1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @section Notes on filtergraph escaping
226 Filtergraph description composition entails several levels of
227 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
228 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
229 information about the employed escaping procedure.
231 A first level escaping affects the content of each filter option
232 value, which may contain the special character @code{:} used to
233 separate values, or one of the escaping characters @code{\'}.
235 A second level escaping affects the whole filter description, which
236 may contain the escaping characters @code{\'} or the special
237 characters @code{[],;} used by the filtergraph description.
239 Finally, when you specify a filtergraph on a shell commandline, you
240 need to perform a third level escaping for the shell special
241 characters contained within it.
243 For example, consider the following string to be embedded in
244 the @ref{drawtext} filter description @option{text} value:
246 this is a 'string': may contain one, or more, special characters
249 This string contains the @code{'} special escaping character, and the
250 @code{:} special character, so it needs to be escaped in this way:
252 text=this is a \'string\'\: may contain one, or more, special characters
255 A second level of escaping is required when embedding the filter
256 description in a filtergraph description, in order to escape all the
257 filtergraph special characters. Thus the example above becomes:
259 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
261 (note that in addition to the @code{\'} escaping special characters,
262 also @code{,} needs to be escaped).
264 Finally an additional level of escaping is needed when writing the
265 filtergraph description in a shell command, which depends on the
266 escaping rules of the adopted shell. For example, assuming that
267 @code{\} is special and needs to be escaped with another @code{\}, the
268 previous string will finally result in:
270 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
273 @chapter Timeline editing
275 Some filters support a generic @option{enable} option. For the filters
276 supporting timeline editing, this option can be set to an expression which is
277 evaluated before sending a frame to the filter. If the evaluation is non-zero,
278 the filter will be enabled, otherwise the frame will be sent unchanged to the
279 next filter in the filtergraph.
281 The expression accepts the following values:
284 timestamp expressed in seconds, NAN if the input timestamp is unknown
287 sequential number of the input frame, starting from 0
290 the position in the file of the input frame, NAN if unknown
294 width and height of the input frame if video
297 Additionally, these filters support an @option{enable} command that can be used
298 to re-define the expression.
300 Like any other filtering option, the @option{enable} option follows the same
303 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
304 minutes, and a @ref{curves} filter starting at 3 seconds:
306 smartblur = enable='between(t,10,3*60)',
307 curves = enable='gte(t,3)' : preset=cross_process
310 See @code{ffmpeg -filters} to view which filters have timeline support.
312 @c man end FILTERGRAPH DESCRIPTION
315 @chapter Options for filters with several inputs (framesync)
316 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
318 Some filters with several inputs support a common set of options.
319 These options can only be set by name, not with the short notation.
323 The action to take when EOF is encountered on the secondary input; it accepts
324 one of the following values:
328 Repeat the last frame (the default).
332 Pass the main input through.
336 If set to 1, force the output to terminate when the shortest input
337 terminates. Default value is 0.
340 If set to 1, force the filter to extend the last frame of secondary streams
341 until the end of the primary stream. A value of 0 disables this behavior.
345 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
347 @chapter Audio Filters
348 @c man begin AUDIO FILTERS
350 When you configure your FFmpeg build, you can disable any of the
351 existing filters using @code{--disable-filters}.
352 The configure output will show the audio filters included in your
355 Below is a description of the currently available audio filters.
359 A compressor is mainly used to reduce the dynamic range of a signal.
360 Especially modern music is mostly compressed at a high ratio to
361 improve the overall loudness. It's done to get the highest attention
362 of a listener, "fatten" the sound and bring more "power" to the track.
363 If a signal is compressed too much it may sound dull or "dead"
364 afterwards or it may start to "pump" (which could be a powerful effect
365 but can also destroy a track completely).
366 The right compression is the key to reach a professional sound and is
367 the high art of mixing and mastering. Because of its complex settings
368 it may take a long time to get the right feeling for this kind of effect.
370 Compression is done by detecting the volume above a chosen level
371 @code{threshold} and dividing it by the factor set with @code{ratio}.
372 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
373 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
374 the signal would cause distortion of the waveform the reduction can be
375 levelled over the time. This is done by setting "Attack" and "Release".
376 @code{attack} determines how long the signal has to rise above the threshold
377 before any reduction will occur and @code{release} sets the time the signal
378 has to fall below the threshold to reduce the reduction again. Shorter signals
379 than the chosen attack time will be left untouched.
380 The overall reduction of the signal can be made up afterwards with the
381 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
382 raising the makeup to this level results in a signal twice as loud than the
383 source. To gain a softer entry in the compression the @code{knee} flattens the
384 hard edge at the threshold in the range of the chosen decibels.
386 The filter accepts the following options:
390 Set input gain. Default is 1. Range is between 0.015625 and 64.
393 If a signal of stream rises above this level it will affect the gain
395 By default it is 0.125. Range is between 0.00097563 and 1.
398 Set a ratio by which the signal is reduced. 1:2 means that if the level
399 rose 4dB above the threshold, it will be only 2dB above after the reduction.
400 Default is 2. Range is between 1 and 20.
403 Amount of milliseconds the signal has to rise above the threshold before gain
404 reduction starts. Default is 20. Range is between 0.01 and 2000.
407 Amount of milliseconds the signal has to fall below the threshold before
408 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
411 Set the amount by how much signal will be amplified after processing.
412 Default is 1. Range is from 1 to 64.
415 Curve the sharp knee around the threshold to enter gain reduction more softly.
416 Default is 2.82843. Range is between 1 and 8.
419 Choose if the @code{average} level between all channels of input stream
420 or the louder(@code{maximum}) channel of input stream affects the
421 reduction. Default is @code{average}.
424 Should the exact signal be taken in case of @code{peak} or an RMS one in case
425 of @code{rms}. Default is @code{rms} which is mostly smoother.
428 How much to use compressed signal in output. Default is 1.
429 Range is between 0 and 1.
433 Simple audio dynamic range commpression/expansion filter.
435 The filter accepts the following options:
439 Set contrast. Default is 33. Allowed range is between 0 and 100.
444 Copy the input audio source unchanged to the output. This is mainly useful for
449 Apply cross fade from one input audio stream to another input audio stream.
450 The cross fade is applied for specified duration near the end of first stream.
452 The filter accepts the following options:
456 Specify the number of samples for which the cross fade effect has to last.
457 At the end of the cross fade effect the first input audio will be completely
458 silent. Default is 44100.
461 Specify the duration of the cross fade effect. See
462 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
463 for the accepted syntax.
464 By default the duration is determined by @var{nb_samples}.
465 If set this option is used instead of @var{nb_samples}.
468 Should first stream end overlap with second stream start. Default is enabled.
471 Set curve for cross fade transition for first stream.
474 Set curve for cross fade transition for second stream.
476 For description of available curve types see @ref{afade} filter description.
483 Cross fade from one input to another:
485 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
489 Cross fade from one input to another but without overlapping:
491 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Reduce audio bit resolution.
499 This filter is bit crusher with enhanced functionality. A bit crusher
500 is used to audibly reduce number of bits an audio signal is sampled
501 with. This doesn't change the bit depth at all, it just produces the
502 effect. Material reduced in bit depth sounds more harsh and "digital".
503 This filter is able to even round to continuous values instead of discrete
505 Additionally it has a D/C offset which results in different crushing of
506 the lower and the upper half of the signal.
507 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
509 Another feature of this filter is the logarithmic mode.
510 This setting switches from linear distances between bits to logarithmic ones.
511 The result is a much more "natural" sounding crusher which doesn't gate low
512 signals for example. The human ear has a logarithmic perception, too
513 so this kind of crushing is much more pleasant.
514 Logarithmic crushing is also able to get anti-aliased.
516 The filter accepts the following options:
532 Can be linear: @code{lin} or logarithmic: @code{log}.
541 Set sample reduction.
544 Enable LFO. By default disabled.
555 Delay one or more audio channels.
557 Samples in delayed channel are filled with silence.
559 The filter accepts the following option:
563 Set list of delays in milliseconds for each channel separated by '|'.
564 Unused delays will be silently ignored. If number of given delays is
565 smaller than number of channels all remaining channels will not be delayed.
566 If you want to delay exact number of samples, append 'S' to number.
573 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
574 the second channel (and any other channels that may be present) unchanged.
580 Delay second channel by 500 samples, the third channel by 700 samples and leave
581 the first channel (and any other channels that may be present) unchanged.
589 Apply echoing to the input audio.
591 Echoes are reflected sound and can occur naturally amongst mountains
592 (and sometimes large buildings) when talking or shouting; digital echo
593 effects emulate this behaviour and are often used to help fill out the
594 sound of a single instrument or vocal. The time difference between the
595 original signal and the reflection is the @code{delay}, and the
596 loudness of the reflected signal is the @code{decay}.
597 Multiple echoes can have different delays and decays.
599 A description of the accepted parameters follows.
603 Set input gain of reflected signal. Default is @code{0.6}.
606 Set output gain of reflected signal. Default is @code{0.3}.
609 Set list of time intervals in milliseconds between original signal and reflections
610 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
611 Default is @code{1000}.
614 Set list of loudness of reflected signals separated by '|'.
615 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
616 Default is @code{0.5}.
623 Make it sound as if there are twice as many instruments as are actually playing:
625 aecho=0.8:0.88:60:0.4
629 If delay is very short, then it sound like a (metallic) robot playing music:
635 A longer delay will sound like an open air concert in the mountains:
637 aecho=0.8:0.9:1000:0.3
641 Same as above but with one more mountain:
643 aecho=0.8:0.9:1000|1800:0.3|0.25
648 Audio emphasis filter creates or restores material directly taken from LPs or
649 emphased CDs with different filter curves. E.g. to store music on vinyl the
650 signal has to be altered by a filter first to even out the disadvantages of
651 this recording medium.
652 Once the material is played back the inverse filter has to be applied to
653 restore the distortion of the frequency response.
655 The filter accepts the following options:
665 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
666 use @code{production} mode. Default is @code{reproduction} mode.
669 Set filter type. Selects medium. Can be one of the following:
681 select Compact Disc (CD).
687 select 50µs (FM-KF).
689 select 75µs (FM-KF).
695 Modify an audio signal according to the specified expressions.
697 This filter accepts one or more expressions (one for each channel),
698 which are evaluated and used to modify a corresponding audio signal.
700 It accepts the following parameters:
704 Set the '|'-separated expressions list for each separate channel. If
705 the number of input channels is greater than the number of
706 expressions, the last specified expression is used for the remaining
709 @item channel_layout, c
710 Set output channel layout. If not specified, the channel layout is
711 specified by the number of expressions. If set to @samp{same}, it will
712 use by default the same input channel layout.
715 Each expression in @var{exprs} can contain the following constants and functions:
719 channel number of the current expression
722 number of the evaluated sample, starting from 0
728 time of the evaluated sample expressed in seconds
731 @item nb_out_channels
732 input and output number of channels
735 the value of input channel with number @var{CH}
738 Note: this filter is slow. For faster processing you should use a
747 aeval=val(ch)/2:c=same
751 Invert phase of the second channel:
760 Apply fade-in/out effect to input audio.
762 A description of the accepted parameters follows.
766 Specify the effect type, can be either @code{in} for fade-in, or
767 @code{out} for a fade-out effect. Default is @code{in}.
769 @item start_sample, ss
770 Specify the number of the start sample for starting to apply the fade
771 effect. Default is 0.
774 Specify the number of samples for which the fade effect has to last. At
775 the end of the fade-in effect the output audio will have the same
776 volume as the input audio, at the end of the fade-out transition
777 the output audio will be silence. Default is 44100.
780 Specify the start time of the fade effect. Default is 0.
781 The value must be specified as a time duration; see
782 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
783 for the accepted syntax.
784 If set this option is used instead of @var{start_sample}.
787 Specify the duration of the fade effect. See
788 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
789 for the accepted syntax.
790 At the end of the fade-in effect the output audio will have the same
791 volume as the input audio, at the end of the fade-out transition
792 the output audio will be silence.
793 By default the duration is determined by @var{nb_samples}.
794 If set this option is used instead of @var{nb_samples}.
797 Set curve for fade transition.
799 It accepts the following values:
802 select triangular, linear slope (default)
804 select quarter of sine wave
806 select half of sine wave
808 select exponential sine wave
812 select inverted parabola
826 select inverted quarter of sine wave
828 select inverted half of sine wave
830 select double-exponential seat
832 select double-exponential sigmoid
840 Fade in first 15 seconds of audio:
846 Fade out last 25 seconds of a 900 seconds audio:
848 afade=t=out:st=875:d=25
853 Apply arbitrary expressions to samples in frequency domain.
857 Set frequency domain real expression for each separate channel separated
858 by '|'. Default is "1".
859 If the number of input channels is greater than the number of
860 expressions, the last specified expression is used for the remaining
864 Set frequency domain imaginary expression for each separate channel
865 separated by '|'. If not set, @var{real} option is used.
867 Each expression in @var{real} and @var{imag} can contain the following
875 current frequency bin number
878 number of available bins
881 channel number of the current expression
893 It accepts the following values:
909 Default is @code{w4096}
912 Set window function. Default is @code{hann}.
915 Set window overlap. If set to 1, the recommended overlap for selected
916 window function will be picked. Default is @code{0.75}.
923 Leave almost only low frequencies in audio:
925 afftfilt="1-clip((b/nb)*b,0,1)"
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 For example, assuming a 5.1+downmix input MOV file,
2193 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2195 will create an output WAV file tagged as stereo from the downmix channels of
2198 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2200 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2203 @section channelsplit
2205 Split each channel from an input audio stream into a separate output stream.
2207 It accepts the following parameters:
2209 @item channel_layout
2210 The channel layout of the input stream. The default is "stereo".
2213 For example, assuming a stereo input MP3 file,
2215 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2217 will create an output Matroska file with two audio streams, one containing only
2218 the left channel and the other the right channel.
2220 Split a 5.1 WAV file into per-channel files:
2222 ffmpeg -i in.wav -filter_complex
2223 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2224 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2225 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2230 Add a chorus effect to the audio.
2232 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2234 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2235 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2236 The modulation depth defines the range the modulated delay is played before or after
2237 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2238 sound tuned around the original one, like in a chorus where some vocals are slightly
2241 It accepts the following parameters:
2244 Set input gain. Default is 0.4.
2247 Set output gain. Default is 0.4.
2250 Set delays. A typical delay is around 40ms to 60ms.
2262 @subsection Examples
2268 chorus=0.7:0.9:55:0.4:0.25:2
2274 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2278 Fuller sounding chorus with three delays:
2280 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
2285 Compress or expand the audio's dynamic range.
2287 It accepts the following parameters:
2293 A list of times in seconds for each channel over which the instantaneous level
2294 of the input signal is averaged to determine its volume. @var{attacks} refers to
2295 increase of volume and @var{decays} refers to decrease of volume. For most
2296 situations, the attack time (response to the audio getting louder) should be
2297 shorter than the decay time, because the human ear is more sensitive to sudden
2298 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2299 a typical value for decay is 0.8 seconds.
2300 If specified number of attacks & decays is lower than number of channels, the last
2301 set attack/decay will be used for all remaining channels.
2304 A list of points for the transfer function, specified in dB relative to the
2305 maximum possible signal amplitude. Each key points list must be defined using
2306 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2307 @code{x0/y0 x1/y1 x2/y2 ....}
2309 The input values must be in strictly increasing order but the transfer function
2310 does not have to be monotonically rising. The point @code{0/0} is assumed but
2311 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2312 function are @code{-70/-70|-60/-20|1/0}.
2315 Set the curve radius in dB for all joints. It defaults to 0.01.
2318 Set the additional gain in dB to be applied at all points on the transfer
2319 function. This allows for easy adjustment of the overall gain.
2323 Set an initial volume, in dB, to be assumed for each channel when filtering
2324 starts. This permits the user to supply a nominal level initially, so that, for
2325 example, a very large gain is not applied to initial signal levels before the
2326 companding has begun to operate. A typical value for audio which is initially
2327 quiet is -90 dB. It defaults to 0.
2330 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2331 delayed before being fed to the volume adjuster. Specifying a delay
2332 approximately equal to the attack/decay times allows the filter to effectively
2333 operate in predictive rather than reactive mode. It defaults to 0.
2337 @subsection Examples
2341 Make music with both quiet and loud passages suitable for listening to in a
2344 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2347 Another example for audio with whisper and explosion parts:
2349 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2353 A noise gate for when the noise is at a lower level than the signal:
2355 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2359 Here is another noise gate, this time for when the noise is at a higher level
2360 than the signal (making it, in some ways, similar to squelch):
2362 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2366 2:1 compression starting at -6dB:
2368 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2372 2:1 compression starting at -9dB:
2374 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2378 2:1 compression starting at -12dB:
2380 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2384 2:1 compression starting at -18dB:
2386 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2390 3:1 compression starting at -15dB:
2392 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2398 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2404 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
2408 Hard limiter at -6dB:
2410 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2414 Hard limiter at -12dB:
2416 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2420 Hard noise gate at -35 dB:
2422 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2428 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2432 @section compensationdelay
2434 Compensation Delay Line is a metric based delay to compensate differing
2435 positions of microphones or speakers.
2437 For example, you have recorded guitar with two microphones placed in
2438 different location. Because the front of sound wave has fixed speed in
2439 normal conditions, the phasing of microphones can vary and depends on
2440 their location and interposition. The best sound mix can be achieved when
2441 these microphones are in phase (synchronized). Note that distance of
2442 ~30 cm between microphones makes one microphone to capture signal in
2443 antiphase to another microphone. That makes the final mix sounding moody.
2444 This filter helps to solve phasing problems by adding different delays
2445 to each microphone track and make them synchronized.
2447 The best result can be reached when you take one track as base and
2448 synchronize other tracks one by one with it.
2449 Remember that synchronization/delay tolerance depends on sample rate, too.
2450 Higher sample rates will give more tolerance.
2452 It accepts the following parameters:
2456 Set millimeters distance. This is compensation distance for fine tuning.
2460 Set cm distance. This is compensation distance for tightening distance setup.
2464 Set meters distance. This is compensation distance for hard distance setup.
2468 Set dry amount. Amount of unprocessed (dry) signal.
2472 Set wet amount. Amount of processed (wet) signal.
2476 Set temperature degree in Celsius. This is the temperature of the environment.
2481 Apply headphone crossfeed filter.
2483 Crossfeed is the process of blending the left and right channels of stereo
2485 It is mainly used to reduce extreme stereo separation of low frequencies.
2487 The intent is to produce more speaker like sound to the listener.
2489 The filter accepts the following options:
2493 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2494 This sets gain of low shelf filter for side part of stereo image.
2495 Default is -6dB. Max allowed is -30db when strength is set to 1.
2498 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2499 This sets cut off frequency of low shelf filter. Default is cut off near
2500 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2503 Set input gain. Default is 0.9.
2506 Set output gain. Default is 1.
2509 @section crystalizer
2510 Simple algorithm to expand audio dynamic range.
2512 The filter accepts the following options:
2516 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2517 (unchanged sound) to 10.0 (maximum effect).
2520 Enable clipping. By default is enabled.
2524 Apply a DC shift to the audio.
2526 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2527 in the recording chain) from the audio. The effect of a DC offset is reduced
2528 headroom and hence volume. The @ref{astats} filter can be used to determine if
2529 a signal has a DC offset.
2533 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2537 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2538 used to prevent clipping.
2542 Measure audio dynamic range.
2544 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2545 is found in transition material. And anything less that 8 have very poor dynamics
2546 and is very compressed.
2548 The filter accepts the following options:
2552 Set window length in seconds used to split audio into segments of equal length.
2553 Default is 3 seconds.
2557 Dynamic Audio Normalizer.
2559 This filter applies a certain amount of gain to the input audio in order
2560 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2561 contrast to more "simple" normalization algorithms, the Dynamic Audio
2562 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2563 This allows for applying extra gain to the "quiet" sections of the audio
2564 while avoiding distortions or clipping the "loud" sections. In other words:
2565 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2566 sections, in the sense that the volume of each section is brought to the
2567 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2568 this goal *without* applying "dynamic range compressing". It will retain 100%
2569 of the dynamic range *within* each section of the audio file.
2573 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2574 Default is 500 milliseconds.
2575 The Dynamic Audio Normalizer processes the input audio in small chunks,
2576 referred to as frames. This is required, because a peak magnitude has no
2577 meaning for just a single sample value. Instead, we need to determine the
2578 peak magnitude for a contiguous sequence of sample values. While a "standard"
2579 normalizer would simply use the peak magnitude of the complete file, the
2580 Dynamic Audio Normalizer determines the peak magnitude individually for each
2581 frame. The length of a frame is specified in milliseconds. By default, the
2582 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2583 been found to give good results with most files.
2584 Note that the exact frame length, in number of samples, will be determined
2585 automatically, based on the sampling rate of the individual input audio file.
2588 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2589 number. Default is 31.
2590 Probably the most important parameter of the Dynamic Audio Normalizer is the
2591 @code{window size} of the Gaussian smoothing filter. The filter's window size
2592 is specified in frames, centered around the current frame. For the sake of
2593 simplicity, this must be an odd number. Consequently, the default value of 31
2594 takes into account the current frame, as well as the 15 preceding frames and
2595 the 15 subsequent frames. Using a larger window results in a stronger
2596 smoothing effect and thus in less gain variation, i.e. slower gain
2597 adaptation. Conversely, using a smaller window results in a weaker smoothing
2598 effect and thus in more gain variation, i.e. faster gain adaptation.
2599 In other words, the more you increase this value, the more the Dynamic Audio
2600 Normalizer will behave like a "traditional" normalization filter. On the
2601 contrary, the more you decrease this value, the more the Dynamic Audio
2602 Normalizer will behave like a dynamic range compressor.
2605 Set the target peak value. This specifies the highest permissible magnitude
2606 level for the normalized audio input. This filter will try to approach the
2607 target peak magnitude as closely as possible, but at the same time it also
2608 makes sure that the normalized signal will never exceed the peak magnitude.
2609 A frame's maximum local gain factor is imposed directly by the target peak
2610 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2611 It is not recommended to go above this value.
2614 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2615 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2616 factor for each input frame, i.e. the maximum gain factor that does not
2617 result in clipping or distortion. The maximum gain factor is determined by
2618 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2619 additionally bounds the frame's maximum gain factor by a predetermined
2620 (global) maximum gain factor. This is done in order to avoid excessive gain
2621 factors in "silent" or almost silent frames. By default, the maximum gain
2622 factor is 10.0, For most inputs the default value should be sufficient and
2623 it usually is not recommended to increase this value. Though, for input
2624 with an extremely low overall volume level, it may be necessary to allow even
2625 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2626 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2627 Instead, a "sigmoid" threshold function will be applied. This way, the
2628 gain factors will smoothly approach the threshold value, but never exceed that
2632 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2633 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2634 This means that the maximum local gain factor for each frame is defined
2635 (only) by the frame's highest magnitude sample. This way, the samples can
2636 be amplified as much as possible without exceeding the maximum signal
2637 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2638 Normalizer can also take into account the frame's root mean square,
2639 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2640 determine the power of a time-varying signal. It is therefore considered
2641 that the RMS is a better approximation of the "perceived loudness" than
2642 just looking at the signal's peak magnitude. Consequently, by adjusting all
2643 frames to a constant RMS value, a uniform "perceived loudness" can be
2644 established. If a target RMS value has been specified, a frame's local gain
2645 factor is defined as the factor that would result in exactly that RMS value.
2646 Note, however, that the maximum local gain factor is still restricted by the
2647 frame's highest magnitude sample, in order to prevent clipping.
2650 Enable channels coupling. By default is enabled.
2651 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2652 amount. This means the same gain factor will be applied to all channels, i.e.
2653 the maximum possible gain factor is determined by the "loudest" channel.
2654 However, in some recordings, it may happen that the volume of the different
2655 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2656 In this case, this option can be used to disable the channel coupling. This way,
2657 the gain factor will be determined independently for each channel, depending
2658 only on the individual channel's highest magnitude sample. This allows for
2659 harmonizing the volume of the different channels.
2662 Enable DC bias correction. By default is disabled.
2663 An audio signal (in the time domain) is a sequence of sample values.
2664 In the Dynamic Audio Normalizer these sample values are represented in the
2665 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2666 audio signal, or "waveform", should be centered around the zero point.
2667 That means if we calculate the mean value of all samples in a file, or in a
2668 single frame, then the result should be 0.0 or at least very close to that
2669 value. If, however, there is a significant deviation of the mean value from
2670 0.0, in either positive or negative direction, this is referred to as a
2671 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2672 Audio Normalizer provides optional DC bias correction.
2673 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2674 the mean value, or "DC correction" offset, of each input frame and subtract
2675 that value from all of the frame's sample values which ensures those samples
2676 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2677 boundaries, the DC correction offset values will be interpolated smoothly
2678 between neighbouring frames.
2681 Enable alternative boundary mode. By default is disabled.
2682 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2683 around each frame. This includes the preceding frames as well as the
2684 subsequent frames. However, for the "boundary" frames, located at the very
2685 beginning and at the very end of the audio file, not all neighbouring
2686 frames are available. In particular, for the first few frames in the audio
2687 file, the preceding frames are not known. And, similarly, for the last few
2688 frames in the audio file, the subsequent frames are not known. Thus, the
2689 question arises which gain factors should be assumed for the missing frames
2690 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2691 to deal with this situation. The default boundary mode assumes a gain factor
2692 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2693 "fade out" at the beginning and at the end of the input, respectively.
2696 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2697 By default, the Dynamic Audio Normalizer does not apply "traditional"
2698 compression. This means that signal peaks will not be pruned and thus the
2699 full dynamic range will be retained within each local neighbourhood. However,
2700 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2701 normalization algorithm with a more "traditional" compression.
2702 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2703 (thresholding) function. If (and only if) the compression feature is enabled,
2704 all input frames will be processed by a soft knee thresholding function prior
2705 to the actual normalization process. Put simply, the thresholding function is
2706 going to prune all samples whose magnitude exceeds a certain threshold value.
2707 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2708 value. Instead, the threshold value will be adjusted for each individual
2710 In general, smaller parameters result in stronger compression, and vice versa.
2711 Values below 3.0 are not recommended, because audible distortion may appear.
2716 Make audio easier to listen to on headphones.
2718 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2719 so that when listened to on headphones the stereo image is moved from
2720 inside your head (standard for headphones) to outside and in front of
2721 the listener (standard for speakers).
2727 Apply a two-pole peaking equalisation (EQ) filter. With this
2728 filter, the signal-level at and around a selected frequency can
2729 be increased or decreased, whilst (unlike bandpass and bandreject
2730 filters) that at all other frequencies is unchanged.
2732 In order to produce complex equalisation curves, this filter can
2733 be given several times, each with a different central frequency.
2735 The filter accepts the following options:
2739 Set the filter's central frequency in Hz.
2742 Set method to specify band-width of filter.
2757 Specify the band-width of a filter in width_type units.
2760 Set the required gain or attenuation in dB.
2761 Beware of clipping when using a positive gain.
2764 Specify which channels to filter, by default all available are filtered.
2767 @subsection Examples
2770 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2772 equalizer=f=1000:t=h:width=200:g=-10
2776 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2778 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2782 @subsection Commands
2784 This filter supports the following commands:
2787 Change equalizer frequency.
2788 Syntax for the command is : "@var{frequency}"
2791 Change equalizer width_type.
2792 Syntax for the command is : "@var{width_type}"
2795 Change equalizer width.
2796 Syntax for the command is : "@var{width}"
2799 Change equalizer gain.
2800 Syntax for the command is : "@var{gain}"
2803 @section extrastereo
2805 Linearly increases the difference between left and right channels which
2806 adds some sort of "live" effect to playback.
2808 The filter accepts the following options:
2812 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2813 (average of both channels), with 1.0 sound will be unchanged, with
2814 -1.0 left and right channels will be swapped.
2817 Enable clipping. By default is enabled.
2820 @section firequalizer
2821 Apply FIR Equalization using arbitrary frequency response.
2823 The filter accepts the following option:
2827 Set gain curve equation (in dB). The expression can contain variables:
2830 the evaluated frequency
2834 channel number, set to 0 when multichannels evaluation is disabled
2836 channel id, see libavutil/channel_layout.h, set to the first channel id when
2837 multichannels evaluation is disabled
2841 channel_layout, see libavutil/channel_layout.h
2846 @item gain_interpolate(f)
2847 interpolate gain on frequency f based on gain_entry
2848 @item cubic_interpolate(f)
2849 same as gain_interpolate, but smoother
2851 This option is also available as command. Default is @code{gain_interpolate(f)}.
2854 Set gain entry for gain_interpolate function. The expression can
2858 store gain entry at frequency f with value g
2860 This option is also available as command.
2863 Set filter delay in seconds. Higher value means more accurate.
2864 Default is @code{0.01}.
2867 Set filter accuracy in Hz. Lower value means more accurate.
2868 Default is @code{5}.
2871 Set window function. Acceptable values are:
2874 rectangular window, useful when gain curve is already smooth
2876 hann window (default)
2882 3-terms continuous 1st derivative nuttall window
2884 minimum 3-terms discontinuous nuttall window
2886 4-terms continuous 1st derivative nuttall window
2888 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2890 blackman-harris window
2896 If enabled, use fixed number of audio samples. This improves speed when
2897 filtering with large delay. Default is disabled.
2900 Enable multichannels evaluation on gain. Default is disabled.
2903 Enable zero phase mode by subtracting timestamp to compensate delay.
2904 Default is disabled.
2907 Set scale used by gain. Acceptable values are:
2910 linear frequency, linear gain
2912 linear frequency, logarithmic (in dB) gain (default)
2914 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2916 logarithmic frequency, logarithmic gain
2920 Set file for dumping, suitable for gnuplot.
2923 Set scale for dumpfile. Acceptable values are same with scale option.
2927 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2928 Default is disabled.
2931 Enable minimum phase impulse response. Default is disabled.
2934 @subsection Examples
2939 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2942 lowpass at 1000 Hz with gain_entry:
2944 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2947 custom equalization:
2949 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2952 higher delay with zero phase to compensate delay:
2954 firequalizer=delay=0.1:fixed=on:zero_phase=on
2957 lowpass on left channel, highpass on right channel:
2959 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2960 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2965 Apply a flanging effect to the audio.
2967 The filter accepts the following options:
2971 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2974 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
2977 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2981 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2982 Default value is 71.
2985 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2988 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2989 Default value is @var{sinusoidal}.
2992 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2993 Default value is 25.
2996 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2997 Default is @var{linear}.
3001 Apply Haas effect to audio.
3003 Note that this makes most sense to apply on mono signals.
3004 With this filter applied to mono signals it give some directionality and
3005 stretches its stereo image.
3007 The filter accepts the following options:
3011 Set input level. By default is @var{1}, or 0dB
3014 Set output level. By default is @var{1}, or 0dB.
3017 Set gain applied to side part of signal. By default is @var{1}.
3020 Set kind of middle source. Can be one of the following:
3030 Pick middle part signal of stereo image.
3033 Pick side part signal of stereo image.
3037 Change middle phase. By default is disabled.
3040 Set left channel delay. By default is @var{2.05} milliseconds.
3043 Set left channel balance. By default is @var{-1}.
3046 Set left channel gain. By default is @var{1}.
3049 Change left phase. By default is disabled.
3052 Set right channel delay. By defaults is @var{2.12} milliseconds.
3055 Set right channel balance. By default is @var{1}.
3058 Set right channel gain. By default is @var{1}.
3061 Change right phase. By default is enabled.
3066 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3067 embedded HDCD codes is expanded into a 20-bit PCM stream.
3069 The filter supports the Peak Extend and Low-level Gain Adjustment features
3070 of HDCD, and detects the Transient Filter flag.
3073 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3076 When using the filter with wav, note the default encoding for wav is 16-bit,
3077 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3078 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3080 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3081 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3084 The filter accepts the following options:
3087 @item disable_autoconvert
3088 Disable any automatic format conversion or resampling in the filter graph.
3090 @item process_stereo
3091 Process the stereo channels together. If target_gain does not match between
3092 channels, consider it invalid and use the last valid target_gain.
3095 Set the code detect timer period in ms.
3098 Always extend peaks above -3dBFS even if PE isn't signaled.
3101 Replace audio with a solid tone and adjust the amplitude to signal some
3102 specific aspect of the decoding process. The output file can be loaded in
3103 an audio editor alongside the original to aid analysis.
3105 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3112 Gain adjustment level at each sample
3114 Samples where peak extend occurs
3116 Samples where the code detect timer is active
3118 Samples where the target gain does not match between channels
3124 Apply head-related transfer functions (HRTFs) to create virtual
3125 loudspeakers around the user for binaural listening via headphones.
3126 The HRIRs are provided via additional streams, for each channel
3127 one stereo input stream is needed.
3129 The filter accepts the following options:
3133 Set mapping of input streams for convolution.
3134 The argument is a '|'-separated list of channel names in order as they
3135 are given as additional stream inputs for filter.
3136 This also specify number of input streams. Number of input streams
3137 must be not less than number of channels in first stream plus one.
3140 Set gain applied to audio. Value is in dB. Default is 0.
3143 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3144 processing audio in time domain which is slow.
3145 @var{freq} is processing audio in frequency domain which is fast.
3146 Default is @var{freq}.
3149 Set custom gain for LFE channels. Value is in dB. Default is 0.
3152 @subsection Examples
3156 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3157 each amovie filter use stereo file with IR coefficients as input.
3158 The files give coefficients for each position of virtual loudspeaker:
3160 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"
3167 Apply a high-pass filter with 3dB point frequency.
3168 The filter can be either single-pole, or double-pole (the default).
3169 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3171 The filter accepts the following options:
3175 Set frequency in Hz. Default is 3000.
3178 Set number of poles. Default is 2.
3181 Set method to specify band-width of filter.
3196 Specify the band-width of a filter in width_type units.
3197 Applies only to double-pole filter.
3198 The default is 0.707q and gives a Butterworth response.
3201 Specify which channels to filter, by default all available are filtered.
3204 @subsection Commands
3206 This filter supports the following commands:
3209 Change highpass frequency.
3210 Syntax for the command is : "@var{frequency}"
3213 Change highpass width_type.
3214 Syntax for the command is : "@var{width_type}"
3217 Change highpass width.
3218 Syntax for the command is : "@var{width}"
3223 Join multiple input streams into one multi-channel stream.
3225 It accepts the following parameters:
3229 The number of input streams. It defaults to 2.
3231 @item channel_layout
3232 The desired output channel layout. It defaults to stereo.
3235 Map channels from inputs to output. The argument is a '|'-separated list of
3236 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3237 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3238 can be either the name of the input channel (e.g. FL for front left) or its
3239 index in the specified input stream. @var{out_channel} is the name of the output
3243 The filter will attempt to guess the mappings when they are not specified
3244 explicitly. It does so by first trying to find an unused matching input channel
3245 and if that fails it picks the first unused input channel.
3247 Join 3 inputs (with properly set channel layouts):
3249 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3252 Build a 5.1 output from 6 single-channel streams:
3254 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3255 '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'
3261 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3263 To enable compilation of this filter you need to configure FFmpeg with
3264 @code{--enable-ladspa}.
3268 Specifies the name of LADSPA plugin library to load. If the environment
3269 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3270 each one of the directories specified by the colon separated list in
3271 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3272 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3273 @file{/usr/lib/ladspa/}.
3276 Specifies the plugin within the library. Some libraries contain only
3277 one plugin, but others contain many of them. If this is not set filter
3278 will list all available plugins within the specified library.
3281 Set the '|' separated list of controls which are zero or more floating point
3282 values that determine the behavior of the loaded plugin (for example delay,
3284 Controls need to be defined using the following syntax:
3285 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3286 @var{valuei} is the value set on the @var{i}-th control.
3287 Alternatively they can be also defined using the following syntax:
3288 @var{value0}|@var{value1}|@var{value2}|..., where
3289 @var{valuei} is the value set on the @var{i}-th control.
3290 If @option{controls} is set to @code{help}, all available controls and
3291 their valid ranges are printed.
3293 @item sample_rate, s
3294 Specify the sample rate, default to 44100. Only used if plugin have
3298 Set the number of samples per channel per each output frame, default
3299 is 1024. Only used if plugin have zero inputs.
3302 Set the minimum duration of the sourced audio. See
3303 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3304 for the accepted syntax.
3305 Note that the resulting duration may be greater than the specified duration,
3306 as the generated audio is always cut at the end of a complete frame.
3307 If not specified, or the expressed duration is negative, the audio is
3308 supposed to be generated forever.
3309 Only used if plugin have zero inputs.
3313 @subsection Examples
3317 List all available plugins within amp (LADSPA example plugin) library:
3323 List all available controls and their valid ranges for @code{vcf_notch}
3324 plugin from @code{VCF} library:
3326 ladspa=f=vcf:p=vcf_notch:c=help
3330 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3333 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3337 Add reverberation to the audio using TAP-plugins
3338 (Tom's Audio Processing plugins):
3340 ladspa=file=tap_reverb:tap_reverb
3344 Generate white noise, with 0.2 amplitude:
3346 ladspa=file=cmt:noise_source_white:c=c0=.2
3350 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3351 @code{C* Audio Plugin Suite} (CAPS) library:
3353 ladspa=file=caps:Click:c=c1=20'
3357 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3359 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3363 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3364 @code{SWH Plugins} collection:
3366 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3370 Attenuate low frequencies using Multiband EQ from Steve Harris
3371 @code{SWH Plugins} collection:
3373 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3377 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3380 ladspa=caps:Narrower
3384 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3386 ladspa=caps:White:.2
3390 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3392 ladspa=caps:Fractal:c=c1=1
3396 Dynamic volume normalization using @code{VLevel} plugin:
3398 ladspa=vlevel-ladspa:vlevel_mono
3402 @subsection Commands
3404 This filter supports the following commands:
3407 Modify the @var{N}-th control value.
3409 If the specified value is not valid, it is ignored and prior one is kept.
3414 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3415 Support for both single pass (livestreams, files) and double pass (files) modes.
3416 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3417 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3418 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3420 The filter accepts the following options:
3424 Set integrated loudness target.
3425 Range is -70.0 - -5.0. Default value is -24.0.
3428 Set loudness range target.
3429 Range is 1.0 - 20.0. Default value is 7.0.
3432 Set maximum true peak.
3433 Range is -9.0 - +0.0. Default value is -2.0.
3435 @item measured_I, measured_i
3436 Measured IL of input file.
3437 Range is -99.0 - +0.0.
3439 @item measured_LRA, measured_lra
3440 Measured LRA of input file.
3441 Range is 0.0 - 99.0.
3443 @item measured_TP, measured_tp
3444 Measured true peak of input file.
3445 Range is -99.0 - +99.0.
3447 @item measured_thresh
3448 Measured threshold of input file.
3449 Range is -99.0 - +0.0.
3452 Set offset gain. Gain is applied before the true-peak limiter.
3453 Range is -99.0 - +99.0. Default is +0.0.
3456 Normalize linearly if possible.
3457 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3458 to be specified in order to use this mode.
3459 Options are true or false. Default is true.
3462 Treat mono input files as "dual-mono". If a mono file is intended for playback
3463 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3464 If set to @code{true}, this option will compensate for this effect.
3465 Multi-channel input files are not affected by this option.
3466 Options are true or false. Default is false.
3469 Set print format for stats. Options are summary, json, or none.
3470 Default value is none.
3475 Apply a low-pass filter with 3dB point frequency.
3476 The filter can be either single-pole or double-pole (the default).
3477 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3479 The filter accepts the following options:
3483 Set frequency in Hz. Default is 500.
3486 Set number of poles. Default is 2.
3489 Set method to specify band-width of filter.
3504 Specify the band-width of a filter in width_type units.
3505 Applies only to double-pole filter.
3506 The default is 0.707q and gives a Butterworth response.
3509 Specify which channels to filter, by default all available are filtered.
3512 @subsection Examples
3515 Lowpass only LFE channel, it LFE is not present it does nothing:
3521 @subsection Commands
3523 This filter supports the following commands:
3526 Change lowpass frequency.
3527 Syntax for the command is : "@var{frequency}"
3530 Change lowpass width_type.
3531 Syntax for the command is : "@var{width_type}"
3534 Change lowpass width.
3535 Syntax for the command is : "@var{width}"
3540 Load a LV2 (LADSPA Version 2) plugin.
3542 To enable compilation of this filter you need to configure FFmpeg with
3543 @code{--enable-lv2}.
3547 Specifies the plugin URI. You may need to escape ':'.
3550 Set the '|' separated list of controls which are zero or more floating point
3551 values that determine the behavior of the loaded plugin (for example delay,
3553 If @option{controls} is set to @code{help}, all available controls and
3554 their valid ranges are printed.
3556 @item sample_rate, s
3557 Specify the sample rate, default to 44100. Only used if plugin have
3561 Set the number of samples per channel per each output frame, default
3562 is 1024. Only used if plugin have zero inputs.
3565 Set the minimum duration of the sourced audio. See
3566 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3567 for the accepted syntax.
3568 Note that the resulting duration may be greater than the specified duration,
3569 as the generated audio is always cut at the end of a complete frame.
3570 If not specified, or the expressed duration is negative, the audio is
3571 supposed to be generated forever.
3572 Only used if plugin have zero inputs.
3575 @subsection Examples
3579 Apply bass enhancer plugin from Calf:
3581 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3585 Apply bass vinyl plugin from Calf:
3587 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3591 Apply bit crusher plugin from ArtyFX:
3593 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3598 Multiband Compress or expand the audio's dynamic range.
3600 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3601 This is akin to the crossover of a loudspeaker, and results in flat frequency
3602 response when absent compander action.
3604 It accepts the following parameters:
3608 This option syntax is:
3609 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3610 For explanation of each item refer to compand filter documentation.
3616 Mix channels with specific gain levels. The filter accepts the output
3617 channel layout followed by a set of channels definitions.
3619 This filter is also designed to efficiently remap the channels of an audio
3622 The filter accepts parameters of the form:
3623 "@var{l}|@var{outdef}|@var{outdef}|..."
3627 output channel layout or number of channels
3630 output channel specification, of the form:
3631 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3634 output channel to define, either a channel name (FL, FR, etc.) or a channel
3635 number (c0, c1, etc.)
3638 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3641 input channel to use, see out_name for details; it is not possible to mix
3642 named and numbered input channels
3645 If the `=' in a channel specification is replaced by `<', then the gains for
3646 that specification will be renormalized so that the total is 1, thus
3647 avoiding clipping noise.
3649 @subsection Mixing examples
3651 For example, if you want to down-mix from stereo to mono, but with a bigger
3652 factor for the left channel:
3654 pan=1c|c0=0.9*c0+0.1*c1
3657 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3658 7-channels surround:
3660 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3663 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3664 that should be preferred (see "-ac" option) unless you have very specific
3667 @subsection Remapping examples
3669 The channel remapping will be effective if, and only if:
3672 @item gain coefficients are zeroes or ones,
3673 @item only one input per channel output,
3676 If all these conditions are satisfied, the filter will notify the user ("Pure
3677 channel mapping detected"), and use an optimized and lossless method to do the
3680 For example, if you have a 5.1 source and want a stereo audio stream by
3681 dropping the extra channels:
3683 pan="stereo| c0=FL | c1=FR"
3686 Given the same source, you can also switch front left and front right channels
3687 and keep the input channel layout:
3689 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3692 If the input is a stereo audio stream, you can mute the front left channel (and
3693 still keep the stereo channel layout) with:
3698 Still with a stereo audio stream input, you can copy the right channel in both
3699 front left and right:
3701 pan="stereo| c0=FR | c1=FR"
3706 ReplayGain scanner filter. This filter takes an audio stream as an input and
3707 outputs it unchanged.
3708 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3712 Convert the audio sample format, sample rate and channel layout. It is
3713 not meant to be used directly.
3716 Apply time-stretching and pitch-shifting with librubberband.
3718 The filter accepts the following options:
3722 Set tempo scale factor.
3725 Set pitch scale factor.
3728 Set transients detector.
3729 Possible values are:
3738 Possible values are:
3747 Possible values are:
3754 Set processing window size.
3755 Possible values are:
3764 Possible values are:
3771 Enable formant preservation when shift pitching.
3772 Possible values are:
3780 Possible values are:
3789 Possible values are:
3796 @section sidechaincompress
3798 This filter acts like normal compressor but has the ability to compress
3799 detected signal using second input signal.
3800 It needs two input streams and returns one output stream.
3801 First input stream will be processed depending on second stream signal.
3802 The filtered signal then can be filtered with other filters in later stages of
3803 processing. See @ref{pan} and @ref{amerge} filter.
3805 The filter accepts the following options:
3809 Set input gain. Default is 1. Range is between 0.015625 and 64.
3812 If a signal of second stream raises above this level it will affect the gain
3813 reduction of first stream.
3814 By default is 0.125. Range is between 0.00097563 and 1.
3817 Set a ratio about which the signal is reduced. 1:2 means that if the level
3818 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3819 Default is 2. Range is between 1 and 20.
3822 Amount of milliseconds the signal has to rise above the threshold before gain
3823 reduction starts. Default is 20. Range is between 0.01 and 2000.
3826 Amount of milliseconds the signal has to fall below the threshold before
3827 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3830 Set the amount by how much signal will be amplified after processing.
3831 Default is 1. Range is from 1 to 64.
3834 Curve the sharp knee around the threshold to enter gain reduction more softly.
3835 Default is 2.82843. Range is between 1 and 8.
3838 Choose if the @code{average} level between all channels of side-chain stream
3839 or the louder(@code{maximum}) channel of side-chain stream affects the
3840 reduction. Default is @code{average}.
3843 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3844 of @code{rms}. Default is @code{rms} which is mainly smoother.
3847 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3850 How much to use compressed signal in output. Default is 1.
3851 Range is between 0 and 1.
3854 @subsection Examples
3858 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3859 depending on the signal of 2nd input and later compressed signal to be
3860 merged with 2nd input:
3862 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3866 @section sidechaingate
3868 A sidechain gate acts like a normal (wideband) gate but has the ability to
3869 filter the detected signal before sending it to the gain reduction stage.
3870 Normally a gate uses the full range signal to detect a level above the
3872 For example: If you cut all lower frequencies from your sidechain signal
3873 the gate will decrease the volume of your track only if not enough highs
3874 appear. With this technique you are able to reduce the resonation of a
3875 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3877 It needs two input streams and returns one output stream.
3878 First input stream will be processed depending on second stream signal.
3880 The filter accepts the following options:
3884 Set input level before filtering.
3885 Default is 1. Allowed range is from 0.015625 to 64.
3888 Set the level of gain reduction when the signal is below the threshold.
3889 Default is 0.06125. Allowed range is from 0 to 1.
3892 If a signal rises above this level the gain reduction is released.
3893 Default is 0.125. Allowed range is from 0 to 1.
3896 Set a ratio about which the signal is reduced.
3897 Default is 2. Allowed range is from 1 to 9000.
3900 Amount of milliseconds the signal has to rise above the threshold before gain
3902 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3905 Amount of milliseconds the signal has to fall below the threshold before the
3906 reduction is increased again. Default is 250 milliseconds.
3907 Allowed range is from 0.01 to 9000.
3910 Set amount of amplification of signal after processing.
3911 Default is 1. Allowed range is from 1 to 64.
3914 Curve the sharp knee around the threshold to enter gain reduction more softly.
3915 Default is 2.828427125. Allowed range is from 1 to 8.
3918 Choose if exact signal should be taken for detection or an RMS like one.
3919 Default is rms. Can be peak or rms.
3922 Choose if the average level between all channels or the louder channel affects
3924 Default is average. Can be average or maximum.
3927 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3930 @section silencedetect
3932 Detect silence in an audio stream.
3934 This filter logs a message when it detects that the input audio volume is less
3935 or equal to a noise tolerance value for a duration greater or equal to the
3936 minimum detected noise duration.
3938 The printed times and duration are expressed in seconds.
3940 The filter accepts the following options:
3944 Set silence duration until notification (default is 2 seconds).
3947 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3948 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3951 @subsection Examples
3955 Detect 5 seconds of silence with -50dB noise tolerance:
3957 silencedetect=n=-50dB:d=5
3961 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3962 tolerance in @file{silence.mp3}:
3964 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3968 @section silenceremove
3970 Remove silence from the beginning, middle or end of the audio.
3972 The filter accepts the following options:
3976 This value is used to indicate if audio should be trimmed at beginning of
3977 the audio. A value of zero indicates no silence should be trimmed from the
3978 beginning. When specifying a non-zero value, it trims audio up until it
3979 finds non-silence. Normally, when trimming silence from beginning of audio
3980 the @var{start_periods} will be @code{1} but it can be increased to higher
3981 values to trim all audio up to specific count of non-silence periods.
3982 Default value is @code{0}.
3984 @item start_duration
3985 Specify the amount of time that non-silence must be detected before it stops
3986 trimming audio. By increasing the duration, bursts of noises can be treated
3987 as silence and trimmed off. Default value is @code{0}.
3989 @item start_threshold
3990 This indicates what sample value should be treated as silence. For digital
3991 audio, a value of @code{0} may be fine but for audio recorded from analog,
3992 you may wish to increase the value to account for background noise.
3993 Can be specified in dB (in case "dB" is appended to the specified value)
3994 or amplitude ratio. Default value is @code{0}.
3997 Set the count for trimming silence from the end of audio.
3998 To remove silence from the middle of a file, specify a @var{stop_periods}
3999 that is negative. This value is then treated as a positive value and is
4000 used to indicate the effect should restart processing as specified by
4001 @var{start_periods}, making it suitable for removing periods of silence
4002 in the middle of the audio.
4003 Default value is @code{0}.
4006 Specify a duration of silence that must exist before audio is not copied any
4007 more. By specifying a higher duration, silence that is wanted can be left in
4009 Default value is @code{0}.
4011 @item stop_threshold
4012 This is the same as @option{start_threshold} but for trimming silence from
4014 Can be specified in dB (in case "dB" is appended to the specified value)
4015 or amplitude ratio. Default value is @code{0}.
4018 This indicates that @var{stop_duration} length of audio should be left intact
4019 at the beginning of each period of silence.
4020 For example, if you want to remove long pauses between words but do not want
4021 to remove the pauses completely. Default value is @code{0}.
4024 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4025 and works better with digital silence which is exactly 0.
4026 Default value is @code{rms}.
4029 Set ratio used to calculate size of window for detecting silence.
4030 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4033 @subsection Examples
4037 The following example shows how this filter can be used to start a recording
4038 that does not contain the delay at the start which usually occurs between
4039 pressing the record button and the start of the performance:
4041 silenceremove=1:5:0.02
4045 Trim all silence encountered from beginning to end where there is more than 1
4046 second of silence in audio:
4048 silenceremove=0:0:0:-1:1:-90dB
4054 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4055 loudspeakers around the user for binaural listening via headphones (audio
4056 formats up to 9 channels supported).
4057 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4058 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4059 Austrian Academy of Sciences.
4061 To enable compilation of this filter you need to configure FFmpeg with
4062 @code{--enable-libmysofa}.
4064 The filter accepts the following options:
4068 Set the SOFA file used for rendering.
4071 Set gain applied to audio. Value is in dB. Default is 0.
4074 Set rotation of virtual loudspeakers in deg. Default is 0.
4077 Set elevation of virtual speakers in deg. Default is 0.
4080 Set distance in meters between loudspeakers and the listener with near-field
4081 HRTFs. Default is 1.
4084 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4085 processing audio in time domain which is slow.
4086 @var{freq} is processing audio in frequency domain which is fast.
4087 Default is @var{freq}.
4090 Set custom positions of virtual loudspeakers. Syntax for this option is:
4091 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4092 Each virtual loudspeaker is described with short channel name following with
4093 azimuth and elevation in degrees.
4094 Each virtual loudspeaker description is separated by '|'.
4095 For example to override front left and front right channel positions use:
4096 'speakers=FL 45 15|FR 345 15'.
4097 Descriptions with unrecognised channel names are ignored.
4100 Set custom gain for LFE channels. Value is in dB. Default is 0.
4103 @subsection Examples
4107 Using ClubFritz6 sofa file:
4109 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4113 Using ClubFritz12 sofa file and bigger radius with small rotation:
4115 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4119 Similar as above but with custom speaker positions for front left, front right, back left and back right
4120 and also with custom gain:
4122 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4126 @section stereotools
4128 This filter has some handy utilities to manage stereo signals, for converting
4129 M/S stereo recordings to L/R signal while having control over the parameters
4130 or spreading the stereo image of master track.
4132 The filter accepts the following options:
4136 Set input level before filtering for both channels. Defaults is 1.
4137 Allowed range is from 0.015625 to 64.
4140 Set output level after filtering for both channels. Defaults is 1.
4141 Allowed range is from 0.015625 to 64.
4144 Set input balance between both channels. Default is 0.
4145 Allowed range is from -1 to 1.
4148 Set output balance between both channels. Default is 0.
4149 Allowed range is from -1 to 1.
4152 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4153 clipping. Disabled by default.
4156 Mute the left channel. Disabled by default.
4159 Mute the right channel. Disabled by default.
4162 Change the phase of the left channel. Disabled by default.
4165 Change the phase of the right channel. Disabled by default.
4168 Set stereo mode. Available values are:
4172 Left/Right to Left/Right, this is default.
4175 Left/Right to Mid/Side.
4178 Mid/Side to Left/Right.
4181 Left/Right to Left/Left.
4184 Left/Right to Right/Right.
4187 Left/Right to Left + Right.
4190 Left/Right to Right/Left.
4193 Mid/Side to Left/Left.
4196 Mid/Side to Right/Right.
4200 Set level of side signal. Default is 1.
4201 Allowed range is from 0.015625 to 64.
4204 Set balance of side signal. Default is 0.
4205 Allowed range is from -1 to 1.
4208 Set level of the middle signal. Default is 1.
4209 Allowed range is from 0.015625 to 64.
4212 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4215 Set stereo base between mono and inversed channels. Default is 0.
4216 Allowed range is from -1 to 1.
4219 Set delay in milliseconds how much to delay left from right channel and
4220 vice versa. Default is 0. Allowed range is from -20 to 20.
4223 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4226 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4228 @item bmode_in, bmode_out
4229 Set balance mode for balance_in/balance_out option.
4231 Can be one of the following:
4235 Classic balance mode. Attenuate one channel at time.
4236 Gain is raised up to 1.
4239 Similar as classic mode above but gain is raised up to 2.
4242 Equal power distribution, from -6dB to +6dB range.
4246 @subsection Examples
4250 Apply karaoke like effect:
4252 stereotools=mlev=0.015625
4256 Convert M/S signal to L/R:
4258 "stereotools=mode=ms>lr"
4262 @section stereowiden
4264 This filter enhance the stereo effect by suppressing signal common to both
4265 channels and by delaying the signal of left into right and vice versa,
4266 thereby widening the stereo effect.
4268 The filter accepts the following options:
4272 Time in milliseconds of the delay of left signal into right and vice versa.
4273 Default is 20 milliseconds.
4276 Amount of gain in delayed signal into right and vice versa. Gives a delay
4277 effect of left signal in right output and vice versa which gives widening
4278 effect. Default is 0.3.
4281 Cross feed of left into right with inverted phase. This helps in suppressing
4282 the mono. If the value is 1 it will cancel all the signal common to both
4283 channels. Default is 0.3.
4286 Set level of input signal of original channel. Default is 0.8.
4289 @section superequalizer
4290 Apply 18 band equalizer.
4292 The filter accepts the following options:
4299 Set 131Hz band gain.
4301 Set 185Hz band gain.
4303 Set 262Hz band gain.
4305 Set 370Hz band gain.
4307 Set 523Hz band gain.
4309 Set 740Hz band gain.
4311 Set 1047Hz band gain.
4313 Set 1480Hz band gain.
4315 Set 2093Hz band gain.
4317 Set 2960Hz band gain.
4319 Set 4186Hz band gain.
4321 Set 5920Hz band gain.
4323 Set 8372Hz band gain.
4325 Set 11840Hz band gain.
4327 Set 16744Hz band gain.
4329 Set 20000Hz band gain.
4333 Apply audio surround upmix filter.
4335 This filter allows to produce multichannel output from audio stream.
4337 The filter accepts the following options:
4341 Set output channel layout. By default, this is @var{5.1}.
4343 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4344 for the required syntax.
4347 Set input channel layout. By default, this is @var{stereo}.
4349 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4350 for the required syntax.
4353 Set input volume level. By default, this is @var{1}.
4356 Set output volume level. By default, this is @var{1}.
4359 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4362 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4365 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4368 Set front center input volume. By default, this is @var{1}.
4371 Set front center output volume. By default, this is @var{1}.
4374 Set LFE input volume. By default, this is @var{1}.
4377 Set LFE output volume. By default, this is @var{1}.
4382 Boost or cut treble (upper) frequencies of the audio using a two-pole
4383 shelving filter with a response similar to that of a standard
4384 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4386 The filter accepts the following options:
4390 Give the gain at whichever is the lower of ~22 kHz and the
4391 Nyquist frequency. Its useful range is about -20 (for a large cut)
4392 to +20 (for a large boost). Beware of clipping when using a positive gain.
4395 Set the filter's central frequency and so can be used
4396 to extend or reduce the frequency range to be boosted or cut.
4397 The default value is @code{3000} Hz.
4400 Set method to specify band-width of filter.
4415 Determine how steep is the filter's shelf transition.
4418 Specify which channels to filter, by default all available are filtered.
4421 @subsection Commands
4423 This filter supports the following commands:
4426 Change treble frequency.
4427 Syntax for the command is : "@var{frequency}"
4430 Change treble width_type.
4431 Syntax for the command is : "@var{width_type}"
4434 Change treble width.
4435 Syntax for the command is : "@var{width}"
4439 Syntax for the command is : "@var{gain}"
4444 Sinusoidal amplitude modulation.
4446 The filter accepts the following options:
4450 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4451 (20 Hz or lower) will result in a tremolo effect.
4452 This filter may also be used as a ring modulator by specifying
4453 a modulation frequency higher than 20 Hz.
4454 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4457 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4458 Default value is 0.5.
4463 Sinusoidal phase modulation.
4465 The filter accepts the following options:
4469 Modulation frequency in Hertz.
4470 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4473 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4474 Default value is 0.5.
4479 Adjust the input audio volume.
4481 It accepts the following parameters:
4485 Set audio volume expression.
4487 Output values are clipped to the maximum value.
4489 The output audio volume is given by the relation:
4491 @var{output_volume} = @var{volume} * @var{input_volume}
4494 The default value for @var{volume} is "1.0".
4497 This parameter represents the mathematical precision.
4499 It determines which input sample formats will be allowed, which affects the
4500 precision of the volume scaling.
4504 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4506 32-bit floating-point; this limits input sample format to FLT. (default)
4508 64-bit floating-point; this limits input sample format to DBL.
4512 Choose the behaviour on encountering ReplayGain side data in input frames.
4516 Remove ReplayGain side data, ignoring its contents (the default).
4519 Ignore ReplayGain side data, but leave it in the frame.
4522 Prefer the track gain, if present.
4525 Prefer the album gain, if present.
4528 @item replaygain_preamp
4529 Pre-amplification gain in dB to apply to the selected replaygain gain.
4531 Default value for @var{replaygain_preamp} is 0.0.
4534 Set when the volume expression is evaluated.
4536 It accepts the following values:
4539 only evaluate expression once during the filter initialization, or
4540 when the @samp{volume} command is sent
4543 evaluate expression for each incoming frame
4546 Default value is @samp{once}.
4549 The volume expression can contain the following parameters.
4553 frame number (starting at zero)
4556 @item nb_consumed_samples
4557 number of samples consumed by the filter
4559 number of samples in the current frame
4561 original frame position in the file
4567 PTS at start of stream
4569 time at start of stream
4575 last set volume value
4578 Note that when @option{eval} is set to @samp{once} only the
4579 @var{sample_rate} and @var{tb} variables are available, all other
4580 variables will evaluate to NAN.
4582 @subsection Commands
4584 This filter supports the following commands:
4587 Modify the volume expression.
4588 The command accepts the same syntax of the corresponding option.
4590 If the specified expression is not valid, it is kept at its current
4592 @item replaygain_noclip
4593 Prevent clipping by limiting the gain applied.
4595 Default value for @var{replaygain_noclip} is 1.
4599 @subsection Examples
4603 Halve the input audio volume:
4607 volume=volume=-6.0206dB
4610 In all the above example the named key for @option{volume} can be
4611 omitted, for example like in:
4617 Increase input audio power by 6 decibels using fixed-point precision:
4619 volume=volume=6dB:precision=fixed
4623 Fade volume after time 10 with an annihilation period of 5 seconds:
4625 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4629 @section volumedetect
4631 Detect the volume of the input video.
4633 The filter has no parameters. The input is not modified. Statistics about
4634 the volume will be printed in the log when the input stream end is reached.
4636 In particular it will show the mean volume (root mean square), maximum
4637 volume (on a per-sample basis), and the beginning of a histogram of the
4638 registered volume values (from the maximum value to a cumulated 1/1000 of
4641 All volumes are in decibels relative to the maximum PCM value.
4643 @subsection Examples
4645 Here is an excerpt of the output:
4647 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4648 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4649 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4650 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4651 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4652 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4653 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4654 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4655 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4661 The mean square energy is approximately -27 dB, or 10^-2.7.
4663 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4665 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4668 In other words, raising the volume by +4 dB does not cause any clipping,
4669 raising it by +5 dB causes clipping for 6 samples, etc.
4671 @c man end AUDIO FILTERS
4673 @chapter Audio Sources
4674 @c man begin AUDIO SOURCES
4676 Below is a description of the currently available audio sources.
4680 Buffer audio frames, and make them available to the filter chain.
4682 This source is mainly intended for a programmatic use, in particular
4683 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4685 It accepts the following parameters:
4689 The timebase which will be used for timestamps of submitted frames. It must be
4690 either a floating-point number or in @var{numerator}/@var{denominator} form.
4693 The sample rate of the incoming audio buffers.
4696 The sample format of the incoming audio buffers.
4697 Either a sample format name or its corresponding integer representation from
4698 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4700 @item channel_layout
4701 The channel layout of the incoming audio buffers.
4702 Either a channel layout name from channel_layout_map in
4703 @file{libavutil/channel_layout.c} or its corresponding integer representation
4704 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4707 The number of channels of the incoming audio buffers.
4708 If both @var{channels} and @var{channel_layout} are specified, then they
4713 @subsection Examples
4716 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4719 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4720 Since the sample format with name "s16p" corresponds to the number
4721 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4724 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4729 Generate an audio signal specified by an expression.
4731 This source accepts in input one or more expressions (one for each
4732 channel), which are evaluated and used to generate a corresponding
4735 This source accepts the following options:
4739 Set the '|'-separated expressions list for each separate channel. In case the
4740 @option{channel_layout} option is not specified, the selected channel layout
4741 depends on the number of provided expressions. Otherwise the last
4742 specified expression is applied to the remaining output channels.
4744 @item channel_layout, c
4745 Set the channel layout. The number of channels in the specified layout
4746 must be equal to the number of specified expressions.
4749 Set the minimum duration of the sourced audio. See
4750 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4751 for the accepted syntax.
4752 Note that the resulting duration may be greater than the specified
4753 duration, as the generated audio is always cut at the end of a
4756 If not specified, or the expressed duration is negative, the audio is
4757 supposed to be generated forever.
4760 Set the number of samples per channel per each output frame,
4763 @item sample_rate, s
4764 Specify the sample rate, default to 44100.
4767 Each expression in @var{exprs} can contain the following constants:
4771 number of the evaluated sample, starting from 0
4774 time of the evaluated sample expressed in seconds, starting from 0
4781 @subsection Examples
4791 Generate a sin signal with frequency of 440 Hz, set sample rate to
4794 aevalsrc="sin(440*2*PI*t):s=8000"
4798 Generate a two channels signal, specify the channel layout (Front
4799 Center + Back Center) explicitly:
4801 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4805 Generate white noise:
4807 aevalsrc="-2+random(0)"
4811 Generate an amplitude modulated signal:
4813 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4817 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4819 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4826 The null audio source, return unprocessed audio frames. It is mainly useful
4827 as a template and to be employed in analysis / debugging tools, or as
4828 the source for filters which ignore the input data (for example the sox
4831 This source accepts the following options:
4835 @item channel_layout, cl
4837 Specifies the channel layout, and can be either an integer or a string
4838 representing a channel layout. The default value of @var{channel_layout}
4841 Check the channel_layout_map definition in
4842 @file{libavutil/channel_layout.c} for the mapping between strings and
4843 channel layout values.
4845 @item sample_rate, r
4846 Specifies the sample rate, and defaults to 44100.
4849 Set the number of samples per requested frames.
4853 @subsection Examples
4857 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4859 anullsrc=r=48000:cl=4
4863 Do the same operation with a more obvious syntax:
4865 anullsrc=r=48000:cl=mono
4869 All the parameters need to be explicitly defined.
4873 Synthesize a voice utterance using the libflite library.
4875 To enable compilation of this filter you need to configure FFmpeg with
4876 @code{--enable-libflite}.
4878 Note that versions of the flite library prior to 2.0 are not thread-safe.
4880 The filter accepts the following options:
4885 If set to 1, list the names of the available voices and exit
4886 immediately. Default value is 0.
4889 Set the maximum number of samples per frame. Default value is 512.
4892 Set the filename containing the text to speak.
4895 Set the text to speak.
4898 Set the voice to use for the speech synthesis. Default value is
4899 @code{kal}. See also the @var{list_voices} option.
4902 @subsection Examples
4906 Read from file @file{speech.txt}, and synthesize the text using the
4907 standard flite voice:
4909 flite=textfile=speech.txt
4913 Read the specified text selecting the @code{slt} voice:
4915 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4919 Input text to ffmpeg:
4921 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4925 Make @file{ffplay} speak the specified text, using @code{flite} and
4926 the @code{lavfi} device:
4928 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4932 For more information about libflite, check:
4933 @url{http://www.festvox.org/flite/}
4937 Generate a noise audio signal.
4939 The filter accepts the following options:
4942 @item sample_rate, r
4943 Specify the sample rate. Default value is 48000 Hz.
4946 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4950 Specify the duration of the generated audio stream. Not specifying this option
4951 results in noise with an infinite length.
4953 @item color, colour, c
4954 Specify the color of noise. Available noise colors are white, pink, brown,
4955 blue and violet. Default color is white.
4958 Specify a value used to seed the PRNG.
4961 Set the number of samples per each output frame, default is 1024.
4964 @subsection Examples
4969 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4971 anoisesrc=d=60:c=pink:r=44100:a=0.5
4977 Generate odd-tap Hilbert transform FIR coefficients.
4979 The resulting stream can be used with @ref{afir} filter for phase-shifting
4980 the signal by 90 degrees.
4982 This is used in many matrix coding schemes and for analytic signal generation.
4983 The process is often written as a multiplication by i (or j), the imaginary unit.
4985 The filter accepts the following options:
4989 @item sample_rate, s
4990 Set sample rate, default is 44100.
4993 Set length of FIR filter, default is 22051.
4996 Set number of samples per each frame.
4999 Set window function to be used when generating FIR coefficients.
5004 Generate an audio signal made of a sine wave with amplitude 1/8.
5006 The audio signal is bit-exact.
5008 The filter accepts the following options:
5013 Set the carrier frequency. Default is 440 Hz.
5015 @item beep_factor, b
5016 Enable a periodic beep every second with frequency @var{beep_factor} times
5017 the carrier frequency. Default is 0, meaning the beep is disabled.
5019 @item sample_rate, r
5020 Specify the sample rate, default is 44100.
5023 Specify the duration of the generated audio stream.
5025 @item samples_per_frame
5026 Set the number of samples per output frame.
5028 The expression can contain the following constants:
5032 The (sequential) number of the output audio frame, starting from 0.
5035 The PTS (Presentation TimeStamp) of the output audio frame,
5036 expressed in @var{TB} units.
5039 The PTS of the output audio frame, expressed in seconds.
5042 The timebase of the output audio frames.
5045 Default is @code{1024}.
5048 @subsection Examples
5053 Generate a simple 440 Hz sine wave:
5059 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5063 sine=frequency=220:beep_factor=4:duration=5
5067 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5070 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5074 @c man end AUDIO SOURCES
5076 @chapter Audio Sinks
5077 @c man begin AUDIO SINKS
5079 Below is a description of the currently available audio sinks.
5081 @section abuffersink
5083 Buffer audio frames, and make them available to the end of filter chain.
5085 This sink is mainly intended for programmatic use, in particular
5086 through the interface defined in @file{libavfilter/buffersink.h}
5087 or the options system.
5089 It accepts a pointer to an AVABufferSinkContext structure, which
5090 defines the incoming buffers' formats, to be passed as the opaque
5091 parameter to @code{avfilter_init_filter} for initialization.
5094 Null audio sink; do absolutely nothing with the input audio. It is
5095 mainly useful as a template and for use in analysis / debugging
5098 @c man end AUDIO SINKS
5100 @chapter Video Filters
5101 @c man begin VIDEO FILTERS
5103 When you configure your FFmpeg build, you can disable any of the
5104 existing filters using @code{--disable-filters}.
5105 The configure output will show the video filters included in your
5108 Below is a description of the currently available video filters.
5110 @section alphaextract
5112 Extract the alpha component from the input as a grayscale video. This
5113 is especially useful with the @var{alphamerge} filter.
5117 Add or replace the alpha component of the primary input with the
5118 grayscale value of a second input. This is intended for use with
5119 @var{alphaextract} to allow the transmission or storage of frame
5120 sequences that have alpha in a format that doesn't support an alpha
5123 For example, to reconstruct full frames from a normal YUV-encoded video
5124 and a separate video created with @var{alphaextract}, you might use:
5126 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5129 Since this filter is designed for reconstruction, it operates on frame
5130 sequences without considering timestamps, and terminates when either
5131 input reaches end of stream. This will cause problems if your encoding
5132 pipeline drops frames. If you're trying to apply an image as an
5133 overlay to a video stream, consider the @var{overlay} filter instead.
5137 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5138 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5139 Substation Alpha) subtitles files.
5141 This filter accepts the following option in addition to the common options from
5142 the @ref{subtitles} filter:
5146 Set the shaping engine
5148 Available values are:
5151 The default libass shaping engine, which is the best available.
5153 Fast, font-agnostic shaper that can do only substitutions
5155 Slower shaper using OpenType for substitutions and positioning
5158 The default is @code{auto}.
5162 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5164 The filter accepts the following options:
5168 Set threshold A for 1st plane. Default is 0.02.
5169 Valid range is 0 to 0.3.
5172 Set threshold B for 1st plane. Default is 0.04.
5173 Valid range is 0 to 5.
5176 Set threshold A for 2nd plane. Default is 0.02.
5177 Valid range is 0 to 0.3.
5180 Set threshold B for 2nd plane. Default is 0.04.
5181 Valid range is 0 to 5.
5184 Set threshold A for 3rd plane. Default is 0.02.
5185 Valid range is 0 to 0.3.
5188 Set threshold B for 3rd plane. Default is 0.04.
5189 Valid range is 0 to 5.
5191 Threshold A is designed to react on abrupt changes in the input signal and
5192 threshold B is designed to react on continuous changes in the input signal.
5195 Set number of frames filter will use for averaging. Default is 33. Must be odd
5196 number in range [5, 129].
5199 Set what planes of frame filter will use for averaging. Default is all.
5204 Apply average blur filter.
5206 The filter accepts the following options:
5210 Set horizontal kernel size.
5213 Set which planes to filter. By default all planes are filtered.
5216 Set vertical kernel size, if zero it will be same as @code{sizeX}.
5217 Default is @code{0}.
5222 Compute the bounding box for the non-black pixels in the input frame
5225 This filter computes the bounding box containing all the pixels with a
5226 luminance value greater than the minimum allowed value.
5227 The parameters describing the bounding box are printed on the filter
5230 The filter accepts the following option:
5234 Set the minimal luminance value. Default is @code{16}.
5237 @section bitplanenoise
5239 Show and measure bit plane noise.
5241 The filter accepts the following options:
5245 Set which plane to analyze. Default is @code{1}.
5248 Filter out noisy pixels from @code{bitplane} set above.
5249 Default is disabled.
5252 @section blackdetect
5254 Detect video intervals that are (almost) completely black. Can be
5255 useful to detect chapter transitions, commercials, or invalid
5256 recordings. Output lines contains the time for the start, end and
5257 duration of the detected black interval expressed in seconds.
5259 In order to display the output lines, you need to set the loglevel at
5260 least to the AV_LOG_INFO value.
5262 The filter accepts the following options:
5265 @item black_min_duration, d
5266 Set the minimum detected black duration expressed in seconds. It must
5267 be a non-negative floating point number.
5269 Default value is 2.0.
5271 @item picture_black_ratio_th, pic_th
5272 Set the threshold for considering a picture "black".
5273 Express the minimum value for the ratio:
5275 @var{nb_black_pixels} / @var{nb_pixels}
5278 for which a picture is considered black.
5279 Default value is 0.98.
5281 @item pixel_black_th, pix_th
5282 Set the threshold for considering a pixel "black".
5284 The threshold expresses the maximum pixel luminance value for which a
5285 pixel is considered "black". The provided value is scaled according to
5286 the following equation:
5288 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5291 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5292 the input video format, the range is [0-255] for YUV full-range
5293 formats and [16-235] for YUV non full-range formats.
5295 Default value is 0.10.
5298 The following example sets the maximum pixel threshold to the minimum
5299 value, and detects only black intervals of 2 or more seconds:
5301 blackdetect=d=2:pix_th=0.00
5306 Detect frames that are (almost) completely black. Can be useful to
5307 detect chapter transitions or commercials. Output lines consist of
5308 the frame number of the detected frame, the percentage of blackness,
5309 the position in the file if known or -1 and the timestamp in seconds.
5311 In order to display the output lines, you need to set the loglevel at
5312 least to the AV_LOG_INFO value.
5314 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5315 The value represents the percentage of pixels in the picture that
5316 are below the threshold value.
5318 It accepts the following parameters:
5323 The percentage of the pixels that have to be below the threshold; it defaults to
5326 @item threshold, thresh
5327 The threshold below which a pixel value is considered black; it defaults to
5332 @section blend, tblend
5334 Blend two video frames into each other.
5336 The @code{blend} filter takes two input streams and outputs one
5337 stream, the first input is the "top" layer and second input is
5338 "bottom" layer. By default, the output terminates when the longest input terminates.
5340 The @code{tblend} (time blend) filter takes two consecutive frames
5341 from one single stream, and outputs the result obtained by blending
5342 the new frame on top of the old frame.
5344 A description of the accepted options follows.
5352 Set blend mode for specific pixel component or all pixel components in case
5353 of @var{all_mode}. Default value is @code{normal}.
5355 Available values for component modes are:
5397 Set blend opacity for specific pixel component or all pixel components in case
5398 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5405 Set blend expression for specific pixel component or all pixel components in case
5406 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5408 The expressions can use the following variables:
5412 The sequential number of the filtered frame, starting from @code{0}.
5416 the coordinates of the current sample
5420 the width and height of currently filtered plane
5424 Width and height scale depending on the currently filtered plane. It is the
5425 ratio between the corresponding luma plane number of pixels and the current
5426 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5427 @code{0.5,0.5} for chroma planes.
5430 Time of the current frame, expressed in seconds.
5433 Value of pixel component at current location for first video frame (top layer).
5436 Value of pixel component at current location for second video frame (bottom layer).
5440 The @code{blend} filter also supports the @ref{framesync} options.
5442 @subsection Examples
5446 Apply transition from bottom layer to top layer in first 10 seconds:
5448 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5452 Apply linear horizontal transition from top layer to bottom layer:
5454 blend=all_expr='A*(X/W)+B*(1-X/W)'
5458 Apply 1x1 checkerboard effect:
5460 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5464 Apply uncover left effect:
5466 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5470 Apply uncover down effect:
5472 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5476 Apply uncover up-left effect:
5478 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5482 Split diagonally video and shows top and bottom layer on each side:
5484 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5488 Display differences between the current and the previous frame:
5490 tblend=all_mode=grainextract
5496 Apply a boxblur algorithm to the input video.
5498 It accepts the following parameters:
5502 @item luma_radius, lr
5503 @item luma_power, lp
5504 @item chroma_radius, cr
5505 @item chroma_power, cp
5506 @item alpha_radius, ar
5507 @item alpha_power, ap
5511 A description of the accepted options follows.
5514 @item luma_radius, lr
5515 @item chroma_radius, cr
5516 @item alpha_radius, ar
5517 Set an expression for the box radius in pixels used for blurring the
5518 corresponding input plane.
5520 The radius value must be a non-negative number, and must not be
5521 greater than the value of the expression @code{min(w,h)/2} for the
5522 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5525 Default value for @option{luma_radius} is "2". If not specified,
5526 @option{chroma_radius} and @option{alpha_radius} default to the
5527 corresponding value set for @option{luma_radius}.
5529 The expressions can contain the following constants:
5533 The input width and height in pixels.
5537 The input chroma image width and height in pixels.
5541 The horizontal and vertical chroma subsample values. For example, for the
5542 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5545 @item luma_power, lp
5546 @item chroma_power, cp
5547 @item alpha_power, ap
5548 Specify how many times the boxblur filter is applied to the
5549 corresponding plane.
5551 Default value for @option{luma_power} is 2. If not specified,
5552 @option{chroma_power} and @option{alpha_power} default to the
5553 corresponding value set for @option{luma_power}.
5555 A value of 0 will disable the effect.
5558 @subsection Examples
5562 Apply a boxblur filter with the luma, chroma, and alpha radii
5565 boxblur=luma_radius=2:luma_power=1
5570 Set the luma radius to 2, and alpha and chroma radius to 0:
5572 boxblur=2:1:cr=0:ar=0
5576 Set the luma and chroma radii to a fraction of the video dimension:
5578 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5584 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5585 Deinterlacing Filter").
5587 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5588 interpolation algorithms.
5589 It accepts the following parameters:
5593 The interlacing mode to adopt. It accepts one of the following values:
5597 Output one frame for each frame.
5599 Output one frame for each field.
5602 The default value is @code{send_field}.
5605 The picture field parity assumed for the input interlaced video. It accepts one
5606 of the following values:
5610 Assume the top field is first.
5612 Assume the bottom field is first.
5614 Enable automatic detection of field parity.
5617 The default value is @code{auto}.
5618 If the interlacing is unknown or the decoder does not export this information,
5619 top field first will be assumed.
5622 Specify which frames to deinterlace. Accept one of the following
5627 Deinterlace all frames.
5629 Only deinterlace frames marked as interlaced.
5632 The default value is @code{all}.
5636 YUV colorspace color/chroma keying.
5638 The filter accepts the following options:
5642 The color which will be replaced with transparency.
5645 Similarity percentage with the key color.
5647 0.01 matches only the exact key color, while 1.0 matches everything.
5652 0.0 makes pixels either fully transparent, or not transparent at all.
5654 Higher values result in semi-transparent pixels, with a higher transparency
5655 the more similar the pixels color is to the key color.
5658 Signals that the color passed is already in YUV instead of RGB.
5660 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5661 This can be used to pass exact YUV values as hexadecimal numbers.
5664 @subsection Examples
5668 Make every green pixel in the input image transparent:
5670 ffmpeg -i input.png -vf chromakey=green out.png
5674 Overlay a greenscreen-video on top of a static black background.
5676 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
5682 Display CIE color diagram with pixels overlaid onto it.
5684 The filter accepts the following options:
5699 @item uhdtv, rec2020
5712 Set what gamuts to draw.
5714 See @code{system} option for available values.
5717 Set ciescope size, by default set to 512.
5720 Set intensity used to map input pixel values to CIE diagram.
5723 Set contrast used to draw tongue colors that are out of active color system gamut.
5726 Correct gamma displayed on scope, by default enabled.
5729 Show white point on CIE diagram, by default disabled.
5732 Set input gamma. Used only with XYZ input color space.
5737 Visualize information exported by some codecs.
5739 Some codecs can export information through frames using side-data or other
5740 means. For example, some MPEG based codecs export motion vectors through the
5741 @var{export_mvs} flag in the codec @option{flags2} option.
5743 The filter accepts the following option:
5747 Set motion vectors to visualize.
5749 Available flags for @var{mv} are:
5753 forward predicted MVs of P-frames
5755 forward predicted MVs of B-frames
5757 backward predicted MVs of B-frames
5761 Display quantization parameters using the chroma planes.
5764 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5766 Available flags for @var{mv_type} are:
5770 forward predicted MVs
5772 backward predicted MVs
5775 @item frame_type, ft
5776 Set frame type to visualize motion vectors of.
5778 Available flags for @var{frame_type} are:
5782 intra-coded frames (I-frames)
5784 predicted frames (P-frames)
5786 bi-directionally predicted frames (B-frames)
5790 @subsection Examples
5794 Visualize forward predicted MVs of all frames using @command{ffplay}:
5796 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5800 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5802 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5806 @section colorbalance
5807 Modify intensity of primary colors (red, green and blue) of input frames.
5809 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5810 regions for the red-cyan, green-magenta or blue-yellow balance.
5812 A positive adjustment value shifts the balance towards the primary color, a negative
5813 value towards the complementary color.
5815 The filter accepts the following options:
5821 Adjust red, green and blue shadows (darkest pixels).
5826 Adjust red, green and blue midtones (medium pixels).
5831 Adjust red, green and blue highlights (brightest pixels).
5833 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5836 @subsection Examples
5840 Add red color cast to shadows:
5847 RGB colorspace color keying.
5849 The filter accepts the following options:
5853 The color which will be replaced with transparency.
5856 Similarity percentage with the key color.
5858 0.01 matches only the exact key color, while 1.0 matches everything.
5863 0.0 makes pixels either fully transparent, or not transparent at all.
5865 Higher values result in semi-transparent pixels, with a higher transparency
5866 the more similar the pixels color is to the key color.
5869 @subsection Examples
5873 Make every green pixel in the input image transparent:
5875 ffmpeg -i input.png -vf colorkey=green out.png
5879 Overlay a greenscreen-video on top of a static background image.
5881 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
5885 @section colorlevels
5887 Adjust video input frames using levels.
5889 The filter accepts the following options:
5896 Adjust red, green, blue and alpha input black point.
5897 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5903 Adjust red, green, blue and alpha input white point.
5904 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5906 Input levels are used to lighten highlights (bright tones), darken shadows
5907 (dark tones), change the balance of bright and dark tones.
5913 Adjust red, green, blue and alpha output black point.
5914 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5920 Adjust red, green, blue and alpha output white point.
5921 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5923 Output levels allows manual selection of a constrained output level range.
5926 @subsection Examples
5930 Make video output darker:
5932 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5938 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5942 Make video output lighter:
5944 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5948 Increase brightness:
5950 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5954 @section colorchannelmixer
5956 Adjust video input frames by re-mixing color channels.
5958 This filter modifies a color channel by adding the values associated to
5959 the other channels of the same pixels. For example if the value to
5960 modify is red, the output value will be:
5962 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5965 The filter accepts the following options:
5972 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5973 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5979 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5980 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5986 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5987 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5993 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5994 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5996 Allowed ranges for options are @code{[-2.0, 2.0]}.
5999 @subsection Examples
6003 Convert source to grayscale:
6005 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6008 Simulate sepia tones:
6010 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6014 @section colormatrix
6016 Convert color matrix.
6018 The filter accepts the following options:
6023 Specify the source and destination color matrix. Both values must be
6026 The accepted values are:
6054 For example to convert from BT.601 to SMPTE-240M, use the command:
6056 colormatrix=bt601:smpte240m
6061 Convert colorspace, transfer characteristics or color primaries.
6062 Input video needs to have an even size.
6064 The filter accepts the following options:
6069 Specify all color properties at once.
6071 The accepted values are:
6101 Specify output colorspace.
6103 The accepted values are:
6112 BT.470BG or BT.601-6 625
6115 SMPTE-170M or BT.601-6 525
6124 BT.2020 with non-constant luminance
6130 Specify output transfer characteristics.
6132 The accepted values are:
6144 Constant gamma of 2.2
6147 Constant gamma of 2.8
6150 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6168 BT.2020 for 10-bits content
6171 BT.2020 for 12-bits content
6177 Specify output color primaries.
6179 The accepted values are:
6188 BT.470BG or BT.601-6 625
6191 SMPTE-170M or BT.601-6 525
6215 Specify output color range.
6217 The accepted values are:
6220 TV (restricted) range
6223 MPEG (restricted) range
6234 Specify output color format.
6236 The accepted values are:
6239 YUV 4:2:0 planar 8-bits
6242 YUV 4:2:0 planar 10-bits
6245 YUV 4:2:0 planar 12-bits
6248 YUV 4:2:2 planar 8-bits
6251 YUV 4:2:2 planar 10-bits
6254 YUV 4:2:2 planar 12-bits
6257 YUV 4:4:4 planar 8-bits
6260 YUV 4:4:4 planar 10-bits
6263 YUV 4:4:4 planar 12-bits
6268 Do a fast conversion, which skips gamma/primary correction. This will take
6269 significantly less CPU, but will be mathematically incorrect. To get output
6270 compatible with that produced by the colormatrix filter, use fast=1.
6273 Specify dithering mode.
6275 The accepted values are:
6281 Floyd-Steinberg dithering
6285 Whitepoint adaptation mode.
6287 The accepted values are:
6290 Bradford whitepoint adaptation
6293 von Kries whitepoint adaptation
6296 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6300 Override all input properties at once. Same accepted values as @ref{all}.
6303 Override input colorspace. Same accepted values as @ref{space}.
6306 Override input color primaries. Same accepted values as @ref{primaries}.
6309 Override input transfer characteristics. Same accepted values as @ref{trc}.
6312 Override input color range. Same accepted values as @ref{range}.
6316 The filter converts the transfer characteristics, color space and color
6317 primaries to the specified user values. The output value, if not specified,
6318 is set to a default value based on the "all" property. If that property is
6319 also not specified, the filter will log an error. The output color range and
6320 format default to the same value as the input color range and format. The
6321 input transfer characteristics, color space, color primaries and color range
6322 should be set on the input data. If any of these are missing, the filter will
6323 log an error and no conversion will take place.
6325 For example to convert the input to SMPTE-240M, use the command:
6327 colorspace=smpte240m
6330 @section convolution
6332 Apply convolution 3x3, 5x5 or 7x7 filter.
6334 The filter accepts the following options:
6341 Set matrix for each plane.
6342 Matrix is sequence of 9, 25 or 49 signed integers.
6348 Set multiplier for calculated value for each plane.
6354 Set bias for each plane. This value is added to the result of the multiplication.
6355 Useful for making the overall image brighter or darker. Default is 0.0.
6358 @subsection Examples
6364 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"
6370 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"
6376 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"
6382 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"
6386 Apply laplacian edge detector which includes diagonals:
6388 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"
6394 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"
6400 Apply 2D convolution of video stream in frequency domain using second stream
6403 The filter accepts the following options:
6407 Set which planes to process.
6410 Set which impulse video frames will be processed, can be @var{first}
6411 or @var{all}. Default is @var{all}.
6414 The @code{convolve} filter also supports the @ref{framesync} options.
6418 Copy the input video source unchanged to the output. This is mainly useful for
6423 Video filtering on GPU using Apple's CoreImage API on OSX.
6425 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6426 processed by video hardware. However, software-based OpenGL implementations
6427 exist which means there is no guarantee for hardware processing. It depends on
6430 There are many filters and image generators provided by Apple that come with a
6431 large variety of options. The filter has to be referenced by its name along
6434 The coreimage filter accepts the following options:
6437 List all available filters and generators along with all their respective
6438 options as well as possible minimum and maximum values along with the default
6445 Specify all filters by their respective name and options.
6446 Use @var{list_filters} to determine all valid filter names and options.
6447 Numerical options are specified by a float value and are automatically clamped
6448 to their respective value range. Vector and color options have to be specified
6449 by a list of space separated float values. Character escaping has to be done.
6450 A special option name @code{default} is available to use default options for a
6453 It is required to specify either @code{default} or at least one of the filter options.
6454 All omitted options are used with their default values.
6455 The syntax of the filter string is as follows:
6457 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6461 Specify a rectangle where the output of the filter chain is copied into the
6462 input image. It is given by a list of space separated float values:
6464 output_rect=x\ y\ width\ height
6466 If not given, the output rectangle equals the dimensions of the input image.
6467 The output rectangle is automatically cropped at the borders of the input
6468 image. Negative values are valid for each component.
6470 output_rect=25\ 25\ 100\ 100
6474 Several filters can be chained for successive processing without GPU-HOST
6475 transfers allowing for fast processing of complex filter chains.
6476 Currently, only filters with zero (generators) or exactly one (filters) input
6477 image and one output image are supported. Also, transition filters are not yet
6480 Some filters generate output images with additional padding depending on the
6481 respective filter kernel. The padding is automatically removed to ensure the
6482 filter output has the same size as the input image.
6484 For image generators, the size of the output image is determined by the
6485 previous output image of the filter chain or the input image of the whole
6486 filterchain, respectively. The generators do not use the pixel information of
6487 this image to generate their output. However, the generated output is
6488 blended onto this image, resulting in partial or complete coverage of the
6491 The @ref{coreimagesrc} video source can be used for generating input images
6492 which are directly fed into the filter chain. By using it, providing input
6493 images by another video source or an input video is not required.
6495 @subsection Examples
6500 List all filters available:
6502 coreimage=list_filters=true
6506 Use the CIBoxBlur filter with default options to blur an image:
6508 coreimage=filter=CIBoxBlur@@default
6512 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6513 its center at 100x100 and a radius of 50 pixels:
6515 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6519 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6520 given as complete and escaped command-line for Apple's standard bash shell:
6522 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6528 Crop the input video to given dimensions.
6530 It accepts the following parameters:
6534 The width of the output video. It defaults to @code{iw}.
6535 This expression is evaluated only once during the filter
6536 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6539 The height of the output video. It defaults to @code{ih}.
6540 This expression is evaluated only once during the filter
6541 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6544 The horizontal position, in the input video, of the left edge of the output
6545 video. It defaults to @code{(in_w-out_w)/2}.
6546 This expression is evaluated per-frame.
6549 The vertical position, in the input video, of the top edge of the output video.
6550 It defaults to @code{(in_h-out_h)/2}.
6551 This expression is evaluated per-frame.
6554 If set to 1 will force the output display aspect ratio
6555 to be the same of the input, by changing the output sample aspect
6556 ratio. It defaults to 0.
6559 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6560 width/height/x/y as specified and will not be rounded to nearest smaller value.
6564 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6565 expressions containing the following constants:
6570 The computed values for @var{x} and @var{y}. They are evaluated for
6575 The input width and height.
6579 These are the same as @var{in_w} and @var{in_h}.
6583 The output (cropped) width and height.
6587 These are the same as @var{out_w} and @var{out_h}.
6590 same as @var{iw} / @var{ih}
6593 input sample aspect ratio
6596 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6600 horizontal and vertical chroma subsample values. For example for the
6601 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6604 The number of the input frame, starting from 0.
6607 the position in the file of the input frame, NAN if unknown
6610 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6614 The expression for @var{out_w} may depend on the value of @var{out_h},
6615 and the expression for @var{out_h} may depend on @var{out_w}, but they
6616 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6617 evaluated after @var{out_w} and @var{out_h}.
6619 The @var{x} and @var{y} parameters specify the expressions for the
6620 position of the top-left corner of the output (non-cropped) area. They
6621 are evaluated for each frame. If the evaluated value is not valid, it
6622 is approximated to the nearest valid value.
6624 The expression for @var{x} may depend on @var{y}, and the expression
6625 for @var{y} may depend on @var{x}.
6627 @subsection Examples
6631 Crop area with size 100x100 at position (12,34).
6636 Using named options, the example above becomes:
6638 crop=w=100:h=100:x=12:y=34
6642 Crop the central input area with size 100x100:
6648 Crop the central input area with size 2/3 of the input video:
6650 crop=2/3*in_w:2/3*in_h
6654 Crop the input video central square:
6661 Delimit the rectangle with the top-left corner placed at position
6662 100:100 and the right-bottom corner corresponding to the right-bottom
6663 corner of the input image.
6665 crop=in_w-100:in_h-100:100:100
6669 Crop 10 pixels from the left and right borders, and 20 pixels from
6670 the top and bottom borders
6672 crop=in_w-2*10:in_h-2*20
6676 Keep only the bottom right quarter of the input image:
6678 crop=in_w/2:in_h/2:in_w/2:in_h/2
6682 Crop height for getting Greek harmony:
6684 crop=in_w:1/PHI*in_w
6688 Apply trembling effect:
6690 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)
6694 Apply erratic camera effect depending on timestamp:
6696 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)"
6700 Set x depending on the value of y:
6702 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6706 @subsection Commands
6708 This filter supports the following commands:
6714 Set width/height of the output video and the horizontal/vertical position
6716 The command accepts the same syntax of the corresponding option.
6718 If the specified expression is not valid, it is kept at its current
6724 Auto-detect the crop size.
6726 It calculates the necessary cropping parameters and prints the
6727 recommended parameters via the logging system. The detected dimensions
6728 correspond to the non-black area of the input video.
6730 It accepts the following parameters:
6735 Set higher black value threshold, which can be optionally specified
6736 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6737 value greater to the set value is considered non-black. It defaults to 24.
6738 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6739 on the bitdepth of the pixel format.
6742 The value which the width/height should be divisible by. It defaults to
6743 16. The offset is automatically adjusted to center the video. Use 2 to
6744 get only even dimensions (needed for 4:2:2 video). 16 is best when
6745 encoding to most video codecs.
6747 @item reset_count, reset
6748 Set the counter that determines after how many frames cropdetect will
6749 reset the previously detected largest video area and start over to
6750 detect the current optimal crop area. Default value is 0.
6752 This can be useful when channel logos distort the video area. 0
6753 indicates 'never reset', and returns the largest area encountered during
6760 Apply color adjustments using curves.
6762 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6763 component (red, green and blue) has its values defined by @var{N} key points
6764 tied from each other using a smooth curve. The x-axis represents the pixel
6765 values from the input frame, and the y-axis the new pixel values to be set for
6768 By default, a component curve is defined by the two points @var{(0;0)} and
6769 @var{(1;1)}. This creates a straight line where each original pixel value is
6770 "adjusted" to its own value, which means no change to the image.
6772 The filter allows you to redefine these two points and add some more. A new
6773 curve (using a natural cubic spline interpolation) will be define to pass
6774 smoothly through all these new coordinates. The new defined points needs to be
6775 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6776 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6777 the vector spaces, the values will be clipped accordingly.
6779 The filter accepts the following options:
6783 Select one of the available color presets. This option can be used in addition
6784 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6785 options takes priority on the preset values.
6786 Available presets are:
6789 @item color_negative
6792 @item increase_contrast
6794 @item linear_contrast
6795 @item medium_contrast
6797 @item strong_contrast
6800 Default is @code{none}.
6802 Set the master key points. These points will define a second pass mapping. It
6803 is sometimes called a "luminance" or "value" mapping. It can be used with
6804 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6805 post-processing LUT.
6807 Set the key points for the red component.
6809 Set the key points for the green component.
6811 Set the key points for the blue component.
6813 Set the key points for all components (not including master).
6814 Can be used in addition to the other key points component
6815 options. In this case, the unset component(s) will fallback on this
6816 @option{all} setting.
6818 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6820 Save Gnuplot script of the curves in specified file.
6823 To avoid some filtergraph syntax conflicts, each key points list need to be
6824 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6826 @subsection Examples
6830 Increase slightly the middle level of blue:
6832 curves=blue='0/0 0.5/0.58 1/1'
6838 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'
6840 Here we obtain the following coordinates for each components:
6843 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6845 @code{(0;0) (0.50;0.48) (1;1)}
6847 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6851 The previous example can also be achieved with the associated built-in preset:
6853 curves=preset=vintage
6863 Use a Photoshop preset and redefine the points of the green component:
6865 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6869 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6870 and @command{gnuplot}:
6872 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6873 gnuplot -p /tmp/curves.plt
6879 Video data analysis filter.
6881 This filter shows hexadecimal pixel values of part of video.
6883 The filter accepts the following options:
6887 Set output video size.
6890 Set x offset from where to pick pixels.
6893 Set y offset from where to pick pixels.
6896 Set scope mode, can be one of the following:
6899 Draw hexadecimal pixel values with white color on black background.
6902 Draw hexadecimal pixel values with input video pixel color on black
6906 Draw hexadecimal pixel values on color background picked from input video,
6907 the text color is picked in such way so its always visible.
6911 Draw rows and columns numbers on left and top of video.
6914 Set background opacity.
6919 Denoise frames using 2D DCT (frequency domain filtering).
6921 This filter is not designed for real time.
6923 The filter accepts the following options:
6927 Set the noise sigma constant.
6929 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6930 coefficient (absolute value) below this threshold with be dropped.
6932 If you need a more advanced filtering, see @option{expr}.
6934 Default is @code{0}.
6937 Set number overlapping pixels for each block. Since the filter can be slow, you
6938 may want to reduce this value, at the cost of a less effective filter and the
6939 risk of various artefacts.
6941 If the overlapping value doesn't permit processing the whole input width or
6942 height, a warning will be displayed and according borders won't be denoised.
6944 Default value is @var{blocksize}-1, which is the best possible setting.
6947 Set the coefficient factor expression.
6949 For each coefficient of a DCT block, this expression will be evaluated as a
6950 multiplier value for the coefficient.
6952 If this is option is set, the @option{sigma} option will be ignored.
6954 The absolute value of the coefficient can be accessed through the @var{c}
6958 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6959 @var{blocksize}, which is the width and height of the processed blocks.
6961 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6962 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6963 on the speed processing. Also, a larger block size does not necessarily means a
6967 @subsection Examples
6969 Apply a denoise with a @option{sigma} of @code{4.5}:
6974 The same operation can be achieved using the expression system:
6976 dctdnoiz=e='gte(c, 4.5*3)'
6979 Violent denoise using a block size of @code{16x16}:
6986 Remove banding artifacts from input video.
6987 It works by replacing banded pixels with average value of referenced pixels.
6989 The filter accepts the following options:
6996 Set banding detection threshold for each plane. Default is 0.02.
6997 Valid range is 0.00003 to 0.5.
6998 If difference between current pixel and reference pixel is less than threshold,
6999 it will be considered as banded.
7002 Banding detection range in pixels. Default is 16. If positive, random number
7003 in range 0 to set value will be used. If negative, exact absolute value
7005 The range defines square of four pixels around current pixel.
7008 Set direction in radians from which four pixel will be compared. If positive,
7009 random direction from 0 to set direction will be picked. If negative, exact of
7010 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7011 will pick only pixels on same row and -PI/2 will pick only pixels on same
7015 If enabled, current pixel is compared with average value of all four
7016 surrounding pixels. The default is enabled. If disabled current pixel is
7017 compared with all four surrounding pixels. The pixel is considered banded
7018 if only all four differences with surrounding pixels are less than threshold.
7021 If enabled, current pixel is changed if and only if all pixel components are banded,
7022 e.g. banding detection threshold is triggered for all color components.
7023 The default is disabled.
7029 Drop duplicated frames at regular intervals.
7031 The filter accepts the following options:
7035 Set the number of frames from which one will be dropped. Setting this to
7036 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7037 Default is @code{5}.
7040 Set the threshold for duplicate detection. If the difference metric for a frame
7041 is less than or equal to this value, then it is declared as duplicate. Default
7045 Set scene change threshold. Default is @code{15}.
7049 Set the size of the x and y-axis blocks used during metric calculations.
7050 Larger blocks give better noise suppression, but also give worse detection of
7051 small movements. Must be a power of two. Default is @code{32}.
7054 Mark main input as a pre-processed input and activate clean source input
7055 stream. This allows the input to be pre-processed with various filters to help
7056 the metrics calculation while keeping the frame selection lossless. When set to
7057 @code{1}, the first stream is for the pre-processed input, and the second
7058 stream is the clean source from where the kept frames are chosen. Default is
7062 Set whether or not chroma is considered in the metric calculations. Default is
7068 Apply 2D deconvolution of video stream in frequency domain using second stream
7071 The filter accepts the following options:
7075 Set which planes to process.
7078 Set which impulse video frames will be processed, can be @var{first}
7079 or @var{all}. Default is @var{all}.
7082 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7083 and height are not same and not power of 2 or if stream prior to convolving
7087 The @code{deconvolve} filter also supports the @ref{framesync} options.
7091 Apply deflate effect to the video.
7093 This filter replaces the pixel by the local(3x3) average by taking into account
7094 only values lower than the pixel.
7096 It accepts the following options:
7103 Limit the maximum change for each plane, default is 65535.
7104 If 0, plane will remain unchanged.
7109 Remove temporal frame luminance variations.
7111 It accepts the following options:
7115 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7118 Set averaging mode to smooth temporal luminance variations.
7120 Available values are:
7145 Do not actually modify frame. Useful when one only wants metadata.
7150 Remove judder produced by partially interlaced telecined content.
7152 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7153 source was partially telecined content then the output of @code{pullup,dejudder}
7154 will have a variable frame rate. May change the recorded frame rate of the
7155 container. Aside from that change, this filter will not affect constant frame
7158 The option available in this filter is:
7162 Specify the length of the window over which the judder repeats.
7164 Accepts any integer greater than 1. Useful values are:
7168 If the original was telecined from 24 to 30 fps (Film to NTSC).
7171 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7174 If a mixture of the two.
7177 The default is @samp{4}.
7182 Suppress a TV station logo by a simple interpolation of the surrounding
7183 pixels. Just set a rectangle covering the logo and watch it disappear
7184 (and sometimes something even uglier appear - your mileage may vary).
7186 It accepts the following parameters:
7191 Specify the top left corner coordinates of the logo. They must be
7196 Specify the width and height of the logo to clear. They must be
7200 Specify the thickness of the fuzzy edge of the rectangle (added to
7201 @var{w} and @var{h}). The default value is 1. This option is
7202 deprecated, setting higher values should no longer be necessary and
7206 When set to 1, a green rectangle is drawn on the screen to simplify
7207 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7208 The default value is 0.
7210 The rectangle is drawn on the outermost pixels which will be (partly)
7211 replaced with interpolated values. The values of the next pixels
7212 immediately outside this rectangle in each direction will be used to
7213 compute the interpolated pixel values inside the rectangle.
7217 @subsection Examples
7221 Set a rectangle covering the area with top left corner coordinates 0,0
7222 and size 100x77, and a band of size 10:
7224 delogo=x=0:y=0:w=100:h=77:band=10
7231 Attempt to fix small changes in horizontal and/or vertical shift. This
7232 filter helps remove camera shake from hand-holding a camera, bumping a
7233 tripod, moving on a vehicle, etc.
7235 The filter accepts the following options:
7243 Specify a rectangular area where to limit the search for motion
7245 If desired the search for motion vectors can be limited to a
7246 rectangular area of the frame defined by its top left corner, width
7247 and height. These parameters have the same meaning as the drawbox
7248 filter which can be used to visualise the position of the bounding
7251 This is useful when simultaneous movement of subjects within the frame
7252 might be confused for camera motion by the motion vector search.
7254 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7255 then the full frame is used. This allows later options to be set
7256 without specifying the bounding box for the motion vector search.
7258 Default - search the whole frame.
7262 Specify the maximum extent of movement in x and y directions in the
7263 range 0-64 pixels. Default 16.
7266 Specify how to generate pixels to fill blanks at the edge of the
7267 frame. Available values are:
7270 Fill zeroes at blank locations
7272 Original image at blank locations
7274 Extruded edge value at blank locations
7276 Mirrored edge at blank locations
7278 Default value is @samp{mirror}.
7281 Specify the blocksize to use for motion search. Range 4-128 pixels,
7285 Specify the contrast threshold for blocks. Only blocks with more than
7286 the specified contrast (difference between darkest and lightest
7287 pixels) will be considered. Range 1-255, default 125.
7290 Specify the search strategy. Available values are:
7293 Set exhaustive search
7295 Set less exhaustive search.
7297 Default value is @samp{exhaustive}.
7300 If set then a detailed log of the motion search is written to the
7307 Remove unwanted contamination of foreground colors, caused by reflected color of
7308 greenscreen or bluescreen.
7310 This filter accepts the following options:
7314 Set what type of despill to use.
7317 Set how spillmap will be generated.
7320 Set how much to get rid of still remaining spill.
7323 Controls amount of red in spill area.
7326 Controls amount of green in spill area.
7327 Should be -1 for greenscreen.
7330 Controls amount of blue in spill area.
7331 Should be -1 for bluescreen.
7334 Controls brightness of spill area, preserving colors.
7337 Modify alpha from generated spillmap.
7342 Apply an exact inverse of the telecine operation. It requires a predefined
7343 pattern specified using the pattern option which must be the same as that passed
7344 to the telecine filter.
7346 This filter accepts the following options:
7355 The default value is @code{top}.
7359 A string of numbers representing the pulldown pattern you wish to apply.
7360 The default value is @code{23}.
7363 A number representing position of the first frame with respect to the telecine
7364 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7369 Apply dilation effect to the video.
7371 This filter replaces the pixel by the local(3x3) maximum.
7373 It accepts the following options:
7380 Limit the maximum change for each plane, default is 65535.
7381 If 0, plane will remain unchanged.
7384 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7387 Flags to local 3x3 coordinates maps like this:
7396 Displace pixels as indicated by second and third input stream.
7398 It takes three input streams and outputs one stream, the first input is the
7399 source, and second and third input are displacement maps.
7401 The second input specifies how much to displace pixels along the
7402 x-axis, while the third input specifies how much to displace pixels
7404 If one of displacement map streams terminates, last frame from that
7405 displacement map will be used.
7407 Note that once generated, displacements maps can be reused over and over again.
7409 A description of the accepted options follows.
7413 Set displace behavior for pixels that are out of range.
7415 Available values are:
7418 Missing pixels are replaced by black pixels.
7421 Adjacent pixels will spread out to replace missing pixels.
7424 Out of range pixels are wrapped so they point to pixels of other side.
7427 Out of range pixels will be replaced with mirrored pixels.
7429 Default is @samp{smear}.
7433 @subsection Examples
7437 Add ripple effect to rgb input of video size hd720:
7439 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
7443 Add wave effect to rgb input of video size hd720:
7445 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
7451 Draw a colored box on the input image.
7453 It accepts the following parameters:
7458 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7462 The expressions which specify the width and height of the box; if 0 they are interpreted as
7463 the input width and height. It defaults to 0.
7466 Specify the color of the box to write. For the general syntax of this option,
7467 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7468 value @code{invert} is used, the box edge color is the same as the
7469 video with inverted luma.
7472 The expression which sets the thickness of the box edge.
7473 A value of @code{fill} will create a filled box. Default value is @code{3}.
7475 See below for the list of accepted constants.
7478 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7479 will overwrite the video's color and alpha pixels.
7480 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7483 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7484 following constants:
7488 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7492 horizontal and vertical chroma subsample values. For example for the
7493 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7497 The input width and height.
7500 The input sample aspect ratio.
7504 The x and y offset coordinates where the box is drawn.
7508 The width and height of the drawn box.
7511 The thickness of the drawn box.
7513 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7514 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7518 @subsection Examples
7522 Draw a black box around the edge of the input image:
7528 Draw a box with color red and an opacity of 50%:
7530 drawbox=10:20:200:60:red@@0.5
7533 The previous example can be specified as:
7535 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7539 Fill the box with pink color:
7541 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7545 Draw a 2-pixel red 2.40:1 mask:
7547 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
7553 Draw a grid on the input image.
7555 It accepts the following parameters:
7560 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7564 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7565 input width and height, respectively, minus @code{thickness}, so image gets
7566 framed. Default to 0.
7569 Specify the color of the grid. For the general syntax of this option,
7570 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7571 value @code{invert} is used, the grid color is the same as the
7572 video with inverted luma.
7575 The expression which sets the thickness of the grid line. Default value is @code{1}.
7577 See below for the list of accepted constants.
7580 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7581 will overwrite the video's color and alpha pixels.
7582 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7585 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7586 following constants:
7590 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7594 horizontal and vertical chroma subsample values. For example for the
7595 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7599 The input grid cell width and height.
7602 The input sample aspect ratio.
7606 The x and y coordinates of some point of grid intersection (meant to configure offset).
7610 The width and height of the drawn cell.
7613 The thickness of the drawn cell.
7615 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7616 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7620 @subsection Examples
7624 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7626 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7630 Draw a white 3x3 grid with an opacity of 50%:
7632 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7639 Draw a text string or text from a specified file on top of a video, using the
7640 libfreetype library.
7642 To enable compilation of this filter, you need to configure FFmpeg with
7643 @code{--enable-libfreetype}.
7644 To enable default font fallback and the @var{font} option you need to
7645 configure FFmpeg with @code{--enable-libfontconfig}.
7646 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7647 @code{--enable-libfribidi}.
7651 It accepts the following parameters:
7656 Used to draw a box around text using the background color.
7657 The value must be either 1 (enable) or 0 (disable).
7658 The default value of @var{box} is 0.
7661 Set the width of the border to be drawn around the box using @var{boxcolor}.
7662 The default value of @var{boxborderw} is 0.
7665 The color to be used for drawing box around text. For the syntax of this
7666 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7668 The default value of @var{boxcolor} is "white".
7671 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7672 The default value of @var{line_spacing} is 0.
7675 Set the width of the border to be drawn around the text using @var{bordercolor}.
7676 The default value of @var{borderw} is 0.
7679 Set the color to be used for drawing border around text. For the syntax of this
7680 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7682 The default value of @var{bordercolor} is "black".
7685 Select how the @var{text} is expanded. Can be either @code{none},
7686 @code{strftime} (deprecated) or
7687 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7691 Set a start time for the count. Value is in microseconds. Only applied
7692 in the deprecated strftime expansion mode. To emulate in normal expansion
7693 mode use the @code{pts} function, supplying the start time (in seconds)
7694 as the second argument.
7697 If true, check and fix text coords to avoid clipping.
7700 The color to be used for drawing fonts. For the syntax of this option, check
7701 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7703 The default value of @var{fontcolor} is "black".
7705 @item fontcolor_expr
7706 String which is expanded the same way as @var{text} to obtain dynamic
7707 @var{fontcolor} value. By default this option has empty value and is not
7708 processed. When this option is set, it overrides @var{fontcolor} option.
7711 The font family to be used for drawing text. By default Sans.
7714 The font file to be used for drawing text. The path must be included.
7715 This parameter is mandatory if the fontconfig support is disabled.
7718 Draw the text applying alpha blending. The value can
7719 be a number between 0.0 and 1.0.
7720 The expression accepts the same variables @var{x, y} as well.
7721 The default value is 1.
7722 Please see @var{fontcolor_expr}.
7725 The font size to be used for drawing text.
7726 The default value of @var{fontsize} is 16.
7729 If set to 1, attempt to shape the text (for example, reverse the order of
7730 right-to-left text and join Arabic characters) before drawing it.
7731 Otherwise, just draw the text exactly as given.
7732 By default 1 (if supported).
7735 The flags to be used for loading the fonts.
7737 The flags map the corresponding flags supported by libfreetype, and are
7738 a combination of the following values:
7745 @item vertical_layout
7746 @item force_autohint
7749 @item ignore_global_advance_width
7751 @item ignore_transform
7757 Default value is "default".
7759 For more information consult the documentation for the FT_LOAD_*
7763 The color to be used for drawing a shadow behind the drawn text. For the
7764 syntax of this option, check the @ref{color syntax,,"Color" section in the
7765 ffmpeg-utils manual,ffmpeg-utils}.
7767 The default value of @var{shadowcolor} is "black".
7771 The x and y offsets for the text shadow position with respect to the
7772 position of the text. They can be either positive or negative
7773 values. The default value for both is "0".
7776 The starting frame number for the n/frame_num variable. The default value
7780 The size in number of spaces to use for rendering the tab.
7784 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7785 format. It can be used with or without text parameter. @var{timecode_rate}
7786 option must be specified.
7788 @item timecode_rate, rate, r
7789 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7790 integer. Minimum value is "1".
7791 Drop-frame timecode is supported for frame rates 30 & 60.
7794 If set to 1, the output of the timecode option will wrap around at 24 hours.
7795 Default is 0 (disabled).
7798 The text string to be drawn. The text must be a sequence of UTF-8
7800 This parameter is mandatory if no file is specified with the parameter
7804 A text file containing text to be drawn. The text must be a sequence
7805 of UTF-8 encoded characters.
7807 This parameter is mandatory if no text string is specified with the
7808 parameter @var{text}.
7810 If both @var{text} and @var{textfile} are specified, an error is thrown.
7813 If set to 1, the @var{textfile} will be reloaded before each frame.
7814 Be sure to update it atomically, or it may be read partially, or even fail.
7818 The expressions which specify the offsets where text will be drawn
7819 within the video frame. They are relative to the top/left border of the
7822 The default value of @var{x} and @var{y} is "0".
7824 See below for the list of accepted constants and functions.
7827 The parameters for @var{x} and @var{y} are expressions containing the
7828 following constants and functions:
7832 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7836 horizontal and vertical chroma subsample values. For example for the
7837 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7840 the height of each text line
7848 @item max_glyph_a, ascent
7849 the maximum distance from the baseline to the highest/upper grid
7850 coordinate used to place a glyph outline point, for all the rendered
7852 It is a positive value, due to the grid's orientation with the Y axis
7855 @item max_glyph_d, descent
7856 the maximum distance from the baseline to the lowest grid coordinate
7857 used to place a glyph outline point, for all the rendered glyphs.
7858 This is a negative value, due to the grid's orientation, with the Y axis
7862 maximum glyph height, that is the maximum height for all the glyphs
7863 contained in the rendered text, it is equivalent to @var{ascent} -
7867 maximum glyph width, that is the maximum width for all the glyphs
7868 contained in the rendered text
7871 the number of input frame, starting from 0
7873 @item rand(min, max)
7874 return a random number included between @var{min} and @var{max}
7877 The input sample aspect ratio.
7880 timestamp expressed in seconds, NAN if the input timestamp is unknown
7883 the height of the rendered text
7886 the width of the rendered text
7890 the x and y offset coordinates where the text is drawn.
7892 These parameters allow the @var{x} and @var{y} expressions to refer
7893 each other, so you can for example specify @code{y=x/dar}.
7896 @anchor{drawtext_expansion}
7897 @subsection Text expansion
7899 If @option{expansion} is set to @code{strftime},
7900 the filter recognizes strftime() sequences in the provided text and
7901 expands them accordingly. Check the documentation of strftime(). This
7902 feature is deprecated.
7904 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7906 If @option{expansion} is set to @code{normal} (which is the default),
7907 the following expansion mechanism is used.
7909 The backslash character @samp{\}, followed by any character, always expands to
7910 the second character.
7912 Sequences of the form @code{%@{...@}} are expanded. The text between the
7913 braces is a function name, possibly followed by arguments separated by ':'.
7914 If the arguments contain special characters or delimiters (':' or '@}'),
7915 they should be escaped.
7917 Note that they probably must also be escaped as the value for the
7918 @option{text} option in the filter argument string and as the filter
7919 argument in the filtergraph description, and possibly also for the shell,
7920 that makes up to four levels of escaping; using a text file avoids these
7923 The following functions are available:
7928 The expression evaluation result.
7930 It must take one argument specifying the expression to be evaluated,
7931 which accepts the same constants and functions as the @var{x} and
7932 @var{y} values. Note that not all constants should be used, for
7933 example the text size is not known when evaluating the expression, so
7934 the constants @var{text_w} and @var{text_h} will have an undefined
7937 @item expr_int_format, eif
7938 Evaluate the expression's value and output as formatted integer.
7940 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7941 The second argument specifies the output format. Allowed values are @samp{x},
7942 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7943 @code{printf} function.
7944 The third parameter is optional and sets the number of positions taken by the output.
7945 It can be used to add padding with zeros from the left.
7948 The time at which the filter is running, expressed in UTC.
7949 It can accept an argument: a strftime() format string.
7952 The time at which the filter is running, expressed in the local time zone.
7953 It can accept an argument: a strftime() format string.
7956 Frame metadata. Takes one or two arguments.
7958 The first argument is mandatory and specifies the metadata key.
7960 The second argument is optional and specifies a default value, used when the
7961 metadata key is not found or empty.
7964 The frame number, starting from 0.
7967 A 1 character description of the current picture type.
7970 The timestamp of the current frame.
7971 It can take up to three arguments.
7973 The first argument is the format of the timestamp; it defaults to @code{flt}
7974 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7975 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7976 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7977 @code{localtime} stands for the timestamp of the frame formatted as
7978 local time zone time.
7980 The second argument is an offset added to the timestamp.
7982 If the format is set to @code{localtime} or @code{gmtime},
7983 a third argument may be supplied: a strftime() format string.
7984 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7987 @subsection Examples
7991 Draw "Test Text" with font FreeSerif, using the default values for the
7992 optional parameters.
7995 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7999 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8000 and y=50 (counting from the top-left corner of the screen), text is
8001 yellow with a red box around it. Both the text and the box have an
8005 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8006 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8009 Note that the double quotes are not necessary if spaces are not used
8010 within the parameter list.
8013 Show the text at the center of the video frame:
8015 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8019 Show the text at a random position, switching to a new position every 30 seconds:
8021 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)"
8025 Show a text line sliding from right to left in the last row of the video
8026 frame. The file @file{LONG_LINE} is assumed to contain a single line
8029 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8033 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8035 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8039 Draw a single green letter "g", at the center of the input video.
8040 The glyph baseline is placed at half screen height.
8042 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8046 Show text for 1 second every 3 seconds:
8048 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8052 Use fontconfig to set the font. Note that the colons need to be escaped.
8054 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8058 Print the date of a real-time encoding (see strftime(3)):
8060 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8064 Show text fading in and out (appearing/disappearing):
8067 DS=1.0 # display start
8068 DE=10.0 # display end
8069 FID=1.5 # fade in duration
8070 FOD=5 # fade out duration
8071 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 @}"
8075 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8076 and the @option{fontsize} value are included in the @option{y} offset.
8078 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8079 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8084 For more information about libfreetype, check:
8085 @url{http://www.freetype.org/}.
8087 For more information about fontconfig, check:
8088 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8090 For more information about libfribidi, check:
8091 @url{http://fribidi.org/}.
8095 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8097 The filter accepts the following options:
8102 Set low and high threshold values used by the Canny thresholding
8105 The high threshold selects the "strong" edge pixels, which are then
8106 connected through 8-connectivity with the "weak" edge pixels selected
8107 by the low threshold.
8109 @var{low} and @var{high} threshold values must be chosen in the range
8110 [0,1], and @var{low} should be lesser or equal to @var{high}.
8112 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8116 Define the drawing mode.
8120 Draw white/gray wires on black background.
8123 Mix the colors to create a paint/cartoon effect.
8126 Default value is @var{wires}.
8129 @subsection Examples
8133 Standard edge detection with custom values for the hysteresis thresholding:
8135 edgedetect=low=0.1:high=0.4
8139 Painting effect without thresholding:
8141 edgedetect=mode=colormix:high=0
8146 Set brightness, contrast, saturation and approximate gamma adjustment.
8148 The filter accepts the following options:
8152 Set the contrast expression. The value must be a float value in range
8153 @code{-2.0} to @code{2.0}. The default value is "1".
8156 Set the brightness expression. The value must be a float value in
8157 range @code{-1.0} to @code{1.0}. The default value is "0".
8160 Set the saturation expression. The value must be a float in
8161 range @code{0.0} to @code{3.0}. The default value is "1".
8164 Set the gamma expression. The value must be a float in range
8165 @code{0.1} to @code{10.0}. The default value is "1".
8168 Set the gamma expression for red. The value must be a float in
8169 range @code{0.1} to @code{10.0}. The default value is "1".
8172 Set the gamma expression for green. The value must be a float in range
8173 @code{0.1} to @code{10.0}. The default value is "1".
8176 Set the gamma expression for blue. The value must be a float in range
8177 @code{0.1} to @code{10.0}. The default value is "1".
8180 Set the gamma weight expression. It can be used to reduce the effect
8181 of a high gamma value on bright image areas, e.g. keep them from
8182 getting overamplified and just plain white. The value must be a float
8183 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8184 gamma correction all the way down while @code{1.0} leaves it at its
8185 full strength. Default is "1".
8188 Set when the expressions for brightness, contrast, saturation and
8189 gamma expressions are evaluated.
8191 It accepts the following values:
8194 only evaluate expressions once during the filter initialization or
8195 when a command is processed
8198 evaluate expressions for each incoming frame
8201 Default value is @samp{init}.
8204 The expressions accept the following parameters:
8207 frame count of the input frame starting from 0
8210 byte position of the corresponding packet in the input file, NAN if
8214 frame rate of the input video, NAN if the input frame rate is unknown
8217 timestamp expressed in seconds, NAN if the input timestamp is unknown
8220 @subsection Commands
8221 The filter supports the following commands:
8225 Set the contrast expression.
8228 Set the brightness expression.
8231 Set the saturation expression.
8234 Set the gamma expression.
8237 Set the gamma_r expression.
8240 Set gamma_g expression.
8243 Set gamma_b expression.
8246 Set gamma_weight expression.
8248 The command accepts the same syntax of the corresponding option.
8250 If the specified expression is not valid, it is kept at its current
8257 Apply erosion effect to the video.
8259 This filter replaces the pixel by the local(3x3) minimum.
8261 It accepts the following options:
8268 Limit the maximum change for each plane, default is 65535.
8269 If 0, plane will remain unchanged.
8272 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8275 Flags to local 3x3 coordinates maps like this:
8282 @section extractplanes
8284 Extract color channel components from input video stream into
8285 separate grayscale video streams.
8287 The filter accepts the following option:
8291 Set plane(s) to extract.
8293 Available values for planes are:
8304 Choosing planes not available in the input will result in an error.
8305 That means you cannot select @code{r}, @code{g}, @code{b} planes
8306 with @code{y}, @code{u}, @code{v} planes at same time.
8309 @subsection Examples
8313 Extract luma, u and v color channel component from input video frame
8314 into 3 grayscale outputs:
8316 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
8322 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8324 For each input image, the filter will compute the optimal mapping from
8325 the input to the output given the codebook length, that is the number
8326 of distinct output colors.
8328 This filter accepts the following options.
8331 @item codebook_length, l
8332 Set codebook length. The value must be a positive integer, and
8333 represents the number of distinct output colors. Default value is 256.
8336 Set the maximum number of iterations to apply for computing the optimal
8337 mapping. The higher the value the better the result and the higher the
8338 computation time. Default value is 1.
8341 Set a random seed, must be an integer included between 0 and
8342 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8343 will try to use a good random seed on a best effort basis.
8346 Set pal8 output pixel format. This option does not work with codebook
8347 length greater than 256.
8352 Measure graylevel entropy in histogram of color channels of video frames.
8354 It accepts the following parameters:
8358 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8360 @var{diff} mode measures entropy of histogram delta values, absolute differences
8361 between neighbour histogram values.
8366 Apply a fade-in/out effect to the input video.
8368 It accepts the following parameters:
8372 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8374 Default is @code{in}.
8376 @item start_frame, s
8377 Specify the number of the frame to start applying the fade
8378 effect at. Default is 0.
8381 The number of frames that the fade effect lasts. At the end of the
8382 fade-in effect, the output video will have the same intensity as the input video.
8383 At the end of the fade-out transition, the output video will be filled with the
8384 selected @option{color}.
8388 If set to 1, fade only alpha channel, if one exists on the input.
8391 @item start_time, st
8392 Specify the timestamp (in seconds) of the frame to start to apply the fade
8393 effect. If both start_frame and start_time are specified, the fade will start at
8394 whichever comes last. Default is 0.
8397 The number of seconds for which the fade effect has to last. At the end of the
8398 fade-in effect the output video will have the same intensity as the input video,
8399 at the end of the fade-out transition the output video will be filled with the
8400 selected @option{color}.
8401 If both duration and nb_frames are specified, duration is used. Default is 0
8402 (nb_frames is used by default).
8405 Specify the color of the fade. Default is "black".
8408 @subsection Examples
8412 Fade in the first 30 frames of video:
8417 The command above is equivalent to:
8423 Fade out the last 45 frames of a 200-frame video:
8426 fade=type=out:start_frame=155:nb_frames=45
8430 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8432 fade=in:0:25, fade=out:975:25
8436 Make the first 5 frames yellow, then fade in from frame 5-24:
8438 fade=in:5:20:color=yellow
8442 Fade in alpha over first 25 frames of video:
8444 fade=in:0:25:alpha=1
8448 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8450 fade=t=in:st=5.5:d=0.5
8456 Apply arbitrary expressions to samples in frequency domain
8460 Adjust the dc value (gain) of the luma plane of the image. The filter
8461 accepts an integer value in range @code{0} to @code{1000}. The default
8462 value is set to @code{0}.
8465 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8466 filter accepts an integer value in range @code{0} to @code{1000}. The
8467 default value is set to @code{0}.
8470 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8471 filter accepts an integer value in range @code{0} to @code{1000}. The
8472 default value is set to @code{0}.
8475 Set the frequency domain weight expression for the luma plane.
8478 Set the frequency domain weight expression for the 1st chroma plane.
8481 Set the frequency domain weight expression for the 2nd chroma plane.
8484 Set when the expressions are evaluated.
8486 It accepts the following values:
8489 Only evaluate expressions once during the filter initialization.
8492 Evaluate expressions for each incoming frame.
8495 Default value is @samp{init}.
8497 The filter accepts the following variables:
8500 The coordinates of the current sample.
8504 The width and height of the image.
8507 The number of input frame, starting from 0.
8510 @subsection Examples
8516 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8522 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8528 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8534 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8541 Extract a single field from an interlaced image using stride
8542 arithmetic to avoid wasting CPU time. The output frames are marked as
8545 The filter accepts the following options:
8549 Specify whether to extract the top (if the value is @code{0} or
8550 @code{top}) or the bottom field (if the value is @code{1} or
8556 Create new frames by copying the top and bottom fields from surrounding frames
8557 supplied as numbers by the hint file.
8561 Set file containing hints: absolute/relative frame numbers.
8563 There must be one line for each frame in a clip. Each line must contain two
8564 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8565 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8566 is current frame number for @code{absolute} mode or out of [-1, 1] range
8567 for @code{relative} mode. First number tells from which frame to pick up top
8568 field and second number tells from which frame to pick up bottom field.
8570 If optionally followed by @code{+} output frame will be marked as interlaced,
8571 else if followed by @code{-} output frame will be marked as progressive, else
8572 it will be marked same as input frame.
8573 If line starts with @code{#} or @code{;} that line is skipped.
8576 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8579 Example of first several lines of @code{hint} file for @code{relative} mode:
8582 1,0 - # second frame, use third's frame top field and second's frame bottom field
8583 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8600 Field matching filter for inverse telecine. It is meant to reconstruct the
8601 progressive frames from a telecined stream. The filter does not drop duplicated
8602 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8603 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8605 The separation of the field matching and the decimation is notably motivated by
8606 the possibility of inserting a de-interlacing filter fallback between the two.
8607 If the source has mixed telecined and real interlaced content,
8608 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8609 But these remaining combed frames will be marked as interlaced, and thus can be
8610 de-interlaced by a later filter such as @ref{yadif} before decimation.
8612 In addition to the various configuration options, @code{fieldmatch} can take an
8613 optional second stream, activated through the @option{ppsrc} option. If
8614 enabled, the frames reconstruction will be based on the fields and frames from
8615 this second stream. This allows the first input to be pre-processed in order to
8616 help the various algorithms of the filter, while keeping the output lossless
8617 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8618 or brightness/contrast adjustments can help.
8620 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8621 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8622 which @code{fieldmatch} is based on. While the semantic and usage are very
8623 close, some behaviour and options names can differ.
8625 The @ref{decimate} filter currently only works for constant frame rate input.
8626 If your input has mixed telecined (30fps) and progressive content with a lower
8627 framerate like 24fps use the following filterchain to produce the necessary cfr
8628 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8630 The filter accepts the following options:
8634 Specify the assumed field order of the input stream. Available values are:
8638 Auto detect parity (use FFmpeg's internal parity value).
8640 Assume bottom field first.
8642 Assume top field first.
8645 Note that it is sometimes recommended not to trust the parity announced by the
8648 Default value is @var{auto}.
8651 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8652 sense that it won't risk creating jerkiness due to duplicate frames when
8653 possible, but if there are bad edits or blended fields it will end up
8654 outputting combed frames when a good match might actually exist. On the other
8655 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8656 but will almost always find a good frame if there is one. The other values are
8657 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8658 jerkiness and creating duplicate frames versus finding good matches in sections
8659 with bad edits, orphaned fields, blended fields, etc.
8661 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8663 Available values are:
8667 2-way matching (p/c)
8669 2-way matching, and trying 3rd match if still combed (p/c + n)
8671 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8673 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8674 still combed (p/c + n + u/b)
8676 3-way matching (p/c/n)
8678 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8679 detected as combed (p/c/n + u/b)
8682 The parenthesis at the end indicate the matches that would be used for that
8683 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8686 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8689 Default value is @var{pc_n}.
8692 Mark the main input stream as a pre-processed input, and enable the secondary
8693 input stream as the clean source to pick the fields from. See the filter
8694 introduction for more details. It is similar to the @option{clip2} feature from
8697 Default value is @code{0} (disabled).
8700 Set the field to match from. It is recommended to set this to the same value as
8701 @option{order} unless you experience matching failures with that setting. In
8702 certain circumstances changing the field that is used to match from can have a
8703 large impact on matching performance. Available values are:
8707 Automatic (same value as @option{order}).
8709 Match from the bottom field.
8711 Match from the top field.
8714 Default value is @var{auto}.
8717 Set whether or not chroma is included during the match comparisons. In most
8718 cases it is recommended to leave this enabled. You should set this to @code{0}
8719 only if your clip has bad chroma problems such as heavy rainbowing or other
8720 artifacts. Setting this to @code{0} could also be used to speed things up at
8721 the cost of some accuracy.
8723 Default value is @code{1}.
8727 These define an exclusion band which excludes the lines between @option{y0} and
8728 @option{y1} from being included in the field matching decision. An exclusion
8729 band can be used to ignore subtitles, a logo, or other things that may
8730 interfere with the matching. @option{y0} sets the starting scan line and
8731 @option{y1} sets the ending line; all lines in between @option{y0} and
8732 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8733 @option{y0} and @option{y1} to the same value will disable the feature.
8734 @option{y0} and @option{y1} defaults to @code{0}.
8737 Set the scene change detection threshold as a percentage of maximum change on
8738 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8739 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8740 @option{scthresh} is @code{[0.0, 100.0]}.
8742 Default value is @code{12.0}.
8745 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8746 account the combed scores of matches when deciding what match to use as the
8747 final match. Available values are:
8751 No final matching based on combed scores.
8753 Combed scores are only used when a scene change is detected.
8755 Use combed scores all the time.
8758 Default is @var{sc}.
8761 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8762 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8763 Available values are:
8767 No forced calculation.
8769 Force p/c/n calculations.
8771 Force p/c/n/u/b calculations.
8774 Default value is @var{none}.
8777 This is the area combing threshold used for combed frame detection. This
8778 essentially controls how "strong" or "visible" combing must be to be detected.
8779 Larger values mean combing must be more visible and smaller values mean combing
8780 can be less visible or strong and still be detected. Valid settings are from
8781 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8782 be detected as combed). This is basically a pixel difference value. A good
8783 range is @code{[8, 12]}.
8785 Default value is @code{9}.
8788 Sets whether or not chroma is considered in the combed frame decision. Only
8789 disable this if your source has chroma problems (rainbowing, etc.) that are
8790 causing problems for the combed frame detection with chroma enabled. Actually,
8791 using @option{chroma}=@var{0} is usually more reliable, except for the case
8792 where there is chroma only combing in the source.
8794 Default value is @code{0}.
8798 Respectively set the x-axis and y-axis size of the window used during combed
8799 frame detection. This has to do with the size of the area in which
8800 @option{combpel} pixels are required to be detected as combed for a frame to be
8801 declared combed. See the @option{combpel} parameter description for more info.
8802 Possible values are any number that is a power of 2 starting at 4 and going up
8805 Default value is @code{16}.
8808 The number of combed pixels inside any of the @option{blocky} by
8809 @option{blockx} size blocks on the frame for the frame to be detected as
8810 combed. While @option{cthresh} controls how "visible" the combing must be, this
8811 setting controls "how much" combing there must be in any localized area (a
8812 window defined by the @option{blockx} and @option{blocky} settings) on the
8813 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8814 which point no frames will ever be detected as combed). This setting is known
8815 as @option{MI} in TFM/VFM vocabulary.
8817 Default value is @code{80}.
8820 @anchor{p/c/n/u/b meaning}
8821 @subsection p/c/n/u/b meaning
8823 @subsubsection p/c/n
8825 We assume the following telecined stream:
8828 Top fields: 1 2 2 3 4
8829 Bottom fields: 1 2 3 4 4
8832 The numbers correspond to the progressive frame the fields relate to. Here, the
8833 first two frames are progressive, the 3rd and 4th are combed, and so on.
8835 When @code{fieldmatch} is configured to run a matching from bottom
8836 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8841 B 1 2 3 4 4 <-- matching reference
8850 As a result of the field matching, we can see that some frames get duplicated.
8851 To perform a complete inverse telecine, you need to rely on a decimation filter
8852 after this operation. See for instance the @ref{decimate} filter.
8854 The same operation now matching from top fields (@option{field}=@var{top})
8859 T 1 2 2 3 4 <-- matching reference
8869 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8870 basically, they refer to the frame and field of the opposite parity:
8873 @item @var{p} matches the field of the opposite parity in the previous frame
8874 @item @var{c} matches the field of the opposite parity in the current frame
8875 @item @var{n} matches the field of the opposite parity in the next frame
8880 The @var{u} and @var{b} matching are a bit special in the sense that they match
8881 from the opposite parity flag. In the following examples, we assume that we are
8882 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8883 'x' is placed above and below each matched fields.
8885 With bottom matching (@option{field}=@var{bottom}):
8890 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8891 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8899 With top matching (@option{field}=@var{top}):
8904 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8905 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8913 @subsection Examples
8915 Simple IVTC of a top field first telecined stream:
8917 fieldmatch=order=tff:combmatch=none, decimate
8920 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8922 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8927 Transform the field order of the input video.
8929 It accepts the following parameters:
8934 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8935 for bottom field first.
8938 The default value is @samp{tff}.
8940 The transformation is done by shifting the picture content up or down
8941 by one line, and filling the remaining line with appropriate picture content.
8942 This method is consistent with most broadcast field order converters.
8944 If the input video is not flagged as being interlaced, or it is already
8945 flagged as being of the required output field order, then this filter does
8946 not alter the incoming video.
8948 It is very useful when converting to or from PAL DV material,
8949 which is bottom field first.
8953 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8956 @section fifo, afifo
8958 Buffer input images and send them when they are requested.
8960 It is mainly useful when auto-inserted by the libavfilter
8963 It does not take parameters.
8965 @section fillborders
8967 Fill borders of the input video, without changing video stream dimensions.
8968 Sometimes video can have garbage at the four edges and you may not want to
8969 crop video input to keep size multiple of some number.
8971 This filter accepts the following options:
8975 Number of pixels to fill from left border.
8978 Number of pixels to fill from right border.
8981 Number of pixels to fill from top border.
8984 Number of pixels to fill from bottom border.
8989 It accepts the following values:
8992 fill pixels using outermost pixels
8995 fill pixels using mirroring
8998 fill pixels with constant value
9001 Default is @var{smear}.
9004 Set color for pixels in fixed mode. Default is @var{black}.
9009 Find a rectangular object
9011 It accepts the following options:
9015 Filepath of the object image, needs to be in gray8.
9018 Detection threshold, default is 0.5.
9021 Number of mipmaps, default is 3.
9023 @item xmin, ymin, xmax, ymax
9024 Specifies the rectangle in which to search.
9027 @subsection Examples
9031 Generate a representative palette of a given video using @command{ffmpeg}:
9033 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9039 Cover a rectangular object
9041 It accepts the following options:
9045 Filepath of the optional cover image, needs to be in yuv420.
9050 It accepts the following values:
9053 cover it by the supplied image
9055 cover it by interpolating the surrounding pixels
9058 Default value is @var{blur}.
9061 @subsection Examples
9065 Generate a representative palette of a given video using @command{ffmpeg}:
9067 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9073 Flood area with values of same pixel components with another values.
9075 It accepts the following options:
9078 Set pixel x coordinate.
9081 Set pixel y coordinate.
9084 Set source #0 component value.
9087 Set source #1 component value.
9090 Set source #2 component value.
9093 Set source #3 component value.
9096 Set destination #0 component value.
9099 Set destination #1 component value.
9102 Set destination #2 component value.
9105 Set destination #3 component value.
9111 Convert the input video to one of the specified pixel formats.
9112 Libavfilter will try to pick one that is suitable as input to
9115 It accepts the following parameters:
9119 A '|'-separated list of pixel format names, such as
9120 "pix_fmts=yuv420p|monow|rgb24".
9124 @subsection Examples
9128 Convert the input video to the @var{yuv420p} format
9130 format=pix_fmts=yuv420p
9133 Convert the input video to any of the formats in the list
9135 format=pix_fmts=yuv420p|yuv444p|yuv410p
9142 Convert the video to specified constant frame rate by duplicating or dropping
9143 frames as necessary.
9145 It accepts the following parameters:
9149 The desired output frame rate. The default is @code{25}.
9152 Assume the first PTS should be the given value, in seconds. This allows for
9153 padding/trimming at the start of stream. By default, no assumption is made
9154 about the first frame's expected PTS, so no padding or trimming is done.
9155 For example, this could be set to 0 to pad the beginning with duplicates of
9156 the first frame if a video stream starts after the audio stream or to trim any
9157 frames with a negative PTS.
9160 Timestamp (PTS) rounding method.
9162 Possible values are:
9169 round towards -infinity
9171 round towards +infinity
9175 The default is @code{near}.
9178 Action performed when reading the last frame.
9180 Possible values are:
9183 Use same timestamp rounding method as used for other frames.
9185 Pass through last frame if input duration has not been reached yet.
9187 The default is @code{round}.
9191 Alternatively, the options can be specified as a flat string:
9192 @var{fps}[:@var{start_time}[:@var{round}]].
9194 See also the @ref{setpts} filter.
9196 @subsection Examples
9200 A typical usage in order to set the fps to 25:
9206 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9208 fps=fps=film:round=near
9214 Pack two different video streams into a stereoscopic video, setting proper
9215 metadata on supported codecs. The two views should have the same size and
9216 framerate and processing will stop when the shorter video ends. Please note
9217 that you may conveniently adjust view properties with the @ref{scale} and
9220 It accepts the following parameters:
9224 The desired packing format. Supported values are:
9229 The views are next to each other (default).
9232 The views are on top of each other.
9235 The views are packed by line.
9238 The views are packed by column.
9241 The views are temporally interleaved.
9250 # Convert left and right views into a frame-sequential video
9251 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9253 # Convert views into a side-by-side video with the same output resolution as the input
9254 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
9259 Change the frame rate by interpolating new video output frames from the source
9262 This filter is not designed to function correctly with interlaced media. If
9263 you wish to change the frame rate of interlaced media then you are required
9264 to deinterlace before this filter and re-interlace after this filter.
9266 A description of the accepted options follows.
9270 Specify the output frames per second. This option can also be specified
9271 as a value alone. The default is @code{50}.
9274 Specify the start of a range where the output frame will be created as a
9275 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9276 the default is @code{15}.
9279 Specify the end of a range where the output frame will be created as a
9280 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9281 the default is @code{240}.
9284 Specify the level at which a scene change is detected as a value between
9285 0 and 100 to indicate a new scene; a low value reflects a low
9286 probability for the current frame to introduce a new scene, while a higher
9287 value means the current frame is more likely to be one.
9288 The default is @code{8.2}.
9291 Specify flags influencing the filter process.
9293 Available value for @var{flags} is:
9296 @item scene_change_detect, scd
9297 Enable scene change detection using the value of the option @var{scene}.
9298 This flag is enabled by default.
9304 Select one frame every N-th frame.
9306 This filter accepts the following option:
9309 Select frame after every @code{step} frames.
9310 Allowed values are positive integers higher than 0. Default value is @code{1}.
9316 Apply a frei0r effect to the input video.
9318 To enable the compilation of this filter, you need to install the frei0r
9319 header and configure FFmpeg with @code{--enable-frei0r}.
9321 It accepts the following parameters:
9326 The name of the frei0r effect to load. If the environment variable
9327 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9328 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9329 Otherwise, the standard frei0r paths are searched, in this order:
9330 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9331 @file{/usr/lib/frei0r-1/}.
9334 A '|'-separated list of parameters to pass to the frei0r effect.
9338 A frei0r effect parameter can be a boolean (its value is either
9339 "y" or "n"), a double, a color (specified as
9340 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9341 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9342 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9343 a position (specified as @var{X}/@var{Y}, where
9344 @var{X} and @var{Y} are floating point numbers) and/or a string.
9346 The number and types of parameters depend on the loaded effect. If an
9347 effect parameter is not specified, the default value is set.
9349 @subsection Examples
9353 Apply the distort0r effect, setting the first two double parameters:
9355 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9359 Apply the colordistance effect, taking a color as the first parameter:
9361 frei0r=colordistance:0.2/0.3/0.4
9362 frei0r=colordistance:violet
9363 frei0r=colordistance:0x112233
9367 Apply the perspective effect, specifying the top left and top right image
9370 frei0r=perspective:0.2/0.2|0.8/0.2
9374 For more information, see
9375 @url{http://frei0r.dyne.org}
9379 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9381 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9382 processing filter, one of them is performed once per block, not per pixel.
9383 This allows for much higher speed.
9385 The filter accepts the following options:
9389 Set quality. This option defines the number of levels for averaging. It accepts
9390 an integer in the range 4-5. Default value is @code{4}.
9393 Force a constant quantization parameter. It accepts an integer in range 0-63.
9394 If not set, the filter will use the QP from the video stream (if available).
9397 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9398 more details but also more artifacts, while higher values make the image smoother
9399 but also blurrier. Default value is @code{0} − PSNR optimal.
9402 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9403 option may cause flicker since the B-Frames have often larger QP. Default is
9404 @code{0} (not enabled).
9410 Apply Gaussian blur filter.
9412 The filter accepts the following options:
9416 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9419 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9422 Set which planes to filter. By default all planes are filtered.
9425 Set vertical sigma, if negative it will be same as @code{sigma}.
9426 Default is @code{-1}.
9431 The filter accepts the following options:
9435 Set the luminance expression.
9437 Set the chrominance blue expression.
9439 Set the chrominance red expression.
9441 Set the alpha expression.
9443 Set the red expression.
9445 Set the green expression.
9447 Set the blue expression.
9450 The colorspace is selected according to the specified options. If one
9451 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9452 options is specified, the filter will automatically select a YCbCr
9453 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9454 @option{blue_expr} options is specified, it will select an RGB
9457 If one of the chrominance expression is not defined, it falls back on the other
9458 one. If no alpha expression is specified it will evaluate to opaque value.
9459 If none of chrominance expressions are specified, they will evaluate
9460 to the luminance expression.
9462 The expressions can use the following variables and functions:
9466 The sequential number of the filtered frame, starting from @code{0}.
9470 The coordinates of the current sample.
9474 The width and height of the image.
9478 Width and height scale depending on the currently filtered plane. It is the
9479 ratio between the corresponding luma plane number of pixels and the current
9480 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9481 @code{0.5,0.5} for chroma planes.
9484 Time of the current frame, expressed in seconds.
9487 Return the value of the pixel at location (@var{x},@var{y}) of the current
9491 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9495 Return the value of the pixel at location (@var{x},@var{y}) of the
9496 blue-difference chroma plane. Return 0 if there is no such plane.
9499 Return the value of the pixel at location (@var{x},@var{y}) of the
9500 red-difference chroma plane. Return 0 if there is no such plane.
9505 Return the value of the pixel at location (@var{x},@var{y}) of the
9506 red/green/blue component. Return 0 if there is no such component.
9509 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9510 plane. Return 0 if there is no such plane.
9513 For functions, if @var{x} and @var{y} are outside the area, the value will be
9514 automatically clipped to the closer edge.
9516 @subsection Examples
9520 Flip the image horizontally:
9526 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9527 wavelength of 100 pixels:
9529 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9533 Generate a fancy enigmatic moving light:
9535 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
9539 Generate a quick emboss effect:
9541 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9545 Modify RGB components depending on pixel position:
9547 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9551 Create a radial gradient that is the same size as the input (also see
9552 the @ref{vignette} filter):
9554 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9560 Fix the banding artifacts that are sometimes introduced into nearly flat
9561 regions by truncation to 8-bit color depth.
9562 Interpolate the gradients that should go where the bands are, and
9565 It is designed for playback only. Do not use it prior to
9566 lossy compression, because compression tends to lose the dither and
9567 bring back the bands.
9569 It accepts the following parameters:
9574 The maximum amount by which the filter will change any one pixel. This is also
9575 the threshold for detecting nearly flat regions. Acceptable values range from
9576 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9580 The neighborhood to fit the gradient to. A larger radius makes for smoother
9581 gradients, but also prevents the filter from modifying the pixels near detailed
9582 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9583 values will be clipped to the valid range.
9587 Alternatively, the options can be specified as a flat string:
9588 @var{strength}[:@var{radius}]
9590 @subsection Examples
9594 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9600 Specify radius, omitting the strength (which will fall-back to the default
9611 Apply a Hald CLUT to a video stream.
9613 First input is the video stream to process, and second one is the Hald CLUT.
9614 The Hald CLUT input can be a simple picture or a complete video stream.
9616 The filter accepts the following options:
9620 Force termination when the shortest input terminates. Default is @code{0}.
9622 Continue applying the last CLUT after the end of the stream. A value of
9623 @code{0} disable the filter after the last frame of the CLUT is reached.
9624 Default is @code{1}.
9627 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9628 filters share the same internals).
9630 More information about the Hald CLUT can be found on Eskil Steenberg's website
9631 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9633 @subsection Workflow examples
9635 @subsubsection Hald CLUT video stream
9637 Generate an identity Hald CLUT stream altered with various effects:
9639 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
9642 Note: make sure you use a lossless codec.
9644 Then use it with @code{haldclut} to apply it on some random stream:
9646 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9649 The Hald CLUT will be applied to the 10 first seconds (duration of
9650 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9651 to the remaining frames of the @code{mandelbrot} stream.
9653 @subsubsection Hald CLUT with preview
9655 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9656 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9657 biggest possible square starting at the top left of the picture. The remaining
9658 padding pixels (bottom or right) will be ignored. This area can be used to add
9659 a preview of the Hald CLUT.
9661 Typically, the following generated Hald CLUT will be supported by the
9662 @code{haldclut} filter:
9665 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9666 pad=iw+320 [padded_clut];
9667 smptebars=s=320x256, split [a][b];
9668 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9669 [main][b] overlay=W-320" -frames:v 1 clut.png
9672 It contains the original and a preview of the effect of the CLUT: SMPTE color
9673 bars are displayed on the right-top, and below the same color bars processed by
9676 Then, the effect of this Hald CLUT can be visualized with:
9678 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9683 Flip the input video horizontally.
9685 For example, to horizontally flip the input video with @command{ffmpeg}:
9687 ffmpeg -i in.avi -vf "hflip" out.avi
9691 This filter applies a global color histogram equalization on a
9694 It can be used to correct video that has a compressed range of pixel
9695 intensities. The filter redistributes the pixel intensities to
9696 equalize their distribution across the intensity range. It may be
9697 viewed as an "automatically adjusting contrast filter". This filter is
9698 useful only for correcting degraded or poorly captured source
9701 The filter accepts the following options:
9705 Determine the amount of equalization to be applied. As the strength
9706 is reduced, the distribution of pixel intensities more-and-more
9707 approaches that of the input frame. The value must be a float number
9708 in the range [0,1] and defaults to 0.200.
9711 Set the maximum intensity that can generated and scale the output
9712 values appropriately. The strength should be set as desired and then
9713 the intensity can be limited if needed to avoid washing-out. The value
9714 must be a float number in the range [0,1] and defaults to 0.210.
9717 Set the antibanding level. If enabled the filter will randomly vary
9718 the luminance of output pixels by a small amount to avoid banding of
9719 the histogram. Possible values are @code{none}, @code{weak} or
9720 @code{strong}. It defaults to @code{none}.
9725 Compute and draw a color distribution histogram for the input video.
9727 The computed histogram is a representation of the color component
9728 distribution in an image.
9730 Standard histogram displays the color components distribution in an image.
9731 Displays color graph for each color component. Shows distribution of
9732 the Y, U, V, A or R, G, B components, depending on input format, in the
9733 current frame. Below each graph a color component scale meter is shown.
9735 The filter accepts the following options:
9739 Set height of level. Default value is @code{200}.
9740 Allowed range is [50, 2048].
9743 Set height of color scale. Default value is @code{12}.
9744 Allowed range is [0, 40].
9748 It accepts the following values:
9751 Per color component graphs are placed below each other.
9754 Per color component graphs are placed side by side.
9757 Presents information identical to that in the @code{parade}, except
9758 that the graphs representing color components are superimposed directly
9761 Default is @code{stack}.
9764 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9765 Default is @code{linear}.
9768 Set what color components to display.
9769 Default is @code{7}.
9772 Set foreground opacity. Default is @code{0.7}.
9775 Set background opacity. Default is @code{0.5}.
9778 @subsection Examples
9783 Calculate and draw histogram:
9785 ffplay -i input -vf histogram
9793 This is a high precision/quality 3d denoise filter. It aims to reduce
9794 image noise, producing smooth images and making still images really
9795 still. It should enhance compressibility.
9797 It accepts the following optional parameters:
9801 A non-negative floating point number which specifies spatial luma strength.
9804 @item chroma_spatial
9805 A non-negative floating point number which specifies spatial chroma strength.
9806 It defaults to 3.0*@var{luma_spatial}/4.0.
9809 A floating point number which specifies luma temporal strength. It defaults to
9810 6.0*@var{luma_spatial}/4.0.
9813 A floating point number which specifies chroma temporal strength. It defaults to
9814 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9819 Download hardware frames to system memory.
9821 The input must be in hardware frames, and the output a non-hardware format.
9822 Not all formats will be supported on the output - it may be necessary to insert
9823 an additional @option{format} filter immediately following in the graph to get
9824 the output in a supported format.
9828 Map hardware frames to system memory or to another device.
9830 This filter has several different modes of operation; which one is used depends
9831 on the input and output formats:
9834 Hardware frame input, normal frame output
9836 Map the input frames to system memory and pass them to the output. If the
9837 original hardware frame is later required (for example, after overlaying
9838 something else on part of it), the @option{hwmap} filter can be used again
9839 in the next mode to retrieve it.
9841 Normal frame input, hardware frame output
9843 If the input is actually a software-mapped hardware frame, then unmap it -
9844 that is, return the original hardware frame.
9846 Otherwise, a device must be provided. Create new hardware surfaces on that
9847 device for the output, then map them back to the software format at the input
9848 and give those frames to the preceding filter. This will then act like the
9849 @option{hwupload} filter, but may be able to avoid an additional copy when
9850 the input is already in a compatible format.
9852 Hardware frame input and output
9854 A device must be supplied for the output, either directly or with the
9855 @option{derive_device} option. The input and output devices must be of
9856 different types and compatible - the exact meaning of this is
9857 system-dependent, but typically it means that they must refer to the same
9858 underlying hardware context (for example, refer to the same graphics card).
9860 If the input frames were originally created on the output device, then unmap
9861 to retrieve the original frames.
9863 Otherwise, map the frames to the output device - create new hardware frames
9864 on the output corresponding to the frames on the input.
9867 The following additional parameters are accepted:
9871 Set the frame mapping mode. Some combination of:
9874 The mapped frame should be readable.
9876 The mapped frame should be writeable.
9878 The mapping will always overwrite the entire frame.
9880 This may improve performance in some cases, as the original contents of the
9881 frame need not be loaded.
9883 The mapping must not involve any copying.
9885 Indirect mappings to copies of frames are created in some cases where either
9886 direct mapping is not possible or it would have unexpected properties.
9887 Setting this flag ensures that the mapping is direct and will fail if that is
9890 Defaults to @var{read+write} if not specified.
9892 @item derive_device @var{type}
9893 Rather than using the device supplied at initialisation, instead derive a new
9894 device of type @var{type} from the device the input frames exist on.
9897 In a hardware to hardware mapping, map in reverse - create frames in the sink
9898 and map them back to the source. This may be necessary in some cases where
9899 a mapping in one direction is required but only the opposite direction is
9900 supported by the devices being used.
9902 This option is dangerous - it may break the preceding filter in undefined
9903 ways if there are any additional constraints on that filter's output.
9904 Do not use it without fully understanding the implications of its use.
9909 Upload system memory frames to hardware surfaces.
9911 The device to upload to must be supplied when the filter is initialised. If
9912 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9915 @anchor{hwupload_cuda}
9916 @section hwupload_cuda
9918 Upload system memory frames to a CUDA device.
9920 It accepts the following optional parameters:
9924 The number of the CUDA device to use
9929 Apply a high-quality magnification filter designed for pixel art. This filter
9930 was originally created by Maxim Stepin.
9932 It accepts the following option:
9936 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9937 @code{hq3x} and @code{4} for @code{hq4x}.
9938 Default is @code{3}.
9942 Stack input videos horizontally.
9944 All streams must be of same pixel format and of same height.
9946 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9947 to create same output.
9949 The filter accept the following option:
9953 Set number of input streams. Default is 2.
9956 If set to 1, force the output to terminate when the shortest input
9957 terminates. Default value is 0.
9962 Modify the hue and/or the saturation of the input.
9964 It accepts the following parameters:
9968 Specify the hue angle as a number of degrees. It accepts an expression,
9969 and defaults to "0".
9972 Specify the saturation in the [-10,10] range. It accepts an expression and
9976 Specify the hue angle as a number of radians. It accepts an
9977 expression, and defaults to "0".
9980 Specify the brightness in the [-10,10] range. It accepts an expression and
9984 @option{h} and @option{H} are mutually exclusive, and can't be
9985 specified at the same time.
9987 The @option{b}, @option{h}, @option{H} and @option{s} option values are
9988 expressions containing the following constants:
9992 frame count of the input frame starting from 0
9995 presentation timestamp of the input frame expressed in time base units
9998 frame rate of the input video, NAN if the input frame rate is unknown
10001 timestamp expressed in seconds, NAN if the input timestamp is unknown
10004 time base of the input video
10007 @subsection Examples
10011 Set the hue to 90 degrees and the saturation to 1.0:
10017 Same command but expressing the hue in radians:
10023 Rotate hue and make the saturation swing between 0
10024 and 2 over a period of 1 second:
10026 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10030 Apply a 3 seconds saturation fade-in effect starting at 0:
10032 hue="s=min(t/3\,1)"
10035 The general fade-in expression can be written as:
10037 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10041 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10043 hue="s=max(0\, min(1\, (8-t)/3))"
10046 The general fade-out expression can be written as:
10048 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10053 @subsection Commands
10055 This filter supports the following commands:
10061 Modify the hue and/or the saturation and/or brightness of the input video.
10062 The command accepts the same syntax of the corresponding option.
10064 If the specified expression is not valid, it is kept at its current
10068 @section hysteresis
10070 Grow first stream into second stream by connecting components.
10071 This makes it possible to build more robust edge masks.
10073 This filter accepts the following options:
10077 Set which planes will be processed as bitmap, unprocessed planes will be
10078 copied from first stream.
10079 By default value 0xf, all planes will be processed.
10082 Set threshold which is used in filtering. If pixel component value is higher than
10083 this value filter algorithm for connecting components is activated.
10084 By default value is 0.
10089 Detect video interlacing type.
10091 This filter tries to detect if the input frames are interlaced, progressive,
10092 top or bottom field first. It will also try to detect fields that are
10093 repeated between adjacent frames (a sign of telecine).
10095 Single frame detection considers only immediately adjacent frames when classifying each frame.
10096 Multiple frame detection incorporates the classification history of previous frames.
10098 The filter will log these metadata values:
10101 @item single.current_frame
10102 Detected type of current frame using single-frame detection. One of:
10103 ``tff'' (top field first), ``bff'' (bottom field first),
10104 ``progressive'', or ``undetermined''
10107 Cumulative number of frames detected as top field first using single-frame detection.
10110 Cumulative number of frames detected as top field first using multiple-frame detection.
10113 Cumulative number of frames detected as bottom field first using single-frame detection.
10115 @item multiple.current_frame
10116 Detected type of current frame using multiple-frame detection. One of:
10117 ``tff'' (top field first), ``bff'' (bottom field first),
10118 ``progressive'', or ``undetermined''
10121 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10123 @item single.progressive
10124 Cumulative number of frames detected as progressive using single-frame detection.
10126 @item multiple.progressive
10127 Cumulative number of frames detected as progressive using multiple-frame detection.
10129 @item single.undetermined
10130 Cumulative number of frames that could not be classified using single-frame detection.
10132 @item multiple.undetermined
10133 Cumulative number of frames that could not be classified using multiple-frame detection.
10135 @item repeated.current_frame
10136 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10138 @item repeated.neither
10139 Cumulative number of frames with no repeated field.
10142 Cumulative number of frames with the top field repeated from the previous frame's top field.
10144 @item repeated.bottom
10145 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10148 The filter accepts the following options:
10152 Set interlacing threshold.
10154 Set progressive threshold.
10156 Threshold for repeated field detection.
10158 Number of frames after which a given frame's contribution to the
10159 statistics is halved (i.e., it contributes only 0.5 to its
10160 classification). The default of 0 means that all frames seen are given
10161 full weight of 1.0 forever.
10162 @item analyze_interlaced_flag
10163 When this is not 0 then idet will use the specified number of frames to determine
10164 if the interlaced flag is accurate, it will not count undetermined frames.
10165 If the flag is found to be accurate it will be used without any further
10166 computations, if it is found to be inaccurate it will be cleared without any
10167 further computations. This allows inserting the idet filter as a low computational
10168 method to clean up the interlaced flag
10173 Deinterleave or interleave fields.
10175 This filter allows one to process interlaced images fields without
10176 deinterlacing them. Deinterleaving splits the input frame into 2
10177 fields (so called half pictures). Odd lines are moved to the top
10178 half of the output image, even lines to the bottom half.
10179 You can process (filter) them independently and then re-interleave them.
10181 The filter accepts the following options:
10185 @item chroma_mode, c
10186 @item alpha_mode, a
10187 Available values for @var{luma_mode}, @var{chroma_mode} and
10188 @var{alpha_mode} are:
10194 @item deinterleave, d
10195 Deinterleave fields, placing one above the other.
10197 @item interleave, i
10198 Interleave fields. Reverse the effect of deinterleaving.
10200 Default value is @code{none}.
10202 @item luma_swap, ls
10203 @item chroma_swap, cs
10204 @item alpha_swap, as
10205 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10210 Apply inflate effect to the video.
10212 This filter replaces the pixel by the local(3x3) average by taking into account
10213 only values higher than the pixel.
10215 It accepts the following options:
10222 Limit the maximum change for each plane, default is 65535.
10223 If 0, plane will remain unchanged.
10228 Simple interlacing filter from progressive contents. This interleaves upper (or
10229 lower) lines from odd frames with lower (or upper) lines from even frames,
10230 halving the frame rate and preserving image height.
10233 Original Original New Frame
10234 Frame 'j' Frame 'j+1' (tff)
10235 ========== =========== ==================
10236 Line 0 --------------------> Frame 'j' Line 0
10237 Line 1 Line 1 ----> Frame 'j+1' Line 1
10238 Line 2 ---------------------> Frame 'j' Line 2
10239 Line 3 Line 3 ----> Frame 'j+1' Line 3
10241 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10244 It accepts the following optional parameters:
10248 This determines whether the interlaced frame is taken from the even
10249 (tff - default) or odd (bff) lines of the progressive frame.
10252 Vertical lowpass filter to avoid twitter interlacing and
10253 reduce moire patterns.
10257 Disable vertical lowpass filter
10260 Enable linear filter (default)
10263 Enable complex filter. This will slightly less reduce twitter and moire
10264 but better retain detail and subjective sharpness impression.
10271 Deinterlace input video by applying Donald Graft's adaptive kernel
10272 deinterling. Work on interlaced parts of a video to produce
10273 progressive frames.
10275 The description of the accepted parameters follows.
10279 Set the threshold which affects the filter's tolerance when
10280 determining if a pixel line must be processed. It must be an integer
10281 in the range [0,255] and defaults to 10. A value of 0 will result in
10282 applying the process on every pixels.
10285 Paint pixels exceeding the threshold value to white if set to 1.
10289 Set the fields order. Swap fields if set to 1, leave fields alone if
10293 Enable additional sharpening if set to 1. Default is 0.
10296 Enable twoway sharpening if set to 1. Default is 0.
10299 @subsection Examples
10303 Apply default values:
10305 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10309 Enable additional sharpening:
10315 Paint processed pixels in white:
10321 @section lenscorrection
10323 Correct radial lens distortion
10325 This filter can be used to correct for radial distortion as can result from the use
10326 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10327 one can use tools available for example as part of opencv or simply trial-and-error.
10328 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10329 and extract the k1 and k2 coefficients from the resulting matrix.
10331 Note that effectively the same filter is available in the open-source tools Krita and
10332 Digikam from the KDE project.
10334 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10335 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10336 brightness distribution, so you may want to use both filters together in certain
10337 cases, though you will have to take care of ordering, i.e. whether vignetting should
10338 be applied before or after lens correction.
10340 @subsection Options
10342 The filter accepts the following options:
10346 Relative x-coordinate of the focal point of the image, and thereby the center of the
10347 distortion. This value has a range [0,1] and is expressed as fractions of the image
10350 Relative y-coordinate of the focal point of the image, and thereby the center of the
10351 distortion. This value has a range [0,1] and is expressed as fractions of the image
10354 Coefficient of the quadratic correction term. 0.5 means no correction.
10356 Coefficient of the double quadratic correction term. 0.5 means no correction.
10359 The formula that generates the correction is:
10361 @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)
10363 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10364 distances from the focal point in the source and target images, respectively.
10368 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10369 score between two input videos.
10371 The obtained VMAF score is printed through the logging system.
10373 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10374 After installing the library it can be enabled using:
10375 @code{./configure --enable-libvmaf}.
10376 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10378 The filter has following options:
10382 Set the model path which is to be used for SVM.
10383 Default value: @code{"vmaf_v0.6.1.pkl"}
10386 Set the file path to be used to store logs.
10389 Set the format of the log file (xml or json).
10391 @item enable_transform
10392 Enables transform for computing vmaf.
10395 Invokes the phone model which will generate VMAF scores higher than in the
10396 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10399 Enables computing psnr along with vmaf.
10402 Enables computing ssim along with vmaf.
10405 Enables computing ms_ssim along with vmaf.
10408 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10411 This filter also supports the @ref{framesync} options.
10413 On the below examples the input file @file{main.mpg} being processed is
10414 compared with the reference file @file{ref.mpg}.
10417 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10420 Example with options:
10422 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10427 Limits the pixel components values to the specified range [min, max].
10429 The filter accepts the following options:
10433 Lower bound. Defaults to the lowest allowed value for the input.
10436 Upper bound. Defaults to the highest allowed value for the input.
10439 Specify which planes will be processed. Defaults to all available.
10446 The filter accepts the following options:
10450 Set the number of loops. Setting this value to -1 will result in infinite loops.
10454 Set maximal size in number of frames. Default is 0.
10457 Set first frame of loop. Default is 0.
10463 Apply a 3D LUT to an input video.
10465 The filter accepts the following options:
10469 Set the 3D LUT file name.
10471 Currently supported formats:
10483 Select interpolation mode.
10485 Available values are:
10489 Use values from the nearest defined point.
10491 Interpolate values using the 8 points defining a cube.
10493 Interpolate values using a tetrahedron.
10497 This filter also supports the @ref{framesync} options.
10501 Turn certain luma values into transparency.
10503 The filter accepts the following options:
10507 Set the luma which will be used as base for transparency.
10508 Default value is @code{0}.
10511 Set the range of luma values to be keyed out.
10512 Default value is @code{0}.
10515 Set the range of softness. Default value is @code{0}.
10516 Use this to control gradual transition from zero to full transparency.
10519 @section lut, lutrgb, lutyuv
10521 Compute a look-up table for binding each pixel component input value
10522 to an output value, and apply it to the input video.
10524 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10525 to an RGB input video.
10527 These filters accept the following parameters:
10530 set first pixel component expression
10532 set second pixel component expression
10534 set third pixel component expression
10536 set fourth pixel component expression, corresponds to the alpha component
10539 set red component expression
10541 set green component expression
10543 set blue component expression
10545 alpha component expression
10548 set Y/luminance component expression
10550 set U/Cb component expression
10552 set V/Cr component expression
10555 Each of them specifies the expression to use for computing the lookup table for
10556 the corresponding pixel component values.
10558 The exact component associated to each of the @var{c*} options depends on the
10561 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10562 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10564 The expressions can contain the following constants and functions:
10569 The input width and height.
10572 The input value for the pixel component.
10575 The input value, clipped to the @var{minval}-@var{maxval} range.
10578 The maximum value for the pixel component.
10581 The minimum value for the pixel component.
10584 The negated value for the pixel component value, clipped to the
10585 @var{minval}-@var{maxval} range; it corresponds to the expression
10586 "maxval-clipval+minval".
10589 The computed value in @var{val}, clipped to the
10590 @var{minval}-@var{maxval} range.
10592 @item gammaval(gamma)
10593 The computed gamma correction value of the pixel component value,
10594 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10596 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10600 All expressions default to "val".
10602 @subsection Examples
10606 Negate input video:
10608 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10609 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10612 The above is the same as:
10614 lutrgb="r=negval:g=negval:b=negval"
10615 lutyuv="y=negval:u=negval:v=negval"
10625 Remove chroma components, turning the video into a graytone image:
10627 lutyuv="u=128:v=128"
10631 Apply a luma burning effect:
10637 Remove green and blue components:
10643 Set a constant alpha channel value on input:
10645 format=rgba,lutrgb=a="maxval-minval/2"
10649 Correct luminance gamma by a factor of 0.5:
10651 lutyuv=y=gammaval(0.5)
10655 Discard least significant bits of luma:
10657 lutyuv=y='bitand(val, 128+64+32)'
10661 Technicolor like effect:
10663 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10667 @section lut2, tlut2
10669 The @code{lut2} filter takes two input streams and outputs one
10672 The @code{tlut2} (time lut2) filter takes two consecutive frames
10673 from one single stream.
10675 This filter accepts the following parameters:
10678 set first pixel component expression
10680 set second pixel component expression
10682 set third pixel component expression
10684 set fourth pixel component expression, corresponds to the alpha component
10687 Each of them specifies the expression to use for computing the lookup table for
10688 the corresponding pixel component values.
10690 The exact component associated to each of the @var{c*} options depends on the
10693 The expressions can contain the following constants:
10698 The input width and height.
10701 The first input value for the pixel component.
10704 The second input value for the pixel component.
10707 The first input video bit depth.
10710 The second input video bit depth.
10713 All expressions default to "x".
10715 @subsection Examples
10719 Highlight differences between two RGB video streams:
10721 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)'
10725 Highlight differences between two YUV video streams:
10727 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)'
10731 Show max difference between two video streams:
10733 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)))'
10737 @section maskedclamp
10739 Clamp the first input stream with the second input and third input stream.
10741 Returns the value of first stream to be between second input
10742 stream - @code{undershoot} and third input stream + @code{overshoot}.
10744 This filter accepts the following options:
10747 Default value is @code{0}.
10750 Default value is @code{0}.
10753 Set which planes will be processed as bitmap, unprocessed planes will be
10754 copied from first stream.
10755 By default value 0xf, all planes will be processed.
10758 @section maskedmerge
10760 Merge the first input stream with the second input stream using per pixel
10761 weights in the third input stream.
10763 A value of 0 in the third stream pixel component means that pixel component
10764 from first stream is returned unchanged, while maximum value (eg. 255 for
10765 8-bit videos) means that pixel component from second stream is returned
10766 unchanged. Intermediate values define the amount of merging between both
10767 input stream's pixel components.
10769 This filter accepts the following options:
10772 Set which planes will be processed as bitmap, unprocessed planes will be
10773 copied from first stream.
10774 By default value 0xf, all planes will be processed.
10779 Apply motion-compensation deinterlacing.
10781 It needs one field per frame as input and must thus be used together
10782 with yadif=1/3 or equivalent.
10784 This filter accepts the following options:
10787 Set the deinterlacing mode.
10789 It accepts one of the following values:
10794 use iterative motion estimation
10796 like @samp{slow}, but use multiple reference frames.
10798 Default value is @samp{fast}.
10801 Set the picture field parity assumed for the input video. It must be
10802 one of the following values:
10806 assume top field first
10808 assume bottom field first
10811 Default value is @samp{bff}.
10814 Set per-block quantization parameter (QP) used by the internal
10817 Higher values should result in a smoother motion vector field but less
10818 optimal individual vectors. Default value is 1.
10821 @section mergeplanes
10823 Merge color channel components from several video streams.
10825 The filter accepts up to 4 input streams, and merge selected input
10826 planes to the output video.
10828 This filter accepts the following options:
10831 Set input to output plane mapping. Default is @code{0}.
10833 The mappings is specified as a bitmap. It should be specified as a
10834 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10835 mapping for the first plane of the output stream. 'A' sets the number of
10836 the input stream to use (from 0 to 3), and 'a' the plane number of the
10837 corresponding input to use (from 0 to 3). The rest of the mappings is
10838 similar, 'Bb' describes the mapping for the output stream second
10839 plane, 'Cc' describes the mapping for the output stream third plane and
10840 'Dd' describes the mapping for the output stream fourth plane.
10843 Set output pixel format. Default is @code{yuva444p}.
10846 @subsection Examples
10850 Merge three gray video streams of same width and height into single video stream:
10852 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10856 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10858 [a0][a1]mergeplanes=0x00010210:yuva444p
10862 Swap Y and A plane in yuva444p stream:
10864 format=yuva444p,mergeplanes=0x03010200:yuva444p
10868 Swap U and V plane in yuv420p stream:
10870 format=yuv420p,mergeplanes=0x000201:yuv420p
10874 Cast a rgb24 clip to yuv444p:
10876 format=rgb24,mergeplanes=0x000102:yuv444p
10882 Estimate and export motion vectors using block matching algorithms.
10883 Motion vectors are stored in frame side data to be used by other filters.
10885 This filter accepts the following options:
10888 Specify the motion estimation method. Accepts one of the following values:
10892 Exhaustive search algorithm.
10894 Three step search algorithm.
10896 Two dimensional logarithmic search algorithm.
10898 New three step search algorithm.
10900 Four step search algorithm.
10902 Diamond search algorithm.
10904 Hexagon-based search algorithm.
10906 Enhanced predictive zonal search algorithm.
10908 Uneven multi-hexagon search algorithm.
10910 Default value is @samp{esa}.
10913 Macroblock size. Default @code{16}.
10916 Search parameter. Default @code{7}.
10919 @section midequalizer
10921 Apply Midway Image Equalization effect using two video streams.
10923 Midway Image Equalization adjusts a pair of images to have the same
10924 histogram, while maintaining their dynamics as much as possible. It's
10925 useful for e.g. matching exposures from a pair of stereo cameras.
10927 This filter has two inputs and one output, which must be of same pixel format, but
10928 may be of different sizes. The output of filter is first input adjusted with
10929 midway histogram of both inputs.
10931 This filter accepts the following option:
10935 Set which planes to process. Default is @code{15}, which is all available planes.
10938 @section minterpolate
10940 Convert the video to specified frame rate using motion interpolation.
10942 This filter accepts the following options:
10945 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}.
10948 Motion interpolation mode. Following values are accepted:
10951 Duplicate previous or next frame for interpolating new ones.
10953 Blend source frames. Interpolated frame is mean of previous and next frames.
10955 Motion compensated interpolation. Following options are effective when this mode is selected:
10959 Motion compensation mode. Following values are accepted:
10962 Overlapped block motion compensation.
10964 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10966 Default mode is @samp{obmc}.
10969 Motion estimation mode. Following values are accepted:
10972 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
10974 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
10976 Default mode is @samp{bilat}.
10979 The algorithm to be used for motion estimation. Following values are accepted:
10982 Exhaustive search algorithm.
10984 Three step search algorithm.
10986 Two dimensional logarithmic search algorithm.
10988 New three step search algorithm.
10990 Four step search algorithm.
10992 Diamond search algorithm.
10994 Hexagon-based search algorithm.
10996 Enhanced predictive zonal search algorithm.
10998 Uneven multi-hexagon search algorithm.
11000 Default algorithm is @samp{epzs}.
11003 Macroblock size. Default @code{16}.
11006 Motion estimation search parameter. Default @code{32}.
11009 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).
11014 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:
11017 Disable scene change detection.
11019 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11021 Default method is @samp{fdiff}.
11023 @item scd_threshold
11024 Scene change detection threshold. Default is @code{5.0}.
11029 Mix several video input streams into one video stream.
11031 A description of the accepted options follows.
11035 The number of inputs. If unspecified, it defaults to 2.
11038 Specify weight of each input video stream as sequence.
11039 Each weight is separated by space.
11042 Specify how end of stream is determined.
11045 The duration of the longest input. (default)
11048 The duration of the shortest input.
11051 The duration of the first input.
11055 @section mpdecimate
11057 Drop frames that do not differ greatly from the previous frame in
11058 order to reduce frame rate.
11060 The main use of this filter is for very-low-bitrate encoding
11061 (e.g. streaming over dialup modem), but it could in theory be used for
11062 fixing movies that were inverse-telecined incorrectly.
11064 A description of the accepted options follows.
11068 Set the maximum number of consecutive frames which can be dropped (if
11069 positive), or the minimum interval between dropped frames (if
11070 negative). If the value is 0, the frame is dropped disregarding the
11071 number of previous sequentially dropped frames.
11073 Default value is 0.
11078 Set the dropping threshold values.
11080 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11081 represent actual pixel value differences, so a threshold of 64
11082 corresponds to 1 unit of difference for each pixel, or the same spread
11083 out differently over the block.
11085 A frame is a candidate for dropping if no 8x8 blocks differ by more
11086 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11087 meaning the whole image) differ by more than a threshold of @option{lo}.
11089 Default value for @option{hi} is 64*12, default value for @option{lo} is
11090 64*5, and default value for @option{frac} is 0.33.
11096 Negate input video.
11098 It accepts an integer in input; if non-zero it negates the
11099 alpha component (if available). The default value in input is 0.
11103 Denoise frames using Non-Local Means algorithm.
11105 Each pixel is adjusted by looking for other pixels with similar contexts. This
11106 context similarity is defined by comparing their surrounding patches of size
11107 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11110 Note that the research area defines centers for patches, which means some
11111 patches will be made of pixels outside that research area.
11113 The filter accepts the following options.
11117 Set denoising strength.
11123 Same as @option{p} but for chroma planes.
11125 The default value is @var{0} and means automatic.
11131 Same as @option{r} but for chroma planes.
11133 The default value is @var{0} and means automatic.
11138 Deinterlace video using neural network edge directed interpolation.
11140 This filter accepts the following options:
11144 Mandatory option, without binary file filter can not work.
11145 Currently file can be found here:
11146 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11149 Set which frames to deinterlace, by default it is @code{all}.
11150 Can be @code{all} or @code{interlaced}.
11153 Set mode of operation.
11155 Can be one of the following:
11159 Use frame flags, both fields.
11161 Use frame flags, single field.
11163 Use top field only.
11165 Use bottom field only.
11167 Use both fields, top first.
11169 Use both fields, bottom first.
11173 Set which planes to process, by default filter process all frames.
11176 Set size of local neighborhood around each pixel, used by the predictor neural
11179 Can be one of the following:
11192 Set the number of neurons in predictor neural network.
11193 Can be one of the following:
11204 Controls the number of different neural network predictions that are blended
11205 together to compute the final output value. Can be @code{fast}, default or
11209 Set which set of weights to use in the predictor.
11210 Can be one of the following:
11214 weights trained to minimize absolute error
11216 weights trained to minimize squared error
11220 Controls whether or not the prescreener neural network is used to decide
11221 which pixels should be processed by the predictor neural network and which
11222 can be handled by simple cubic interpolation.
11223 The prescreener is trained to know whether cubic interpolation will be
11224 sufficient for a pixel or whether it should be predicted by the predictor nn.
11225 The computational complexity of the prescreener nn is much less than that of
11226 the predictor nn. Since most pixels can be handled by cubic interpolation,
11227 using the prescreener generally results in much faster processing.
11228 The prescreener is pretty accurate, so the difference between using it and not
11229 using it is almost always unnoticeable.
11231 Can be one of the following:
11239 Default is @code{new}.
11242 Set various debugging flags.
11247 Force libavfilter not to use any of the specified pixel formats for the
11248 input to the next filter.
11250 It accepts the following parameters:
11254 A '|'-separated list of pixel format names, such as
11255 pix_fmts=yuv420p|monow|rgb24".
11259 @subsection Examples
11263 Force libavfilter to use a format different from @var{yuv420p} for the
11264 input to the vflip filter:
11266 noformat=pix_fmts=yuv420p,vflip
11270 Convert the input video to any of the formats not contained in the list:
11272 noformat=yuv420p|yuv444p|yuv410p
11278 Add noise on video input frame.
11280 The filter accepts the following options:
11288 Set noise seed for specific pixel component or all pixel components in case
11289 of @var{all_seed}. Default value is @code{123457}.
11291 @item all_strength, alls
11292 @item c0_strength, c0s
11293 @item c1_strength, c1s
11294 @item c2_strength, c2s
11295 @item c3_strength, c3s
11296 Set noise strength for specific pixel component or all pixel components in case
11297 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11299 @item all_flags, allf
11300 @item c0_flags, c0f
11301 @item c1_flags, c1f
11302 @item c2_flags, c2f
11303 @item c3_flags, c3f
11304 Set pixel component flags or set flags for all components if @var{all_flags}.
11305 Available values for component flags are:
11308 averaged temporal noise (smoother)
11310 mix random noise with a (semi)regular pattern
11312 temporal noise (noise pattern changes between frames)
11314 uniform noise (gaussian otherwise)
11318 @subsection Examples
11320 Add temporal and uniform noise to input video:
11322 noise=alls=20:allf=t+u
11327 Normalize RGB video (aka histogram stretching, contrast stretching).
11328 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11330 For each channel of each frame, the filter computes the input range and maps
11331 it linearly to the user-specified output range. The output range defaults
11332 to the full dynamic range from pure black to pure white.
11334 Temporal smoothing can be used on the input range to reduce flickering (rapid
11335 changes in brightness) caused when small dark or bright objects enter or leave
11336 the scene. This is similar to the auto-exposure (automatic gain control) on a
11337 video camera, and, like a video camera, it may cause a period of over- or
11338 under-exposure of the video.
11340 The R,G,B channels can be normalized independently, which may cause some
11341 color shifting, or linked together as a single channel, which prevents
11342 color shifting. Linked normalization preserves hue. Independent normalization
11343 does not, so it can be used to remove some color casts. Independent and linked
11344 normalization can be combined in any ratio.
11346 The normalize filter accepts the following options:
11351 Colors which define the output range. The minimum input value is mapped to
11352 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11353 The defaults are black and white respectively. Specifying white for
11354 @var{blackpt} and black for @var{whitept} will give color-inverted,
11355 normalized video. Shades of grey can be used to reduce the dynamic range
11356 (contrast). Specifying saturated colors here can create some interesting
11360 The number of previous frames to use for temporal smoothing. The input range
11361 of each channel is smoothed using a rolling average over the current frame
11362 and the @var{smoothing} previous frames. The default is 0 (no temporal
11366 Controls the ratio of independent (color shifting) channel normalization to
11367 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11368 independent. Defaults to 1.0 (fully independent).
11371 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11372 expensive no-op. Defaults to 1.0 (full strength).
11376 @subsection Examples
11378 Stretch video contrast to use the full dynamic range, with no temporal
11379 smoothing; may flicker depending on the source content:
11381 normalize=blackpt=black:whitept=white:smoothing=0
11384 As above, but with 50 frames of temporal smoothing; flicker should be
11385 reduced, depending on the source content:
11387 normalize=blackpt=black:whitept=white:smoothing=50
11390 As above, but with hue-preserving linked channel normalization:
11392 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11395 As above, but with half strength:
11397 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11400 Map the darkest input color to red, the brightest input color to cyan:
11402 normalize=blackpt=red:whitept=cyan
11407 Pass the video source unchanged to the output.
11410 Optical Character Recognition
11412 This filter uses Tesseract for optical character recognition.
11414 It accepts the following options:
11418 Set datapath to tesseract data. Default is to use whatever was
11419 set at installation.
11422 Set language, default is "eng".
11425 Set character whitelist.
11428 Set character blacklist.
11431 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11435 Apply a video transform using libopencv.
11437 To enable this filter, install the libopencv library and headers and
11438 configure FFmpeg with @code{--enable-libopencv}.
11440 It accepts the following parameters:
11445 The name of the libopencv filter to apply.
11447 @item filter_params
11448 The parameters to pass to the libopencv filter. If not specified, the default
11449 values are assumed.
11453 Refer to the official libopencv documentation for more precise
11455 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11457 Several libopencv filters are supported; see the following subsections.
11462 Dilate an image by using a specific structuring element.
11463 It corresponds to the libopencv function @code{cvDilate}.
11465 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11467 @var{struct_el} represents a structuring element, and has the syntax:
11468 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11470 @var{cols} and @var{rows} represent the number of columns and rows of
11471 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11472 point, and @var{shape} the shape for the structuring element. @var{shape}
11473 must be "rect", "cross", "ellipse", or "custom".
11475 If the value for @var{shape} is "custom", it must be followed by a
11476 string of the form "=@var{filename}". The file with name
11477 @var{filename} is assumed to represent a binary image, with each
11478 printable character corresponding to a bright pixel. When a custom
11479 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11480 or columns and rows of the read file are assumed instead.
11482 The default value for @var{struct_el} is "3x3+0x0/rect".
11484 @var{nb_iterations} specifies the number of times the transform is
11485 applied to the image, and defaults to 1.
11489 # Use the default values
11492 # Dilate using a structuring element with a 5x5 cross, iterating two times
11493 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11495 # Read the shape from the file diamond.shape, iterating two times.
11496 # The file diamond.shape may contain a pattern of characters like this
11502 # The specified columns and rows are ignored
11503 # but the anchor point coordinates are not
11504 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11509 Erode an image by using a specific structuring element.
11510 It corresponds to the libopencv function @code{cvErode}.
11512 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11513 with the same syntax and semantics as the @ref{dilate} filter.
11517 Smooth the input video.
11519 The filter takes the following parameters:
11520 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11522 @var{type} is the type of smooth filter to apply, and must be one of
11523 the following values: "blur", "blur_no_scale", "median", "gaussian",
11524 or "bilateral". The default value is "gaussian".
11526 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11527 depend on the smooth type. @var{param1} and
11528 @var{param2} accept integer positive values or 0. @var{param3} and
11529 @var{param4} accept floating point values.
11531 The default value for @var{param1} is 3. The default value for the
11532 other parameters is 0.
11534 These parameters correspond to the parameters assigned to the
11535 libopencv function @code{cvSmooth}.
11537 @section oscilloscope
11539 2D Video Oscilloscope.
11541 Useful to measure spatial impulse, step responses, chroma delays, etc.
11543 It accepts the following parameters:
11547 Set scope center x position.
11550 Set scope center y position.
11553 Set scope size, relative to frame diagonal.
11556 Set scope tilt/rotation.
11562 Set trace center x position.
11565 Set trace center y position.
11568 Set trace width, relative to width of frame.
11571 Set trace height, relative to height of frame.
11574 Set which components to trace. By default it traces first three components.
11577 Draw trace grid. By default is enabled.
11580 Draw some statistics. By default is enabled.
11583 Draw scope. By default is enabled.
11586 @subsection Examples
11590 Inspect full first row of video frame.
11592 oscilloscope=x=0.5:y=0:s=1
11596 Inspect full last row of video frame.
11598 oscilloscope=x=0.5:y=1:s=1
11602 Inspect full 5th line of video frame of height 1080.
11604 oscilloscope=x=0.5:y=5/1080:s=1
11608 Inspect full last column of video frame.
11610 oscilloscope=x=1:y=0.5:s=1:t=1
11618 Overlay one video on top of another.
11620 It takes two inputs and has one output. The first input is the "main"
11621 video on which the second input is overlaid.
11623 It accepts the following parameters:
11625 A description of the accepted options follows.
11630 Set the expression for the x and y coordinates of the overlaid video
11631 on the main video. Default value is "0" for both expressions. In case
11632 the expression is invalid, it is set to a huge value (meaning that the
11633 overlay will not be displayed within the output visible area).
11636 See @ref{framesync}.
11639 Set when the expressions for @option{x}, and @option{y} are evaluated.
11641 It accepts the following values:
11644 only evaluate expressions once during the filter initialization or
11645 when a command is processed
11648 evaluate expressions for each incoming frame
11651 Default value is @samp{frame}.
11654 See @ref{framesync}.
11657 Set the format for the output video.
11659 It accepts the following values:
11662 force YUV420 output
11665 force YUV422 output
11668 force YUV444 output
11671 force packed RGB output
11674 force planar RGB output
11677 automatically pick format
11680 Default value is @samp{yuv420}.
11683 See @ref{framesync}.
11686 Set format of alpha of the overlaid video, it can be @var{straight} or
11687 @var{premultiplied}. Default is @var{straight}.
11690 The @option{x}, and @option{y} expressions can contain the following
11696 The main input width and height.
11700 The overlay input width and height.
11704 The computed values for @var{x} and @var{y}. They are evaluated for
11709 horizontal and vertical chroma subsample values of the output
11710 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11714 the number of input frame, starting from 0
11717 the position in the file of the input frame, NAN if unknown
11720 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11724 This filter also supports the @ref{framesync} options.
11726 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11727 when evaluation is done @emph{per frame}, and will evaluate to NAN
11728 when @option{eval} is set to @samp{init}.
11730 Be aware that frames are taken from each input video in timestamp
11731 order, hence, if their initial timestamps differ, it is a good idea
11732 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11733 have them begin in the same zero timestamp, as the example for
11734 the @var{movie} filter does.
11736 You can chain together more overlays but you should test the
11737 efficiency of such approach.
11739 @subsection Commands
11741 This filter supports the following commands:
11745 Modify the x and y of the overlay input.
11746 The command accepts the same syntax of the corresponding option.
11748 If the specified expression is not valid, it is kept at its current
11752 @subsection Examples
11756 Draw the overlay at 10 pixels from the bottom right corner of the main
11759 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11762 Using named options the example above becomes:
11764 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11768 Insert a transparent PNG logo in the bottom left corner of the input,
11769 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11771 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11775 Insert 2 different transparent PNG logos (second logo on bottom
11776 right corner) using the @command{ffmpeg} tool:
11778 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
11782 Add a transparent color layer on top of the main video; @code{WxH}
11783 must specify the size of the main input to the overlay filter:
11785 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11789 Play an original video and a filtered version (here with the deshake
11790 filter) side by side using the @command{ffplay} tool:
11792 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11795 The above command is the same as:
11797 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11801 Make a sliding overlay appearing from the left to the right top part of the
11802 screen starting since time 2:
11804 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11808 Compose output by putting two input videos side to side:
11810 ffmpeg -i left.avi -i right.avi -filter_complex "
11811 nullsrc=size=200x100 [background];
11812 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11813 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11814 [background][left] overlay=shortest=1 [background+left];
11815 [background+left][right] overlay=shortest=1:x=100 [left+right]
11820 Mask 10-20 seconds of a video by applying the delogo filter to a section
11822 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11823 -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]'
11828 Chain several overlays in cascade:
11830 nullsrc=s=200x200 [bg];
11831 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11832 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11833 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11834 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11835 [in3] null, [mid2] overlay=100:100 [out0]
11842 Apply Overcomplete Wavelet denoiser.
11844 The filter accepts the following options:
11850 Larger depth values will denoise lower frequency components more, but
11851 slow down filtering.
11853 Must be an int in the range 8-16, default is @code{8}.
11855 @item luma_strength, ls
11858 Must be a double value in the range 0-1000, default is @code{1.0}.
11860 @item chroma_strength, cs
11861 Set chroma strength.
11863 Must be a double value in the range 0-1000, default is @code{1.0}.
11869 Add paddings to the input image, and place the original input at the
11870 provided @var{x}, @var{y} coordinates.
11872 It accepts the following parameters:
11877 Specify an expression for the size of the output image with the
11878 paddings added. If the value for @var{width} or @var{height} is 0, the
11879 corresponding input size is used for the output.
11881 The @var{width} expression can reference the value set by the
11882 @var{height} expression, and vice versa.
11884 The default value of @var{width} and @var{height} is 0.
11888 Specify the offsets to place the input image at within the padded area,
11889 with respect to the top/left border of the output image.
11891 The @var{x} expression can reference the value set by the @var{y}
11892 expression, and vice versa.
11894 The default value of @var{x} and @var{y} is 0.
11896 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11897 so the input image is centered on the padded area.
11900 Specify the color of the padded area. For the syntax of this option,
11901 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
11902 manual,ffmpeg-utils}.
11904 The default value of @var{color} is "black".
11907 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11909 It accepts the following values:
11913 Only evaluate expressions once during the filter initialization or when
11914 a command is processed.
11917 Evaluate expressions for each incoming frame.
11921 Default value is @samp{init}.
11924 Pad to aspect instead to a resolution.
11928 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11929 options are expressions containing the following constants:
11934 The input video width and height.
11938 These are the same as @var{in_w} and @var{in_h}.
11942 The output width and height (the size of the padded area), as
11943 specified by the @var{width} and @var{height} expressions.
11947 These are the same as @var{out_w} and @var{out_h}.
11951 The x and y offsets as specified by the @var{x} and @var{y}
11952 expressions, or NAN if not yet specified.
11955 same as @var{iw} / @var{ih}
11958 input sample aspect ratio
11961 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11965 The horizontal and vertical chroma subsample values. For example for the
11966 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11969 @subsection Examples
11973 Add paddings with the color "violet" to the input video. The output video
11974 size is 640x480, and the top-left corner of the input video is placed at
11977 pad=640:480:0:40:violet
11980 The example above is equivalent to the following command:
11982 pad=width=640:height=480:x=0:y=40:color=violet
11986 Pad the input to get an output with dimensions increased by 3/2,
11987 and put the input video at the center of the padded area:
11989 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
11993 Pad the input to get a squared output with size equal to the maximum
11994 value between the input width and height, and put the input video at
11995 the center of the padded area:
11997 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12001 Pad the input to get a final w/h ratio of 16:9:
12003 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12007 In case of anamorphic video, in order to set the output display aspect
12008 correctly, it is necessary to use @var{sar} in the expression,
12009 according to the relation:
12011 (ih * X / ih) * sar = output_dar
12012 X = output_dar / sar
12015 Thus the previous example needs to be modified to:
12017 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12021 Double the output size and put the input video in the bottom-right
12022 corner of the output padded area:
12024 pad="2*iw:2*ih:ow-iw:oh-ih"
12028 @anchor{palettegen}
12029 @section palettegen
12031 Generate one palette for a whole video stream.
12033 It accepts the following options:
12037 Set the maximum number of colors to quantize in the palette.
12038 Note: the palette will still contain 256 colors; the unused palette entries
12041 @item reserve_transparent
12042 Create a palette of 255 colors maximum and reserve the last one for
12043 transparency. Reserving the transparency color is useful for GIF optimization.
12044 If not set, the maximum of colors in the palette will be 256. You probably want
12045 to disable this option for a standalone image.
12048 @item transparency_color
12049 Set the color that will be used as background for transparency.
12052 Set statistics mode.
12054 It accepts the following values:
12057 Compute full frame histograms.
12059 Compute histograms only for the part that differs from previous frame. This
12060 might be relevant to give more importance to the moving part of your input if
12061 the background is static.
12063 Compute new histogram for each frame.
12066 Default value is @var{full}.
12069 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12070 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12071 color quantization of the palette. This information is also visible at
12072 @var{info} logging level.
12074 @subsection Examples
12078 Generate a representative palette of a given video using @command{ffmpeg}:
12080 ffmpeg -i input.mkv -vf palettegen palette.png
12084 @section paletteuse
12086 Use a palette to downsample an input video stream.
12088 The filter takes two inputs: one video stream and a palette. The palette must
12089 be a 256 pixels image.
12091 It accepts the following options:
12095 Select dithering mode. Available algorithms are:
12098 Ordered 8x8 bayer dithering (deterministic)
12100 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12101 Note: this dithering is sometimes considered "wrong" and is included as a
12103 @item floyd_steinberg
12104 Floyd and Steingberg dithering (error diffusion)
12106 Frankie Sierra dithering v2 (error diffusion)
12108 Frankie Sierra dithering v2 "Lite" (error diffusion)
12111 Default is @var{sierra2_4a}.
12114 When @var{bayer} dithering is selected, this option defines the scale of the
12115 pattern (how much the crosshatch pattern is visible). A low value means more
12116 visible pattern for less banding, and higher value means less visible pattern
12117 at the cost of more banding.
12119 The option must be an integer value in the range [0,5]. Default is @var{2}.
12122 If set, define the zone to process
12126 Only the changing rectangle will be reprocessed. This is similar to GIF
12127 cropping/offsetting compression mechanism. This option can be useful for speed
12128 if only a part of the image is changing, and has use cases such as limiting the
12129 scope of the error diffusal @option{dither} to the rectangle that bounds the
12130 moving scene (it leads to more deterministic output if the scene doesn't change
12131 much, and as a result less moving noise and better GIF compression).
12134 Default is @var{none}.
12137 Take new palette for each output frame.
12139 @item alpha_threshold
12140 Sets the alpha threshold for transparency. Alpha values above this threshold
12141 will be treated as completely opaque, and values below this threshold will be
12142 treated as completely transparent.
12144 The option must be an integer value in the range [0,255]. Default is @var{128}.
12147 @subsection Examples
12151 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12152 using @command{ffmpeg}:
12154 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12158 @section perspective
12160 Correct perspective of video not recorded perpendicular to the screen.
12162 A description of the accepted parameters follows.
12173 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12174 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12175 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12176 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12177 then the corners of the source will be sent to the specified coordinates.
12179 The expressions can use the following variables:
12184 the width and height of video frame.
12188 Output frame count.
12191 @item interpolation
12192 Set interpolation for perspective correction.
12194 It accepts the following values:
12200 Default value is @samp{linear}.
12203 Set interpretation of coordinate options.
12205 It accepts the following values:
12209 Send point in the source specified by the given coordinates to
12210 the corners of the destination.
12212 @item 1, destination
12214 Send the corners of the source to the point in the destination specified
12215 by the given coordinates.
12217 Default value is @samp{source}.
12221 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12223 It accepts the following values:
12226 only evaluate expressions once during the filter initialization or
12227 when a command is processed
12230 evaluate expressions for each incoming frame
12233 Default value is @samp{init}.
12238 Delay interlaced video by one field time so that the field order changes.
12240 The intended use is to fix PAL movies that have been captured with the
12241 opposite field order to the film-to-video transfer.
12243 A description of the accepted parameters follows.
12249 It accepts the following values:
12252 Capture field order top-first, transfer bottom-first.
12253 Filter will delay the bottom field.
12256 Capture field order bottom-first, transfer top-first.
12257 Filter will delay the top field.
12260 Capture and transfer with the same field order. This mode only exists
12261 for the documentation of the other options to refer to, but if you
12262 actually select it, the filter will faithfully do nothing.
12265 Capture field order determined automatically by field flags, transfer
12267 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12268 basis using field flags. If no field information is available,
12269 then this works just like @samp{u}.
12272 Capture unknown or varying, transfer opposite.
12273 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12274 analyzing the images and selecting the alternative that produces best
12275 match between the fields.
12278 Capture top-first, transfer unknown or varying.
12279 Filter selects among @samp{t} and @samp{p} using image analysis.
12282 Capture bottom-first, transfer unknown or varying.
12283 Filter selects among @samp{b} and @samp{p} using image analysis.
12286 Capture determined by field flags, transfer unknown or varying.
12287 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12288 image analysis. If no field information is available, then this works just
12289 like @samp{U}. This is the default mode.
12292 Both capture and transfer unknown or varying.
12293 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12297 @section pixdesctest
12299 Pixel format descriptor test filter, mainly useful for internal
12300 testing. The output video should be equal to the input video.
12304 format=monow, pixdesctest
12307 can be used to test the monowhite pixel format descriptor definition.
12311 Display sample values of color channels. Mainly useful for checking color
12312 and levels. Minimum supported resolution is 640x480.
12314 The filters accept the following options:
12318 Set scope X position, relative offset on X axis.
12321 Set scope Y position, relative offset on Y axis.
12330 Set window opacity. This window also holds statistics about pixel area.
12333 Set window X position, relative offset on X axis.
12336 Set window Y position, relative offset on Y axis.
12341 Enable the specified chain of postprocessing subfilters using libpostproc. This
12342 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12343 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12344 Each subfilter and some options have a short and a long name that can be used
12345 interchangeably, i.e. dr/dering are the same.
12347 The filters accept the following options:
12351 Set postprocessing subfilters string.
12354 All subfilters share common options to determine their scope:
12358 Honor the quality commands for this subfilter.
12361 Do chrominance filtering, too (default).
12364 Do luminance filtering only (no chrominance).
12367 Do chrominance filtering only (no luminance).
12370 These options can be appended after the subfilter name, separated by a '|'.
12372 Available subfilters are:
12375 @item hb/hdeblock[|difference[|flatness]]
12376 Horizontal deblocking filter
12379 Difference factor where higher values mean more deblocking (default: @code{32}).
12381 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12384 @item vb/vdeblock[|difference[|flatness]]
12385 Vertical deblocking filter
12388 Difference factor where higher values mean more deblocking (default: @code{32}).
12390 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12393 @item ha/hadeblock[|difference[|flatness]]
12394 Accurate horizontal deblocking filter
12397 Difference factor where higher values mean more deblocking (default: @code{32}).
12399 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12402 @item va/vadeblock[|difference[|flatness]]
12403 Accurate vertical deblocking filter
12406 Difference factor where higher values mean more deblocking (default: @code{32}).
12408 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12412 The horizontal and vertical deblocking filters share the difference and
12413 flatness values so you cannot set different horizontal and vertical
12417 @item h1/x1hdeblock
12418 Experimental horizontal deblocking filter
12420 @item v1/x1vdeblock
12421 Experimental vertical deblocking filter
12426 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12429 larger -> stronger filtering
12431 larger -> stronger filtering
12433 larger -> stronger filtering
12436 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12439 Stretch luminance to @code{0-255}.
12442 @item lb/linblenddeint
12443 Linear blend deinterlacing filter that deinterlaces the given block by
12444 filtering all lines with a @code{(1 2 1)} filter.
12446 @item li/linipoldeint
12447 Linear interpolating deinterlacing filter that deinterlaces the given block by
12448 linearly interpolating every second line.
12450 @item ci/cubicipoldeint
12451 Cubic interpolating deinterlacing filter deinterlaces the given block by
12452 cubically interpolating every second line.
12454 @item md/mediandeint
12455 Median deinterlacing filter that deinterlaces the given block by applying a
12456 median filter to every second line.
12458 @item fd/ffmpegdeint
12459 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12460 second line with a @code{(-1 4 2 4 -1)} filter.
12463 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12464 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12466 @item fq/forceQuant[|quantizer]
12467 Overrides the quantizer table from the input with the constant quantizer you
12475 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12478 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12481 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12484 @subsection Examples
12488 Apply horizontal and vertical deblocking, deringing and automatic
12489 brightness/contrast:
12495 Apply default filters without brightness/contrast correction:
12501 Apply default filters and temporal denoiser:
12503 pp=default/tmpnoise|1|2|3
12507 Apply deblocking on luminance only, and switch vertical deblocking on or off
12508 automatically depending on available CPU time:
12515 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12516 similar to spp = 6 with 7 point DCT, where only the center sample is
12519 The filter accepts the following options:
12523 Force a constant quantization parameter. It accepts an integer in range
12524 0 to 63. If not set, the filter will use the QP from the video stream
12528 Set thresholding mode. Available modes are:
12532 Set hard thresholding.
12534 Set soft thresholding (better de-ringing effect, but likely blurrier).
12536 Set medium thresholding (good results, default).
12540 @section premultiply
12541 Apply alpha premultiply effect to input video stream using first plane
12542 of second stream as alpha.
12544 Both streams must have same dimensions and same pixel format.
12546 The filter accepts the following option:
12550 Set which planes will be processed, unprocessed planes will be copied.
12551 By default value 0xf, all planes will be processed.
12554 Do not require 2nd input for processing, instead use alpha plane from input stream.
12558 Apply prewitt operator to input video stream.
12560 The filter accepts the following option:
12564 Set which planes will be processed, unprocessed planes will be copied.
12565 By default value 0xf, all planes will be processed.
12568 Set value which will be multiplied with filtered result.
12571 Set value which will be added to filtered result.
12574 @anchor{program_opencl}
12575 @section program_opencl
12577 Filter video using an OpenCL program.
12582 OpenCL program source file.
12585 Kernel name in program.
12588 Number of inputs to the filter. Defaults to 1.
12591 Size of output frames. Defaults to the same as the first input.
12595 The program source file must contain a kernel function with the given name,
12596 which will be run once for each plane of the output. Each run on a plane
12597 gets enqueued as a separate 2D global NDRange with one work-item for each
12598 pixel to be generated. The global ID offset for each work-item is therefore
12599 the coordinates of a pixel in the destination image.
12601 The kernel function needs to take the following arguments:
12604 Destination image, @var{__write_only image2d_t}.
12606 This image will become the output; the kernel should write all of it.
12608 Frame index, @var{unsigned int}.
12610 This is a counter starting from zero and increasing by one for each frame.
12612 Source images, @var{__read_only image2d_t}.
12614 These are the most recent images on each input. The kernel may read from
12615 them to generate the output, but they can't be written to.
12622 Copy the input to the output (output must be the same size as the input).
12624 __kernel void copy(__write_only image2d_t destination,
12625 unsigned int index,
12626 __read_only image2d_t source)
12628 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
12630 int2 location = (int2)(get_global_id(0), get_global_id(1));
12632 float4 value = read_imagef(source, sampler, location);
12634 write_imagef(destination, location, value);
12639 Apply a simple transformation, rotating the input by an amount increasing
12640 with the index counter. Pixel values are linearly interpolated by the
12641 sampler, and the output need not have the same dimensions as the input.
12643 __kernel void rotate_image(__write_only image2d_t dst,
12644 unsigned int index,
12645 __read_only image2d_t src)
12647 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12648 CLK_FILTER_LINEAR);
12650 float angle = (float)index / 100.0f;
12652 float2 dst_dim = convert_float2(get_image_dim(dst));
12653 float2 src_dim = convert_float2(get_image_dim(src));
12655 float2 dst_cen = dst_dim / 2.0f;
12656 float2 src_cen = src_dim / 2.0f;
12658 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12660 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
12662 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
12663 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
12665 src_pos = src_pos * src_dim / dst_dim;
12667 float2 src_loc = src_pos + src_cen;
12669 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
12670 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
12671 write_imagef(dst, dst_loc, 0.5f);
12673 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
12678 Blend two inputs together, with the amount of each input used varying
12679 with the index counter.
12681 __kernel void blend_images(__write_only image2d_t dst,
12682 unsigned int index,
12683 __read_only image2d_t src1,
12684 __read_only image2d_t src2)
12686 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12687 CLK_FILTER_LINEAR);
12689 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
12691 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12692 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
12693 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
12695 float4 val1 = read_imagef(src1, sampler, src1_loc);
12696 float4 val2 = read_imagef(src2, sampler, src2_loc);
12698 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
12704 @section pseudocolor
12706 Alter frame colors in video with pseudocolors.
12708 This filter accept the following options:
12712 set pixel first component expression
12715 set pixel second component expression
12718 set pixel third component expression
12721 set pixel fourth component expression, corresponds to the alpha component
12724 set component to use as base for altering colors
12727 Each of them specifies the expression to use for computing the lookup table for
12728 the corresponding pixel component values.
12730 The expressions can contain the following constants and functions:
12735 The input width and height.
12738 The input value for the pixel component.
12740 @item ymin, umin, vmin, amin
12741 The minimum allowed component value.
12743 @item ymax, umax, vmax, amax
12744 The maximum allowed component value.
12747 All expressions default to "val".
12749 @subsection Examples
12753 Change too high luma values to gradient:
12755 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'"
12761 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12762 Ratio) between two input videos.
12764 This filter takes in input two input videos, the first input is
12765 considered the "main" source and is passed unchanged to the
12766 output. The second input is used as a "reference" video for computing
12769 Both video inputs must have the same resolution and pixel format for
12770 this filter to work correctly. Also it assumes that both inputs
12771 have the same number of frames, which are compared one by one.
12773 The obtained average PSNR is printed through the logging system.
12775 The filter stores the accumulated MSE (mean squared error) of each
12776 frame, and at the end of the processing it is averaged across all frames
12777 equally, and the following formula is applied to obtain the PSNR:
12780 PSNR = 10*log10(MAX^2/MSE)
12783 Where MAX is the average of the maximum values of each component of the
12786 The description of the accepted parameters follows.
12789 @item stats_file, f
12790 If specified the filter will use the named file to save the PSNR of
12791 each individual frame. When filename equals "-" the data is sent to
12794 @item stats_version
12795 Specifies which version of the stats file format to use. Details of
12796 each format are written below.
12797 Default value is 1.
12799 @item stats_add_max
12800 Determines whether the max value is output to the stats log.
12801 Default value is 0.
12802 Requires stats_version >= 2. If this is set and stats_version < 2,
12803 the filter will return an error.
12806 This filter also supports the @ref{framesync} options.
12808 The file printed if @var{stats_file} is selected, contains a sequence of
12809 key/value pairs of the form @var{key}:@var{value} for each compared
12812 If a @var{stats_version} greater than 1 is specified, a header line precedes
12813 the list of per-frame-pair stats, with key value pairs following the frame
12814 format with the following parameters:
12817 @item psnr_log_version
12818 The version of the log file format. Will match @var{stats_version}.
12821 A comma separated list of the per-frame-pair parameters included in
12825 A description of each shown per-frame-pair parameter follows:
12829 sequential number of the input frame, starting from 1
12832 Mean Square Error pixel-by-pixel average difference of the compared
12833 frames, averaged over all the image components.
12835 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
12836 Mean Square Error pixel-by-pixel average difference of the compared
12837 frames for the component specified by the suffix.
12839 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12840 Peak Signal to Noise ratio of the compared frames for the component
12841 specified by the suffix.
12843 @item max_avg, max_y, max_u, max_v
12844 Maximum allowed value for each channel, and average over all
12850 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12851 [main][ref] psnr="stats_file=stats.log" [out]
12854 On this example the input file being processed is compared with the
12855 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12856 is stored in @file{stats.log}.
12861 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12862 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12865 The pullup filter is designed to take advantage of future context in making
12866 its decisions. This filter is stateless in the sense that it does not lock
12867 onto a pattern to follow, but it instead looks forward to the following
12868 fields in order to identify matches and rebuild progressive frames.
12870 To produce content with an even framerate, insert the fps filter after
12871 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12872 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12874 The filter accepts the following options:
12881 These options set the amount of "junk" to ignore at the left, right, top, and
12882 bottom of the image, respectively. Left and right are in units of 8 pixels,
12883 while top and bottom are in units of 2 lines.
12884 The default is 8 pixels on each side.
12887 Set the strict breaks. Setting this option to 1 will reduce the chances of
12888 filter generating an occasional mismatched frame, but it may also cause an
12889 excessive number of frames to be dropped during high motion sequences.
12890 Conversely, setting it to -1 will make filter match fields more easily.
12891 This may help processing of video where there is slight blurring between
12892 the fields, but may also cause there to be interlaced frames in the output.
12893 Default value is @code{0}.
12896 Set the metric plane to use. It accepts the following values:
12902 Use chroma blue plane.
12905 Use chroma red plane.
12908 This option may be set to use chroma plane instead of the default luma plane
12909 for doing filter's computations. This may improve accuracy on very clean
12910 source material, but more likely will decrease accuracy, especially if there
12911 is chroma noise (rainbow effect) or any grayscale video.
12912 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12913 load and make pullup usable in realtime on slow machines.
12916 For best results (without duplicated frames in the output file) it is
12917 necessary to change the output frame rate. For example, to inverse
12918 telecine NTSC input:
12920 ffmpeg -i input -vf pullup -r 24000/1001 ...
12925 Change video quantization parameters (QP).
12927 The filter accepts the following option:
12931 Set expression for quantization parameter.
12934 The expression is evaluated through the eval API and can contain, among others,
12935 the following constants:
12939 1 if index is not 129, 0 otherwise.
12942 Sequential index starting from -129 to 128.
12945 @subsection Examples
12949 Some equation like:
12957 Flush video frames from internal cache of frames into a random order.
12958 No frame is discarded.
12959 Inspired by @ref{frei0r} nervous filter.
12963 Set size in number of frames of internal cache, in range from @code{2} to
12964 @code{512}. Default is @code{30}.
12967 Set seed for random number generator, must be an integer included between
12968 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12969 less than @code{0}, the filter will try to use a good random seed on a
12973 @section readeia608
12975 Read closed captioning (EIA-608) information from the top lines of a video frame.
12977 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
12978 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
12979 with EIA-608 data (starting from 0). A description of each metadata value follows:
12982 @item lavfi.readeia608.X.cc
12983 The two bytes stored as EIA-608 data (printed in hexadecimal).
12985 @item lavfi.readeia608.X.line
12986 The number of the line on which the EIA-608 data was identified and read.
12989 This filter accepts the following options:
12993 Set the line to start scanning for EIA-608 data. Default is @code{0}.
12996 Set the line to end scanning for EIA-608 data. Default is @code{29}.
12999 Set minimal acceptable amplitude change for sync codes detection.
13000 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13003 Set the ratio of width reserved for sync code detection.
13004 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13007 Set the max peaks height difference for sync code detection.
13008 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13011 Set max peaks period difference for sync code detection.
13012 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13015 Set the first two max start code bits differences.
13016 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13019 Set the minimum ratio of bits height compared to 3rd start code bit.
13020 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13023 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13026 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13029 Enable checking the parity bit. In the event of a parity error, the filter will output
13030 @code{0x00} for that character. Default is false.
13033 @subsection Examples
13037 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13039 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
13045 Read vertical interval timecode (VITC) information from the top lines of a
13048 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13049 timecode value, if a valid timecode has been detected. Further metadata key
13050 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13051 timecode data has been found or not.
13053 This filter accepts the following options:
13057 Set the maximum number of lines to scan for VITC data. If the value is set to
13058 @code{-1} the full video frame is scanned. Default is @code{45}.
13061 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13062 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13065 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13066 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13069 @subsection Examples
13073 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13074 draw @code{--:--:--:--} as a placeholder:
13076 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13082 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13084 Destination pixel at position (X, Y) will be picked from source (x, y) position
13085 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13086 value for pixel will be used for destination pixel.
13088 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13089 will have Xmap/Ymap video stream dimensions.
13090 Xmap and Ymap input video streams are 16bit depth, single channel.
13092 @section removegrain
13094 The removegrain filter is a spatial denoiser for progressive video.
13098 Set mode for the first plane.
13101 Set mode for the second plane.
13104 Set mode for the third plane.
13107 Set mode for the fourth plane.
13110 Range of mode is from 0 to 24. Description of each mode follows:
13114 Leave input plane unchanged. Default.
13117 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13120 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13123 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13126 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13127 This is equivalent to a median filter.
13130 Line-sensitive clipping giving the minimal change.
13133 Line-sensitive clipping, intermediate.
13136 Line-sensitive clipping, intermediate.
13139 Line-sensitive clipping, intermediate.
13142 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13145 Replaces the target pixel with the closest neighbour.
13148 [1 2 1] horizontal and vertical kernel blur.
13154 Bob mode, interpolates top field from the line where the neighbours
13155 pixels are the closest.
13158 Bob mode, interpolates bottom field from the line where the neighbours
13159 pixels are the closest.
13162 Bob mode, interpolates top field. Same as 13 but with a more complicated
13163 interpolation formula.
13166 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13167 interpolation formula.
13170 Clips the pixel with the minimum and maximum of respectively the maximum and
13171 minimum of each pair of opposite neighbour pixels.
13174 Line-sensitive clipping using opposite neighbours whose greatest distance from
13175 the current pixel is minimal.
13178 Replaces the pixel with the average of its 8 neighbours.
13181 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13184 Clips pixels using the averages of opposite neighbour.
13187 Same as mode 21 but simpler and faster.
13190 Small edge and halo removal, but reputed useless.
13196 @section removelogo
13198 Suppress a TV station logo, using an image file to determine which
13199 pixels comprise the logo. It works by filling in the pixels that
13200 comprise the logo with neighboring pixels.
13202 The filter accepts the following options:
13206 Set the filter bitmap file, which can be any image format supported by
13207 libavformat. The width and height of the image file must match those of the
13208 video stream being processed.
13211 Pixels in the provided bitmap image with a value of zero are not
13212 considered part of the logo, non-zero pixels are considered part of
13213 the logo. If you use white (255) for the logo and black (0) for the
13214 rest, you will be safe. For making the filter bitmap, it is
13215 recommended to take a screen capture of a black frame with the logo
13216 visible, and then using a threshold filter followed by the erode
13217 filter once or twice.
13219 If needed, little splotches can be fixed manually. Remember that if
13220 logo pixels are not covered, the filter quality will be much
13221 reduced. Marking too many pixels as part of the logo does not hurt as
13222 much, but it will increase the amount of blurring needed to cover over
13223 the image and will destroy more information than necessary, and extra
13224 pixels will slow things down on a large logo.
13226 @section repeatfields
13228 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13229 fields based on its value.
13233 Reverse a video clip.
13235 Warning: This filter requires memory to buffer the entire clip, so trimming
13238 @subsection Examples
13242 Take the first 5 seconds of a clip, and reverse it.
13249 Apply roberts cross operator to input video stream.
13251 The filter accepts the following option:
13255 Set which planes will be processed, unprocessed planes will be copied.
13256 By default value 0xf, all planes will be processed.
13259 Set value which will be multiplied with filtered result.
13262 Set value which will be added to filtered result.
13267 Rotate video by an arbitrary angle expressed in radians.
13269 The filter accepts the following options:
13271 A description of the optional parameters follows.
13274 Set an expression for the angle by which to rotate the input video
13275 clockwise, expressed as a number of radians. A negative value will
13276 result in a counter-clockwise rotation. By default it is set to "0".
13278 This expression is evaluated for each frame.
13281 Set the output width expression, default value is "iw".
13282 This expression is evaluated just once during configuration.
13285 Set the output height expression, default value is "ih".
13286 This expression is evaluated just once during configuration.
13289 Enable bilinear interpolation if set to 1, a value of 0 disables
13290 it. Default value is 1.
13293 Set the color used to fill the output area not covered by the rotated
13294 image. For the general syntax of this option, check the
13295 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13296 If the special value "none" is selected then no
13297 background is printed (useful for example if the background is never shown).
13299 Default value is "black".
13302 The expressions for the angle and the output size can contain the
13303 following constants and functions:
13307 sequential number of the input frame, starting from 0. It is always NAN
13308 before the first frame is filtered.
13311 time in seconds of the input frame, it is set to 0 when the filter is
13312 configured. It is always NAN before the first frame is filtered.
13316 horizontal and vertical chroma subsample values. For example for the
13317 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13321 the input video width and height
13325 the output width and height, that is the size of the padded area as
13326 specified by the @var{width} and @var{height} expressions
13330 the minimal width/height required for completely containing the input
13331 video rotated by @var{a} radians.
13333 These are only available when computing the @option{out_w} and
13334 @option{out_h} expressions.
13337 @subsection Examples
13341 Rotate the input by PI/6 radians clockwise:
13347 Rotate the input by PI/6 radians counter-clockwise:
13353 Rotate the input by 45 degrees clockwise:
13359 Apply a constant rotation with period T, starting from an angle of PI/3:
13361 rotate=PI/3+2*PI*t/T
13365 Make the input video rotation oscillating with a period of T
13366 seconds and an amplitude of A radians:
13368 rotate=A*sin(2*PI/T*t)
13372 Rotate the video, output size is chosen so that the whole rotating
13373 input video is always completely contained in the output:
13375 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13379 Rotate the video, reduce the output size so that no background is ever
13382 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13386 @subsection Commands
13388 The filter supports the following commands:
13392 Set the angle expression.
13393 The command accepts the same syntax of the corresponding option.
13395 If the specified expression is not valid, it is kept at its current
13401 Apply Shape Adaptive Blur.
13403 The filter accepts the following options:
13406 @item luma_radius, lr
13407 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13408 value is 1.0. A greater value will result in a more blurred image, and
13409 in slower processing.
13411 @item luma_pre_filter_radius, lpfr
13412 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13415 @item luma_strength, ls
13416 Set luma maximum difference between pixels to still be considered, must
13417 be a value in the 0.1-100.0 range, default value is 1.0.
13419 @item chroma_radius, cr
13420 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13421 greater value will result in a more blurred image, and in slower
13424 @item chroma_pre_filter_radius, cpfr
13425 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13427 @item chroma_strength, cs
13428 Set chroma maximum difference between pixels to still be considered,
13429 must be a value in the -0.9-100.0 range.
13432 Each chroma option value, if not explicitly specified, is set to the
13433 corresponding luma option value.
13438 Scale (resize) the input video, using the libswscale library.
13440 The scale filter forces the output display aspect ratio to be the same
13441 of the input, by changing the output sample aspect ratio.
13443 If the input image format is different from the format requested by
13444 the next filter, the scale filter will convert the input to the
13447 @subsection Options
13448 The filter accepts the following options, or any of the options
13449 supported by the libswscale scaler.
13451 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13452 the complete list of scaler options.
13457 Set the output video dimension expression. Default value is the input
13460 If the @var{width} or @var{w} value is 0, the input width is used for
13461 the output. If the @var{height} or @var{h} value is 0, the input height
13462 is used for the output.
13464 If one and only one of the values is -n with n >= 1, the scale filter
13465 will use a value that maintains the aspect ratio of the input image,
13466 calculated from the other specified dimension. After that it will,
13467 however, make sure that the calculated dimension is divisible by n and
13468 adjust the value if necessary.
13470 If both values are -n with n >= 1, the behavior will be identical to
13471 both values being set to 0 as previously detailed.
13473 See below for the list of accepted constants for use in the dimension
13477 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
13481 Only evaluate expressions once during the filter initialization or when a command is processed.
13484 Evaluate expressions for each incoming frame.
13488 Default value is @samp{init}.
13492 Set the interlacing mode. It accepts the following values:
13496 Force interlaced aware scaling.
13499 Do not apply interlaced scaling.
13502 Select interlaced aware scaling depending on whether the source frames
13503 are flagged as interlaced or not.
13506 Default value is @samp{0}.
13509 Set libswscale scaling flags. See
13510 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13511 complete list of values. If not explicitly specified the filter applies
13515 @item param0, param1
13516 Set libswscale input parameters for scaling algorithms that need them. See
13517 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13518 complete documentation. If not explicitly specified the filter applies
13524 Set the video size. For the syntax of this option, check the
13525 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13527 @item in_color_matrix
13528 @item out_color_matrix
13529 Set in/output YCbCr color space type.
13531 This allows the autodetected value to be overridden as well as allows forcing
13532 a specific value used for the output and encoder.
13534 If not specified, the color space type depends on the pixel format.
13540 Choose automatically.
13543 Format conforming to International Telecommunication Union (ITU)
13544 Recommendation BT.709.
13547 Set color space conforming to the United States Federal Communications
13548 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
13551 Set color space conforming to:
13555 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
13558 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
13561 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
13566 Set color space conforming to SMPTE ST 240:1999.
13571 Set in/output YCbCr sample range.
13573 This allows the autodetected value to be overridden as well as allows forcing
13574 a specific value used for the output and encoder. If not specified, the
13575 range depends on the pixel format. Possible values:
13579 Choose automatically.
13582 Set full range (0-255 in case of 8-bit luma).
13584 @item mpeg/limited/tv
13585 Set "MPEG" range (16-235 in case of 8-bit luma).
13588 @item force_original_aspect_ratio
13589 Enable decreasing or increasing output video width or height if necessary to
13590 keep the original aspect ratio. Possible values:
13594 Scale the video as specified and disable this feature.
13597 The output video dimensions will automatically be decreased if needed.
13600 The output video dimensions will automatically be increased if needed.
13604 One useful instance of this option is that when you know a specific device's
13605 maximum allowed resolution, you can use this to limit the output video to
13606 that, while retaining the aspect ratio. For example, device A allows
13607 1280x720 playback, and your video is 1920x800. Using this option (set it to
13608 decrease) and specifying 1280x720 to the command line makes the output
13611 Please note that this is a different thing than specifying -1 for @option{w}
13612 or @option{h}, you still need to specify the output resolution for this option
13617 The values of the @option{w} and @option{h} options are expressions
13618 containing the following constants:
13623 The input width and height
13627 These are the same as @var{in_w} and @var{in_h}.
13631 The output (scaled) width and height
13635 These are the same as @var{out_w} and @var{out_h}
13638 The same as @var{iw} / @var{ih}
13641 input sample aspect ratio
13644 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13648 horizontal and vertical input chroma subsample values. For example for the
13649 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13653 horizontal and vertical output chroma subsample values. For example for the
13654 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13657 @subsection Examples
13661 Scale the input video to a size of 200x100
13666 This is equivalent to:
13677 Specify a size abbreviation for the output size:
13682 which can also be written as:
13688 Scale the input to 2x:
13690 scale=w=2*iw:h=2*ih
13694 The above is the same as:
13696 scale=2*in_w:2*in_h
13700 Scale the input to 2x with forced interlaced scaling:
13702 scale=2*iw:2*ih:interl=1
13706 Scale the input to half size:
13708 scale=w=iw/2:h=ih/2
13712 Increase the width, and set the height to the same size:
13718 Seek Greek harmony:
13725 Increase the height, and set the width to 3/2 of the height:
13727 scale=w=3/2*oh:h=3/5*ih
13731 Increase the size, making the size a multiple of the chroma
13734 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13738 Increase the width to a maximum of 500 pixels,
13739 keeping the same aspect ratio as the input:
13741 scale=w='min(500\, iw*3/2):h=-1'
13745 Make pixels square by combining scale and setsar:
13747 scale='trunc(ih*dar):ih',setsar=1/1
13751 Make pixels square by combining scale and setsar,
13752 making sure the resulting resolution is even (required by some codecs):
13754 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
13758 @subsection Commands
13760 This filter supports the following commands:
13764 Set the output video dimension expression.
13765 The command accepts the same syntax of the corresponding option.
13767 If the specified expression is not valid, it is kept at its current
13773 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13774 format conversion on CUDA video frames. Setting the output width and height
13775 works in the same way as for the @var{scale} filter.
13777 The following additional options are accepted:
13780 The pixel format of the output CUDA frames. If set to the string "same" (the
13781 default), the input format will be kept. Note that automatic format negotiation
13782 and conversion is not yet supported for hardware frames
13785 The interpolation algorithm used for resizing. One of the following:
13792 @item cubic2p_bspline
13793 2-parameter cubic (B=1, C=0)
13795 @item cubic2p_catmullrom
13796 2-parameter cubic (B=0, C=1/2)
13798 @item cubic2p_b05c03
13799 2-parameter cubic (B=1/2, C=3/10)
13811 Scale (resize) the input video, based on a reference video.
13813 See the scale filter for available options, scale2ref supports the same but
13814 uses the reference video instead of the main input as basis. scale2ref also
13815 supports the following additional constants for the @option{w} and
13816 @option{h} options:
13821 The main input video's width and height
13824 The same as @var{main_w} / @var{main_h}
13827 The main input video's sample aspect ratio
13829 @item main_dar, mdar
13830 The main input video's display aspect ratio. Calculated from
13831 @code{(main_w / main_h) * main_sar}.
13835 The main input video's horizontal and vertical chroma subsample values.
13836 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13840 @subsection Examples
13844 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13846 'scale2ref[b][a];[a][b]overlay'
13850 @anchor{selectivecolor}
13851 @section selectivecolor
13853 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13854 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13855 by the "purity" of the color (that is, how saturated it already is).
13857 This filter is similar to the Adobe Photoshop Selective Color tool.
13859 The filter accepts the following options:
13862 @item correction_method
13863 Select color correction method.
13865 Available values are:
13868 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13871 Specified adjustments are relative to the original component value.
13873 Default is @code{absolute}.
13875 Adjustments for red pixels (pixels where the red component is the maximum)
13877 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13879 Adjustments for green pixels (pixels where the green component is the maximum)
13881 Adjustments for cyan pixels (pixels where the red component is the minimum)
13883 Adjustments for blue pixels (pixels where the blue component is the maximum)
13885 Adjustments for magenta pixels (pixels where the green component is the minimum)
13887 Adjustments for white pixels (pixels where all components are greater than 128)
13889 Adjustments for all pixels except pure black and pure white
13891 Adjustments for black pixels (pixels where all components are lesser than 128)
13893 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13896 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13897 4 space separated floating point adjustment values in the [-1,1] range,
13898 respectively to adjust the amount of cyan, magenta, yellow and black for the
13899 pixels of its range.
13901 @subsection Examples
13905 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13906 increase magenta by 27% in blue areas:
13908 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13912 Use a Photoshop selective color preset:
13914 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13918 @anchor{separatefields}
13919 @section separatefields
13921 The @code{separatefields} takes a frame-based video input and splits
13922 each frame into its components fields, producing a new half height clip
13923 with twice the frame rate and twice the frame count.
13925 This filter use field-dominance information in frame to decide which
13926 of each pair of fields to place first in the output.
13927 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13929 @section setdar, setsar
13931 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13934 This is done by changing the specified Sample (aka Pixel) Aspect
13935 Ratio, according to the following equation:
13937 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13940 Keep in mind that the @code{setdar} filter does not modify the pixel
13941 dimensions of the video frame. Also, the display aspect ratio set by
13942 this filter may be changed by later filters in the filterchain,
13943 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13946 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13947 the filter output video.
13949 Note that as a consequence of the application of this filter, the
13950 output display aspect ratio will change according to the equation
13953 Keep in mind that the sample aspect ratio set by the @code{setsar}
13954 filter may be changed by later filters in the filterchain, e.g. if
13955 another "setsar" or a "setdar" filter is applied.
13957 It accepts the following parameters:
13960 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13961 Set the aspect ratio used by the filter.
13963 The parameter can be a floating point number string, an expression, or
13964 a string of the form @var{num}:@var{den}, where @var{num} and
13965 @var{den} are the numerator and denominator of the aspect ratio. If
13966 the parameter is not specified, it is assumed the value "0".
13967 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13971 Set the maximum integer value to use for expressing numerator and
13972 denominator when reducing the expressed aspect ratio to a rational.
13973 Default value is @code{100}.
13977 The parameter @var{sar} is an expression containing
13978 the following constants:
13982 These are approximated values for the mathematical constants e
13983 (Euler's number), pi (Greek pi), and phi (the golden ratio).
13986 The input width and height.
13989 These are the same as @var{w} / @var{h}.
13992 The input sample aspect ratio.
13995 The input display aspect ratio. It is the same as
13996 (@var{w} / @var{h}) * @var{sar}.
13999 Horizontal and vertical chroma subsample values. For example, for the
14000 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14003 @subsection Examples
14008 To change the display aspect ratio to 16:9, specify one of the following:
14015 To change the sample aspect ratio to 10:11, specify:
14021 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14022 1000 in the aspect ratio reduction, use the command:
14024 setdar=ratio=16/9:max=1000
14032 Force field for the output video frame.
14034 The @code{setfield} filter marks the interlace type field for the
14035 output frames. It does not change the input frame, but only sets the
14036 corresponding property, which affects how the frame is treated by
14037 following filters (e.g. @code{fieldorder} or @code{yadif}).
14039 The filter accepts the following options:
14044 Available values are:
14048 Keep the same field property.
14051 Mark the frame as bottom-field-first.
14054 Mark the frame as top-field-first.
14057 Mark the frame as progressive.
14063 Show a line containing various information for each input video frame.
14064 The input video is not modified.
14066 The shown line contains a sequence of key/value pairs of the form
14067 @var{key}:@var{value}.
14069 The following values are shown in the output:
14073 The (sequential) number of the input frame, starting from 0.
14076 The Presentation TimeStamp of the input frame, expressed as a number of
14077 time base units. The time base unit depends on the filter input pad.
14080 The Presentation TimeStamp of the input frame, expressed as a number of
14084 The position of the frame in the input stream, or -1 if this information is
14085 unavailable and/or meaningless (for example in case of synthetic video).
14088 The pixel format name.
14091 The sample aspect ratio of the input frame, expressed in the form
14092 @var{num}/@var{den}.
14095 The size of the input frame. For the syntax of this option, check the
14096 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14099 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14100 for bottom field first).
14103 This is 1 if the frame is a key frame, 0 otherwise.
14106 The picture type of the input frame ("I" for an I-frame, "P" for a
14107 P-frame, "B" for a B-frame, or "?" for an unknown type).
14108 Also refer to the documentation of the @code{AVPictureType} enum and of
14109 the @code{av_get_picture_type_char} function defined in
14110 @file{libavutil/avutil.h}.
14113 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14115 @item plane_checksum
14116 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14117 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14120 @section showpalette
14122 Displays the 256 colors palette of each frame. This filter is only relevant for
14123 @var{pal8} pixel format frames.
14125 It accepts the following option:
14129 Set the size of the box used to represent one palette color entry. Default is
14130 @code{30} (for a @code{30x30} pixel box).
14133 @section shuffleframes
14135 Reorder and/or duplicate and/or drop video frames.
14137 It accepts the following parameters:
14141 Set the destination indexes of input frames.
14142 This is space or '|' separated list of indexes that maps input frames to output
14143 frames. Number of indexes also sets maximal value that each index may have.
14144 '-1' index have special meaning and that is to drop frame.
14147 The first frame has the index 0. The default is to keep the input unchanged.
14149 @subsection Examples
14153 Swap second and third frame of every three frames of the input:
14155 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14159 Swap 10th and 1st frame of every ten frames of the input:
14161 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14165 @section shuffleplanes
14167 Reorder and/or duplicate video planes.
14169 It accepts the following parameters:
14174 The index of the input plane to be used as the first output plane.
14177 The index of the input plane to be used as the second output plane.
14180 The index of the input plane to be used as the third output plane.
14183 The index of the input plane to be used as the fourth output plane.
14187 The first plane has the index 0. The default is to keep the input unchanged.
14189 @subsection Examples
14193 Swap the second and third planes of the input:
14195 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14199 @anchor{signalstats}
14200 @section signalstats
14201 Evaluate various visual metrics that assist in determining issues associated
14202 with the digitization of analog video media.
14204 By default the filter will log these metadata values:
14208 Display the minimal Y value contained within the input frame. Expressed in
14212 Display the Y value at the 10% percentile within the input frame. Expressed in
14216 Display the average Y value within the input frame. Expressed in range of
14220 Display the Y value at the 90% percentile within the input frame. Expressed in
14224 Display the maximum Y value contained within the input frame. Expressed in
14228 Display the minimal U value contained within the input frame. Expressed in
14232 Display the U value at the 10% percentile within the input frame. Expressed in
14236 Display the average U value within the input frame. Expressed in range of
14240 Display the U value at the 90% percentile within the input frame. Expressed in
14244 Display the maximum U value contained within the input frame. Expressed in
14248 Display the minimal V value contained within the input frame. Expressed in
14252 Display the V value at the 10% percentile within the input frame. Expressed in
14256 Display the average V value within the input frame. Expressed in range of
14260 Display the V value at the 90% percentile within the input frame. Expressed in
14264 Display the maximum V value contained within the input frame. Expressed in
14268 Display the minimal saturation value contained within the input frame.
14269 Expressed in range of [0-~181.02].
14272 Display the saturation value at the 10% percentile within the input frame.
14273 Expressed in range of [0-~181.02].
14276 Display the average saturation value within the input frame. Expressed in range
14280 Display the saturation value at the 90% percentile within the input frame.
14281 Expressed in range of [0-~181.02].
14284 Display the maximum saturation value contained within the input frame.
14285 Expressed in range of [0-~181.02].
14288 Display the median value for hue within the input frame. Expressed in range of
14292 Display the average value for hue within the input frame. Expressed in range of
14296 Display the average of sample value difference between all values of the Y
14297 plane in the current frame and corresponding values of the previous input frame.
14298 Expressed in range of [0-255].
14301 Display the average of sample value difference between all values of the U
14302 plane in the current frame and corresponding values of the previous input frame.
14303 Expressed in range of [0-255].
14306 Display the average of sample value difference between all values of the V
14307 plane in the current frame and corresponding values of the previous input frame.
14308 Expressed in range of [0-255].
14311 Display bit depth of Y plane in current frame.
14312 Expressed in range of [0-16].
14315 Display bit depth of U plane in current frame.
14316 Expressed in range of [0-16].
14319 Display bit depth of V plane in current frame.
14320 Expressed in range of [0-16].
14323 The filter accepts the following options:
14329 @option{stat} specify an additional form of image analysis.
14330 @option{out} output video with the specified type of pixel highlighted.
14332 Both options accept the following values:
14336 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14337 unlike the neighboring pixels of the same field. Examples of temporal outliers
14338 include the results of video dropouts, head clogs, or tape tracking issues.
14341 Identify @var{vertical line repetition}. Vertical line repetition includes
14342 similar rows of pixels within a frame. In born-digital video vertical line
14343 repetition is common, but this pattern is uncommon in video digitized from an
14344 analog source. When it occurs in video that results from the digitization of an
14345 analog source it can indicate concealment from a dropout compensator.
14348 Identify pixels that fall outside of legal broadcast range.
14352 Set the highlight color for the @option{out} option. The default color is
14356 @subsection Examples
14360 Output data of various video metrics:
14362 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14366 Output specific data about the minimum and maximum values of the Y plane per frame:
14368 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14372 Playback video while highlighting pixels that are outside of broadcast range in red.
14374 ffplay example.mov -vf signalstats="out=brng:color=red"
14378 Playback video with signalstats metadata drawn over the frame.
14380 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14383 The contents of signalstat_drawtext.txt used in the command are:
14386 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14387 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14388 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14389 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14397 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14398 input. In this case the matching between the inputs can be calculated additionally.
14399 The filter always passes through the first input. The signature of each stream can
14400 be written into a file.
14402 It accepts the following options:
14406 Enable or disable the matching process.
14408 Available values are:
14412 Disable the calculation of a matching (default).
14414 Calculate the matching for the whole video and output whether the whole video
14415 matches or only parts.
14417 Calculate only until a matching is found or the video ends. Should be faster in
14422 Set the number of inputs. The option value must be a non negative integer.
14423 Default value is 1.
14426 Set the path to which the output is written. If there is more than one input,
14427 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14428 integer), that will be replaced with the input number. If no filename is
14429 specified, no output will be written. This is the default.
14432 Choose the output format.
14434 Available values are:
14438 Use the specified binary representation (default).
14440 Use the specified xml representation.
14444 Set threshold to detect one word as similar. The option value must be an integer
14445 greater than zero. The default value is 9000.
14448 Set threshold to detect all words as similar. The option value must be an integer
14449 greater than zero. The default value is 60000.
14452 Set threshold to detect frames as similar. The option value must be an integer
14453 greater than zero. The default value is 116.
14456 Set the minimum length of a sequence in frames to recognize it as matching
14457 sequence. The option value must be a non negative integer value.
14458 The default value is 0.
14461 Set the minimum relation, that matching frames to all frames must have.
14462 The option value must be a double value between 0 and 1. The default value is 0.5.
14465 @subsection Examples
14469 To calculate the signature of an input video and store it in signature.bin:
14471 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14475 To detect whether two videos match and store the signatures in XML format in
14476 signature0.xml and signature1.xml:
14478 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 -
14486 Blur the input video without impacting the outlines.
14488 It accepts the following options:
14491 @item luma_radius, lr
14492 Set the luma radius. The option value must be a float number in
14493 the range [0.1,5.0] that specifies the variance of the gaussian filter
14494 used to blur the image (slower if larger). Default value is 1.0.
14496 @item luma_strength, ls
14497 Set the luma strength. The option value must be a float number
14498 in the range [-1.0,1.0] that configures the blurring. A value included
14499 in [0.0,1.0] will blur the image whereas a value included in
14500 [-1.0,0.0] will sharpen the image. Default value is 1.0.
14502 @item luma_threshold, lt
14503 Set the luma threshold used as a coefficient to determine
14504 whether a pixel should be blurred or not. The option value must be an
14505 integer in the range [-30,30]. A value of 0 will filter all the image,
14506 a value included in [0,30] will filter flat areas and a value included
14507 in [-30,0] will filter edges. Default value is 0.
14509 @item chroma_radius, cr
14510 Set the chroma radius. The option value must be a float number in
14511 the range [0.1,5.0] that specifies the variance of the gaussian filter
14512 used to blur the image (slower if larger). Default value is @option{luma_radius}.
14514 @item chroma_strength, cs
14515 Set the chroma strength. The option value must be a float number
14516 in the range [-1.0,1.0] that configures the blurring. A value included
14517 in [0.0,1.0] will blur the image whereas a value included in
14518 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
14520 @item chroma_threshold, ct
14521 Set the chroma threshold used as a coefficient to determine
14522 whether a pixel should be blurred or not. The option value must be an
14523 integer in the range [-30,30]. A value of 0 will filter all the image,
14524 a value included in [0,30] will filter flat areas and a value included
14525 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
14528 If a chroma option is not explicitly set, the corresponding luma value
14533 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
14535 This filter takes in input two input videos, the first input is
14536 considered the "main" source and is passed unchanged to the
14537 output. The second input is used as a "reference" video for computing
14540 Both video inputs must have the same resolution and pixel format for
14541 this filter to work correctly. Also it assumes that both inputs
14542 have the same number of frames, which are compared one by one.
14544 The filter stores the calculated SSIM of each frame.
14546 The description of the accepted parameters follows.
14549 @item stats_file, f
14550 If specified the filter will use the named file to save the SSIM of
14551 each individual frame. When filename equals "-" the data is sent to
14555 The file printed if @var{stats_file} is selected, contains a sequence of
14556 key/value pairs of the form @var{key}:@var{value} for each compared
14559 A description of each shown parameter follows:
14563 sequential number of the input frame, starting from 1
14565 @item Y, U, V, R, G, B
14566 SSIM of the compared frames for the component specified by the suffix.
14569 SSIM of the compared frames for the whole frame.
14572 Same as above but in dB representation.
14575 This filter also supports the @ref{framesync} options.
14579 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14580 [main][ref] ssim="stats_file=stats.log" [out]
14583 On this example the input file being processed is compared with the
14584 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
14585 is stored in @file{stats.log}.
14587 Another example with both psnr and ssim at same time:
14589 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
14594 Convert between different stereoscopic image formats.
14596 The filters accept the following options:
14600 Set stereoscopic image format of input.
14602 Available values for input image formats are:
14605 side by side parallel (left eye left, right eye right)
14608 side by side crosseye (right eye left, left eye right)
14611 side by side parallel with half width resolution
14612 (left eye left, right eye right)
14615 side by side crosseye with half width resolution
14616 (right eye left, left eye right)
14619 above-below (left eye above, right eye below)
14622 above-below (right eye above, left eye below)
14625 above-below with half height resolution
14626 (left eye above, right eye below)
14629 above-below with half height resolution
14630 (right eye above, left eye below)
14633 alternating frames (left eye first, right eye second)
14636 alternating frames (right eye first, left eye second)
14639 interleaved rows (left eye has top row, right eye starts on next row)
14642 interleaved rows (right eye has top row, left eye starts on next row)
14645 interleaved columns, left eye first
14648 interleaved columns, right eye first
14650 Default value is @samp{sbsl}.
14654 Set stereoscopic image format of output.
14658 side by side parallel (left eye left, right eye right)
14661 side by side crosseye (right eye left, left eye right)
14664 side by side parallel with half width resolution
14665 (left eye left, right eye right)
14668 side by side crosseye with half width resolution
14669 (right eye left, left eye right)
14672 above-below (left eye above, right eye below)
14675 above-below (right eye above, left eye below)
14678 above-below with half height resolution
14679 (left eye above, right eye below)
14682 above-below with half height resolution
14683 (right eye above, left eye below)
14686 alternating frames (left eye first, right eye second)
14689 alternating frames (right eye first, left eye second)
14692 interleaved rows (left eye has top row, right eye starts on next row)
14695 interleaved rows (right eye has top row, left eye starts on next row)
14698 anaglyph red/blue gray
14699 (red filter on left eye, blue filter on right eye)
14702 anaglyph red/green gray
14703 (red filter on left eye, green filter on right eye)
14706 anaglyph red/cyan gray
14707 (red filter on left eye, cyan filter on right eye)
14710 anaglyph red/cyan half colored
14711 (red filter on left eye, cyan filter on right eye)
14714 anaglyph red/cyan color
14715 (red filter on left eye, cyan filter on right eye)
14718 anaglyph red/cyan color optimized with the least squares projection of dubois
14719 (red filter on left eye, cyan filter on right eye)
14722 anaglyph green/magenta gray
14723 (green filter on left eye, magenta filter on right eye)
14726 anaglyph green/magenta half colored
14727 (green filter on left eye, magenta filter on right eye)
14730 anaglyph green/magenta colored
14731 (green filter on left eye, magenta filter on right eye)
14734 anaglyph green/magenta color optimized with the least squares projection of dubois
14735 (green filter on left eye, magenta filter on right eye)
14738 anaglyph yellow/blue gray
14739 (yellow filter on left eye, blue filter on right eye)
14742 anaglyph yellow/blue half colored
14743 (yellow filter on left eye, blue filter on right eye)
14746 anaglyph yellow/blue colored
14747 (yellow filter on left eye, blue filter on right eye)
14750 anaglyph yellow/blue color optimized with the least squares projection of dubois
14751 (yellow filter on left eye, blue filter on right eye)
14754 mono output (left eye only)
14757 mono output (right eye only)
14760 checkerboard, left eye first
14763 checkerboard, right eye first
14766 interleaved columns, left eye first
14769 interleaved columns, right eye first
14775 Default value is @samp{arcd}.
14778 @subsection Examples
14782 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14788 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14794 @section streamselect, astreamselect
14795 Select video or audio streams.
14797 The filter accepts the following options:
14801 Set number of inputs. Default is 2.
14804 Set input indexes to remap to outputs.
14807 @subsection Commands
14809 The @code{streamselect} and @code{astreamselect} filter supports the following
14814 Set input indexes to remap to outputs.
14817 @subsection Examples
14821 Select first 5 seconds 1st stream and rest of time 2nd stream:
14823 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14827 Same as above, but for audio:
14829 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14834 Apply sobel operator to input video stream.
14836 The filter accepts the following option:
14840 Set which planes will be processed, unprocessed planes will be copied.
14841 By default value 0xf, all planes will be processed.
14844 Set value which will be multiplied with filtered result.
14847 Set value which will be added to filtered result.
14853 Apply a simple postprocessing filter that compresses and decompresses the image
14854 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14855 and average the results.
14857 The filter accepts the following options:
14861 Set quality. This option defines the number of levels for averaging. It accepts
14862 an integer in the range 0-6. If set to @code{0}, the filter will have no
14863 effect. A value of @code{6} means the higher quality. For each increment of
14864 that value the speed drops by a factor of approximately 2. Default value is
14868 Force a constant quantization parameter. If not set, the filter will use the QP
14869 from the video stream (if available).
14872 Set thresholding mode. Available modes are:
14876 Set hard thresholding (default).
14878 Set soft thresholding (better de-ringing effect, but likely blurrier).
14881 @item use_bframe_qp
14882 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14883 option may cause flicker since the B-Frames have often larger QP. Default is
14884 @code{0} (not enabled).
14890 Draw subtitles on top of input video using the libass library.
14892 To enable compilation of this filter you need to configure FFmpeg with
14893 @code{--enable-libass}. This filter also requires a build with libavcodec and
14894 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14895 Alpha) subtitles format.
14897 The filter accepts the following options:
14901 Set the filename of the subtitle file to read. It must be specified.
14903 @item original_size
14904 Specify the size of the original video, the video for which the ASS file
14905 was composed. For the syntax of this option, check the
14906 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14907 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14908 correctly scale the fonts if the aspect ratio has been changed.
14911 Set a directory path containing fonts that can be used by the filter.
14912 These fonts will be used in addition to whatever the font provider uses.
14915 Process alpha channel, by default alpha channel is untouched.
14918 Set subtitles input character encoding. @code{subtitles} filter only. Only
14919 useful if not UTF-8.
14921 @item stream_index, si
14922 Set subtitles stream index. @code{subtitles} filter only.
14925 Override default style or script info parameters of the subtitles. It accepts a
14926 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14929 If the first key is not specified, it is assumed that the first value
14930 specifies the @option{filename}.
14932 For example, to render the file @file{sub.srt} on top of the input
14933 video, use the command:
14938 which is equivalent to:
14940 subtitles=filename=sub.srt
14943 To render the default subtitles stream from file @file{video.mkv}, use:
14945 subtitles=video.mkv
14948 To render the second subtitles stream from that file, use:
14950 subtitles=video.mkv:si=1
14953 To make the subtitles stream from @file{sub.srt} appear in transparent green
14954 @code{DejaVu Serif}, use:
14956 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14959 @section super2xsai
14961 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14962 Interpolate) pixel art scaling algorithm.
14964 Useful for enlarging pixel art images without reducing sharpness.
14968 Swap two rectangular objects in video.
14970 This filter accepts the following options:
14980 Set 1st rect x coordinate.
14983 Set 1st rect y coordinate.
14986 Set 2nd rect x coordinate.
14989 Set 2nd rect y coordinate.
14991 All expressions are evaluated once for each frame.
14994 The all options are expressions containing the following constants:
14999 The input width and height.
15002 same as @var{w} / @var{h}
15005 input sample aspect ratio
15008 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15011 The number of the input frame, starting from 0.
15014 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15017 the position in the file of the input frame, NAN if unknown
15025 Apply telecine process to the video.
15027 This filter accepts the following options:
15036 The default value is @code{top}.
15040 A string of numbers representing the pulldown pattern you wish to apply.
15041 The default value is @code{23}.
15045 Some typical patterns:
15050 24p: 2332 (preferred)
15057 24p: 222222222223 ("Euro pulldown")
15064 Apply threshold effect to video stream.
15066 This filter needs four video streams to perform thresholding.
15067 First stream is stream we are filtering.
15068 Second stream is holding threshold values, third stream is holding min values,
15069 and last, fourth stream is holding max values.
15071 The filter accepts the following option:
15075 Set which planes will be processed, unprocessed planes will be copied.
15076 By default value 0xf, all planes will be processed.
15079 For example if first stream pixel's component value is less then threshold value
15080 of pixel component from 2nd threshold stream, third stream value will picked,
15081 otherwise fourth stream pixel component value will be picked.
15083 Using color source filter one can perform various types of thresholding:
15085 @subsection Examples
15089 Binary threshold, using gray color as threshold:
15091 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15095 Inverted binary threshold, using gray color as threshold:
15097 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15101 Truncate binary threshold, using gray color as threshold:
15103 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15107 Threshold to zero, using gray color as threshold:
15109 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15113 Inverted threshold to zero, using gray color as threshold:
15115 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15120 Select the most representative frame in a given sequence of consecutive frames.
15122 The filter accepts the following options:
15126 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15127 will pick one of them, and then handle the next batch of @var{n} frames until
15128 the end. Default is @code{100}.
15131 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15132 value will result in a higher memory usage, so a high value is not recommended.
15134 @subsection Examples
15138 Extract one picture each 50 frames:
15144 Complete example of a thumbnail creation with @command{ffmpeg}:
15146 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15152 Tile several successive frames together.
15154 The filter accepts the following options:
15159 Set the grid size (i.e. the number of lines and columns). For the syntax of
15160 this option, check the
15161 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15164 Set the maximum number of frames to render in the given area. It must be less
15165 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15166 the area will be used.
15169 Set the outer border margin in pixels.
15172 Set the inner border thickness (i.e. the number of pixels between frames). For
15173 more advanced padding options (such as having different values for the edges),
15174 refer to the pad video filter.
15177 Specify the color of the unused area. For the syntax of this option, check the
15178 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15179 The default value of @var{color} is "black".
15182 Set the number of frames to overlap when tiling several successive frames together.
15183 The value must be between @code{0} and @var{nb_frames - 1}.
15186 Set the number of frames to initially be empty before displaying first output frame.
15187 This controls how soon will one get first output frame.
15188 The value must be between @code{0} and @var{nb_frames - 1}.
15191 @subsection Examples
15195 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15197 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15199 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15200 duplicating each output frame to accommodate the originally detected frame
15204 Display @code{5} pictures in an area of @code{3x2} frames,
15205 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15206 mixed flat and named options:
15208 tile=3x2:nb_frames=5:padding=7:margin=2
15212 @section tinterlace
15214 Perform various types of temporal field interlacing.
15216 Frames are counted starting from 1, so the first input frame is
15219 The filter accepts the following options:
15224 Specify the mode of the interlacing. This option can also be specified
15225 as a value alone. See below for a list of values for this option.
15227 Available values are:
15231 Move odd frames into the upper field, even into the lower field,
15232 generating a double height frame at half frame rate.
15236 Frame 1 Frame 2 Frame 3 Frame 4
15238 11111 22222 33333 44444
15239 11111 22222 33333 44444
15240 11111 22222 33333 44444
15241 11111 22222 33333 44444
15255 Only output odd frames, even frames are dropped, generating a frame with
15256 unchanged height at half frame rate.
15261 Frame 1 Frame 2 Frame 3 Frame 4
15263 11111 22222 33333 44444
15264 11111 22222 33333 44444
15265 11111 22222 33333 44444
15266 11111 22222 33333 44444
15276 Only output even frames, odd frames are dropped, generating a frame with
15277 unchanged height at half frame rate.
15282 Frame 1 Frame 2 Frame 3 Frame 4
15284 11111 22222 33333 44444
15285 11111 22222 33333 44444
15286 11111 22222 33333 44444
15287 11111 22222 33333 44444
15297 Expand each frame to full height, but pad alternate lines with black,
15298 generating a frame with double height at the same input frame rate.
15303 Frame 1 Frame 2 Frame 3 Frame 4
15305 11111 22222 33333 44444
15306 11111 22222 33333 44444
15307 11111 22222 33333 44444
15308 11111 22222 33333 44444
15311 11111 ..... 33333 .....
15312 ..... 22222 ..... 44444
15313 11111 ..... 33333 .....
15314 ..... 22222 ..... 44444
15315 11111 ..... 33333 .....
15316 ..... 22222 ..... 44444
15317 11111 ..... 33333 .....
15318 ..... 22222 ..... 44444
15322 @item interleave_top, 4
15323 Interleave the upper field from odd frames with the lower field from
15324 even frames, generating a frame with unchanged height at half frame rate.
15329 Frame 1 Frame 2 Frame 3 Frame 4
15331 11111<- 22222 33333<- 44444
15332 11111 22222<- 33333 44444<-
15333 11111<- 22222 33333<- 44444
15334 11111 22222<- 33333 44444<-
15344 @item interleave_bottom, 5
15345 Interleave the lower field from odd frames with the upper field from
15346 even frames, generating a frame with unchanged height at half frame rate.
15351 Frame 1 Frame 2 Frame 3 Frame 4
15353 11111 22222<- 33333 44444<-
15354 11111<- 22222 33333<- 44444
15355 11111 22222<- 33333 44444<-
15356 11111<- 22222 33333<- 44444
15366 @item interlacex2, 6
15367 Double frame rate with unchanged height. Frames are inserted each
15368 containing the second temporal field from the previous input frame and
15369 the first temporal field from the next input frame. This mode relies on
15370 the top_field_first flag. Useful for interlaced video displays with no
15371 field synchronisation.
15376 Frame 1 Frame 2 Frame 3 Frame 4
15378 11111 22222 33333 44444
15379 11111 22222 33333 44444
15380 11111 22222 33333 44444
15381 11111 22222 33333 44444
15384 11111 22222 22222 33333 33333 44444 44444
15385 11111 11111 22222 22222 33333 33333 44444
15386 11111 22222 22222 33333 33333 44444 44444
15387 11111 11111 22222 22222 33333 33333 44444
15392 Move odd frames into the upper field, even into the lower field,
15393 generating a double height frame at same frame rate.
15398 Frame 1 Frame 2 Frame 3 Frame 4
15400 11111 22222 33333 44444
15401 11111 22222 33333 44444
15402 11111 22222 33333 44444
15403 11111 22222 33333 44444
15406 11111 33333 33333 55555
15407 22222 22222 44444 44444
15408 11111 33333 33333 55555
15409 22222 22222 44444 44444
15410 11111 33333 33333 55555
15411 22222 22222 44444 44444
15412 11111 33333 33333 55555
15413 22222 22222 44444 44444
15418 Numeric values are deprecated but are accepted for backward
15419 compatibility reasons.
15421 Default mode is @code{merge}.
15424 Specify flags influencing the filter process.
15426 Available value for @var{flags} is:
15429 @item low_pass_filter, vlfp
15430 Enable linear vertical low-pass filtering in the filter.
15431 Vertical low-pass filtering is required when creating an interlaced
15432 destination from a progressive source which contains high-frequency
15433 vertical detail. Filtering will reduce interlace 'twitter' and Moire
15436 @item complex_filter, cvlfp
15437 Enable complex vertical low-pass filtering.
15438 This will slightly less reduce interlace 'twitter' and Moire
15439 patterning but better retain detail and subjective sharpness impression.
15443 Vertical low-pass filtering can only be enabled for @option{mode}
15444 @var{interleave_top} and @var{interleave_bottom}.
15449 Tone map colors from different dynamic ranges.
15451 This filter expects data in single precision floating point, as it needs to
15452 operate on (and can output) out-of-range values. Another filter, such as
15453 @ref{zscale}, is needed to convert the resulting frame to a usable format.
15455 The tonemapping algorithms implemented only work on linear light, so input
15456 data should be linearized beforehand (and possibly correctly tagged).
15459 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
15462 @subsection Options
15463 The filter accepts the following options.
15467 Set the tone map algorithm to use.
15469 Possible values are:
15472 Do not apply any tone map, only desaturate overbright pixels.
15475 Hard-clip any out-of-range values. Use it for perfect color accuracy for
15476 in-range values, while distorting out-of-range values.
15479 Stretch the entire reference gamut to a linear multiple of the display.
15482 Fit a logarithmic transfer between the tone curves.
15485 Preserve overall image brightness with a simple curve, using nonlinear
15486 contrast, which results in flattening details and degrading color accuracy.
15489 Preserve both dark and bright details better than @var{reinhard}, at the cost
15490 of slightly darkening everything. Use it when detail preservation is more
15491 important than color and brightness accuracy.
15494 Smoothly map out-of-range values, while retaining contrast and colors for
15495 in-range material as much as possible. Use it when color accuracy is more
15496 important than detail preservation.
15502 Tune the tone mapping algorithm.
15504 This affects the following algorithms:
15510 Specifies the scale factor to use while stretching.
15514 Specifies the exponent of the function.
15518 Specify an extra linear coefficient to multiply into the signal before clipping.
15522 Specify the local contrast coefficient at the display peak.
15523 Default to 0.5, which means that in-gamut values will be about half as bright
15530 Specify the transition point from linear to mobius transform. Every value
15531 below this point is guaranteed to be mapped 1:1. The higher the value, the
15532 more accurate the result will be, at the cost of losing bright details.
15533 Default to 0.3, which due to the steep initial slope still preserves in-range
15534 colors fairly accurately.
15538 Apply desaturation for highlights that exceed this level of brightness. The
15539 higher the parameter, the more color information will be preserved. This
15540 setting helps prevent unnaturally blown-out colors for super-highlights, by
15541 (smoothly) turning into white instead. This makes images feel more natural,
15542 at the cost of reducing information about out-of-range colors.
15544 The default of 2.0 is somewhat conservative and will mostly just apply to
15545 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
15547 This option works only if the input frame has a supported color tag.
15550 Override signal/nominal/reference peak with this value. Useful when the
15551 embedded peak information in display metadata is not reliable or when tone
15552 mapping from a lower range to a higher range.
15557 Transpose rows with columns in the input video and optionally flip it.
15559 It accepts the following parameters:
15564 Specify the transposition direction.
15566 Can assume the following values:
15568 @item 0, 4, cclock_flip
15569 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
15577 Rotate by 90 degrees clockwise, that is:
15585 Rotate by 90 degrees counterclockwise, that is:
15592 @item 3, 7, clock_flip
15593 Rotate by 90 degrees clockwise and vertically flip, that is:
15601 For values between 4-7, the transposition is only done if the input
15602 video geometry is portrait and not landscape. These values are
15603 deprecated, the @code{passthrough} option should be used instead.
15605 Numerical values are deprecated, and should be dropped in favor of
15606 symbolic constants.
15609 Do not apply the transposition if the input geometry matches the one
15610 specified by the specified value. It accepts the following values:
15613 Always apply transposition.
15615 Preserve portrait geometry (when @var{height} >= @var{width}).
15617 Preserve landscape geometry (when @var{width} >= @var{height}).
15620 Default value is @code{none}.
15623 For example to rotate by 90 degrees clockwise and preserve portrait
15626 transpose=dir=1:passthrough=portrait
15629 The command above can also be specified as:
15631 transpose=1:portrait
15635 Trim the input so that the output contains one continuous subpart of the input.
15637 It accepts the following parameters:
15640 Specify the time of the start of the kept section, i.e. the frame with the
15641 timestamp @var{start} will be the first frame in the output.
15644 Specify the time of the first frame that will be dropped, i.e. the frame
15645 immediately preceding the one with the timestamp @var{end} will be the last
15646 frame in the output.
15649 This is the same as @var{start}, except this option sets the start timestamp
15650 in timebase units instead of seconds.
15653 This is the same as @var{end}, except this option sets the end timestamp
15654 in timebase units instead of seconds.
15657 The maximum duration of the output in seconds.
15660 The number of the first frame that should be passed to the output.
15663 The number of the first frame that should be dropped.
15666 @option{start}, @option{end}, and @option{duration} are expressed as time
15667 duration specifications; see
15668 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15669 for the accepted syntax.
15671 Note that the first two sets of the start/end options and the @option{duration}
15672 option look at the frame timestamp, while the _frame variants simply count the
15673 frames that pass through the filter. Also note that this filter does not modify
15674 the timestamps. If you wish for the output timestamps to start at zero, insert a
15675 setpts filter after the trim filter.
15677 If multiple start or end options are set, this filter tries to be greedy and
15678 keep all the frames that match at least one of the specified constraints. To keep
15679 only the part that matches all the constraints at once, chain multiple trim
15682 The defaults are such that all the input is kept. So it is possible to set e.g.
15683 just the end values to keep everything before the specified time.
15688 Drop everything except the second minute of input:
15690 ffmpeg -i INPUT -vf trim=60:120
15694 Keep only the first second:
15696 ffmpeg -i INPUT -vf trim=duration=1
15701 @section unpremultiply
15702 Apply alpha unpremultiply effect to input video stream using first plane
15703 of second stream as alpha.
15705 Both streams must have same dimensions and same pixel format.
15707 The filter accepts the following option:
15711 Set which planes will be processed, unprocessed planes will be copied.
15712 By default value 0xf, all planes will be processed.
15714 If the format has 1 or 2 components, then luma is bit 0.
15715 If the format has 3 or 4 components:
15716 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15717 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15718 If present, the alpha channel is always the last bit.
15721 Do not require 2nd input for processing, instead use alpha plane from input stream.
15727 Sharpen or blur the input video.
15729 It accepts the following parameters:
15732 @item luma_msize_x, lx
15733 Set the luma matrix horizontal size. It must be an odd integer between
15734 3 and 23. The default value is 5.
15736 @item luma_msize_y, ly
15737 Set the luma matrix vertical size. It must be an odd integer between 3
15738 and 23. The default value is 5.
15740 @item luma_amount, la
15741 Set the luma effect strength. It must be a floating point number, reasonable
15742 values lay between -1.5 and 1.5.
15744 Negative values will blur the input video, while positive values will
15745 sharpen it, a value of zero will disable the effect.
15747 Default value is 1.0.
15749 @item chroma_msize_x, cx
15750 Set the chroma matrix horizontal size. It must be an odd integer
15751 between 3 and 23. The default value is 5.
15753 @item chroma_msize_y, cy
15754 Set the chroma matrix vertical size. It must be an odd integer
15755 between 3 and 23. The default value is 5.
15757 @item chroma_amount, ca
15758 Set the chroma effect strength. It must be a floating point number, reasonable
15759 values lay between -1.5 and 1.5.
15761 Negative values will blur the input video, while positive values will
15762 sharpen it, a value of zero will disable the effect.
15764 Default value is 0.0.
15768 All parameters are optional and default to the equivalent of the
15769 string '5:5:1.0:5:5:0.0'.
15771 @subsection Examples
15775 Apply strong luma sharpen effect:
15777 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15781 Apply a strong blur of both luma and chroma parameters:
15783 unsharp=7:7:-2:7:7:-2
15789 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15790 the image at several (or - in the case of @option{quality} level @code{8} - all)
15791 shifts and average the results.
15793 The way this differs from the behavior of spp is that uspp actually encodes &
15794 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15795 DCT similar to MJPEG.
15797 The filter accepts the following options:
15801 Set quality. This option defines the number of levels for averaging. It accepts
15802 an integer in the range 0-8. If set to @code{0}, the filter will have no
15803 effect. A value of @code{8} means the higher quality. For each increment of
15804 that value the speed drops by a factor of approximately 2. Default value is
15808 Force a constant quantization parameter. If not set, the filter will use the QP
15809 from the video stream (if available).
15812 @section vaguedenoiser
15814 Apply a wavelet based denoiser.
15816 It transforms each frame from the video input into the wavelet domain,
15817 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15818 the obtained coefficients. It does an inverse wavelet transform after.
15819 Due to wavelet properties, it should give a nice smoothed result, and
15820 reduced noise, without blurring picture features.
15822 This filter accepts the following options:
15826 The filtering strength. The higher, the more filtered the video will be.
15827 Hard thresholding can use a higher threshold than soft thresholding
15828 before the video looks overfiltered. Default value is 2.
15831 The filtering method the filter will use.
15833 It accepts the following values:
15836 All values under the threshold will be zeroed.
15839 All values under the threshold will be zeroed. All values above will be
15840 reduced by the threshold.
15843 Scales or nullifies coefficients - intermediary between (more) soft and
15844 (less) hard thresholding.
15847 Default is garrote.
15850 Number of times, the wavelet will decompose the picture. Picture can't
15851 be decomposed beyond a particular point (typically, 8 for a 640x480
15852 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15855 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15858 A list of the planes to process. By default all planes are processed.
15861 @section vectorscope
15863 Display 2 color component values in the two dimensional graph (which is called
15866 This filter accepts the following options:
15870 Set vectorscope mode.
15872 It accepts the following values:
15875 Gray values are displayed on graph, higher brightness means more pixels have
15876 same component color value on location in graph. This is the default mode.
15879 Gray values are displayed on graph. Surrounding pixels values which are not
15880 present in video frame are drawn in gradient of 2 color components which are
15881 set by option @code{x} and @code{y}. The 3rd color component is static.
15884 Actual color components values present in video frame are displayed on graph.
15887 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15888 on graph increases value of another color component, which is luminance by
15889 default values of @code{x} and @code{y}.
15892 Actual colors present in video frame are displayed on graph. If two different
15893 colors map to same position on graph then color with higher value of component
15894 not present in graph is picked.
15897 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15898 component picked from radial gradient.
15902 Set which color component will be represented on X-axis. Default is @code{1}.
15905 Set which color component will be represented on Y-axis. Default is @code{2}.
15908 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15909 of color component which represents frequency of (X, Y) location in graph.
15914 No envelope, this is default.
15917 Instant envelope, even darkest single pixel will be clearly highlighted.
15920 Hold maximum and minimum values presented in graph over time. This way you
15921 can still spot out of range values without constantly looking at vectorscope.
15924 Peak and instant envelope combined together.
15928 Set what kind of graticule to draw.
15936 Set graticule opacity.
15939 Set graticule flags.
15943 Draw graticule for white point.
15946 Draw graticule for black point.
15949 Draw color points short names.
15953 Set background opacity.
15955 @item lthreshold, l
15956 Set low threshold for color component not represented on X or Y axis.
15957 Values lower than this value will be ignored. Default is 0.
15958 Note this value is multiplied with actual max possible value one pixel component
15959 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15962 @item hthreshold, h
15963 Set high threshold for color component not represented on X or Y axis.
15964 Values higher than this value will be ignored. Default is 1.
15965 Note this value is multiplied with actual max possible value one pixel component
15966 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15967 is 0.9 * 255 = 230.
15969 @item colorspace, c
15970 Set what kind of colorspace to use when drawing graticule.
15979 @anchor{vidstabdetect}
15980 @section vidstabdetect
15982 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
15983 @ref{vidstabtransform} for pass 2.
15985 This filter generates a file with relative translation and rotation
15986 transform information about subsequent frames, which is then used by
15987 the @ref{vidstabtransform} filter.
15989 To enable compilation of this filter you need to configure FFmpeg with
15990 @code{--enable-libvidstab}.
15992 This filter accepts the following options:
15996 Set the path to the file used to write the transforms information.
15997 Default value is @file{transforms.trf}.
16000 Set how shaky the video is and how quick the camera is. It accepts an
16001 integer in the range 1-10, a value of 1 means little shakiness, a
16002 value of 10 means strong shakiness. Default value is 5.
16005 Set the accuracy of the detection process. It must be a value in the
16006 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16007 accuracy. Default value is 15.
16010 Set stepsize of the search process. The region around minimum is
16011 scanned with 1 pixel resolution. Default value is 6.
16014 Set minimum contrast. Below this value a local measurement field is
16015 discarded. Must be a floating point value in the range 0-1. Default
16019 Set reference frame number for tripod mode.
16021 If enabled, the motion of the frames is compared to a reference frame
16022 in the filtered stream, identified by the specified number. The idea
16023 is to compensate all movements in a more-or-less static scene and keep
16024 the camera view absolutely still.
16026 If set to 0, it is disabled. The frames are counted starting from 1.
16029 Show fields and transforms in the resulting frames. It accepts an
16030 integer in the range 0-2. Default value is 0, which disables any
16034 @subsection Examples
16038 Use default values:
16044 Analyze strongly shaky movie and put the results in file
16045 @file{mytransforms.trf}:
16047 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16051 Visualize the result of internal transformations in the resulting
16054 vidstabdetect=show=1
16058 Analyze a video with medium shakiness using @command{ffmpeg}:
16060 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16064 @anchor{vidstabtransform}
16065 @section vidstabtransform
16067 Video stabilization/deshaking: pass 2 of 2,
16068 see @ref{vidstabdetect} for pass 1.
16070 Read a file with transform information for each frame and
16071 apply/compensate them. Together with the @ref{vidstabdetect}
16072 filter this can be used to deshake videos. See also
16073 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16074 the @ref{unsharp} filter, see below.
16076 To enable compilation of this filter you need to configure FFmpeg with
16077 @code{--enable-libvidstab}.
16079 @subsection Options
16083 Set path to the file used to read the transforms. Default value is
16084 @file{transforms.trf}.
16087 Set the number of frames (value*2 + 1) used for lowpass filtering the
16088 camera movements. Default value is 10.
16090 For example a number of 10 means that 21 frames are used (10 in the
16091 past and 10 in the future) to smoothen the motion in the video. A
16092 larger value leads to a smoother video, but limits the acceleration of
16093 the camera (pan/tilt movements). 0 is a special case where a static
16094 camera is simulated.
16097 Set the camera path optimization algorithm.
16099 Accepted values are:
16102 gaussian kernel low-pass filter on camera motion (default)
16104 averaging on transformations
16108 Set maximal number of pixels to translate frames. Default value is -1,
16112 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16113 value is -1, meaning no limit.
16116 Specify how to deal with borders that may be visible due to movement
16119 Available values are:
16122 keep image information from previous frame (default)
16124 fill the border black
16128 Invert transforms if set to 1. Default value is 0.
16131 Consider transforms as relative to previous frame if set to 1,
16132 absolute if set to 0. Default value is 0.
16135 Set percentage to zoom. A positive value will result in a zoom-in
16136 effect, a negative value in a zoom-out effect. Default value is 0 (no
16140 Set optimal zooming to avoid borders.
16142 Accepted values are:
16147 optimal static zoom value is determined (only very strong movements
16148 will lead to visible borders) (default)
16150 optimal adaptive zoom value is determined (no borders will be
16151 visible), see @option{zoomspeed}
16154 Note that the value given at zoom is added to the one calculated here.
16157 Set percent to zoom maximally each frame (enabled when
16158 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16162 Specify type of interpolation.
16164 Available values are:
16169 linear only horizontal
16171 linear in both directions (default)
16173 cubic in both directions (slow)
16177 Enable virtual tripod mode if set to 1, which is equivalent to
16178 @code{relative=0:smoothing=0}. Default value is 0.
16180 Use also @code{tripod} option of @ref{vidstabdetect}.
16183 Increase log verbosity if set to 1. Also the detected global motions
16184 are written to the temporary file @file{global_motions.trf}. Default
16188 @subsection Examples
16192 Use @command{ffmpeg} for a typical stabilization with default values:
16194 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16197 Note the use of the @ref{unsharp} filter which is always recommended.
16200 Zoom in a bit more and load transform data from a given file:
16202 vidstabtransform=zoom=5:input="mytransforms.trf"
16206 Smoothen the video even more:
16208 vidstabtransform=smoothing=30
16214 Flip the input video vertically.
16216 For example, to vertically flip a video with @command{ffmpeg}:
16218 ffmpeg -i in.avi -vf "vflip" out.avi
16224 Make or reverse a natural vignetting effect.
16226 The filter accepts the following options:
16230 Set lens angle expression as a number of radians.
16232 The value is clipped in the @code{[0,PI/2]} range.
16234 Default value: @code{"PI/5"}
16238 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16242 Set forward/backward mode.
16244 Available modes are:
16247 The larger the distance from the central point, the darker the image becomes.
16250 The larger the distance from the central point, the brighter the image becomes.
16251 This can be used to reverse a vignette effect, though there is no automatic
16252 detection to extract the lens @option{angle} and other settings (yet). It can
16253 also be used to create a burning effect.
16256 Default value is @samp{forward}.
16259 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16261 It accepts the following values:
16264 Evaluate expressions only once during the filter initialization.
16267 Evaluate expressions for each incoming frame. This is way slower than the
16268 @samp{init} mode since it requires all the scalers to be re-computed, but it
16269 allows advanced dynamic expressions.
16272 Default value is @samp{init}.
16275 Set dithering to reduce the circular banding effects. Default is @code{1}
16279 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16280 Setting this value to the SAR of the input will make a rectangular vignetting
16281 following the dimensions of the video.
16283 Default is @code{1/1}.
16286 @subsection Expressions
16288 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16289 following parameters.
16294 input width and height
16297 the number of input frame, starting from 0
16300 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16301 @var{TB} units, NAN if undefined
16304 frame rate of the input video, NAN if the input frame rate is unknown
16307 the PTS (Presentation TimeStamp) of the filtered video frame,
16308 expressed in seconds, NAN if undefined
16311 time base of the input video
16315 @subsection Examples
16319 Apply simple strong vignetting effect:
16325 Make a flickering vignetting:
16327 vignette='PI/4+random(1)*PI/50':eval=frame
16332 @section vmafmotion
16334 Obtain the average vmaf motion score of a video.
16335 It is one of the component filters of VMAF.
16337 The obtained average motion score is printed through the logging system.
16339 In the below example the input file @file{ref.mpg} is being processed and score
16343 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16347 Stack input videos vertically.
16349 All streams must be of same pixel format and of same width.
16351 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16352 to create same output.
16354 The filter accept the following option:
16358 Set number of input streams. Default is 2.
16361 If set to 1, force the output to terminate when the shortest input
16362 terminates. Default value is 0.
16367 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
16368 Deinterlacing Filter").
16370 Based on the process described by Martin Weston for BBC R&D, and
16371 implemented based on the de-interlace algorithm written by Jim
16372 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
16373 uses filter coefficients calculated by BBC R&D.
16375 There are two sets of filter coefficients, so called "simple":
16376 and "complex". Which set of filter coefficients is used can
16377 be set by passing an optional parameter:
16381 Set the interlacing filter coefficients. Accepts one of the following values:
16385 Simple filter coefficient set.
16387 More-complex filter coefficient set.
16389 Default value is @samp{complex}.
16392 Specify which frames to deinterlace. Accept one of the following values:
16396 Deinterlace all frames,
16398 Only deinterlace frames marked as interlaced.
16401 Default value is @samp{all}.
16405 Video waveform monitor.
16407 The waveform monitor plots color component intensity. By default luminance
16408 only. Each column of the waveform corresponds to a column of pixels in the
16411 It accepts the following options:
16415 Can be either @code{row}, or @code{column}. Default is @code{column}.
16416 In row mode, the graph on the left side represents color component value 0 and
16417 the right side represents value = 255. In column mode, the top side represents
16418 color component value = 0 and bottom side represents value = 255.
16421 Set intensity. Smaller values are useful to find out how many values of the same
16422 luminance are distributed across input rows/columns.
16423 Default value is @code{0.04}. Allowed range is [0, 1].
16426 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
16427 In mirrored mode, higher values will be represented on the left
16428 side for @code{row} mode and at the top for @code{column} mode. Default is
16429 @code{1} (mirrored).
16433 It accepts the following values:
16436 Presents information identical to that in the @code{parade}, except
16437 that the graphs representing color components are superimposed directly
16440 This display mode makes it easier to spot relative differences or similarities
16441 in overlapping areas of the color components that are supposed to be identical,
16442 such as neutral whites, grays, or blacks.
16445 Display separate graph for the color components side by side in
16446 @code{row} mode or one below the other in @code{column} mode.
16449 Display separate graph for the color components side by side in
16450 @code{column} mode or one below the other in @code{row} mode.
16452 Using this display mode makes it easy to spot color casts in the highlights
16453 and shadows of an image, by comparing the contours of the top and the bottom
16454 graphs of each waveform. Since whites, grays, and blacks are characterized
16455 by exactly equal amounts of red, green, and blue, neutral areas of the picture
16456 should display three waveforms of roughly equal width/height. If not, the
16457 correction is easy to perform by making level adjustments the three waveforms.
16459 Default is @code{stack}.
16461 @item components, c
16462 Set which color components to display. Default is 1, which means only luminance
16463 or red color component if input is in RGB colorspace. If is set for example to
16464 7 it will display all 3 (if) available color components.
16469 No envelope, this is default.
16472 Instant envelope, minimum and maximum values presented in graph will be easily
16473 visible even with small @code{step} value.
16476 Hold minimum and maximum values presented in graph across time. This way you
16477 can still spot out of range values without constantly looking at waveforms.
16480 Peak and instant envelope combined together.
16486 No filtering, this is default.
16489 Luma and chroma combined together.
16492 Similar as above, but shows difference between blue and red chroma.
16495 Displays only chroma.
16498 Displays actual color value on waveform.
16501 Similar as above, but with luma showing frequency of chroma values.
16505 Set which graticule to display.
16509 Do not display graticule.
16512 Display green graticule showing legal broadcast ranges.
16516 Set graticule opacity.
16519 Set graticule flags.
16523 Draw numbers above lines. By default enabled.
16526 Draw dots instead of lines.
16530 Set scale used for displaying graticule.
16537 Default is digital.
16540 Set background opacity.
16543 @section weave, doubleweave
16545 The @code{weave} takes a field-based video input and join
16546 each two sequential fields into single frame, producing a new double
16547 height clip with half the frame rate and half the frame count.
16549 The @code{doubleweave} works same as @code{weave} but without
16550 halving frame rate and frame count.
16552 It accepts the following option:
16556 Set first field. Available values are:
16560 Set the frame as top-field-first.
16563 Set the frame as bottom-field-first.
16567 @subsection Examples
16571 Interlace video using @ref{select} and @ref{separatefields} filter:
16573 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
16578 Apply the xBR high-quality magnification filter which is designed for pixel
16579 art. It follows a set of edge-detection rules, see
16580 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
16582 It accepts the following option:
16586 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
16587 @code{3xBR} and @code{4} for @code{4xBR}.
16588 Default is @code{3}.
16594 Deinterlace the input video ("yadif" means "yet another deinterlacing
16597 It accepts the following parameters:
16603 The interlacing mode to adopt. It accepts one of the following values:
16606 @item 0, send_frame
16607 Output one frame for each frame.
16608 @item 1, send_field
16609 Output one frame for each field.
16610 @item 2, send_frame_nospatial
16611 Like @code{send_frame}, but it skips the spatial interlacing check.
16612 @item 3, send_field_nospatial
16613 Like @code{send_field}, but it skips the spatial interlacing check.
16616 The default value is @code{send_frame}.
16619 The picture field parity assumed for the input interlaced video. It accepts one
16620 of the following values:
16624 Assume the top field is first.
16626 Assume the bottom field is first.
16628 Enable automatic detection of field parity.
16631 The default value is @code{auto}.
16632 If the interlacing is unknown or the decoder does not export this information,
16633 top field first will be assumed.
16636 Specify which frames to deinterlace. Accept one of the following
16641 Deinterlace all frames.
16642 @item 1, interlaced
16643 Only deinterlace frames marked as interlaced.
16646 The default value is @code{all}.
16651 Apply Zoom & Pan effect.
16653 This filter accepts the following options:
16657 Set the zoom expression. Default is 1.
16661 Set the x and y expression. Default is 0.
16664 Set the duration expression in number of frames.
16665 This sets for how many number of frames effect will last for
16666 single input image.
16669 Set the output image size, default is 'hd720'.
16672 Set the output frame rate, default is '25'.
16675 Each expression can contain the following constants:
16694 Output frame count.
16698 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16699 for current input frame.
16703 'x' and 'y' of last output frame of previous input frame or 0 when there was
16704 not yet such frame (first input frame).
16707 Last calculated zoom from 'z' expression for current input frame.
16710 Last calculated zoom of last output frame of previous input frame.
16713 Number of output frames for current input frame. Calculated from 'd' expression
16714 for each input frame.
16717 number of output frames created for previous input frame
16720 Rational number: input width / input height
16723 sample aspect ratio
16726 display aspect ratio
16730 @subsection Examples
16734 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16736 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
16740 Zoom-in up to 1.5 and pan always at center of picture:
16742 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16746 Same as above but without pausing:
16748 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16754 Scale (resize) the input video, using the z.lib library:
16755 https://github.com/sekrit-twc/zimg.
16757 The zscale filter forces the output display aspect ratio to be the same
16758 as the input, by changing the output sample aspect ratio.
16760 If the input image format is different from the format requested by
16761 the next filter, the zscale filter will convert the input to the
16764 @subsection Options
16765 The filter accepts the following options.
16770 Set the output video dimension expression. Default value is the input
16773 If the @var{width} or @var{w} value is 0, the input width is used for
16774 the output. If the @var{height} or @var{h} value is 0, the input height
16775 is used for the output.
16777 If one and only one of the values is -n with n >= 1, the zscale filter
16778 will use a value that maintains the aspect ratio of the input image,
16779 calculated from the other specified dimension. After that it will,
16780 however, make sure that the calculated dimension is divisible by n and
16781 adjust the value if necessary.
16783 If both values are -n with n >= 1, the behavior will be identical to
16784 both values being set to 0 as previously detailed.
16786 See below for the list of accepted constants for use in the dimension
16790 Set the video size. For the syntax of this option, check the
16791 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16794 Set the dither type.
16796 Possible values are:
16801 @item error_diffusion
16807 Set the resize filter type.
16809 Possible values are:
16819 Default is bilinear.
16822 Set the color range.
16824 Possible values are:
16831 Default is same as input.
16834 Set the color primaries.
16836 Possible values are:
16846 Default is same as input.
16849 Set the transfer characteristics.
16851 Possible values are:
16865 Default is same as input.
16868 Set the colorspace matrix.
16870 Possible value are:
16881 Default is same as input.
16884 Set the input color range.
16886 Possible values are:
16893 Default is same as input.
16895 @item primariesin, pin
16896 Set the input color primaries.
16898 Possible values are:
16908 Default is same as input.
16910 @item transferin, tin
16911 Set the input transfer characteristics.
16913 Possible values are:
16924 Default is same as input.
16926 @item matrixin, min
16927 Set the input colorspace matrix.
16929 Possible value are:
16941 Set the output chroma location.
16943 Possible values are:
16954 @item chromalin, cin
16955 Set the input chroma location.
16957 Possible values are:
16969 Set the nominal peak luminance.
16972 The values of the @option{w} and @option{h} options are expressions
16973 containing the following constants:
16978 The input width and height
16982 These are the same as @var{in_w} and @var{in_h}.
16986 The output (scaled) width and height
16990 These are the same as @var{out_w} and @var{out_h}
16993 The same as @var{iw} / @var{ih}
16996 input sample aspect ratio
16999 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17003 horizontal and vertical input chroma subsample values. For example for the
17004 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17008 horizontal and vertical output chroma subsample values. For example for the
17009 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17015 @c man end VIDEO FILTERS
17017 @chapter Video Sources
17018 @c man begin VIDEO SOURCES
17020 Below is a description of the currently available video sources.
17024 Buffer video frames, and make them available to the filter chain.
17026 This source is mainly intended for a programmatic use, in particular
17027 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17029 It accepts the following parameters:
17034 Specify the size (width and height) of the buffered video frames. For the
17035 syntax of this option, check the
17036 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17039 The input video width.
17042 The input video height.
17045 A string representing the pixel format of the buffered video frames.
17046 It may be a number corresponding to a pixel format, or a pixel format
17050 Specify the timebase assumed by the timestamps of the buffered frames.
17053 Specify the frame rate expected for the video stream.
17055 @item pixel_aspect, sar
17056 The sample (pixel) aspect ratio of the input video.
17059 Specify the optional parameters to be used for the scale filter which
17060 is automatically inserted when an input change is detected in the
17061 input size or format.
17063 @item hw_frames_ctx
17064 When using a hardware pixel format, this should be a reference to an
17065 AVHWFramesContext describing input frames.
17070 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17073 will instruct the source to accept video frames with size 320x240 and
17074 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17075 square pixels (1:1 sample aspect ratio).
17076 Since the pixel format with name "yuv410p" corresponds to the number 6
17077 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17078 this example corresponds to:
17080 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17083 Alternatively, the options can be specified as a flat string, but this
17084 syntax is deprecated:
17086 @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}]
17090 Create a pattern generated by an elementary cellular automaton.
17092 The initial state of the cellular automaton can be defined through the
17093 @option{filename} and @option{pattern} options. If such options are
17094 not specified an initial state is created randomly.
17096 At each new frame a new row in the video is filled with the result of
17097 the cellular automaton next generation. The behavior when the whole
17098 frame is filled is defined by the @option{scroll} option.
17100 This source accepts the following options:
17104 Read the initial cellular automaton state, i.e. the starting row, from
17105 the specified file.
17106 In the file, each non-whitespace character is considered an alive
17107 cell, a newline will terminate the row, and further characters in the
17108 file will be ignored.
17111 Read the initial cellular automaton state, i.e. the starting row, from
17112 the specified string.
17114 Each non-whitespace character in the string is considered an alive
17115 cell, a newline will terminate the row, and further characters in the
17116 string will be ignored.
17119 Set the video rate, that is the number of frames generated per second.
17122 @item random_fill_ratio, ratio
17123 Set the random fill ratio for the initial cellular automaton row. It
17124 is a floating point number value ranging from 0 to 1, defaults to
17127 This option is ignored when a file or a pattern is specified.
17129 @item random_seed, seed
17130 Set the seed for filling randomly the initial row, must be an integer
17131 included between 0 and UINT32_MAX. If not specified, or if explicitly
17132 set to -1, the filter will try to use a good random seed on a best
17136 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17137 Default value is 110.
17140 Set the size of the output video. For the syntax of this option, check the
17141 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17143 If @option{filename} or @option{pattern} is specified, the size is set
17144 by default to the width of the specified initial state row, and the
17145 height is set to @var{width} * PHI.
17147 If @option{size} is set, it must contain the width of the specified
17148 pattern string, and the specified pattern will be centered in the
17151 If a filename or a pattern string is not specified, the size value
17152 defaults to "320x518" (used for a randomly generated initial state).
17155 If set to 1, scroll the output upward when all the rows in the output
17156 have been already filled. If set to 0, the new generated row will be
17157 written over the top row just after the bottom row is filled.
17160 @item start_full, full
17161 If set to 1, completely fill the output with generated rows before
17162 outputting the first frame.
17163 This is the default behavior, for disabling set the value to 0.
17166 If set to 1, stitch the left and right row edges together.
17167 This is the default behavior, for disabling set the value to 0.
17170 @subsection Examples
17174 Read the initial state from @file{pattern}, and specify an output of
17177 cellauto=f=pattern:s=200x400
17181 Generate a random initial row with a width of 200 cells, with a fill
17184 cellauto=ratio=2/3:s=200x200
17188 Create a pattern generated by rule 18 starting by a single alive cell
17189 centered on an initial row with width 100:
17191 cellauto=p=@@:s=100x400:full=0:rule=18
17195 Specify a more elaborated initial pattern:
17197 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17202 @anchor{coreimagesrc}
17203 @section coreimagesrc
17204 Video source generated on GPU using Apple's CoreImage API on OSX.
17206 This video source is a specialized version of the @ref{coreimage} video filter.
17207 Use a core image generator at the beginning of the applied filterchain to
17208 generate the content.
17210 The coreimagesrc video source accepts the following options:
17212 @item list_generators
17213 List all available generators along with all their respective options as well as
17214 possible minimum and maximum values along with the default values.
17216 list_generators=true
17220 Specify the size of the sourced video. For the syntax of this option, check the
17221 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17222 The default value is @code{320x240}.
17225 Specify the frame rate of the sourced video, as the number of frames
17226 generated per second. It has to be a string in the format
17227 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17228 number or a valid video frame rate abbreviation. The default value is
17232 Set the sample aspect ratio of the sourced video.
17235 Set the duration of the sourced video. See
17236 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17237 for the accepted syntax.
17239 If not specified, or the expressed duration is negative, the video is
17240 supposed to be generated forever.
17243 Additionally, all options of the @ref{coreimage} video filter are accepted.
17244 A complete filterchain can be used for further processing of the
17245 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17246 and examples for details.
17248 @subsection Examples
17253 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17254 given as complete and escaped command-line for Apple's standard bash shell:
17256 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17258 This example is equivalent to the QRCode example of @ref{coreimage} without the
17259 need for a nullsrc video source.
17263 @section mandelbrot
17265 Generate a Mandelbrot set fractal, and progressively zoom towards the
17266 point specified with @var{start_x} and @var{start_y}.
17268 This source accepts the following options:
17273 Set the terminal pts value. Default value is 400.
17276 Set the terminal scale value.
17277 Must be a floating point value. Default value is 0.3.
17280 Set the inner coloring mode, that is the algorithm used to draw the
17281 Mandelbrot fractal internal region.
17283 It shall assume one of the following values:
17288 Show time until convergence.
17290 Set color based on point closest to the origin of the iterations.
17295 Default value is @var{mincol}.
17298 Set the bailout value. Default value is 10.0.
17301 Set the maximum of iterations performed by the rendering
17302 algorithm. Default value is 7189.
17305 Set outer coloring mode.
17306 It shall assume one of following values:
17308 @item iteration_count
17309 Set iteration cound mode.
17310 @item normalized_iteration_count
17311 set normalized iteration count mode.
17313 Default value is @var{normalized_iteration_count}.
17316 Set frame rate, expressed as number of frames per second. Default
17320 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17321 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17324 Set the initial scale value. Default value is 3.0.
17327 Set the initial x position. Must be a floating point value between
17328 -100 and 100. Default value is -0.743643887037158704752191506114774.
17331 Set the initial y position. Must be a floating point value between
17332 -100 and 100. Default value is -0.131825904205311970493132056385139.
17337 Generate various test patterns, as generated by the MPlayer test filter.
17339 The size of the generated video is fixed, and is 256x256.
17340 This source is useful in particular for testing encoding features.
17342 This source accepts the following options:
17347 Specify the frame rate of the sourced video, as the number of frames
17348 generated per second. It has to be a string in the format
17349 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17350 number or a valid video frame rate abbreviation. The default value is
17354 Set the duration of the sourced video. See
17355 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17356 for the accepted syntax.
17358 If not specified, or the expressed duration is negative, the video is
17359 supposed to be generated forever.
17363 Set the number or the name of the test to perform. Supported tests are:
17379 Default value is "all", which will cycle through the list of all tests.
17384 mptestsrc=t=dc_luma
17387 will generate a "dc_luma" test pattern.
17389 @section frei0r_src
17391 Provide a frei0r source.
17393 To enable compilation of this filter you need to install the frei0r
17394 header and configure FFmpeg with @code{--enable-frei0r}.
17396 This source accepts the following parameters:
17401 The size of the video to generate. For the syntax of this option, check the
17402 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17405 The framerate of the generated video. It may be a string of the form
17406 @var{num}/@var{den} or a frame rate abbreviation.
17409 The name to the frei0r source to load. For more information regarding frei0r and
17410 how to set the parameters, read the @ref{frei0r} section in the video filters
17413 @item filter_params
17414 A '|'-separated list of parameters to pass to the frei0r source.
17418 For example, to generate a frei0r partik0l source with size 200x200
17419 and frame rate 10 which is overlaid on the overlay filter main input:
17421 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
17426 Generate a life pattern.
17428 This source is based on a generalization of John Conway's life game.
17430 The sourced input represents a life grid, each pixel represents a cell
17431 which can be in one of two possible states, alive or dead. Every cell
17432 interacts with its eight neighbours, which are the cells that are
17433 horizontally, vertically, or diagonally adjacent.
17435 At each interaction the grid evolves according to the adopted rule,
17436 which specifies the number of neighbor alive cells which will make a
17437 cell stay alive or born. The @option{rule} option allows one to specify
17440 This source accepts the following options:
17444 Set the file from which to read the initial grid state. In the file,
17445 each non-whitespace character is considered an alive cell, and newline
17446 is used to delimit the end of each row.
17448 If this option is not specified, the initial grid is generated
17452 Set the video rate, that is the number of frames generated per second.
17455 @item random_fill_ratio, ratio
17456 Set the random fill ratio for the initial random grid. It is a
17457 floating point number value ranging from 0 to 1, defaults to 1/PHI.
17458 It is ignored when a file is specified.
17460 @item random_seed, seed
17461 Set the seed for filling the initial random grid, must be an integer
17462 included between 0 and UINT32_MAX. If not specified, or if explicitly
17463 set to -1, the filter will try to use a good random seed on a best
17469 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
17470 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
17471 @var{NS} specifies the number of alive neighbor cells which make a
17472 live cell stay alive, and @var{NB} the number of alive neighbor cells
17473 which make a dead cell to become alive (i.e. to "born").
17474 "s" and "b" can be used in place of "S" and "B", respectively.
17476 Alternatively a rule can be specified by an 18-bits integer. The 9
17477 high order bits are used to encode the next cell state if it is alive
17478 for each number of neighbor alive cells, the low order bits specify
17479 the rule for "borning" new cells. Higher order bits encode for an
17480 higher number of neighbor cells.
17481 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
17482 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
17484 Default value is "S23/B3", which is the original Conway's game of life
17485 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
17486 cells, and will born a new cell if there are three alive cells around
17490 Set the size of the output video. For the syntax of this option, check the
17491 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17493 If @option{filename} is specified, the size is set by default to the
17494 same size of the input file. If @option{size} is set, it must contain
17495 the size specified in the input file, and the initial grid defined in
17496 that file is centered in the larger resulting area.
17498 If a filename is not specified, the size value defaults to "320x240"
17499 (used for a randomly generated initial grid).
17502 If set to 1, stitch the left and right grid edges together, and the
17503 top and bottom edges also. Defaults to 1.
17506 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
17507 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
17508 value from 0 to 255.
17511 Set the color of living (or new born) cells.
17514 Set the color of dead cells. If @option{mold} is set, this is the first color
17515 used to represent a dead cell.
17518 Set mold color, for definitely dead and moldy cells.
17520 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
17521 ffmpeg-utils manual,ffmpeg-utils}.
17524 @subsection Examples
17528 Read a grid from @file{pattern}, and center it on a grid of size
17531 life=f=pattern:s=300x300
17535 Generate a random grid of size 200x200, with a fill ratio of 2/3:
17537 life=ratio=2/3:s=200x200
17541 Specify a custom rule for evolving a randomly generated grid:
17547 Full example with slow death effect (mold) using @command{ffplay}:
17549 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
17556 @anchor{haldclutsrc}
17558 @anchor{rgbtestsrc}
17560 @anchor{smptehdbars}
17563 @anchor{yuvtestsrc}
17564 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
17566 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
17568 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
17570 The @code{color} source provides an uniformly colored input.
17572 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
17573 @ref{haldclut} filter.
17575 The @code{nullsrc} source returns unprocessed video frames. It is
17576 mainly useful to be employed in analysis / debugging tools, or as the
17577 source for filters which ignore the input data.
17579 The @code{rgbtestsrc} source generates an RGB test pattern useful for
17580 detecting RGB vs BGR issues. You should see a red, green and blue
17581 stripe from top to bottom.
17583 The @code{smptebars} source generates a color bars pattern, based on
17584 the SMPTE Engineering Guideline EG 1-1990.
17586 The @code{smptehdbars} source generates a color bars pattern, based on
17587 the SMPTE RP 219-2002.
17589 The @code{testsrc} source generates a test video pattern, showing a
17590 color pattern, a scrolling gradient and a timestamp. This is mainly
17591 intended for testing purposes.
17593 The @code{testsrc2} source is similar to testsrc, but supports more
17594 pixel formats instead of just @code{rgb24}. This allows using it as an
17595 input for other tests without requiring a format conversion.
17597 The @code{yuvtestsrc} source generates an YUV test pattern. You should
17598 see a y, cb and cr stripe from top to bottom.
17600 The sources accept the following parameters:
17605 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17606 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17607 pixels to be used as identity matrix for 3D lookup tables. Each component is
17608 coded on a @code{1/(N*N)} scale.
17611 Specify the color of the source, only available in the @code{color}
17612 source. For the syntax of this option, check the
17613 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17616 Specify the size of the sourced video. For the syntax of this option, check the
17617 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17618 The default value is @code{320x240}.
17620 This option is not available with the @code{allrgb}, @code{allyuv}, and
17621 @code{haldclutsrc} filters.
17624 Specify the frame rate of the sourced video, as the number of frames
17625 generated per second. It has to be a string in the format
17626 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17627 number or a valid video frame rate abbreviation. The default value is
17631 Set the duration of the sourced video. See
17632 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17633 for the accepted syntax.
17635 If not specified, or the expressed duration is negative, the video is
17636 supposed to be generated forever.
17639 Set the sample aspect ratio of the sourced video.
17642 Specify the alpha (opacity) of the background, only available in the
17643 @code{testsrc2} source. The value must be between 0 (fully transparent) and
17644 255 (fully opaque, the default).
17647 Set the number of decimals to show in the timestamp, only available in the
17648 @code{testsrc} source.
17650 The displayed timestamp value will correspond to the original
17651 timestamp value multiplied by the power of 10 of the specified
17652 value. Default value is 0.
17655 @subsection Examples
17659 Generate a video with a duration of 5.3 seconds, with size
17660 176x144 and a frame rate of 10 frames per second:
17662 testsrc=duration=5.3:size=qcif:rate=10
17666 The following graph description will generate a red source
17667 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17670 color=c=red@@0.2:s=qcif:r=10
17674 If the input content is to be ignored, @code{nullsrc} can be used. The
17675 following command generates noise in the luminance plane by employing
17676 the @code{geq} filter:
17678 nullsrc=s=256x256, geq=random(1)*255:128:128
17682 @subsection Commands
17684 The @code{color} source supports the following commands:
17688 Set the color of the created image. Accepts the same syntax of the
17689 corresponding @option{color} option.
17694 Generate video using an OpenCL program.
17699 OpenCL program source file.
17702 Kernel name in program.
17705 Size of frames to generate. This must be set.
17708 Pixel format to use for the generated frames. This must be set.
17711 Number of frames generated every second. Default value is '25'.
17715 For details of how the program loading works, see the @ref{program_opencl}
17722 Generate a colour ramp by setting pixel values from the position of the pixel
17723 in the output image. (Note that this will work with all pixel formats, but
17724 the generated output will not be the same.)
17726 __kernel void ramp(__write_only image2d_t dst,
17727 unsigned int index)
17729 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17732 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
17734 write_imagef(dst, loc, val);
17739 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
17741 __kernel void sierpinski_carpet(__write_only image2d_t dst,
17742 unsigned int index)
17744 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17746 float4 value = 0.0f;
17747 int x = loc.x + index;
17748 int y = loc.y + index;
17749 while (x > 0 || y > 0) {
17750 if (x % 3 == 1 && y % 3 == 1) {
17758 write_imagef(dst, loc, value);
17764 @c man end VIDEO SOURCES
17766 @chapter Video Sinks
17767 @c man begin VIDEO SINKS
17769 Below is a description of the currently available video sinks.
17771 @section buffersink
17773 Buffer video frames, and make them available to the end of the filter
17776 This sink is mainly intended for programmatic use, in particular
17777 through the interface defined in @file{libavfilter/buffersink.h}
17778 or the options system.
17780 It accepts a pointer to an AVBufferSinkContext structure, which
17781 defines the incoming buffers' formats, to be passed as the opaque
17782 parameter to @code{avfilter_init_filter} for initialization.
17786 Null video sink: do absolutely nothing with the input video. It is
17787 mainly useful as a template and for use in analysis / debugging
17790 @c man end VIDEO SINKS
17792 @chapter Multimedia Filters
17793 @c man begin MULTIMEDIA FILTERS
17795 Below is a description of the currently available multimedia filters.
17799 Convert input audio to a video output, displaying the audio bit scope.
17801 The filter accepts the following options:
17805 Set frame rate, expressed as number of frames per second. Default
17809 Specify the video size for the output. For the syntax of this option, check the
17810 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17811 Default value is @code{1024x256}.
17814 Specify list of colors separated by space or by '|' which will be used to
17815 draw channels. Unrecognized or missing colors will be replaced
17819 @section ahistogram
17821 Convert input audio to a video output, displaying the volume histogram.
17823 The filter accepts the following options:
17827 Specify how histogram is calculated.
17829 It accepts the following values:
17832 Use single histogram for all channels.
17834 Use separate histogram for each channel.
17836 Default is @code{single}.
17839 Set frame rate, expressed as number of frames per second. Default
17843 Specify the video size for the output. For the syntax of this option, check the
17844 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17845 Default value is @code{hd720}.
17850 It accepts the following values:
17861 reverse logarithmic
17863 Default is @code{log}.
17866 Set amplitude scale.
17868 It accepts the following values:
17875 Default is @code{log}.
17878 Set how much frames to accumulate in histogram.
17879 Defauls is 1. Setting this to -1 accumulates all frames.
17882 Set histogram ratio of window height.
17885 Set sonogram sliding.
17887 It accepts the following values:
17890 replace old rows with new ones.
17892 scroll from top to bottom.
17894 Default is @code{replace}.
17897 @section aphasemeter
17899 Convert input audio to a video output, displaying the audio phase.
17901 The filter accepts the following options:
17905 Set the output frame rate. Default value is @code{25}.
17908 Set the video size for the output. For the syntax of this option, check the
17909 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17910 Default value is @code{800x400}.
17915 Specify the red, green, blue contrast. Default values are @code{2},
17916 @code{7} and @code{1}.
17917 Allowed range is @code{[0, 255]}.
17920 Set color which will be used for drawing median phase. If color is
17921 @code{none} which is default, no median phase value will be drawn.
17924 Enable video output. Default is enabled.
17927 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17928 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17929 The @code{-1} means left and right channels are completely out of phase and
17930 @code{1} means channels are in phase.
17932 @section avectorscope
17934 Convert input audio to a video output, representing the audio vector
17937 The filter is used to measure the difference between channels of stereo
17938 audio stream. A monoaural signal, consisting of identical left and right
17939 signal, results in straight vertical line. Any stereo separation is visible
17940 as a deviation from this line, creating a Lissajous figure.
17941 If the straight (or deviation from it) but horizontal line appears this
17942 indicates that the left and right channels are out of phase.
17944 The filter accepts the following options:
17948 Set the vectorscope mode.
17950 Available values are:
17953 Lissajous rotated by 45 degrees.
17956 Same as above but not rotated.
17959 Shape resembling half of circle.
17962 Default value is @samp{lissajous}.
17965 Set the video size for the output. For the syntax of this option, check the
17966 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17967 Default value is @code{400x400}.
17970 Set the output frame rate. Default value is @code{25}.
17976 Specify the red, green, blue and alpha contrast. Default values are @code{40},
17977 @code{160}, @code{80} and @code{255}.
17978 Allowed range is @code{[0, 255]}.
17984 Specify the red, green, blue and alpha fade. Default values are @code{15},
17985 @code{10}, @code{5} and @code{5}.
17986 Allowed range is @code{[0, 255]}.
17989 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
17990 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
17993 Set the vectorscope drawing mode.
17995 Available values are:
17998 Draw dot for each sample.
18001 Draw line between previous and current sample.
18004 Default value is @samp{dot}.
18007 Specify amplitude scale of audio samples.
18009 Available values are:
18025 Swap left channel axis with right channel axis.
18035 Mirror only x axis.
18038 Mirror only y axis.
18046 @subsection Examples
18050 Complete example using @command{ffplay}:
18052 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18053 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18057 @section bench, abench
18059 Benchmark part of a filtergraph.
18061 The filter accepts the following options:
18065 Start or stop a timer.
18067 Available values are:
18070 Get the current time, set it as frame metadata (using the key
18071 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18074 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18075 the input frame metadata to get the time difference. Time difference, average,
18076 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18077 @code{min}) are then printed. The timestamps are expressed in seconds.
18081 @subsection Examples
18085 Benchmark @ref{selectivecolor} filter:
18087 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18093 Concatenate audio and video streams, joining them together one after the
18096 The filter works on segments of synchronized video and audio streams. All
18097 segments must have the same number of streams of each type, and that will
18098 also be the number of streams at output.
18100 The filter accepts the following options:
18105 Set the number of segments. Default is 2.
18108 Set the number of output video streams, that is also the number of video
18109 streams in each segment. Default is 1.
18112 Set the number of output audio streams, that is also the number of audio
18113 streams in each segment. Default is 0.
18116 Activate unsafe mode: do not fail if segments have a different format.
18120 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18121 @var{a} audio outputs.
18123 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18124 segment, in the same order as the outputs, then the inputs for the second
18127 Related streams do not always have exactly the same duration, for various
18128 reasons including codec frame size or sloppy authoring. For that reason,
18129 related synchronized streams (e.g. a video and its audio track) should be
18130 concatenated at once. The concat filter will use the duration of the longest
18131 stream in each segment (except the last one), and if necessary pad shorter
18132 audio streams with silence.
18134 For this filter to work correctly, all segments must start at timestamp 0.
18136 All corresponding streams must have the same parameters in all segments; the
18137 filtering system will automatically select a common pixel format for video
18138 streams, and a common sample format, sample rate and channel layout for
18139 audio streams, but other settings, such as resolution, must be converted
18140 explicitly by the user.
18142 Different frame rates are acceptable but will result in variable frame rate
18143 at output; be sure to configure the output file to handle it.
18145 @subsection Examples
18149 Concatenate an opening, an episode and an ending, all in bilingual version
18150 (video in stream 0, audio in streams 1 and 2):
18152 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18153 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18154 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18155 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18159 Concatenate two parts, handling audio and video separately, using the
18160 (a)movie sources, and adjusting the resolution:
18162 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18163 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18164 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18166 Note that a desync will happen at the stitch if the audio and video streams
18167 do not have exactly the same duration in the first file.
18171 @subsection Commands
18173 This filter supports the following commands:
18176 Close the current segment and step to the next one
18179 @section drawgraph, adrawgraph
18181 Draw a graph using input video or audio metadata.
18183 It accepts the following parameters:
18187 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18190 Set 1st foreground color expression.
18193 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18196 Set 2nd foreground color expression.
18199 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18202 Set 3rd foreground color expression.
18205 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18208 Set 4th foreground color expression.
18211 Set minimal value of metadata value.
18214 Set maximal value of metadata value.
18217 Set graph background color. Default is white.
18222 Available values for mode is:
18229 Default is @code{line}.
18234 Available values for slide is:
18237 Draw new frame when right border is reached.
18240 Replace old columns with new ones.
18243 Scroll from right to left.
18246 Scroll from left to right.
18249 Draw single picture.
18252 Default is @code{frame}.
18255 Set size of graph video. For the syntax of this option, check the
18256 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18257 The default value is @code{900x256}.
18259 The foreground color expressions can use the following variables:
18262 Minimal value of metadata value.
18265 Maximal value of metadata value.
18268 Current metadata key value.
18271 The color is defined as 0xAABBGGRR.
18274 Example using metadata from @ref{signalstats} filter:
18276 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18279 Example using metadata from @ref{ebur128} filter:
18281 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18287 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18288 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18289 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18290 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18292 The filter also has a video output (see the @var{video} option) with a real
18293 time graph to observe the loudness evolution. The graphic contains the logged
18294 message mentioned above, so it is not printed anymore when this option is set,
18295 unless the verbose logging is set. The main graphing area contains the
18296 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18297 the momentary loudness (400 milliseconds).
18299 More information about the Loudness Recommendation EBU R128 on
18300 @url{http://tech.ebu.ch/loudness}.
18302 The filter accepts the following options:
18307 Activate the video output. The audio stream is passed unchanged whether this
18308 option is set or no. The video stream will be the first output stream if
18309 activated. Default is @code{0}.
18312 Set the video size. This option is for video only. For the syntax of this
18314 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18315 Default and minimum resolution is @code{640x480}.
18318 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18319 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18320 other integer value between this range is allowed.
18323 Set metadata injection. If set to @code{1}, the audio input will be segmented
18324 into 100ms output frames, each of them containing various loudness information
18325 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18327 Default is @code{0}.
18330 Force the frame logging level.
18332 Available values are:
18335 information logging level
18337 verbose logging level
18340 By default, the logging level is set to @var{info}. If the @option{video} or
18341 the @option{metadata} options are set, it switches to @var{verbose}.
18346 Available modes can be cumulated (the option is a @code{flag} type). Possible
18350 Disable any peak mode (default).
18352 Enable sample-peak mode.
18354 Simple peak mode looking for the higher sample value. It logs a message
18355 for sample-peak (identified by @code{SPK}).
18357 Enable true-peak mode.
18359 If enabled, the peak lookup is done on an over-sampled version of the input
18360 stream for better peak accuracy. It logs a message for true-peak.
18361 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
18362 This mode requires a build with @code{libswresample}.
18366 Treat mono input files as "dual mono". If a mono file is intended for playback
18367 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
18368 If set to @code{true}, this option will compensate for this effect.
18369 Multi-channel input files are not affected by this option.
18372 Set a specific pan law to be used for the measurement of dual mono files.
18373 This parameter is optional, and has a default value of -3.01dB.
18376 @subsection Examples
18380 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
18382 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
18386 Run an analysis with @command{ffmpeg}:
18388 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
18392 @section interleave, ainterleave
18394 Temporally interleave frames from several inputs.
18396 @code{interleave} works with video inputs, @code{ainterleave} with audio.
18398 These filters read frames from several inputs and send the oldest
18399 queued frame to the output.
18401 Input streams must have well defined, monotonically increasing frame
18404 In order to submit one frame to output, these filters need to enqueue
18405 at least one frame for each input, so they cannot work in case one
18406 input is not yet terminated and will not receive incoming frames.
18408 For example consider the case when one input is a @code{select} filter
18409 which always drops input frames. The @code{interleave} filter will keep
18410 reading from that input, but it will never be able to send new frames
18411 to output until the input sends an end-of-stream signal.
18413 Also, depending on inputs synchronization, the filters will drop
18414 frames in case one input receives more frames than the other ones, and
18415 the queue is already filled.
18417 These filters accept the following options:
18421 Set the number of different inputs, it is 2 by default.
18424 @subsection Examples
18428 Interleave frames belonging to different streams using @command{ffmpeg}:
18430 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
18434 Add flickering blur effect:
18436 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
18440 @section metadata, ametadata
18442 Manipulate frame metadata.
18444 This filter accepts the following options:
18448 Set mode of operation of the filter.
18450 Can be one of the following:
18454 If both @code{value} and @code{key} is set, select frames
18455 which have such metadata. If only @code{key} is set, select
18456 every frame that has such key in metadata.
18459 Add new metadata @code{key} and @code{value}. If key is already available
18463 Modify value of already present key.
18466 If @code{value} is set, delete only keys that have such value.
18467 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
18471 Print key and its value if metadata was found. If @code{key} is not set print all
18472 metadata values available in frame.
18476 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
18479 Set metadata value which will be used. This option is mandatory for
18480 @code{modify} and @code{add} mode.
18483 Which function to use when comparing metadata value and @code{value}.
18485 Can be one of following:
18489 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
18492 Values are interpreted as strings, returns true if metadata value starts with
18493 the @code{value} option string.
18496 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
18499 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
18502 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
18505 Values are interpreted as floats, returns true if expression from option @code{expr}
18510 Set expression which is used when @code{function} is set to @code{expr}.
18511 The expression is evaluated through the eval API and can contain the following
18516 Float representation of @code{value} from metadata key.
18519 Float representation of @code{value} as supplied by user in @code{value} option.
18523 If specified in @code{print} mode, output is written to the named file. Instead of
18524 plain filename any writable url can be specified. Filename ``-'' is a shorthand
18525 for standard output. If @code{file} option is not set, output is written to the log
18526 with AV_LOG_INFO loglevel.
18530 @subsection Examples
18534 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
18537 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
18540 Print silencedetect output to file @file{metadata.txt}.
18542 silencedetect,ametadata=mode=print:file=metadata.txt
18545 Direct all metadata to a pipe with file descriptor 4.
18547 metadata=mode=print:file='pipe\:4'
18551 @section perms, aperms
18553 Set read/write permissions for the output frames.
18555 These filters are mainly aimed at developers to test direct path in the
18556 following filter in the filtergraph.
18558 The filters accept the following options:
18562 Select the permissions mode.
18564 It accepts the following values:
18567 Do nothing. This is the default.
18569 Set all the output frames read-only.
18571 Set all the output frames directly writable.
18573 Make the frame read-only if writable, and writable if read-only.
18575 Set each output frame read-only or writable randomly.
18579 Set the seed for the @var{random} mode, must be an integer included between
18580 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
18581 @code{-1}, the filter will try to use a good random seed on a best effort
18585 Note: in case of auto-inserted filter between the permission filter and the
18586 following one, the permission might not be received as expected in that
18587 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
18588 perms/aperms filter can avoid this problem.
18590 @section realtime, arealtime
18592 Slow down filtering to match real time approximately.
18594 These filters will pause the filtering for a variable amount of time to
18595 match the output rate with the input timestamps.
18596 They are similar to the @option{re} option to @code{ffmpeg}.
18598 They accept the following options:
18602 Time limit for the pauses. Any pause longer than that will be considered
18603 a timestamp discontinuity and reset the timer. Default is 2 seconds.
18607 @section select, aselect
18609 Select frames to pass in output.
18611 This filter accepts the following options:
18616 Set expression, which is evaluated for each input frame.
18618 If the expression is evaluated to zero, the frame is discarded.
18620 If the evaluation result is negative or NaN, the frame is sent to the
18621 first output; otherwise it is sent to the output with index
18622 @code{ceil(val)-1}, assuming that the input index starts from 0.
18624 For example a value of @code{1.2} corresponds to the output with index
18625 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
18628 Set the number of outputs. The output to which to send the selected
18629 frame is based on the result of the evaluation. Default value is 1.
18632 The expression can contain the following constants:
18636 The (sequential) number of the filtered frame, starting from 0.
18639 The (sequential) number of the selected frame, starting from 0.
18641 @item prev_selected_n
18642 The sequential number of the last selected frame. It's NAN if undefined.
18645 The timebase of the input timestamps.
18648 The PTS (Presentation TimeStamp) of the filtered video frame,
18649 expressed in @var{TB} units. It's NAN if undefined.
18652 The PTS of the filtered video frame,
18653 expressed in seconds. It's NAN if undefined.
18656 The PTS of the previously filtered video frame. It's NAN if undefined.
18658 @item prev_selected_pts
18659 The PTS of the last previously filtered video frame. It's NAN if undefined.
18661 @item prev_selected_t
18662 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
18665 The PTS of the first video frame in the video. It's NAN if undefined.
18668 The time of the first video frame in the video. It's NAN if undefined.
18670 @item pict_type @emph{(video only)}
18671 The type of the filtered frame. It can assume one of the following
18683 @item interlace_type @emph{(video only)}
18684 The frame interlace type. It can assume one of the following values:
18687 The frame is progressive (not interlaced).
18689 The frame is top-field-first.
18691 The frame is bottom-field-first.
18694 @item consumed_sample_n @emph{(audio only)}
18695 the number of selected samples before the current frame
18697 @item samples_n @emph{(audio only)}
18698 the number of samples in the current frame
18700 @item sample_rate @emph{(audio only)}
18701 the input sample rate
18704 This is 1 if the filtered frame is a key-frame, 0 otherwise.
18707 the position in the file of the filtered frame, -1 if the information
18708 is not available (e.g. for synthetic video)
18710 @item scene @emph{(video only)}
18711 value between 0 and 1 to indicate a new scene; a low value reflects a low
18712 probability for the current frame to introduce a new scene, while a higher
18713 value means the current frame is more likely to be one (see the example below)
18715 @item concatdec_select
18716 The concat demuxer can select only part of a concat input file by setting an
18717 inpoint and an outpoint, but the output packets may not be entirely contained
18718 in the selected interval. By using this variable, it is possible to skip frames
18719 generated by the concat demuxer which are not exactly contained in the selected
18722 This works by comparing the frame pts against the @var{lavf.concat.start_time}
18723 and the @var{lavf.concat.duration} packet metadata values which are also
18724 present in the decoded frames.
18726 The @var{concatdec_select} variable is -1 if the frame pts is at least
18727 start_time and either the duration metadata is missing or the frame pts is less
18728 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
18731 That basically means that an input frame is selected if its pts is within the
18732 interval set by the concat demuxer.
18736 The default value of the select expression is "1".
18738 @subsection Examples
18742 Select all frames in input:
18747 The example above is the same as:
18759 Select only I-frames:
18761 select='eq(pict_type\,I)'
18765 Select one frame every 100:
18767 select='not(mod(n\,100))'
18771 Select only frames contained in the 10-20 time interval:
18773 select=between(t\,10\,20)
18777 Select only I-frames contained in the 10-20 time interval:
18779 select=between(t\,10\,20)*eq(pict_type\,I)
18783 Select frames with a minimum distance of 10 seconds:
18785 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18789 Use aselect to select only audio frames with samples number > 100:
18791 aselect='gt(samples_n\,100)'
18795 Create a mosaic of the first scenes:
18797 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18800 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18804 Send even and odd frames to separate outputs, and compose them:
18806 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18810 Select useful frames from an ffconcat file which is using inpoints and
18811 outpoints but where the source files are not intra frame only.
18813 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18817 @section sendcmd, asendcmd
18819 Send commands to filters in the filtergraph.
18821 These filters read commands to be sent to other filters in the
18824 @code{sendcmd} must be inserted between two video filters,
18825 @code{asendcmd} must be inserted between two audio filters, but apart
18826 from that they act the same way.
18828 The specification of commands can be provided in the filter arguments
18829 with the @var{commands} option, or in a file specified by the
18830 @var{filename} option.
18832 These filters accept the following options:
18835 Set the commands to be read and sent to the other filters.
18837 Set the filename of the commands to be read and sent to the other
18841 @subsection Commands syntax
18843 A commands description consists of a sequence of interval
18844 specifications, comprising a list of commands to be executed when a
18845 particular event related to that interval occurs. The occurring event
18846 is typically the current frame time entering or leaving a given time
18849 An interval is specified by the following syntax:
18851 @var{START}[-@var{END}] @var{COMMANDS};
18854 The time interval is specified by the @var{START} and @var{END} times.
18855 @var{END} is optional and defaults to the maximum time.
18857 The current frame time is considered within the specified interval if
18858 it is included in the interval [@var{START}, @var{END}), that is when
18859 the time is greater or equal to @var{START} and is lesser than
18862 @var{COMMANDS} consists of a sequence of one or more command
18863 specifications, separated by ",", relating to that interval. The
18864 syntax of a command specification is given by:
18866 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18869 @var{FLAGS} is optional and specifies the type of events relating to
18870 the time interval which enable sending the specified command, and must
18871 be a non-null sequence of identifier flags separated by "+" or "|" and
18872 enclosed between "[" and "]".
18874 The following flags are recognized:
18877 The command is sent when the current frame timestamp enters the
18878 specified interval. In other words, the command is sent when the
18879 previous frame timestamp was not in the given interval, and the
18883 The command is sent when the current frame timestamp leaves the
18884 specified interval. In other words, the command is sent when the
18885 previous frame timestamp was in the given interval, and the
18889 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18892 @var{TARGET} specifies the target of the command, usually the name of
18893 the filter class or a specific filter instance name.
18895 @var{COMMAND} specifies the name of the command for the target filter.
18897 @var{ARG} is optional and specifies the optional list of argument for
18898 the given @var{COMMAND}.
18900 Between one interval specification and another, whitespaces, or
18901 sequences of characters starting with @code{#} until the end of line,
18902 are ignored and can be used to annotate comments.
18904 A simplified BNF description of the commands specification syntax
18907 @var{COMMAND_FLAG} ::= "enter" | "leave"
18908 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18909 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18910 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18911 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18912 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18915 @subsection Examples
18919 Specify audio tempo change at second 4:
18921 asendcmd=c='4.0 atempo tempo 1.5',atempo
18925 Target a specific filter instance:
18927 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18931 Specify a list of drawtext and hue commands in a file.
18933 # show text in the interval 5-10
18934 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18935 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18937 # desaturate the image in the interval 15-20
18938 15.0-20.0 [enter] hue s 0,
18939 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18941 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18943 # apply an exponential saturation fade-out effect, starting from time 25
18944 25 [enter] hue s exp(25-t)
18947 A filtergraph allowing to read and process the above command list
18948 stored in a file @file{test.cmd}, can be specified with:
18950 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18955 @section setpts, asetpts
18957 Change the PTS (presentation timestamp) of the input frames.
18959 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18961 This filter accepts the following options:
18966 The expression which is evaluated for each frame to construct its timestamp.
18970 The expression is evaluated through the eval API and can contain the following
18975 frame rate, only defined for constant frame-rate video
18978 The presentation timestamp in input
18981 The count of the input frame for video or the number of consumed samples,
18982 not including the current frame for audio, starting from 0.
18984 @item NB_CONSUMED_SAMPLES
18985 The number of consumed samples, not including the current frame (only
18988 @item NB_SAMPLES, S
18989 The number of samples in the current frame (only audio)
18991 @item SAMPLE_RATE, SR
18992 The audio sample rate.
18995 The PTS of the first frame.
18998 the time in seconds of the first frame
19001 State whether the current frame is interlaced.
19004 the time in seconds of the current frame
19007 original position in the file of the frame, or undefined if undefined
19008 for the current frame
19011 The previous input PTS.
19014 previous input time in seconds
19017 The previous output PTS.
19020 previous output time in seconds
19023 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19027 The wallclock (RTC) time at the start of the movie in microseconds.
19030 The timebase of the input timestamps.
19034 @subsection Examples
19038 Start counting PTS from zero
19040 setpts=PTS-STARTPTS
19044 Apply fast motion effect:
19050 Apply slow motion effect:
19056 Set fixed rate of 25 frames per second:
19062 Set fixed rate 25 fps with some jitter:
19064 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19068 Apply an offset of 10 seconds to the input PTS:
19074 Generate timestamps from a "live source" and rebase onto the current timebase:
19076 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19080 Generate timestamps by counting samples:
19089 Force color range for the output video frame.
19091 The @code{setrange} filter marks the color range property for the
19092 output frames. It does not change the input frame, but only sets the
19093 corresponding property, which affects how the frame is treated by
19096 The filter accepts the following options:
19101 Available values are:
19105 Keep the same color range property.
19107 @item unspecified, unknown
19108 Set the color range as unspecified.
19110 @item limited, tv, mpeg
19111 Set the color range as limited.
19113 @item full, pc, jpeg
19114 Set the color range as full.
19118 @section settb, asettb
19120 Set the timebase to use for the output frames timestamps.
19121 It is mainly useful for testing timebase configuration.
19123 It accepts the following parameters:
19128 The expression which is evaluated into the output timebase.
19132 The value for @option{tb} is an arithmetic expression representing a
19133 rational. The expression can contain the constants "AVTB" (the default
19134 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19135 audio only). Default value is "intb".
19137 @subsection Examples
19141 Set the timebase to 1/25:
19147 Set the timebase to 1/10:
19153 Set the timebase to 1001/1000:
19159 Set the timebase to 2*intb:
19165 Set the default timebase value:
19172 Convert input audio to a video output representing frequency spectrum
19173 logarithmically using Brown-Puckette constant Q transform algorithm with
19174 direct frequency domain coefficient calculation (but the transform itself
19175 is not really constant Q, instead the Q factor is actually variable/clamped),
19176 with musical tone scale, from E0 to D#10.
19178 The filter accepts the following options:
19182 Specify the video size for the output. It must be even. For the syntax of this option,
19183 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19184 Default value is @code{1920x1080}.
19187 Set the output frame rate. Default value is @code{25}.
19190 Set the bargraph height. It must be even. Default value is @code{-1} which
19191 computes the bargraph height automatically.
19194 Set the axis height. It must be even. Default value is @code{-1} which computes
19195 the axis height automatically.
19198 Set the sonogram height. It must be even. Default value is @code{-1} which
19199 computes the sonogram height automatically.
19202 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19203 instead. Default value is @code{1}.
19205 @item sono_v, volume
19206 Specify the sonogram volume expression. It can contain variables:
19209 the @var{bar_v} evaluated expression
19210 @item frequency, freq, f
19211 the frequency where it is evaluated
19212 @item timeclamp, tc
19213 the value of @var{timeclamp} option
19217 @item a_weighting(f)
19218 A-weighting of equal loudness
19219 @item b_weighting(f)
19220 B-weighting of equal loudness
19221 @item c_weighting(f)
19222 C-weighting of equal loudness.
19224 Default value is @code{16}.
19226 @item bar_v, volume2
19227 Specify the bargraph volume expression. It can contain variables:
19230 the @var{sono_v} evaluated expression
19231 @item frequency, freq, f
19232 the frequency where it is evaluated
19233 @item timeclamp, tc
19234 the value of @var{timeclamp} option
19238 @item a_weighting(f)
19239 A-weighting of equal loudness
19240 @item b_weighting(f)
19241 B-weighting of equal loudness
19242 @item c_weighting(f)
19243 C-weighting of equal loudness.
19245 Default value is @code{sono_v}.
19247 @item sono_g, gamma
19248 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19249 higher gamma makes the spectrum having more range. Default value is @code{3}.
19250 Acceptable range is @code{[1, 7]}.
19252 @item bar_g, gamma2
19253 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19257 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19258 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19260 @item timeclamp, tc
19261 Specify the transform timeclamp. At low frequency, there is trade-off between
19262 accuracy in time domain and frequency domain. If timeclamp is lower,
19263 event in time domain is represented more accurately (such as fast bass drum),
19264 otherwise event in frequency domain is represented more accurately
19265 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19268 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19269 limits future samples by applying asymmetric windowing in time domain, useful
19270 when low latency is required. Accepted range is @code{[0, 1]}.
19273 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19274 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19277 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19278 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19281 This option is deprecated and ignored.
19284 Specify the transform length in time domain. Use this option to control accuracy
19285 trade-off between time domain and frequency domain at every frequency sample.
19286 It can contain variables:
19288 @item frequency, freq, f
19289 the frequency where it is evaluated
19290 @item timeclamp, tc
19291 the value of @var{timeclamp} option.
19293 Default value is @code{384*tc/(384+tc*f)}.
19296 Specify the transform count for every video frame. Default value is @code{6}.
19297 Acceptable range is @code{[1, 30]}.
19300 Specify the transform count for every single pixel. Default value is @code{0},
19301 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19304 Specify font file for use with freetype to draw the axis. If not specified,
19305 use embedded font. Note that drawing with font file or embedded font is not
19306 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19310 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19311 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19314 Specify font color expression. This is arithmetic expression that should return
19315 integer value 0xRRGGBB. It can contain variables:
19317 @item frequency, freq, f
19318 the frequency where it is evaluated
19319 @item timeclamp, tc
19320 the value of @var{timeclamp} option
19325 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19326 @item r(x), g(x), b(x)
19327 red, green, and blue value of intensity x.
19329 Default value is @code{st(0, (midi(f)-59.5)/12);
19330 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19331 r(1-ld(1)) + b(ld(1))}.
19334 Specify image file to draw the axis. This option override @var{fontfile} and
19335 @var{fontcolor} option.
19338 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19339 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19340 Default value is @code{1}.
19343 Set colorspace. The accepted values are:
19346 Unspecified (default)
19355 BT.470BG or BT.601-6 625
19358 SMPTE-170M or BT.601-6 525
19364 BT.2020 with non-constant luminance
19369 Set spectrogram color scheme. This is list of floating point values with format
19370 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
19371 The default is @code{1|0.5|0|0|0.5|1}.
19375 @subsection Examples
19379 Playing audio while showing the spectrum:
19381 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
19385 Same as above, but with frame rate 30 fps:
19387 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
19391 Playing at 1280x720:
19393 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
19397 Disable sonogram display:
19403 A1 and its harmonics: A1, A2, (near)E3, A3:
19405 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),
19406 asplit[a][out1]; [a] showcqt [out0]'
19410 Same as above, but with more accuracy in frequency domain:
19412 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),
19413 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
19419 bar_v=10:sono_v=bar_v*a_weighting(f)
19423 Custom gamma, now spectrum is linear to the amplitude.
19429 Custom tlength equation:
19431 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)))'
19435 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
19437 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
19441 Custom font using fontconfig:
19443 font='Courier New,Monospace,mono|bold'
19447 Custom frequency range with custom axis using image file:
19449 axisfile=myaxis.png:basefreq=40:endfreq=10000
19455 Convert input audio to video output representing the audio power spectrum.
19456 Audio amplitude is on Y-axis while frequency is on X-axis.
19458 The filter accepts the following options:
19462 Specify size of video. For the syntax of this option, check the
19463 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19464 Default is @code{1024x512}.
19468 This set how each frequency bin will be represented.
19470 It accepts the following values:
19476 Default is @code{bar}.
19479 Set amplitude scale.
19481 It accepts the following values:
19495 Default is @code{log}.
19498 Set frequency scale.
19500 It accepts the following values:
19509 Reverse logarithmic scale.
19511 Default is @code{lin}.
19516 It accepts the following values:
19532 Default is @code{w2048}
19535 Set windowing function.
19537 It accepts the following values:
19559 Default is @code{hanning}.
19562 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19563 which means optimal overlap for selected window function will be picked.
19566 Set time averaging. Setting this to 0 will display current maximal peaks.
19567 Default is @code{1}, which means time averaging is disabled.
19570 Specify list of colors separated by space or by '|' which will be used to
19571 draw channel frequencies. Unrecognized or missing colors will be replaced
19575 Set channel display mode.
19577 It accepts the following values:
19582 Default is @code{combined}.
19585 Set minimum amplitude used in @code{log} amplitude scaler.
19589 @anchor{showspectrum}
19590 @section showspectrum
19592 Convert input audio to a video output, representing the audio frequency
19595 The filter accepts the following options:
19599 Specify the video size for the output. For the syntax of this option, check the
19600 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19601 Default value is @code{640x512}.
19604 Specify how the spectrum should slide along the window.
19606 It accepts the following values:
19609 the samples start again on the left when they reach the right
19611 the samples scroll from right to left
19613 frames are only produced when the samples reach the right
19615 the samples scroll from left to right
19618 Default value is @code{replace}.
19621 Specify display mode.
19623 It accepts the following values:
19626 all channels are displayed in the same row
19628 all channels are displayed in separate rows
19631 Default value is @samp{combined}.
19634 Specify display color mode.
19636 It accepts the following values:
19639 each channel is displayed in a separate color
19641 each channel is displayed using the same color scheme
19643 each channel is displayed using the rainbow color scheme
19645 each channel is displayed using the moreland color scheme
19647 each channel is displayed using the nebulae color scheme
19649 each channel is displayed using the fire color scheme
19651 each channel is displayed using the fiery color scheme
19653 each channel is displayed using the fruit color scheme
19655 each channel is displayed using the cool color scheme
19658 Default value is @samp{channel}.
19661 Specify scale used for calculating intensity color values.
19663 It accepts the following values:
19668 square root, default
19679 Default value is @samp{sqrt}.
19682 Set saturation modifier for displayed colors. Negative values provide
19683 alternative color scheme. @code{0} is no saturation at all.
19684 Saturation must be in [-10.0, 10.0] range.
19685 Default value is @code{1}.
19688 Set window function.
19690 It accepts the following values:
19714 Default value is @code{hann}.
19717 Set orientation of time vs frequency axis. Can be @code{vertical} or
19718 @code{horizontal}. Default is @code{vertical}.
19721 Set ratio of overlap window. Default value is @code{0}.
19722 When value is @code{1} overlap is set to recommended size for specific
19723 window function currently used.
19726 Set scale gain for calculating intensity color values.
19727 Default value is @code{1}.
19730 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
19733 Set color rotation, must be in [-1.0, 1.0] range.
19734 Default value is @code{0}.
19737 The usage is very similar to the showwaves filter; see the examples in that
19740 @subsection Examples
19744 Large window with logarithmic color scaling:
19746 showspectrum=s=1280x480:scale=log
19750 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
19752 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19753 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
19757 @section showspectrumpic
19759 Convert input audio to a single video frame, representing the audio frequency
19762 The filter accepts the following options:
19766 Specify the video size for the output. For the syntax of this option, check the
19767 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19768 Default value is @code{4096x2048}.
19771 Specify display mode.
19773 It accepts the following values:
19776 all channels are displayed in the same row
19778 all channels are displayed in separate rows
19780 Default value is @samp{combined}.
19783 Specify display color mode.
19785 It accepts the following values:
19788 each channel is displayed in a separate color
19790 each channel is displayed using the same color scheme
19792 each channel is displayed using the rainbow color scheme
19794 each channel is displayed using the moreland color scheme
19796 each channel is displayed using the nebulae color scheme
19798 each channel is displayed using the fire color scheme
19800 each channel is displayed using the fiery color scheme
19802 each channel is displayed using the fruit color scheme
19804 each channel is displayed using the cool color scheme
19806 Default value is @samp{intensity}.
19809 Specify scale used for calculating intensity color values.
19811 It accepts the following values:
19816 square root, default
19826 Default value is @samp{log}.
19829 Set saturation modifier for displayed colors. Negative values provide
19830 alternative color scheme. @code{0} is no saturation at all.
19831 Saturation must be in [-10.0, 10.0] range.
19832 Default value is @code{1}.
19835 Set window function.
19837 It accepts the following values:
19860 Default value is @code{hann}.
19863 Set orientation of time vs frequency axis. Can be @code{vertical} or
19864 @code{horizontal}. Default is @code{vertical}.
19867 Set scale gain for calculating intensity color values.
19868 Default value is @code{1}.
19871 Draw time and frequency axes and legends. Default is enabled.
19874 Set color rotation, must be in [-1.0, 1.0] range.
19875 Default value is @code{0}.
19878 @subsection Examples
19882 Extract an audio spectrogram of a whole audio track
19883 in a 1024x1024 picture using @command{ffmpeg}:
19885 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19889 @section showvolume
19891 Convert input audio volume to a video output.
19893 The filter accepts the following options:
19900 Set border width, allowed range is [0, 5]. Default is 1.
19903 Set channel width, allowed range is [80, 8192]. Default is 400.
19906 Set channel height, allowed range is [1, 900]. Default is 20.
19909 Set fade, allowed range is [0.001, 1]. Default is 0.95.
19912 Set volume color expression.
19914 The expression can use the following variables:
19918 Current max volume of channel in dB.
19924 Current channel number, starting from 0.
19928 If set, displays channel names. Default is enabled.
19931 If set, displays volume values. Default is enabled.
19934 Set orientation, can be @code{horizontal} or @code{vertical},
19935 default is @code{horizontal}.
19938 Set step size, allowed range s [0, 5]. Default is 0, which means
19944 Convert input audio to a video output, representing the samples waves.
19946 The filter accepts the following options:
19950 Specify the video size for the output. For the syntax of this option, check the
19951 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19952 Default value is @code{600x240}.
19957 Available values are:
19960 Draw a point for each sample.
19963 Draw a vertical line for each sample.
19966 Draw a point for each sample and a line between them.
19969 Draw a centered vertical line for each sample.
19972 Default value is @code{point}.
19975 Set the number of samples which are printed on the same column. A
19976 larger value will decrease the frame rate. Must be a positive
19977 integer. This option can be set only if the value for @var{rate}
19978 is not explicitly specified.
19981 Set the (approximate) output frame rate. This is done by setting the
19982 option @var{n}. Default value is "25".
19984 @item split_channels
19985 Set if channels should be drawn separately or overlap. Default value is 0.
19988 Set colors separated by '|' which are going to be used for drawing of each channel.
19991 Set amplitude scale.
19993 Available values are:
20011 @subsection Examples
20015 Output the input file audio and the corresponding video representation
20018 amovie=a.mp3,asplit[out0],showwaves[out1]
20022 Create a synthetic signal and show it with showwaves, forcing a
20023 frame rate of 30 frames per second:
20025 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20029 @section showwavespic
20031 Convert input audio to a single video frame, representing the samples waves.
20033 The filter accepts the following options:
20037 Specify the video size for the output. For the syntax of this option, check the
20038 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20039 Default value is @code{600x240}.
20041 @item split_channels
20042 Set if channels should be drawn separately or overlap. Default value is 0.
20045 Set colors separated by '|' which are going to be used for drawing of each channel.
20048 Set amplitude scale.
20050 Available values are:
20068 @subsection Examples
20072 Extract a channel split representation of the wave form of a whole audio track
20073 in a 1024x800 picture using @command{ffmpeg}:
20075 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20079 @section sidedata, asidedata
20081 Delete frame side data, or select frames based on it.
20083 This filter accepts the following options:
20087 Set mode of operation of the filter.
20089 Can be one of the following:
20093 Select every frame with side data of @code{type}.
20096 Delete side data of @code{type}. If @code{type} is not set, delete all side
20102 Set side data type used with all modes. Must be set for @code{select} mode. For
20103 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20104 in @file{libavutil/frame.h}. For example, to choose
20105 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20109 @section spectrumsynth
20111 Sythesize audio from 2 input video spectrums, first input stream represents
20112 magnitude across time and second represents phase across time.
20113 The filter will transform from frequency domain as displayed in videos back
20114 to time domain as presented in audio output.
20116 This filter is primarily created for reversing processed @ref{showspectrum}
20117 filter outputs, but can synthesize sound from other spectrograms too.
20118 But in such case results are going to be poor if the phase data is not
20119 available, because in such cases phase data need to be recreated, usually
20120 its just recreated from random noise.
20121 For best results use gray only output (@code{channel} color mode in
20122 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20123 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20124 @code{data} option. Inputs videos should generally use @code{fullframe}
20125 slide mode as that saves resources needed for decoding video.
20127 The filter accepts the following options:
20131 Specify sample rate of output audio, the sample rate of audio from which
20132 spectrum was generated may differ.
20135 Set number of channels represented in input video spectrums.
20138 Set scale which was used when generating magnitude input spectrum.
20139 Can be @code{lin} or @code{log}. Default is @code{log}.
20142 Set slide which was used when generating inputs spectrums.
20143 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20144 Default is @code{fullframe}.
20147 Set window function used for resynthesis.
20150 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20151 which means optimal overlap for selected window function will be picked.
20154 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20155 Default is @code{vertical}.
20158 @subsection Examples
20162 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20163 then resynthesize videos back to audio with spectrumsynth:
20165 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
20166 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
20167 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20171 @section split, asplit
20173 Split input into several identical outputs.
20175 @code{asplit} works with audio input, @code{split} with video.
20177 The filter accepts a single parameter which specifies the number of outputs. If
20178 unspecified, it defaults to 2.
20180 @subsection Examples
20184 Create two separate outputs from the same input:
20186 [in] split [out0][out1]
20190 To create 3 or more outputs, you need to specify the number of
20193 [in] asplit=3 [out0][out1][out2]
20197 Create two separate outputs from the same input, one cropped and
20200 [in] split [splitout1][splitout2];
20201 [splitout1] crop=100:100:0:0 [cropout];
20202 [splitout2] pad=200:200:100:100 [padout];
20206 Create 5 copies of the input audio with @command{ffmpeg}:
20208 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20214 Receive commands sent through a libzmq client, and forward them to
20215 filters in the filtergraph.
20217 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20218 must be inserted between two video filters, @code{azmq} between two
20221 To enable these filters you need to install the libzmq library and
20222 headers and configure FFmpeg with @code{--enable-libzmq}.
20224 For more information about libzmq see:
20225 @url{http://www.zeromq.org/}
20227 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20228 receives messages sent through a network interface defined by the
20229 @option{bind_address} option.
20231 The received message must be in the form:
20233 @var{TARGET} @var{COMMAND} [@var{ARG}]
20236 @var{TARGET} specifies the target of the command, usually the name of
20237 the filter class or a specific filter instance name.
20239 @var{COMMAND} specifies the name of the command for the target filter.
20241 @var{ARG} is optional and specifies the optional argument list for the
20242 given @var{COMMAND}.
20244 Upon reception, the message is processed and the corresponding command
20245 is injected into the filtergraph. Depending on the result, the filter
20246 will send a reply to the client, adopting the format:
20248 @var{ERROR_CODE} @var{ERROR_REASON}
20252 @var{MESSAGE} is optional.
20254 @subsection Examples
20256 Look at @file{tools/zmqsend} for an example of a zmq client which can
20257 be used to send commands processed by these filters.
20259 Consider the following filtergraph generated by @command{ffplay}
20261 ffplay -dumpgraph 1 -f lavfi "
20262 color=s=100x100:c=red [l];
20263 color=s=100x100:c=blue [r];
20264 nullsrc=s=200x100, zmq [bg];
20265 [bg][l] overlay [bg+l];
20266 [bg+l][r] overlay=x=100 "
20269 To change the color of the left side of the video, the following
20270 command can be used:
20272 echo Parsed_color_0 c yellow | tools/zmqsend
20275 To change the right side:
20277 echo Parsed_color_1 c pink | tools/zmqsend
20280 @c man end MULTIMEDIA FILTERS
20282 @chapter Multimedia Sources
20283 @c man begin MULTIMEDIA SOURCES
20285 Below is a description of the currently available multimedia sources.
20289 This is the same as @ref{movie} source, except it selects an audio
20295 Read audio and/or video stream(s) from a movie container.
20297 It accepts the following parameters:
20301 The name of the resource to read (not necessarily a file; it can also be a
20302 device or a stream accessed through some protocol).
20304 @item format_name, f
20305 Specifies the format assumed for the movie to read, and can be either
20306 the name of a container or an input device. If not specified, the
20307 format is guessed from @var{movie_name} or by probing.
20309 @item seek_point, sp
20310 Specifies the seek point in seconds. The frames will be output
20311 starting from this seek point. The parameter is evaluated with
20312 @code{av_strtod}, so the numerical value may be suffixed by an IS
20313 postfix. The default value is "0".
20316 Specifies the streams to read. Several streams can be specified,
20317 separated by "+". The source will then have as many outputs, in the
20318 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
20319 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
20320 respectively the default (best suited) video and audio stream. Default
20321 is "dv", or "da" if the filter is called as "amovie".
20323 @item stream_index, si
20324 Specifies the index of the video stream to read. If the value is -1,
20325 the most suitable video stream will be automatically selected. The default
20326 value is "-1". Deprecated. If the filter is called "amovie", it will select
20327 audio instead of video.
20330 Specifies how many times to read the stream in sequence.
20331 If the value is 0, the stream will be looped infinitely.
20332 Default value is "1".
20334 Note that when the movie is looped the source timestamps are not
20335 changed, so it will generate non monotonically increasing timestamps.
20337 @item discontinuity
20338 Specifies the time difference between frames above which the point is
20339 considered a timestamp discontinuity which is removed by adjusting the later
20343 It allows overlaying a second video on top of the main input of
20344 a filtergraph, as shown in this graph:
20346 input -----------> deltapts0 --> overlay --> output
20349 movie --> scale--> deltapts1 -------+
20351 @subsection Examples
20355 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
20356 on top of the input labelled "in":
20358 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
20359 [in] setpts=PTS-STARTPTS [main];
20360 [main][over] overlay=16:16 [out]
20364 Read from a video4linux2 device, and overlay it on top of the input
20367 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
20368 [in] setpts=PTS-STARTPTS [main];
20369 [main][over] overlay=16:16 [out]
20373 Read the first video stream and the audio stream with id 0x81 from
20374 dvd.vob; the video is connected to the pad named "video" and the audio is
20375 connected to the pad named "audio":
20377 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
20381 @subsection Commands
20383 Both movie and amovie support the following commands:
20386 Perform seek using "av_seek_frame".
20387 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
20390 @var{stream_index}: If stream_index is -1, a default
20391 stream is selected, and @var{timestamp} is automatically converted
20392 from AV_TIME_BASE units to the stream specific time_base.
20394 @var{timestamp}: Timestamp in AVStream.time_base units
20395 or, if no stream is specified, in AV_TIME_BASE units.
20397 @var{flags}: Flags which select direction and seeking mode.
20401 Get movie duration in AV_TIME_BASE units.
20405 @c man end MULTIMEDIA SOURCES