subsection Examples
2365 Make music with both quiet and loud passages suitable for listening to in a
2368 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2371 Another example for audio with whisper and explosion parts:
2373 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2377 A noise gate for when the noise is at a lower level than the signal:
2379 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2383 Here is another noise gate, this time for when the noise is at a higher level
2384 than the signal (making it, in some ways, similar to squelch):
2386 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2390 2:1 compression starting at -6dB:
2392 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2396 2:1 compression starting at -9dB:
2398 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2402 2:1 compression starting at -12dB:
2404 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2408 2:1 compression starting at -18dB:
2410 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2414 3:1 compression starting at -15dB:
2416 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2422 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2428 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
2432 Hard limiter at -6dB:
2434 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2438 Hard limiter at -12dB:
2440 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2444 Hard noise gate at -35 dB:
2446 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2452 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2456 @section compensationdelay
2458 Compensation Delay Line is a metric based delay to compensate differing
2459 positions of microphones or speakers.
2461 For example, you have recorded guitar with two microphones placed in
2462 different location. Because the front of sound wave has fixed speed in
2463 normal conditions, the phasing of microphones can vary and depends on
2464 their location and interposition. The best sound mix can be achieved when
2465 these microphones are in phase (synchronized). Note that distance of
2466 ~30 cm between microphones makes one microphone to capture signal in
2467 antiphase to another microphone. That makes the final mix sounding moody.
2468 This filter helps to solve phasing problems by adding different delays
2469 to each microphone track and make them synchronized.
2471 The best result can be reached when you take one track as base and
2472 synchronize other tracks one by one with it.
2473 Remember that synchronization/delay tolerance depends on sample rate, too.
2474 Higher sample rates will give more tolerance.
2476 It accepts the following parameters:
2480 Set millimeters distance. This is compensation distance for fine tuning.
2484 Set cm distance. This is compensation distance for tightening distance setup.
2488 Set meters distance. This is compensation distance for hard distance setup.
2492 Set dry amount. Amount of unprocessed (dry) signal.
2496 Set wet amount. Amount of processed (wet) signal.
2500 Set temperature degree in Celsius. This is the temperature of the environment.
2505 Apply headphone crossfeed filter.
2507 Crossfeed is the process of blending the left and right channels of stereo
2509 It is mainly used to reduce extreme stereo separation of low frequencies.
2511 The intent is to produce more speaker like sound to the listener.
2513 The filter accepts the following options:
2517 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2518 This sets gain of low shelf filter for side part of stereo image.
2519 Default is -6dB. Max allowed is -30db when strength is set to 1.
2522 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2523 This sets cut off frequency of low shelf filter. Default is cut off near
2524 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2527 Set input gain. Default is 0.9.
2530 Set output gain. Default is 1.
2533 @section crystalizer
2534 Simple algorithm to expand audio dynamic range.
2536 The filter accepts the following options:
2540 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2541 (unchanged sound) to 10.0 (maximum effect).
2544 Enable clipping. By default is enabled.
2548 Apply a DC shift to the audio.
2550 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2551 in the recording chain) from the audio. The effect of a DC offset is reduced
2552 headroom and hence volume. The @ref{astats} filter can be used to determine if
2553 a signal has a DC offset.
2557 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2561 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2562 used to prevent clipping.
2566 Measure audio dynamic range.
2568 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2569 is found in transition material. And anything less that 8 have very poor dynamics
2570 and is very compressed.
2572 The filter accepts the following options:
2576 Set window length in seconds used to split audio into segments of equal length.
2577 Default is 3 seconds.
2581 Dynamic Audio Normalizer.
2583 This filter applies a certain amount of gain to the input audio in order
2584 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2585 contrast to more "simple" normalization algorithms, the Dynamic Audio
2586 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2587 This allows for applying extra gain to the "quiet" sections of the audio
2588 while avoiding distortions or clipping the "loud" sections. In other words:
2589 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2590 sections, in the sense that the volume of each section is brought to the
2591 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2592 this goal *without* applying "dynamic range compressing". It will retain 100%
2593 of the dynamic range *within* each section of the audio file.
2597 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2598 Default is 500 milliseconds.
2599 The Dynamic Audio Normalizer processes the input audio in small chunks,
2600 referred to as frames. This is required, because a peak magnitude has no
2601 meaning for just a single sample value. Instead, we need to determine the
2602 peak magnitude for a contiguous sequence of sample values. While a "standard"
2603 normalizer would simply use the peak magnitude of the complete file, the
2604 Dynamic Audio Normalizer determines the peak magnitude individually for each
2605 frame. The length of a frame is specified in milliseconds. By default, the
2606 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2607 been found to give good results with most files.
2608 Note that the exact frame length, in number of samples, will be determined
2609 automatically, based on the sampling rate of the individual input audio file.
2612 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2613 number. Default is 31.
2614 Probably the most important parameter of the Dynamic Audio Normalizer is the
2615 @code{window size} of the Gaussian smoothing filter. The filter's window size
2616 is specified in frames, centered around the current frame. For the sake of
2617 simplicity, this must be an odd number. Consequently, the default value of 31
2618 takes into account the current frame, as well as the 15 preceding frames and
2619 the 15 subsequent frames. Using a larger window results in a stronger
2620 smoothing effect and thus in less gain variation, i.e. slower gain
2621 adaptation. Conversely, using a smaller window results in a weaker smoothing
2622 effect and thus in more gain variation, i.e. faster gain adaptation.
2623 In other words, the more you increase this value, the more the Dynamic Audio
2624 Normalizer will behave like a "traditional" normalization filter. On the
2625 contrary, the more you decrease this value, the more the Dynamic Audio
2626 Normalizer will behave like a dynamic range compressor.
2629 Set the target peak value. This specifies the highest permissible magnitude
2630 level for the normalized audio input. This filter will try to approach the
2631 target peak magnitude as closely as possible, but at the same time it also
2632 makes sure that the normalized signal will never exceed the peak magnitude.
2633 A frame's maximum local gain factor is imposed directly by the target peak
2634 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2635 It is not recommended to go above this value.
2638 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2639 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2640 factor for each input frame, i.e. the maximum gain factor that does not
2641 result in clipping or distortion. The maximum gain factor is determined by
2642 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2643 additionally bounds the frame's maximum gain factor by a predetermined
2644 (global) maximum gain factor. This is done in order to avoid excessive gain
2645 factors in "silent" or almost silent frames. By default, the maximum gain
2646 factor is 10.0, For most inputs the default value should be sufficient and
2647 it usually is not recommended to increase this value. Though, for input
2648 with an extremely low overall volume level, it may be necessary to allow even
2649 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2650 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2651 Instead, a "sigmoid" threshold function will be applied. This way, the
2652 gain factors will smoothly approach the threshold value, but never exceed that
2656 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2657 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2658 This means that the maximum local gain factor for each frame is defined
2659 (only) by the frame's highest magnitude sample. This way, the samples can
2660 be amplified as much as possible without exceeding the maximum signal
2661 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2662 Normalizer can also take into account the frame's root mean square,
2663 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2664 determine the power of a time-varying signal. It is therefore considered
2665 that the RMS is a better approximation of the "perceived loudness" than
2666 just looking at the signal's peak magnitude. Consequently, by adjusting all
2667 frames to a constant RMS value, a uniform "perceived loudness" can be
2668 established. If a target RMS value has been specified, a frame's local gain
2669 factor is defined as the factor that would result in exactly that RMS value.
2670 Note, however, that the maximum local gain factor is still restricted by the
2671 frame's highest magnitude sample, in order to prevent clipping.
2674 Enable channels coupling. By default is enabled.
2675 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2676 amount. This means the same gain factor will be applied to all channels, i.e.
2677 the maximum possible gain factor is determined by the "loudest" channel.
2678 However, in some recordings, it may happen that the volume of the different
2679 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2680 In this case, this option can be used to disable the channel coupling. This way,
2681 the gain factor will be determined independently for each channel, depending
2682 only on the individual channel's highest magnitude sample. This allows for
2683 harmonizing the volume of the different channels.
2686 Enable DC bias correction. By default is disabled.
2687 An audio signal (in the time domain) is a sequence of sample values.
2688 In the Dynamic Audio Normalizer these sample values are represented in the
2689 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2690 audio signal, or "waveform", should be centered around the zero point.
2691 That means if we calculate the mean value of all samples in a file, or in a
2692 single frame, then the result should be 0.0 or at least very close to that
2693 value. If, however, there is a significant deviation of the mean value from
2694 0.0, in either positive or negative direction, this is referred to as a
2695 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2696 Audio Normalizer provides optional DC bias correction.
2697 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2698 the mean value, or "DC correction" offset, of each input frame and subtract
2699 that value from all of the frame's sample values which ensures those samples
2700 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2701 boundaries, the DC correction offset values will be interpolated smoothly
2702 between neighbouring frames.
2705 Enable alternative boundary mode. By default is disabled.
2706 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2707 around each frame. This includes the preceding frames as well as the
2708 subsequent frames. However, for the "boundary" frames, located at the very
2709 beginning and at the very end of the audio file, not all neighbouring
2710 frames are available. In particular, for the first few frames in the audio
2711 file, the preceding frames are not known. And, similarly, for the last few
2712 frames in the audio file, the subsequent frames are not known. Thus, the
2713 question arises which gain factors should be assumed for the missing frames
2714 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2715 to deal with this situation. The default boundary mode assumes a gain factor
2716 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2717 "fade out" at the beginning and at the end of the input, respectively.
2720 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2721 By default, the Dynamic Audio Normalizer does not apply "traditional"
2722 compression. This means that signal peaks will not be pruned and thus the
2723 full dynamic range will be retained within each local neighbourhood. However,
2724 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2725 normalization algorithm with a more "traditional" compression.
2726 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2727 (thresholding) function. If (and only if) the compression feature is enabled,
2728 all input frames will be processed by a soft knee thresholding function prior
2729 to the actual normalization process. Put simply, the thresholding function is
2730 going to prune all samples whose magnitude exceeds a certain threshold value.
2731 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2732 value. Instead, the threshold value will be adjusted for each individual
2734 In general, smaller parameters result in stronger compression, and vice versa.
2735 Values below 3.0 are not recommended, because audible distortion may appear.
2740 Make audio easier to listen to on headphones.
2742 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2743 so that when listened to on headphones the stereo image is moved from
2744 inside your head (standard for headphones) to outside and in front of
2745 the listener (standard for speakers).
2751 Apply a two-pole peaking equalisation (EQ) filter. With this
2752 filter, the signal-level at and around a selected frequency can
2753 be increased or decreased, whilst (unlike bandpass and bandreject
2754 filters) that at all other frequencies is unchanged.
2756 In order to produce complex equalisation curves, this filter can
2757 be given several times, each with a different central frequency.
2759 The filter accepts the following options:
2763 Set the filter's central frequency in Hz.
2766 Set method to specify band-width of filter.
2781 Specify the band-width of a filter in width_type units.
2784 Set the required gain or attenuation in dB.
2785 Beware of clipping when using a positive gain.
2788 Specify which channels to filter, by default all available are filtered.
2791 @subsection Examples
2794 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2796 equalizer=f=1000:t=h:width=200:g=-10
2800 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2802 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2806 @subsection Commands
2808 This filter supports the following commands:
2811 Change equalizer frequency.
2812 Syntax for the command is : "@var{frequency}"
2815 Change equalizer width_type.
2816 Syntax for the command is : "@var{width_type}"
2819 Change equalizer width.
2820 Syntax for the command is : "@var{width}"
2823 Change equalizer gain.
2824 Syntax for the command is : "@var{gain}"
2827 @section extrastereo
2829 Linearly increases the difference between left and right channels which
2830 adds some sort of "live" effect to playback.
2832 The filter accepts the following options:
2836 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2837 (average of both channels), with 1.0 sound will be unchanged, with
2838 -1.0 left and right channels will be swapped.
2841 Enable clipping. By default is enabled.
2844 @section firequalizer
2845 Apply FIR Equalization using arbitrary frequency response.
2847 The filter accepts the following option:
2851 Set gain curve equation (in dB). The expression can contain variables:
2854 the evaluated frequency
2858 channel number, set to 0 when multichannels evaluation is disabled
2860 channel id, see libavutil/channel_layout.h, set to the first channel id when
2861 multichannels evaluation is disabled
2865 channel_layout, see libavutil/channel_layout.h
2870 @item gain_interpolate(f)
2871 interpolate gain on frequency f based on gain_entry
2872 @item cubic_interpolate(f)
2873 same as gain_interpolate, but smoother
2875 This option is also available as command. Default is @code{gain_interpolate(f)}.
2878 Set gain entry for gain_interpolate function. The expression can
2882 store gain entry at frequency f with value g
2884 This option is also available as command.
2887 Set filter delay in seconds. Higher value means more accurate.
2888 Default is @code{0.01}.
2891 Set filter accuracy in Hz. Lower value means more accurate.
2892 Default is @code{5}.
2895 Set window function. Acceptable values are:
2898 rectangular window, useful when gain curve is already smooth
2900 hann window (default)
2906 3-terms continuous 1st derivative nuttall window
2908 minimum 3-terms discontinuous nuttall window
2910 4-terms continuous 1st derivative nuttall window
2912 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2914 blackman-harris window
2920 If enabled, use fixed number of audio samples. This improves speed when
2921 filtering with large delay. Default is disabled.
2924 Enable multichannels evaluation on gain. Default is disabled.
2927 Enable zero phase mode by subtracting timestamp to compensate delay.
2928 Default is disabled.
2931 Set scale used by gain. Acceptable values are:
2934 linear frequency, linear gain
2936 linear frequency, logarithmic (in dB) gain (default)
2938 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2940 logarithmic frequency, logarithmic gain
2944 Set file for dumping, suitable for gnuplot.
2947 Set scale for dumpfile. Acceptable values are same with scale option.
2951 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2952 Default is disabled.
2955 Enable minimum phase impulse response. Default is disabled.
2958 @subsection Examples
2963 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2966 lowpass at 1000 Hz with gain_entry:
2968 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2971 custom equalization:
2973 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2976 higher delay with zero phase to compensate delay:
2978 firequalizer=delay=0.1:fixed=on:zero_phase=on
2981 lowpass on left channel, highpass on right channel:
2983 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2984 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2989 Apply a flanging effect to the audio.
2991 The filter accepts the following options:
2995 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2998 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3001 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3005 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3006 Default value is 71.
3009 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3012 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3013 Default value is @var{sinusoidal}.
3016 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3017 Default value is 25.
3020 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3021 Default is @var{linear}.
3025 Apply Haas effect to audio.
3027 Note that this makes most sense to apply on mono signals.
3028 With this filter applied to mono signals it give some directionality and
3029 stretches its stereo image.
3031 The filter accepts the following options:
3035 Set input level. By default is @var{1}, or 0dB
3038 Set output level. By default is @var{1}, or 0dB.
3041 Set gain applied to side part of signal. By default is @var{1}.
3044 Set kind of middle source. Can be one of the following:
3054 Pick middle part signal of stereo image.
3057 Pick side part signal of stereo image.
3061 Change middle phase. By default is disabled.
3064 Set left channel delay. By default is @var{2.05} milliseconds.
3067 Set left channel balance. By default is @var{-1}.
3070 Set left channel gain. By default is @var{1}.
3073 Change left phase. By default is disabled.
3076 Set right channel delay. By defaults is @var{2.12} milliseconds.
3079 Set right channel balance. By default is @var{1}.
3082 Set right channel gain. By default is @var{1}.
3085 Change right phase. By default is enabled.
3090 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3091 embedded HDCD codes is expanded into a 20-bit PCM stream.
3093 The filter supports the Peak Extend and Low-level Gain Adjustment features
3094 of HDCD, and detects the Transient Filter flag.
3097 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3100 When using the filter with wav, note the default encoding for wav is 16-bit,
3101 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3102 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3104 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3105 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3108 The filter accepts the following options:
3111 @item disable_autoconvert
3112 Disable any automatic format conversion or resampling in the filter graph.
3114 @item process_stereo
3115 Process the stereo channels together. If target_gain does not match between
3116 channels, consider it invalid and use the last valid target_gain.
3119 Set the code detect timer period in ms.
3122 Always extend peaks above -3dBFS even if PE isn't signaled.
3125 Replace audio with a solid tone and adjust the amplitude to signal some
3126 specific aspect of the decoding process. The output file can be loaded in
3127 an audio editor alongside the original to aid analysis.
3129 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3136 Gain adjustment level at each sample
3138 Samples where peak extend occurs
3140 Samples where the code detect timer is active
3142 Samples where the target gain does not match between channels
3148 Apply head-related transfer functions (HRTFs) to create virtual
3149 loudspeakers around the user for binaural listening via headphones.
3150 The HRIRs are provided via additional streams, for each channel
3151 one stereo input stream is needed.
3153 The filter accepts the following options:
3157 Set mapping of input streams for convolution.
3158 The argument is a '|'-separated list of channel names in order as they
3159 are given as additional stream inputs for filter.
3160 This also specify number of input streams. Number of input streams
3161 must be not less than number of channels in first stream plus one.
3164 Set gain applied to audio. Value is in dB. Default is 0.
3167 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3168 processing audio in time domain which is slow.
3169 @var{freq} is processing audio in frequency domain which is fast.
3170 Default is @var{freq}.
3173 Set custom gain for LFE channels. Value is in dB. Default is 0.
3176 @subsection Examples
3180 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3181 each amovie filter use stereo file with IR coefficients as input.
3182 The files give coefficients for each position of virtual loudspeaker:
3184 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"
3191 Apply a high-pass filter with 3dB point frequency.
3192 The filter can be either single-pole, or double-pole (the default).
3193 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3195 The filter accepts the following options:
3199 Set frequency in Hz. Default is 3000.
3202 Set number of poles. Default is 2.
3205 Set method to specify band-width of filter.
3220 Specify the band-width of a filter in width_type units.
3221 Applies only to double-pole filter.
3222 The default is 0.707q and gives a Butterworth response.
3225 Specify which channels to filter, by default all available are filtered.
3228 @subsection Commands
3230 This filter supports the following commands:
3233 Change highpass frequency.
3234 Syntax for the command is : "@var{frequency}"
3237 Change highpass width_type.
3238 Syntax for the command is : "@var{width_type}"
3241 Change highpass width.
3242 Syntax for the command is : "@var{width}"
3247 Join multiple input streams into one multi-channel stream.
3249 It accepts the following parameters:
3253 The number of input streams. It defaults to 2.
3255 @item channel_layout
3256 The desired output channel layout. It defaults to stereo.
3259 Map channels from inputs to output. The argument is a '|'-separated list of
3260 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3261 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3262 can be either the name of the input channel (e.g. FL for front left) or its
3263 index in the specified input stream. @var{out_channel} is the name of the output
3267 The filter will attempt to guess the mappings when they are not specified
3268 explicitly. It does so by first trying to find an unused matching input channel
3269 and if that fails it picks the first unused input channel.
3271 Join 3 inputs (with properly set channel layouts):
3273 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3276 Build a 5.1 output from 6 single-channel streams:
3278 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3279 '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'
3285 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3287 To enable compilation of this filter you need to configure FFmpeg with
3288 @code{--enable-ladspa}.
3292 Specifies the name of LADSPA plugin library to load. If the environment
3293 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3294 each one of the directories specified by the colon separated list in
3295 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3296 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3297 @file{/usr/lib/ladspa/}.
3300 Specifies the plugin within the library. Some libraries contain only
3301 one plugin, but others contain many of them. If this is not set filter
3302 will list all available plugins within the specified library.
3305 Set the '|' separated list of controls which are zero or more floating point
3306 values that determine the behavior of the loaded plugin (for example delay,
3308 Controls need to be defined using the following syntax:
3309 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3310 @var{valuei} is the value set on the @var{i}-th control.
3311 Alternatively they can be also defined using the following syntax:
3312 @var{value0}|@var{value1}|@var{value2}|..., where
3313 @var{valuei} is the value set on the @var{i}-th control.
3314 If @option{controls} is set to @code{help}, all available controls and
3315 their valid ranges are printed.
3317 @item sample_rate, s
3318 Specify the sample rate, default to 44100. Only used if plugin have
3322 Set the number of samples per channel per each output frame, default
3323 is 1024. Only used if plugin have zero inputs.
3326 Set the minimum duration of the sourced audio. See
3327 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3328 for the accepted syntax.
3329 Note that the resulting duration may be greater than the specified duration,
3330 as the generated audio is always cut at the end of a complete frame.
3331 If not specified, or the expressed duration is negative, the audio is
3332 supposed to be generated forever.
3333 Only used if plugin have zero inputs.
3337 @subsection Examples
3341 List all available plugins within amp (LADSPA example plugin) library:
3347 List all available controls and their valid ranges for @code{vcf_notch}
3348 plugin from @code{VCF} library:
3350 ladspa=f=vcf:p=vcf_notch:c=help
3354 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3357 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3361 Add reverberation to the audio using TAP-plugins
3362 (Tom's Audio Processing plugins):
3364 ladspa=file=tap_reverb:tap_reverb
3368 Generate white noise, with 0.2 amplitude:
3370 ladspa=file=cmt:noise_source_white:c=c0=.2
3374 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3375 @code{C* Audio Plugin Suite} (CAPS) library:
3377 ladspa=file=caps:Click:c=c1=20'
3381 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3383 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3387 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3388 @code{SWH Plugins} collection:
3390 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3394 Attenuate low frequencies using Multiband EQ from Steve Harris
3395 @code{SWH Plugins} collection:
3397 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3401 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3404 ladspa=caps:Narrower
3408 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3410 ladspa=caps:White:.2
3414 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3416 ladspa=caps:Fractal:c=c1=1
3420 Dynamic volume normalization using @code{VLevel} plugin:
3422 ladspa=vlevel-ladspa:vlevel_mono
3426 @subsection Commands
3428 This filter supports the following commands:
3431 Modify the @var{N}-th control value.
3433 If the specified value is not valid, it is ignored and prior one is kept.
3438 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3439 Support for both single pass (livestreams, files) and double pass (files) modes.
3440 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3441 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3442 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3444 The filter accepts the following options:
3448 Set integrated loudness target.
3449 Range is -70.0 - -5.0. Default value is -24.0.
3452 Set loudness range target.
3453 Range is 1.0 - 20.0. Default value is 7.0.
3456 Set maximum true peak.
3457 Range is -9.0 - +0.0. Default value is -2.0.
3459 @item measured_I, measured_i
3460 Measured IL of input file.
3461 Range is -99.0 - +0.0.
3463 @item measured_LRA, measured_lra
3464 Measured LRA of input file.
3465 Range is 0.0 - 99.0.
3467 @item measured_TP, measured_tp
3468 Measured true peak of input file.
3469 Range is -99.0 - +99.0.
3471 @item measured_thresh
3472 Measured threshold of input file.
3473 Range is -99.0 - +0.0.
3476 Set offset gain. Gain is applied before the true-peak limiter.
3477 Range is -99.0 - +99.0. Default is +0.0.
3480 Normalize linearly if possible.
3481 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3482 to be specified in order to use this mode.
3483 Options are true or false. Default is true.
3486 Treat mono input files as "dual-mono". If a mono file is intended for playback
3487 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3488 If set to @code{true}, this option will compensate for this effect.
3489 Multi-channel input files are not affected by this option.
3490 Options are true or false. Default is false.
3493 Set print format for stats. Options are summary, json, or none.
3494 Default value is none.
3499 Apply a low-pass filter with 3dB point frequency.
3500 The filter can be either single-pole or double-pole (the default).
3501 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3503 The filter accepts the following options:
3507 Set frequency in Hz. Default is 500.
3510 Set number of poles. Default is 2.
3513 Set method to specify band-width of filter.
3528 Specify the band-width of a filter in width_type units.
3529 Applies only to double-pole filter.
3530 The default is 0.707q and gives a Butterworth response.
3533 Specify which channels to filter, by default all available are filtered.
3536 @subsection Examples
3539 Lowpass only LFE channel, it LFE is not present it does nothing:
3545 @subsection Commands
3547 This filter supports the following commands:
3550 Change lowpass frequency.
3551 Syntax for the command is : "@var{frequency}"
3554 Change lowpass width_type.
3555 Syntax for the command is : "@var{width_type}"
3558 Change lowpass width.
3559 Syntax for the command is : "@var{width}"
3564 Load a LV2 (LADSPA Version 2) plugin.
3566 To enable compilation of this filter you need to configure FFmpeg with
3567 @code{--enable-lv2}.
3571 Specifies the plugin URI. You may need to escape ':'.
3574 Set the '|' separated list of controls which are zero or more floating point
3575 values that determine the behavior of the loaded plugin (for example delay,
3577 If @option{controls} is set to @code{help}, all available controls and
3578 their valid ranges are printed.
3580 @item sample_rate, s
3581 Specify the sample rate, default to 44100. Only used if plugin have
3585 Set the number of samples per channel per each output frame, default
3586 is 1024. Only used if plugin have zero inputs.
3589 Set the minimum duration of the sourced audio. See
3590 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3591 for the accepted syntax.
3592 Note that the resulting duration may be greater than the specified duration,
3593 as the generated audio is always cut at the end of a complete frame.
3594 If not specified, or the expressed duration is negative, the audio is
3595 supposed to be generated forever.
3596 Only used if plugin have zero inputs.
3599 @subsection Examples
3603 Apply bass enhancer plugin from Calf:
3605 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3609 Apply vinyl plugin from Calf:
3611 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3615 Apply bit crusher plugin from ArtyFX:
3617 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3622 Multiband Compress or expand the audio's dynamic range.
3624 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3625 This is akin to the crossover of a loudspeaker, and results in flat frequency
3626 response when absent compander action.
3628 It accepts the following parameters:
3632 This option syntax is:
3633 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3634 For explanation of each item refer to compand filter documentation.
3640 Mix channels with specific gain levels. The filter accepts the output
3641 channel layout followed by a set of channels definitions.
3643 This filter is also designed to efficiently remap the channels of an audio
3646 The filter accepts parameters of the form:
3647 "@var{l}|@var{outdef}|@var{outdef}|..."
3651 output channel layout or number of channels
3654 output channel specification, of the form:
3655 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3658 output channel to define, either a channel name (FL, FR, etc.) or a channel
3659 number (c0, c1, etc.)
3662 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3665 input channel to use, see out_name for details; it is not possible to mix
3666 named and numbered input channels
3669 If the `=' in a channel specification is replaced by `<', then the gains for
3670 that specification will be renormalized so that the total is 1, thus
3671 avoiding clipping noise.
3673 @subsection Mixing examples
3675 For example, if you want to down-mix from stereo to mono, but with a bigger
3676 factor for the left channel:
3678 pan=1c|c0=0.9*c0+0.1*c1
3681 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3682 7-channels surround:
3684 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3687 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3688 that should be preferred (see "-ac" option) unless you have very specific
3691 @subsection Remapping examples
3693 The channel remapping will be effective if, and only if:
3696 @item gain coefficients are zeroes or ones,
3697 @item only one input per channel output,
3700 If all these conditions are satisfied, the filter will notify the user ("Pure
3701 channel mapping detected"), and use an optimized and lossless method to do the
3704 For example, if you have a 5.1 source and want a stereo audio stream by
3705 dropping the extra channels:
3707 pan="stereo| c0=FL | c1=FR"
3710 Given the same source, you can also switch front left and front right channels
3711 and keep the input channel layout:
3713 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3716 If the input is a stereo audio stream, you can mute the front left channel (and
3717 still keep the stereo channel layout) with:
3722 Still with a stereo audio stream input, you can copy the right channel in both
3723 front left and right:
3725 pan="stereo| c0=FR | c1=FR"
3730 ReplayGain scanner filter. This filter takes an audio stream as an input and
3731 outputs it unchanged.
3732 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3736 Convert the audio sample format, sample rate and channel layout. It is
3737 not meant to be used directly.
3740 Apply time-stretching and pitch-shifting with librubberband.
3742 The filter accepts the following options:
3746 Set tempo scale factor.
3749 Set pitch scale factor.
3752 Set transients detector.
3753 Possible values are:
3762 Possible values are:
3771 Possible values are:
3778 Set processing window size.
3779 Possible values are:
3788 Possible values are:
3795 Enable formant preservation when shift pitching.
3796 Possible values are:
3804 Possible values are:
3813 Possible values are:
3820 @section sidechaincompress
3822 This filter acts like normal compressor but has the ability to compress
3823 detected signal using second input signal.
3824 It needs two input streams and returns one output stream.
3825 First input stream will be processed depending on second stream signal.
3826 The filtered signal then can be filtered with other filters in later stages of
3827 processing. See @ref{pan} and @ref{amerge} filter.
3829 The filter accepts the following options:
3833 Set input gain. Default is 1. Range is between 0.015625 and 64.
3836 If a signal of second stream raises above this level it will affect the gain
3837 reduction of first stream.
3838 By default is 0.125. Range is between 0.00097563 and 1.
3841 Set a ratio about which the signal is reduced. 1:2 means that if the level
3842 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3843 Default is 2. Range is between 1 and 20.
3846 Amount of milliseconds the signal has to rise above the threshold before gain
3847 reduction starts. Default is 20. Range is between 0.01 and 2000.
3850 Amount of milliseconds the signal has to fall below the threshold before
3851 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3854 Set the amount by how much signal will be amplified after processing.
3855 Default is 1. Range is from 1 to 64.
3858 Curve the sharp knee around the threshold to enter gain reduction more softly.
3859 Default is 2.82843. Range is between 1 and 8.
3862 Choose if the @code{average} level between all channels of side-chain stream
3863 or the louder(@code{maximum}) channel of side-chain stream affects the
3864 reduction. Default is @code{average}.
3867 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3868 of @code{rms}. Default is @code{rms} which is mainly smoother.
3871 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3874 How much to use compressed signal in output. Default is 1.
3875 Range is between 0 and 1.
3878 @subsection Examples
3882 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3883 depending on the signal of 2nd input and later compressed signal to be
3884 merged with 2nd input:
3886 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3890 @section sidechaingate
3892 A sidechain gate acts like a normal (wideband) gate but has the ability to
3893 filter the detected signal before sending it to the gain reduction stage.
3894 Normally a gate uses the full range signal to detect a level above the
3896 For example: If you cut all lower frequencies from your sidechain signal
3897 the gate will decrease the volume of your track only if not enough highs
3898 appear. With this technique you are able to reduce the resonation of a
3899 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3901 It needs two input streams and returns one output stream.
3902 First input stream will be processed depending on second stream signal.
3904 The filter accepts the following options:
3908 Set input level before filtering.
3909 Default is 1. Allowed range is from 0.015625 to 64.
3912 Set the level of gain reduction when the signal is below the threshold.
3913 Default is 0.06125. Allowed range is from 0 to 1.
3916 If a signal rises above this level the gain reduction is released.
3917 Default is 0.125. Allowed range is from 0 to 1.
3920 Set a ratio about which the signal is reduced.
3921 Default is 2. Allowed range is from 1 to 9000.
3924 Amount of milliseconds the signal has to rise above the threshold before gain
3926 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3929 Amount of milliseconds the signal has to fall below the threshold before the
3930 reduction is increased again. Default is 250 milliseconds.
3931 Allowed range is from 0.01 to 9000.
3934 Set amount of amplification of signal after processing.
3935 Default is 1. Allowed range is from 1 to 64.
3938 Curve the sharp knee around the threshold to enter gain reduction more softly.
3939 Default is 2.828427125. Allowed range is from 1 to 8.
3942 Choose if exact signal should be taken for detection or an RMS like one.
3943 Default is rms. Can be peak or rms.
3946 Choose if the average level between all channels or the louder channel affects
3948 Default is average. Can be average or maximum.
3951 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3954 @section silencedetect
3956 Detect silence in an audio stream.
3958 This filter logs a message when it detects that the input audio volume is less
3959 or equal to a noise tolerance value for a duration greater or equal to the
3960 minimum detected noise duration.
3962 The printed times and duration are expressed in seconds.
3964 The filter accepts the following options:
3968 Set silence duration until notification (default is 2 seconds).
3971 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3972 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3975 @subsection Examples
3979 Detect 5 seconds of silence with -50dB noise tolerance:
3981 silencedetect=n=-50dB:d=5
3985 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3986 tolerance in @file{silence.mp3}:
3988 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3992 @section silenceremove
3994 Remove silence from the beginning, middle or end of the audio.
3996 The filter accepts the following options:
4000 This value is used to indicate if audio should be trimmed at beginning of
4001 the audio. A value of zero indicates no silence should be trimmed from the
4002 beginning. When specifying a non-zero value, it trims audio up until it
4003 finds non-silence. Normally, when trimming silence from beginning of audio
4004 the @var{start_periods} will be @code{1} but it can be increased to higher
4005 values to trim all audio up to specific count of non-silence periods.
4006 Default value is @code{0}.
4008 @item start_duration
4009 Specify the amount of time that non-silence must be detected before it stops
4010 trimming audio. By increasing the duration, bursts of noises can be treated
4011 as silence and trimmed off. Default value is @code{0}.
4013 @item start_threshold
4014 This indicates what sample value should be treated as silence. For digital
4015 audio, a value of @code{0} may be fine but for audio recorded from analog,
4016 you may wish to increase the value to account for background noise.
4017 Can be specified in dB (in case "dB" is appended to the specified value)
4018 or amplitude ratio. Default value is @code{0}.
4021 Set the count for trimming silence from the end of audio.
4022 To remove silence from the middle of a file, specify a @var{stop_periods}
4023 that is negative. This value is then treated as a positive value and is
4024 used to indicate the effect should restart processing as specified by
4025 @var{start_periods}, making it suitable for removing periods of silence
4026 in the middle of the audio.
4027 Default value is @code{0}.
4030 Specify a duration of silence that must exist before audio is not copied any
4031 more. By specifying a higher duration, silence that is wanted can be left in
4033 Default value is @code{0}.
4035 @item stop_threshold
4036 This is the same as @option{start_threshold} but for trimming silence from
4038 Can be specified in dB (in case "dB" is appended to the specified value)
4039 or amplitude ratio. Default value is @code{0}.
4042 This indicates that @var{stop_duration} length of audio should be left intact
4043 at the beginning of each period of silence.
4044 For example, if you want to remove long pauses between words but do not want
4045 to remove the pauses completely. Default value is @code{0}.
4048 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4049 and works better with digital silence which is exactly 0.
4050 Default value is @code{rms}.
4053 Set ratio used to calculate size of window for detecting silence.
4054 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4057 @subsection Examples
4061 The following example shows how this filter can be used to start a recording
4062 that does not contain the delay at the start which usually occurs between
4063 pressing the record button and the start of the performance:
4065 silenceremove=1:5:0.02
4069 Trim all silence encountered from beginning to end where there is more than 1
4070 second of silence in audio:
4072 silenceremove=0:0:0:-1:1:-90dB
4078 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4079 loudspeakers around the user for binaural listening via headphones (audio
4080 formats up to 9 channels supported).
4081 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4082 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4083 Austrian Academy of Sciences.
4085 To enable compilation of this filter you need to configure FFmpeg with
4086 @code{--enable-libmysofa}.
4088 The filter accepts the following options:
4092 Set the SOFA file used for rendering.
4095 Set gain applied to audio. Value is in dB. Default is 0.
4098 Set rotation of virtual loudspeakers in deg. Default is 0.
4101 Set elevation of virtual speakers in deg. Default is 0.
4104 Set distance in meters between loudspeakers and the listener with near-field
4105 HRTFs. Default is 1.
4108 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4109 processing audio in time domain which is slow.
4110 @var{freq} is processing audio in frequency domain which is fast.
4111 Default is @var{freq}.
4114 Set custom positions of virtual loudspeakers. Syntax for this option is:
4115 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4116 Each virtual loudspeaker is described with short channel name following with
4117 azimuth and elevation in degrees.
4118 Each virtual loudspeaker description is separated by '|'.
4119 For example to override front left and front right channel positions use:
4120 'speakers=FL 45 15|FR 345 15'.
4121 Descriptions with unrecognised channel names are ignored.
4124 Set custom gain for LFE channels. Value is in dB. Default is 0.
4127 @subsection Examples
4131 Using ClubFritz6 sofa file:
4133 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4137 Using ClubFritz12 sofa file and bigger radius with small rotation:
4139 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4143 Similar as above but with custom speaker positions for front left, front right, back left and back right
4144 and also with custom gain:
4146 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4150 @section stereotools
4152 This filter has some handy utilities to manage stereo signals, for converting
4153 M/S stereo recordings to L/R signal while having control over the parameters
4154 or spreading the stereo image of master track.
4156 The filter accepts the following options:
4160 Set input level before filtering for both channels. Defaults is 1.
4161 Allowed range is from 0.015625 to 64.
4164 Set output level after filtering for both channels. Defaults is 1.
4165 Allowed range is from 0.015625 to 64.
4168 Set input balance between both channels. Default is 0.
4169 Allowed range is from -1 to 1.
4172 Set output balance between both channels. Default is 0.
4173 Allowed range is from -1 to 1.
4176 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4177 clipping. Disabled by default.
4180 Mute the left channel. Disabled by default.
4183 Mute the right channel. Disabled by default.
4186 Change the phase of the left channel. Disabled by default.
4189 Change the phase of the right channel. Disabled by default.
4192 Set stereo mode. Available values are:
4196 Left/Right to Left/Right, this is default.
4199 Left/Right to Mid/Side.
4202 Mid/Side to Left/Right.
4205 Left/Right to Left/Left.
4208 Left/Right to Right/Right.
4211 Left/Right to Left + Right.
4214 Left/Right to Right/Left.
4217 Mid/Side to Left/Left.
4220 Mid/Side to Right/Right.
4224 Set level of side signal. Default is 1.
4225 Allowed range is from 0.015625 to 64.
4228 Set balance of side signal. Default is 0.
4229 Allowed range is from -1 to 1.
4232 Set level of the middle signal. Default is 1.
4233 Allowed range is from 0.015625 to 64.
4236 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4239 Set stereo base between mono and inversed channels. Default is 0.
4240 Allowed range is from -1 to 1.
4243 Set delay in milliseconds how much to delay left from right channel and
4244 vice versa. Default is 0. Allowed range is from -20 to 20.
4247 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4250 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4252 @item bmode_in, bmode_out
4253 Set balance mode for balance_in/balance_out option.
4255 Can be one of the following:
4259 Classic balance mode. Attenuate one channel at time.
4260 Gain is raised up to 1.
4263 Similar as classic mode above but gain is raised up to 2.
4266 Equal power distribution, from -6dB to +6dB range.
4270 @subsection Examples
4274 Apply karaoke like effect:
4276 stereotools=mlev=0.015625
4280 Convert M/S signal to L/R:
4282 "stereotools=mode=ms>lr"
4286 @section stereowiden
4288 This filter enhance the stereo effect by suppressing signal common to both
4289 channels and by delaying the signal of left into right and vice versa,
4290 thereby widening the stereo effect.
4292 The filter accepts the following options:
4296 Time in milliseconds of the delay of left signal into right and vice versa.
4297 Default is 20 milliseconds.
4300 Amount of gain in delayed signal into right and vice versa. Gives a delay
4301 effect of left signal in right output and vice versa which gives widening
4302 effect. Default is 0.3.
4305 Cross feed of left into right with inverted phase. This helps in suppressing
4306 the mono. If the value is 1 it will cancel all the signal common to both
4307 channels. Default is 0.3.
4310 Set level of input signal of original channel. Default is 0.8.
4313 @section superequalizer
4314 Apply 18 band equalizer.
4316 The filter accepts the following options:
4323 Set 131Hz band gain.
4325 Set 185Hz band gain.
4327 Set 262Hz band gain.
4329 Set 370Hz band gain.
4331 Set 523Hz band gain.
4333 Set 740Hz band gain.
4335 Set 1047Hz band gain.
4337 Set 1480Hz band gain.
4339 Set 2093Hz band gain.
4341 Set 2960Hz band gain.
4343 Set 4186Hz band gain.
4345 Set 5920Hz band gain.
4347 Set 8372Hz band gain.
4349 Set 11840Hz band gain.
4351 Set 16744Hz band gain.
4353 Set 20000Hz band gain.
4357 Apply audio surround upmix filter.
4359 This filter allows to produce multichannel output from audio stream.
4361 The filter accepts the following options:
4365 Set output channel layout. By default, this is @var{5.1}.
4367 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4368 for the required syntax.
4371 Set input channel layout. By default, this is @var{stereo}.
4373 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4374 for the required syntax.
4377 Set input volume level. By default, this is @var{1}.
4380 Set output volume level. By default, this is @var{1}.
4383 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4386 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4389 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4392 Set front center input volume. By default, this is @var{1}.
4395 Set front center output volume. By default, this is @var{1}.
4398 Set LFE input volume. By default, this is @var{1}.
4401 Set LFE output volume. By default, this is @var{1}.
4406 Boost or cut treble (upper) frequencies of the audio using a two-pole
4407 shelving filter with a response similar to that of a standard
4408 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4410 The filter accepts the following options:
4414 Give the gain at whichever is the lower of ~22 kHz and the
4415 Nyquist frequency. Its useful range is about -20 (for a large cut)
4416 to +20 (for a large boost). Beware of clipping when using a positive gain.
4419 Set the filter's central frequency and so can be used
4420 to extend or reduce the frequency range to be boosted or cut.
4421 The default value is @code{3000} Hz.
4424 Set method to specify band-width of filter.
4439 Determine how steep is the filter's shelf transition.
4442 Specify which channels to filter, by default all available are filtered.
4445 @subsection Commands
4447 This filter supports the following commands:
4450 Change treble frequency.
4451 Syntax for the command is : "@var{frequency}"
4454 Change treble width_type.
4455 Syntax for the command is : "@var{width_type}"
4458 Change treble width.
4459 Syntax for the command is : "@var{width}"
4463 Syntax for the command is : "@var{gain}"
4468 Sinusoidal amplitude modulation.
4470 The filter accepts the following options:
4474 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4475 (20 Hz or lower) will result in a tremolo effect.
4476 This filter may also be used as a ring modulator by specifying
4477 a modulation frequency higher than 20 Hz.
4478 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4481 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4482 Default value is 0.5.
4487 Sinusoidal phase modulation.
4489 The filter accepts the following options:
4493 Modulation frequency in Hertz.
4494 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4497 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4498 Default value is 0.5.
4503 Adjust the input audio volume.
4505 It accepts the following parameters:
4509 Set audio volume expression.
4511 Output values are clipped to the maximum value.
4513 The output audio volume is given by the relation:
4515 @var{output_volume} = @var{volume} * @var{input_volume}
4518 The default value for @var{volume} is "1.0".
4521 This parameter represents the mathematical precision.
4523 It determines which input sample formats will be allowed, which affects the
4524 precision of the volume scaling.
4528 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4530 32-bit floating-point; this limits input sample format to FLT. (default)
4532 64-bit floating-point; this limits input sample format to DBL.
4536 Choose the behaviour on encountering ReplayGain side data in input frames.
4540 Remove ReplayGain side data, ignoring its contents (the default).
4543 Ignore ReplayGain side data, but leave it in the frame.
4546 Prefer the track gain, if present.
4549 Prefer the album gain, if present.
4552 @item replaygain_preamp
4553 Pre-amplification gain in dB to apply to the selected replaygain gain.
4555 Default value for @var{replaygain_preamp} is 0.0.
4558 Set when the volume expression is evaluated.
4560 It accepts the following values:
4563 only evaluate expression once during the filter initialization, or
4564 when the @samp{volume} command is sent
4567 evaluate expression for each incoming frame
4570 Default value is @samp{once}.
4573 The volume expression can contain the following parameters.
4577 frame number (starting at zero)
4580 @item nb_consumed_samples
4581 number of samples consumed by the filter
4583 number of samples in the current frame
4585 original frame position in the file
4591 PTS at start of stream
4593 time at start of stream
4599 last set volume value
4602 Note that when @option{eval} is set to @samp{once} only the
4603 @var{sample_rate} and @var{tb} variables are available, all other
4604 variables will evaluate to NAN.
4606 @subsection Commands
4608 This filter supports the following commands:
4611 Modify the volume expression.
4612 The command accepts the same syntax of the corresponding option.
4614 If the specified expression is not valid, it is kept at its current
4616 @item replaygain_noclip
4617 Prevent clipping by limiting the gain applied.
4619 Default value for @var{replaygain_noclip} is 1.
4623 @subsection Examples
4627 Halve the input audio volume:
4631 volume=volume=-6.0206dB
4634 In all the above example the named key for @option{volume} can be
4635 omitted, for example like in:
4641 Increase input audio power by 6 decibels using fixed-point precision:
4643 volume=volume=6dB:precision=fixed
4647 Fade volume after time 10 with an annihilation period of 5 seconds:
4649 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4653 @section volumedetect
4655 Detect the volume of the input video.
4657 The filter has no parameters. The input is not modified. Statistics about
4658 the volume will be printed in the log when the input stream end is reached.
4660 In particular it will show the mean volume (root mean square), maximum
4661 volume (on a per-sample basis), and the beginning of a histogram of the
4662 registered volume values (from the maximum value to a cumulated 1/1000 of
4665 All volumes are in decibels relative to the maximum PCM value.
4667 @subsection Examples
4669 Here is an excerpt of the output:
4671 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4672 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4673 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4674 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4675 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4676 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4677 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4678 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4679 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4685 The mean square energy is approximately -27 dB, or 10^-2.7.
4687 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4689 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4692 In other words, raising the volume by +4 dB does not cause any clipping,
4693 raising it by +5 dB causes clipping for 6 samples, etc.
4695 @c man end AUDIO FILTERS
4697 @chapter Audio Sources
4698 @c man begin AUDIO SOURCES
4700 Below is a description of the currently available audio sources.
4704 Buffer audio frames, and make them available to the filter chain.
4706 This source is mainly intended for a programmatic use, in particular
4707 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4709 It accepts the following parameters:
4713 The timebase which will be used for timestamps of submitted frames. It must be
4714 either a floating-point number or in @var{numerator}/@var{denominator} form.
4717 The sample rate of the incoming audio buffers.
4720 The sample format of the incoming audio buffers.
4721 Either a sample format name or its corresponding integer representation from
4722 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4724 @item channel_layout
4725 The channel layout of the incoming audio buffers.
4726 Either a channel layout name from channel_layout_map in
4727 @file{libavutil/channel_layout.c} or its corresponding integer representation
4728 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4731 The number of channels of the incoming audio buffers.
4732 If both @var{channels} and @var{channel_layout} are specified, then they
4737 @subsection Examples
4740 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4743 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4744 Since the sample format with name "s16p" corresponds to the number
4745 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4748 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4753 Generate an audio signal specified by an expression.
4755 This source accepts in input one or more expressions (one for each
4756 channel), which are evaluated and used to generate a corresponding
4759 This source accepts the following options:
4763 Set the '|'-separated expressions list for each separate channel. In case the
4764 @option{channel_layout} option is not specified, the selected channel layout
4765 depends on the number of provided expressions. Otherwise the last
4766 specified expression is applied to the remaining output channels.
4768 @item channel_layout, c
4769 Set the channel layout. The number of channels in the specified layout
4770 must be equal to the number of specified expressions.
4773 Set the minimum duration of the sourced audio. See
4774 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4775 for the accepted syntax.
4776 Note that the resulting duration may be greater than the specified
4777 duration, as the generated audio is always cut at the end of a
4780 If not specified, or the expressed duration is negative, the audio is
4781 supposed to be generated forever.
4784 Set the number of samples per channel per each output frame,
4787 @item sample_rate, s
4788 Specify the sample rate, default to 44100.
4791 Each expression in @var{exprs} can contain the following constants:
4795 number of the evaluated sample, starting from 0
4798 time of the evaluated sample expressed in seconds, starting from 0
4805 @subsection Examples
4815 Generate a sin signal with frequency of 440 Hz, set sample rate to
4818 aevalsrc="sin(440*2*PI*t):s=8000"
4822 Generate a two channels signal, specify the channel layout (Front
4823 Center + Back Center) explicitly:
4825 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4829 Generate white noise:
4831 aevalsrc="-2+random(0)"
4835 Generate an amplitude modulated signal:
4837 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4841 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4843 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4850 The null audio source, return unprocessed audio frames. It is mainly useful
4851 as a template and to be employed in analysis / debugging tools, or as
4852 the source for filters which ignore the input data (for example the sox
4855 This source accepts the following options:
4859 @item channel_layout, cl
4861 Specifies the channel layout, and can be either an integer or a string
4862 representing a channel layout. The default value of @var{channel_layout}
4865 Check the channel_layout_map definition in
4866 @file{libavutil/channel_layout.c} for the mapping between strings and
4867 channel layout values.
4869 @item sample_rate, r
4870 Specifies the sample rate, and defaults to 44100.
4873 Set the number of samples per requested frames.
4877 @subsection Examples
4881 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4883 anullsrc=r=48000:cl=4
4887 Do the same operation with a more obvious syntax:
4889 anullsrc=r=48000:cl=mono
4893 All the parameters need to be explicitly defined.
4897 Synthesize a voice utterance using the libflite library.
4899 To enable compilation of this filter you need to configure FFmpeg with
4900 @code{--enable-libflite}.
4902 Note that versions of the flite library prior to 2.0 are not thread-safe.
4904 The filter accepts the following options:
4909 If set to 1, list the names of the available voices and exit
4910 immediately. Default value is 0.
4913 Set the maximum number of samples per frame. Default value is 512.
4916 Set the filename containing the text to speak.
4919 Set the text to speak.
4922 Set the voice to use for the speech synthesis. Default value is
4923 @code{kal}. See also the @var{list_voices} option.
4926 @subsection Examples
4930 Read from file @file{speech.txt}, and synthesize the text using the
4931 standard flite voice:
4933 flite=textfile=speech.txt
4937 Read the specified text selecting the @code{slt} voice:
4939 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4943 Input text to ffmpeg:
4945 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4949 Make @file{ffplay} speak the specified text, using @code{flite} and
4950 the @code{lavfi} device:
4952 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4956 For more information about libflite, check:
4957 @url{http://www.festvox.org/flite/}
4961 Generate a noise audio signal.
4963 The filter accepts the following options:
4966 @item sample_rate, r
4967 Specify the sample rate. Default value is 48000 Hz.
4970 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4974 Specify the duration of the generated audio stream. Not specifying this option
4975 results in noise with an infinite length.
4977 @item color, colour, c
4978 Specify the color of noise. Available noise colors are white, pink, brown,
4979 blue and violet. Default color is white.
4982 Specify a value used to seed the PRNG.
4985 Set the number of samples per each output frame, default is 1024.
4988 @subsection Examples
4993 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4995 anoisesrc=d=60:c=pink:r=44100:a=0.5
5001 Generate odd-tap Hilbert transform FIR coefficients.
5003 The resulting stream can be used with @ref{afir} filter for phase-shifting
5004 the signal by 90 degrees.
5006 This is used in many matrix coding schemes and for analytic signal generation.
5007 The process is often written as a multiplication by i (or j), the imaginary unit.
5009 The filter accepts the following options:
5013 @item sample_rate, s
5014 Set sample rate, default is 44100.
5017 Set length of FIR filter, default is 22051.
5020 Set number of samples per each frame.
5023 Set window function to be used when generating FIR coefficients.
5028 Generate an audio signal made of a sine wave with amplitude 1/8.
5030 The audio signal is bit-exact.
5032 The filter accepts the following options:
5037 Set the carrier frequency. Default is 440 Hz.
5039 @item beep_factor, b
5040 Enable a periodic beep every second with frequency @var{beep_factor} times
5041 the carrier frequency. Default is 0, meaning the beep is disabled.
5043 @item sample_rate, r
5044 Specify the sample rate, default is 44100.
5047 Specify the duration of the generated audio stream.
5049 @item samples_per_frame
5050 Set the number of samples per output frame.
5052 The expression can contain the following constants:
5056 The (sequential) number of the output audio frame, starting from 0.
5059 The PTS (Presentation TimeStamp) of the output audio frame,
5060 expressed in @var{TB} units.
5063 The PTS of the output audio frame, expressed in seconds.
5066 The timebase of the output audio frames.
5069 Default is @code{1024}.
5072 @subsection Examples
5077 Generate a simple 440 Hz sine wave:
5083 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5087 sine=frequency=220:beep_factor=4:duration=5
5091 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5094 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5098 @c man end AUDIO SOURCES
5100 @chapter Audio Sinks
5101 @c man begin AUDIO SINKS
5103 Below is a description of the currently available audio sinks.
5105 @section abuffersink
5107 Buffer audio frames, and make them available to the end of filter chain.
5109 This sink is mainly intended for programmatic use, in particular
5110 through the interface defined in @file{libavfilter/buffersink.h}
5111 or the options system.
5113 It accepts a pointer to an AVABufferSinkContext structure, which
5114 defines the incoming buffers' formats, to be passed as the opaque
5115 parameter to @code{avfilter_init_filter} for initialization.
5118 Null audio sink; do absolutely nothing with the input audio. It is
5119 mainly useful as a template and for use in analysis / debugging
5122 @c man end AUDIO SINKS
5124 @chapter Video Filters
5125 @c man begin VIDEO FILTERS
5127 When you configure your FFmpeg build, you can disable any of the
5128 existing filters using @code{--disable-filters}.
5129 The configure output will show the video filters included in your
5132 Below is a description of the currently available video filters.
5134 @section alphaextract
5136 Extract the alpha component from the input as a grayscale video. This
5137 is especially useful with the @var{alphamerge} filter.
5141 Add or replace the alpha component of the primary input with the
5142 grayscale value of a second input. This is intended for use with
5143 @var{alphaextract} to allow the transmission or storage of frame
5144 sequences that have alpha in a format that doesn't support an alpha
5147 For example, to reconstruct full frames from a normal YUV-encoded video
5148 and a separate video created with @var{alphaextract}, you might use:
5150 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5153 Since this filter is designed for reconstruction, it operates on frame
5154 sequences without considering timestamps, and terminates when either
5155 input reaches end of stream. This will cause problems if your encoding
5156 pipeline drops frames. If you're trying to apply an image as an
5157 overlay to a video stream, consider the @var{overlay} filter instead.
5161 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5162 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5163 Substation Alpha) subtitles files.
5165 This filter accepts the following option in addition to the common options from
5166 the @ref{subtitles} filter:
5170 Set the shaping engine
5172 Available values are:
5175 The default libass shaping engine, which is the best available.
5177 Fast, font-agnostic shaper that can do only substitutions
5179 Slower shaper using OpenType for substitutions and positioning
5182 The default is @code{auto}.
5186 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5188 The filter accepts the following options:
5192 Set threshold A for 1st plane. Default is 0.02.
5193 Valid range is 0 to 0.3.
5196 Set threshold B for 1st plane. Default is 0.04.
5197 Valid range is 0 to 5.
5200 Set threshold A for 2nd plane. Default is 0.02.
5201 Valid range is 0 to 0.3.
5204 Set threshold B for 2nd plane. Default is 0.04.
5205 Valid range is 0 to 5.
5208 Set threshold A for 3rd plane. Default is 0.02.
5209 Valid range is 0 to 0.3.
5212 Set threshold B for 3rd plane. Default is 0.04.
5213 Valid range is 0 to 5.
5215 Threshold A is designed to react on abrupt changes in the input signal and
5216 threshold B is designed to react on continuous changes in the input signal.
5219 Set number of frames filter will use for averaging. Default is 33. Must be odd
5220 number in range [5, 129].
5223 Set what planes of frame filter will use for averaging. Default is all.
5228 Apply average blur filter.
5230 The filter accepts the following options:
5234 Set horizontal kernel size.
5237 Set which planes to filter. By default all planes are filtered.
5240 Set vertical kernel size, if zero it will be same as @code{sizeX}.
5241 Default is @code{0}.
5246 Compute the bounding box for the non-black pixels in the input frame
5249 This filter computes the bounding box containing all the pixels with a
5250 luminance value greater than the minimum allowed value.
5251 The parameters describing the bounding box are printed on the filter
5254 The filter accepts the following option:
5258 Set the minimal luminance value. Default is @code{16}.
5261 @section bitplanenoise
5263 Show and measure bit plane noise.
5265 The filter accepts the following options:
5269 Set which plane to analyze. Default is @code{1}.
5272 Filter out noisy pixels from @code{bitplane} set above.
5273 Default is disabled.
5276 @section blackdetect
5278 Detect video intervals that are (almost) completely black. Can be
5279 useful to detect chapter transitions, commercials, or invalid
5280 recordings. Output lines contains the time for the start, end and
5281 duration of the detected black interval expressed in seconds.
5283 In order to display the output lines, you need to set the loglevel at
5284 least to the AV_LOG_INFO value.
5286 The filter accepts the following options:
5289 @item black_min_duration, d
5290 Set the minimum detected black duration expressed in seconds. It must
5291 be a non-negative floating point number.
5293 Default value is 2.0.
5295 @item picture_black_ratio_th, pic_th
5296 Set the threshold for considering a picture "black".
5297 Express the minimum value for the ratio:
5299 @var{nb_black_pixels} / @var{nb_pixels}
5302 for which a picture is considered black.
5303 Default value is 0.98.
5305 @item pixel_black_th, pix_th
5306 Set the threshold for considering a pixel "black".
5308 The threshold expresses the maximum pixel luminance value for which a
5309 pixel is considered "black". The provided value is scaled according to
5310 the following equation:
5312 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5315 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5316 the input video format, the range is [0-255] for YUV full-range
5317 formats and [16-235] for YUV non full-range formats.
5319 Default value is 0.10.
5322 The following example sets the maximum pixel threshold to the minimum
5323 value, and detects only black intervals of 2 or more seconds:
5325 blackdetect=d=2:pix_th=0.00
5330 Detect frames that are (almost) completely black. Can be useful to
5331 detect chapter transitions or commercials. Output lines consist of
5332 the frame number of the detected frame, the percentage of blackness,
5333 the position in the file if known or -1 and the timestamp in seconds.
5335 In order to display the output lines, you need to set the loglevel at
5336 least to the AV_LOG_INFO value.
5338 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5339 The value represents the percentage of pixels in the picture that
5340 are below the threshold value.
5342 It accepts the following parameters:
5347 The percentage of the pixels that have to be below the threshold; it defaults to
5350 @item threshold, thresh
5351 The threshold below which a pixel value is considered black; it defaults to
5356 @section blend, tblend
5358 Blend two video frames into each other.
5360 The @code{blend} filter takes two input streams and outputs one
5361 stream, the first input is the "top" layer and second input is
5362 "bottom" layer. By default, the output terminates when the longest input terminates.
5364 The @code{tblend} (time blend) filter takes two consecutive frames
5365 from one single stream, and outputs the result obtained by blending
5366 the new frame on top of the old frame.
5368 A description of the accepted options follows.
5376 Set blend mode for specific pixel component or all pixel components in case
5377 of @var{all_mode}. Default value is @code{normal}.
5379 Available values for component modes are:
5421 Set blend opacity for specific pixel component or all pixel components in case
5422 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5429 Set blend expression for specific pixel component or all pixel components in case
5430 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5432 The expressions can use the following variables:
5436 The sequential number of the filtered frame, starting from @code{0}.
5440 the coordinates of the current sample
5444 the width and height of currently filtered plane
5448 Width and height scale depending on the currently filtered plane. It is the
5449 ratio between the corresponding luma plane number of pixels and the current
5450 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5451 @code{0.5,0.5} for chroma planes.
5454 Time of the current frame, expressed in seconds.
5457 Value of pixel component at current location for first video frame (top layer).
5460 Value of pixel component at current location for second video frame (bottom layer).
5464 The @code{blend} filter also supports the @ref{framesync} options.
5466 @subsection Examples
5470 Apply transition from bottom layer to top layer in first 10 seconds:
5472 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5476 Apply linear horizontal transition from top layer to bottom layer:
5478 blend=all_expr='A*(X/W)+B*(1-X/W)'
5482 Apply 1x1 checkerboard effect:
5484 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5488 Apply uncover left effect:
5490 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5494 Apply uncover down effect:
5496 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5500 Apply uncover up-left effect:
5502 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5506 Split diagonally video and shows top and bottom layer on each side:
5508 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5512 Display differences between the current and the previous frame:
5514 tblend=all_mode=grainextract
5520 Apply a boxblur algorithm to the input video.
5522 It accepts the following parameters:
5526 @item luma_radius, lr
5527 @item luma_power, lp
5528 @item chroma_radius, cr
5529 @item chroma_power, cp
5530 @item alpha_radius, ar
5531 @item alpha_power, ap
5535 A description of the accepted options follows.
5538 @item luma_radius, lr
5539 @item chroma_radius, cr
5540 @item alpha_radius, ar
5541 Set an expression for the box radius in pixels used for blurring the
5542 corresponding input plane.
5544 The radius value must be a non-negative number, and must not be
5545 greater than the value of the expression @code{min(w,h)/2} for the
5546 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5549 Default value for @option{luma_radius} is "2". If not specified,
5550 @option{chroma_radius} and @option{alpha_radius} default to the
5551 corresponding value set for @option{luma_radius}.
5553 The expressions can contain the following constants:
5557 The input width and height in pixels.
5561 The input chroma image width and height in pixels.
5565 The horizontal and vertical chroma subsample values. For example, for the
5566 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5569 @item luma_power, lp
5570 @item chroma_power, cp
5571 @item alpha_power, ap
5572 Specify how many times the boxblur filter is applied to the
5573 corresponding plane.
5575 Default value for @option{luma_power} is 2. If not specified,
5576 @option{chroma_power} and @option{alpha_power} default to the
5577 corresponding value set for @option{luma_power}.
5579 A value of 0 will disable the effect.
5582 @subsection Examples
5586 Apply a boxblur filter with the luma, chroma, and alpha radii
5589 boxblur=luma_radius=2:luma_power=1
5594 Set the luma radius to 2, and alpha and chroma radius to 0:
5596 boxblur=2:1:cr=0:ar=0
5600 Set the luma and chroma radii to a fraction of the video dimension:
5602 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5608 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5609 Deinterlacing Filter").
5611 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5612 interpolation algorithms.
5613 It accepts the following parameters:
5617 The interlacing mode to adopt. It accepts one of the following values:
5621 Output one frame for each frame.
5623 Output one frame for each field.
5626 The default value is @code{send_field}.
5629 The picture field parity assumed for the input interlaced video. It accepts one
5630 of the following values:
5634 Assume the top field is first.
5636 Assume the bottom field is first.
5638 Enable automatic detection of field parity.
5641 The default value is @code{auto}.
5642 If the interlacing is unknown or the decoder does not export this information,
5643 top field first will be assumed.
5646 Specify which frames to deinterlace. Accept one of the following
5651 Deinterlace all frames.
5653 Only deinterlace frames marked as interlaced.
5656 The default value is @code{all}.
5660 YUV colorspace color/chroma keying.
5662 The filter accepts the following options:
5666 The color which will be replaced with transparency.
5669 Similarity percentage with the key color.
5671 0.01 matches only the exact key color, while 1.0 matches everything.
5676 0.0 makes pixels either fully transparent, or not transparent at all.
5678 Higher values result in semi-transparent pixels, with a higher transparency
5679 the more similar the pixels color is to the key color.
5682 Signals that the color passed is already in YUV instead of RGB.
5684 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5685 This can be used to pass exact YUV values as hexadecimal numbers.
5688 @subsection Examples
5692 Make every green pixel in the input image transparent:
5694 ffmpeg -i input.png -vf chromakey=green out.png
5698 Overlay a greenscreen-video on top of a static black background.
5700 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
5706 Display CIE color diagram with pixels overlaid onto it.
5708 The filter accepts the following options:
5723 @item uhdtv, rec2020
5736 Set what gamuts to draw.
5738 See @code{system} option for available values.
5741 Set ciescope size, by default set to 512.
5744 Set intensity used to map input pixel values to CIE diagram.
5747 Set contrast used to draw tongue colors that are out of active color system gamut.
5750 Correct gamma displayed on scope, by default enabled.
5753 Show white point on CIE diagram, by default disabled.
5756 Set input gamma. Used only with XYZ input color space.
5761 Visualize information exported by some codecs.
5763 Some codecs can export information through frames using side-data or other
5764 means. For example, some MPEG based codecs export motion vectors through the
5765 @var{export_mvs} flag in the codec @option{flags2} option.
5767 The filter accepts the following option:
5771 Set motion vectors to visualize.
5773 Available flags for @var{mv} are:
5777 forward predicted MVs of P-frames
5779 forward predicted MVs of B-frames
5781 backward predicted MVs of B-frames
5785 Display quantization parameters using the chroma planes.
5788 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5790 Available flags for @var{mv_type} are:
5794 forward predicted MVs
5796 backward predicted MVs
5799 @item frame_type, ft
5800 Set frame type to visualize motion vectors of.
5802 Available flags for @var{frame_type} are:
5806 intra-coded frames (I-frames)
5808 predicted frames (P-frames)
5810 bi-directionally predicted frames (B-frames)
5814 @subsection Examples
5818 Visualize forward predicted MVs of all frames using @command{ffplay}:
5820 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5824 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5826 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5830 @section colorbalance
5831 Modify intensity of primary colors (red, green and blue) of input frames.
5833 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5834 regions for the red-cyan, green-magenta or blue-yellow balance.
5836 A positive adjustment value shifts the balance towards the primary color, a negative
5837 value towards the complementary color.
5839 The filter accepts the following options:
5845 Adjust red, green and blue shadows (darkest pixels).
5850 Adjust red, green and blue midtones (medium pixels).
5855 Adjust red, green and blue highlights (brightest pixels).
5857 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5860 @subsection Examples
5864 Add red color cast to shadows:
5871 RGB colorspace color keying.
5873 The filter accepts the following options:
5877 The color which will be replaced with transparency.
5880 Similarity percentage with the key color.
5882 0.01 matches only the exact key color, while 1.0 matches everything.
5887 0.0 makes pixels either fully transparent, or not transparent at all.
5889 Higher values result in semi-transparent pixels, with a higher transparency
5890 the more similar the pixels color is to the key color.
5893 @subsection Examples
5897 Make every green pixel in the input image transparent:
5899 ffmpeg -i input.png -vf colorkey=green out.png
5903 Overlay a greenscreen-video on top of a static background image.
5905 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
5909 @section colorlevels
5911 Adjust video input frames using levels.
5913 The filter accepts the following options:
5920 Adjust red, green, blue and alpha input black point.
5921 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5927 Adjust red, green, blue and alpha input white point.
5928 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5930 Input levels are used to lighten highlights (bright tones), darken shadows
5931 (dark tones), change the balance of bright and dark tones.
5937 Adjust red, green, blue and alpha output black point.
5938 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5944 Adjust red, green, blue and alpha output white point.
5945 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5947 Output levels allows manual selection of a constrained output level range.
5950 @subsection Examples
5954 Make video output darker:
5956 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5962 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5966 Make video output lighter:
5968 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5972 Increase brightness:
5974 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5978 @section colorchannelmixer
5980 Adjust video input frames by re-mixing color channels.
5982 This filter modifies a color channel by adding the values associated to
5983 the other channels of the same pixels. For example if the value to
5984 modify is red, the output value will be:
5986 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5989 The filter accepts the following options:
5996 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5997 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6003 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6004 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6010 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6011 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6017 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6018 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6020 Allowed ranges for options are @code{[-2.0, 2.0]}.
6023 @subsection Examples
6027 Convert source to grayscale:
6029 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6032 Simulate sepia tones:
6034 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6038 @section colormatrix
6040 Convert color matrix.
6042 The filter accepts the following options:
6047 Specify the source and destination color matrix. Both values must be
6050 The accepted values are:
6078 For example to convert from BT.601 to SMPTE-240M, use the command:
6080 colormatrix=bt601:smpte240m
6085 Convert colorspace, transfer characteristics or color primaries.
6086 Input video needs to have an even size.
6088 The filter accepts the following options:
6093 Specify all color properties at once.
6095 The accepted values are:
6125 Specify output colorspace.
6127 The accepted values are:
6136 BT.470BG or BT.601-6 625
6139 SMPTE-170M or BT.601-6 525
6148 BT.2020 with non-constant luminance
6154 Specify output transfer characteristics.
6156 The accepted values are:
6168 Constant gamma of 2.2
6171 Constant gamma of 2.8
6174 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6192 BT.2020 for 10-bits content
6195 BT.2020 for 12-bits content
6201 Specify output color primaries.
6203 The accepted values are:
6212 BT.470BG or BT.601-6 625
6215 SMPTE-170M or BT.601-6 525
6239 Specify output color range.
6241 The accepted values are:
6244 TV (restricted) range
6247 MPEG (restricted) range
6258 Specify output color format.
6260 The accepted values are:
6263 YUV 4:2:0 planar 8-bits
6266 YUV 4:2:0 planar 10-bits
6269 YUV 4:2:0 planar 12-bits
6272 YUV 4:2:2 planar 8-bits
6275 YUV 4:2:2 planar 10-bits
6278 YUV 4:2:2 planar 12-bits
6281 YUV 4:4:4 planar 8-bits
6284 YUV 4:4:4 planar 10-bits
6287 YUV 4:4:4 planar 12-bits
6292 Do a fast conversion, which skips gamma/primary correction. This will take
6293 significantly less CPU, but will be mathematically incorrect. To get output
6294 compatible with that produced by the colormatrix filter, use fast=1.
6297 Specify dithering mode.
6299 The accepted values are:
6305 Floyd-Steinberg dithering
6309 Whitepoint adaptation mode.
6311 The accepted values are:
6314 Bradford whitepoint adaptation
6317 von Kries whitepoint adaptation
6320 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6324 Override all input properties at once. Same accepted values as @ref{all}.
6327 Override input colorspace. Same accepted values as @ref{space}.
6330 Override input color primaries. Same accepted values as @ref{primaries}.
6333 Override input transfer characteristics. Same accepted values as @ref{trc}.
6336 Override input color range. Same accepted values as @ref{range}.
6340 The filter converts the transfer characteristics, color space and color
6341 primaries to the specified user values. The output value, if not specified,
6342 is set to a default value based on the "all" property. If that property is
6343 also not specified, the filter will log an error. The output color range and
6344 format default to the same value as the input color range and format. The
6345 input transfer characteristics, color space, color primaries and color range
6346 should be set on the input data. If any of these are missing, the filter will
6347 log an error and no conversion will take place.
6349 For example to convert the input to SMPTE-240M, use the command:
6351 colorspace=smpte240m
6354 @section convolution
6356 Apply convolution 3x3, 5x5 or 7x7 filter.
6358 The filter accepts the following options:
6365 Set matrix for each plane.
6366 Matrix is sequence of 9, 25 or 49 signed integers.
6372 Set multiplier for calculated value for each plane.
6378 Set bias for each plane. This value is added to the result of the multiplication.
6379 Useful for making the overall image brighter or darker. Default is 0.0.
6382 @subsection Examples
6388 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"
6394 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"
6400 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"
6406 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"
6410 Apply laplacian edge detector which includes diagonals:
6412 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"
6418 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"
6424 Apply 2D convolution of video stream in frequency domain using second stream
6427 The filter accepts the following options:
6431 Set which planes to process.
6434 Set which impulse video frames will be processed, can be @var{first}
6435 or @var{all}. Default is @var{all}.
6438 The @code{convolve} filter also supports the @ref{framesync} options.
6442 Copy the input video source unchanged to the output. This is mainly useful for
6447 Video filtering on GPU using Apple's CoreImage API on OSX.
6449 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6450 processed by video hardware. However, software-based OpenGL implementations
6451 exist which means there is no guarantee for hardware processing. It depends on
6454 There are many filters and image generators provided by Apple that come with a
6455 large variety of options. The filter has to be referenced by its name along
6458 The coreimage filter accepts the following options:
6461 List all available filters and generators along with all their respective
6462 options as well as possible minimum and maximum values along with the default
6469 Specify all filters by their respective name and options.
6470 Use @var{list_filters} to determine all valid filter names and options.
6471 Numerical options are specified by a float value and are automatically clamped
6472 to their respective value range. Vector and color options have to be specified
6473 by a list of space separated float values. Character escaping has to be done.
6474 A special option name @code{default} is available to use default options for a
6477 It is required to specify either @code{default} or at least one of the filter options.
6478 All omitted options are used with their default values.
6479 The syntax of the filter string is as follows:
6481 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6485 Specify a rectangle where the output of the filter chain is copied into the
6486 input image. It is given by a list of space separated float values:
6488 output_rect=x\ y\ width\ height
6490 If not given, the output rectangle equals the dimensions of the input image.
6491 The output rectangle is automatically cropped at the borders of the input
6492 image. Negative values are valid for each component.
6494 output_rect=25\ 25\ 100\ 100
6498 Several filters can be chained for successive processing without GPU-HOST
6499 transfers allowing for fast processing of complex filter chains.
6500 Currently, only filters with zero (generators) or exactly one (filters) input
6501 image and one output image are supported. Also, transition filters are not yet
6504 Some filters generate output images with additional padding depending on the
6505 respective filter kernel. The padding is automatically removed to ensure the
6506 filter output has the same size as the input image.
6508 For image generators, the size of the output image is determined by the
6509 previous output image of the filter chain or the input image of the whole
6510 filterchain, respectively. The generators do not use the pixel information of
6511 this image to generate their output. However, the generated output is
6512 blended onto this image, resulting in partial or complete coverage of the
6515 The @ref{coreimagesrc} video source can be used for generating input images
6516 which are directly fed into the filter chain. By using it, providing input
6517 images by another video source or an input video is not required.
6519 @subsection Examples
6524 List all filters available:
6526 coreimage=list_filters=true
6530 Use the CIBoxBlur filter with default options to blur an image:
6532 coreimage=filter=CIBoxBlur@@default
6536 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6537 its center at 100x100 and a radius of 50 pixels:
6539 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6543 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6544 given as complete and escaped command-line for Apple's standard bash shell:
6546 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6552 Crop the input video to given dimensions.
6554 It accepts the following parameters:
6558 The width of the output video. It defaults to @code{iw}.
6559 This expression is evaluated only once during the filter
6560 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6563 The height of the output video. It defaults to @code{ih}.
6564 This expression is evaluated only once during the filter
6565 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6568 The horizontal position, in the input video, of the left edge of the output
6569 video. It defaults to @code{(in_w-out_w)/2}.
6570 This expression is evaluated per-frame.
6573 The vertical position, in the input video, of the top edge of the output video.
6574 It defaults to @code{(in_h-out_h)/2}.
6575 This expression is evaluated per-frame.
6578 If set to 1 will force the output display aspect ratio
6579 to be the same of the input, by changing the output sample aspect
6580 ratio. It defaults to 0.
6583 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6584 width/height/x/y as specified and will not be rounded to nearest smaller value.
6588 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6589 expressions containing the following constants:
6594 The computed values for @var{x} and @var{y}. They are evaluated for
6599 The input width and height.
6603 These are the same as @var{in_w} and @var{in_h}.
6607 The output (cropped) width and height.
6611 These are the same as @var{out_w} and @var{out_h}.
6614 same as @var{iw} / @var{ih}
6617 input sample aspect ratio
6620 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6624 horizontal and vertical chroma subsample values. For example for the
6625 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6628 The number of the input frame, starting from 0.
6631 the position in the file of the input frame, NAN if unknown
6634 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6638 The expression for @var{out_w} may depend on the value of @var{out_h},
6639 and the expression for @var{out_h} may depend on @var{out_w}, but they
6640 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6641 evaluated after @var{out_w} and @var{out_h}.
6643 The @var{x} and @var{y} parameters specify the expressions for the
6644 position of the top-left corner of the output (non-cropped) area. They
6645 are evaluated for each frame. If the evaluated value is not valid, it
6646 is approximated to the nearest valid value.
6648 The expression for @var{x} may depend on @var{y}, and the expression
6649 for @var{y} may depend on @var{x}.
6651 @subsection Examples
6655 Crop area with size 100x100 at position (12,34).
6660 Using named options, the example above becomes:
6662 crop=w=100:h=100:x=12:y=34
6666 Crop the central input area with size 100x100:
6672 Crop the central input area with size 2/3 of the input video:
6674 crop=2/3*in_w:2/3*in_h
6678 Crop the input video central square:
6685 Delimit the rectangle with the top-left corner placed at position
6686 100:100 and the right-bottom corner corresponding to the right-bottom
6687 corner of the input image.
6689 crop=in_w-100:in_h-100:100:100
6693 Crop 10 pixels from the left and right borders, and 20 pixels from
6694 the top and bottom borders
6696 crop=in_w-2*10:in_h-2*20
6700 Keep only the bottom right quarter of the input image:
6702 crop=in_w/2:in_h/2:in_w/2:in_h/2
6706 Crop height for getting Greek harmony:
6708 crop=in_w:1/PHI*in_w
6712 Apply trembling effect:
6714 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)
6718 Apply erratic camera effect depending on timestamp:
6720 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)"
6724 Set x depending on the value of y:
6726 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6730 @subsection Commands
6732 This filter supports the following commands:
6738 Set width/height of the output video and the horizontal/vertical position
6740 The command accepts the same syntax of the corresponding option.
6742 If the specified expression is not valid, it is kept at its current
6748 Auto-detect the crop size.
6750 It calculates the necessary cropping parameters and prints the
6751 recommended parameters via the logging system. The detected dimensions
6752 correspond to the non-black area of the input video.
6754 It accepts the following parameters:
6759 Set higher black value threshold, which can be optionally specified
6760 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6761 value greater to the set value is considered non-black. It defaults to 24.
6762 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6763 on the bitdepth of the pixel format.
6766 The value which the width/height should be divisible by. It defaults to
6767 16. The offset is automatically adjusted to center the video. Use 2 to
6768 get only even dimensions (needed for 4:2:2 video). 16 is best when
6769 encoding to most video codecs.
6771 @item reset_count, reset
6772 Set the counter that determines after how many frames cropdetect will
6773 reset the previously detected largest video area and start over to
6774 detect the current optimal crop area. Default value is 0.
6776 This can be useful when channel logos distort the video area. 0
6777 indicates 'never reset', and returns the largest area encountered during
6784 Apply color adjustments using curves.
6786 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6787 component (red, green and blue) has its values defined by @var{N} key points
6788 tied from each other using a smooth curve. The x-axis represents the pixel
6789 values from the input frame, and the y-axis the new pixel values to be set for
6792 By default, a component curve is defined by the two points @var{(0;0)} and
6793 @var{(1;1)}. This creates a straight line where each original pixel value is
6794 "adjusted" to its own value, which means no change to the image.
6796 The filter allows you to redefine these two points and add some more. A new
6797 curve (using a natural cubic spline interpolation) will be define to pass
6798 smoothly through all these new coordinates. The new defined points needs to be
6799 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6800 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6801 the vector spaces, the values will be clipped accordingly.
6803 The filter accepts the following options:
6807 Select one of the available color presets. This option can be used in addition
6808 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6809 options takes priority on the preset values.
6810 Available presets are:
6813 @item color_negative
6816 @item increase_contrast
6818 @item linear_contrast
6819 @item medium_contrast
6821 @item strong_contrast
6824 Default is @code{none}.
6826 Set the master key points. These points will define a second pass mapping. It
6827 is sometimes called a "luminance" or "value" mapping. It can be used with
6828 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6829 post-processing LUT.
6831 Set the key points for the red component.
6833 Set the key points for the green component.
6835 Set the key points for the blue component.
6837 Set the key points for all components (not including master).
6838 Can be used in addition to the other key points component
6839 options. In this case, the unset component(s) will fallback on this
6840 @option{all} setting.
6842 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6844 Save Gnuplot script of the curves in specified file.
6847 To avoid some filtergraph syntax conflicts, each key points list need to be
6848 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6850 @subsection Examples
6854 Increase slightly the middle level of blue:
6856 curves=blue='0/0 0.5/0.58 1/1'
6862 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'
6864 Here we obtain the following coordinates for each components:
6867 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6869 @code{(0;0) (0.50;0.48) (1;1)}
6871 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6875 The previous example can also be achieved with the associated built-in preset:
6877 curves=preset=vintage
6887 Use a Photoshop preset and redefine the points of the green component:
6889 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6893 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6894 and @command{gnuplot}:
6896 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6897 gnuplot -p /tmp/curves.plt
6903 Video data analysis filter.
6905 This filter shows hexadecimal pixel values of part of video.
6907 The filter accepts the following options:
6911 Set output video size.
6914 Set x offset from where to pick pixels.
6917 Set y offset from where to pick pixels.
6920 Set scope mode, can be one of the following:
6923 Draw hexadecimal pixel values with white color on black background.
6926 Draw hexadecimal pixel values with input video pixel color on black
6930 Draw hexadecimal pixel values on color background picked from input video,
6931 the text color is picked in such way so its always visible.
6935 Draw rows and columns numbers on left and top of video.
6938 Set background opacity.
6943 Denoise frames using 2D DCT (frequency domain filtering).
6945 This filter is not designed for real time.
6947 The filter accepts the following options:
6951 Set the noise sigma constant.
6953 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6954 coefficient (absolute value) below this threshold with be dropped.
6956 If you need a more advanced filtering, see @option{expr}.
6958 Default is @code{0}.
6961 Set number overlapping pixels for each block. Since the filter can be slow, you
6962 may want to reduce this value, at the cost of a less effective filter and the
6963 risk of various artefacts.
6965 If the overlapping value doesn't permit processing the whole input width or
6966 height, a warning will be displayed and according borders won't be denoised.
6968 Default value is @var{blocksize}-1, which is the best possible setting.
6971 Set the coefficient factor expression.
6973 For each coefficient of a DCT block, this expression will be evaluated as a
6974 multiplier value for the coefficient.
6976 If this is option is set, the @option{sigma} option will be ignored.
6978 The absolute value of the coefficient can be accessed through the @var{c}
6982 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6983 @var{blocksize}, which is the width and height of the processed blocks.
6985 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6986 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6987 on the speed processing. Also, a larger block size does not necessarily means a
6991 @subsection Examples
6993 Apply a denoise with a @option{sigma} of @code{4.5}:
6998 The same operation can be achieved using the expression system:
7000 dctdnoiz=e='gte(c, 4.5*3)'
7003 Violent denoise using a block size of @code{16x16}:
7010 Remove banding artifacts from input video.
7011 It works by replacing banded pixels with average value of referenced pixels.
7013 The filter accepts the following options:
7020 Set banding detection threshold for each plane. Default is 0.02.
7021 Valid range is 0.00003 to 0.5.
7022 If difference between current pixel and reference pixel is less than threshold,
7023 it will be considered as banded.
7026 Banding detection range in pixels. Default is 16. If positive, random number
7027 in range 0 to set value will be used. If negative, exact absolute value
7029 The range defines square of four pixels around current pixel.
7032 Set direction in radians from which four pixel will be compared. If positive,
7033 random direction from 0 to set direction will be picked. If negative, exact of
7034 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7035 will pick only pixels on same row and -PI/2 will pick only pixels on same
7039 If enabled, current pixel is compared with average value of all four
7040 surrounding pixels. The default is enabled. If disabled current pixel is
7041 compared with all four surrounding pixels. The pixel is considered banded
7042 if only all four differences with surrounding pixels are less than threshold.
7045 If enabled, current pixel is changed if and only if all pixel components are banded,
7046 e.g. banding detection threshold is triggered for all color components.
7047 The default is disabled.
7053 Drop duplicated frames at regular intervals.
7055 The filter accepts the following options:
7059 Set the number of frames from which one will be dropped. Setting this to
7060 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7061 Default is @code{5}.
7064 Set the threshold for duplicate detection. If the difference metric for a frame
7065 is less than or equal to this value, then it is declared as duplicate. Default
7069 Set scene change threshold. Default is @code{15}.
7073 Set the size of the x and y-axis blocks used during metric calculations.
7074 Larger blocks give better noise suppression, but also give worse detection of
7075 small movements. Must be a power of two. Default is @code{32}.
7078 Mark main input as a pre-processed input and activate clean source input
7079 stream. This allows the input to be pre-processed with various filters to help
7080 the metrics calculation while keeping the frame selection lossless. When set to
7081 @code{1}, the first stream is for the pre-processed input, and the second
7082 stream is the clean source from where the kept frames are chosen. Default is
7086 Set whether or not chroma is considered in the metric calculations. Default is
7092 Apply 2D deconvolution of video stream in frequency domain using second stream
7095 The filter accepts the following options:
7099 Set which planes to process.
7102 Set which impulse video frames will be processed, can be @var{first}
7103 or @var{all}. Default is @var{all}.
7106 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7107 and height are not same and not power of 2 or if stream prior to convolving
7111 The @code{deconvolve} filter also supports the @ref{framesync} options.
7115 Apply deflate effect to the video.
7117 This filter replaces the pixel by the local(3x3) average by taking into account
7118 only values lower than the pixel.
7120 It accepts the following options:
7127 Limit the maximum change for each plane, default is 65535.
7128 If 0, plane will remain unchanged.
7133 Remove temporal frame luminance variations.
7135 It accepts the following options:
7139 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7142 Set averaging mode to smooth temporal luminance variations.
7144 Available values are:
7169 Do not actually modify frame. Useful when one only wants metadata.
7174 Remove judder produced by partially interlaced telecined content.
7176 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7177 source was partially telecined content then the output of @code{pullup,dejudder}
7178 will have a variable frame rate. May change the recorded frame rate of the
7179 container. Aside from that change, this filter will not affect constant frame
7182 The option available in this filter is:
7186 Specify the length of the window over which the judder repeats.
7188 Accepts any integer greater than 1. Useful values are:
7192 If the original was telecined from 24 to 30 fps (Film to NTSC).
7195 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7198 If a mixture of the two.
7201 The default is @samp{4}.
7206 Suppress a TV station logo by a simple interpolation of the surrounding
7207 pixels. Just set a rectangle covering the logo and watch it disappear
7208 (and sometimes something even uglier appear - your mileage may vary).
7210 It accepts the following parameters:
7215 Specify the top left corner coordinates of the logo. They must be
7220 Specify the width and height of the logo to clear. They must be
7224 Specify the thickness of the fuzzy edge of the rectangle (added to
7225 @var{w} and @var{h}). The default value is 1. This option is
7226 deprecated, setting higher values should no longer be necessary and
7230 When set to 1, a green rectangle is drawn on the screen to simplify
7231 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7232 The default value is 0.
7234 The rectangle is drawn on the outermost pixels which will be (partly)
7235 replaced with interpolated values. The values of the next pixels
7236 immediately outside this rectangle in each direction will be used to
7237 compute the interpolated pixel values inside the rectangle.
7241 @subsection Examples
7245 Set a rectangle covering the area with top left corner coordinates 0,0
7246 and size 100x77, and a band of size 10:
7248 delogo=x=0:y=0:w=100:h=77:band=10
7255 Attempt to fix small changes in horizontal and/or vertical shift. This
7256 filter helps remove camera shake from hand-holding a camera, bumping a
7257 tripod, moving on a vehicle, etc.
7259 The filter accepts the following options:
7267 Specify a rectangular area where to limit the search for motion
7269 If desired the search for motion vectors can be limited to a
7270 rectangular area of the frame defined by its top left corner, width
7271 and height. These parameters have the same meaning as the drawbox
7272 filter which can be used to visualise the position of the bounding
7275 This is useful when simultaneous movement of subjects within the frame
7276 might be confused for camera motion by the motion vector search.
7278 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7279 then the full frame is used. This allows later options to be set
7280 without specifying the bounding box for the motion vector search.
7282 Default - search the whole frame.
7286 Specify the maximum extent of movement in x and y directions in the
7287 range 0-64 pixels. Default 16.
7290 Specify how to generate pixels to fill blanks at the edge of the
7291 frame. Available values are:
7294 Fill zeroes at blank locations
7296 Original image at blank locations
7298 Extruded edge value at blank locations
7300 Mirrored edge at blank locations
7302 Default value is @samp{mirror}.
7305 Specify the blocksize to use for motion search. Range 4-128 pixels,
7309 Specify the contrast threshold for blocks. Only blocks with more than
7310 the specified contrast (difference between darkest and lightest
7311 pixels) will be considered. Range 1-255, default 125.
7314 Specify the search strategy. Available values are:
7317 Set exhaustive search
7319 Set less exhaustive search.
7321 Default value is @samp{exhaustive}.
7324 If set then a detailed log of the motion search is written to the
7331 Remove unwanted contamination of foreground colors, caused by reflected color of
7332 greenscreen or bluescreen.
7334 This filter accepts the following options:
7338 Set what type of despill to use.
7341 Set how spillmap will be generated.
7344 Set how much to get rid of still remaining spill.
7347 Controls amount of red in spill area.
7350 Controls amount of green in spill area.
7351 Should be -1 for greenscreen.
7354 Controls amount of blue in spill area.
7355 Should be -1 for bluescreen.
7358 Controls brightness of spill area, preserving colors.
7361 Modify alpha from generated spillmap.
7366 Apply an exact inverse of the telecine operation. It requires a predefined
7367 pattern specified using the pattern option which must be the same as that passed
7368 to the telecine filter.
7370 This filter accepts the following options:
7379 The default value is @code{top}.
7383 A string of numbers representing the pulldown pattern you wish to apply.
7384 The default value is @code{23}.
7387 A number representing position of the first frame with respect to the telecine
7388 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7393 Apply dilation effect to the video.
7395 This filter replaces the pixel by the local(3x3) maximum.
7397 It accepts the following options:
7404 Limit the maximum change for each plane, default is 65535.
7405 If 0, plane will remain unchanged.
7408 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7411 Flags to local 3x3 coordinates maps like this:
7420 Displace pixels as indicated by second and third input stream.
7422 It takes three input streams and outputs one stream, the first input is the
7423 source, and second and third input are displacement maps.
7425 The second input specifies how much to displace pixels along the
7426 x-axis, while the third input specifies how much to displace pixels
7428 If one of displacement map streams terminates, last frame from that
7429 displacement map will be used.
7431 Note that once generated, displacements maps can be reused over and over again.
7433 A description of the accepted options follows.
7437 Set displace behavior for pixels that are out of range.
7439 Available values are:
7442 Missing pixels are replaced by black pixels.
7445 Adjacent pixels will spread out to replace missing pixels.
7448 Out of range pixels are wrapped so they point to pixels of other side.
7451 Out of range pixels will be replaced with mirrored pixels.
7453 Default is @samp{smear}.
7457 @subsection Examples
7461 Add ripple effect to rgb input of video size hd720:
7463 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
7467 Add wave effect to rgb input of video size hd720:
7469 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
7475 Draw a colored box on the input image.
7477 It accepts the following parameters:
7482 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7486 The expressions which specify the width and height of the box; if 0 they are interpreted as
7487 the input width and height. It defaults to 0.
7490 Specify the color of the box to write. For the general syntax of this option,
7491 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7492 value @code{invert} is used, the box edge color is the same as the
7493 video with inverted luma.
7496 The expression which sets the thickness of the box edge.
7497 A value of @code{fill} will create a filled box. Default value is @code{3}.
7499 See below for the list of accepted constants.
7502 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7503 will overwrite the video's color and alpha pixels.
7504 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7507 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7508 following constants:
7512 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7516 horizontal and vertical chroma subsample values. For example for the
7517 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7521 The input width and height.
7524 The input sample aspect ratio.
7528 The x and y offset coordinates where the box is drawn.
7532 The width and height of the drawn box.
7535 The thickness of the drawn box.
7537 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7538 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7542 @subsection Examples
7546 Draw a black box around the edge of the input image:
7552 Draw a box with color red and an opacity of 50%:
7554 drawbox=10:20:200:60:red@@0.5
7557 The previous example can be specified as:
7559 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7563 Fill the box with pink color:
7565 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7569 Draw a 2-pixel red 2.40:1 mask:
7571 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
7577 Draw a grid on the input image.
7579 It accepts the following parameters:
7584 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7588 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7589 input width and height, respectively, minus @code{thickness}, so image gets
7590 framed. Default to 0.
7593 Specify the color of the grid. For the general syntax of this option,
7594 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7595 value @code{invert} is used, the grid color is the same as the
7596 video with inverted luma.
7599 The expression which sets the thickness of the grid line. Default value is @code{1}.
7601 See below for the list of accepted constants.
7604 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7605 will overwrite the video's color and alpha pixels.
7606 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7609 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7610 following constants:
7614 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7618 horizontal and vertical chroma subsample values. For example for the
7619 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7623 The input grid cell width and height.
7626 The input sample aspect ratio.
7630 The x and y coordinates of some point of grid intersection (meant to configure offset).
7634 The width and height of the drawn cell.
7637 The thickness of the drawn cell.
7639 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7640 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7644 @subsection Examples
7648 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7650 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7654 Draw a white 3x3 grid with an opacity of 50%:
7656 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7663 Draw a text string or text from a specified file on top of a video, using the
7664 libfreetype library.
7666 To enable compilation of this filter, you need to configure FFmpeg with
7667 @code{--enable-libfreetype}.
7668 To enable default font fallback and the @var{font} option you need to
7669 configure FFmpeg with @code{--enable-libfontconfig}.
7670 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7671 @code{--enable-libfribidi}.
7675 It accepts the following parameters:
7680 Used to draw a box around text using the background color.
7681 The value must be either 1 (enable) or 0 (disable).
7682 The default value of @var{box} is 0.
7685 Set the width of the border to be drawn around the box using @var{boxcolor}.
7686 The default value of @var{boxborderw} is 0.
7689 The color to be used for drawing box around text. For the syntax of this
7690 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7692 The default value of @var{boxcolor} is "white".
7695 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7696 The default value of @var{line_spacing} is 0.
7699 Set the width of the border to be drawn around the text using @var{bordercolor}.
7700 The default value of @var{borderw} is 0.
7703 Set the color to be used for drawing border around text. For the syntax of this
7704 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7706 The default value of @var{bordercolor} is "black".
7709 Select how the @var{text} is expanded. Can be either @code{none},
7710 @code{strftime} (deprecated) or
7711 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7715 Set a start time for the count. Value is in microseconds. Only applied
7716 in the deprecated strftime expansion mode. To emulate in normal expansion
7717 mode use the @code{pts} function, supplying the start time (in seconds)
7718 as the second argument.
7721 If true, check and fix text coords to avoid clipping.
7724 The color to be used for drawing fonts. For the syntax of this option, check
7725 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7727 The default value of @var{fontcolor} is "black".
7729 @item fontcolor_expr
7730 String which is expanded the same way as @var{text} to obtain dynamic
7731 @var{fontcolor} value. By default this option has empty value and is not
7732 processed. When this option is set, it overrides @var{fontcolor} option.
7735 The font family to be used for drawing text. By default Sans.
7738 The font file to be used for drawing text. The path must be included.
7739 This parameter is mandatory if the fontconfig support is disabled.
7742 Draw the text applying alpha blending. The value can
7743 be a number between 0.0 and 1.0.
7744 The expression accepts the same variables @var{x, y} as well.
7745 The default value is 1.
7746 Please see @var{fontcolor_expr}.
7749 The font size to be used for drawing text.
7750 The default value of @var{fontsize} is 16.
7753 If set to 1, attempt to shape the text (for example, reverse the order of
7754 right-to-left text and join Arabic characters) before drawing it.
7755 Otherwise, just draw the text exactly as given.
7756 By default 1 (if supported).
7759 The flags to be used for loading the fonts.
7761 The flags map the corresponding flags supported by libfreetype, and are
7762 a combination of the following values:
7769 @item vertical_layout
7770 @item force_autohint
7773 @item ignore_global_advance_width
7775 @item ignore_transform
7781 Default value is "default".
7783 For more information consult the documentation for the FT_LOAD_*
7787 The color to be used for drawing a shadow behind the drawn text. For the
7788 syntax of this option, check the @ref{color syntax,,"Color" section in the
7789 ffmpeg-utils manual,ffmpeg-utils}.
7791 The default value of @var{shadowcolor} is "black".
7795 The x and y offsets for the text shadow position with respect to the
7796 position of the text. They can be either positive or negative
7797 values. The default value for both is "0".
7800 The starting frame number for the n/frame_num variable. The default value
7804 The size in number of spaces to use for rendering the tab.
7808 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7809 format. It can be used with or without text parameter. @var{timecode_rate}
7810 option must be specified.
7812 @item timecode_rate, rate, r
7813 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7814 integer. Minimum value is "1".
7815 Drop-frame timecode is supported for frame rates 30 & 60.
7818 If set to 1, the output of the timecode option will wrap around at 24 hours.
7819 Default is 0 (disabled).
7822 The text string to be drawn. The text must be a sequence of UTF-8
7824 This parameter is mandatory if no file is specified with the parameter
7828 A text file containing text to be drawn. The text must be a sequence
7829 of UTF-8 encoded characters.
7831 This parameter is mandatory if no text string is specified with the
7832 parameter @var{text}.
7834 If both @var{text} and @var{textfile} are specified, an error is thrown.
7837 If set to 1, the @var{textfile} will be reloaded before each frame.
7838 Be sure to update it atomically, or it may be read partially, or even fail.
7842 The expressions which specify the offsets where text will be drawn
7843 within the video frame. They are relative to the top/left border of the
7846 The default value of @var{x} and @var{y} is "0".
7848 See below for the list of accepted constants and functions.
7851 The parameters for @var{x} and @var{y} are expressions containing the
7852 following constants and functions:
7856 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7860 horizontal and vertical chroma subsample values. For example for the
7861 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7864 the height of each text line
7872 @item max_glyph_a, ascent
7873 the maximum distance from the baseline to the highest/upper grid
7874 coordinate used to place a glyph outline point, for all the rendered
7876 It is a positive value, due to the grid's orientation with the Y axis
7879 @item max_glyph_d, descent
7880 the maximum distance from the baseline to the lowest grid coordinate
7881 used to place a glyph outline point, for all the rendered glyphs.
7882 This is a negative value, due to the grid's orientation, with the Y axis
7886 maximum glyph height, that is the maximum height for all the glyphs
7887 contained in the rendered text, it is equivalent to @var{ascent} -
7891 maximum glyph width, that is the maximum width for all the glyphs
7892 contained in the rendered text
7895 the number of input frame, starting from 0
7897 @item rand(min, max)
7898 return a random number included between @var{min} and @var{max}
7901 The input sample aspect ratio.
7904 timestamp expressed in seconds, NAN if the input timestamp is unknown
7907 the height of the rendered text
7910 the width of the rendered text
7914 the x and y offset coordinates where the text is drawn.
7916 These parameters allow the @var{x} and @var{y} expressions to refer
7917 each other, so you can for example specify @code{y=x/dar}.
7920 @anchor{drawtext_expansion}
7921 @subsection Text expansion
7923 If @option{expansion} is set to @code{strftime},
7924 the filter recognizes strftime() sequences in the provided text and
7925 expands them accordingly. Check the documentation of strftime(). This
7926 feature is deprecated.
7928 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7930 If @option{expansion} is set to @code{normal} (which is the default),
7931 the following expansion mechanism is used.
7933 The backslash character @samp{\}, followed by any character, always expands to
7934 the second character.
7936 Sequences of the form @code{%@{...@}} are expanded. The text between the
7937 braces is a function name, possibly followed by arguments separated by ':'.
7938 If the arguments contain special characters or delimiters (':' or '@}'),
7939 they should be escaped.
7941 Note that they probably must also be escaped as the value for the
7942 @option{text} option in the filter argument string and as the filter
7943 argument in the filtergraph description, and possibly also for the shell,
7944 that makes up to four levels of escaping; using a text file avoids these
7947 The following functions are available:
7952 The expression evaluation result.
7954 It must take one argument specifying the expression to be evaluated,
7955 which accepts the same constants and functions as the @var{x} and
7956 @var{y} values. Note that not all constants should be used, for
7957 example the text size is not known when evaluating the expression, so
7958 the constants @var{text_w} and @var{text_h} will have an undefined
7961 @item expr_int_format, eif
7962 Evaluate the expression's value and output as formatted integer.
7964 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7965 The second argument specifies the output format. Allowed values are @samp{x},
7966 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7967 @code{printf} function.
7968 The third parameter is optional and sets the number of positions taken by the output.
7969 It can be used to add padding with zeros from the left.
7972 The time at which the filter is running, expressed in UTC.
7973 It can accept an argument: a strftime() format string.
7976 The time at which the filter is running, expressed in the local time zone.
7977 It can accept an argument: a strftime() format string.
7980 Frame metadata. Takes one or two arguments.
7982 The first argument is mandatory and specifies the metadata key.
7984 The second argument is optional and specifies a default value, used when the
7985 metadata key is not found or empty.
7988 The frame number, starting from 0.
7991 A 1 character description of the current picture type.
7994 The timestamp of the current frame.
7995 It can take up to three arguments.
7997 The first argument is the format of the timestamp; it defaults to @code{flt}
7998 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7999 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8000 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8001 @code{localtime} stands for the timestamp of the frame formatted as
8002 local time zone time.
8004 The second argument is an offset added to the timestamp.
8006 If the format is set to @code{localtime} or @code{gmtime},
8007 a third argument may be supplied: a strftime() format string.
8008 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8011 @subsection Examples
8015 Draw "Test Text" with font FreeSerif, using the default values for the
8016 optional parameters.
8019 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8023 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8024 and y=50 (counting from the top-left corner of the screen), text is
8025 yellow with a red box around it. Both the text and the box have an
8029 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8030 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8033 Note that the double quotes are not necessary if spaces are not used
8034 within the parameter list.
8037 Show the text at the center of the video frame:
8039 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8043 Show the text at a random position, switching to a new position every 30 seconds:
8045 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)"
8049 Show a text line sliding from right to left in the last row of the video
8050 frame. The file @file{LONG_LINE} is assumed to contain a single line
8053 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8057 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8059 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8063 Draw a single green letter "g", at the center of the input video.
8064 The glyph baseline is placed at half screen height.
8066 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8070 Show text for 1 second every 3 seconds:
8072 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8076 Use fontconfig to set the font. Note that the colons need to be escaped.
8078 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8082 Print the date of a real-time encoding (see strftime(3)):
8084 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8088 Show text fading in and out (appearing/disappearing):
8091 DS=1.0 # display start
8092 DE=10.0 # display end
8093 FID=1.5 # fade in duration
8094 FOD=5 # fade out duration
8095 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 @}"
8099 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8100 and the @option{fontsize} value are included in the @option{y} offset.
8102 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8103 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8108 For more information about libfreetype, check:
8109 @url{http://www.freetype.org/}.
8111 For more information about fontconfig, check:
8112 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8114 For more information about libfribidi, check:
8115 @url{http://fribidi.org/}.
8119 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8121 The filter accepts the following options:
8126 Set low and high threshold values used by the Canny thresholding
8129 The high threshold selects the "strong" edge pixels, which are then
8130 connected through 8-connectivity with the "weak" edge pixels selected
8131 by the low threshold.
8133 @var{low} and @var{high} threshold values must be chosen in the range
8134 [0,1], and @var{low} should be lesser or equal to @var{high}.
8136 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8140 Define the drawing mode.
8144 Draw white/gray wires on black background.
8147 Mix the colors to create a paint/cartoon effect.
8150 Default value is @var{wires}.
8153 @subsection Examples
8157 Standard edge detection with custom values for the hysteresis thresholding:
8159 edgedetect=low=0.1:high=0.4
8163 Painting effect without thresholding:
8165 edgedetect=mode=colormix:high=0
8170 Set brightness, contrast, saturation and approximate gamma adjustment.
8172 The filter accepts the following options:
8176 Set the contrast expression. The value must be a float value in range
8177 @code{-2.0} to @code{2.0}. The default value is "1".
8180 Set the brightness expression. The value must be a float value in
8181 range @code{-1.0} to @code{1.0}. The default value is "0".
8184 Set the saturation expression. The value must be a float in
8185 range @code{0.0} to @code{3.0}. The default value is "1".
8188 Set the gamma expression. The value must be a float in range
8189 @code{0.1} to @code{10.0}. The default value is "1".
8192 Set the gamma expression for red. The value must be a float in
8193 range @code{0.1} to @code{10.0}. The default value is "1".
8196 Set the gamma expression for green. The value must be a float in range
8197 @code{0.1} to @code{10.0}. The default value is "1".
8200 Set the gamma expression for blue. The value must be a float in range
8201 @code{0.1} to @code{10.0}. The default value is "1".
8204 Set the gamma weight expression. It can be used to reduce the effect
8205 of a high gamma value on bright image areas, e.g. keep them from
8206 getting overamplified and just plain white. The value must be a float
8207 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8208 gamma correction all the way down while @code{1.0} leaves it at its
8209 full strength. Default is "1".
8212 Set when the expressions for brightness, contrast, saturation and
8213 gamma expressions are evaluated.
8215 It accepts the following values:
8218 only evaluate expressions once during the filter initialization or
8219 when a command is processed
8222 evaluate expressions for each incoming frame
8225 Default value is @samp{init}.
8228 The expressions accept the following parameters:
8231 frame count of the input frame starting from 0
8234 byte position of the corresponding packet in the input file, NAN if
8238 frame rate of the input video, NAN if the input frame rate is unknown
8241 timestamp expressed in seconds, NAN if the input timestamp is unknown
8244 @subsection Commands
8245 The filter supports the following commands:
8249 Set the contrast expression.
8252 Set the brightness expression.
8255 Set the saturation expression.
8258 Set the gamma expression.
8261 Set the gamma_r expression.
8264 Set gamma_g expression.
8267 Set gamma_b expression.
8270 Set gamma_weight expression.
8272 The command accepts the same syntax of the corresponding option.
8274 If the specified expression is not valid, it is kept at its current
8281 Apply erosion effect to the video.
8283 This filter replaces the pixel by the local(3x3) minimum.
8285 It accepts the following options:
8292 Limit the maximum change for each plane, default is 65535.
8293 If 0, plane will remain unchanged.
8296 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8299 Flags to local 3x3 coordinates maps like this:
8306 @section extractplanes
8308 Extract color channel components from input video stream into
8309 separate grayscale video streams.
8311 The filter accepts the following option:
8315 Set plane(s) to extract.
8317 Available values for planes are:
8328 Choosing planes not available in the input will result in an error.
8329 That means you cannot select @code{r}, @code{g}, @code{b} planes
8330 with @code{y}, @code{u}, @code{v} planes at same time.
8333 @subsection Examples
8337 Extract luma, u and v color channel component from input video frame
8338 into 3 grayscale outputs:
8340 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
8346 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8348 For each input image, the filter will compute the optimal mapping from
8349 the input to the output given the codebook length, that is the number
8350 of distinct output colors.
8352 This filter accepts the following options.
8355 @item codebook_length, l
8356 Set codebook length. The value must be a positive integer, and
8357 represents the number of distinct output colors. Default value is 256.
8360 Set the maximum number of iterations to apply for computing the optimal
8361 mapping. The higher the value the better the result and the higher the
8362 computation time. Default value is 1.
8365 Set a random seed, must be an integer included between 0 and
8366 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8367 will try to use a good random seed on a best effort basis.
8370 Set pal8 output pixel format. This option does not work with codebook
8371 length greater than 256.
8376 Measure graylevel entropy in histogram of color channels of video frames.
8378 It accepts the following parameters:
8382 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8384 @var{diff} mode measures entropy of histogram delta values, absolute differences
8385 between neighbour histogram values.
8390 Apply a fade-in/out effect to the input video.
8392 It accepts the following parameters:
8396 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8398 Default is @code{in}.
8400 @item start_frame, s
8401 Specify the number of the frame to start applying the fade
8402 effect at. Default is 0.
8405 The number of frames that the fade effect lasts. At the end of the
8406 fade-in effect, the output video will have the same intensity as the input video.
8407 At the end of the fade-out transition, the output video will be filled with the
8408 selected @option{color}.
8412 If set to 1, fade only alpha channel, if one exists on the input.
8415 @item start_time, st
8416 Specify the timestamp (in seconds) of the frame to start to apply the fade
8417 effect. If both start_frame and start_time are specified, the fade will start at
8418 whichever comes last. Default is 0.
8421 The number of seconds for which the fade effect has to last. At the end of the
8422 fade-in effect the output video will have the same intensity as the input video,
8423 at the end of the fade-out transition the output video will be filled with the
8424 selected @option{color}.
8425 If both duration and nb_frames are specified, duration is used. Default is 0
8426 (nb_frames is used by default).
8429 Specify the color of the fade. Default is "black".
8432 @subsection Examples
8436 Fade in the first 30 frames of video:
8441 The command above is equivalent to:
8447 Fade out the last 45 frames of a 200-frame video:
8450 fade=type=out:start_frame=155:nb_frames=45
8454 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8456 fade=in:0:25, fade=out:975:25
8460 Make the first 5 frames yellow, then fade in from frame 5-24:
8462 fade=in:5:20:color=yellow
8466 Fade in alpha over first 25 frames of video:
8468 fade=in:0:25:alpha=1
8472 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8474 fade=t=in:st=5.5:d=0.5
8480 Apply arbitrary expressions to samples in frequency domain
8484 Adjust the dc value (gain) of the luma plane of the image. The filter
8485 accepts an integer value in range @code{0} to @code{1000}. The default
8486 value is set to @code{0}.
8489 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8490 filter accepts an integer value in range @code{0} to @code{1000}. The
8491 default value is set to @code{0}.
8494 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8495 filter accepts an integer value in range @code{0} to @code{1000}. The
8496 default value is set to @code{0}.
8499 Set the frequency domain weight expression for the luma plane.
8502 Set the frequency domain weight expression for the 1st chroma plane.
8505 Set the frequency domain weight expression for the 2nd chroma plane.
8508 Set when the expressions are evaluated.
8510 It accepts the following values:
8513 Only evaluate expressions once during the filter initialization.
8516 Evaluate expressions for each incoming frame.
8519 Default value is @samp{init}.
8521 The filter accepts the following variables:
8524 The coordinates of the current sample.
8528 The width and height of the image.
8531 The number of input frame, starting from 0.
8534 @subsection Examples
8540 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8546 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8552 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8558 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8565 Extract a single field from an interlaced image using stride
8566 arithmetic to avoid wasting CPU time. The output frames are marked as
8569 The filter accepts the following options:
8573 Specify whether to extract the top (if the value is @code{0} or
8574 @code{top}) or the bottom field (if the value is @code{1} or
8580 Create new frames by copying the top and bottom fields from surrounding frames
8581 supplied as numbers by the hint file.
8585 Set file containing hints: absolute/relative frame numbers.
8587 There must be one line for each frame in a clip. Each line must contain two
8588 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8589 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8590 is current frame number for @code{absolute} mode or out of [-1, 1] range
8591 for @code{relative} mode. First number tells from which frame to pick up top
8592 field and second number tells from which frame to pick up bottom field.
8594 If optionally followed by @code{+} output frame will be marked as interlaced,
8595 else if followed by @code{-} output frame will be marked as progressive, else
8596 it will be marked same as input frame.
8597 If line starts with @code{#} or @code{;} that line is skipped.
8600 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8603 Example of first several lines of @code{hint} file for @code{relative} mode:
8606 1,0 - # second frame, use third's frame top field and second's frame bottom field
8607 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8624 Field matching filter for inverse telecine. It is meant to reconstruct the
8625 progressive frames from a telecined stream. The filter does not drop duplicated
8626 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8627 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8629 The separation of the field matching and the decimation is notably motivated by
8630 the possibility of inserting a de-interlacing filter fallback between the two.
8631 If the source has mixed telecined and real interlaced content,
8632 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8633 But these remaining combed frames will be marked as interlaced, and thus can be
8634 de-interlaced by a later filter such as @ref{yadif} before decimation.
8636 In addition to the various configuration options, @code{fieldmatch} can take an
8637 optional second stream, activated through the @option{ppsrc} option. If
8638 enabled, the frames reconstruction will be based on the fields and frames from
8639 this second stream. This allows the first input to be pre-processed in order to
8640 help the various algorithms of the filter, while keeping the output lossless
8641 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8642 or brightness/contrast adjustments can help.
8644 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8645 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8646 which @code{fieldmatch} is based on. While the semantic and usage are very
8647 close, some behaviour and options names can differ.
8649 The @ref{decimate} filter currently only works for constant frame rate input.
8650 If your input has mixed telecined (30fps) and progressive content with a lower
8651 framerate like 24fps use the following filterchain to produce the necessary cfr
8652 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8654 The filter accepts the following options:
8658 Specify the assumed field order of the input stream. Available values are:
8662 Auto detect parity (use FFmpeg's internal parity value).
8664 Assume bottom field first.
8666 Assume top field first.
8669 Note that it is sometimes recommended not to trust the parity announced by the
8672 Default value is @var{auto}.
8675 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8676 sense that it won't risk creating jerkiness due to duplicate frames when
8677 possible, but if there are bad edits or blended fields it will end up
8678 outputting combed frames when a good match might actually exist. On the other
8679 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8680 but will almost always find a good frame if there is one. The other values are
8681 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8682 jerkiness and creating duplicate frames versus finding good matches in sections
8683 with bad edits, orphaned fields, blended fields, etc.
8685 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8687 Available values are:
8691 2-way matching (p/c)
8693 2-way matching, and trying 3rd match if still combed (p/c + n)
8695 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8697 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8698 still combed (p/c + n + u/b)
8700 3-way matching (p/c/n)
8702 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8703 detected as combed (p/c/n + u/b)
8706 The parenthesis at the end indicate the matches that would be used for that
8707 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8710 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8713 Default value is @var{pc_n}.
8716 Mark the main input stream as a pre-processed input, and enable the secondary
8717 input stream as the clean source to pick the fields from. See the filter
8718 introduction for more details. It is similar to the @option{clip2} feature from
8721 Default value is @code{0} (disabled).
8724 Set the field to match from. It is recommended to set this to the same value as
8725 @option{order} unless you experience matching failures with that setting. In
8726 certain circumstances changing the field that is used to match from can have a
8727 large impact on matching performance. Available values are:
8731 Automatic (same value as @option{order}).
8733 Match from the bottom field.
8735 Match from the top field.
8738 Default value is @var{auto}.
8741 Set whether or not chroma is included during the match comparisons. In most
8742 cases it is recommended to leave this enabled. You should set this to @code{0}
8743 only if your clip has bad chroma problems such as heavy rainbowing or other
8744 artifacts. Setting this to @code{0} could also be used to speed things up at
8745 the cost of some accuracy.
8747 Default value is @code{1}.
8751 These define an exclusion band which excludes the lines between @option{y0} and
8752 @option{y1} from being included in the field matching decision. An exclusion
8753 band can be used to ignore subtitles, a logo, or other things that may
8754 interfere with the matching. @option{y0} sets the starting scan line and
8755 @option{y1} sets the ending line; all lines in between @option{y0} and
8756 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8757 @option{y0} and @option{y1} to the same value will disable the feature.
8758 @option{y0} and @option{y1} defaults to @code{0}.
8761 Set the scene change detection threshold as a percentage of maximum change on
8762 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8763 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8764 @option{scthresh} is @code{[0.0, 100.0]}.
8766 Default value is @code{12.0}.
8769 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8770 account the combed scores of matches when deciding what match to use as the
8771 final match. Available values are:
8775 No final matching based on combed scores.
8777 Combed scores are only used when a scene change is detected.
8779 Use combed scores all the time.
8782 Default is @var{sc}.
8785 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8786 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8787 Available values are:
8791 No forced calculation.
8793 Force p/c/n calculations.
8795 Force p/c/n/u/b calculations.
8798 Default value is @var{none}.
8801 This is the area combing threshold used for combed frame detection. This
8802 essentially controls how "strong" or "visible" combing must be to be detected.
8803 Larger values mean combing must be more visible and smaller values mean combing
8804 can be less visible or strong and still be detected. Valid settings are from
8805 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8806 be detected as combed). This is basically a pixel difference value. A good
8807 range is @code{[8, 12]}.
8809 Default value is @code{9}.
8812 Sets whether or not chroma is considered in the combed frame decision. Only
8813 disable this if your source has chroma problems (rainbowing, etc.) that are
8814 causing problems for the combed frame detection with chroma enabled. Actually,
8815 using @option{chroma}=@var{0} is usually more reliable, except for the case
8816 where there is chroma only combing in the source.
8818 Default value is @code{0}.
8822 Respectively set the x-axis and y-axis size of the window used during combed
8823 frame detection. This has to do with the size of the area in which
8824 @option{combpel} pixels are required to be detected as combed for a frame to be
8825 declared combed. See the @option{combpel} parameter description for more info.
8826 Possible values are any number that is a power of 2 starting at 4 and going up
8829 Default value is @code{16}.
8832 The number of combed pixels inside any of the @option{blocky} by
8833 @option{blockx} size blocks on the frame for the frame to be detected as
8834 combed. While @option{cthresh} controls how "visible" the combing must be, this
8835 setting controls "how much" combing there must be in any localized area (a
8836 window defined by the @option{blockx} and @option{blocky} settings) on the
8837 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8838 which point no frames will ever be detected as combed). This setting is known
8839 as @option{MI} in TFM/VFM vocabulary.
8841 Default value is @code{80}.
8844 @anchor{p/c/n/u/b meaning}
8845 @subsection p/c/n/u/b meaning
8847 @subsubsection p/c/n
8849 We assume the following telecined stream:
8852 Top fields: 1 2 2 3 4
8853 Bottom fields: 1 2 3 4 4
8856 The numbers correspond to the progressive frame the fields relate to. Here, the
8857 first two frames are progressive, the 3rd and 4th are combed, and so on.
8859 When @code{fieldmatch} is configured to run a matching from bottom
8860 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8865 B 1 2 3 4 4 <-- matching reference
8874 As a result of the field matching, we can see that some frames get duplicated.
8875 To perform a complete inverse telecine, you need to rely on a decimation filter
8876 after this operation. See for instance the @ref{decimate} filter.
8878 The same operation now matching from top fields (@option{field}=@var{top})
8883 T 1 2 2 3 4 <-- matching reference
8893 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8894 basically, they refer to the frame and field of the opposite parity:
8897 @item @var{p} matches the field of the opposite parity in the previous frame
8898 @item @var{c} matches the field of the opposite parity in the current frame
8899 @item @var{n} matches the field of the opposite parity in the next frame
8904 The @var{u} and @var{b} matching are a bit special in the sense that they match
8905 from the opposite parity flag. In the following examples, we assume that we are
8906 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8907 'x' is placed above and below each matched fields.
8909 With bottom matching (@option{field}=@var{bottom}):
8914 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8915 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8923 With top matching (@option{field}=@var{top}):
8928 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8929 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8937 @subsection Examples
8939 Simple IVTC of a top field first telecined stream:
8941 fieldmatch=order=tff:combmatch=none, decimate
8944 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8946 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8951 Transform the field order of the input video.
8953 It accepts the following parameters:
8958 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8959 for bottom field first.
8962 The default value is @samp{tff}.
8964 The transformation is done by shifting the picture content up or down
8965 by one line, and filling the remaining line with appropriate picture content.
8966 This method is consistent with most broadcast field order converters.
8968 If the input video is not flagged as being interlaced, or it is already
8969 flagged as being of the required output field order, then this filter does
8970 not alter the incoming video.
8972 It is very useful when converting to or from PAL DV material,
8973 which is bottom field first.
8977 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8980 @section fifo, afifo
8982 Buffer input images and send them when they are requested.
8984 It is mainly useful when auto-inserted by the libavfilter
8987 It does not take parameters.
8989 @section fillborders
8991 Fill borders of the input video, without changing video stream dimensions.
8992 Sometimes video can have garbage at the four edges and you may not want to
8993 crop video input to keep size multiple of some number.
8995 This filter accepts the following options:
8999 Number of pixels to fill from left border.
9002 Number of pixels to fill from right border.
9005 Number of pixels to fill from top border.
9008 Number of pixels to fill from bottom border.
9013 It accepts the following values:
9016 fill pixels using outermost pixels
9019 fill pixels using mirroring
9022 fill pixels with constant value
9025 Default is @var{smear}.
9028 Set color for pixels in fixed mode. Default is @var{black}.
9033 Find a rectangular object
9035 It accepts the following options:
9039 Filepath of the object image, needs to be in gray8.
9042 Detection threshold, default is 0.5.
9045 Number of mipmaps, default is 3.
9047 @item xmin, ymin, xmax, ymax
9048 Specifies the rectangle in which to search.
9051 @subsection Examples
9055 Generate a representative palette of a given video using @command{ffmpeg}:
9057 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9063 Cover a rectangular object
9065 It accepts the following options:
9069 Filepath of the optional cover image, needs to be in yuv420.
9074 It accepts the following values:
9077 cover it by the supplied image
9079 cover it by interpolating the surrounding pixels
9082 Default value is @var{blur}.
9085 @subsection Examples
9089 Generate a representative palette of a given video using @command{ffmpeg}:
9091 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9097 Flood area with values of same pixel components with another values.
9099 It accepts the following options:
9102 Set pixel x coordinate.
9105 Set pixel y coordinate.
9108 Set source #0 component value.
9111 Set source #1 component value.
9114 Set source #2 component value.
9117 Set source #3 component value.
9120 Set destination #0 component value.
9123 Set destination #1 component value.
9126 Set destination #2 component value.
9129 Set destination #3 component value.
9135 Convert the input video to one of the specified pixel formats.
9136 Libavfilter will try to pick one that is suitable as input to
9139 It accepts the following parameters:
9143 A '|'-separated list of pixel format names, such as
9144 "pix_fmts=yuv420p|monow|rgb24".
9148 @subsection Examples
9152 Convert the input video to the @var{yuv420p} format
9154 format=pix_fmts=yuv420p
9157 Convert the input video to any of the formats in the list
9159 format=pix_fmts=yuv420p|yuv444p|yuv410p
9166 Convert the video to specified constant frame rate by duplicating or dropping
9167 frames as necessary.
9169 It accepts the following parameters:
9173 The desired output frame rate. The default is @code{25}.
9176 Assume the first PTS should be the given value, in seconds. This allows for
9177 padding/trimming at the start of stream. By default, no assumption is made
9178 about the first frame's expected PTS, so no padding or trimming is done.
9179 For example, this could be set to 0 to pad the beginning with duplicates of
9180 the first frame if a video stream starts after the audio stream or to trim any
9181 frames with a negative PTS.
9184 Timestamp (PTS) rounding method.
9186 Possible values are:
9193 round towards -infinity
9195 round towards +infinity
9199 The default is @code{near}.
9202 Action performed when reading the last frame.
9204 Possible values are:
9207 Use same timestamp rounding method as used for other frames.
9209 Pass through last frame if input duration has not been reached yet.
9211 The default is @code{round}.
9215 Alternatively, the options can be specified as a flat string:
9216 @var{fps}[:@var{start_time}[:@var{round}]].
9218 See also the @ref{setpts} filter.
9220 @subsection Examples
9224 A typical usage in order to set the fps to 25:
9230 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9232 fps=fps=film:round=near
9238 Pack two different video streams into a stereoscopic video, setting proper
9239 metadata on supported codecs. The two views should have the same size and
9240 framerate and processing will stop when the shorter video ends. Please note
9241 that you may conveniently adjust view properties with the @ref{scale} and
9244 It accepts the following parameters:
9248 The desired packing format. Supported values are:
9253 The views are next to each other (default).
9256 The views are on top of each other.
9259 The views are packed by line.
9262 The views are packed by column.
9265 The views are temporally interleaved.
9274 # Convert left and right views into a frame-sequential video
9275 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9277 # Convert views into a side-by-side video with the same output resolution as the input
9278 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
9283 Change the frame rate by interpolating new video output frames from the source
9286 This filter is not designed to function correctly with interlaced media. If
9287 you wish to change the frame rate of interlaced media then you are required
9288 to deinterlace before this filter and re-interlace after this filter.
9290 A description of the accepted options follows.
9294 Specify the output frames per second. This option can also be specified
9295 as a value alone. The default is @code{50}.
9298 Specify the start of a range where the output frame will be created as a
9299 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9300 the default is @code{15}.
9303 Specify the end of a range where the output frame will be created as a
9304 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9305 the default is @code{240}.
9308 Specify the level at which a scene change is detected as a value between
9309 0 and 100 to indicate a new scene; a low value reflects a low
9310 probability for the current frame to introduce a new scene, while a higher
9311 value means the current frame is more likely to be one.
9312 The default is @code{8.2}.
9315 Specify flags influencing the filter process.
9317 Available value for @var{flags} is:
9320 @item scene_change_detect, scd
9321 Enable scene change detection using the value of the option @var{scene}.
9322 This flag is enabled by default.
9328 Select one frame every N-th frame.
9330 This filter accepts the following option:
9333 Select frame after every @code{step} frames.
9334 Allowed values are positive integers higher than 0. Default value is @code{1}.
9340 Apply a frei0r effect to the input video.
9342 To enable the compilation of this filter, you need to install the frei0r
9343 header and configure FFmpeg with @code{--enable-frei0r}.
9345 It accepts the following parameters:
9350 The name of the frei0r effect to load. If the environment variable
9351 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9352 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9353 Otherwise, the standard frei0r paths are searched, in this order:
9354 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9355 @file{/usr/lib/frei0r-1/}.
9358 A '|'-separated list of parameters to pass to the frei0r effect.
9362 A frei0r effect parameter can be a boolean (its value is either
9363 "y" or "n"), a double, a color (specified as
9364 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9365 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9366 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9367 a position (specified as @var{X}/@var{Y}, where
9368 @var{X} and @var{Y} are floating point numbers) and/or a string.
9370 The number and types of parameters depend on the loaded effect. If an
9371 effect parameter is not specified, the default value is set.
9373 @subsection Examples
9377 Apply the distort0r effect, setting the first two double parameters:
9379 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9383 Apply the colordistance effect, taking a color as the first parameter:
9385 frei0r=colordistance:0.2/0.3/0.4
9386 frei0r=colordistance:violet
9387 frei0r=colordistance:0x112233
9391 Apply the perspective effect, specifying the top left and top right image
9394 frei0r=perspective:0.2/0.2|0.8/0.2
9398 For more information, see
9399 @url{http://frei0r.dyne.org}
9403 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9405 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9406 processing filter, one of them is performed once per block, not per pixel.
9407 This allows for much higher speed.
9409 The filter accepts the following options:
9413 Set quality. This option defines the number of levels for averaging. It accepts
9414 an integer in the range 4-5. Default value is @code{4}.
9417 Force a constant quantization parameter. It accepts an integer in range 0-63.
9418 If not set, the filter will use the QP from the video stream (if available).
9421 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9422 more details but also more artifacts, while higher values make the image smoother
9423 but also blurrier. Default value is @code{0} − PSNR optimal.
9426 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9427 option may cause flicker since the B-Frames have often larger QP. Default is
9428 @code{0} (not enabled).
9434 Apply Gaussian blur filter.
9436 The filter accepts the following options:
9440 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9443 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9446 Set which planes to filter. By default all planes are filtered.
9449 Set vertical sigma, if negative it will be same as @code{sigma}.
9450 Default is @code{-1}.
9455 The filter accepts the following options:
9459 Set the luminance expression.
9461 Set the chrominance blue expression.
9463 Set the chrominance red expression.
9465 Set the alpha expression.
9467 Set the red expression.
9469 Set the green expression.
9471 Set the blue expression.
9474 The colorspace is selected according to the specified options. If one
9475 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9476 options is specified, the filter will automatically select a YCbCr
9477 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9478 @option{blue_expr} options is specified, it will select an RGB
9481 If one of the chrominance expression is not defined, it falls back on the other
9482 one. If no alpha expression is specified it will evaluate to opaque value.
9483 If none of chrominance expressions are specified, they will evaluate
9484 to the luminance expression.
9486 The expressions can use the following variables and functions:
9490 The sequential number of the filtered frame, starting from @code{0}.
9494 The coordinates of the current sample.
9498 The width and height of the image.
9502 Width and height scale depending on the currently filtered plane. It is the
9503 ratio between the corresponding luma plane number of pixels and the current
9504 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9505 @code{0.5,0.5} for chroma planes.
9508 Time of the current frame, expressed in seconds.
9511 Return the value of the pixel at location (@var{x},@var{y}) of the current
9515 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9519 Return the value of the pixel at location (@var{x},@var{y}) of the
9520 blue-difference chroma plane. Return 0 if there is no such plane.
9523 Return the value of the pixel at location (@var{x},@var{y}) of the
9524 red-difference chroma plane. Return 0 if there is no such plane.
9529 Return the value of the pixel at location (@var{x},@var{y}) of the
9530 red/green/blue component. Return 0 if there is no such component.
9533 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9534 plane. Return 0 if there is no such plane.
9537 For functions, if @var{x} and @var{y} are outside the area, the value will be
9538 automatically clipped to the closer edge.
9540 @subsection Examples
9544 Flip the image horizontally:
9550 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9551 wavelength of 100 pixels:
9553 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9557 Generate a fancy enigmatic moving light:
9559 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
9563 Generate a quick emboss effect:
9565 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9569 Modify RGB components depending on pixel position:
9571 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9575 Create a radial gradient that is the same size as the input (also see
9576 the @ref{vignette} filter):
9578 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9584 Fix the banding artifacts that are sometimes introduced into nearly flat
9585 regions by truncation to 8-bit color depth.
9586 Interpolate the gradients that should go where the bands are, and
9589 It is designed for playback only. Do not use it prior to
9590 lossy compression, because compression tends to lose the dither and
9591 bring back the bands.
9593 It accepts the following parameters:
9598 The maximum amount by which the filter will change any one pixel. This is also
9599 the threshold for detecting nearly flat regions. Acceptable values range from
9600 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9604 The neighborhood to fit the gradient to. A larger radius makes for smoother
9605 gradients, but also prevents the filter from modifying the pixels near detailed
9606 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9607 values will be clipped to the valid range.
9611 Alternatively, the options can be specified as a flat string:
9612 @var{strength}[:@var{radius}]
9614 @subsection Examples
9618 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9624 Specify radius, omitting the strength (which will fall-back to the default
9635 Apply a Hald CLUT to a video stream.
9637 First input is the video stream to process, and second one is the Hald CLUT.
9638 The Hald CLUT input can be a simple picture or a complete video stream.
9640 The filter accepts the following options:
9644 Force termination when the shortest input terminates. Default is @code{0}.
9646 Continue applying the last CLUT after the end of the stream. A value of
9647 @code{0} disable the filter after the last frame of the CLUT is reached.
9648 Default is @code{1}.
9651 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9652 filters share the same internals).
9654 More information about the Hald CLUT can be found on Eskil Steenberg's website
9655 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9657 @subsection Workflow examples
9659 @subsubsection Hald CLUT video stream
9661 Generate an identity Hald CLUT stream altered with various effects:
9663 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
9666 Note: make sure you use a lossless codec.
9668 Then use it with @code{haldclut} to apply it on some random stream:
9670 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9673 The Hald CLUT will be applied to the 10 first seconds (duration of
9674 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9675 to the remaining frames of the @code{mandelbrot} stream.
9677 @subsubsection Hald CLUT with preview
9679 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9680 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9681 biggest possible square starting at the top left of the picture. The remaining
9682 padding pixels (bottom or right) will be ignored. This area can be used to add
9683 a preview of the Hald CLUT.
9685 Typically, the following generated Hald CLUT will be supported by the
9686 @code{haldclut} filter:
9689 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9690 pad=iw+320 [padded_clut];
9691 smptebars=s=320x256, split [a][b];
9692 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9693 [main][b] overlay=W-320" -frames:v 1 clut.png
9696 It contains the original and a preview of the effect of the CLUT: SMPTE color
9697 bars are displayed on the right-top, and below the same color bars processed by
9700 Then, the effect of this Hald CLUT can be visualized with:
9702 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9707 Flip the input video horizontally.
9709 For example, to horizontally flip the input video with @command{ffmpeg}:
9711 ffmpeg -i in.avi -vf "hflip" out.avi
9715 This filter applies a global color histogram equalization on a
9718 It can be used to correct video that has a compressed range of pixel
9719 intensities. The filter redistributes the pixel intensities to
9720 equalize their distribution across the intensity range. It may be
9721 viewed as an "automatically adjusting contrast filter". This filter is
9722 useful only for correcting degraded or poorly captured source
9725 The filter accepts the following options:
9729 Determine the amount of equalization to be applied. As the strength
9730 is reduced, the distribution of pixel intensities more-and-more
9731 approaches that of the input frame. The value must be a float number
9732 in the range [0,1] and defaults to 0.200.
9735 Set the maximum intensity that can generated and scale the output
9736 values appropriately. The strength should be set as desired and then
9737 the intensity can be limited if needed to avoid washing-out. The value
9738 must be a float number in the range [0,1] and defaults to 0.210.
9741 Set the antibanding level. If enabled the filter will randomly vary
9742 the luminance of output pixels by a small amount to avoid banding of
9743 the histogram. Possible values are @code{none}, @code{weak} or
9744 @code{strong}. It defaults to @code{none}.
9749 Compute and draw a color distribution histogram for the input video.
9751 The computed histogram is a representation of the color component
9752 distribution in an image.
9754 Standard histogram displays the color components distribution in an image.
9755 Displays color graph for each color component. Shows distribution of
9756 the Y, U, V, A or R, G, B components, depending on input format, in the
9757 current frame. Below each graph a color component scale meter is shown.
9759 The filter accepts the following options:
9763 Set height of level. Default value is @code{200}.
9764 Allowed range is [50, 2048].
9767 Set height of color scale. Default value is @code{12}.
9768 Allowed range is [0, 40].
9772 It accepts the following values:
9775 Per color component graphs are placed below each other.
9778 Per color component graphs are placed side by side.
9781 Presents information identical to that in the @code{parade}, except
9782 that the graphs representing color components are superimposed directly
9785 Default is @code{stack}.
9788 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9789 Default is @code{linear}.
9792 Set what color components to display.
9793 Default is @code{7}.
9796 Set foreground opacity. Default is @code{0.7}.
9799 Set background opacity. Default is @code{0.5}.
9802 @subsection Examples
9807 Calculate and draw histogram:
9809 ffplay -i input -vf histogram
9817 This is a high precision/quality 3d denoise filter. It aims to reduce
9818 image noise, producing smooth images and making still images really
9819 still. It should enhance compressibility.
9821 It accepts the following optional parameters:
9825 A non-negative floating point number which specifies spatial luma strength.
9828 @item chroma_spatial
9829 A non-negative floating point number which specifies spatial chroma strength.
9830 It defaults to 3.0*@var{luma_spatial}/4.0.
9833 A floating point number which specifies luma temporal strength. It defaults to
9834 6.0*@var{luma_spatial}/4.0.
9837 A floating point number which specifies chroma temporal strength. It defaults to
9838 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9843 Download hardware frames to system memory.
9845 The input must be in hardware frames, and the output a non-hardware format.
9846 Not all formats will be supported on the output - it may be necessary to insert
9847 an additional @option{format} filter immediately following in the graph to get
9848 the output in a supported format.
9852 Map hardware frames to system memory or to another device.
9854 This filter has several different modes of operation; which one is used depends
9855 on the input and output formats:
9858 Hardware frame input, normal frame output
9860 Map the input frames to system memory and pass them to the output. If the
9861 original hardware frame is later required (for example, after overlaying
9862 something else on part of it), the @option{hwmap} filter can be used again
9863 in the next mode to retrieve it.
9865 Normal frame input, hardware frame output
9867 If the input is actually a software-mapped hardware frame, then unmap it -
9868 that is, return the original hardware frame.
9870 Otherwise, a device must be provided. Create new hardware surfaces on that
9871 device for the output, then map them back to the software format at the input
9872 and give those frames to the preceding filter. This will then act like the
9873 @option{hwupload} filter, but may be able to avoid an additional copy when
9874 the input is already in a compatible format.
9876 Hardware frame input and output
9878 A device must be supplied for the output, either directly or with the
9879 @option{derive_device} option. The input and output devices must be of
9880 different types and compatible - the exact meaning of this is
9881 system-dependent, but typically it means that they must refer to the same
9882 underlying hardware context (for example, refer to the same graphics card).
9884 If the input frames were originally created on the output device, then unmap
9885 to retrieve the original frames.
9887 Otherwise, map the frames to the output device - create new hardware frames
9888 on the output corresponding to the frames on the input.
9891 The following additional parameters are accepted:
9895 Set the frame mapping mode. Some combination of:
9898 The mapped frame should be readable.
9900 The mapped frame should be writeable.
9902 The mapping will always overwrite the entire frame.
9904 This may improve performance in some cases, as the original contents of the
9905 frame need not be loaded.
9907 The mapping must not involve any copying.
9909 Indirect mappings to copies of frames are created in some cases where either
9910 direct mapping is not possible or it would have unexpected properties.
9911 Setting this flag ensures that the mapping is direct and will fail if that is
9914 Defaults to @var{read+write} if not specified.
9916 @item derive_device @var{type}
9917 Rather than using the device supplied at initialisation, instead derive a new
9918 device of type @var{type} from the device the input frames exist on.
9921 In a hardware to hardware mapping, map in reverse - create frames in the sink
9922 and map them back to the source. This may be necessary in some cases where
9923 a mapping in one direction is required but only the opposite direction is
9924 supported by the devices being used.
9926 This option is dangerous - it may break the preceding filter in undefined
9927 ways if there are any additional constraints on that filter's output.
9928 Do not use it without fully understanding the implications of its use.
9933 Upload system memory frames to hardware surfaces.
9935 The device to upload to must be supplied when the filter is initialised. If
9936 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9939 @anchor{hwupload_cuda}
9940 @section hwupload_cuda
9942 Upload system memory frames to a CUDA device.
9944 It accepts the following optional parameters:
9948 The number of the CUDA device to use
9953 Apply a high-quality magnification filter designed for pixel art. This filter
9954 was originally created by Maxim Stepin.
9956 It accepts the following option:
9960 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9961 @code{hq3x} and @code{4} for @code{hq4x}.
9962 Default is @code{3}.
9966 Stack input videos horizontally.
9968 All streams must be of same pixel format and of same height.
9970 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9971 to create same output.
9973 The filter accept the following option:
9977 Set number of input streams. Default is 2.
9980 If set to 1, force the output to terminate when the shortest input
9981 terminates. Default value is 0.
9986 Modify the hue and/or the saturation of the input.
9988 It accepts the following parameters:
9992 Specify the hue angle as a number of degrees. It accepts an expression,
9993 and defaults to "0".
9996 Specify the saturation in the [-10,10] range. It accepts an expression and
10000 Specify the hue angle as a number of radians. It accepts an
10001 expression, and defaults to "0".
10004 Specify the brightness in the [-10,10] range. It accepts an expression and
10008 @option{h} and @option{H} are mutually exclusive, and can't be
10009 specified at the same time.
10011 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10012 expressions containing the following constants:
10016 frame count of the input frame starting from 0
10019 presentation timestamp of the input frame expressed in time base units
10022 frame rate of the input video, NAN if the input frame rate is unknown
10025 timestamp expressed in seconds, NAN if the input timestamp is unknown
10028 time base of the input video
10031 @subsection Examples
10035 Set the hue to 90 degrees and the saturation to 1.0:
10041 Same command but expressing the hue in radians:
10047 Rotate hue and make the saturation swing between 0
10048 and 2 over a period of 1 second:
10050 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10054 Apply a 3 seconds saturation fade-in effect starting at 0:
10056 hue="s=min(t/3\,1)"
10059 The general fade-in expression can be written as:
10061 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10065 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10067 hue="s=max(0\, min(1\, (8-t)/3))"
10070 The general fade-out expression can be written as:
10072 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10077 @subsection Commands
10079 This filter supports the following commands:
10085 Modify the hue and/or the saturation and/or brightness of the input video.
10086 The command accepts the same syntax of the corresponding option.
10088 If the specified expression is not valid, it is kept at its current
10092 @section hysteresis
10094 Grow first stream into second stream by connecting components.
10095 This makes it possible to build more robust edge masks.
10097 This filter accepts the following options:
10101 Set which planes will be processed as bitmap, unprocessed planes will be
10102 copied from first stream.
10103 By default value 0xf, all planes will be processed.
10106 Set threshold which is used in filtering. If pixel component value is higher than
10107 this value filter algorithm for connecting components is activated.
10108 By default value is 0.
10113 Detect video interlacing type.
10115 This filter tries to detect if the input frames are interlaced, progressive,
10116 top or bottom field first. It will also try to detect fields that are
10117 repeated between adjacent frames (a sign of telecine).
10119 Single frame detection considers only immediately adjacent frames when classifying each frame.
10120 Multiple frame detection incorporates the classification history of previous frames.
10122 The filter will log these metadata values:
10125 @item single.current_frame
10126 Detected type of current frame using single-frame detection. One of:
10127 ``tff'' (top field first), ``bff'' (bottom field first),
10128 ``progressive'', or ``undetermined''
10131 Cumulative number of frames detected as top field first using single-frame detection.
10134 Cumulative number of frames detected as top field first using multiple-frame detection.
10137 Cumulative number of frames detected as bottom field first using single-frame detection.
10139 @item multiple.current_frame
10140 Detected type of current frame using multiple-frame detection. One of:
10141 ``tff'' (top field first), ``bff'' (bottom field first),
10142 ``progressive'', or ``undetermined''
10145 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10147 @item single.progressive
10148 Cumulative number of frames detected as progressive using single-frame detection.
10150 @item multiple.progressive
10151 Cumulative number of frames detected as progressive using multiple-frame detection.
10153 @item single.undetermined
10154 Cumulative number of frames that could not be classified using single-frame detection.
10156 @item multiple.undetermined
10157 Cumulative number of frames that could not be classified using multiple-frame detection.
10159 @item repeated.current_frame
10160 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10162 @item repeated.neither
10163 Cumulative number of frames with no repeated field.
10166 Cumulative number of frames with the top field repeated from the previous frame's top field.
10168 @item repeated.bottom
10169 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10172 The filter accepts the following options:
10176 Set interlacing threshold.
10178 Set progressive threshold.
10180 Threshold for repeated field detection.
10182 Number of frames after which a given frame's contribution to the
10183 statistics is halved (i.e., it contributes only 0.5 to its
10184 classification). The default of 0 means that all frames seen are given
10185 full weight of 1.0 forever.
10186 @item analyze_interlaced_flag
10187 When this is not 0 then idet will use the specified number of frames to determine
10188 if the interlaced flag is accurate, it will not count undetermined frames.
10189 If the flag is found to be accurate it will be used without any further
10190 computations, if it is found to be inaccurate it will be cleared without any
10191 further computations. This allows inserting the idet filter as a low computational
10192 method to clean up the interlaced flag
10197 Deinterleave or interleave fields.
10199 This filter allows one to process interlaced images fields without
10200 deinterlacing them. Deinterleaving splits the input frame into 2
10201 fields (so called half pictures). Odd lines are moved to the top
10202 half of the output image, even lines to the bottom half.
10203 You can process (filter) them independently and then re-interleave them.
10205 The filter accepts the following options:
10209 @item chroma_mode, c
10210 @item alpha_mode, a
10211 Available values for @var{luma_mode}, @var{chroma_mode} and
10212 @var{alpha_mode} are:
10218 @item deinterleave, d
10219 Deinterleave fields, placing one above the other.
10221 @item interleave, i
10222 Interleave fields. Reverse the effect of deinterleaving.
10224 Default value is @code{none}.
10226 @item luma_swap, ls
10227 @item chroma_swap, cs
10228 @item alpha_swap, as
10229 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10234 Apply inflate effect to the video.
10236 This filter replaces the pixel by the local(3x3) average by taking into account
10237 only values higher than the pixel.
10239 It accepts the following options:
10246 Limit the maximum change for each plane, default is 65535.
10247 If 0, plane will remain unchanged.
10252 Simple interlacing filter from progressive contents. This interleaves upper (or
10253 lower) lines from odd frames with lower (or upper) lines from even frames,
10254 halving the frame rate and preserving image height.
10257 Original Original New Frame
10258 Frame 'j' Frame 'j+1' (tff)
10259 ========== =========== ==================
10260 Line 0 --------------------> Frame 'j' Line 0
10261 Line 1 Line 1 ----> Frame 'j+1' Line 1
10262 Line 2 ---------------------> Frame 'j' Line 2
10263 Line 3 Line 3 ----> Frame 'j+1' Line 3
10265 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10268 It accepts the following optional parameters:
10272 This determines whether the interlaced frame is taken from the even
10273 (tff - default) or odd (bff) lines of the progressive frame.
10276 Vertical lowpass filter to avoid twitter interlacing and
10277 reduce moire patterns.
10281 Disable vertical lowpass filter
10284 Enable linear filter (default)
10287 Enable complex filter. This will slightly less reduce twitter and moire
10288 but better retain detail and subjective sharpness impression.
10295 Deinterlace input video by applying Donald Graft's adaptive kernel
10296 deinterling. Work on interlaced parts of a video to produce
10297 progressive frames.
10299 The description of the accepted parameters follows.
10303 Set the threshold which affects the filter's tolerance when
10304 determining if a pixel line must be processed. It must be an integer
10305 in the range [0,255] and defaults to 10. A value of 0 will result in
10306 applying the process on every pixels.
10309 Paint pixels exceeding the threshold value to white if set to 1.
10313 Set the fields order. Swap fields if set to 1, leave fields alone if
10317 Enable additional sharpening if set to 1. Default is 0.
10320 Enable twoway sharpening if set to 1. Default is 0.
10323 @subsection Examples
10327 Apply default values:
10329 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10333 Enable additional sharpening:
10339 Paint processed pixels in white:
10345 @section lenscorrection
10347 Correct radial lens distortion
10349 This filter can be used to correct for radial distortion as can result from the use
10350 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10351 one can use tools available for example as part of opencv or simply trial-and-error.
10352 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10353 and extract the k1 and k2 coefficients from the resulting matrix.
10355 Note that effectively the same filter is available in the open-source tools Krita and
10356 Digikam from the KDE project.
10358 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10359 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10360 brightness distribution, so you may want to use both filters together in certain
10361 cases, though you will have to take care of ordering, i.e. whether vignetting should
10362 be applied before or after lens correction.
10364 @subsection Options
10366 The filter accepts the following options:
10370 Relative x-coordinate of the focal point of the image, and thereby the center of the
10371 distortion. This value has a range [0,1] and is expressed as fractions of the image
10374 Relative y-coordinate of the focal point of the image, and thereby the center of the
10375 distortion. This value has a range [0,1] and is expressed as fractions of the image
10378 Coefficient of the quadratic correction term. 0.5 means no correction.
10380 Coefficient of the double quadratic correction term. 0.5 means no correction.
10383 The formula that generates the correction is:
10385 @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)
10387 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10388 distances from the focal point in the source and target images, respectively.
10392 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10393 score between two input videos.
10395 The obtained VMAF score is printed through the logging system.
10397 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10398 After installing the library it can be enabled using:
10399 @code{./configure --enable-libvmaf}.
10400 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10402 The filter has following options:
10406 Set the model path which is to be used for SVM.
10407 Default value: @code{"vmaf_v0.6.1.pkl"}
10410 Set the file path to be used to store logs.
10413 Set the format of the log file (xml or json).
10415 @item enable_transform
10416 Enables transform for computing vmaf.
10419 Invokes the phone model which will generate VMAF scores higher than in the
10420 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10423 Enables computing psnr along with vmaf.
10426 Enables computing ssim along with vmaf.
10429 Enables computing ms_ssim along with vmaf.
10432 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10435 This filter also supports the @ref{framesync} options.
10437 On the below examples the input file @file{main.mpg} being processed is
10438 compared with the reference file @file{ref.mpg}.
10441 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10444 Example with options:
10446 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10451 Limits the pixel components values to the specified range [min, max].
10453 The filter accepts the following options:
10457 Lower bound. Defaults to the lowest allowed value for the input.
10460 Upper bound. Defaults to the highest allowed value for the input.
10463 Specify which planes will be processed. Defaults to all available.
10470 The filter accepts the following options:
10474 Set the number of loops. Setting this value to -1 will result in infinite loops.
10478 Set maximal size in number of frames. Default is 0.
10481 Set first frame of loop. Default is 0.
10487 Apply a 3D LUT to an input video.
10489 The filter accepts the following options:
10493 Set the 3D LUT file name.
10495 Currently supported formats:
10507 Select interpolation mode.
10509 Available values are:
10513 Use values from the nearest defined point.
10515 Interpolate values using the 8 points defining a cube.
10517 Interpolate values using a tetrahedron.
10521 This filter also supports the @ref{framesync} options.
10525 Turn certain luma values into transparency.
10527 The filter accepts the following options:
10531 Set the luma which will be used as base for transparency.
10532 Default value is @code{0}.
10535 Set the range of luma values to be keyed out.
10536 Default value is @code{0}.
10539 Set the range of softness. Default value is @code{0}.
10540 Use this to control gradual transition from zero to full transparency.
10543 @section lut, lutrgb, lutyuv
10545 Compute a look-up table for binding each pixel component input value
10546 to an output value, and apply it to the input video.
10548 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10549 to an RGB input video.
10551 These filters accept the following parameters:
10554 set first pixel component expression
10556 set second pixel component expression
10558 set third pixel component expression
10560 set fourth pixel component expression, corresponds to the alpha component
10563 set red component expression
10565 set green component expression
10567 set blue component expression
10569 alpha component expression
10572 set Y/luminance component expression
10574 set U/Cb component expression
10576 set V/Cr component expression
10579 Each of them specifies the expression to use for computing the lookup table for
10580 the corresponding pixel component values.
10582 The exact component associated to each of the @var{c*} options depends on the
10585 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10586 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10588 The expressions can contain the following constants and functions:
10593 The input width and height.
10596 The input value for the pixel component.
10599 The input value, clipped to the @var{minval}-@var{maxval} range.
10602 The maximum value for the pixel component.
10605 The minimum value for the pixel component.
10608 The negated value for the pixel component value, clipped to the
10609 @var{minval}-@var{maxval} range; it corresponds to the expression
10610 "maxval-clipval+minval".
10613 The computed value in @var{val}, clipped to the
10614 @var{minval}-@var{maxval} range.
10616 @item gammaval(gamma)
10617 The computed gamma correction value of the pixel component value,
10618 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10620 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10624 All expressions default to "val".
10626 @subsection Examples
10630 Negate input video:
10632 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10633 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10636 The above is the same as:
10638 lutrgb="r=negval:g=negval:b=negval"
10639 lutyuv="y=negval:u=negval:v=negval"
10649 Remove chroma components, turning the video into a graytone image:
10651 lutyuv="u=128:v=128"
10655 Apply a luma burning effect:
10661 Remove green and blue components:
10667 Set a constant alpha channel value on input:
10669 format=rgba,lutrgb=a="maxval-minval/2"
10673 Correct luminance gamma by a factor of 0.5:
10675 lutyuv=y=gammaval(0.5)
10679 Discard least significant bits of luma:
10681 lutyuv=y='bitand(val, 128+64+32)'
10685 Technicolor like effect:
10687 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10691 @section lut2, tlut2
10693 The @code{lut2} filter takes two input streams and outputs one
10696 The @code{tlut2} (time lut2) filter takes two consecutive frames
10697 from one single stream.
10699 This filter accepts the following parameters:
10702 set first pixel component expression
10704 set second pixel component expression
10706 set third pixel component expression
10708 set fourth pixel component expression, corresponds to the alpha component
10711 Each of them specifies the expression to use for computing the lookup table for
10712 the corresponding pixel component values.
10714 The exact component associated to each of the @var{c*} options depends on the
10717 The expressions can contain the following constants:
10722 The input width and height.
10725 The first input value for the pixel component.
10728 The second input value for the pixel component.
10731 The first input video bit depth.
10734 The second input video bit depth.
10737 All expressions default to "x".
10739 @subsection Examples
10743 Highlight differences between two RGB video streams:
10745 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)'
10749 Highlight differences between two YUV video streams:
10751 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)'
10755 Show max difference between two video streams:
10757 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)))'
10761 @section maskedclamp
10763 Clamp the first input stream with the second input and third input stream.
10765 Returns the value of first stream to be between second input
10766 stream - @code{undershoot} and third input stream + @code{overshoot}.
10768 This filter accepts the following options:
10771 Default value is @code{0}.
10774 Default value is @code{0}.
10777 Set which planes will be processed as bitmap, unprocessed planes will be
10778 copied from first stream.
10779 By default value 0xf, all planes will be processed.
10782 @section maskedmerge
10784 Merge the first input stream with the second input stream using per pixel
10785 weights in the third input stream.
10787 A value of 0 in the third stream pixel component means that pixel component
10788 from first stream is returned unchanged, while maximum value (eg. 255 for
10789 8-bit videos) means that pixel component from second stream is returned
10790 unchanged. Intermediate values define the amount of merging between both
10791 input stream's pixel components.
10793 This filter accepts the following options:
10796 Set which planes will be processed as bitmap, unprocessed planes will be
10797 copied from first stream.
10798 By default value 0xf, all planes will be processed.
10803 Apply motion-compensation deinterlacing.
10805 It needs one field per frame as input and must thus be used together
10806 with yadif=1/3 or equivalent.
10808 This filter accepts the following options:
10811 Set the deinterlacing mode.
10813 It accepts one of the following values:
10818 use iterative motion estimation
10820 like @samp{slow}, but use multiple reference frames.
10822 Default value is @samp{fast}.
10825 Set the picture field parity assumed for the input video. It must be
10826 one of the following values:
10830 assume top field first
10832 assume bottom field first
10835 Default value is @samp{bff}.
10838 Set per-block quantization parameter (QP) used by the internal
10841 Higher values should result in a smoother motion vector field but less
10842 optimal individual vectors. Default value is 1.
10845 @section mergeplanes
10847 Merge color channel components from several video streams.
10849 The filter accepts up to 4 input streams, and merge selected input
10850 planes to the output video.
10852 This filter accepts the following options:
10855 Set input to output plane mapping. Default is @code{0}.
10857 The mappings is specified as a bitmap. It should be specified as a
10858 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10859 mapping for the first plane of the output stream. 'A' sets the number of
10860 the input stream to use (from 0 to 3), and 'a' the plane number of the
10861 corresponding input to use (from 0 to 3). The rest of the mappings is
10862 similar, 'Bb' describes the mapping for the output stream second
10863 plane, 'Cc' describes the mapping for the output stream third plane and
10864 'Dd' describes the mapping for the output stream fourth plane.
10867 Set output pixel format. Default is @code{yuva444p}.
10870 @subsection Examples
10874 Merge three gray video streams of same width and height into single video stream:
10876 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10880 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10882 [a0][a1]mergeplanes=0x00010210:yuva444p
10886 Swap Y and A plane in yuva444p stream:
10888 format=yuva444p,mergeplanes=0x03010200:yuva444p
10892 Swap U and V plane in yuv420p stream:
10894 format=yuv420p,mergeplanes=0x000201:yuv420p
10898 Cast a rgb24 clip to yuv444p:
10900 format=rgb24,mergeplanes=0x000102:yuv444p
10906 Estimate and export motion vectors using block matching algorithms.
10907 Motion vectors are stored in frame side data to be used by other filters.
10909 This filter accepts the following options:
10912 Specify the motion estimation method. Accepts one of the following values:
10916 Exhaustive search algorithm.
10918 Three step search algorithm.
10920 Two dimensional logarithmic search algorithm.
10922 New three step search algorithm.
10924 Four step search algorithm.
10926 Diamond search algorithm.
10928 Hexagon-based search algorithm.
10930 Enhanced predictive zonal search algorithm.
10932 Uneven multi-hexagon search algorithm.
10934 Default value is @samp{esa}.
10937 Macroblock size. Default @code{16}.
10940 Search parameter. Default @code{7}.
10943 @section midequalizer
10945 Apply Midway Image Equalization effect using two video streams.
10947 Midway Image Equalization adjusts a pair of images to have the same
10948 histogram, while maintaining their dynamics as much as possible. It's
10949 useful for e.g. matching exposures from a pair of stereo cameras.
10951 This filter has two inputs and one output, which must be of same pixel format, but
10952 may be of different sizes. The output of filter is first input adjusted with
10953 midway histogram of both inputs.
10955 This filter accepts the following option:
10959 Set which planes to process. Default is @code{15}, which is all available planes.
10962 @section minterpolate
10964 Convert the video to specified frame rate using motion interpolation.
10966 This filter accepts the following options:
10969 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}.
10972 Motion interpolation mode. Following values are accepted:
10975 Duplicate previous or next frame for interpolating new ones.
10977 Blend source frames. Interpolated frame is mean of previous and next frames.
10979 Motion compensated interpolation. Following options are effective when this mode is selected:
10983 Motion compensation mode. Following values are accepted:
10986 Overlapped block motion compensation.
10988 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10990 Default mode is @samp{obmc}.
10993 Motion estimation mode. Following values are accepted:
10996 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
10998 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11000 Default mode is @samp{bilat}.
11003 The algorithm to be used for motion estimation. Following values are accepted:
11006 Exhaustive search algorithm.
11008 Three step search algorithm.
11010 Two dimensional logarithmic search algorithm.
11012 New three step search algorithm.
11014 Four step search algorithm.
11016 Diamond search algorithm.
11018 Hexagon-based search algorithm.
11020 Enhanced predictive zonal search algorithm.
11022 Uneven multi-hexagon search algorithm.
11024 Default algorithm is @samp{epzs}.
11027 Macroblock size. Default @code{16}.
11030 Motion estimation search parameter. Default @code{32}.
11033 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).
11038 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:
11041 Disable scene change detection.
11043 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11045 Default method is @samp{fdiff}.
11047 @item scd_threshold
11048 Scene change detection threshold. Default is @code{5.0}.
11053 Mix several video input streams into one video stream.
11055 A description of the accepted options follows.
11059 The number of inputs. If unspecified, it defaults to 2.
11062 Specify weight of each input video stream as sequence.
11063 Each weight is separated by space.
11066 Specify how end of stream is determined.
11069 The duration of the longest input. (default)
11072 The duration of the shortest input.
11075 The duration of the first input.
11079 @section mpdecimate
11081 Drop frames that do not differ greatly from the previous frame in
11082 order to reduce frame rate.
11084 The main use of this filter is for very-low-bitrate encoding
11085 (e.g. streaming over dialup modem), but it could in theory be used for
11086 fixing movies that were inverse-telecined incorrectly.
11088 A description of the accepted options follows.
11092 Set the maximum number of consecutive frames which can be dropped (if
11093 positive), or the minimum interval between dropped frames (if
11094 negative). If the value is 0, the frame is dropped disregarding the
11095 number of previous sequentially dropped frames.
11097 Default value is 0.
11102 Set the dropping threshold values.
11104 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11105 represent actual pixel value differences, so a threshold of 64
11106 corresponds to 1 unit of difference for each pixel, or the same spread
11107 out differently over the block.
11109 A frame is a candidate for dropping if no 8x8 blocks differ by more
11110 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11111 meaning the whole image) differ by more than a threshold of @option{lo}.
11113 Default value for @option{hi} is 64*12, default value for @option{lo} is
11114 64*5, and default value for @option{frac} is 0.33.
11120 Negate input video.
11122 It accepts an integer in input; if non-zero it negates the
11123 alpha component (if available). The default value in input is 0.
11127 Denoise frames using Non-Local Means algorithm.
11129 Each pixel is adjusted by looking for other pixels with similar contexts. This
11130 context similarity is defined by comparing their surrounding patches of size
11131 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11134 Note that the research area defines centers for patches, which means some
11135 patches will be made of pixels outside that research area.
11137 The filter accepts the following options.
11141 Set denoising strength.
11147 Same as @option{p} but for chroma planes.
11149 The default value is @var{0} and means automatic.
11155 Same as @option{r} but for chroma planes.
11157 The default value is @var{0} and means automatic.
11162 Deinterlace video using neural network edge directed interpolation.
11164 This filter accepts the following options:
11168 Mandatory option, without binary file filter can not work.
11169 Currently file can be found here:
11170 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11173 Set which frames to deinterlace, by default it is @code{all}.
11174 Can be @code{all} or @code{interlaced}.
11177 Set mode of operation.
11179 Can be one of the following:
11183 Use frame flags, both fields.
11185 Use frame flags, single field.
11187 Use top field only.
11189 Use bottom field only.
11191 Use both fields, top first.
11193 Use both fields, bottom first.
11197 Set which planes to process, by default filter process all frames.
11200 Set size of local neighborhood around each pixel, used by the predictor neural
11203 Can be one of the following:
11216 Set the number of neurons in predictor neural network.
11217 Can be one of the following:
11228 Controls the number of different neural network predictions that are blended
11229 together to compute the final output value. Can be @code{fast}, default or
11233 Set which set of weights to use in the predictor.
11234 Can be one of the following:
11238 weights trained to minimize absolute error
11240 weights trained to minimize squared error
11244 Controls whether or not the prescreener neural network is used to decide
11245 which pixels should be processed by the predictor neural network and which
11246 can be handled by simple cubic interpolation.
11247 The prescreener is trained to know whether cubic interpolation will be
11248 sufficient for a pixel or whether it should be predicted by the predictor nn.
11249 The computational complexity of the prescreener nn is much less than that of
11250 the predictor nn. Since most pixels can be handled by cubic interpolation,
11251 using the prescreener generally results in much faster processing.
11252 The prescreener is pretty accurate, so the difference between using it and not
11253 using it is almost always unnoticeable.
11255 Can be one of the following:
11263 Default is @code{new}.
11266 Set various debugging flags.
11271 Force libavfilter not to use any of the specified pixel formats for the
11272 input to the next filter.
11274 It accepts the following parameters:
11278 A '|'-separated list of pixel format names, such as
11279 pix_fmts=yuv420p|monow|rgb24".
11283 @subsection Examples
11287 Force libavfilter to use a format different from @var{yuv420p} for the
11288 input to the vflip filter:
11290 noformat=pix_fmts=yuv420p,vflip
11294 Convert the input video to any of the formats not contained in the list:
11296 noformat=yuv420p|yuv444p|yuv410p
11302 Add noise on video input frame.
11304 The filter accepts the following options:
11312 Set noise seed for specific pixel component or all pixel components in case
11313 of @var{all_seed}. Default value is @code{123457}.
11315 @item all_strength, alls
11316 @item c0_strength, c0s
11317 @item c1_strength, c1s
11318 @item c2_strength, c2s
11319 @item c3_strength, c3s
11320 Set noise strength for specific pixel component or all pixel components in case
11321 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11323 @item all_flags, allf
11324 @item c0_flags, c0f
11325 @item c1_flags, c1f
11326 @item c2_flags, c2f
11327 @item c3_flags, c3f
11328 Set pixel component flags or set flags for all components if @var{all_flags}.
11329 Available values for component flags are:
11332 averaged temporal noise (smoother)
11334 mix random noise with a (semi)regular pattern
11336 temporal noise (noise pattern changes between frames)
11338 uniform noise (gaussian otherwise)
11342 @subsection Examples
11344 Add temporal and uniform noise to input video:
11346 noise=alls=20:allf=t+u
11351 Normalize RGB video (aka histogram stretching, contrast stretching).
11352 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11354 For each channel of each frame, the filter computes the input range and maps
11355 it linearly to the user-specified output range. The output range defaults
11356 to the full dynamic range from pure black to pure white.
11358 Temporal smoothing can be used on the input range to reduce flickering (rapid
11359 changes in brightness) caused when small dark or bright objects enter or leave
11360 the scene. This is similar to the auto-exposure (automatic gain control) on a
11361 video camera, and, like a video camera, it may cause a period of over- or
11362 under-exposure of the video.
11364 The R,G,B channels can be normalized independently, which may cause some
11365 color shifting, or linked together as a single channel, which prevents
11366 color shifting. Linked normalization preserves hue. Independent normalization
11367 does not, so it can be used to remove some color casts. Independent and linked
11368 normalization can be combined in any ratio.
11370 The normalize filter accepts the following options:
11375 Colors which define the output range. The minimum input value is mapped to
11376 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11377 The defaults are black and white respectively. Specifying white for
11378 @var{blackpt} and black for @var{whitept} will give color-inverted,
11379 normalized video. Shades of grey can be used to reduce the dynamic range
11380 (contrast). Specifying saturated colors here can create some interesting
11384 The number of previous frames to use for temporal smoothing. The input range
11385 of each channel is smoothed using a rolling average over the current frame
11386 and the @var{smoothing} previous frames. The default is 0 (no temporal
11390 Controls the ratio of independent (color shifting) channel normalization to
11391 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11392 independent. Defaults to 1.0 (fully independent).
11395 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11396 expensive no-op. Defaults to 1.0 (full strength).
11400 @subsection Examples
11402 Stretch video contrast to use the full dynamic range, with no temporal
11403 smoothing; may flicker depending on the source content:
11405 normalize=blackpt=black:whitept=white:smoothing=0
11408 As above, but with 50 frames of temporal smoothing; flicker should be
11409 reduced, depending on the source content:
11411 normalize=blackpt=black:whitept=white:smoothing=50
11414 As above, but with hue-preserving linked channel normalization:
11416 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11419 As above, but with half strength:
11421 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11424 Map the darkest input color to red, the brightest input color to cyan:
11426 normalize=blackpt=red:whitept=cyan
11431 Pass the video source unchanged to the output.
11434 Optical Character Recognition
11436 This filter uses Tesseract for optical character recognition.
11438 It accepts the following options:
11442 Set datapath to tesseract data. Default is to use whatever was
11443 set at installation.
11446 Set language, default is "eng".
11449 Set character whitelist.
11452 Set character blacklist.
11455 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11459 Apply a video transform using libopencv.
11461 To enable this filter, install the libopencv library and headers and
11462 configure FFmpeg with @code{--enable-libopencv}.
11464 It accepts the following parameters:
11469 The name of the libopencv filter to apply.
11471 @item filter_params
11472 The parameters to pass to the libopencv filter. If not specified, the default
11473 values are assumed.
11477 Refer to the official libopencv documentation for more precise
11479 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11481 Several libopencv filters are supported; see the following subsections.
11486 Dilate an image by using a specific structuring element.
11487 It corresponds to the libopencv function @code{cvDilate}.
11489 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11491 @var{struct_el} represents a structuring element, and has the syntax:
11492 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11494 @var{cols} and @var{rows} represent the number of columns and rows of
11495 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11496 point, and @var{shape} the shape for the structuring element. @var{shape}
11497 must be "rect", "cross", "ellipse", or "custom".
11499 If the value for @var{shape} is "custom", it must be followed by a
11500 string of the form "=@var{filename}". The file with name
11501 @var{filename} is assumed to represent a binary image, with each
11502 printable character corresponding to a bright pixel. When a custom
11503 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11504 or columns and rows of the read file are assumed instead.
11506 The default value for @var{struct_el} is "3x3+0x0/rect".
11508 @var{nb_iterations} specifies the number of times the transform is
11509 applied to the image, and defaults to 1.
11513 # Use the default values
11516 # Dilate using a structuring element with a 5x5 cross, iterating two times
11517 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11519 # Read the shape from the file diamond.shape, iterating two times.
11520 # The file diamond.shape may contain a pattern of characters like this
11526 # The specified columns and rows are ignored
11527 # but the anchor point coordinates are not
11528 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11533 Erode an image by using a specific structuring element.
11534 It corresponds to the libopencv function @code{cvErode}.
11536 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11537 with the same syntax and semantics as the @ref{dilate} filter.
11541 Smooth the input video.
11543 The filter takes the following parameters:
11544 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11546 @var{type} is the type of smooth filter to apply, and must be one of
11547 the following values: "blur", "blur_no_scale", "median", "gaussian",
11548 or "bilateral". The default value is "gaussian".
11550 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11551 depend on the smooth type. @var{param1} and
11552 @var{param2} accept integer positive values or 0. @var{param3} and
11553 @var{param4} accept floating point values.
11555 The default value for @var{param1} is 3. The default value for the
11556 other parameters is 0.
11558 These parameters correspond to the parameters assigned to the
11559 libopencv function @code{cvSmooth}.
11561 @section oscilloscope
11563 2D Video Oscilloscope.
11565 Useful to measure spatial impulse, step responses, chroma delays, etc.
11567 It accepts the following parameters:
11571 Set scope center x position.
11574 Set scope center y position.
11577 Set scope size, relative to frame diagonal.
11580 Set scope tilt/rotation.
11586 Set trace center x position.
11589 Set trace center y position.
11592 Set trace width, relative to width of frame.
11595 Set trace height, relative to height of frame.
11598 Set which components to trace. By default it traces first three components.
11601 Draw trace grid. By default is enabled.
11604 Draw some statistics. By default is enabled.
11607 Draw scope. By default is enabled.
11610 @subsection Examples
11614 Inspect full first row of video frame.
11616 oscilloscope=x=0.5:y=0:s=1
11620 Inspect full last row of video frame.
11622 oscilloscope=x=0.5:y=1:s=1
11626 Inspect full 5th line of video frame of height 1080.
11628 oscilloscope=x=0.5:y=5/1080:s=1
11632 Inspect full last column of video frame.
11634 oscilloscope=x=1:y=0.5:s=1:t=1
11642 Overlay one video on top of another.
11644 It takes two inputs and has one output. The first input is the "main"
11645 video on which the second input is overlaid.
11647 It accepts the following parameters:
11649 A description of the accepted options follows.
11654 Set the expression for the x and y coordinates of the overlaid video
11655 on the main video. Default value is "0" for both expressions. In case
11656 the expression is invalid, it is set to a huge value (meaning that the
11657 overlay will not be displayed within the output visible area).
11660 See @ref{framesync}.
11663 Set when the expressions for @option{x}, and @option{y} are evaluated.
11665 It accepts the following values:
11668 only evaluate expressions once during the filter initialization or
11669 when a command is processed
11672 evaluate expressions for each incoming frame
11675 Default value is @samp{frame}.
11678 See @ref{framesync}.
11681 Set the format for the output video.
11683 It accepts the following values:
11686 force YUV420 output
11689 force YUV422 output
11692 force YUV444 output
11695 force packed RGB output
11698 force planar RGB output
11701 automatically pick format
11704 Default value is @samp{yuv420}.
11707 See @ref{framesync}.
11710 Set format of alpha of the overlaid video, it can be @var{straight} or
11711 @var{premultiplied}. Default is @var{straight}.
11714 The @option{x}, and @option{y} expressions can contain the following
11720 The main input width and height.
11724 The overlay input width and height.
11728 The computed values for @var{x} and @var{y}. They are evaluated for
11733 horizontal and vertical chroma subsample values of the output
11734 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11738 the number of input frame, starting from 0
11741 the position in the file of the input frame, NAN if unknown
11744 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11748 This filter also supports the @ref{framesync} options.
11750 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11751 when evaluation is done @emph{per frame}, and will evaluate to NAN
11752 when @option{eval} is set to @samp{init}.
11754 Be aware that frames are taken from each input video in timestamp
11755 order, hence, if their initial timestamps differ, it is a good idea
11756 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11757 have them begin in the same zero timestamp, as the example for
11758 the @var{movie} filter does.
11760 You can chain together more overlays but you should test the
11761 efficiency of such approach.
11763 @subsection Commands
11765 This filter supports the following commands:
11769 Modify the x and y of the overlay input.
11770 The command accepts the same syntax of the corresponding option.
11772 If the specified expression is not valid, it is kept at its current
11776 @subsection Examples
11780 Draw the overlay at 10 pixels from the bottom right corner of the main
11783 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11786 Using named options the example above becomes:
11788 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11792 Insert a transparent PNG logo in the bottom left corner of the input,
11793 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11795 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11799 Insert 2 different transparent PNG logos (second logo on bottom
11800 right corner) using the @command{ffmpeg} tool:
11802 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
11806 Add a transparent color layer on top of the main video; @code{WxH}
11807 must specify the size of the main input to the overlay filter:
11809 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11813 Play an original video and a filtered version (here with the deshake
11814 filter) side by side using the @command{ffplay} tool:
11816 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11819 The above command is the same as:
11821 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11825 Make a sliding overlay appearing from the left to the right top part of the
11826 screen starting since time 2:
11828 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11832 Compose output by putting two input videos side to side:
11834 ffmpeg -i left.avi -i right.avi -filter_complex "
11835 nullsrc=size=200x100 [background];
11836 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11837 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11838 [background][left] overlay=shortest=1 [background+left];
11839 [background+left][right] overlay=shortest=1:x=100 [left+right]
11844 Mask 10-20 seconds of a video by applying the delogo filter to a section
11846 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11847 -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]'
11852 Chain several overlays in cascade:
11854 nullsrc=s=200x200 [bg];
11855 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11856 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11857 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11858 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11859 [in3] null, [mid2] overlay=100:100 [out0]
11866 Apply Overcomplete Wavelet denoiser.
11868 The filter accepts the following options:
11874 Larger depth values will denoise lower frequency components more, but
11875 slow down filtering.
11877 Must be an int in the range 8-16, default is @code{8}.
11879 @item luma_strength, ls
11882 Must be a double value in the range 0-1000, default is @code{1.0}.
11884 @item chroma_strength, cs
11885 Set chroma strength.
11887 Must be a double value in the range 0-1000, default is @code{1.0}.
11893 Add paddings to the input image, and place the original input at the
11894 provided @var{x}, @var{y} coordinates.
11896 It accepts the following parameters:
11901 Specify an expression for the size of the output image with the
11902 paddings added. If the value for @var{width} or @var{height} is 0, the
11903 corresponding input size is used for the output.
11905 The @var{width} expression can reference the value set by the
11906 @var{height} expression, and vice versa.
11908 The default value of @var{width} and @var{height} is 0.
11912 Specify the offsets to place the input image at within the padded area,
11913 with respect to the top/left border of the output image.
11915 The @var{x} expression can reference the value set by the @var{y}
11916 expression, and vice versa.
11918 The default value of @var{x} and @var{y} is 0.
11920 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11921 so the input image is centered on the padded area.
11924 Specify the color of the padded area. For the syntax of this option,
11925 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
11926 manual,ffmpeg-utils}.
11928 The default value of @var{color} is "black".
11931 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11933 It accepts the following values:
11937 Only evaluate expressions once during the filter initialization or when
11938 a command is processed.
11941 Evaluate expressions for each incoming frame.
11945 Default value is @samp{init}.
11948 Pad to aspect instead to a resolution.
11952 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11953 options are expressions containing the following constants:
11958 The input video width and height.
11962 These are the same as @var{in_w} and @var{in_h}.
11966 The output width and height (the size of the padded area), as
11967 specified by the @var{width} and @var{height} expressions.
11971 These are the same as @var{out_w} and @var{out_h}.
11975 The x and y offsets as specified by the @var{x} and @var{y}
11976 expressions, or NAN if not yet specified.
11979 same as @var{iw} / @var{ih}
11982 input sample aspect ratio
11985 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11989 The horizontal and vertical chroma subsample values. For example for the
11990 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11993 @subsection Examples
11997 Add paddings with the color "violet" to the input video. The output video
11998 size is 640x480, and the top-left corner of the input video is placed at
12001 pad=640:480:0:40:violet
12004 The example above is equivalent to the following command:
12006 pad=width=640:height=480:x=0:y=40:color=violet
12010 Pad the input to get an output with dimensions increased by 3/2,
12011 and put the input video at the center of the padded area:
12013 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12017 Pad the input to get a squared output with size equal to the maximum
12018 value between the input width and height, and put the input video at
12019 the center of the padded area:
12021 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12025 Pad the input to get a final w/h ratio of 16:9:
12027 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12031 In case of anamorphic video, in order to set the output display aspect
12032 correctly, it is necessary to use @var{sar} in the expression,
12033 according to the relation:
12035 (ih * X / ih) * sar = output_dar
12036 X = output_dar / sar
12039 Thus the previous example needs to be modified to:
12041 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12045 Double the output size and put the input video in the bottom-right
12046 corner of the output padded area:
12048 pad="2*iw:2*ih:ow-iw:oh-ih"
12052 @anchor{palettegen}
12053 @section palettegen
12055 Generate one palette for a whole video stream.
12057 It accepts the following options:
12061 Set the maximum number of colors to quantize in the palette.
12062 Note: the palette will still contain 256 colors; the unused palette entries
12065 @item reserve_transparent
12066 Create a palette of 255 colors maximum and reserve the last one for
12067 transparency. Reserving the transparency color is useful for GIF optimization.
12068 If not set, the maximum of colors in the palette will be 256. You probably want
12069 to disable this option for a standalone image.
12072 @item transparency_color
12073 Set the color that will be used as background for transparency.
12076 Set statistics mode.
12078 It accepts the following values:
12081 Compute full frame histograms.
12083 Compute histograms only for the part that differs from previous frame. This
12084 might be relevant to give more importance to the moving part of your input if
12085 the background is static.
12087 Compute new histogram for each frame.
12090 Default value is @var{full}.
12093 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12094 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12095 color quantization of the palette. This information is also visible at
12096 @var{info} logging level.
12098 @subsection Examples
12102 Generate a representative palette of a given video using @command{ffmpeg}:
12104 ffmpeg -i input.mkv -vf palettegen palette.png
12108 @section paletteuse
12110 Use a palette to downsample an input video stream.
12112 The filter takes two inputs: one video stream and a palette. The palette must
12113 be a 256 pixels image.
12115 It accepts the following options:
12119 Select dithering mode. Available algorithms are:
12122 Ordered 8x8 bayer dithering (deterministic)
12124 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12125 Note: this dithering is sometimes considered "wrong" and is included as a
12127 @item floyd_steinberg
12128 Floyd and Steingberg dithering (error diffusion)
12130 Frankie Sierra dithering v2 (error diffusion)
12132 Frankie Sierra dithering v2 "Lite" (error diffusion)
12135 Default is @var{sierra2_4a}.
12138 When @var{bayer} dithering is selected, this option defines the scale of the
12139 pattern (how much the crosshatch pattern is visible). A low value means more
12140 visible pattern for less banding, and higher value means less visible pattern
12141 at the cost of more banding.
12143 The option must be an integer value in the range [0,5]. Default is @var{2}.
12146 If set, define the zone to process
12150 Only the changing rectangle will be reprocessed. This is similar to GIF
12151 cropping/offsetting compression mechanism. This option can be useful for speed
12152 if only a part of the image is changing, and has use cases such as limiting the
12153 scope of the error diffusal @option{dither} to the rectangle that bounds the
12154 moving scene (it leads to more deterministic output if the scene doesn't change
12155 much, and as a result less moving noise and better GIF compression).
12158 Default is @var{none}.
12161 Take new palette for each output frame.
12163 @item alpha_threshold
12164 Sets the alpha threshold for transparency. Alpha values above this threshold
12165 will be treated as completely opaque, and values below this threshold will be
12166 treated as completely transparent.
12168 The option must be an integer value in the range [0,255]. Default is @var{128}.
12171 @subsection Examples
12175 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12176 using @command{ffmpeg}:
12178 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12182 @section perspective
12184 Correct perspective of video not recorded perpendicular to the screen.
12186 A description of the accepted parameters follows.
12197 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12198 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12199 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12200 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12201 then the corners of the source will be sent to the specified coordinates.
12203 The expressions can use the following variables:
12208 the width and height of video frame.
12212 Output frame count.
12215 @item interpolation
12216 Set interpolation for perspective correction.
12218 It accepts the following values:
12224 Default value is @samp{linear}.
12227 Set interpretation of coordinate options.
12229 It accepts the following values:
12233 Send point in the source specified by the given coordinates to
12234 the corners of the destination.
12236 @item 1, destination
12238 Send the corners of the source to the point in the destination specified
12239 by the given coordinates.
12241 Default value is @samp{source}.
12245 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12247 It accepts the following values:
12250 only evaluate expressions once during the filter initialization or
12251 when a command is processed
12254 evaluate expressions for each incoming frame
12257 Default value is @samp{init}.
12262 Delay interlaced video by one field time so that the field order changes.
12264 The intended use is to fix PAL movies that have been captured with the
12265 opposite field order to the film-to-video transfer.
12267 A description of the accepted parameters follows.
12273 It accepts the following values:
12276 Capture field order top-first, transfer bottom-first.
12277 Filter will delay the bottom field.
12280 Capture field order bottom-first, transfer top-first.
12281 Filter will delay the top field.
12284 Capture and transfer with the same field order. This mode only exists
12285 for the documentation of the other options to refer to, but if you
12286 actually select it, the filter will faithfully do nothing.
12289 Capture field order determined automatically by field flags, transfer
12291 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12292 basis using field flags. If no field information is available,
12293 then this works just like @samp{u}.
12296 Capture unknown or varying, transfer opposite.
12297 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12298 analyzing the images and selecting the alternative that produces best
12299 match between the fields.
12302 Capture top-first, transfer unknown or varying.
12303 Filter selects among @samp{t} and @samp{p} using image analysis.
12306 Capture bottom-first, transfer unknown or varying.
12307 Filter selects among @samp{b} and @samp{p} using image analysis.
12310 Capture determined by field flags, transfer unknown or varying.
12311 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12312 image analysis. If no field information is available, then this works just
12313 like @samp{U}. This is the default mode.
12316 Both capture and transfer unknown or varying.
12317 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12321 @section pixdesctest
12323 Pixel format descriptor test filter, mainly useful for internal
12324 testing. The output video should be equal to the input video.
12328 format=monow, pixdesctest
12331 can be used to test the monowhite pixel format descriptor definition.
12335 Display sample values of color channels. Mainly useful for checking color
12336 and levels. Minimum supported resolution is 640x480.
12338 The filters accept the following options:
12342 Set scope X position, relative offset on X axis.
12345 Set scope Y position, relative offset on Y axis.
12354 Set window opacity. This window also holds statistics about pixel area.
12357 Set window X position, relative offset on X axis.
12360 Set window Y position, relative offset on Y axis.
12365 Enable the specified chain of postprocessing subfilters using libpostproc. This
12366 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12367 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12368 Each subfilter and some options have a short and a long name that can be used
12369 interchangeably, i.e. dr/dering are the same.
12371 The filters accept the following options:
12375 Set postprocessing subfilters string.
12378 All subfilters share common options to determine their scope:
12382 Honor the quality commands for this subfilter.
12385 Do chrominance filtering, too (default).
12388 Do luminance filtering only (no chrominance).
12391 Do chrominance filtering only (no luminance).
12394 These options can be appended after the subfilter name, separated by a '|'.
12396 Available subfilters are:
12399 @item hb/hdeblock[|difference[|flatness]]
12400 Horizontal deblocking filter
12403 Difference factor where higher values mean more deblocking (default: @code{32}).
12405 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12408 @item vb/vdeblock[|difference[|flatness]]
12409 Vertical deblocking filter
12412 Difference factor where higher values mean more deblocking (default: @code{32}).
12414 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12417 @item ha/hadeblock[|difference[|flatness]]
12418 Accurate horizontal deblocking filter
12421 Difference factor where higher values mean more deblocking (default: @code{32}).
12423 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12426 @item va/vadeblock[|difference[|flatness]]
12427 Accurate vertical deblocking filter
12430 Difference factor where higher values mean more deblocking (default: @code{32}).
12432 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12436 The horizontal and vertical deblocking filters share the difference and
12437 flatness values so you cannot set different horizontal and vertical
12441 @item h1/x1hdeblock
12442 Experimental horizontal deblocking filter
12444 @item v1/x1vdeblock
12445 Experimental vertical deblocking filter
12450 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12453 larger -> stronger filtering
12455 larger -> stronger filtering
12457 larger -> stronger filtering
12460 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12463 Stretch luminance to @code{0-255}.
12466 @item lb/linblenddeint
12467 Linear blend deinterlacing filter that deinterlaces the given block by
12468 filtering all lines with a @code{(1 2 1)} filter.
12470 @item li/linipoldeint
12471 Linear interpolating deinterlacing filter that deinterlaces the given block by
12472 linearly interpolating every second line.
12474 @item ci/cubicipoldeint
12475 Cubic interpolating deinterlacing filter deinterlaces the given block by
12476 cubically interpolating every second line.
12478 @item md/mediandeint
12479 Median deinterlacing filter that deinterlaces the given block by applying a
12480 median filter to every second line.
12482 @item fd/ffmpegdeint
12483 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12484 second line with a @code{(-1 4 2 4 -1)} filter.
12487 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12488 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12490 @item fq/forceQuant[|quantizer]
12491 Overrides the quantizer table from the input with the constant quantizer you
12499 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12502 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12505 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12508 @subsection Examples
12512 Apply horizontal and vertical deblocking, deringing and automatic
12513 brightness/contrast:
12519 Apply default filters without brightness/contrast correction:
12525 Apply default filters and temporal denoiser:
12527 pp=default/tmpnoise|1|2|3
12531 Apply deblocking on luminance only, and switch vertical deblocking on or off
12532 automatically depending on available CPU time:
12539 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12540 similar to spp = 6 with 7 point DCT, where only the center sample is
12543 The filter accepts the following options:
12547 Force a constant quantization parameter. It accepts an integer in range
12548 0 to 63. If not set, the filter will use the QP from the video stream
12552 Set thresholding mode. Available modes are:
12556 Set hard thresholding.
12558 Set soft thresholding (better de-ringing effect, but likely blurrier).
12560 Set medium thresholding (good results, default).
12564 @section premultiply
12565 Apply alpha premultiply effect to input video stream using first plane
12566 of second stream as alpha.
12568 Both streams must have same dimensions and same pixel format.
12570 The filter accepts the following option:
12574 Set which planes will be processed, unprocessed planes will be copied.
12575 By default value 0xf, all planes will be processed.
12578 Do not require 2nd input for processing, instead use alpha plane from input stream.
12582 Apply prewitt operator to input video stream.
12584 The filter accepts the following option:
12588 Set which planes will be processed, unprocessed planes will be copied.
12589 By default value 0xf, all planes will be processed.
12592 Set value which will be multiplied with filtered result.
12595 Set value which will be added to filtered result.
12598 @anchor{program_opencl}
12599 @section program_opencl
12601 Filter video using an OpenCL program.
12606 OpenCL program source file.
12609 Kernel name in program.
12612 Number of inputs to the filter. Defaults to 1.
12615 Size of output frames. Defaults to the same as the first input.
12619 The program source file must contain a kernel function with the given name,
12620 which will be run once for each plane of the output. Each run on a plane
12621 gets enqueued as a separate 2D global NDRange with one work-item for each
12622 pixel to be generated. The global ID offset for each work-item is therefore
12623 the coordinates of a pixel in the destination image.
12625 The kernel function needs to take the following arguments:
12628 Destination image, @var{__write_only image2d_t}.
12630 This image will become the output; the kernel should write all of it.
12632 Frame index, @var{unsigned int}.
12634 This is a counter starting from zero and increasing by one for each frame.
12636 Source images, @var{__read_only image2d_t}.
12638 These are the most recent images on each input. The kernel may read from
12639 them to generate the output, but they can't be written to.
12646 Copy the input to the output (output must be the same size as the input).
12648 __kernel void copy(__write_only image2d_t destination,
12649 unsigned int index,
12650 __read_only image2d_t source)
12652 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
12654 int2 location = (int2)(get_global_id(0), get_global_id(1));
12656 float4 value = read_imagef(source, sampler, location);
12658 write_imagef(destination, location, value);
12663 Apply a simple transformation, rotating the input by an amount increasing
12664 with the index counter. Pixel values are linearly interpolated by the
12665 sampler, and the output need not have the same dimensions as the input.
12667 __kernel void rotate_image(__write_only image2d_t dst,
12668 unsigned int index,
12669 __read_only image2d_t src)
12671 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12672 CLK_FILTER_LINEAR);
12674 float angle = (float)index / 100.0f;
12676 float2 dst_dim = convert_float2(get_image_dim(dst));
12677 float2 src_dim = convert_float2(get_image_dim(src));
12679 float2 dst_cen = dst_dim / 2.0f;
12680 float2 src_cen = src_dim / 2.0f;
12682 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12684 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
12686 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
12687 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
12689 src_pos = src_pos * src_dim / dst_dim;
12691 float2 src_loc = src_pos + src_cen;
12693 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
12694 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
12695 write_imagef(dst, dst_loc, 0.5f);
12697 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
12702 Blend two inputs together, with the amount of each input used varying
12703 with the index counter.
12705 __kernel void blend_images(__write_only image2d_t dst,
12706 unsigned int index,
12707 __read_only image2d_t src1,
12708 __read_only image2d_t src2)
12710 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12711 CLK_FILTER_LINEAR);
12713 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
12715 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12716 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
12717 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
12719 float4 val1 = read_imagef(src1, sampler, src1_loc);
12720 float4 val2 = read_imagef(src2, sampler, src2_loc);
12722 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
12728 @section pseudocolor
12730 Alter frame colors in video with pseudocolors.
12732 This filter accept the following options:
12736 set pixel first component expression
12739 set pixel second component expression
12742 set pixel third component expression
12745 set pixel fourth component expression, corresponds to the alpha component
12748 set component to use as base for altering colors
12751 Each of them specifies the expression to use for computing the lookup table for
12752 the corresponding pixel component values.
12754 The expressions can contain the following constants and functions:
12759 The input width and height.
12762 The input value for the pixel component.
12764 @item ymin, umin, vmin, amin
12765 The minimum allowed component value.
12767 @item ymax, umax, vmax, amax
12768 The maximum allowed component value.
12771 All expressions default to "val".
12773 @subsection Examples
12777 Change too high luma values to gradient:
12779 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'"
12785 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12786 Ratio) between two input videos.
12788 This filter takes in input two input videos, the first input is
12789 considered the "main" source and is passed unchanged to the
12790 output. The second input is used as a "reference" video for computing
12793 Both video inputs must have the same resolution and pixel format for
12794 this filter to work correctly. Also it assumes that both inputs
12795 have the same number of frames, which are compared one by one.
12797 The obtained average PSNR is printed through the logging system.
12799 The filter stores the accumulated MSE (mean squared error) of each
12800 frame, and at the end of the processing it is averaged across all frames
12801 equally, and the following formula is applied to obtain the PSNR:
12804 PSNR = 10*log10(MAX^2/MSE)
12807 Where MAX is the average of the maximum values of each component of the
12810 The description of the accepted parameters follows.
12813 @item stats_file, f
12814 If specified the filter will use the named file to save the PSNR of
12815 each individual frame. When filename equals "-" the data is sent to
12818 @item stats_version
12819 Specifies which version of the stats file format to use. Details of
12820 each format are written below.
12821 Default value is 1.
12823 @item stats_add_max
12824 Determines whether the max value is output to the stats log.
12825 Default value is 0.
12826 Requires stats_version >= 2. If this is set and stats_version < 2,
12827 the filter will return an error.
12830 This filter also supports the @ref{framesync} options.
12832 The file printed if @var{stats_file} is selected, contains a sequence of
12833 key/value pairs of the form @var{key}:@var{value} for each compared
12836 If a @var{stats_version} greater than 1 is specified, a header line precedes
12837 the list of per-frame-pair stats, with key value pairs following the frame
12838 format with the following parameters:
12841 @item psnr_log_version
12842 The version of the log file format. Will match @var{stats_version}.
12845 A comma separated list of the per-frame-pair parameters included in
12849 A description of each shown per-frame-pair parameter follows:
12853 sequential number of the input frame, starting from 1
12856 Mean Square Error pixel-by-pixel average difference of the compared
12857 frames, averaged over all the image components.
12859 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
12860 Mean Square Error pixel-by-pixel average difference of the compared
12861 frames for the component specified by the suffix.
12863 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12864 Peak Signal to Noise ratio of the compared frames for the component
12865 specified by the suffix.
12867 @item max_avg, max_y, max_u, max_v
12868 Maximum allowed value for each channel, and average over all
12874 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12875 [main][ref] psnr="stats_file=stats.log" [out]
12878 On this example the input file being processed is compared with the
12879 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12880 is stored in @file{stats.log}.
12885 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12886 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12889 The pullup filter is designed to take advantage of future context in making
12890 its decisions. This filter is stateless in the sense that it does not lock
12891 onto a pattern to follow, but it instead looks forward to the following
12892 fields in order to identify matches and rebuild progressive frames.
12894 To produce content with an even framerate, insert the fps filter after
12895 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12896 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12898 The filter accepts the following options:
12905 These options set the amount of "junk" to ignore at the left, right, top, and
12906 bottom of the image, respectively. Left and right are in units of 8 pixels,
12907 while top and bottom are in units of 2 lines.
12908 The default is 8 pixels on each side.
12911 Set the strict breaks. Setting this option to 1 will reduce the chances of
12912 filter generating an occasional mismatched frame, but it may also cause an
12913 excessive number of frames to be dropped during high motion sequences.
12914 Conversely, setting it to -1 will make filter match fields more easily.
12915 This may help processing of video where there is slight blurring between
12916 the fields, but may also cause there to be interlaced frames in the output.
12917 Default value is @code{0}.
12920 Set the metric plane to use. It accepts the following values:
12926 Use chroma blue plane.
12929 Use chroma red plane.
12932 This option may be set to use chroma plane instead of the default luma plane
12933 for doing filter's computations. This may improve accuracy on very clean
12934 source material, but more likely will decrease accuracy, especially if there
12935 is chroma noise (rainbow effect) or any grayscale video.
12936 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12937 load and make pullup usable in realtime on slow machines.
12940 For best results (without duplicated frames in the output file) it is
12941 necessary to change the output frame rate. For example, to inverse
12942 telecine NTSC input:
12944 ffmpeg -i input -vf pullup -r 24000/1001 ...
12949 Change video quantization parameters (QP).
12951 The filter accepts the following option:
12955 Set expression for quantization parameter.
12958 The expression is evaluated through the eval API and can contain, among others,
12959 the following constants:
12963 1 if index is not 129, 0 otherwise.
12966 Sequential index starting from -129 to 128.
12969 @subsection Examples
12973 Some equation like:
12981 Flush video frames from internal cache of frames into a random order.
12982 No frame is discarded.
12983 Inspired by @ref{frei0r} nervous filter.
12987 Set size in number of frames of internal cache, in range from @code{2} to
12988 @code{512}. Default is @code{30}.
12991 Set seed for random number generator, must be an integer included between
12992 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12993 less than @code{0}, the filter will try to use a good random seed on a
12997 @section readeia608
12999 Read closed captioning (EIA-608) information from the top lines of a video frame.
13001 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13002 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13003 with EIA-608 data (starting from 0). A description of each metadata value follows:
13006 @item lavfi.readeia608.X.cc
13007 The two bytes stored as EIA-608 data (printed in hexadecimal).
13009 @item lavfi.readeia608.X.line
13010 The number of the line on which the EIA-608 data was identified and read.
13013 This filter accepts the following options:
13017 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13020 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13023 Set minimal acceptable amplitude change for sync codes detection.
13024 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13027 Set the ratio of width reserved for sync code detection.
13028 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13031 Set the max peaks height difference for sync code detection.
13032 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13035 Set max peaks period difference for sync code detection.
13036 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13039 Set the first two max start code bits differences.
13040 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13043 Set the minimum ratio of bits height compared to 3rd start code bit.
13044 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13047 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13050 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13053 Enable checking the parity bit. In the event of a parity error, the filter will output
13054 @code{0x00} for that character. Default is false.
13057 @subsection Examples
13061 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13063 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
13069 Read vertical interval timecode (VITC) information from the top lines of a
13072 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13073 timecode value, if a valid timecode has been detected. Further metadata key
13074 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13075 timecode data has been found or not.
13077 This filter accepts the following options:
13081 Set the maximum number of lines to scan for VITC data. If the value is set to
13082 @code{-1} the full video frame is scanned. Default is @code{45}.
13085 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13086 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13089 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13090 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13093 @subsection Examples
13097 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13098 draw @code{--:--:--:--} as a placeholder:
13100 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13106 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13108 Destination pixel at position (X, Y) will be picked from source (x, y) position
13109 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13110 value for pixel will be used for destination pixel.
13112 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13113 will have Xmap/Ymap video stream dimensions.
13114 Xmap and Ymap input video streams are 16bit depth, single channel.
13116 @section removegrain
13118 The removegrain filter is a spatial denoiser for progressive video.
13122 Set mode for the first plane.
13125 Set mode for the second plane.
13128 Set mode for the third plane.
13131 Set mode for the fourth plane.
13134 Range of mode is from 0 to 24. Description of each mode follows:
13138 Leave input plane unchanged. Default.
13141 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13144 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13147 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13150 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13151 This is equivalent to a median filter.
13154 Line-sensitive clipping giving the minimal change.
13157 Line-sensitive clipping, intermediate.
13160 Line-sensitive clipping, intermediate.
13163 Line-sensitive clipping, intermediate.
13166 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13169 Replaces the target pixel with the closest neighbour.
13172 [1 2 1] horizontal and vertical kernel blur.
13178 Bob mode, interpolates top field from the line where the neighbours
13179 pixels are the closest.
13182 Bob mode, interpolates bottom field from the line where the neighbours
13183 pixels are the closest.
13186 Bob mode, interpolates top field. Same as 13 but with a more complicated
13187 interpolation formula.
13190 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13191 interpolation formula.
13194 Clips the pixel with the minimum and maximum of respectively the maximum and
13195 minimum of each pair of opposite neighbour pixels.
13198 Line-sensitive clipping using opposite neighbours whose greatest distance from
13199 the current pixel is minimal.
13202 Replaces the pixel with the average of its 8 neighbours.
13205 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13208 Clips pixels using the averages of opposite neighbour.
13211 Same as mode 21 but simpler and faster.
13214 Small edge and halo removal, but reputed useless.
13220 @section removelogo
13222 Suppress a TV station logo, using an image file to determine which
13223 pixels comprise the logo. It works by filling in the pixels that
13224 comprise the logo with neighboring pixels.
13226 The filter accepts the following options:
13230 Set the filter bitmap file, which can be any image format supported by
13231 libavformat. The width and height of the image file must match those of the
13232 video stream being processed.
13235 Pixels in the provided bitmap image with a value of zero are not
13236 considered part of the logo, non-zero pixels are considered part of
13237 the logo. If you use white (255) for the logo and black (0) for the
13238 rest, you will be safe. For making the filter bitmap, it is
13239 recommended to take a screen capture of a black frame with the logo
13240 visible, and then using a threshold filter followed by the erode
13241 filter once or twice.
13243 If needed, little splotches can be fixed manually. Remember that if
13244 logo pixels are not covered, the filter quality will be much
13245 reduced. Marking too many pixels as part of the logo does not hurt as
13246 much, but it will increase the amount of blurring needed to cover over
13247 the image and will destroy more information than necessary, and extra
13248 pixels will slow things down on a large logo.
13250 @section repeatfields
13252 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13253 fields based on its value.
13257 Reverse a video clip.
13259 Warning: This filter requires memory to buffer the entire clip, so trimming
13262 @subsection Examples
13266 Take the first 5 seconds of a clip, and reverse it.
13273 Apply roberts cross operator to input video stream.
13275 The filter accepts the following option:
13279 Set which planes will be processed, unprocessed planes will be copied.
13280 By default value 0xf, all planes will be processed.
13283 Set value which will be multiplied with filtered result.
13286 Set value which will be added to filtered result.
13291 Rotate video by an arbitrary angle expressed in radians.
13293 The filter accepts the following options:
13295 A description of the optional parameters follows.
13298 Set an expression for the angle by which to rotate the input video
13299 clockwise, expressed as a number of radians. A negative value will
13300 result in a counter-clockwise rotation. By default it is set to "0".
13302 This expression is evaluated for each frame.
13305 Set the output width expression, default value is "iw".
13306 This expression is evaluated just once during configuration.
13309 Set the output height expression, default value is "ih".
13310 This expression is evaluated just once during configuration.
13313 Enable bilinear interpolation if set to 1, a value of 0 disables
13314 it. Default value is 1.
13317 Set the color used to fill the output area not covered by the rotated
13318 image. For the general syntax of this option, check the
13319 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13320 If the special value "none" is selected then no
13321 background is printed (useful for example if the background is never shown).
13323 Default value is "black".
13326 The expressions for the angle and the output size can contain the
13327 following constants and functions:
13331 sequential number of the input frame, starting from 0. It is always NAN
13332 before the first frame is filtered.
13335 time in seconds of the input frame, it is set to 0 when the filter is
13336 configured. It is always NAN before the first frame is filtered.
13340 horizontal and vertical chroma subsample values. For example for the
13341 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13345 the input video width and height
13349 the output width and height, that is the size of the padded area as
13350 specified by the @var{width} and @var{height} expressions
13354 the minimal width/height required for completely containing the input
13355 video rotated by @var{a} radians.
13357 These are only available when computing the @option{out_w} and
13358 @option{out_h} expressions.
13361 @subsection Examples
13365 Rotate the input by PI/6 radians clockwise:
13371 Rotate the input by PI/6 radians counter-clockwise:
13377 Rotate the input by 45 degrees clockwise:
13383 Apply a constant rotation with period T, starting from an angle of PI/3:
13385 rotate=PI/3+2*PI*t/T
13389 Make the input video rotation oscillating with a period of T
13390 seconds and an amplitude of A radians:
13392 rotate=A*sin(2*PI/T*t)
13396 Rotate the video, output size is chosen so that the whole rotating
13397 input video is always completely contained in the output:
13399 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13403 Rotate the video, reduce the output size so that no background is ever
13406 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13410 @subsection Commands
13412 The filter supports the following commands:
13416 Set the angle expression.
13417 The command accepts the same syntax of the corresponding option.
13419 If the specified expression is not valid, it is kept at its current
13425 Apply Shape Adaptive Blur.
13427 The filter accepts the following options:
13430 @item luma_radius, lr
13431 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13432 value is 1.0. A greater value will result in a more blurred image, and
13433 in slower processing.
13435 @item luma_pre_filter_radius, lpfr
13436 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13439 @item luma_strength, ls
13440 Set luma maximum difference between pixels to still be considered, must
13441 be a value in the 0.1-100.0 range, default value is 1.0.
13443 @item chroma_radius, cr
13444 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13445 greater value will result in a more blurred image, and in slower
13448 @item chroma_pre_filter_radius, cpfr
13449 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13451 @item chroma_strength, cs
13452 Set chroma maximum difference between pixels to still be considered,
13453 must be a value in the -0.9-100.0 range.
13456 Each chroma option value, if not explicitly specified, is set to the
13457 corresponding luma option value.
13462 Scale (resize) the input video, using the libswscale library.
13464 The scale filter forces the output display aspect ratio to be the same
13465 of the input, by changing the output sample aspect ratio.
13467 If the input image format is different from the format requested by
13468 the next filter, the scale filter will convert the input to the
13471 @subsection Options
13472 The filter accepts the following options, or any of the options
13473 supported by the libswscale scaler.
13475 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13476 the complete list of scaler options.
13481 Set the output video dimension expression. Default value is the input
13484 If the @var{width} or @var{w} value is 0, the input width is used for
13485 the output. If the @var{height} or @var{h} value is 0, the input height
13486 is used for the output.
13488 If one and only one of the values is -n with n >= 1, the scale filter
13489 will use a value that maintains the aspect ratio of the input image,
13490 calculated from the other specified dimension. After that it will,
13491 however, make sure that the calculated dimension is divisible by n and
13492 adjust the value if necessary.
13494 If both values are -n with n >= 1, the behavior will be identical to
13495 both values being set to 0 as previously detailed.
13497 See below for the list of accepted constants for use in the dimension
13501 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
13505 Only evaluate expressions once during the filter initialization or when a command is processed.
13508 Evaluate expressions for each incoming frame.
13512 Default value is @samp{init}.
13516 Set the interlacing mode. It accepts the following values:
13520 Force interlaced aware scaling.
13523 Do not apply interlaced scaling.
13526 Select interlaced aware scaling depending on whether the source frames
13527 are flagged as interlaced or not.
13530 Default value is @samp{0}.
13533 Set libswscale scaling flags. See
13534 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13535 complete list of values. If not explicitly specified the filter applies
13539 @item param0, param1
13540 Set libswscale input parameters for scaling algorithms that need them. See
13541 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13542 complete documentation. If not explicitly specified the filter applies
13548 Set the video size. For the syntax of this option, check the
13549 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13551 @item in_color_matrix
13552 @item out_color_matrix
13553 Set in/output YCbCr color space type.
13555 This allows the autodetected value to be overridden as well as allows forcing
13556 a specific value used for the output and encoder.
13558 If not specified, the color space type depends on the pixel format.
13564 Choose automatically.
13567 Format conforming to International Telecommunication Union (ITU)
13568 Recommendation BT.709.
13571 Set color space conforming to the United States Federal Communications
13572 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
13575 Set color space conforming to:
13579 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
13582 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
13585 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
13590 Set color space conforming to SMPTE ST 240:1999.
13595 Set in/output YCbCr sample range.
13597 This allows the autodetected value to be overridden as well as allows forcing
13598 a specific value used for the output and encoder. If not specified, the
13599 range depends on the pixel format. Possible values:
13603 Choose automatically.
13606 Set full range (0-255 in case of 8-bit luma).
13608 @item mpeg/limited/tv
13609 Set "MPEG" range (16-235 in case of 8-bit luma).
13612 @item force_original_aspect_ratio
13613 Enable decreasing or increasing output video width or height if necessary to
13614 keep the original aspect ratio. Possible values:
13618 Scale the video as specified and disable this feature.
13621 The output video dimensions will automatically be decreased if needed.
13624 The output video dimensions will automatically be increased if needed.
13628 One useful instance of this option is that when you know a specific device's
13629 maximum allowed resolution, you can use this to limit the output video to
13630 that, while retaining the aspect ratio. For example, device A allows
13631 1280x720 playback, and your video is 1920x800. Using this option (set it to
13632 decrease) and specifying 1280x720 to the command line makes the output
13635 Please note that this is a different thing than specifying -1 for @option{w}
13636 or @option{h}, you still need to specify the output resolution for this option
13641 The values of the @option{w} and @option{h} options are expressions
13642 containing the following constants:
13647 The input width and height
13651 These are the same as @var{in_w} and @var{in_h}.
13655 The output (scaled) width and height
13659 These are the same as @var{out_w} and @var{out_h}
13662 The same as @var{iw} / @var{ih}
13665 input sample aspect ratio
13668 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13672 horizontal and vertical input chroma subsample values. For example for the
13673 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13677 horizontal and vertical output chroma subsample values. For example for the
13678 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13681 @subsection Examples
13685 Scale the input video to a size of 200x100
13690 This is equivalent to:
13701 Specify a size abbreviation for the output size:
13706 which can also be written as:
13712 Scale the input to 2x:
13714 scale=w=2*iw:h=2*ih
13718 The above is the same as:
13720 scale=2*in_w:2*in_h
13724 Scale the input to 2x with forced interlaced scaling:
13726 scale=2*iw:2*ih:interl=1
13730 Scale the input to half size:
13732 scale=w=iw/2:h=ih/2
13736 Increase the width, and set the height to the same size:
13742 Seek Greek harmony:
13749 Increase the height, and set the width to 3/2 of the height:
13751 scale=w=3/2*oh:h=3/5*ih
13755 Increase the size, making the size a multiple of the chroma
13758 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13762 Increase the width to a maximum of 500 pixels,
13763 keeping the same aspect ratio as the input:
13765 scale=w='min(500\, iw*3/2):h=-1'
13769 Make pixels square by combining scale and setsar:
13771 scale='trunc(ih*dar):ih',setsar=1/1
13775 Make pixels square by combining scale and setsar,
13776 making sure the resulting resolution is even (required by some codecs):
13778 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
13782 @subsection Commands
13784 This filter supports the following commands:
13788 Set the output video dimension expression.
13789 The command accepts the same syntax of the corresponding option.
13791 If the specified expression is not valid, it is kept at its current
13797 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13798 format conversion on CUDA video frames. Setting the output width and height
13799 works in the same way as for the @var{scale} filter.
13801 The following additional options are accepted:
13804 The pixel format of the output CUDA frames. If set to the string "same" (the
13805 default), the input format will be kept. Note that automatic format negotiation
13806 and conversion is not yet supported for hardware frames
13809 The interpolation algorithm used for resizing. One of the following:
13816 @item cubic2p_bspline
13817 2-parameter cubic (B=1, C=0)
13819 @item cubic2p_catmullrom
13820 2-parameter cubic (B=0, C=1/2)
13822 @item cubic2p_b05c03
13823 2-parameter cubic (B=1/2, C=3/10)
13835 Scale (resize) the input video, based on a reference video.
13837 See the scale filter for available options, scale2ref supports the same but
13838 uses the reference video instead of the main input as basis. scale2ref also
13839 supports the following additional constants for the @option{w} and
13840 @option{h} options:
13845 The main input video's width and height
13848 The same as @var{main_w} / @var{main_h}
13851 The main input video's sample aspect ratio
13853 @item main_dar, mdar
13854 The main input video's display aspect ratio. Calculated from
13855 @code{(main_w / main_h) * main_sar}.
13859 The main input video's horizontal and vertical chroma subsample values.
13860 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13864 @subsection Examples
13868 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13870 'scale2ref[b][a];[a][b]overlay'
13874 @anchor{selectivecolor}
13875 @section selectivecolor
13877 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13878 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13879 by the "purity" of the color (that is, how saturated it already is).
13881 This filter is similar to the Adobe Photoshop Selective Color tool.
13883 The filter accepts the following options:
13886 @item correction_method
13887 Select color correction method.
13889 Available values are:
13892 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13895 Specified adjustments are relative to the original component value.
13897 Default is @code{absolute}.
13899 Adjustments for red pixels (pixels where the red component is the maximum)
13901 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13903 Adjustments for green pixels (pixels where the green component is the maximum)
13905 Adjustments for cyan pixels (pixels where the red component is the minimum)
13907 Adjustments for blue pixels (pixels where the blue component is the maximum)
13909 Adjustments for magenta pixels (pixels where the green component is the minimum)
13911 Adjustments for white pixels (pixels where all components are greater than 128)
13913 Adjustments for all pixels except pure black and pure white
13915 Adjustments for black pixels (pixels where all components are lesser than 128)
13917 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13920 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13921 4 space separated floating point adjustment values in the [-1,1] range,
13922 respectively to adjust the amount of cyan, magenta, yellow and black for the
13923 pixels of its range.
13925 @subsection Examples
13929 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13930 increase magenta by 27% in blue areas:
13932 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13936 Use a Photoshop selective color preset:
13938 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13942 @anchor{separatefields}
13943 @section separatefields
13945 The @code{separatefields} takes a frame-based video input and splits
13946 each frame into its components fields, producing a new half height clip
13947 with twice the frame rate and twice the frame count.
13949 This filter use field-dominance information in frame to decide which
13950 of each pair of fields to place first in the output.
13951 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13953 @section setdar, setsar
13955 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13958 This is done by changing the specified Sample (aka Pixel) Aspect
13959 Ratio, according to the following equation:
13961 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13964 Keep in mind that the @code{setdar} filter does not modify the pixel
13965 dimensions of the video frame. Also, the display aspect ratio set by
13966 this filter may be changed by later filters in the filterchain,
13967 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13970 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13971 the filter output video.
13973 Note that as a consequence of the application of this filter, the
13974 output display aspect ratio will change according to the equation
13977 Keep in mind that the sample aspect ratio set by the @code{setsar}
13978 filter may be changed by later filters in the filterchain, e.g. if
13979 another "setsar" or a "setdar" filter is applied.
13981 It accepts the following parameters:
13984 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13985 Set the aspect ratio used by the filter.
13987 The parameter can be a floating point number string, an expression, or
13988 a string of the form @var{num}:@var{den}, where @var{num} and
13989 @var{den} are the numerator and denominator of the aspect ratio. If
13990 the parameter is not specified, it is assumed the value "0".
13991 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13995 Set the maximum integer value to use for expressing numerator and
13996 denominator when reducing the expressed aspect ratio to a rational.
13997 Default value is @code{100}.
14001 The parameter @var{sar} is an expression containing
14002 the following constants:
14006 These are approximated values for the mathematical constants e
14007 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14010 The input width and height.
14013 These are the same as @var{w} / @var{h}.
14016 The input sample aspect ratio.
14019 The input display aspect ratio. It is the same as
14020 (@var{w} / @var{h}) * @var{sar}.
14023 Horizontal and vertical chroma subsample values. For example, for the
14024 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14027 @subsection Examples
14032 To change the display aspect ratio to 16:9, specify one of the following:
14039 To change the sample aspect ratio to 10:11, specify:
14045 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14046 1000 in the aspect ratio reduction, use the command:
14048 setdar=ratio=16/9:max=1000
14056 Force field for the output video frame.
14058 The @code{setfield} filter marks the interlace type field for the
14059 output frames. It does not change the input frame, but only sets the
14060 corresponding property, which affects how the frame is treated by
14061 following filters (e.g. @code{fieldorder} or @code{yadif}).
14063 The filter accepts the following options:
14068 Available values are:
14072 Keep the same field property.
14075 Mark the frame as bottom-field-first.
14078 Mark the frame as top-field-first.
14081 Mark the frame as progressive.
14087 Show a line containing various information for each input video frame.
14088 The input video is not modified.
14090 The shown line contains a sequence of key/value pairs of the form
14091 @var{key}:@var{value}.
14093 The following values are shown in the output:
14097 The (sequential) number of the input frame, starting from 0.
14100 The Presentation TimeStamp of the input frame, expressed as a number of
14101 time base units. The time base unit depends on the filter input pad.
14104 The Presentation TimeStamp of the input frame, expressed as a number of
14108 The position of the frame in the input stream, or -1 if this information is
14109 unavailable and/or meaningless (for example in case of synthetic video).
14112 The pixel format name.
14115 The sample aspect ratio of the input frame, expressed in the form
14116 @var{num}/@var{den}.
14119 The size of the input frame. For the syntax of this option, check the
14120 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14123 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14124 for bottom field first).
14127 This is 1 if the frame is a key frame, 0 otherwise.
14130 The picture type of the input frame ("I" for an I-frame, "P" for a
14131 P-frame, "B" for a B-frame, or "?" for an unknown type).
14132 Also refer to the documentation of the @code{AVPictureType} enum and of
14133 the @code{av_get_picture_type_char} function defined in
14134 @file{libavutil/avutil.h}.
14137 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14139 @item plane_checksum
14140 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14141 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14144 @section showpalette
14146 Displays the 256 colors palette of each frame. This filter is only relevant for
14147 @var{pal8} pixel format frames.
14149 It accepts the following option:
14153 Set the size of the box used to represent one palette color entry. Default is
14154 @code{30} (for a @code{30x30} pixel box).
14157 @section shuffleframes
14159 Reorder and/or duplicate and/or drop video frames.
14161 It accepts the following parameters:
14165 Set the destination indexes of input frames.
14166 This is space or '|' separated list of indexes that maps input frames to output
14167 frames. Number of indexes also sets maximal value that each index may have.
14168 '-1' index have special meaning and that is to drop frame.
14171 The first frame has the index 0. The default is to keep the input unchanged.
14173 @subsection Examples
14177 Swap second and third frame of every three frames of the input:
14179 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14183 Swap 10th and 1st frame of every ten frames of the input:
14185 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14189 @section shuffleplanes
14191 Reorder and/or duplicate video planes.
14193 It accepts the following parameters:
14198 The index of the input plane to be used as the first output plane.
14201 The index of the input plane to be used as the second output plane.
14204 The index of the input plane to be used as the third output plane.
14207 The index of the input plane to be used as the fourth output plane.
14211 The first plane has the index 0. The default is to keep the input unchanged.
14213 @subsection Examples
14217 Swap the second and third planes of the input:
14219 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14223 @anchor{signalstats}
14224 @section signalstats
14225 Evaluate various visual metrics that assist in determining issues associated
14226 with the digitization of analog video media.
14228 By default the filter will log these metadata values:
14232 Display the minimal Y value contained within the input frame. Expressed in
14236 Display the Y value at the 10% percentile within the input frame. Expressed in
14240 Display the average Y value within the input frame. Expressed in range of
14244 Display the Y value at the 90% percentile within the input frame. Expressed in
14248 Display the maximum Y value contained within the input frame. Expressed in
14252 Display the minimal U value contained within the input frame. Expressed in
14256 Display the U value at the 10% percentile within the input frame. Expressed in
14260 Display the average U value within the input frame. Expressed in range of
14264 Display the U value at the 90% percentile within the input frame. Expressed in
14268 Display the maximum U value contained within the input frame. Expressed in
14272 Display the minimal V value contained within the input frame. Expressed in
14276 Display the V value at the 10% percentile within the input frame. Expressed in
14280 Display the average V value within the input frame. Expressed in range of
14284 Display the V value at the 90% percentile within the input frame. Expressed in
14288 Display the maximum V value contained within the input frame. Expressed in
14292 Display the minimal saturation value contained within the input frame.
14293 Expressed in range of [0-~181.02].
14296 Display the saturation value at the 10% percentile within the input frame.
14297 Expressed in range of [0-~181.02].
14300 Display the average saturation value within the input frame. Expressed in range
14304 Display the saturation value at the 90% percentile within the input frame.
14305 Expressed in range of [0-~181.02].
14308 Display the maximum saturation value contained within the input frame.
14309 Expressed in range of [0-~181.02].
14312 Display the median value for hue within the input frame. Expressed in range of
14316 Display the average value for hue within the input frame. Expressed in range of
14320 Display the average of sample value difference between all values of the Y
14321 plane in the current frame and corresponding values of the previous input frame.
14322 Expressed in range of [0-255].
14325 Display the average of sample value difference between all values of the U
14326 plane in the current frame and corresponding values of the previous input frame.
14327 Expressed in range of [0-255].
14330 Display the average of sample value difference between all values of the V
14331 plane in the current frame and corresponding values of the previous input frame.
14332 Expressed in range of [0-255].
14335 Display bit depth of Y plane in current frame.
14336 Expressed in range of [0-16].
14339 Display bit depth of U plane in current frame.
14340 Expressed in range of [0-16].
14343 Display bit depth of V plane in current frame.
14344 Expressed in range of [0-16].
14347 The filter accepts the following options:
14353 @option{stat} specify an additional form of image analysis.
14354 @option{out} output video with the specified type of pixel highlighted.
14356 Both options accept the following values:
14360 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14361 unlike the neighboring pixels of the same field. Examples of temporal outliers
14362 include the results of video dropouts, head clogs, or tape tracking issues.
14365 Identify @var{vertical line repetition}. Vertical line repetition includes
14366 similar rows of pixels within a frame. In born-digital video vertical line
14367 repetition is common, but this pattern is uncommon in video digitized from an
14368 analog source. When it occurs in video that results from the digitization of an
14369 analog source it can indicate concealment from a dropout compensator.
14372 Identify pixels that fall outside of legal broadcast range.
14376 Set the highlight color for the @option{out} option. The default color is
14380 @subsection Examples
14384 Output data of various video metrics:
14386 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14390 Output specific data about the minimum and maximum values of the Y plane per frame:
14392 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14396 Playback video while highlighting pixels that are outside of broadcast range in red.
14398 ffplay example.mov -vf signalstats="out=brng:color=red"
14402 Playback video with signalstats metadata drawn over the frame.
14404 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14407 The contents of signalstat_drawtext.txt used in the command are:
14410 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14411 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14412 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14413 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14421 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14422 input. In this case the matching between the inputs can be calculated additionally.
14423 The filter always passes through the first input. The signature of each stream can
14424 be written into a file.
14426 It accepts the following options:
14430 Enable or disable the matching process.
14432 Available values are:
14436 Disable the calculation of a matching (default).
14438 Calculate the matching for the whole video and output whether the whole video
14439 matches or only parts.
14441 Calculate only until a matching is found or the video ends. Should be faster in
14446 Set the number of inputs. The option value must be a non negative integer.
14447 Default value is 1.
14450 Set the path to which the output is written. If there is more than one input,
14451 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14452 integer), that will be replaced with the input number. If no filename is
14453 specified, no output will be written. This is the default.
14456 Choose the output format.
14458 Available values are:
14462 Use the specified binary representation (default).
14464 Use the specified xml representation.
14468 Set threshold to detect one word as similar. The option value must be an integer
14469 greater than zero. The default value is 9000.
14472 Set threshold to detect all words as similar. The option value must be an integer
14473 greater than zero. The default value is 60000.
14476 Set threshold to detect frames as similar. The option value must be an integer
14477 greater than zero. The default value is 116.
14480 Set the minimum length of a sequence in frames to recognize it as matching
14481 sequence. The option value must be a non negative integer value.
14482 The default value is 0.
14485 Set the minimum relation, that matching frames to all frames must have.
14486 The option value must be a double value between 0 and 1. The default value is 0.5.
14489 @subsection Examples
14493 To calculate the signature of an input video and store it in signature.bin:
14495 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14499 To detect whether two videos match and store the signatures in XML format in
14500 signature0.xml and signature1.xml:
14502 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 -
14510 Blur the input video without impacting the outlines.
14512 It accepts the following options:
14515 @item luma_radius, lr
14516 Set the luma radius. The option value must be a float number in
14517 the range [0.1,5.0] that specifies the variance of the gaussian filter
14518 used to blur the image (slower if larger). Default value is 1.0.
14520 @item luma_strength, ls
14521 Set the luma strength. The option value must be a float number
14522 in the range [-1.0,1.0] that configures the blurring. A value included
14523 in [0.0,1.0] will blur the image whereas a value included in
14524 [-1.0,0.0] will sharpen the image. Default value is 1.0.
14526 @item luma_threshold, lt
14527 Set the luma threshold used as a coefficient to determine
14528 whether a pixel should be blurred or not. The option value must be an
14529 integer in the range [-30,30]. A value of 0 will filter all the image,
14530 a value included in [0,30] will filter flat areas and a value included
14531 in [-30,0] will filter edges. Default value is 0.
14533 @item chroma_radius, cr
14534 Set the chroma radius. The option value must be a float number in
14535 the range [0.1,5.0] that specifies the variance of the gaussian filter
14536 used to blur the image (slower if larger). Default value is @option{luma_radius}.
14538 @item chroma_strength, cs
14539 Set the chroma strength. The option value must be a float number
14540 in the range [-1.0,1.0] that configures the blurring. A value included
14541 in [0.0,1.0] will blur the image whereas a value included in
14542 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
14544 @item chroma_threshold, ct
14545 Set the chroma threshold used as a coefficient to determine
14546 whether a pixel should be blurred or not. The option value must be an
14547 integer in the range [-30,30]. A value of 0 will filter all the image,
14548 a value included in [0,30] will filter flat areas and a value included
14549 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
14552 If a chroma option is not explicitly set, the corresponding luma value
14557 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
14559 This filter takes in input two input videos, the first input is
14560 considered the "main" source and is passed unchanged to the
14561 output. The second input is used as a "reference" video for computing
14564 Both video inputs must have the same resolution and pixel format for
14565 this filter to work correctly. Also it assumes that both inputs
14566 have the same number of frames, which are compared one by one.
14568 The filter stores the calculated SSIM of each frame.
14570 The description of the accepted parameters follows.
14573 @item stats_file, f
14574 If specified the filter will use the named file to save the SSIM of
14575 each individual frame. When filename equals "-" the data is sent to
14579 The file printed if @var{stats_file} is selected, contains a sequence of
14580 key/value pairs of the form @var{key}:@var{value} for each compared
14583 A description of each shown parameter follows:
14587 sequential number of the input frame, starting from 1
14589 @item Y, U, V, R, G, B
14590 SSIM of the compared frames for the component specified by the suffix.
14593 SSIM of the compared frames for the whole frame.
14596 Same as above but in dB representation.
14599 This filter also supports the @ref{framesync} options.
14603 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14604 [main][ref] ssim="stats_file=stats.log" [out]
14607 On this example the input file being processed is compared with the
14608 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
14609 is stored in @file{stats.log}.
14611 Another example with both psnr and ssim at same time:
14613 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
14618 Convert between different stereoscopic image formats.
14620 The filters accept the following options:
14624 Set stereoscopic image format of input.
14626 Available values for input image formats are:
14629 side by side parallel (left eye left, right eye right)
14632 side by side crosseye (right eye left, left eye right)
14635 side by side parallel with half width resolution
14636 (left eye left, right eye right)
14639 side by side crosseye with half width resolution
14640 (right eye left, left eye right)
14643 above-below (left eye above, right eye below)
14646 above-below (right eye above, left eye below)
14649 above-below with half height resolution
14650 (left eye above, right eye below)
14653 above-below with half height resolution
14654 (right eye above, left eye below)
14657 alternating frames (left eye first, right eye second)
14660 alternating frames (right eye first, left eye second)
14663 interleaved rows (left eye has top row, right eye starts on next row)
14666 interleaved rows (right eye has top row, left eye starts on next row)
14669 interleaved columns, left eye first
14672 interleaved columns, right eye first
14674 Default value is @samp{sbsl}.
14678 Set stereoscopic image format of output.
14682 side by side parallel (left eye left, right eye right)
14685 side by side crosseye (right eye left, left eye right)
14688 side by side parallel with half width resolution
14689 (left eye left, right eye right)
14692 side by side crosseye with half width resolution
14693 (right eye left, left eye right)
14696 above-below (left eye above, right eye below)
14699 above-below (right eye above, left eye below)
14702 above-below with half height resolution
14703 (left eye above, right eye below)
14706 above-below with half height resolution
14707 (right eye above, left eye below)
14710 alternating frames (left eye first, right eye second)
14713 alternating frames (right eye first, left eye second)
14716 interleaved rows (left eye has top row, right eye starts on next row)
14719 interleaved rows (right eye has top row, left eye starts on next row)
14722 anaglyph red/blue gray
14723 (red filter on left eye, blue filter on right eye)
14726 anaglyph red/green gray
14727 (red filter on left eye, green filter on right eye)
14730 anaglyph red/cyan gray
14731 (red filter on left eye, cyan filter on right eye)
14734 anaglyph red/cyan half colored
14735 (red filter on left eye, cyan filter on right eye)
14738 anaglyph red/cyan color
14739 (red filter on left eye, cyan filter on right eye)
14742 anaglyph red/cyan color optimized with the least squares projection of dubois
14743 (red filter on left eye, cyan filter on right eye)
14746 anaglyph green/magenta gray
14747 (green filter on left eye, magenta filter on right eye)
14750 anaglyph green/magenta half colored
14751 (green filter on left eye, magenta filter on right eye)
14754 anaglyph green/magenta colored
14755 (green filter on left eye, magenta filter on right eye)
14758 anaglyph green/magenta color optimized with the least squares projection of dubois
14759 (green filter on left eye, magenta filter on right eye)
14762 anaglyph yellow/blue gray
14763 (yellow filter on left eye, blue filter on right eye)
14766 anaglyph yellow/blue half colored
14767 (yellow filter on left eye, blue filter on right eye)
14770 anaglyph yellow/blue colored
14771 (yellow filter on left eye, blue filter on right eye)
14774 anaglyph yellow/blue color optimized with the least squares projection of dubois
14775 (yellow filter on left eye, blue filter on right eye)
14778 mono output (left eye only)
14781 mono output (right eye only)
14784 checkerboard, left eye first
14787 checkerboard, right eye first
14790 interleaved columns, left eye first
14793 interleaved columns, right eye first
14799 Default value is @samp{arcd}.
14802 @subsection Examples
14806 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14812 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14818 @section streamselect, astreamselect
14819 Select video or audio streams.
14821 The filter accepts the following options:
14825 Set number of inputs. Default is 2.
14828 Set input indexes to remap to outputs.
14831 @subsection Commands
14833 The @code{streamselect} and @code{astreamselect} filter supports the following
14838 Set input indexes to remap to outputs.
14841 @subsection Examples
14845 Select first 5 seconds 1st stream and rest of time 2nd stream:
14847 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14851 Same as above, but for audio:
14853 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14858 Apply sobel operator to input video stream.
14860 The filter accepts the following option:
14864 Set which planes will be processed, unprocessed planes will be copied.
14865 By default value 0xf, all planes will be processed.
14868 Set value which will be multiplied with filtered result.
14871 Set value which will be added to filtered result.
14877 Apply a simple postprocessing filter that compresses and decompresses the image
14878 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14879 and average the results.
14881 The filter accepts the following options:
14885 Set quality. This option defines the number of levels for averaging. It accepts
14886 an integer in the range 0-6. If set to @code{0}, the filter will have no
14887 effect. A value of @code{6} means the higher quality. For each increment of
14888 that value the speed drops by a factor of approximately 2. Default value is
14892 Force a constant quantization parameter. If not set, the filter will use the QP
14893 from the video stream (if available).
14896 Set thresholding mode. Available modes are:
14900 Set hard thresholding (default).
14902 Set soft thresholding (better de-ringing effect, but likely blurrier).
14905 @item use_bframe_qp
14906 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14907 option may cause flicker since the B-Frames have often larger QP. Default is
14908 @code{0} (not enabled).
14914 Draw subtitles on top of input video using the libass library.
14916 To enable compilation of this filter you need to configure FFmpeg with
14917 @code{--enable-libass}. This filter also requires a build with libavcodec and
14918 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14919 Alpha) subtitles format.
14921 The filter accepts the following options:
14925 Set the filename of the subtitle file to read. It must be specified.
14927 @item original_size
14928 Specify the size of the original video, the video for which the ASS file
14929 was composed. For the syntax of this option, check the
14930 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14931 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14932 correctly scale the fonts if the aspect ratio has been changed.
14935 Set a directory path containing fonts that can be used by the filter.
14936 These fonts will be used in addition to whatever the font provider uses.
14939 Process alpha channel, by default alpha channel is untouched.
14942 Set subtitles input character encoding. @code{subtitles} filter only. Only
14943 useful if not UTF-8.
14945 @item stream_index, si
14946 Set subtitles stream index. @code{subtitles} filter only.
14949 Override default style or script info parameters of the subtitles. It accepts a
14950 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14953 If the first key is not specified, it is assumed that the first value
14954 specifies the @option{filename}.
14956 For example, to render the file @file{sub.srt} on top of the input
14957 video, use the command:
14962 which is equivalent to:
14964 subtitles=filename=sub.srt
14967 To render the default subtitles stream from file @file{video.mkv}, use:
14969 subtitles=video.mkv
14972 To render the second subtitles stream from that file, use:
14974 subtitles=video.mkv:si=1
14977 To make the subtitles stream from @file{sub.srt} appear in transparent green
14978 @code{DejaVu Serif}, use:
14980 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14983 @section super2xsai
14985 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14986 Interpolate) pixel art scaling algorithm.
14988 Useful for enlarging pixel art images without reducing sharpness.
14992 Swap two rectangular objects in video.
14994 This filter accepts the following options:
15004 Set 1st rect x coordinate.
15007 Set 1st rect y coordinate.
15010 Set 2nd rect x coordinate.
15013 Set 2nd rect y coordinate.
15015 All expressions are evaluated once for each frame.
15018 The all options are expressions containing the following constants:
15023 The input width and height.
15026 same as @var{w} / @var{h}
15029 input sample aspect ratio
15032 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15035 The number of the input frame, starting from 0.
15038 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15041 the position in the file of the input frame, NAN if unknown
15049 Apply telecine process to the video.
15051 This filter accepts the following options:
15060 The default value is @code{top}.
15064 A string of numbers representing the pulldown pattern you wish to apply.
15065 The default value is @code{23}.
15069 Some typical patterns:
15074 24p: 2332 (preferred)
15081 24p: 222222222223 ("Euro pulldown")
15088 Apply threshold effect to video stream.
15090 This filter needs four video streams to perform thresholding.
15091 First stream is stream we are filtering.
15092 Second stream is holding threshold values, third stream is holding min values,
15093 and last, fourth stream is holding max values.
15095 The filter accepts the following option:
15099 Set which planes will be processed, unprocessed planes will be copied.
15100 By default value 0xf, all planes will be processed.
15103 For example if first stream pixel's component value is less then threshold value
15104 of pixel component from 2nd threshold stream, third stream value will picked,
15105 otherwise fourth stream pixel component value will be picked.
15107 Using color source filter one can perform various types of thresholding:
15109 @subsection Examples
15113 Binary threshold, using gray color as threshold:
15115 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15119 Inverted binary threshold, using gray color as threshold:
15121 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15125 Truncate binary threshold, using gray color as threshold:
15127 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15131 Threshold to zero, using gray color as threshold:
15133 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15137 Inverted threshold to zero, using gray color as threshold:
15139 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15144 Select the most representative frame in a given sequence of consecutive frames.
15146 The filter accepts the following options:
15150 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15151 will pick one of them, and then handle the next batch of @var{n} frames until
15152 the end. Default is @code{100}.
15155 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15156 value will result in a higher memory usage, so a high value is not recommended.
15158 @subsection Examples
15162 Extract one picture each 50 frames:
15168 Complete example of a thumbnail creation with @command{ffmpeg}:
15170 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15176 Tile several successive frames together.
15178 The filter accepts the following options:
15183 Set the grid size (i.e. the number of lines and columns). For the syntax of
15184 this option, check the
15185 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15188 Set the maximum number of frames to render in the given area. It must be less
15189 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15190 the area will be used.
15193 Set the outer border margin in pixels.
15196 Set the inner border thickness (i.e. the number of pixels between frames). For
15197 more advanced padding options (such as having different values for the edges),
15198 refer to the pad video filter.
15201 Specify the color of the unused area. For the syntax of this option, check the
15202 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15203 The default value of @var{color} is "black".
15206 Set the number of frames to overlap when tiling several successive frames together.
15207 The value must be between @code{0} and @var{nb_frames - 1}.
15210 Set the number of frames to initially be empty before displaying first output frame.
15211 This controls how soon will one get first output frame.
15212 The value must be between @code{0} and @var{nb_frames - 1}.
15215 @subsection Examples
15219 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15221 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15223 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15224 duplicating each output frame to accommodate the originally detected frame
15228 Display @code{5} pictures in an area of @code{3x2} frames,
15229 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15230 mixed flat and named options:
15232 tile=3x2:nb_frames=5:padding=7:margin=2
15236 @section tinterlace
15238 Perform various types of temporal field interlacing.
15240 Frames are counted starting from 1, so the first input frame is
15243 The filter accepts the following options:
15248 Specify the mode of the interlacing. This option can also be specified
15249 as a value alone. See below for a list of values for this option.
15251 Available values are:
15255 Move odd frames into the upper field, even into the lower field,
15256 generating a double height frame at half frame rate.
15260 Frame 1 Frame 2 Frame 3 Frame 4
15262 11111 22222 33333 44444
15263 11111 22222 33333 44444
15264 11111 22222 33333 44444
15265 11111 22222 33333 44444
15279 Only output odd frames, even frames are dropped, generating a frame with
15280 unchanged height at half frame rate.
15285 Frame 1 Frame 2 Frame 3 Frame 4
15287 11111 22222 33333 44444
15288 11111 22222 33333 44444
15289 11111 22222 33333 44444
15290 11111 22222 33333 44444
15300 Only output even frames, odd frames are dropped, generating a frame with
15301 unchanged height at half frame rate.
15306 Frame 1 Frame 2 Frame 3 Frame 4
15308 11111 22222 33333 44444
15309 11111 22222 33333 44444
15310 11111 22222 33333 44444
15311 11111 22222 33333 44444
15321 Expand each frame to full height, but pad alternate lines with black,
15322 generating a frame with double height at the same input frame rate.
15327 Frame 1 Frame 2 Frame 3 Frame 4
15329 11111 22222 33333 44444
15330 11111 22222 33333 44444
15331 11111 22222 33333 44444
15332 11111 22222 33333 44444
15335 11111 ..... 33333 .....
15336 ..... 22222 ..... 44444
15337 11111 ..... 33333 .....
15338 ..... 22222 ..... 44444
15339 11111 ..... 33333 .....
15340 ..... 22222 ..... 44444
15341 11111 ..... 33333 .....
15342 ..... 22222 ..... 44444
15346 @item interleave_top, 4
15347 Interleave the upper field from odd frames with the lower field from
15348 even frames, generating a frame with unchanged height at half frame rate.
15353 Frame 1 Frame 2 Frame 3 Frame 4
15355 11111<- 22222 33333<- 44444
15356 11111 22222<- 33333 44444<-
15357 11111<- 22222 33333<- 44444
15358 11111 22222<- 33333 44444<-
15368 @item interleave_bottom, 5
15369 Interleave the lower field from odd frames with the upper field from
15370 even frames, generating a frame with unchanged height at half frame rate.
15375 Frame 1 Frame 2 Frame 3 Frame 4
15377 11111 22222<- 33333 44444<-
15378 11111<- 22222 33333<- 44444
15379 11111 22222<- 33333 44444<-
15380 11111<- 22222 33333<- 44444
15390 @item interlacex2, 6
15391 Double frame rate with unchanged height. Frames are inserted each
15392 containing the second temporal field from the previous input frame and
15393 the first temporal field from the next input frame. This mode relies on
15394 the top_field_first flag. Useful for interlaced video displays with no
15395 field synchronisation.
15400 Frame 1 Frame 2 Frame 3 Frame 4
15402 11111 22222 33333 44444
15403 11111 22222 33333 44444
15404 11111 22222 33333 44444
15405 11111 22222 33333 44444
15408 11111 22222 22222 33333 33333 44444 44444
15409 11111 11111 22222 22222 33333 33333 44444
15410 11111 22222 22222 33333 33333 44444 44444
15411 11111 11111 22222 22222 33333 33333 44444
15416 Move odd frames into the upper field, even into the lower field,
15417 generating a double height frame at same frame rate.
15422 Frame 1 Frame 2 Frame 3 Frame 4
15424 11111 22222 33333 44444
15425 11111 22222 33333 44444
15426 11111 22222 33333 44444
15427 11111 22222 33333 44444
15430 11111 33333 33333 55555
15431 22222 22222 44444 44444
15432 11111 33333 33333 55555
15433 22222 22222 44444 44444
15434 11111 33333 33333 55555
15435 22222 22222 44444 44444
15436 11111 33333 33333 55555
15437 22222 22222 44444 44444
15442 Numeric values are deprecated but are accepted for backward
15443 compatibility reasons.
15445 Default mode is @code{merge}.
15448 Specify flags influencing the filter process.
15450 Available value for @var{flags} is:
15453 @item low_pass_filter, vlfp
15454 Enable linear vertical low-pass filtering in the filter.
15455 Vertical low-pass filtering is required when creating an interlaced
15456 destination from a progressive source which contains high-frequency
15457 vertical detail. Filtering will reduce interlace 'twitter' and Moire
15460 @item complex_filter, cvlfp
15461 Enable complex vertical low-pass filtering.
15462 This will slightly less reduce interlace 'twitter' and Moire
15463 patterning but better retain detail and subjective sharpness impression.
15467 Vertical low-pass filtering can only be enabled for @option{mode}
15468 @var{interleave_top} and @var{interleave_bottom}.
15473 Tone map colors from different dynamic ranges.
15475 This filter expects data in single precision floating point, as it needs to
15476 operate on (and can output) out-of-range values. Another filter, such as
15477 @ref{zscale}, is needed to convert the resulting frame to a usable format.
15479 The tonemapping algorithms implemented only work on linear light, so input
15480 data should be linearized beforehand (and possibly correctly tagged).
15483 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
15486 @subsection Options
15487 The filter accepts the following options.
15491 Set the tone map algorithm to use.
15493 Possible values are:
15496 Do not apply any tone map, only desaturate overbright pixels.
15499 Hard-clip any out-of-range values. Use it for perfect color accuracy for
15500 in-range values, while distorting out-of-range values.
15503 Stretch the entire reference gamut to a linear multiple of the display.
15506 Fit a logarithmic transfer between the tone curves.
15509 Preserve overall image brightness with a simple curve, using nonlinear
15510 contrast, which results in flattening details and degrading color accuracy.
15513 Preserve both dark and bright details better than @var{reinhard}, at the cost
15514 of slightly darkening everything. Use it when detail preservation is more
15515 important than color and brightness accuracy.
15518 Smoothly map out-of-range values, while retaining contrast and colors for
15519 in-range material as much as possible. Use it when color accuracy is more
15520 important than detail preservation.
15526 Tune the tone mapping algorithm.
15528 This affects the following algorithms:
15534 Specifies the scale factor to use while stretching.
15538 Specifies the exponent of the function.
15542 Specify an extra linear coefficient to multiply into the signal before clipping.
15546 Specify the local contrast coefficient at the display peak.
15547 Default to 0.5, which means that in-gamut values will be about half as bright
15554 Specify the transition point from linear to mobius transform. Every value
15555 below this point is guaranteed to be mapped 1:1. The higher the value, the
15556 more accurate the result will be, at the cost of losing bright details.
15557 Default to 0.3, which due to the steep initial slope still preserves in-range
15558 colors fairly accurately.
15562 Apply desaturation for highlights that exceed this level of brightness. The
15563 higher the parameter, the more color information will be preserved. This
15564 setting helps prevent unnaturally blown-out colors for super-highlights, by
15565 (smoothly) turning into white instead. This makes images feel more natural,
15566 at the cost of reducing information about out-of-range colors.
15568 The default of 2.0 is somewhat conservative and will mostly just apply to
15569 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
15571 This option works only if the input frame has a supported color tag.
15574 Override signal/nominal/reference peak with this value. Useful when the
15575 embedded peak information in display metadata is not reliable or when tone
15576 mapping from a lower range to a higher range.
15581 Transpose rows with columns in the input video and optionally flip it.
15583 It accepts the following parameters:
15588 Specify the transposition direction.
15590 Can assume the following values:
15592 @item 0, 4, cclock_flip
15593 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
15601 Rotate by 90 degrees clockwise, that is:
15609 Rotate by 90 degrees counterclockwise, that is:
15616 @item 3, 7, clock_flip
15617 Rotate by 90 degrees clockwise and vertically flip, that is:
15625 For values between 4-7, the transposition is only done if the input
15626 video geometry is portrait and not landscape. These values are
15627 deprecated, the @code{passthrough} option should be used instead.
15629 Numerical values are deprecated, and should be dropped in favor of
15630 symbolic constants.
15633 Do not apply the transposition if the input geometry matches the one
15634 specified by the specified value. It accepts the following values:
15637 Always apply transposition.
15639 Preserve portrait geometry (when @var{height} >= @var{width}).
15641 Preserve landscape geometry (when @var{width} >= @var{height}).
15644 Default value is @code{none}.
15647 For example to rotate by 90 degrees clockwise and preserve portrait
15650 transpose=dir=1:passthrough=portrait
15653 The command above can also be specified as:
15655 transpose=1:portrait
15659 Trim the input so that the output contains one continuous subpart of the input.
15661 It accepts the following parameters:
15664 Specify the time of the start of the kept section, i.e. the frame with the
15665 timestamp @var{start} will be the first frame in the output.
15668 Specify the time of the first frame that will be dropped, i.e. the frame
15669 immediately preceding the one with the timestamp @var{end} will be the last
15670 frame in the output.
15673 This is the same as @var{start}, except this option sets the start timestamp
15674 in timebase units instead of seconds.
15677 This is the same as @var{end}, except this option sets the end timestamp
15678 in timebase units instead of seconds.
15681 The maximum duration of the output in seconds.
15684 The number of the first frame that should be passed to the output.
15687 The number of the first frame that should be dropped.
15690 @option{start}, @option{end}, and @option{duration} are expressed as time
15691 duration specifications; see
15692 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15693 for the accepted syntax.
15695 Note that the first two sets of the start/end options and the @option{duration}
15696 option look at the frame timestamp, while the _frame variants simply count the
15697 frames that pass through the filter. Also note that this filter does not modify
15698 the timestamps. If you wish for the output timestamps to start at zero, insert a
15699 setpts filter after the trim filter.
15701 If multiple start or end options are set, this filter tries to be greedy and
15702 keep all the frames that match at least one of the specified constraints. To keep
15703 only the part that matches all the constraints at once, chain multiple trim
15706 The defaults are such that all the input is kept. So it is possible to set e.g.
15707 just the end values to keep everything before the specified time.
15712 Drop everything except the second minute of input:
15714 ffmpeg -i INPUT -vf trim=60:120
15718 Keep only the first second:
15720 ffmpeg -i INPUT -vf trim=duration=1
15725 @section unpremultiply
15726 Apply alpha unpremultiply effect to input video stream using first plane
15727 of second stream as alpha.
15729 Both streams must have same dimensions and same pixel format.
15731 The filter accepts the following option:
15735 Set which planes will be processed, unprocessed planes will be copied.
15736 By default value 0xf, all planes will be processed.
15738 If the format has 1 or 2 components, then luma is bit 0.
15739 If the format has 3 or 4 components:
15740 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15741 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15742 If present, the alpha channel is always the last bit.
15745 Do not require 2nd input for processing, instead use alpha plane from input stream.
15751 Sharpen or blur the input video.
15753 It accepts the following parameters:
15756 @item luma_msize_x, lx
15757 Set the luma matrix horizontal size. It must be an odd integer between
15758 3 and 23. The default value is 5.
15760 @item luma_msize_y, ly
15761 Set the luma matrix vertical size. It must be an odd integer between 3
15762 and 23. The default value is 5.
15764 @item luma_amount, la
15765 Set the luma effect strength. It must be a floating point number, reasonable
15766 values lay between -1.5 and 1.5.
15768 Negative values will blur the input video, while positive values will
15769 sharpen it, a value of zero will disable the effect.
15771 Default value is 1.0.
15773 @item chroma_msize_x, cx
15774 Set the chroma matrix horizontal size. It must be an odd integer
15775 between 3 and 23. The default value is 5.
15777 @item chroma_msize_y, cy
15778 Set the chroma matrix vertical size. It must be an odd integer
15779 between 3 and 23. The default value is 5.
15781 @item chroma_amount, ca
15782 Set the chroma effect strength. It must be a floating point number, reasonable
15783 values lay between -1.5 and 1.5.
15785 Negative values will blur the input video, while positive values will
15786 sharpen it, a value of zero will disable the effect.
15788 Default value is 0.0.
15792 All parameters are optional and default to the equivalent of the
15793 string '5:5:1.0:5:5:0.0'.
15795 @subsection Examples
15799 Apply strong luma sharpen effect:
15801 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15805 Apply a strong blur of both luma and chroma parameters:
15807 unsharp=7:7:-2:7:7:-2
15813 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15814 the image at several (or - in the case of @option{quality} level @code{8} - all)
15815 shifts and average the results.
15817 The way this differs from the behavior of spp is that uspp actually encodes &
15818 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15819 DCT similar to MJPEG.
15821 The filter accepts the following options:
15825 Set quality. This option defines the number of levels for averaging. It accepts
15826 an integer in the range 0-8. If set to @code{0}, the filter will have no
15827 effect. A value of @code{8} means the higher quality. For each increment of
15828 that value the speed drops by a factor of approximately 2. Default value is
15832 Force a constant quantization parameter. If not set, the filter will use the QP
15833 from the video stream (if available).
15836 @section vaguedenoiser
15838 Apply a wavelet based denoiser.
15840 It transforms each frame from the video input into the wavelet domain,
15841 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15842 the obtained coefficients. It does an inverse wavelet transform after.
15843 Due to wavelet properties, it should give a nice smoothed result, and
15844 reduced noise, without blurring picture features.
15846 This filter accepts the following options:
15850 The filtering strength. The higher, the more filtered the video will be.
15851 Hard thresholding can use a higher threshold than soft thresholding
15852 before the video looks overfiltered. Default value is 2.
15855 The filtering method the filter will use.
15857 It accepts the following values:
15860 All values under the threshold will be zeroed.
15863 All values under the threshold will be zeroed. All values above will be
15864 reduced by the threshold.
15867 Scales or nullifies coefficients - intermediary between (more) soft and
15868 (less) hard thresholding.
15871 Default is garrote.
15874 Number of times, the wavelet will decompose the picture. Picture can't
15875 be decomposed beyond a particular point (typically, 8 for a 640x480
15876 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15879 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15882 A list of the planes to process. By default all planes are processed.
15885 @section vectorscope
15887 Display 2 color component values in the two dimensional graph (which is called
15890 This filter accepts the following options:
15894 Set vectorscope mode.
15896 It accepts the following values:
15899 Gray values are displayed on graph, higher brightness means more pixels have
15900 same component color value on location in graph. This is the default mode.
15903 Gray values are displayed on graph. Surrounding pixels values which are not
15904 present in video frame are drawn in gradient of 2 color components which are
15905 set by option @code{x} and @code{y}. The 3rd color component is static.
15908 Actual color components values present in video frame are displayed on graph.
15911 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15912 on graph increases value of another color component, which is luminance by
15913 default values of @code{x} and @code{y}.
15916 Actual colors present in video frame are displayed on graph. If two different
15917 colors map to same position on graph then color with higher value of component
15918 not present in graph is picked.
15921 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15922 component picked from radial gradient.
15926 Set which color component will be represented on X-axis. Default is @code{1}.
15929 Set which color component will be represented on Y-axis. Default is @code{2}.
15932 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15933 of color component which represents frequency of (X, Y) location in graph.
15938 No envelope, this is default.
15941 Instant envelope, even darkest single pixel will be clearly highlighted.
15944 Hold maximum and minimum values presented in graph over time. This way you
15945 can still spot out of range values without constantly looking at vectorscope.
15948 Peak and instant envelope combined together.
15952 Set what kind of graticule to draw.
15960 Set graticule opacity.
15963 Set graticule flags.
15967 Draw graticule for white point.
15970 Draw graticule for black point.
15973 Draw color points short names.
15977 Set background opacity.
15979 @item lthreshold, l
15980 Set low threshold for color component not represented on X or Y axis.
15981 Values lower than this value will be ignored. Default is 0.
15982 Note this value is multiplied with actual max possible value one pixel component
15983 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15986 @item hthreshold, h
15987 Set high threshold for color component not represented on X or Y axis.
15988 Values higher than this value will be ignored. Default is 1.
15989 Note this value is multiplied with actual max possible value one pixel component
15990 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15991 is 0.9 * 255 = 230.
15993 @item colorspace, c
15994 Set what kind of colorspace to use when drawing graticule.
16003 @anchor{vidstabdetect}
16004 @section vidstabdetect
16006 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16007 @ref{vidstabtransform} for pass 2.
16009 This filter generates a file with relative translation and rotation
16010 transform information about subsequent frames, which is then used by
16011 the @ref{vidstabtransform} filter.
16013 To enable compilation of this filter you need to configure FFmpeg with
16014 @code{--enable-libvidstab}.
16016 This filter accepts the following options:
16020 Set the path to the file used to write the transforms information.
16021 Default value is @file{transforms.trf}.
16024 Set how shaky the video is and how quick the camera is. It accepts an
16025 integer in the range 1-10, a value of 1 means little shakiness, a
16026 value of 10 means strong shakiness. Default value is 5.
16029 Set the accuracy of the detection process. It must be a value in the
16030 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16031 accuracy. Default value is 15.
16034 Set stepsize of the search process. The region around minimum is
16035 scanned with 1 pixel resolution. Default value is 6.
16038 Set minimum contrast. Below this value a local measurement field is
16039 discarded. Must be a floating point value in the range 0-1. Default
16043 Set reference frame number for tripod mode.
16045 If enabled, the motion of the frames is compared to a reference frame
16046 in the filtered stream, identified by the specified number. The idea
16047 is to compensate all movements in a more-or-less static scene and keep
16048 the camera view absolutely still.
16050 If set to 0, it is disabled. The frames are counted starting from 1.
16053 Show fields and transforms in the resulting frames. It accepts an
16054 integer in the range 0-2. Default value is 0, which disables any
16058 @subsection Examples
16062 Use default values:
16068 Analyze strongly shaky movie and put the results in file
16069 @file{mytransforms.trf}:
16071 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16075 Visualize the result of internal transformations in the resulting
16078 vidstabdetect=show=1
16082 Analyze a video with medium shakiness using @command{ffmpeg}:
16084 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16088 @anchor{vidstabtransform}
16089 @section vidstabtransform
16091 Video stabilization/deshaking: pass 2 of 2,
16092 see @ref{vidstabdetect} for pass 1.
16094 Read a file with transform information for each frame and
16095 apply/compensate them. Together with the @ref{vidstabdetect}
16096 filter this can be used to deshake videos. See also
16097 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16098 the @ref{unsharp} filter, see below.
16100 To enable compilation of this filter you need to configure FFmpeg with
16101 @code{--enable-libvidstab}.
16103 @subsection Options
16107 Set path to the file used to read the transforms. Default value is
16108 @file{transforms.trf}.
16111 Set the number of frames (value*2 + 1) used for lowpass filtering the
16112 camera movements. Default value is 10.
16114 For example a number of 10 means that 21 frames are used (10 in the
16115 past and 10 in the future) to smoothen the motion in the video. A
16116 larger value leads to a smoother video, but limits the acceleration of
16117 the camera (pan/tilt movements). 0 is a special case where a static
16118 camera is simulated.
16121 Set the camera path optimization algorithm.
16123 Accepted values are:
16126 gaussian kernel low-pass filter on camera motion (default)
16128 averaging on transformations
16132 Set maximal number of pixels to translate frames. Default value is -1,
16136 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16137 value is -1, meaning no limit.
16140 Specify how to deal with borders that may be visible due to movement
16143 Available values are:
16146 keep image information from previous frame (default)
16148 fill the border black
16152 Invert transforms if set to 1. Default value is 0.
16155 Consider transforms as relative to previous frame if set to 1,
16156 absolute if set to 0. Default value is 0.
16159 Set percentage to zoom. A positive value will result in a zoom-in
16160 effect, a negative value in a zoom-out effect. Default value is 0 (no
16164 Set optimal zooming to avoid borders.
16166 Accepted values are:
16171 optimal static zoom value is determined (only very strong movements
16172 will lead to visible borders) (default)
16174 optimal adaptive zoom value is determined (no borders will be
16175 visible), see @option{zoomspeed}
16178 Note that the value given at zoom is added to the one calculated here.
16181 Set percent to zoom maximally each frame (enabled when
16182 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16186 Specify type of interpolation.
16188 Available values are:
16193 linear only horizontal
16195 linear in both directions (default)
16197 cubic in both directions (slow)
16201 Enable virtual tripod mode if set to 1, which is equivalent to
16202 @code{relative=0:smoothing=0}. Default value is 0.
16204 Use also @code{tripod} option of @ref{vidstabdetect}.
16207 Increase log verbosity if set to 1. Also the detected global motions
16208 are written to the temporary file @file{global_motions.trf}. Default
16212 @subsection Examples
16216 Use @command{ffmpeg} for a typical stabilization with default values:
16218 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16221 Note the use of the @ref{unsharp} filter which is always recommended.
16224 Zoom in a bit more and load transform data from a given file:
16226 vidstabtransform=zoom=5:input="mytransforms.trf"
16230 Smoothen the video even more:
16232 vidstabtransform=smoothing=30
16238 Flip the input video vertically.
16240 For example, to vertically flip a video with @command{ffmpeg}:
16242 ffmpeg -i in.avi -vf "vflip" out.avi
16248 Make or reverse a natural vignetting effect.
16250 The filter accepts the following options:
16254 Set lens angle expression as a number of radians.
16256 The value is clipped in the @code{[0,PI/2]} range.
16258 Default value: @code{"PI/5"}
16262 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16266 Set forward/backward mode.
16268 Available modes are:
16271 The larger the distance from the central point, the darker the image becomes.
16274 The larger the distance from the central point, the brighter the image becomes.
16275 This can be used to reverse a vignette effect, though there is no automatic
16276 detection to extract the lens @option{angle} and other settings (yet). It can
16277 also be used to create a burning effect.
16280 Default value is @samp{forward}.
16283 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16285 It accepts the following values:
16288 Evaluate expressions only once during the filter initialization.
16291 Evaluate expressions for each incoming frame. This is way slower than the
16292 @samp{init} mode since it requires all the scalers to be re-computed, but it
16293 allows advanced dynamic expressions.
16296 Default value is @samp{init}.
16299 Set dithering to reduce the circular banding effects. Default is @code{1}
16303 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16304 Setting this value to the SAR of the input will make a rectangular vignetting
16305 following the dimensions of the video.
16307 Default is @code{1/1}.
16310 @subsection Expressions
16312 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16313 following parameters.
16318 input width and height
16321 the number of input frame, starting from 0
16324 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16325 @var{TB} units, NAN if undefined
16328 frame rate of the input video, NAN if the input frame rate is unknown
16331 the PTS (Presentation TimeStamp) of the filtered video frame,
16332 expressed in seconds, NAN if undefined
16335 time base of the input video
16339 @subsection Examples
16343 Apply simple strong vignetting effect:
16349 Make a flickering vignetting:
16351 vignette='PI/4+random(1)*PI/50':eval=frame
16356 @section vmafmotion
16358 Obtain the average vmaf motion score of a video.
16359 It is one of the component filters of VMAF.
16361 The obtained average motion score is printed through the logging system.
16363 In the below example the input file @file{ref.mpg} is being processed and score
16367 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16371 Stack input videos vertically.
16373 All streams must be of same pixel format and of same width.
16375 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16376 to create same output.
16378 The filter accept the following option:
16382 Set number of input streams. Default is 2.
16385 If set to 1, force the output to terminate when the shortest input
16386 terminates. Default value is 0.
16391 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
16392 Deinterlacing Filter").
16394 Based on the process described by Martin Weston for BBC R&D, and
16395 implemented based on the de-interlace algorithm written by Jim
16396 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
16397 uses filter coefficients calculated by BBC R&D.
16399 There are two sets of filter coefficients, so called "simple":
16400 and "complex". Which set of filter coefficients is used can
16401 be set by passing an optional parameter:
16405 Set the interlacing filter coefficients. Accepts one of the following values:
16409 Simple filter coefficient set.
16411 More-complex filter coefficient set.
16413 Default value is @samp{complex}.
16416 Specify which frames to deinterlace. Accept one of the following values:
16420 Deinterlace all frames,
16422 Only deinterlace frames marked as interlaced.
16425 Default value is @samp{all}.
16429 Video waveform monitor.
16431 The waveform monitor plots color component intensity. By default luminance
16432 only. Each column of the waveform corresponds to a column of pixels in the
16435 It accepts the following options:
16439 Can be either @code{row}, or @code{column}. Default is @code{column}.
16440 In row mode, the graph on the left side represents color component value 0 and
16441 the right side represents value = 255. In column mode, the top side represents
16442 color component value = 0 and bottom side represents value = 255.
16445 Set intensity. Smaller values are useful to find out how many values of the same
16446 luminance are distributed across input rows/columns.
16447 Default value is @code{0.04}. Allowed range is [0, 1].
16450 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
16451 In mirrored mode, higher values will be represented on the left
16452 side for @code{row} mode and at the top for @code{column} mode. Default is
16453 @code{1} (mirrored).
16457 It accepts the following values:
16460 Presents information identical to that in the @code{parade}, except
16461 that the graphs representing color components are superimposed directly
16464 This display mode makes it easier to spot relative differences or similarities
16465 in overlapping areas of the color components that are supposed to be identical,
16466 such as neutral whites, grays, or blacks.
16469 Display separate graph for the color components side by side in
16470 @code{row} mode or one below the other in @code{column} mode.
16473 Display separate graph for the color components side by side in
16474 @code{column} mode or one below the other in @code{row} mode.
16476 Using this display mode makes it easy to spot color casts in the highlights
16477 and shadows of an image, by comparing the contours of the top and the bottom
16478 graphs of each waveform. Since whites, grays, and blacks are characterized
16479 by exactly equal amounts of red, green, and blue, neutral areas of the picture
16480 should display three waveforms of roughly equal width/height. If not, the
16481 correction is easy to perform by making level adjustments the three waveforms.
16483 Default is @code{stack}.
16485 @item components, c
16486 Set which color components to display. Default is 1, which means only luminance
16487 or red color component if input is in RGB colorspace. If is set for example to
16488 7 it will display all 3 (if) available color components.
16493 No envelope, this is default.
16496 Instant envelope, minimum and maximum values presented in graph will be easily
16497 visible even with small @code{step} value.
16500 Hold minimum and maximum values presented in graph across time. This way you
16501 can still spot out of range values without constantly looking at waveforms.
16504 Peak and instant envelope combined together.
16510 No filtering, this is default.
16513 Luma and chroma combined together.
16516 Similar as above, but shows difference between blue and red chroma.
16519 Similar as above, but use different colors.
16522 Displays only chroma.
16525 Displays actual color value on waveform.
16528 Similar as above, but with luma showing frequency of chroma values.
16532 Set which graticule to display.
16536 Do not display graticule.
16539 Display green graticule showing legal broadcast ranges.
16542 Display orange graticule showing legal broadcast ranges.
16546 Set graticule opacity.
16549 Set graticule flags.
16553 Draw numbers above lines. By default enabled.
16556 Draw dots instead of lines.
16560 Set scale used for displaying graticule.
16567 Default is digital.
16570 Set background opacity.
16573 @section weave, doubleweave
16575 The @code{weave} takes a field-based video input and join
16576 each two sequential fields into single frame, producing a new double
16577 height clip with half the frame rate and half the frame count.
16579 The @code{doubleweave} works same as @code{weave} but without
16580 halving frame rate and frame count.
16582 It accepts the following option:
16586 Set first field. Available values are:
16590 Set the frame as top-field-first.
16593 Set the frame as bottom-field-first.
16597 @subsection Examples
16601 Interlace video using @ref{select} and @ref{separatefields} filter:
16603 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
16608 Apply the xBR high-quality magnification filter which is designed for pixel
16609 art. It follows a set of edge-detection rules, see
16610 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
16612 It accepts the following option:
16616 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
16617 @code{3xBR} and @code{4} for @code{4xBR}.
16618 Default is @code{3}.
16624 Deinterlace the input video ("yadif" means "yet another deinterlacing
16627 It accepts the following parameters:
16633 The interlacing mode to adopt. It accepts one of the following values:
16636 @item 0, send_frame
16637 Output one frame for each frame.
16638 @item 1, send_field
16639 Output one frame for each field.
16640 @item 2, send_frame_nospatial
16641 Like @code{send_frame}, but it skips the spatial interlacing check.
16642 @item 3, send_field_nospatial
16643 Like @code{send_field}, but it skips the spatial interlacing check.
16646 The default value is @code{send_frame}.
16649 The picture field parity assumed for the input interlaced video. It accepts one
16650 of the following values:
16654 Assume the top field is first.
16656 Assume the bottom field is first.
16658 Enable automatic detection of field parity.
16661 The default value is @code{auto}.
16662 If the interlacing is unknown or the decoder does not export this information,
16663 top field first will be assumed.
16666 Specify which frames to deinterlace. Accept one of the following
16671 Deinterlace all frames.
16672 @item 1, interlaced
16673 Only deinterlace frames marked as interlaced.
16676 The default value is @code{all}.
16681 Apply Zoom & Pan effect.
16683 This filter accepts the following options:
16687 Set the zoom expression. Default is 1.
16691 Set the x and y expression. Default is 0.
16694 Set the duration expression in number of frames.
16695 This sets for how many number of frames effect will last for
16696 single input image.
16699 Set the output image size, default is 'hd720'.
16702 Set the output frame rate, default is '25'.
16705 Each expression can contain the following constants:
16724 Output frame count.
16728 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16729 for current input frame.
16733 'x' and 'y' of last output frame of previous input frame or 0 when there was
16734 not yet such frame (first input frame).
16737 Last calculated zoom from 'z' expression for current input frame.
16740 Last calculated zoom of last output frame of previous input frame.
16743 Number of output frames for current input frame. Calculated from 'd' expression
16744 for each input frame.
16747 number of output frames created for previous input frame
16750 Rational number: input width / input height
16753 sample aspect ratio
16756 display aspect ratio
16760 @subsection Examples
16764 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16766 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
16770 Zoom-in up to 1.5 and pan always at center of picture:
16772 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16776 Same as above but without pausing:
16778 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16784 Scale (resize) the input video, using the z.lib library:
16785 https://github.com/sekrit-twc/zimg.
16787 The zscale filter forces the output display aspect ratio to be the same
16788 as the input, by changing the output sample aspect ratio.
16790 If the input image format is different from the format requested by
16791 the next filter, the zscale filter will convert the input to the
16794 @subsection Options
16795 The filter accepts the following options.
16800 Set the output video dimension expression. Default value is the input
16803 If the @var{width} or @var{w} value is 0, the input width is used for
16804 the output. If the @var{height} or @var{h} value is 0, the input height
16805 is used for the output.
16807 If one and only one of the values is -n with n >= 1, the zscale filter
16808 will use a value that maintains the aspect ratio of the input image,
16809 calculated from the other specified dimension. After that it will,
16810 however, make sure that the calculated dimension is divisible by n and
16811 adjust the value if necessary.
16813 If both values are -n with n >= 1, the behavior will be identical to
16814 both values being set to 0 as previously detailed.
16816 See below for the list of accepted constants for use in the dimension
16820 Set the video size. For the syntax of this option, check the
16821 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16824 Set the dither type.
16826 Possible values are:
16831 @item error_diffusion
16837 Set the resize filter type.
16839 Possible values are:
16849 Default is bilinear.
16852 Set the color range.
16854 Possible values are:
16861 Default is same as input.
16864 Set the color primaries.
16866 Possible values are:
16876 Default is same as input.
16879 Set the transfer characteristics.
16881 Possible values are:
16895 Default is same as input.
16898 Set the colorspace matrix.
16900 Possible value are:
16911 Default is same as input.
16914 Set the input color range.
16916 Possible values are:
16923 Default is same as input.
16925 @item primariesin, pin
16926 Set the input color primaries.
16928 Possible values are:
16938 Default is same as input.
16940 @item transferin, tin
16941 Set the input transfer characteristics.
16943 Possible values are:
16954 Default is same as input.
16956 @item matrixin, min
16957 Set the input colorspace matrix.
16959 Possible value are:
16971 Set the output chroma location.
16973 Possible values are:
16984 @item chromalin, cin
16985 Set the input chroma location.
16987 Possible values are:
16999 Set the nominal peak luminance.
17002 The values of the @option{w} and @option{h} options are expressions
17003 containing the following constants:
17008 The input width and height
17012 These are the same as @var{in_w} and @var{in_h}.
17016 The output (scaled) width and height
17020 These are the same as @var{out_w} and @var{out_h}
17023 The same as @var{iw} / @var{ih}
17026 input sample aspect ratio
17029 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17033 horizontal and vertical input chroma subsample values. For example for the
17034 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17038 horizontal and vertical output chroma subsample values. For example for the
17039 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17045 @c man end VIDEO FILTERS
17047 @chapter Video Sources
17048 @c man begin VIDEO SOURCES
17050 Below is a description of the currently available video sources.
17054 Buffer video frames, and make them available to the filter chain.
17056 This source is mainly intended for a programmatic use, in particular
17057 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17059 It accepts the following parameters:
17064 Specify the size (width and height) of the buffered video frames. For the
17065 syntax of this option, check the
17066 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17069 The input video width.
17072 The input video height.
17075 A string representing the pixel format of the buffered video frames.
17076 It may be a number corresponding to a pixel format, or a pixel format
17080 Specify the timebase assumed by the timestamps of the buffered frames.
17083 Specify the frame rate expected for the video stream.
17085 @item pixel_aspect, sar
17086 The sample (pixel) aspect ratio of the input video.
17089 Specify the optional parameters to be used for the scale filter which
17090 is automatically inserted when an input change is detected in the
17091 input size or format.
17093 @item hw_frames_ctx
17094 When using a hardware pixel format, this should be a reference to an
17095 AVHWFramesContext describing input frames.
17100 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17103 will instruct the source to accept video frames with size 320x240 and
17104 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17105 square pixels (1:1 sample aspect ratio).
17106 Since the pixel format with name "yuv410p" corresponds to the number 6
17107 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17108 this example corresponds to:
17110 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17113 Alternatively, the options can be specified as a flat string, but this
17114 syntax is deprecated:
17116 @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}]
17120 Create a pattern generated by an elementary cellular automaton.
17122 The initial state of the cellular automaton can be defined through the
17123 @option{filename} and @option{pattern} options. If such options are
17124 not specified an initial state is created randomly.
17126 At each new frame a new row in the video is filled with the result of
17127 the cellular automaton next generation. The behavior when the whole
17128 frame is filled is defined by the @option{scroll} option.
17130 This source accepts the following options:
17134 Read the initial cellular automaton state, i.e. the starting row, from
17135 the specified file.
17136 In the file, each non-whitespace character is considered an alive
17137 cell, a newline will terminate the row, and further characters in the
17138 file will be ignored.
17141 Read the initial cellular automaton state, i.e. the starting row, from
17142 the specified string.
17144 Each non-whitespace character in the string is considered an alive
17145 cell, a newline will terminate the row, and further characters in the
17146 string will be ignored.
17149 Set the video rate, that is the number of frames generated per second.
17152 @item random_fill_ratio, ratio
17153 Set the random fill ratio for the initial cellular automaton row. It
17154 is a floating point number value ranging from 0 to 1, defaults to
17157 This option is ignored when a file or a pattern is specified.
17159 @item random_seed, seed
17160 Set the seed for filling randomly the initial row, must be an integer
17161 included between 0 and UINT32_MAX. If not specified, or if explicitly
17162 set to -1, the filter will try to use a good random seed on a best
17166 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17167 Default value is 110.
17170 Set the size of the output video. For the syntax of this option, check the
17171 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17173 If @option{filename} or @option{pattern} is specified, the size is set
17174 by default to the width of the specified initial state row, and the
17175 height is set to @var{width} * PHI.
17177 If @option{size} is set, it must contain the width of the specified
17178 pattern string, and the specified pattern will be centered in the
17181 If a filename or a pattern string is not specified, the size value
17182 defaults to "320x518" (used for a randomly generated initial state).
17185 If set to 1, scroll the output upward when all the rows in the output
17186 have been already filled. If set to 0, the new generated row will be
17187 written over the top row just after the bottom row is filled.
17190 @item start_full, full
17191 If set to 1, completely fill the output with generated rows before
17192 outputting the first frame.
17193 This is the default behavior, for disabling set the value to 0.
17196 If set to 1, stitch the left and right row edges together.
17197 This is the default behavior, for disabling set the value to 0.
17200 @subsection Examples
17204 Read the initial state from @file{pattern}, and specify an output of
17207 cellauto=f=pattern:s=200x400
17211 Generate a random initial row with a width of 200 cells, with a fill
17214 cellauto=ratio=2/3:s=200x200
17218 Create a pattern generated by rule 18 starting by a single alive cell
17219 centered on an initial row with width 100:
17221 cellauto=p=@@:s=100x400:full=0:rule=18
17225 Specify a more elaborated initial pattern:
17227 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17232 @anchor{coreimagesrc}
17233 @section coreimagesrc
17234 Video source generated on GPU using Apple's CoreImage API on OSX.
17236 This video source is a specialized version of the @ref{coreimage} video filter.
17237 Use a core image generator at the beginning of the applied filterchain to
17238 generate the content.
17240 The coreimagesrc video source accepts the following options:
17242 @item list_generators
17243 List all available generators along with all their respective options as well as
17244 possible minimum and maximum values along with the default values.
17246 list_generators=true
17250 Specify the size of the sourced video. For the syntax of this option, check the
17251 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17252 The default value is @code{320x240}.
17255 Specify the frame rate of the sourced video, as the number of frames
17256 generated per second. It has to be a string in the format
17257 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17258 number or a valid video frame rate abbreviation. The default value is
17262 Set the sample aspect ratio of the sourced video.
17265 Set the duration of the sourced video. See
17266 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17267 for the accepted syntax.
17269 If not specified, or the expressed duration is negative, the video is
17270 supposed to be generated forever.
17273 Additionally, all options of the @ref{coreimage} video filter are accepted.
17274 A complete filterchain can be used for further processing of the
17275 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17276 and examples for details.
17278 @subsection Examples
17283 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17284 given as complete and escaped command-line for Apple's standard bash shell:
17286 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17288 This example is equivalent to the QRCode example of @ref{coreimage} without the
17289 need for a nullsrc video source.
17293 @section mandelbrot
17295 Generate a Mandelbrot set fractal, and progressively zoom towards the
17296 point specified with @var{start_x} and @var{start_y}.
17298 This source accepts the following options:
17303 Set the terminal pts value. Default value is 400.
17306 Set the terminal scale value.
17307 Must be a floating point value. Default value is 0.3.
17310 Set the inner coloring mode, that is the algorithm used to draw the
17311 Mandelbrot fractal internal region.
17313 It shall assume one of the following values:
17318 Show time until convergence.
17320 Set color based on point closest to the origin of the iterations.
17325 Default value is @var{mincol}.
17328 Set the bailout value. Default value is 10.0.
17331 Set the maximum of iterations performed by the rendering
17332 algorithm. Default value is 7189.
17335 Set outer coloring mode.
17336 It shall assume one of following values:
17338 @item iteration_count
17339 Set iteration cound mode.
17340 @item normalized_iteration_count
17341 set normalized iteration count mode.
17343 Default value is @var{normalized_iteration_count}.
17346 Set frame rate, expressed as number of frames per second. Default
17350 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17351 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17354 Set the initial scale value. Default value is 3.0.
17357 Set the initial x position. Must be a floating point value between
17358 -100 and 100. Default value is -0.743643887037158704752191506114774.
17361 Set the initial y position. Must be a floating point value between
17362 -100 and 100. Default value is -0.131825904205311970493132056385139.
17367 Generate various test patterns, as generated by the MPlayer test filter.
17369 The size of the generated video is fixed, and is 256x256.
17370 This source is useful in particular for testing encoding features.
17372 This source accepts the following options:
17377 Specify the frame rate of the sourced video, as the number of frames
17378 generated per second. It has to be a string in the format
17379 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17380 number or a valid video frame rate abbreviation. The default value is
17384 Set the duration of the sourced video. See
17385 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17386 for the accepted syntax.
17388 If not specified, or the expressed duration is negative, the video is
17389 supposed to be generated forever.
17393 Set the number or the name of the test to perform. Supported tests are:
17409 Default value is "all", which will cycle through the list of all tests.
17414 mptestsrc=t=dc_luma
17417 will generate a "dc_luma" test pattern.
17419 @section frei0r_src
17421 Provide a frei0r source.
17423 To enable compilation of this filter you need to install the frei0r
17424 header and configure FFmpeg with @code{--enable-frei0r}.
17426 This source accepts the following parameters:
17431 The size of the video to generate. For the syntax of this option, check the
17432 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17435 The framerate of the generated video. It may be a string of the form
17436 @var{num}/@var{den} or a frame rate abbreviation.
17439 The name to the frei0r source to load. For more information regarding frei0r and
17440 how to set the parameters, read the @ref{frei0r} section in the video filters
17443 @item filter_params
17444 A '|'-separated list of parameters to pass to the frei0r source.
17448 For example, to generate a frei0r partik0l source with size 200x200
17449 and frame rate 10 which is overlaid on the overlay filter main input:
17451 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
17456 Generate a life pattern.
17458 This source is based on a generalization of John Conway's life game.
17460 The sourced input represents a life grid, each pixel represents a cell
17461 which can be in one of two possible states, alive or dead. Every cell
17462 interacts with its eight neighbours, which are the cells that are
17463 horizontally, vertically, or diagonally adjacent.
17465 At each interaction the grid evolves according to the adopted rule,
17466 which specifies the number of neighbor alive cells which will make a
17467 cell stay alive or born. The @option{rule} option allows one to specify
17470 This source accepts the following options:
17474 Set the file from which to read the initial grid state. In the file,
17475 each non-whitespace character is considered an alive cell, and newline
17476 is used to delimit the end of each row.
17478 If this option is not specified, the initial grid is generated
17482 Set the video rate, that is the number of frames generated per second.
17485 @item random_fill_ratio, ratio
17486 Set the random fill ratio for the initial random grid. It is a
17487 floating point number value ranging from 0 to 1, defaults to 1/PHI.
17488 It is ignored when a file is specified.
17490 @item random_seed, seed
17491 Set the seed for filling the initial random grid, must be an integer
17492 included between 0 and UINT32_MAX. If not specified, or if explicitly
17493 set to -1, the filter will try to use a good random seed on a best
17499 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
17500 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
17501 @var{NS} specifies the number of alive neighbor cells which make a
17502 live cell stay alive, and @var{NB} the number of alive neighbor cells
17503 which make a dead cell to become alive (i.e. to "born").
17504 "s" and "b" can be used in place of "S" and "B", respectively.
17506 Alternatively a rule can be specified by an 18-bits integer. The 9
17507 high order bits are used to encode the next cell state if it is alive
17508 for each number of neighbor alive cells, the low order bits specify
17509 the rule for "borning" new cells. Higher order bits encode for an
17510 higher number of neighbor cells.
17511 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
17512 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
17514 Default value is "S23/B3", which is the original Conway's game of life
17515 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
17516 cells, and will born a new cell if there are three alive cells around
17520 Set the size of the output video. For the syntax of this option, check the
17521 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17523 If @option{filename} is specified, the size is set by default to the
17524 same size of the input file. If @option{size} is set, it must contain
17525 the size specified in the input file, and the initial grid defined in
17526 that file is centered in the larger resulting area.
17528 If a filename is not specified, the size value defaults to "320x240"
17529 (used for a randomly generated initial grid).
17532 If set to 1, stitch the left and right grid edges together, and the
17533 top and bottom edges also. Defaults to 1.
17536 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
17537 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
17538 value from 0 to 255.
17541 Set the color of living (or new born) cells.
17544 Set the color of dead cells. If @option{mold} is set, this is the first color
17545 used to represent a dead cell.
17548 Set mold color, for definitely dead and moldy cells.
17550 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
17551 ffmpeg-utils manual,ffmpeg-utils}.
17554 @subsection Examples
17558 Read a grid from @file{pattern}, and center it on a grid of size
17561 life=f=pattern:s=300x300
17565 Generate a random grid of size 200x200, with a fill ratio of 2/3:
17567 life=ratio=2/3:s=200x200
17571 Specify a custom rule for evolving a randomly generated grid:
17577 Full example with slow death effect (mold) using @command{ffplay}:
17579 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
17586 @anchor{haldclutsrc}
17588 @anchor{rgbtestsrc}
17590 @anchor{smptehdbars}
17593 @anchor{yuvtestsrc}
17594 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
17596 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
17598 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
17600 The @code{color} source provides an uniformly colored input.
17602 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
17603 @ref{haldclut} filter.
17605 The @code{nullsrc} source returns unprocessed video frames. It is
17606 mainly useful to be employed in analysis / debugging tools, or as the
17607 source for filters which ignore the input data.
17609 The @code{rgbtestsrc} source generates an RGB test pattern useful for
17610 detecting RGB vs BGR issues. You should see a red, green and blue
17611 stripe from top to bottom.
17613 The @code{smptebars} source generates a color bars pattern, based on
17614 the SMPTE Engineering Guideline EG 1-1990.
17616 The @code{smptehdbars} source generates a color bars pattern, based on
17617 the SMPTE RP 219-2002.
17619 The @code{testsrc} source generates a test video pattern, showing a
17620 color pattern, a scrolling gradient and a timestamp. This is mainly
17621 intended for testing purposes.
17623 The @code{testsrc2} source is similar to testsrc, but supports more
17624 pixel formats instead of just @code{rgb24}. This allows using it as an
17625 input for other tests without requiring a format conversion.
17627 The @code{yuvtestsrc} source generates an YUV test pattern. You should
17628 see a y, cb and cr stripe from top to bottom.
17630 The sources accept the following parameters:
17635 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17636 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17637 pixels to be used as identity matrix for 3D lookup tables. Each component is
17638 coded on a @code{1/(N*N)} scale.
17641 Specify the color of the source, only available in the @code{color}
17642 source. For the syntax of this option, check the
17643 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17646 Specify the size of the sourced video. For the syntax of this option, check the
17647 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17648 The default value is @code{320x240}.
17650 This option is not available with the @code{allrgb}, @code{allyuv}, and
17651 @code{haldclutsrc} filters.
17654 Specify the frame rate of the sourced video, as the number of frames
17655 generated per second. It has to be a string in the format
17656 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17657 number or a valid video frame rate abbreviation. The default value is
17661 Set the duration of the sourced video. See
17662 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17663 for the accepted syntax.
17665 If not specified, or the expressed duration is negative, the video is
17666 supposed to be generated forever.
17669 Set the sample aspect ratio of the sourced video.
17672 Specify the alpha (opacity) of the background, only available in the
17673 @code{testsrc2} source. The value must be between 0 (fully transparent) and
17674 255 (fully opaque, the default).
17677 Set the number of decimals to show in the timestamp, only available in the
17678 @code{testsrc} source.
17680 The displayed timestamp value will correspond to the original
17681 timestamp value multiplied by the power of 10 of the specified
17682 value. Default value is 0.
17685 @subsection Examples
17689 Generate a video with a duration of 5.3 seconds, with size
17690 176x144 and a frame rate of 10 frames per second:
17692 testsrc=duration=5.3:size=qcif:rate=10
17696 The following graph description will generate a red source
17697 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17700 color=c=red@@0.2:s=qcif:r=10
17704 If the input content is to be ignored, @code{nullsrc} can be used. The
17705 following command generates noise in the luminance plane by employing
17706 the @code{geq} filter:
17708 nullsrc=s=256x256, geq=random(1)*255:128:128
17712 @subsection Commands
17714 The @code{color} source supports the following commands:
17718 Set the color of the created image. Accepts the same syntax of the
17719 corresponding @option{color} option.
17724 Generate video using an OpenCL program.
17729 OpenCL program source file.
17732 Kernel name in program.
17735 Size of frames to generate. This must be set.
17738 Pixel format to use for the generated frames. This must be set.
17741 Number of frames generated every second. Default value is '25'.
17745 For details of how the program loading works, see the @ref{program_opencl}
17752 Generate a colour ramp by setting pixel values from the position of the pixel
17753 in the output image. (Note that this will work with all pixel formats, but
17754 the generated output will not be the same.)
17756 __kernel void ramp(__write_only image2d_t dst,
17757 unsigned int index)
17759 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17762 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
17764 write_imagef(dst, loc, val);
17769 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
17771 __kernel void sierpinski_carpet(__write_only image2d_t dst,
17772 unsigned int index)
17774 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17776 float4 value = 0.0f;
17777 int x = loc.x + index;
17778 int y = loc.y + index;
17779 while (x > 0 || y > 0) {
17780 if (x % 3 == 1 && y % 3 == 1) {
17788 write_imagef(dst, loc, value);
17794 @c man end VIDEO SOURCES
17796 @chapter Video Sinks
17797 @c man begin VIDEO SINKS
17799 Below is a description of the currently available video sinks.
17801 @section buffersink
17803 Buffer video frames, and make them available to the end of the filter
17806 This sink is mainly intended for programmatic use, in particular
17807 through the interface defined in @file{libavfilter/buffersink.h}
17808 or the options system.
17810 It accepts a pointer to an AVBufferSinkContext structure, which
17811 defines the incoming buffers' formats, to be passed as the opaque
17812 parameter to @code{avfilter_init_filter} for initialization.
17816 Null video sink: do absolutely nothing with the input video. It is
17817 mainly useful as a template and for use in analysis / debugging
17820 @c man end VIDEO SINKS
17822 @chapter Multimedia Filters
17823 @c man begin MULTIMEDIA FILTERS
17825 Below is a description of the currently available multimedia filters.
17829 Convert input audio to a video output, displaying the audio bit scope.
17831 The filter accepts the following options:
17835 Set frame rate, expressed as number of frames per second. Default
17839 Specify the video size for the output. For the syntax of this option, check the
17840 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17841 Default value is @code{1024x256}.
17844 Specify list of colors separated by space or by '|' which will be used to
17845 draw channels. Unrecognized or missing colors will be replaced
17849 @section ahistogram
17851 Convert input audio to a video output, displaying the volume histogram.
17853 The filter accepts the following options:
17857 Specify how histogram is calculated.
17859 It accepts the following values:
17862 Use single histogram for all channels.
17864 Use separate histogram for each channel.
17866 Default is @code{single}.
17869 Set frame rate, expressed as number of frames per second. Default
17873 Specify the video size for the output. For the syntax of this option, check the
17874 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17875 Default value is @code{hd720}.
17880 It accepts the following values:
17891 reverse logarithmic
17893 Default is @code{log}.
17896 Set amplitude scale.
17898 It accepts the following values:
17905 Default is @code{log}.
17908 Set how much frames to accumulate in histogram.
17909 Defauls is 1. Setting this to -1 accumulates all frames.
17912 Set histogram ratio of window height.
17915 Set sonogram sliding.
17917 It accepts the following values:
17920 replace old rows with new ones.
17922 scroll from top to bottom.
17924 Default is @code{replace}.
17927 @section aphasemeter
17929 Convert input audio to a video output, displaying the audio phase.
17931 The filter accepts the following options:
17935 Set the output frame rate. Default value is @code{25}.
17938 Set the video size for the output. For the syntax of this option, check the
17939 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17940 Default value is @code{800x400}.
17945 Specify the red, green, blue contrast. Default values are @code{2},
17946 @code{7} and @code{1}.
17947 Allowed range is @code{[0, 255]}.
17950 Set color which will be used for drawing median phase. If color is
17951 @code{none} which is default, no median phase value will be drawn.
17954 Enable video output. Default is enabled.
17957 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17958 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17959 The @code{-1} means left and right channels are completely out of phase and
17960 @code{1} means channels are in phase.
17962 @section avectorscope
17964 Convert input audio to a video output, representing the audio vector
17967 The filter is used to measure the difference between channels of stereo
17968 audio stream. A monoaural signal, consisting of identical left and right
17969 signal, results in straight vertical line. Any stereo separation is visible
17970 as a deviation from this line, creating a Lissajous figure.
17971 If the straight (or deviation from it) but horizontal line appears this
17972 indicates that the left and right channels are out of phase.
17974 The filter accepts the following options:
17978 Set the vectorscope mode.
17980 Available values are:
17983 Lissajous rotated by 45 degrees.
17986 Same as above but not rotated.
17989 Shape resembling half of circle.
17992 Default value is @samp{lissajous}.
17995 Set the video size for the output. For the syntax of this option, check the
17996 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17997 Default value is @code{400x400}.
18000 Set the output frame rate. Default value is @code{25}.
18006 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18007 @code{160}, @code{80} and @code{255}.
18008 Allowed range is @code{[0, 255]}.
18014 Specify the red, green, blue and alpha fade. Default values are @code{15},
18015 @code{10}, @code{5} and @code{5}.
18016 Allowed range is @code{[0, 255]}.
18019 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18020 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18023 Set the vectorscope drawing mode.
18025 Available values are:
18028 Draw dot for each sample.
18031 Draw line between previous and current sample.
18034 Default value is @samp{dot}.
18037 Specify amplitude scale of audio samples.
18039 Available values are:
18055 Swap left channel axis with right channel axis.
18065 Mirror only x axis.
18068 Mirror only y axis.
18076 @subsection Examples
18080 Complete example using @command{ffplay}:
18082 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18083 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18087 @section bench, abench
18089 Benchmark part of a filtergraph.
18091 The filter accepts the following options:
18095 Start or stop a timer.
18097 Available values are:
18100 Get the current time, set it as frame metadata (using the key
18101 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18104 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18105 the input frame metadata to get the time difference. Time difference, average,
18106 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18107 @code{min}) are then printed. The timestamps are expressed in seconds.
18111 @subsection Examples
18115 Benchmark @ref{selectivecolor} filter:
18117 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18123 Concatenate audio and video streams, joining them together one after the
18126 The filter works on segments of synchronized video and audio streams. All
18127 segments must have the same number of streams of each type, and that will
18128 also be the number of streams at output.
18130 The filter accepts the following options:
18135 Set the number of segments. Default is 2.
18138 Set the number of output video streams, that is also the number of video
18139 streams in each segment. Default is 1.
18142 Set the number of output audio streams, that is also the number of audio
18143 streams in each segment. Default is 0.
18146 Activate unsafe mode: do not fail if segments have a different format.
18150 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18151 @var{a} audio outputs.
18153 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18154 segment, in the same order as the outputs, then the inputs for the second
18157 Related streams do not always have exactly the same duration, for various
18158 reasons including codec frame size or sloppy authoring. For that reason,
18159 related synchronized streams (e.g. a video and its audio track) should be
18160 concatenated at once. The concat filter will use the duration of the longest
18161 stream in each segment (except the last one), and if necessary pad shorter
18162 audio streams with silence.
18164 For this filter to work correctly, all segments must start at timestamp 0.
18166 All corresponding streams must have the same parameters in all segments; the
18167 filtering system will automatically select a common pixel format for video
18168 streams, and a common sample format, sample rate and channel layout for
18169 audio streams, but other settings, such as resolution, must be converted
18170 explicitly by the user.
18172 Different frame rates are acceptable but will result in variable frame rate
18173 at output; be sure to configure the output file to handle it.
18175 @subsection Examples
18179 Concatenate an opening, an episode and an ending, all in bilingual version
18180 (video in stream 0, audio in streams 1 and 2):
18182 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18183 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18184 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18185 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18189 Concatenate two parts, handling audio and video separately, using the
18190 (a)movie sources, and adjusting the resolution:
18192 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18193 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18194 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18196 Note that a desync will happen at the stitch if the audio and video streams
18197 do not have exactly the same duration in the first file.
18201 @subsection Commands
18203 This filter supports the following commands:
18206 Close the current segment and step to the next one
18209 @section drawgraph, adrawgraph
18211 Draw a graph using input video or audio metadata.
18213 It accepts the following parameters:
18217 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18220 Set 1st foreground color expression.
18223 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18226 Set 2nd foreground color expression.
18229 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18232 Set 3rd foreground color expression.
18235 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18238 Set 4th foreground color expression.
18241 Set minimal value of metadata value.
18244 Set maximal value of metadata value.
18247 Set graph background color. Default is white.
18252 Available values for mode is:
18259 Default is @code{line}.
18264 Available values for slide is:
18267 Draw new frame when right border is reached.
18270 Replace old columns with new ones.
18273 Scroll from right to left.
18276 Scroll from left to right.
18279 Draw single picture.
18282 Default is @code{frame}.
18285 Set size of graph video. For the syntax of this option, check the
18286 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18287 The default value is @code{900x256}.
18289 The foreground color expressions can use the following variables:
18292 Minimal value of metadata value.
18295 Maximal value of metadata value.
18298 Current metadata key value.
18301 The color is defined as 0xAABBGGRR.
18304 Example using metadata from @ref{signalstats} filter:
18306 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18309 Example using metadata from @ref{ebur128} filter:
18311 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18317 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18318 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18319 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18320 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18322 The filter also has a video output (see the @var{video} option) with a real
18323 time graph to observe the loudness evolution. The graphic contains the logged
18324 message mentioned above, so it is not printed anymore when this option is set,
18325 unless the verbose logging is set. The main graphing area contains the
18326 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18327 the momentary loudness (400 milliseconds).
18329 More information about the Loudness Recommendation EBU R128 on
18330 @url{http://tech.ebu.ch/loudness}.
18332 The filter accepts the following options:
18337 Activate the video output. The audio stream is passed unchanged whether this
18338 option is set or no. The video stream will be the first output stream if
18339 activated. Default is @code{0}.
18342 Set the video size. This option is for video only. For the syntax of this
18344 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18345 Default and minimum resolution is @code{640x480}.
18348 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18349 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18350 other integer value between this range is allowed.
18353 Set metadata injection. If set to @code{1}, the audio input will be segmented
18354 into 100ms output frames, each of them containing various loudness information
18355 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18357 Default is @code{0}.
18360 Force the frame logging level.
18362 Available values are:
18365 information logging level
18367 verbose logging level
18370 By default, the logging level is set to @var{info}. If the @option{video} or
18371 the @option{metadata} options are set, it switches to @var{verbose}.
18376 Available modes can be cumulated (the option is a @code{flag} type). Possible
18380 Disable any peak mode (default).
18382 Enable sample-peak mode.
18384 Simple peak mode looking for the higher sample value. It logs a message
18385 for sample-peak (identified by @code{SPK}).
18387 Enable true-peak mode.
18389 If enabled, the peak lookup is done on an over-sampled version of the input
18390 stream for better peak accuracy. It logs a message for true-peak.
18391 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
18392 This mode requires a build with @code{libswresample}.
18396 Treat mono input files as "dual mono". If a mono file is intended for playback
18397 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
18398 If set to @code{true}, this option will compensate for this effect.
18399 Multi-channel input files are not affected by this option.
18402 Set a specific pan law to be used for the measurement of dual mono files.
18403 This parameter is optional, and has a default value of -3.01dB.
18406 @subsection Examples
18410 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
18412 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
18416 Run an analysis with @command{ffmpeg}:
18418 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
18422 @section interleave, ainterleave
18424 Temporally interleave frames from several inputs.
18426 @code{interleave} works with video inputs, @code{ainterleave} with audio.
18428 These filters read frames from several inputs and send the oldest
18429 queued frame to the output.
18431 Input streams must have well defined, monotonically increasing frame
18434 In order to submit one frame to output, these filters need to enqueue
18435 at least one frame for each input, so they cannot work in case one
18436 input is not yet terminated and will not receive incoming frames.
18438 For example consider the case when one input is a @code{select} filter
18439 which always drops input frames. The @code{interleave} filter will keep
18440 reading from that input, but it will never be able to send new frames
18441 to output until the input sends an end-of-stream signal.
18443 Also, depending on inputs synchronization, the filters will drop
18444 frames in case one input receives more frames than the other ones, and
18445 the queue is already filled.
18447 These filters accept the following options:
18451 Set the number of different inputs, it is 2 by default.
18454 @subsection Examples
18458 Interleave frames belonging to different streams using @command{ffmpeg}:
18460 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
18464 Add flickering blur effect:
18466 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
18470 @section metadata, ametadata
18472 Manipulate frame metadata.
18474 This filter accepts the following options:
18478 Set mode of operation of the filter.
18480 Can be one of the following:
18484 If both @code{value} and @code{key} is set, select frames
18485 which have such metadata. If only @code{key} is set, select
18486 every frame that has such key in metadata.
18489 Add new metadata @code{key} and @code{value}. If key is already available
18493 Modify value of already present key.
18496 If @code{value} is set, delete only keys that have such value.
18497 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
18501 Print key and its value if metadata was found. If @code{key} is not set print all
18502 metadata values available in frame.
18506 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
18509 Set metadata value which will be used. This option is mandatory for
18510 @code{modify} and @code{add} mode.
18513 Which function to use when comparing metadata value and @code{value}.
18515 Can be one of following:
18519 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
18522 Values are interpreted as strings, returns true if metadata value starts with
18523 the @code{value} option string.
18526 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
18529 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
18532 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
18535 Values are interpreted as floats, returns true if expression from option @code{expr}
18540 Set expression which is used when @code{function} is set to @code{expr}.
18541 The expression is evaluated through the eval API and can contain the following
18546 Float representation of @code{value} from metadata key.
18549 Float representation of @code{value} as supplied by user in @code{value} option.
18553 If specified in @code{print} mode, output is written to the named file. Instead of
18554 plain filename any writable url can be specified. Filename ``-'' is a shorthand
18555 for standard output. If @code{file} option is not set, output is written to the log
18556 with AV_LOG_INFO loglevel.
18560 @subsection Examples
18564 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
18567 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
18570 Print silencedetect output to file @file{metadata.txt}.
18572 silencedetect,ametadata=mode=print:file=metadata.txt
18575 Direct all metadata to a pipe with file descriptor 4.
18577 metadata=mode=print:file='pipe\:4'
18581 @section perms, aperms
18583 Set read/write permissions for the output frames.
18585 These filters are mainly aimed at developers to test direct path in the
18586 following filter in the filtergraph.
18588 The filters accept the following options:
18592 Select the permissions mode.
18594 It accepts the following values:
18597 Do nothing. This is the default.
18599 Set all the output frames read-only.
18601 Set all the output frames directly writable.
18603 Make the frame read-only if writable, and writable if read-only.
18605 Set each output frame read-only or writable randomly.
18609 Set the seed for the @var{random} mode, must be an integer included between
18610 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
18611 @code{-1}, the filter will try to use a good random seed on a best effort
18615 Note: in case of auto-inserted filter between the permission filter and the
18616 following one, the permission might not be received as expected in that
18617 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
18618 perms/aperms filter can avoid this problem.
18620 @section realtime, arealtime
18622 Slow down filtering to match real time approximately.
18624 These filters will pause the filtering for a variable amount of time to
18625 match the output rate with the input timestamps.
18626 They are similar to the @option{re} option to @code{ffmpeg}.
18628 They accept the following options:
18632 Time limit for the pauses. Any pause longer than that will be considered
18633 a timestamp discontinuity and reset the timer. Default is 2 seconds.
18637 @section select, aselect
18639 Select frames to pass in output.
18641 This filter accepts the following options:
18646 Set expression, which is evaluated for each input frame.
18648 If the expression is evaluated to zero, the frame is discarded.
18650 If the evaluation result is negative or NaN, the frame is sent to the
18651 first output; otherwise it is sent to the output with index
18652 @code{ceil(val)-1}, assuming that the input index starts from 0.
18654 For example a value of @code{1.2} corresponds to the output with index
18655 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
18658 Set the number of outputs. The output to which to send the selected
18659 frame is based on the result of the evaluation. Default value is 1.
18662 The expression can contain the following constants:
18666 The (sequential) number of the filtered frame, starting from 0.
18669 The (sequential) number of the selected frame, starting from 0.
18671 @item prev_selected_n
18672 The sequential number of the last selected frame. It's NAN if undefined.
18675 The timebase of the input timestamps.
18678 The PTS (Presentation TimeStamp) of the filtered video frame,
18679 expressed in @var{TB} units. It's NAN if undefined.
18682 The PTS of the filtered video frame,
18683 expressed in seconds. It's NAN if undefined.
18686 The PTS of the previously filtered video frame. It's NAN if undefined.
18688 @item prev_selected_pts
18689 The PTS of the last previously filtered video frame. It's NAN if undefined.
18691 @item prev_selected_t
18692 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
18695 The PTS of the first video frame in the video. It's NAN if undefined.
18698 The time of the first video frame in the video. It's NAN if undefined.
18700 @item pict_type @emph{(video only)}
18701 The type of the filtered frame. It can assume one of the following
18713 @item interlace_type @emph{(video only)}
18714 The frame interlace type. It can assume one of the following values:
18717 The frame is progressive (not interlaced).
18719 The frame is top-field-first.
18721 The frame is bottom-field-first.
18724 @item consumed_sample_n @emph{(audio only)}
18725 the number of selected samples before the current frame
18727 @item samples_n @emph{(audio only)}
18728 the number of samples in the current frame
18730 @item sample_rate @emph{(audio only)}
18731 the input sample rate
18734 This is 1 if the filtered frame is a key-frame, 0 otherwise.
18737 the position in the file of the filtered frame, -1 if the information
18738 is not available (e.g. for synthetic video)
18740 @item scene @emph{(video only)}
18741 value between 0 and 1 to indicate a new scene; a low value reflects a low
18742 probability for the current frame to introduce a new scene, while a higher
18743 value means the current frame is more likely to be one (see the example below)
18745 @item concatdec_select
18746 The concat demuxer can select only part of a concat input file by setting an
18747 inpoint and an outpoint, but the output packets may not be entirely contained
18748 in the selected interval. By using this variable, it is possible to skip frames
18749 generated by the concat demuxer which are not exactly contained in the selected
18752 This works by comparing the frame pts against the @var{lavf.concat.start_time}
18753 and the @var{lavf.concat.duration} packet metadata values which are also
18754 present in the decoded frames.
18756 The @var{concatdec_select} variable is -1 if the frame pts is at least
18757 start_time and either the duration metadata is missing or the frame pts is less
18758 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
18761 That basically means that an input frame is selected if its pts is within the
18762 interval set by the concat demuxer.
18766 The default value of the select expression is "1".
18768 @subsection Examples
18772 Select all frames in input:
18777 The example above is the same as:
18789 Select only I-frames:
18791 select='eq(pict_type\,I)'
18795 Select one frame every 100:
18797 select='not(mod(n\,100))'
18801 Select only frames contained in the 10-20 time interval:
18803 select=between(t\,10\,20)
18807 Select only I-frames contained in the 10-20 time interval:
18809 select=between(t\,10\,20)*eq(pict_type\,I)
18813 Select frames with a minimum distance of 10 seconds:
18815 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18819 Use aselect to select only audio frames with samples number > 100:
18821 aselect='gt(samples_n\,100)'
18825 Create a mosaic of the first scenes:
18827 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18830 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18834 Send even and odd frames to separate outputs, and compose them:
18836 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18840 Select useful frames from an ffconcat file which is using inpoints and
18841 outpoints but where the source files are not intra frame only.
18843 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18847 @section sendcmd, asendcmd
18849 Send commands to filters in the filtergraph.
18851 These filters read commands to be sent to other filters in the
18854 @code{sendcmd} must be inserted between two video filters,
18855 @code{asendcmd} must be inserted between two audio filters, but apart
18856 from that they act the same way.
18858 The specification of commands can be provided in the filter arguments
18859 with the @var{commands} option, or in a file specified by the
18860 @var{filename} option.
18862 These filters accept the following options:
18865 Set the commands to be read and sent to the other filters.
18867 Set the filename of the commands to be read and sent to the other
18871 @subsection Commands syntax
18873 A commands description consists of a sequence of interval
18874 specifications, comprising a list of commands to be executed when a
18875 particular event related to that interval occurs. The occurring event
18876 is typically the current frame time entering or leaving a given time
18879 An interval is specified by the following syntax:
18881 @var{START}[-@var{END}] @var{COMMANDS};
18884 The time interval is specified by the @var{START} and @var{END} times.
18885 @var{END} is optional and defaults to the maximum time.
18887 The current frame time is considered within the specified interval if
18888 it is included in the interval [@var{START}, @var{END}), that is when
18889 the time is greater or equal to @var{START} and is lesser than
18892 @var{COMMANDS} consists of a sequence of one or more command
18893 specifications, separated by ",", relating to that interval. The
18894 syntax of a command specification is given by:
18896 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18899 @var{FLAGS} is optional and specifies the type of events relating to
18900 the time interval which enable sending the specified command, and must
18901 be a non-null sequence of identifier flags separated by "+" or "|" and
18902 enclosed between "[" and "]".
18904 The following flags are recognized:
18907 The command is sent when the current frame timestamp enters the
18908 specified interval. In other words, the command is sent when the
18909 previous frame timestamp was not in the given interval, and the
18913 The command is sent when the current frame timestamp leaves the
18914 specified interval. In other words, the command is sent when the
18915 previous frame timestamp was in the given interval, and the
18919 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18922 @var{TARGET} specifies the target of the command, usually the name of
18923 the filter class or a specific filter instance name.
18925 @var{COMMAND} specifies the name of the command for the target filter.
18927 @var{ARG} is optional and specifies the optional list of argument for
18928 the given @var{COMMAND}.
18930 Between one interval specification and another, whitespaces, or
18931 sequences of characters starting with @code{#} until the end of line,
18932 are ignored and can be used to annotate comments.
18934 A simplified BNF description of the commands specification syntax
18937 @var{COMMAND_FLAG} ::= "enter" | "leave"
18938 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18939 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18940 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18941 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18942 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18945 @subsection Examples
18949 Specify audio tempo change at second 4:
18951 asendcmd=c='4.0 atempo tempo 1.5',atempo
18955 Target a specific filter instance:
18957 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18961 Specify a list of drawtext and hue commands in a file.
18963 # show text in the interval 5-10
18964 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18965 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18967 # desaturate the image in the interval 15-20
18968 15.0-20.0 [enter] hue s 0,
18969 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18971 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18973 # apply an exponential saturation fade-out effect, starting from time 25
18974 25 [enter] hue s exp(25-t)
18977 A filtergraph allowing to read and process the above command list
18978 stored in a file @file{test.cmd}, can be specified with:
18980 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18985 @section setpts, asetpts
18987 Change the PTS (presentation timestamp) of the input frames.
18989 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18991 This filter accepts the following options:
18996 The expression which is evaluated for each frame to construct its timestamp.
19000 The expression is evaluated through the eval API and can contain the following
19005 frame rate, only defined for constant frame-rate video
19008 The presentation timestamp in input
19011 The count of the input frame for video or the number of consumed samples,
19012 not including the current frame for audio, starting from 0.
19014 @item NB_CONSUMED_SAMPLES
19015 The number of consumed samples, not including the current frame (only
19018 @item NB_SAMPLES, S
19019 The number of samples in the current frame (only audio)
19021 @item SAMPLE_RATE, SR
19022 The audio sample rate.
19025 The PTS of the first frame.
19028 the time in seconds of the first frame
19031 State whether the current frame is interlaced.
19034 the time in seconds of the current frame
19037 original position in the file of the frame, or undefined if undefined
19038 for the current frame
19041 The previous input PTS.
19044 previous input time in seconds
19047 The previous output PTS.
19050 previous output time in seconds
19053 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19057 The wallclock (RTC) time at the start of the movie in microseconds.
19060 The timebase of the input timestamps.
19064 @subsection Examples
19068 Start counting PTS from zero
19070 setpts=PTS-STARTPTS
19074 Apply fast motion effect:
19080 Apply slow motion effect:
19086 Set fixed rate of 25 frames per second:
19092 Set fixed rate 25 fps with some jitter:
19094 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19098 Apply an offset of 10 seconds to the input PTS:
19104 Generate timestamps from a "live source" and rebase onto the current timebase:
19106 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19110 Generate timestamps by counting samples:
19119 Force color range for the output video frame.
19121 The @code{setrange} filter marks the color range property for the
19122 output frames. It does not change the input frame, but only sets the
19123 corresponding property, which affects how the frame is treated by
19126 The filter accepts the following options:
19131 Available values are:
19135 Keep the same color range property.
19137 @item unspecified, unknown
19138 Set the color range as unspecified.
19140 @item limited, tv, mpeg
19141 Set the color range as limited.
19143 @item full, pc, jpeg
19144 Set the color range as full.
19148 @section settb, asettb
19150 Set the timebase to use for the output frames timestamps.
19151 It is mainly useful for testing timebase configuration.
19153 It accepts the following parameters:
19158 The expression which is evaluated into the output timebase.
19162 The value for @option{tb} is an arithmetic expression representing a
19163 rational. The expression can contain the constants "AVTB" (the default
19164 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19165 audio only). Default value is "intb".
19167 @subsection Examples
19171 Set the timebase to 1/25:
19177 Set the timebase to 1/10:
19183 Set the timebase to 1001/1000:
19189 Set the timebase to 2*intb:
19195 Set the default timebase value:
19202 Convert input audio to a video output representing frequency spectrum
19203 logarithmically using Brown-Puckette constant Q transform algorithm with
19204 direct frequency domain coefficient calculation (but the transform itself
19205 is not really constant Q, instead the Q factor is actually variable/clamped),
19206 with musical tone scale, from E0 to D#10.
19208 The filter accepts the following options:
19212 Specify the video size for the output. It must be even. For the syntax of this option,
19213 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19214 Default value is @code{1920x1080}.
19217 Set the output frame rate. Default value is @code{25}.
19220 Set the bargraph height. It must be even. Default value is @code{-1} which
19221 computes the bargraph height automatically.
19224 Set the axis height. It must be even. Default value is @code{-1} which computes
19225 the axis height automatically.
19228 Set the sonogram height. It must be even. Default value is @code{-1} which
19229 computes the sonogram height automatically.
19232 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19233 instead. Default value is @code{1}.
19235 @item sono_v, volume
19236 Specify the sonogram volume expression. It can contain variables:
19239 the @var{bar_v} evaluated expression
19240 @item frequency, freq, f
19241 the frequency where it is evaluated
19242 @item timeclamp, tc
19243 the value of @var{timeclamp} option
19247 @item a_weighting(f)
19248 A-weighting of equal loudness
19249 @item b_weighting(f)
19250 B-weighting of equal loudness
19251 @item c_weighting(f)
19252 C-weighting of equal loudness.
19254 Default value is @code{16}.
19256 @item bar_v, volume2
19257 Specify the bargraph volume expression. It can contain variables:
19260 the @var{sono_v} evaluated expression
19261 @item frequency, freq, f
19262 the frequency where it is evaluated
19263 @item timeclamp, tc
19264 the value of @var{timeclamp} option
19268 @item a_weighting(f)
19269 A-weighting of equal loudness
19270 @item b_weighting(f)
19271 B-weighting of equal loudness
19272 @item c_weighting(f)
19273 C-weighting of equal loudness.
19275 Default value is @code{sono_v}.
19277 @item sono_g, gamma
19278 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19279 higher gamma makes the spectrum having more range. Default value is @code{3}.
19280 Acceptable range is @code{[1, 7]}.
19282 @item bar_g, gamma2
19283 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19287 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19288 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19290 @item timeclamp, tc
19291 Specify the transform timeclamp. At low frequency, there is trade-off between
19292 accuracy in time domain and frequency domain. If timeclamp is lower,
19293 event in time domain is represented more accurately (such as fast bass drum),
19294 otherwise event in frequency domain is represented more accurately
19295 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19298 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19299 limits future samples by applying asymmetric windowing in time domain, useful
19300 when low latency is required. Accepted range is @code{[0, 1]}.
19303 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19304 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19307 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19308 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19311 This option is deprecated and ignored.
19314 Specify the transform length in time domain. Use this option to control accuracy
19315 trade-off between time domain and frequency domain at every frequency sample.
19316 It can contain variables:
19318 @item frequency, freq, f
19319 the frequency where it is evaluated
19320 @item timeclamp, tc
19321 the value of @var{timeclamp} option.
19323 Default value is @code{384*tc/(384+tc*f)}.
19326 Specify the transform count for every video frame. Default value is @code{6}.
19327 Acceptable range is @code{[1, 30]}.
19330 Specify the transform count for every single pixel. Default value is @code{0},
19331 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19334 Specify font file for use with freetype to draw the axis. If not specified,
19335 use embedded font. Note that drawing with font file or embedded font is not
19336 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19340 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19341 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19344 Specify font color expression. This is arithmetic expression that should return
19345 integer value 0xRRGGBB. It can contain variables:
19347 @item frequency, freq, f
19348 the frequency where it is evaluated
19349 @item timeclamp, tc
19350 the value of @var{timeclamp} option
19355 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19356 @item r(x), g(x), b(x)
19357 red, green, and blue value of intensity x.
19359 Default value is @code{st(0, (midi(f)-59.5)/12);
19360 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19361 r(1-ld(1)) + b(ld(1))}.
19364 Specify image file to draw the axis. This option override @var{fontfile} and
19365 @var{fontcolor} option.
19368 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19369 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19370 Default value is @code{1}.
19373 Set colorspace. The accepted values are:
19376 Unspecified (default)
19385 BT.470BG or BT.601-6 625
19388 SMPTE-170M or BT.601-6 525
19394 BT.2020 with non-constant luminance
19399 Set spectrogram color scheme. This is list of floating point values with format
19400 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
19401 The default is @code{1|0.5|0|0|0.5|1}.
19405 @subsection Examples
19409 Playing audio while showing the spectrum:
19411 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
19415 Same as above, but with frame rate 30 fps:
19417 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
19421 Playing at 1280x720:
19423 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
19427 Disable sonogram display:
19433 A1 and its harmonics: A1, A2, (near)E3, A3:
19435 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),
19436 asplit[a][out1]; [a] showcqt [out0]'
19440 Same as above, but with more accuracy in frequency domain:
19442 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),
19443 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
19449 bar_v=10:sono_v=bar_v*a_weighting(f)
19453 Custom gamma, now spectrum is linear to the amplitude.
19459 Custom tlength equation:
19461 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)))'
19465 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
19467 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
19471 Custom font using fontconfig:
19473 font='Courier New,Monospace,mono|bold'
19477 Custom frequency range with custom axis using image file:
19479 axisfile=myaxis.png:basefreq=40:endfreq=10000
19485 Convert input audio to video output representing the audio power spectrum.
19486 Audio amplitude is on Y-axis while frequency is on X-axis.
19488 The filter accepts the following options:
19492 Specify size of video. For the syntax of this option, check the
19493 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19494 Default is @code{1024x512}.
19498 This set how each frequency bin will be represented.
19500 It accepts the following values:
19506 Default is @code{bar}.
19509 Set amplitude scale.
19511 It accepts the following values:
19525 Default is @code{log}.
19528 Set frequency scale.
19530 It accepts the following values:
19539 Reverse logarithmic scale.
19541 Default is @code{lin}.
19546 It accepts the following values:
19562 Default is @code{w2048}
19565 Set windowing function.
19567 It accepts the following values:
19589 Default is @code{hanning}.
19592 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19593 which means optimal overlap for selected window function will be picked.
19596 Set time averaging. Setting this to 0 will display current maximal peaks.
19597 Default is @code{1}, which means time averaging is disabled.
19600 Specify list of colors separated by space or by '|' which will be used to
19601 draw channel frequencies. Unrecognized or missing colors will be replaced
19605 Set channel display mode.
19607 It accepts the following values:
19612 Default is @code{combined}.
19615 Set minimum amplitude used in @code{log} amplitude scaler.
19619 @anchor{showspectrum}
19620 @section showspectrum
19622 Convert input audio to a video output, representing the audio frequency
19625 The filter accepts the following options:
19629 Specify the video size for the output. For the syntax of this option, check the
19630 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19631 Default value is @code{640x512}.
19634 Specify how the spectrum should slide along the window.
19636 It accepts the following values:
19639 the samples start again on the left when they reach the right
19641 the samples scroll from right to left
19643 frames are only produced when the samples reach the right
19645 the samples scroll from left to right
19648 Default value is @code{replace}.
19651 Specify display mode.
19653 It accepts the following values:
19656 all channels are displayed in the same row
19658 all channels are displayed in separate rows
19661 Default value is @samp{combined}.
19664 Specify display color mode.
19666 It accepts the following values:
19669 each channel is displayed in a separate color
19671 each channel is displayed using the same color scheme
19673 each channel is displayed using the rainbow color scheme
19675 each channel is displayed using the moreland color scheme
19677 each channel is displayed using the nebulae color scheme
19679 each channel is displayed using the fire color scheme
19681 each channel is displayed using the fiery color scheme
19683 each channel is displayed using the fruit color scheme
19685 each channel is displayed using the cool color scheme
19688 Default value is @samp{channel}.
19691 Specify scale used for calculating intensity color values.
19693 It accepts the following values:
19698 square root, default
19709 Default value is @samp{sqrt}.
19712 Set saturation modifier for displayed colors. Negative values provide
19713 alternative color scheme. @code{0} is no saturation at all.
19714 Saturation must be in [-10.0, 10.0] range.
19715 Default value is @code{1}.
19718 Set window function.
19720 It accepts the following values:
19744 Default value is @code{hann}.
19747 Set orientation of time vs frequency axis. Can be @code{vertical} or
19748 @code{horizontal}. Default is @code{vertical}.
19751 Set ratio of overlap window. Default value is @code{0}.
19752 When value is @code{1} overlap is set to recommended size for specific
19753 window function currently used.
19756 Set scale gain for calculating intensity color values.
19757 Default value is @code{1}.
19760 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
19763 Set color rotation, must be in [-1.0, 1.0] range.
19764 Default value is @code{0}.
19767 The usage is very similar to the showwaves filter; see the examples in that
19770 @subsection Examples
19774 Large window with logarithmic color scaling:
19776 showspectrum=s=1280x480:scale=log
19780 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
19782 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19783 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
19787 @section showspectrumpic
19789 Convert input audio to a single video frame, representing the audio frequency
19792 The filter accepts the following options:
19796 Specify the video size for the output. For the syntax of this option, check the
19797 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19798 Default value is @code{4096x2048}.
19801 Specify display mode.
19803 It accepts the following values:
19806 all channels are displayed in the same row
19808 all channels are displayed in separate rows
19810 Default value is @samp{combined}.
19813 Specify display color mode.
19815 It accepts the following values:
19818 each channel is displayed in a separate color
19820 each channel is displayed using the same color scheme
19822 each channel is displayed using the rainbow color scheme
19824 each channel is displayed using the moreland color scheme
19826 each channel is displayed using the nebulae color scheme
19828 each channel is displayed using the fire color scheme
19830 each channel is displayed using the fiery color scheme
19832 each channel is displayed using the fruit color scheme
19834 each channel is displayed using the cool color scheme
19836 Default value is @samp{intensity}.
19839 Specify scale used for calculating intensity color values.
19841 It accepts the following values:
19846 square root, default
19856 Default value is @samp{log}.
19859 Set saturation modifier for displayed colors. Negative values provide
19860 alternative color scheme. @code{0} is no saturation at all.
19861 Saturation must be in [-10.0, 10.0] range.
19862 Default value is @code{1}.
19865 Set window function.
19867 It accepts the following values:
19890 Default value is @code{hann}.
19893 Set orientation of time vs frequency axis. Can be @code{vertical} or
19894 @code{horizontal}. Default is @code{vertical}.
19897 Set scale gain for calculating intensity color values.
19898 Default value is @code{1}.
19901 Draw time and frequency axes and legends. Default is enabled.
19904 Set color rotation, must be in [-1.0, 1.0] range.
19905 Default value is @code{0}.
19908 @subsection Examples
19912 Extract an audio spectrogram of a whole audio track
19913 in a 1024x1024 picture using @command{ffmpeg}:
19915 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19919 @section showvolume
19921 Convert input audio volume to a video output.
19923 The filter accepts the following options:
19930 Set border width, allowed range is [0, 5]. Default is 1.
19933 Set channel width, allowed range is [80, 8192]. Default is 400.
19936 Set channel height, allowed range is [1, 900]. Default is 20.
19939 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19942 Set volume color expression.
19944 The expression can use the following variables:
19948 Current max volume of channel in dB.
19954 Current channel number, starting from 0.
19958 If set, displays channel names. Default is enabled.
19961 If set, displays volume values. Default is enabled.
19964 Set orientation, can be @code{horizontal} or @code{vertical},
19965 default is @code{horizontal}.
19968 Set step size, allowed range is [0, 5]. Default is 0, which means
19972 Set background opacity, allowed range is [0, 1]. Default is 0.
19975 Set metering mode, can be peak: @code{p} or rms: @code{r},
19976 default is @code{p}.
19981 Convert input audio to a video output, representing the samples waves.
19983 The filter accepts the following options:
19987 Specify the video size for the output. For the syntax of this option, check the
19988 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19989 Default value is @code{600x240}.
19994 Available values are:
19997 Draw a point for each sample.
20000 Draw a vertical line for each sample.
20003 Draw a point for each sample and a line between them.
20006 Draw a centered vertical line for each sample.
20009 Default value is @code{point}.
20012 Set the number of samples which are printed on the same column. A
20013 larger value will decrease the frame rate. Must be a positive
20014 integer. This option can be set only if the value for @var{rate}
20015 is not explicitly specified.
20018 Set the (approximate) output frame rate. This is done by setting the
20019 option @var{n}. Default value is "25".
20021 @item split_channels
20022 Set if channels should be drawn separately or overlap. Default value is 0.
20025 Set colors separated by '|' which are going to be used for drawing of each channel.
20028 Set amplitude scale.
20030 Available values are:
20048 Set the draw mode. This is mostly useful to set for high @var{n}.
20050 Available values are:
20053 Scale pixel values for each drawn sample.
20056 Draw every sample directly.
20059 Default value is @code{scale}.
20062 @subsection Examples
20066 Output the input file audio and the corresponding video representation
20069 amovie=a.mp3,asplit[out0],showwaves[out1]
20073 Create a synthetic signal and show it with showwaves, forcing a
20074 frame rate of 30 frames per second:
20076 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20080 @section showwavespic
20082 Convert input audio to a single video frame, representing the samples waves.
20084 The filter accepts the following options:
20088 Specify the video size for the output. For the syntax of this option, check the
20089 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20090 Default value is @code{600x240}.
20092 @item split_channels
20093 Set if channels should be drawn separately or overlap. Default value is 0.
20096 Set colors separated by '|' which are going to be used for drawing of each channel.
20099 Set amplitude scale.
20101 Available values are:
20119 @subsection Examples
20123 Extract a channel split representation of the wave form of a whole audio track
20124 in a 1024x800 picture using @command{ffmpeg}:
20126 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20130 @section sidedata, asidedata
20132 Delete frame side data, or select frames based on it.
20134 This filter accepts the following options:
20138 Set mode of operation of the filter.
20140 Can be one of the following:
20144 Select every frame with side data of @code{type}.
20147 Delete side data of @code{type}. If @code{type} is not set, delete all side
20153 Set side data type used with all modes. Must be set for @code{select} mode. For
20154 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20155 in @file{libavutil/frame.h}. For example, to choose
20156 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20160 @section spectrumsynth
20162 Sythesize audio from 2 input video spectrums, first input stream represents
20163 magnitude across time and second represents phase across time.
20164 The filter will transform from frequency domain as displayed in videos back
20165 to time domain as presented in audio output.
20167 This filter is primarily created for reversing processed @ref{showspectrum}
20168 filter outputs, but can synthesize sound from other spectrograms too.
20169 But in such case results are going to be poor if the phase data is not
20170 available, because in such cases phase data need to be recreated, usually
20171 its just recreated from random noise.
20172 For best results use gray only output (@code{channel} color mode in
20173 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20174 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20175 @code{data} option. Inputs videos should generally use @code{fullframe}
20176 slide mode as that saves resources needed for decoding video.
20178 The filter accepts the following options:
20182 Specify sample rate of output audio, the sample rate of audio from which
20183 spectrum was generated may differ.
20186 Set number of channels represented in input video spectrums.
20189 Set scale which was used when generating magnitude input spectrum.
20190 Can be @code{lin} or @code{log}. Default is @code{log}.
20193 Set slide which was used when generating inputs spectrums.
20194 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20195 Default is @code{fullframe}.
20198 Set window function used for resynthesis.
20201 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20202 which means optimal overlap for selected window function will be picked.
20205 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20206 Default is @code{vertical}.
20209 @subsection Examples
20213 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20214 then resynthesize videos back to audio with spectrumsynth:
20216 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
20217 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
20218 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20222 @section split, asplit
20224 Split input into several identical outputs.
20226 @code{asplit} works with audio input, @code{split} with video.
20228 The filter accepts a single parameter which specifies the number of outputs. If
20229 unspecified, it defaults to 2.
20231 @subsection Examples
20235 Create two separate outputs from the same input:
20237 [in] split [out0][out1]
20241 To create 3 or more outputs, you need to specify the number of
20244 [in] asplit=3 [out0][out1][out2]
20248 Create two separate outputs from the same input, one cropped and
20251 [in] split [splitout1][splitout2];
20252 [splitout1] crop=100:100:0:0 [cropout];
20253 [splitout2] pad=200:200:100:100 [padout];
20257 Create 5 copies of the input audio with @command{ffmpeg}:
20259 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20265 Receive commands sent through a libzmq client, and forward them to
20266 filters in the filtergraph.
20268 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20269 must be inserted between two video filters, @code{azmq} between two
20272 To enable these filters you need to install the libzmq library and
20273 headers and configure FFmpeg with @code{--enable-libzmq}.
20275 For more information about libzmq see:
20276 @url{http://www.zeromq.org/}
20278 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20279 receives messages sent through a network interface defined by the
20280 @option{bind_address} option.
20282 The received message must be in the form:
20284 @var{TARGET} @var{COMMAND} [@var{ARG}]
20287 @var{TARGET} specifies the target of the command, usually the name of
20288 the filter class or a specific filter instance name.
20290 @var{COMMAND} specifies the name of the command for the target filter.
20292 @var{ARG} is optional and specifies the optional argument list for the
20293 given @var{COMMAND}.
20295 Upon reception, the message is processed and the corresponding command
20296 is injected into the filtergraph. Depending on the result, the filter
20297 will send a reply to the client, adopting the format:
20299 @var{ERROR_CODE} @var{ERROR_REASON}
20303 @var{MESSAGE} is optional.
20305 @subsection Examples
20307 Look at @file{tools/zmqsend} for an example of a zmq client which can
20308 be used to send commands processed by these filters.
20310 Consider the following filtergraph generated by @command{ffplay}
20312 ffplay -dumpgraph 1 -f lavfi "
20313 color=s=100x100:c=red [l];
20314 color=s=100x100:c=blue [r];
20315 nullsrc=s=200x100, zmq [bg];
20316 [bg][l] overlay [bg+l];
20317 [bg+l][r] overlay=x=100 "
20320 To change the color of the left side of the video, the following
20321 command can be used:
20323 echo Parsed_color_0 c yellow | tools/zmqsend
20326 To change the right side:
20328 echo Parsed_color_1 c pink | tools/zmqsend
20331 @c man end MULTIMEDIA FILTERS
20333 @chapter Multimedia Sources
20334 @c man begin MULTIMEDIA SOURCES
20336 Below is a description of the currently available multimedia sources.
20340 This is the same as @ref{movie} source, except it selects an audio
20346 Read audio and/or video stream(s) from a movie container.
20348 It accepts the following parameters:
20352 The name of the resource to read (not necessarily a file; it can also be a
20353 device or a stream accessed through some protocol).
20355 @item format_name, f
20356 Specifies the format assumed for the movie to read, and can be either
20357 the name of a container or an input device. If not specified, the
20358 format is guessed from @var{movie_name} or by probing.
20360 @item seek_point, sp
20361 Specifies the seek point in seconds. The frames will be output
20362 starting from this seek point. The parameter is evaluated with
20363 @code{av_strtod}, so the numerical value may be suffixed by an IS
20364 postfix. The default value is "0".
20367 Specifies the streams to read. Several streams can be specified,
20368 separated by "+". The source will then have as many outputs, in the
20369 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
20370 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
20371 respectively the default (best suited) video and audio stream. Default
20372 is "dv", or "da" if the filter is called as "amovie".
20374 @item stream_index, si
20375 Specifies the index of the video stream to read. If the value is -1,
20376 the most suitable video stream will be automatically selected. The default
20377 value is "-1". Deprecated. If the filter is called "amovie", it will select
20378 audio instead of video.
20381 Specifies how many times to read the stream in sequence.
20382 If the value is 0, the stream will be looped infinitely.
20383 Default value is "1".
20385 Note that when the movie is looped the source timestamps are not
20386 changed, so it will generate non monotonically increasing timestamps.
20388 @item discontinuity
20389 Specifies the time difference between frames above which the point is
20390 considered a timestamp discontinuity which is removed by adjusting the later
20394 It allows overlaying a second video on top of the main input of
20395 a filtergraph, as shown in this graph:
20397 input -----------> deltapts0 --> overlay --> output
20400 movie --> scale--> deltapts1 -------+
20402 @subsection Examples
20406 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
20407 on top of the input labelled "in":
20409 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
20410 [in] setpts=PTS-STARTPTS [main];
20411 [main][over] overlay=16:16 [out]
20415 Read from a video4linux2 device, and overlay it on top of the input
20418 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
20419 [in] setpts=PTS-STARTPTS [main];
20420 [main][over] overlay=16:16 [out]
20424 Read the first video stream and the audio stream with id 0x81 from
20425 dvd.vob; the video is connected to the pad named "video" and the audio is
20426 connected to the pad named "audio":
20428 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
20432 @subsection Commands
20434 Both movie and amovie support the following commands:
20437 Perform seek using "av_seek_frame".
20438 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
20441 @var{stream_index}: If stream_index is -1, a default
20442 stream is selected, and @var{timestamp} is automatically converted
20443 from AV_TIME_BASE units to the stream specific time_base.
20445 @var{timestamp}: Timestamp in AVStream.time_base units
20446 or, if no stream is specified, in AV_TIME_BASE units.
20448 @var{flags}: Flags which select direction and seeking mode.
20452 Get movie duration in AV_TIME_BASE units.
20456 @c man end MULTIMEDIA SOURCES