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
12 input --> split ---------------------> overlay --> output
15 +-----> crop --> vflip -------+
18 This filtergraph splits the input stream in two streams, sends one
19 stream through the crop filter and the vflip filter before merging it
20 back with the other stream by overlaying it on top. You can use the
21 following command to achieve this:
24 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
27 The result will be that in output the top half of the video is mirrored
30 Filters in the same linear chain are separated by commas, and distinct
31 linear chains of filters are separated by semicolons. In our example,
32 @var{crop,vflip} are in one linear chain, @var{split} and
33 @var{overlay} are separately in another. The points where the linear
34 chains join are labelled by names enclosed in square brackets. In the
35 example, the split filter generates two outputs that are associated to
36 the labels @var{[main]} and @var{[tmp]}.
38 The stream sent to the second output of @var{split}, labelled as
39 @var{[tmp]}, is processed through the @var{crop} filter, which crops
40 away the lower half part of the video, and then vertically flipped. The
41 @var{overlay} filter takes in input the first unchanged output of the
42 split filter (which was labelled as @var{[main]}), and overlay on its
43 lower half the output generated by the @var{crop,vflip} filterchain.
45 Some filters take in input a list of parameters: they are specified
46 after the filter name and an equal sign, and are separated from each other
49 There exist so-called @var{source filters} that do not have an
50 audio/video input, and @var{sink filters} that will not have audio/video
53 @c man end FILTERING INTRODUCTION
56 @c man begin GRAPH2DOT
58 The @file{graph2dot} program included in the FFmpeg @file{tools}
59 directory can be used to parse a filtergraph description and issue a
60 corresponding textual representation in the dot language.
67 to see how to use @file{graph2dot}.
69 You can then pass the dot description to the @file{dot} program (from
70 the graphviz suite of programs) and obtain a graphical representation
73 For example the sequence of commands:
75 echo @var{GRAPH_DESCRIPTION} | \
76 tools/graph2dot -o graph.tmp && \
77 dot -Tpng graph.tmp -o graph.png && \
81 can be used to create and display an image representing the graph
82 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
83 a complete self-contained graph, with its inputs and outputs explicitly defined.
84 For example if your command line is of the form:
86 ffmpeg -i infile -vf scale=640:360 outfile
88 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
90 nullsrc,scale=640:360,nullsink
92 you may also need to set the @var{nullsrc} parameters and add a @var{format}
93 filter in order to simulate a specific input file.
97 @chapter Filtergraph description
98 @c man begin FILTERGRAPH DESCRIPTION
100 A filtergraph is a directed graph of connected filters. It can contain
101 cycles, and there can be multiple links between a pair of
102 filters. Each link has one input pad on one side connecting it to one
103 filter from which it takes its input, and one output pad on the other
104 side connecting it to the one filter accepting its output.
106 Each filter in a filtergraph is an instance of a filter class
107 registered in the application, which defines the features and the
108 number of input and output pads of the filter.
110 A filter with no input pads is called a "source", a filter with no
111 output pads is called a "sink".
113 @anchor{Filtergraph syntax}
114 @section Filtergraph syntax
116 A filtergraph can be represented using a textual representation, which is
117 recognized by the @option{-filter}/@option{-vf} and @option{-filter_complex}
118 options in @command{ffmpeg} and @option{-vf} in @command{ffplay}, and by the
119 @code{avfilter_graph_parse()}/@code{avfilter_graph_parse2()} function defined in
120 @file{libavfilter/avfilter.h}.
122 A filterchain consists of a sequence of connected filters, each one
123 connected to the previous one in the sequence. A filterchain is
124 represented by a list of ","-separated filter descriptions.
126 A filtergraph consists of a sequence of filterchains. A sequence of
127 filterchains is represented by a list of ";"-separated filterchain
130 A filter is represented by a string of the form:
131 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
133 @var{filter_name} is the name of the filter class of which the
134 described filter is an instance of, and has to be the name of one of
135 the filter classes registered in the program.
136 The name of the filter class is optionally followed by a string
139 @var{arguments} is a string which contains the parameters used to
140 initialize the filter instance. It may have one of the following forms:
144 A ':'-separated list of @var{key=value} pairs.
147 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
148 the option names in the order they are declared. E.g. the @code{fade} filter
149 declares three options in this order -- @option{type}, @option{start_frame} and
150 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
151 @var{in} is assigned to the option @option{type}, @var{0} to
152 @option{start_frame} and @var{30} to @option{nb_frames}.
155 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
156 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
157 follow the same constraints order of the previous point. The following
158 @var{key=value} pairs can be set in any preferred order.
162 If the option value itself is a list of items (e.g. the @code{format} filter
163 takes a list of pixel formats), the items in the list are usually separated by
166 The list of arguments can be quoted using the character "'" as initial
167 and ending mark, and the character '\' for escaping the characters
168 within the quoted text; otherwise the argument string is considered
169 terminated when the next special character (belonging to the set
170 "[]=;,") is encountered.
172 The name and arguments of the filter are optionally preceded and
173 followed by a list of link labels.
174 A link label allows to name a link and associate it to a filter output
175 or input pad. The preceding labels @var{in_link_1}
176 ... @var{in_link_N}, are associated to the filter input pads,
177 the following labels @var{out_link_1} ... @var{out_link_M}, are
178 associated to the output pads.
180 When two link labels with the same name are found in the
181 filtergraph, a link between the corresponding input and output pad is
184 If an output pad is not labelled, it is linked by default to the first
185 unlabelled input pad of the next filter in the filterchain.
186 For example in the filterchain:
188 nullsrc, split[L1], [L2]overlay, nullsink
190 the split filter instance has two output pads, and the overlay filter
191 instance two input pads. The first output pad of split is labelled
192 "L1", the first input pad of overlay is labelled "L2", and the second
193 output pad of split is linked to the second input pad of overlay,
194 which are both unlabelled.
196 In a complete filterchain all the unlabelled filter input and output
197 pads must be connected. A filtergraph is considered valid if all the
198 filter input and output pads of all the filterchains are connected.
200 Libavfilter will automatically insert scale filters where format
201 conversion is required. It is possible to specify swscale flags
202 for those automatically inserted scalers by prepending
203 @code{sws_flags=@var{flags};}
204 to the filtergraph description.
206 Follows a BNF description for the filtergraph syntax:
208 @var{NAME} ::= sequence of alphanumeric characters and '_'
209 @var{LINKLABEL} ::= "[" @var{NAME} "]"
210 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
211 @var{FILTER_ARGUMENTS} ::= sequence of chars (eventually quoted)
212 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
213 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
214 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
217 @section Notes on filtergraph escaping
219 Some filter arguments require the use of special characters, typically
220 @code{:} to separate key=value pairs in a named options list. In this
221 case the user should perform a first level escaping when specifying
222 the filter arguments. For example, consider the following literal
223 string to be embedded in the @ref{drawtext} filter arguments:
225 this is a 'string': may contain one, or more, special characters
228 Since @code{:} is special for the filter arguments syntax, it needs to
229 be escaped, so you get:
231 text=this is a \'string\'\: may contain one, or more, special characters
234 A second level of escaping is required when embedding the filter
235 arguments in a filtergraph description, in order to escape all the
236 filtergraph special characters. Thus the example above becomes:
238 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
241 Finally an additional level of escaping may be needed when writing the
242 filtergraph description in a shell command, which depends on the
243 escaping rules of the adopted shell. For example, assuming that
244 @code{\} is special and needs to be escaped with another @code{\}, the
245 previous string will finally result in:
247 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
250 Sometimes, it might be more convenient to employ quoting in place of
251 escaping. For example the string:
253 Caesar: tu quoque, Brute, fili mi
256 Can be quoted in the filter arguments as:
258 text='Caesar: tu quoque, Brute, fili mi'
261 And finally inserted in a filtergraph like:
263 drawtext=text=\'Caesar: tu quoque\, Brute\, fili mi\'
266 See the ``Quoting and escaping'' section in the ffmpeg-utils manual
267 for more information about the escaping and quoting rules adopted by
270 @chapter Timeline editing
272 Some filters support a generic @option{enable} option. For the filters
273 supporting timeline editing, this option can be set to an expression which is
274 evaluated before sending a frame to the filter. If the evaluation is non-zero,
275 the filter will be enabled, otherwise the frame will be sent unchanged to the
276 next filter in the filtergraph.
278 The expression accepts the following values:
281 timestamp expressed in seconds, NAN if the input timestamp is unknown
284 sequential number of the input frame, starting from 0
287 the position in the file of the input frame, NAN if unknown
290 Additionally, these filters support an @option{enable} command that can be used
291 to re-define the expression.
293 Like any other filtering option, the @option{enable} option follows the same
296 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
297 minutes, and a @ref{curves} filter starting at 3 seconds:
299 smartblur = enable='between(t,10,3*60)',
300 curves = enable='gte(t,3)' : preset=cross_process
303 @c man end FILTERGRAPH DESCRIPTION
305 @chapter Audio Filters
306 @c man begin AUDIO FILTERS
308 When you configure your FFmpeg build, you can disable any of the
309 existing filters using @code{--disable-filters}.
310 The configure output will show the audio filters included in your
313 Below is a description of the currently available audio filters.
317 Convert the input audio format to the specified formats.
319 @emph{This filter is deprecated. Use @ref{aformat} instead.}
321 The filter accepts a string of the form:
322 "@var{sample_format}:@var{channel_layout}".
324 @var{sample_format} specifies the sample format, and can be a string or the
325 corresponding numeric value defined in @file{libavutil/samplefmt.h}. Use 'p'
326 suffix for a planar sample format.
328 @var{channel_layout} specifies the channel layout, and can be a string
329 or the corresponding number value defined in @file{libavutil/channel_layout.h}.
331 The special parameter "auto", signifies that the filter will
332 automatically select the output format depending on the output filter.
338 Convert input to float, planar, stereo:
344 Convert input to unsigned 8-bit, automatically select out channel layout:
352 Apply echoing to the input audio.
354 Echoes are reflected sound and can occur naturally amongst mountains
355 (and sometimes large buildings) when talking or shouting; digital echo
356 effects emulate this behaviour and are often used to help fill out the
357 sound of a single instrument or vocal. The time difference between the
358 original signal and the reflection is the @code{delay}, and the
359 loudness of the reflected signal is the @code{decay}.
360 Multiple echoes can have different delays and decays.
362 A description of the accepted parameters follows.
366 Set input gain of reflected signal. Default is @code{0.6}.
369 Set output gain of reflected signal. Default is @code{0.3}.
372 Set list of time intervals in milliseconds between original signal and reflections
373 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
374 Default is @code{1000}.
377 Set list of loudnesses of reflected signals separated by '|'.
378 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
379 Default is @code{0.5}.
386 Make it sound as if there are twice as many instruments as are actually playing:
388 aecho=0.8:0.88:60:0.4
392 If delay is very short, then it sound like a (metallic) robot playing music:
398 A longer delay will sound like an open air concert in the mountains:
400 aecho=0.8:0.9:1000:0.3
404 Same as above but with one more mountain:
406 aecho=0.8:0.9:1000|1800:0.3|0.25
412 Apply fade-in/out effect to input audio.
414 A description of the accepted parameters follows.
418 Specify the effect type, can be either @code{in} for fade-in, or
419 @code{out} for a fade-out effect. Default is @code{in}.
421 @item start_sample, ss
422 Specify the number of the start sample for starting to apply the fade
423 effect. Default is 0.
426 Specify the number of samples for which the fade effect has to last. At
427 the end of the fade-in effect the output audio will have the same
428 volume as the input audio, at the end of the fade-out transition
429 the output audio will be silence. Default is 44100.
432 Specify time for starting to apply the fade effect. Default is 0.
433 The accepted syntax is:
435 [-]HH[:MM[:SS[.m...]]]
438 See also the function @code{av_parse_time()}.
439 If set this option is used instead of @var{start_sample} one.
442 Specify the duration for which the fade effect has to last. Default is 0.
443 The accepted syntax is:
445 [-]HH[:MM[:SS[.m...]]]
448 See also the function @code{av_parse_time()}.
449 At the end of the fade-in effect the output audio will have the same
450 volume as the input audio, at the end of the fade-out transition
451 the output audio will be silence.
452 If set this option is used instead of @var{nb_samples} one.
455 Set curve for fade transition.
457 It accepts the following values:
460 select triangular, linear slope (default)
462 select quarter of sine wave
464 select half of sine wave
466 select exponential sine wave
470 select inverted parabola
486 Fade in first 15 seconds of audio:
492 Fade out last 25 seconds of a 900 seconds audio:
494 afade=t=out:st=875:d=25
501 Set output format constraints for the input audio. The framework will
502 negotiate the most appropriate format to minimize conversions.
504 The filter accepts the following named parameters:
508 A '|'-separated list of requested sample formats.
511 A '|'-separated list of requested sample rates.
513 @item channel_layouts
514 A '|'-separated list of requested channel layouts.
518 If a parameter is omitted, all values are allowed.
520 For example to force the output to either unsigned 8-bit or signed 16-bit stereo:
522 aformat=sample_fmts=u8|s16:channel_layouts=stereo
527 Apply a two-pole all-pass filter with central frequency (in Hz)
528 @var{frequency}, and filter-width @var{width}.
529 An all-pass filter changes the audio's frequency to phase relationship
530 without changing its frequency to amplitude relationship.
532 The filter accepts the following options:
539 Set method to specify band-width of filter.
552 Specify the band-width of a filter in width_type units.
557 Merge two or more audio streams into a single multi-channel stream.
559 The filter accepts the following options:
564 Set the number of inputs. Default is 2.
568 If the channel layouts of the inputs are disjoint, and therefore compatible,
569 the channel layout of the output will be set accordingly and the channels
570 will be reordered as necessary. If the channel layouts of the inputs are not
571 disjoint, the output will have all the channels of the first input then all
572 the channels of the second input, in that order, and the channel layout of
573 the output will be the default value corresponding to the total number of
576 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
577 is FC+BL+BR, then the output will be in 5.1, with the channels in the
578 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
579 first input, b1 is the first channel of the second input).
581 On the other hand, if both input are in stereo, the output channels will be
582 in the default order: a1, a2, b1, b2, and the channel layout will be
583 arbitrarily set to 4.0, which may or may not be the expected value.
585 All inputs must have the same sample rate, and format.
587 If inputs do not have the same duration, the output will stop with the
594 Merge two mono files into a stereo stream:
596 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
600 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
602 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
608 Mixes multiple audio inputs into a single output.
612 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
614 will mix 3 input audio streams to a single output with the same duration as the
615 first input and a dropout transition time of 3 seconds.
617 The filter accepts the following named parameters:
621 Number of inputs. If unspecified, it defaults to 2.
624 How to determine the end-of-stream.
628 Duration of longest input. (default)
631 Duration of shortest input.
634 Duration of first input.
638 @item dropout_transition
639 Transition time, in seconds, for volume renormalization when an input
640 stream ends. The default value is 2 seconds.
646 Pass the audio source unchanged to the output.
650 Pad the end of a audio stream with silence, this can be used together with
651 -shortest to extend audio streams to the same length as the video stream.
654 Add a phasing effect to the input audio.
656 A phaser filter creates series of peaks and troughs in the frequency spectrum.
657 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
659 A description of the accepted parameters follows.
663 Set input gain. Default is 0.4.
666 Set output gain. Default is 0.74
669 Set delay in milliseconds. Default is 3.0.
672 Set decay. Default is 0.4.
675 Set modulation speed in Hz. Default is 0.5.
678 Set modulation type. Default is triangular.
680 It accepts the following values:
690 Resample the input audio to the specified parameters, using the
691 libswresample library. If none are specified then the filter will
692 automatically convert between its input and output.
694 This filter is also able to stretch/squeeze the audio data to make it match
695 the timestamps or to inject silence / cut out audio to make it match the
696 timestamps, do a combination of both or do neither.
698 The filter accepts the syntax
699 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
700 expresses a sample rate and @var{resampler_options} is a list of
701 @var{key}=@var{value} pairs, separated by ":". See the
702 ffmpeg-resampler manual for the complete list of supported options.
708 Resample the input audio to 44100Hz:
714 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
715 samples per second compensation:
721 @section asetnsamples
723 Set the number of samples per each output audio frame.
725 The last output packet may contain a different number of samples, as
726 the filter will flush all the remaining samples when the input audio
729 The filter accepts the following options:
733 @item nb_out_samples, n
734 Set the number of frames per each output audio frame. The number is
735 intended as the number of samples @emph{per each channel}.
736 Default value is 1024.
739 If set to 1, the filter will pad the last audio frame with zeroes, so
740 that the last frame will contain the same number of samples as the
741 previous ones. Default value is 1.
744 For example, to set the number of per-frame samples to 1234 and
745 disable padding for the last frame, use:
747 asetnsamples=n=1234:p=0
752 Set the sample rate without altering the PCM data.
753 This will result in a change of speed and pitch.
755 The filter accepts the following options:
759 Set the output sample rate. Default is 44100 Hz.
764 Show a line containing various information for each input audio frame.
765 The input audio is not modified.
767 The shown line contains a sequence of key/value pairs of the form
768 @var{key}:@var{value}.
770 A description of each shown parameter follows:
774 sequential number of the input frame, starting from 0
777 Presentation timestamp of the input frame, in time base units; the time base
778 depends on the filter input pad, and is usually 1/@var{sample_rate}.
781 presentation timestamp of the input frame in seconds
784 position of the frame in the input stream, -1 if this information in
785 unavailable and/or meaningless (for example in case of synthetic audio)
794 sample rate for the audio frame
797 number of samples (per channel) in the frame
800 Adler-32 checksum (printed in hexadecimal) of the audio data. For planar audio
801 the data is treated as if all the planes were concatenated.
803 @item plane_checksums
804 A list of Adler-32 checksums for each data plane.
809 Display time domain statistical information about the audio channels.
810 Statistics are calculated and displayed for each audio channel and,
811 where applicable, an overall figure is also given.
813 The filter accepts the following option:
816 Short window length in seconds, used for peak and trough RMS measurement.
817 Default is @code{0.05} (50 miliseconds). Allowed range is @code{[0.1 - 10]}.
820 A description of each shown parameter follows:
824 Mean amplitude displacement from zero.
827 Minimal sample level.
830 Maximal sample level.
834 Standard peak and RMS level measured in dBFS.
838 Peak and trough values for RMS level measured over a short window.
841 Standard ratio of peak to RMS level (note: not in dB).
844 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
845 (i.e. either @var{Min level} or @var{Max level}).
848 Number of occasions (not the number of samples) that the signal attained either
849 @var{Min level} or @var{Max level}.
854 Forward two audio streams and control the order the buffers are forwarded.
856 The filter accepts the following options:
860 Set the expression deciding which stream should be
861 forwarded next: if the result is negative, the first stream is forwarded; if
862 the result is positive or zero, the second stream is forwarded. It can use
863 the following variables:
867 number of buffers forwarded so far on each stream
869 number of samples forwarded so far on each stream
871 current timestamp of each stream
874 The default value is @code{t1-t2}, which means to always forward the stream
875 that has a smaller timestamp.
880 Stress-test @code{amerge} by randomly sending buffers on the wrong
881 input, while avoiding too much of a desynchronization:
883 amovie=file.ogg [a] ; amovie=file.mp3 [b] ;
884 [a] [b] astreamsync=(2*random(1))-1+tanh(5*(t1-t2)) [a2] [b2] ;
890 Synchronize audio data with timestamps by squeezing/stretching it and/or
891 dropping samples/adding silence when needed.
893 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
895 The filter accepts the following named parameters:
899 Enable stretching/squeezing the data to make it match the timestamps. Disabled
900 by default. When disabled, time gaps are covered with silence.
903 Minimum difference between timestamps and audio data (in seconds) to trigger
904 adding/dropping samples. Default value is 0.1. If you get non-perfect sync with
905 this filter, try setting this parameter to 0.
908 Maximum compensation in samples per second. Relevant only with compensate=1.
912 Assume the first pts should be this value. The time base is 1 / sample rate.
913 This allows for padding/trimming at the start of stream. By default, no
914 assumption is made about the first frame's expected pts, so no padding or
915 trimming is done. For example, this could be set to 0 to pad the beginning with
916 silence if an audio stream starts after the video stream or to trim any samples
917 with a negative pts due to encoder delay.
925 The filter accepts exactly one parameter, the audio tempo. If not
926 specified then the filter will assume nominal 1.0 tempo. Tempo must
927 be in the [0.5, 2.0] range.
933 Slow down audio to 80% tempo:
939 To speed up audio to 125% tempo:
947 Trim the input so that the output contains one continuous subpart of the input.
949 This filter accepts the following options:
952 Specify time of the start of the kept section, i.e. the audio sample
953 with the timestamp @var{start} will be the first sample in the output.
956 Specify time of the first audio sample that will be dropped, i.e. the
957 audio sample immediately preceding the one with the timestamp @var{end} will be
958 the last sample in the output.
961 Same as @var{start}, except this option sets the start timestamp in samples
965 Same as @var{end}, except this option sets the end timestamp in samples instead
969 Specify maximum duration of the output.
972 Number of the first sample that should be passed to output.
975 Number of the first sample that should be dropped.
978 @option{start}, @option{end}, @option{duration} are expressed as time
979 duration specifications, check the "Time duration" section in the
982 Note that the first two sets of the start/end options and the @option{duration}
983 option look at the frame timestamp, while the _sample options simply count the
984 samples that pass through the filter. So start/end_pts and start/end_sample will
985 give different results when the timestamps are wrong, inexact or do not start at
986 zero. Also note that this filter does not modify the timestamps. If you wish
987 that the output timestamps start at zero, insert the asetpts filter after the
990 If multiple start or end options are set, this filter tries to be greedy and
991 keep all samples that match at least one of the specified constraints. To keep
992 only the part that matches all the constraints at once, chain multiple atrim
995 The defaults are such that all the input is kept. So it is possible to set e.g.
996 just the end values to keep everything before the specified time.
1001 drop everything except the second minute of input
1003 ffmpeg -i INPUT -af atrim=60:120
1007 keep only the first 1000 samples
1009 ffmpeg -i INPUT -af atrim=end_sample=1000
1016 Apply a two-pole Butterworth band-pass filter with central
1017 frequency @var{frequency}, and (3dB-point) band-width width.
1018 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1019 instead of the default: constant 0dB peak gain.
1020 The filter roll off at 6dB per octave (20dB per decade).
1022 The filter accepts the following options:
1026 Set the filter's central frequency. Default is @code{3000}.
1029 Constant skirt gain if set to 1. Defaults to 0.
1032 Set method to specify band-width of filter.
1045 Specify the band-width of a filter in width_type units.
1050 Apply a two-pole Butterworth band-reject filter with central
1051 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1052 The filter roll off at 6dB per octave (20dB per decade).
1054 The filter accepts the following options:
1058 Set the filter's central frequency. Default is @code{3000}.
1061 Set method to specify band-width of filter.
1074 Specify the band-width of a filter in width_type units.
1079 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1080 shelving filter with a response similar to that of a standard
1081 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1083 The filter accepts the following options:
1087 Give the gain at 0 Hz. Its useful range is about -20
1088 (for a large cut) to +20 (for a large boost).
1089 Beware of clipping when using a positive gain.
1092 Set the filter's central frequency and so can be used
1093 to extend or reduce the frequency range to be boosted or cut.
1094 The default value is @code{100} Hz.
1097 Set method to specify band-width of filter.
1110 Determine how steep is the filter's shelf transition.
1115 Apply a biquad IIR filter with the given coefficients.
1116 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1117 are the numerator and denominator coefficients respectively.
1121 Remap input channels to new locations.
1123 This filter accepts the following named parameters:
1125 @item channel_layout
1126 Channel layout of the output stream.
1129 Map channels from input to output. The argument is a '|'-separated list of
1130 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1131 @var{in_channel} form. @var{in_channel} can be either the name of the input
1132 channel (e.g. FL for front left) or its index in the input channel layout.
1133 @var{out_channel} is the name of the output channel or its index in the output
1134 channel layout. If @var{out_channel} is not given then it is implicitly an
1135 index, starting with zero and increasing by one for each mapping.
1138 If no mapping is present, the filter will implicitly map input channels to
1139 output channels preserving index.
1141 For example, assuming a 5.1+downmix input MOV file
1143 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1145 will create an output WAV file tagged as stereo from the downmix channels of
1148 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1150 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:channel_layout=5.1' out.wav
1153 @section channelsplit
1155 Split each channel in input audio stream into a separate output stream.
1157 This filter accepts the following named parameters:
1159 @item channel_layout
1160 Channel layout of the input stream. Default is "stereo".
1163 For example, assuming a stereo input MP3 file
1165 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1167 will create an output Matroska file with two audio streams, one containing only
1168 the left channel and the other the right channel.
1170 To split a 5.1 WAV file into per-channel files
1172 ffmpeg -i in.wav -filter_complex
1173 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1174 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1175 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1181 Compress or expand audio dynamic range.
1183 A description of the accepted options follows.
1188 Set list of times in seconds for each channel over which the instantaneous
1189 level of the input signal is averaged to determine its volume.
1190 @option{attacks} refers to increase of volume and @option{decays} refers
1191 to decrease of volume.
1192 For most situations, the attack time (response to the audio getting louder)
1193 should be shorter than the decay time because the human ear is more sensitive
1194 to sudden loud audio than sudden soft audio.
1195 Typical value for attack is @code{0.3} seconds and for decay @code{0.8}
1199 Set list of points for transfer function, specified in dB relative to maximum
1200 possible signal amplitude.
1201 Each key points list need to be defined using the following syntax:
1202 @code{x0/y0 x1/y1 x2/y2 ...}.
1204 The input values must be in strictly increasing order but the transfer
1205 function does not have to be monotonically rising.
1206 The point @code{0/0} is assumed but may be overridden (by @code{0/out-dBn}).
1207 Typical values for the transfer function are @code{-70/-70 -60/-20}.
1210 Set amount for which the points at where adjacent line segments on the
1211 transfer function meet will be rounded. Defaults is @code{0.01}.
1214 Set additional gain in dB to be applied at all points on the transfer function
1215 and allows easy adjustment of the overall gain.
1216 Default is @code{0}.
1219 Set initial volume in dB to be assumed for each channel when filtering starts.
1220 This permits the user to supply a nominal level initially, so that,
1221 for example, a very large gain is not applied to initial signal levels before
1222 the companding has begun to operate. A typical value for audio which is
1223 initially quiet is -90 dB. Default is @code{0}.
1226 Set delay in seconds. Default is @code{0}. The input audio
1227 is analysed immediately, but audio is delayed before being fed to the
1228 volume adjuster. Specifying a delay approximately equal to the attack/decay
1229 times allows the filter to effectively operate in predictive rather than
1233 @subsection Examples
1236 Make music with both quiet and loud passages suitable for listening
1237 in a noisy environment:
1239 compand=.3 .3:1 1:-90/-60 -60/-40 -40/-30 -20/-20:6:0:-90:0.2
1243 Noise-gate for when the noise is at a lower level than the signal:
1245 compand=.1 .1:.2 .2:-900/-900 -50.1/-900 -50/-50:.01:0:-90:.1
1249 Here is another noise-gate, this time for when the noise is at a higher level
1250 than the signal (making it, in some ways, similar to squelch):
1252 compand=.1 .1:.1 .1:-45.1/-45.1 -45/-900 0/-900:.01:45:-90:.1
1258 Make audio easier to listen to on headphones.
1260 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
1261 so that when listened to on headphones the stereo image is moved from
1262 inside your head (standard for headphones) to outside and in front of
1263 the listener (standard for speakers).
1269 Apply a two-pole peaking equalisation (EQ) filter. With this
1270 filter, the signal-level at and around a selected frequency can
1271 be increased or decreased, whilst (unlike bandpass and bandreject
1272 filters) that at all other frequencies is unchanged.
1274 In order to produce complex equalisation curves, this filter can
1275 be given several times, each with a different central frequency.
1277 The filter accepts the following options:
1281 Set the filter's central frequency in Hz.
1284 Set method to specify band-width of filter.
1297 Specify the band-width of a filter in width_type units.
1300 Set the required gain or attenuation in dB.
1301 Beware of clipping when using a positive gain.
1306 Apply a high-pass filter with 3dB point frequency.
1307 The filter can be either single-pole, or double-pole (the default).
1308 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
1310 The filter accepts the following options:
1314 Set frequency in Hz. Default is 3000.
1317 Set number of poles. Default is 2.
1320 Set method to specify band-width of filter.
1333 Specify the band-width of a filter in width_type units.
1334 Applies only to double-pole filter.
1335 The default is 0.707q and gives a Butterworth response.
1340 Join multiple input streams into one multi-channel stream.
1342 The filter accepts the following named parameters:
1346 Number of input streams. Defaults to 2.
1348 @item channel_layout
1349 Desired output channel layout. Defaults to stereo.
1352 Map channels from inputs to output. The argument is a '|'-separated list of
1353 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
1354 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
1355 can be either the name of the input channel (e.g. FL for front left) or its
1356 index in the specified input stream. @var{out_channel} is the name of the output
1360 The filter will attempt to guess the mappings when those are not specified
1361 explicitly. It does so by first trying to find an unused matching input channel
1362 and if that fails it picks the first unused input channel.
1364 E.g. to join 3 inputs (with properly set channel layouts)
1366 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
1369 To build a 5.1 output from 6 single-channel streams:
1371 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
1372 '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'
1378 Apply a low-pass filter with 3dB point frequency.
1379 The filter can be either single-pole or double-pole (the default).
1380 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
1382 The filter accepts the following options:
1386 Set frequency in Hz. Default is 500.
1389 Set number of poles. Default is 2.
1392 Set method to specify band-width of filter.
1405 Specify the band-width of a filter in width_type units.
1406 Applies only to double-pole filter.
1407 The default is 0.707q and gives a Butterworth response.
1412 Mix channels with specific gain levels. The filter accepts the output
1413 channel layout followed by a set of channels definitions.
1415 This filter is also designed to remap efficiently the channels of an audio
1418 The filter accepts parameters of the form:
1419 "@var{l}:@var{outdef}:@var{outdef}:..."
1423 output channel layout or number of channels
1426 output channel specification, of the form:
1427 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
1430 output channel to define, either a channel name (FL, FR, etc.) or a channel
1431 number (c0, c1, etc.)
1434 multiplicative coefficient for the channel, 1 leaving the volume unchanged
1437 input channel to use, see out_name for details; it is not possible to mix
1438 named and numbered input channels
1441 If the `=' in a channel specification is replaced by `<', then the gains for
1442 that specification will be renormalized so that the total is 1, thus
1443 avoiding clipping noise.
1445 @subsection Mixing examples
1447 For example, if you want to down-mix from stereo to mono, but with a bigger
1448 factor for the left channel:
1450 pan=1:c0=0.9*c0+0.1*c1
1453 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
1454 7-channels surround:
1456 pan=stereo: FL < FL + 0.5*FC + 0.6*BL + 0.6*SL : FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
1459 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
1460 that should be preferred (see "-ac" option) unless you have very specific
1463 @subsection Remapping examples
1465 The channel remapping will be effective if, and only if:
1468 @item gain coefficients are zeroes or ones,
1469 @item only one input per channel output,
1472 If all these conditions are satisfied, the filter will notify the user ("Pure
1473 channel mapping detected"), and use an optimized and lossless method to do the
1476 For example, if you have a 5.1 source and want a stereo audio stream by
1477 dropping the extra channels:
1479 pan="stereo: c0=FL : c1=FR"
1482 Given the same source, you can also switch front left and front right channels
1483 and keep the input channel layout:
1485 pan="5.1: c0=c1 : c1=c0 : c2=c2 : c3=c3 : c4=c4 : c5=c5"
1488 If the input is a stereo audio stream, you can mute the front left channel (and
1489 still keep the stereo channel layout) with:
1494 Still with a stereo audio stream input, you can copy the right channel in both
1495 front left and right:
1497 pan="stereo: c0=FR : c1=FR"
1502 Convert the audio sample format, sample rate and channel layout. This filter is
1503 not meant to be used directly.
1505 @section silencedetect
1507 Detect silence in an audio stream.
1509 This filter logs a message when it detects that the input audio volume is less
1510 or equal to a noise tolerance value for a duration greater or equal to the
1511 minimum detected noise duration.
1513 The printed times and duration are expressed in seconds.
1515 The filter accepts the following options:
1519 Set silence duration until notification (default is 2 seconds).
1522 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
1523 specified value) or amplitude ratio. Default is -60dB, or 0.001.
1526 @subsection Examples
1530 Detect 5 seconds of silence with -50dB noise tolerance:
1532 silencedetect=n=-50dB:d=5
1536 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
1537 tolerance in @file{silence.mp3}:
1539 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
1545 Boost or cut treble (upper) frequencies of the audio using a two-pole
1546 shelving filter with a response similar to that of a standard
1547 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1549 The filter accepts the following options:
1553 Give the gain at whichever is the lower of ~22 kHz and the
1554 Nyquist frequency. Its useful range is about -20 (for a large cut)
1555 to +20 (for a large boost). Beware of clipping when using a positive gain.
1558 Set the filter's central frequency and so can be used
1559 to extend or reduce the frequency range to be boosted or cut.
1560 The default value is @code{3000} Hz.
1563 Set method to specify band-width of filter.
1576 Determine how steep is the filter's shelf transition.
1581 Adjust the input audio volume.
1583 The filter accepts the following options:
1588 Expresses how the audio volume will be increased or decreased.
1590 Output values are clipped to the maximum value.
1592 The output audio volume is given by the relation:
1594 @var{output_volume} = @var{volume} * @var{input_volume}
1597 Default value for @var{volume} is 1.0.
1600 Set the mathematical precision.
1602 This determines which input sample formats will be allowed, which affects the
1603 precision of the volume scaling.
1607 8-bit fixed-point; limits input sample format to U8, S16, and S32.
1609 32-bit floating-point; limits input sample format to FLT. (default)
1611 64-bit floating-point; limits input sample format to DBL.
1615 @subsection Examples
1619 Halve the input audio volume:
1623 volume=volume=-6.0206dB
1626 In all the above example the named key for @option{volume} can be
1627 omitted, for example like in:
1633 Increase input audio power by 6 decibels using fixed-point precision:
1635 volume=volume=6dB:precision=fixed
1639 @section volumedetect
1641 Detect the volume of the input video.
1643 The filter has no parameters. The input is not modified. Statistics about
1644 the volume will be printed in the log when the input stream end is reached.
1646 In particular it will show the mean volume (root mean square), maximum
1647 volume (on a per-sample basis), and the beginning of a histogram of the
1648 registered volume values (from the maximum value to a cumulated 1/1000 of
1651 All volumes are in decibels relative to the maximum PCM value.
1653 @subsection Examples
1655 Here is an excerpt of the output:
1657 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
1658 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
1659 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
1660 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
1661 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
1662 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
1663 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
1664 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
1665 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
1671 The mean square energy is approximately -27 dB, or 10^-2.7.
1673 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
1675 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
1678 In other words, raising the volume by +4 dB does not cause any clipping,
1679 raising it by +5 dB causes clipping for 6 samples, etc.
1681 @c man end AUDIO FILTERS
1683 @chapter Audio Sources
1684 @c man begin AUDIO SOURCES
1686 Below is a description of the currently available audio sources.
1690 Buffer audio frames, and make them available to the filter chain.
1692 This source is mainly intended for a programmatic use, in particular
1693 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
1695 It accepts the following named parameters:
1700 Timebase which will be used for timestamps of submitted frames. It must be
1701 either a floating-point number or in @var{numerator}/@var{denominator} form.
1704 The sample rate of the incoming audio buffers.
1707 The sample format of the incoming audio buffers.
1708 Either a sample format name or its corresponging integer representation from
1709 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
1711 @item channel_layout
1712 The channel layout of the incoming audio buffers.
1713 Either a channel layout name from channel_layout_map in
1714 @file{libavutil/channel_layout.c} or its corresponding integer representation
1715 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
1718 The number of channels of the incoming audio buffers.
1719 If both @var{channels} and @var{channel_layout} are specified, then they
1724 @subsection Examples
1727 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
1730 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
1731 Since the sample format with name "s16p" corresponds to the number
1732 6 and the "stereo" channel layout corresponds to the value 0x3, this is
1735 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
1740 Generate an audio signal specified by an expression.
1742 This source accepts in input one or more expressions (one for each
1743 channel), which are evaluated and used to generate a corresponding
1746 This source accepts the following options:
1750 Set the '|'-separated expressions list for each separate channel. In case the
1751 @option{channel_layout} option is not specified, the selected channel layout
1752 depends on the number of provided expressions.
1754 @item channel_layout, c
1755 Set the channel layout. The number of channels in the specified layout
1756 must be equal to the number of specified expressions.
1759 Set the minimum duration of the sourced audio. See the function
1760 @code{av_parse_time()} for the accepted format.
1761 Note that the resulting duration may be greater than the specified
1762 duration, as the generated audio is always cut at the end of a
1765 If not specified, or the expressed duration is negative, the audio is
1766 supposed to be generated forever.
1769 Set the number of samples per channel per each output frame,
1772 @item sample_rate, s
1773 Specify the sample rate, default to 44100.
1776 Each expression in @var{exprs} can contain the following constants:
1780 number of the evaluated sample, starting from 0
1783 time of the evaluated sample expressed in seconds, starting from 0
1790 @subsection Examples
1800 Generate a sin signal with frequency of 440 Hz, set sample rate to
1803 aevalsrc="sin(440*2*PI*t):s=8000"
1807 Generate a two channels signal, specify the channel layout (Front
1808 Center + Back Center) explicitly:
1810 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
1814 Generate white noise:
1816 aevalsrc="-2+random(0)"
1820 Generate an amplitude modulated signal:
1822 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
1826 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
1828 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
1835 Null audio source, return unprocessed audio frames. It is mainly useful
1836 as a template and to be employed in analysis / debugging tools, or as
1837 the source for filters which ignore the input data (for example the sox
1840 This source accepts the following options:
1844 @item channel_layout, cl
1846 Specify the channel layout, and can be either an integer or a string
1847 representing a channel layout. The default value of @var{channel_layout}
1850 Check the channel_layout_map definition in
1851 @file{libavutil/channel_layout.c} for the mapping between strings and
1852 channel layout values.
1854 @item sample_rate, r
1855 Specify the sample rate, and defaults to 44100.
1858 Set the number of samples per requested frames.
1862 @subsection Examples
1866 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
1868 anullsrc=r=48000:cl=4
1872 Do the same operation with a more obvious syntax:
1874 anullsrc=r=48000:cl=mono
1878 All the parameters need to be explicitly defined.
1882 Synthesize a voice utterance using the libflite library.
1884 To enable compilation of this filter you need to configure FFmpeg with
1885 @code{--enable-libflite}.
1887 Note that the flite library is not thread-safe.
1889 The filter accepts the following options:
1894 If set to 1, list the names of the available voices and exit
1895 immediately. Default value is 0.
1898 Set the maximum number of samples per frame. Default value is 512.
1901 Set the filename containing the text to speak.
1904 Set the text to speak.
1907 Set the voice to use for the speech synthesis. Default value is
1908 @code{kal}. See also the @var{list_voices} option.
1911 @subsection Examples
1915 Read from file @file{speech.txt}, and synthetize the text using the
1916 standard flite voice:
1918 flite=textfile=speech.txt
1922 Read the specified text selecting the @code{slt} voice:
1924 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
1928 Input text to ffmpeg:
1930 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
1934 Make @file{ffplay} speak the specified text, using @code{flite} and
1935 the @code{lavfi} device:
1937 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
1941 For more information about libflite, check:
1942 @url{http://www.speech.cs.cmu.edu/flite/}
1946 Generate an audio signal made of a sine wave with amplitude 1/8.
1948 The audio signal is bit-exact.
1950 The filter accepts the following options:
1955 Set the carrier frequency. Default is 440 Hz.
1957 @item beep_factor, b
1958 Enable a periodic beep every second with frequency @var{beep_factor} times
1959 the carrier frequency. Default is 0, meaning the beep is disabled.
1961 @item sample_rate, r
1962 Specify the sample rate, default is 44100.
1965 Specify the duration of the generated audio stream.
1967 @item samples_per_frame
1968 Set the number of samples per output frame, default is 1024.
1971 @subsection Examples
1976 Generate a simple 440 Hz sine wave:
1982 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
1986 sine=frequency=220:beep_factor=4:duration=5
1991 @c man end AUDIO SOURCES
1993 @chapter Audio Sinks
1994 @c man begin AUDIO SINKS
1996 Below is a description of the currently available audio sinks.
1998 @section abuffersink
2000 Buffer audio frames, and make them available to the end of filter chain.
2002 This sink is mainly intended for programmatic use, in particular
2003 through the interface defined in @file{libavfilter/buffersink.h}
2004 or the options system.
2006 It accepts a pointer to an AVABufferSinkContext structure, which
2007 defines the incoming buffers' formats, to be passed as the opaque
2008 parameter to @code{avfilter_init_filter} for initialization.
2012 Null audio sink, do absolutely nothing with the input audio. It is
2013 mainly useful as a template and to be employed in analysis / debugging
2016 @c man end AUDIO SINKS
2018 @chapter Video Filters
2019 @c man begin VIDEO FILTERS
2021 When you configure your FFmpeg build, you can disable any of the
2022 existing filters using @code{--disable-filters}.
2023 The configure output will show the video filters included in your
2026 Below is a description of the currently available video filters.
2028 @section alphaextract
2030 Extract the alpha component from the input as a grayscale video. This
2031 is especially useful with the @var{alphamerge} filter.
2035 Add or replace the alpha component of the primary input with the
2036 grayscale value of a second input. This is intended for use with
2037 @var{alphaextract} to allow the transmission or storage of frame
2038 sequences that have alpha in a format that doesn't support an alpha
2041 For example, to reconstruct full frames from a normal YUV-encoded video
2042 and a separate video created with @var{alphaextract}, you might use:
2044 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
2047 Since this filter is designed for reconstruction, it operates on frame
2048 sequences without considering timestamps, and terminates when either
2049 input reaches end of stream. This will cause problems if your encoding
2050 pipeline drops frames. If you're trying to apply an image as an
2051 overlay to a video stream, consider the @var{overlay} filter instead.
2055 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
2056 and libavformat to work. On the other hand, it is limited to ASS (Advanced
2057 Substation Alpha) subtitles files.
2061 Compute the bounding box for the non-black pixels in the input frame
2064 This filter computes the bounding box containing all the pixels with a
2065 luminance value greater than the minimum allowed value.
2066 The parameters describing the bounding box are printed on the filter
2069 The filter accepts the following option:
2073 Set the minimal luminance value. Default is @code{16}.
2076 @section blackdetect
2078 Detect video intervals that are (almost) completely black. Can be
2079 useful to detect chapter transitions, commercials, or invalid
2080 recordings. Output lines contains the time for the start, end and
2081 duration of the detected black interval expressed in seconds.
2083 In order to display the output lines, you need to set the loglevel at
2084 least to the AV_LOG_INFO value.
2086 The filter accepts the following options:
2089 @item black_min_duration, d
2090 Set the minimum detected black duration expressed in seconds. It must
2091 be a non-negative floating point number.
2093 Default value is 2.0.
2095 @item picture_black_ratio_th, pic_th
2096 Set the threshold for considering a picture "black".
2097 Express the minimum value for the ratio:
2099 @var{nb_black_pixels} / @var{nb_pixels}
2102 for which a picture is considered black.
2103 Default value is 0.98.
2105 @item pixel_black_th, pix_th
2106 Set the threshold for considering a pixel "black".
2108 The threshold expresses the maximum pixel luminance value for which a
2109 pixel is considered "black". The provided value is scaled according to
2110 the following equation:
2112 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
2115 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
2116 the input video format, the range is [0-255] for YUV full-range
2117 formats and [16-235] for YUV non full-range formats.
2119 Default value is 0.10.
2122 The following example sets the maximum pixel threshold to the minimum
2123 value, and detects only black intervals of 2 or more seconds:
2125 blackdetect=d=2:pix_th=0.00
2130 Detect frames that are (almost) completely black. Can be useful to
2131 detect chapter transitions or commercials. Output lines consist of
2132 the frame number of the detected frame, the percentage of blackness,
2133 the position in the file if known or -1 and the timestamp in seconds.
2135 In order to display the output lines, you need to set the loglevel at
2136 least to the AV_LOG_INFO value.
2138 The filter accepts the following options:
2143 Set the percentage of the pixels that have to be below the threshold, defaults
2146 @item threshold, thresh
2147 Set the threshold below which a pixel value is considered black, defaults to
2154 Blend two video frames into each other.
2156 It takes two input streams and outputs one stream, the first input is the
2157 "top" layer and second input is "bottom" layer.
2158 Output terminates when shortest input terminates.
2160 A description of the accepted options follows.
2168 Set blend mode for specific pixel component or all pixel components in case
2169 of @var{all_mode}. Default value is @code{normal}.
2171 Available values for component modes are:
2204 Set blend opacity for specific pixel component or all pixel components in case
2205 of @var{all_opacity}. Only used in combination with pixel component blend modes.
2212 Set blend expression for specific pixel component or all pixel components in case
2213 of @var{all_expr}. Note that related mode options will be ignored if those are set.
2215 The expressions can use the following variables:
2219 The sequential number of the filtered frame, starting from @code{0}.
2223 the coordinates of the current sample
2227 the width and height of currently filtered plane
2231 Width and height scale depending on the currently filtered plane. It is the
2232 ratio between the corresponding luma plane number of pixels and the current
2233 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
2234 @code{0.5,0.5} for chroma planes.
2237 Time of the current frame, expressed in seconds.
2240 Value of pixel component at current location for first video frame (top layer).
2243 Value of pixel component at current location for second video frame (bottom layer).
2247 Force termination when the shortest input terminates. Default is @code{0}.
2249 Continue applying the last bottom frame after the end of the stream. A value of
2250 @code{0} disable the filter after the last frame of the bottom layer is reached.
2251 Default is @code{1}.
2254 @subsection Examples
2258 Apply transition from bottom layer to top layer in first 10 seconds:
2260 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
2264 Apply 1x1 checkerboard effect:
2266 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
2272 Apply boxblur algorithm to the input video.
2274 The filter accepts the following options:
2278 @item luma_radius, lr
2279 @item luma_power, lp
2280 @item chroma_radius, cr
2281 @item chroma_power, cp
2282 @item alpha_radius, ar
2283 @item alpha_power, ap
2287 A description of the accepted options follows.
2290 @item luma_radius, lr
2291 @item chroma_radius, cr
2292 @item alpha_radius, ar
2293 Set an expression for the box radius in pixels used for blurring the
2294 corresponding input plane.
2296 The radius value must be a non-negative number, and must not be
2297 greater than the value of the expression @code{min(w,h)/2} for the
2298 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
2301 Default value for @option{luma_radius} is "2". If not specified,
2302 @option{chroma_radius} and @option{alpha_radius} default to the
2303 corresponding value set for @option{luma_radius}.
2305 The expressions can contain the following constants:
2309 the input width and height in pixels
2313 the input chroma image width and height in pixels
2317 horizontal and vertical chroma subsample values. For example for the
2318 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2321 @item luma_power, lp
2322 @item chroma_power, cp
2323 @item alpha_power, ap
2324 Specify how many times the boxblur filter is applied to the
2325 corresponding plane.
2327 Default value for @option{luma_power} is 2. If not specified,
2328 @option{chroma_power} and @option{alpha_power} default to the
2329 corresponding value set for @option{luma_power}.
2331 A value of 0 will disable the effect.
2334 @subsection Examples
2338 Apply a boxblur filter with luma, chroma, and alpha radius
2341 boxblur=luma_radius=2:luma_power=1
2346 Set luma radius to 2, alpha and chroma radius to 0:
2348 boxblur=2:1:cr=0:ar=0
2352 Set luma and chroma radius to a fraction of the video dimension:
2354 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
2358 @section colorbalance
2359 Modify intensity of primary colors (red, green and blue) of input frames.
2361 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
2362 regions for the red-cyan, green-magenta or blue-yellow balance.
2364 A positive adjustment value shifts the balance towards the primary color, a negative
2365 value towards the complementary color.
2367 The filter accepts the following options:
2373 Adjust red, green and blue shadows (darkest pixels).
2378 Adjust red, green and blue midtones (medium pixels).
2383 Adjust red, green and blue highlights (brightest pixels).
2385 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
2388 @subsection Examples
2392 Add red color cast to shadows:
2398 @section colorchannelmixer
2400 Adjust video input frames by re-mixing color channels.
2402 This filter modifies a color channel by adding the values associated to
2403 the other channels of the same pixels. For example if the value to
2404 modify is red, the output value will be:
2406 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
2409 The filter accepts the following options:
2416 Adjust contribution of input red, green, blue and alpha channels for output red channel.
2417 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
2423 Adjust contribution of input red, green, blue and alpha channels for output green channel.
2424 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
2430 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
2431 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
2437 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
2438 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
2440 Allowed ranges for options are @code{[-2.0, 2.0]}.
2443 @subsection Examples
2447 Convert source to grayscale:
2449 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
2452 Simulate sepia tones:
2454 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
2458 @section colormatrix
2460 Convert color matrix.
2462 The filter accepts the following options:
2467 Specify the source and destination color matrix. Both values must be
2470 The accepted values are:
2486 For example to convert from BT.601 to SMPTE-240M, use the command:
2488 colormatrix=bt601:smpte240m
2493 Copy the input source unchanged to the output. Mainly useful for
2498 Crop the input video to given dimensions.
2500 The filter accepts the following options:
2504 Width of the output video. It defaults to @code{iw}.
2505 This expression is evaluated only once during the filter
2509 Height of the output video. It defaults to @code{ih}.
2510 This expression is evaluated only once during the filter
2514 Horizontal position, in the input video, of the left edge of the output video.
2515 It defaults to @code{(in_w-out_w)/2}.
2516 This expression is evaluated per-frame.
2519 Vertical position, in the input video, of the top edge of the output video.
2520 It defaults to @code{(in_h-out_h)/2}.
2521 This expression is evaluated per-frame.
2524 If set to 1 will force the output display aspect ratio
2525 to be the same of the input, by changing the output sample aspect
2526 ratio. It defaults to 0.
2529 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
2530 expressions containing the following constants:
2535 the computed values for @var{x} and @var{y}. They are evaluated for
2540 the input width and height
2544 same as @var{in_w} and @var{in_h}
2548 the output (cropped) width and height
2552 same as @var{out_w} and @var{out_h}
2555 same as @var{iw} / @var{ih}
2558 input sample aspect ratio
2561 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
2565 horizontal and vertical chroma subsample values. For example for the
2566 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2569 the number of input frame, starting from 0
2572 the position in the file of the input frame, NAN if unknown
2575 timestamp expressed in seconds, NAN if the input timestamp is unknown
2579 The expression for @var{out_w} may depend on the value of @var{out_h},
2580 and the expression for @var{out_h} may depend on @var{out_w}, but they
2581 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
2582 evaluated after @var{out_w} and @var{out_h}.
2584 The @var{x} and @var{y} parameters specify the expressions for the
2585 position of the top-left corner of the output (non-cropped) area. They
2586 are evaluated for each frame. If the evaluated value is not valid, it
2587 is approximated to the nearest valid value.
2589 The expression for @var{x} may depend on @var{y}, and the expression
2590 for @var{y} may depend on @var{x}.
2592 @subsection Examples
2596 Crop area with size 100x100 at position (12,34).
2601 Using named options, the example above becomes:
2603 crop=w=100:h=100:x=12:y=34
2607 Crop the central input area with size 100x100:
2613 Crop the central input area with size 2/3 of the input video:
2615 crop=2/3*in_w:2/3*in_h
2619 Crop the input video central square:
2626 Delimit the rectangle with the top-left corner placed at position
2627 100:100 and the right-bottom corner corresponding to the right-bottom
2628 corner of the input image:
2630 crop=in_w-100:in_h-100:100:100
2634 Crop 10 pixels from the left and right borders, and 20 pixels from
2635 the top and bottom borders
2637 crop=in_w-2*10:in_h-2*20
2641 Keep only the bottom right quarter of the input image:
2643 crop=in_w/2:in_h/2:in_w/2:in_h/2
2647 Crop height for getting Greek harmony:
2649 crop=in_w:1/PHI*in_w
2653 Appply trembling effect:
2655 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)
2659 Apply erratic camera effect depending on timestamp:
2661 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)"
2665 Set x depending on the value of y:
2667 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
2673 Auto-detect crop size.
2675 Calculate necessary cropping parameters and prints the recommended
2676 parameters through the logging system. The detected dimensions
2677 correspond to the non-black area of the input video.
2679 The filter accepts the following options:
2684 Set higher black value threshold, which can be optionally specified
2685 from nothing (0) to everything (255). An intensity value greater
2686 to the set value is considered non-black. Default value is 24.
2689 Set the value for which the width/height should be divisible by. The
2690 offset is automatically adjusted to center the video. Use 2 to get
2691 only even dimensions (needed for 4:2:2 video). 16 is best when
2692 encoding to most video codecs. Default value is 16.
2694 @item reset_count, reset
2695 Set the counter that determines after how many frames cropdetect will
2696 reset the previously detected largest video area and start over to
2697 detect the current optimal crop area. Default value is 0.
2699 This can be useful when channel logos distort the video area. 0
2700 indicates never reset and return the largest area encountered during
2707 Apply color adjustments using curves.
2709 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
2710 component (red, green and blue) has its values defined by @var{N} key points
2711 tied from each other using a smooth curve. The x-axis represents the pixel
2712 values from the input frame, and the y-axis the new pixel values to be set for
2715 By default, a component curve is defined by the two points @var{(0;0)} and
2716 @var{(1;1)}. This creates a straight line where each original pixel value is
2717 "adjusted" to its own value, which means no change to the image.
2719 The filter allows you to redefine these two points and add some more. A new
2720 curve (using a natural cubic spline interpolation) will be define to pass
2721 smoothly through all these new coordinates. The new defined points needs to be
2722 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
2723 be in the @var{[0;1]} interval. If the computed curves happened to go outside
2724 the vector spaces, the values will be clipped accordingly.
2726 If there is no key point defined in @code{x=0}, the filter will automatically
2727 insert a @var{(0;0)} point. In the same way, if there is no key point defined
2728 in @code{x=1}, the filter will automatically insert a @var{(1;1)} point.
2730 The filter accepts the following options:
2734 Select one of the available color presets. This option can be used in addition
2735 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
2736 options takes priority on the preset values.
2737 Available presets are:
2740 @item color_negative
2743 @item increase_contrast
2745 @item linear_contrast
2746 @item medium_contrast
2748 @item strong_contrast
2751 Default is @code{none}.
2753 Set the master key points. These points will define a second pass mapping. It
2754 is sometimes called a "luminance" or "value" mapping. It can be used with
2755 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
2756 post-processing LUT.
2758 Set the key points for the red component.
2760 Set the key points for the green component.
2762 Set the key points for the blue component.
2764 Set the key points for all components (not including master).
2765 Can be used in addition to the other key points component
2766 options. In this case, the unset component(s) will fallback on this
2767 @option{all} setting.
2769 Specify a Photoshop curves file (@code{.asv}) to import the settings from.
2772 To avoid some filtergraph syntax conflicts, each key points list need to be
2773 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
2775 @subsection Examples
2779 Increase slightly the middle level of blue:
2781 curves=blue='0.5/0.58'
2787 curves=r='0/0.11 .42/.51 1/0.95':g='0.50/0.48':b='0/0.22 .49/.44 1/0.8'
2789 Here we obtain the following coordinates for each components:
2792 @code{(0;0.11) (0.42;0.51) (1;0.95)}
2794 @code{(0;0) (0.50;0.48) (1;1)}
2796 @code{(0;0.22) (0.49;0.44) (1;0.80)}
2800 The previous example can also be achieved with the associated built-in preset:
2802 curves=preset=vintage
2812 Use a Photoshop preset and redefine the points of the green component:
2814 curves=psfile='MyCurvesPresets/purple.asv':green='0.45/0.53'
2820 Denoise frames using 2D DCT (frequency domain filtering).
2822 This filter is not designed for real time and can be extremely slow.
2824 The filter accepts the following options:
2828 Set the noise sigma constant.
2830 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
2831 coefficient (absolute value) below this threshold with be dropped.
2833 If you need a more advanced filtering, see @option{expr}.
2835 Default is @code{0}.
2838 Set number overlapping pixels for each block. Each block is of size
2839 @code{16x16}. Since the filter can be slow, you may want to reduce this value,
2840 at the cost of a less effective filter and the risk of various artefacts.
2842 If the overlapping value doesn't allow to process the whole input width or
2843 height, a warning will be displayed and according borders won't be denoised.
2845 Default value is @code{15}.
2848 Set the coefficient factor expression.
2850 For each coefficient of a DCT block, this expression will be evaluated as a
2851 multiplier value for the coefficient.
2853 If this is option is set, the @option{sigma} option will be ignored.
2855 The absolute value of the coefficient can be accessed through the @var{c}
2859 @subsection Examples
2861 Apply a denoise with a @option{sigma} of @code{4.5}:
2866 The same operation can be achieved using the expression system:
2868 dctdnoiz=e='gte(c, 4.5*3)'
2874 Drop duplicated frames at regular intervals.
2876 The filter accepts the following options:
2880 Set the number of frames from which one will be dropped. Setting this to
2881 @var{N} means one frame in every batch of @var{N} frames will be dropped.
2882 Default is @code{5}.
2885 Set the threshold for duplicate detection. If the difference metric for a frame
2886 is less than or equal to this value, then it is declared as duplicate. Default
2890 Set scene change threshold. Default is @code{15}.
2894 Set the size of the x and y-axis blocks used during metric calculations.
2895 Larger blocks give better noise suppression, but also give worse detection of
2896 small movements. Must be a power of two. Default is @code{32}.
2899 Mark main input as a pre-processed input and activate clean source input
2900 stream. This allows the input to be pre-processed with various filters to help
2901 the metrics calculation while keeping the frame selection lossless. When set to
2902 @code{1}, the first stream is for the pre-processed input, and the second
2903 stream is the clean source from where the kept frames are chosen. Default is
2907 Set whether or not chroma is considered in the metric calculations. Default is
2913 Suppress a TV station logo by a simple interpolation of the surrounding
2914 pixels. Just set a rectangle covering the logo and watch it disappear
2915 (and sometimes something even uglier appear - your mileage may vary).
2917 This filter accepts the following options:
2922 Specify the top left corner coordinates of the logo. They must be
2927 Specify the width and height of the logo to clear. They must be
2931 Specify the thickness of the fuzzy edge of the rectangle (added to
2932 @var{w} and @var{h}). The default value is 4.
2935 When set to 1, a green rectangle is drawn on the screen to simplify
2936 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
2937 The default value is 0.
2939 The rectangle is drawn on the outermost pixels which will be (partly)
2940 replaced with interpolated values. The values of the next pixels
2941 immediately outside this rectangle in each direction will be used to
2942 compute the interpolated pixel values inside the rectangle.
2946 @subsection Examples
2950 Set a rectangle covering the area with top left corner coordinates 0,0
2951 and size 100x77, setting a band of size 10:
2953 delogo=x=0:y=0:w=100:h=77:band=10
2960 Attempt to fix small changes in horizontal and/or vertical shift. This
2961 filter helps remove camera shake from hand-holding a camera, bumping a
2962 tripod, moving on a vehicle, etc.
2964 The filter accepts the following options:
2972 Specify a rectangular area where to limit the search for motion
2974 If desired the search for motion vectors can be limited to a
2975 rectangular area of the frame defined by its top left corner, width
2976 and height. These parameters have the same meaning as the drawbox
2977 filter which can be used to visualise the position of the bounding
2980 This is useful when simultaneous movement of subjects within the frame
2981 might be confused for camera motion by the motion vector search.
2983 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
2984 then the full frame is used. This allows later options to be set
2985 without specifying the bounding box for the motion vector search.
2987 Default - search the whole frame.
2991 Specify the maximum extent of movement in x and y directions in the
2992 range 0-64 pixels. Default 16.
2995 Specify how to generate pixels to fill blanks at the edge of the
2996 frame. Available values are:
2999 Fill zeroes at blank locations
3001 Original image at blank locations
3003 Extruded edge value at blank locations
3005 Mirrored edge at blank locations
3007 Default value is @samp{mirror}.
3010 Specify the blocksize to use for motion search. Range 4-128 pixels,
3014 Specify the contrast threshold for blocks. Only blocks with more than
3015 the specified contrast (difference between darkest and lightest
3016 pixels) will be considered. Range 1-255, default 125.
3019 Specify the search strategy. Available values are:
3022 Set exhaustive search
3024 Set less exhaustive search.
3026 Default value is @samp{exhaustive}.
3029 If set then a detailed log of the motion search is written to the
3033 If set to 1, specify using OpenCL capabilities, only available if
3034 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
3040 Draw a colored box on the input image.
3042 This filter accepts the following options:
3047 The expressions which specify the top left corner coordinates of the box. Default to 0.
3051 The expressions which specify the width and height of the box, if 0 they are interpreted as
3052 the input width and height. Default to 0.
3055 Specify the color of the box to write, it can be the name of a color
3056 (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
3057 value @code{invert} is used, the box edge color is the same as the
3058 video with inverted luma.
3061 The expression which sets the thickness of the box edge. Default value is @code{3}.
3063 See below for the list of accepted constants.
3066 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
3067 following constants:
3071 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
3075 horizontal and vertical chroma subsample values. For example for the
3076 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3080 The input width and height.
3083 The input sample aspect ratio.
3087 The x and y offset coordinates where the box is drawn.
3091 The width and height of the drawn box.
3094 The thickness of the drawn box.
3096 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
3097 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
3101 @subsection Examples
3105 Draw a black box around the edge of the input image:
3111 Draw a box with color red and an opacity of 50%:
3113 drawbox=10:20:200:60:red@@0.5
3116 The previous example can be specified as:
3118 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
3122 Fill the box with pink color:
3124 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
3128 Draw a 2-pixel red 2.40:1 mask:
3130 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
3136 Draw a grid on the input image.
3138 This filter accepts the following options:
3143 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
3147 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
3148 input width and height, respectively, minus @code{thickness}, so image gets
3149 framed. Default to 0.
3152 Specify the color of the grid, it can be the name of a color
3153 (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
3154 value @code{invert} is used, the grid color is the same as the
3155 video with inverted luma.
3156 Note that you can append opacity value (in range of 0.0 - 1.0)
3157 to color name after @@ sign.
3160 The expression which sets the thickness of the grid line. Default value is @code{1}.
3162 See below for the list of accepted constants.
3165 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
3166 following constants:
3170 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
3174 horizontal and vertical chroma subsample values. For example for the
3175 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3179 The input grid cell width and height.
3182 The input sample aspect ratio.
3186 The x and y coordinates of some point of grid intersection (meant to configure offset).
3190 The width and height of the drawn cell.
3193 The thickness of the drawn cell.
3195 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
3196 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
3200 @subsection Examples
3204 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
3206 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
3210 Draw a white 3x3 grid with an opacity of 50%:
3212 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
3219 Draw text string or text from specified file on top of video using the
3220 libfreetype library.
3222 To enable compilation of this filter you need to configure FFmpeg with
3223 @code{--enable-libfreetype}.
3227 The description of the accepted parameters follows.
3232 Used to draw a box around text using background color.
3233 Value should be either 1 (enable) or 0 (disable).
3234 The default value of @var{box} is 0.
3237 The color to be used for drawing box around text.
3238 Either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
3239 (e.g. "0xff00ff"), possibly followed by an alpha specifier.
3240 The default value of @var{boxcolor} is "white".
3243 Set an expression which specifies if the text should be drawn. If the
3244 expression evaluates to 0, the text is not drawn. This is useful for
3245 specifying that the text should be drawn only when specific conditions
3248 Default value is "1".
3250 See below for the list of accepted constants and functions.
3253 Select how the @var{text} is expanded. Can be either @code{none},
3254 @code{strftime} (deprecated) or
3255 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
3259 If true, check and fix text coords to avoid clipping.
3262 The color to be used for drawing fonts.
3263 Either a string (e.g. "red") or in 0xRRGGBB[AA] format
3264 (e.g. "0xff000033"), possibly followed by an alpha specifier.
3265 The default value of @var{fontcolor} is "black".
3268 The font file to be used for drawing text. Path must be included.
3269 This parameter is mandatory.
3272 The font size to be used for drawing text.
3273 The default value of @var{fontsize} is 16.
3276 Flags to be used for loading the fonts.
3278 The flags map the corresponding flags supported by libfreetype, and are
3279 a combination of the following values:
3286 @item vertical_layout
3287 @item force_autohint
3290 @item ignore_global_advance_width
3292 @item ignore_transform
3298 Default value is "render".
3300 For more information consult the documentation for the FT_LOAD_*
3304 The color to be used for drawing a shadow behind the drawn text. It
3305 can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
3306 form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
3307 The default value of @var{shadowcolor} is "black".
3311 The x and y offsets for the text shadow position with respect to the
3312 position of the text. They can be either positive or negative
3313 values. Default value for both is "0".
3316 The starting frame number for the n/frame_num variable. The default value
3320 The size in number of spaces to use for rendering the tab.
3324 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
3325 format. It can be used with or without text parameter. @var{timecode_rate}
3326 option must be specified.
3328 @item timecode_rate, rate, r
3329 Set the timecode frame rate (timecode only).
3332 The text string to be drawn. The text must be a sequence of UTF-8
3334 This parameter is mandatory if no file is specified with the parameter
3338 A text file containing text to be drawn. The text must be a sequence
3339 of UTF-8 encoded characters.
3341 This parameter is mandatory if no text string is specified with the
3342 parameter @var{text}.
3344 If both @var{text} and @var{textfile} are specified, an error is thrown.
3347 If set to 1, the @var{textfile} will be reloaded before each frame.
3348 Be sure to update it atomically, or it may be read partially, or even fail.
3352 The expressions which specify the offsets where text will be drawn
3353 within the video frame. They are relative to the top/left border of the
3356 The default value of @var{x} and @var{y} is "0".
3358 See below for the list of accepted constants and functions.
3361 The parameters for @var{x} and @var{y} are expressions containing the
3362 following constants and functions:
3366 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
3370 horizontal and vertical chroma subsample values. For example for the
3371 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3374 the height of each text line
3382 @item max_glyph_a, ascent
3383 the maximum distance from the baseline to the highest/upper grid
3384 coordinate used to place a glyph outline point, for all the rendered
3386 It is a positive value, due to the grid's orientation with the Y axis
3389 @item max_glyph_d, descent
3390 the maximum distance from the baseline to the lowest grid coordinate
3391 used to place a glyph outline point, for all the rendered glyphs.
3392 This is a negative value, due to the grid's orientation, with the Y axis
3396 maximum glyph height, that is the maximum height for all the glyphs
3397 contained in the rendered text, it is equivalent to @var{ascent} -
3401 maximum glyph width, that is the maximum width for all the glyphs
3402 contained in the rendered text
3405 the number of input frame, starting from 0
3407 @item rand(min, max)
3408 return a random number included between @var{min} and @var{max}
3411 input sample aspect ratio
3414 timestamp expressed in seconds, NAN if the input timestamp is unknown
3417 the height of the rendered text
3420 the width of the rendered text
3424 the x and y offset coordinates where the text is drawn.
3426 These parameters allow the @var{x} and @var{y} expressions to refer
3427 each other, so you can for example specify @code{y=x/dar}.
3430 If libavfilter was built with @code{--enable-fontconfig}, then
3431 @option{fontfile} can be a fontconfig pattern or omitted.
3433 @anchor{drawtext_expansion}
3434 @subsection Text expansion
3436 If @option{expansion} is set to @code{strftime},
3437 the filter recognizes strftime() sequences in the provided text and
3438 expands them accordingly. Check the documentation of strftime(). This
3439 feature is deprecated.
3441 If @option{expansion} is set to @code{none}, the text is printed verbatim.
3443 If @option{expansion} is set to @code{normal} (which is the default),
3444 the following expansion mechanism is used.
3446 The backslash character '\', followed by any character, always expands to
3447 the second character.
3449 Sequence of the form @code{%@{...@}} are expanded. The text between the
3450 braces is a function name, possibly followed by arguments separated by ':'.
3451 If the arguments contain special characters or delimiters (':' or '@}'),
3452 they should be escaped.
3454 Note that they probably must also be escaped as the value for the
3455 @option{text} option in the filter argument string and as the filter
3456 argument in the filtergraph description, and possibly also for the shell,
3457 that makes up to four levels of escaping; using a text file avoids these
3460 The following functions are available:
3465 The expression evaluation result.
3467 It must take one argument specifying the expression to be evaluated,
3468 which accepts the same constants and functions as the @var{x} and
3469 @var{y} values. Note that not all constants should be used, for
3470 example the text size is not known when evaluating the expression, so
3471 the constants @var{text_w} and @var{text_h} will have an undefined
3475 The time at which the filter is running, expressed in UTC.
3476 It can accept an argument: a strftime() format string.
3479 The time at which the filter is running, expressed in the local time zone.
3480 It can accept an argument: a strftime() format string.
3483 Frame metadata. It must take one argument specifying metadata key.
3486 The frame number, starting from 0.
3489 A 1 character description of the current picture type.
3492 The timestamp of the current frame, in seconds, with microsecond accuracy.
3496 @subsection Examples
3500 Draw "Test Text" with font FreeSerif, using the default values for the
3501 optional parameters.
3504 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
3508 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
3509 and y=50 (counting from the top-left corner of the screen), text is
3510 yellow with a red box around it. Both the text and the box have an
3514 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
3515 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
3518 Note that the double quotes are not necessary if spaces are not used
3519 within the parameter list.
3522 Show the text at the center of the video frame:
3524 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h-line_h)/2"
3528 Show a text line sliding from right to left in the last row of the video
3529 frame. The file @file{LONG_LINE} is assumed to contain a single line
3532 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
3536 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
3538 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
3542 Draw a single green letter "g", at the center of the input video.
3543 The glyph baseline is placed at half screen height.
3545 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
3549 Show text for 1 second every 3 seconds:
3551 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:draw=lt(mod(t\,3)\,1):text='blink'"
3555 Use fontconfig to set the font. Note that the colons need to be escaped.
3557 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
3561 Print the date of a real-time encoding (see strftime(3)):
3563 drawtext='fontfile=FreeSans.ttf:text=%@{localtime:%a %b %d %Y@}'
3568 For more information about libfreetype, check:
3569 @url{http://www.freetype.org/}.
3571 For more information about fontconfig, check:
3572 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
3576 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
3578 The filter accepts the following options:
3583 Set low and high threshold values used by the Canny thresholding
3586 The high threshold selects the "strong" edge pixels, which are then
3587 connected through 8-connectivity with the "weak" edge pixels selected
3588 by the low threshold.
3590 @var{low} and @var{high} threshold values must be choosen in the range
3591 [0,1], and @var{low} should be lesser or equal to @var{high}.
3593 Default value for @var{low} is @code{20/255}, and default value for @var{high}
3599 edgedetect=low=0.1:high=0.4
3602 @section extractplanes
3604 Extract color channel components from input video stream into
3605 separate grayscale video streams.
3607 The filter accepts the following option:
3611 Set plane(s) to extract.
3613 Available values for planes are:
3624 Choosing planes not available in the input will result in an error.
3625 That means you cannot select @code{r}, @code{g}, @code{b} planes
3626 with @code{y}, @code{u}, @code{v} planes at same time.
3629 @subsection Examples
3633 Extract luma, u and v color channel component from input video frame
3634 into 3 grayscale outputs:
3636 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
3642 Apply fade-in/out effect to input video.
3644 This filter accepts the following options:
3648 The effect type -- can be either "in" for fade-in, or "out" for a fade-out
3650 Default is @code{in}.
3652 @item start_frame, s
3653 Specify the number of the start frame for starting to apply the fade
3654 effect. Default is 0.
3657 The number of frames for which the fade effect has to last. At the end of the
3658 fade-in effect the output video will have the same intensity as the input video,
3659 at the end of the fade-out transition the output video will be completely black.
3663 If set to 1, fade only alpha channel, if one exists on the input.
3666 @item start_time, st
3667 Specify the timestamp (in seconds) of the frame to start to apply the fade
3668 effect. If both start_frame and start_time are specified, the fade will start at
3669 whichever comes last. Default is 0.
3672 The number of seconds for which the fade effect has to last. At the end of the
3673 fade-in effect the output video will have the same intensity as the input video,
3674 at the end of the fade-out transition the output video will be completely black.
3675 If both duration and nb_frames are specified, duration is used. Default is 0.
3678 @subsection Examples
3682 Fade in first 30 frames of video:
3687 The command above is equivalent to:
3693 Fade out last 45 frames of a 200-frame video:
3696 fade=type=out:start_frame=155:nb_frames=45
3700 Fade in first 25 frames and fade out last 25 frames of a 1000-frame video:
3702 fade=in:0:25, fade=out:975:25
3706 Make first 5 frames black, then fade in from frame 5-24:
3712 Fade in alpha over first 25 frames of video:
3714 fade=in:0:25:alpha=1
3718 Make first 5.5 seconds black, then fade in for 0.5 seconds:
3720 fade=t=in:st=5.5:d=0.5
3727 Extract a single field from an interlaced image using stride
3728 arithmetic to avoid wasting CPU time. The output frames are marked as
3731 The filter accepts the following options:
3735 Specify whether to extract the top (if the value is @code{0} or
3736 @code{top}) or the bottom field (if the value is @code{1} or
3742 Field matching filter for inverse telecine. It is meant to reconstruct the
3743 progressive frames from a telecined stream. The filter does not drop duplicated
3744 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
3745 followed by a decimation filter such as @ref{decimate} in the filtergraph.
3747 The separation of the field matching and the decimation is notably motivated by
3748 the possibility of inserting a de-interlacing filter fallback between the two.
3749 If the source has mixed telecined and real interlaced content,
3750 @code{fieldmatch} will not be able to match fields for the interlaced parts.
3751 But these remaining combed frames will be marked as interlaced, and thus can be
3752 de-interlaced by a later filter such as @ref{yadif} before decimation.
3754 In addition to the various configuration options, @code{fieldmatch} can take an
3755 optional second stream, activated through the @option{ppsrc} option. If
3756 enabled, the frames reconstruction will be based on the fields and frames from
3757 this second stream. This allows the first input to be pre-processed in order to
3758 help the various algorithms of the filter, while keeping the output lossless
3759 (assuming the fields are matched properly). Typically, a field-aware denoiser,
3760 or brightness/contrast adjustments can help.
3762 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
3763 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
3764 which @code{fieldmatch} is based on. While the semantic and usage are very
3765 close, some behaviour and options names can differ.
3767 The filter accepts the following options:
3771 Specify the assumed field order of the input stream. Available values are:
3775 Auto detect parity (use FFmpeg's internal parity value).
3777 Assume bottom field first.
3779 Assume top field first.
3782 Note that it is sometimes recommended not to trust the parity announced by the
3785 Default value is @var{auto}.
3788 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
3789 sense that it won't risk creating jerkiness due to duplicate frames when
3790 possible, but if there are bad edits or blended fields it will end up
3791 outputting combed frames when a good match might actually exist. On the other
3792 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
3793 but will almost always find a good frame if there is one. The other values are
3794 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
3795 jerkiness and creating duplicate frames versus finding good matches in sections
3796 with bad edits, orphaned fields, blended fields, etc.
3798 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
3800 Available values are:
3804 2-way matching (p/c)
3806 2-way matching, and trying 3rd match if still combed (p/c + n)
3808 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
3810 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
3811 still combed (p/c + n + u/b)
3813 3-way matching (p/c/n)
3815 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
3816 detected as combed (p/c/n + u/b)
3819 The parenthesis at the end indicate the matches that would be used for that
3820 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
3823 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
3826 Default value is @var{pc_n}.
3829 Mark the main input stream as a pre-processed input, and enable the secondary
3830 input stream as the clean source to pick the fields from. See the filter
3831 introduction for more details. It is similar to the @option{clip2} feature from
3834 Default value is @code{0} (disabled).
3837 Set the field to match from. It is recommended to set this to the same value as
3838 @option{order} unless you experience matching failures with that setting. In
3839 certain circumstances changing the field that is used to match from can have a
3840 large impact on matching performance. Available values are:
3844 Automatic (same value as @option{order}).
3846 Match from the bottom field.
3848 Match from the top field.
3851 Default value is @var{auto}.
3854 Set whether or not chroma is included during the match comparisons. In most
3855 cases it is recommended to leave this enabled. You should set this to @code{0}
3856 only if your clip has bad chroma problems such as heavy rainbowing or other
3857 artifacts. Setting this to @code{0} could also be used to speed things up at
3858 the cost of some accuracy.
3860 Default value is @code{1}.
3864 These define an exclusion band which excludes the lines between @option{y0} and
3865 @option{y1} from being included in the field matching decision. An exclusion
3866 band can be used to ignore subtitles, a logo, or other things that may
3867 interfere with the matching. @option{y0} sets the starting scan line and
3868 @option{y1} sets the ending line; all lines in between @option{y0} and
3869 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
3870 @option{y0} and @option{y1} to the same value will disable the feature.
3871 @option{y0} and @option{y1} defaults to @code{0}.
3874 Set the scene change detection threshold as a percentage of maximum change on
3875 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
3876 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
3877 @option{scthresh} is @code{[0.0, 100.0]}.
3879 Default value is @code{12.0}.
3882 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
3883 account the combed scores of matches when deciding what match to use as the
3884 final match. Available values are:
3888 No final matching based on combed scores.
3890 Combed scores are only used when a scene change is detected.
3892 Use combed scores all the time.
3895 Default is @var{sc}.
3898 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
3899 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
3900 Available values are:
3904 No forced calculation.
3906 Force p/c/n calculations.
3908 Force p/c/n/u/b calculations.
3911 Default value is @var{none}.
3914 This is the area combing threshold used for combed frame detection. This
3915 essentially controls how "strong" or "visible" combing must be to be detected.
3916 Larger values mean combing must be more visible and smaller values mean combing
3917 can be less visible or strong and still be detected. Valid settings are from
3918 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
3919 be detected as combed). This is basically a pixel difference value. A good
3920 range is @code{[8, 12]}.
3922 Default value is @code{9}.
3925 Sets whether or not chroma is considered in the combed frame decision. Only
3926 disable this if your source has chroma problems (rainbowing, etc.) that are
3927 causing problems for the combed frame detection with chroma enabled. Actually,
3928 using @option{chroma}=@var{0} is usually more reliable, except for the case
3929 where there is chroma only combing in the source.
3931 Default value is @code{0}.
3935 Respectively set the x-axis and y-axis size of the window used during combed
3936 frame detection. This has to do with the size of the area in which
3937 @option{combpel} pixels are required to be detected as combed for a frame to be
3938 declared combed. See the @option{combpel} parameter description for more info.
3939 Possible values are any number that is a power of 2 starting at 4 and going up
3942 Default value is @code{16}.
3945 The number of combed pixels inside any of the @option{blocky} by
3946 @option{blockx} size blocks on the frame for the frame to be detected as
3947 combed. While @option{cthresh} controls how "visible" the combing must be, this
3948 setting controls "how much" combing there must be in any localized area (a
3949 window defined by the @option{blockx} and @option{blocky} settings) on the
3950 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
3951 which point no frames will ever be detected as combed). This setting is known
3952 as @option{MI} in TFM/VFM vocabulary.
3954 Default value is @code{80}.
3957 @anchor{p/c/n/u/b meaning}
3958 @subsection p/c/n/u/b meaning
3960 @subsubsection p/c/n
3962 We assume the following telecined stream:
3965 Top fields: 1 2 2 3 4
3966 Bottom fields: 1 2 3 4 4
3969 The numbers correspond to the progressive frame the fields relate to. Here, the
3970 first two frames are progressive, the 3rd and 4th are combed, and so on.
3972 When @code{fieldmatch} is configured to run a matching from bottom
3973 (@option{field}=@var{bottom}) this is how this input stream get transformed:
3978 B 1 2 3 4 4 <-- matching reference
3987 As a result of the field matching, we can see that some frames get duplicated.
3988 To perform a complete inverse telecine, you need to rely on a decimation filter
3989 after this operation. See for instance the @ref{decimate} filter.
3991 The same operation now matching from top fields (@option{field}=@var{top})
3996 T 1 2 2 3 4 <-- matching reference
4006 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
4007 basically, they refer to the frame and field of the opposite parity:
4010 @item @var{p} matches the field of the opposite parity in the previous frame
4011 @item @var{c} matches the field of the opposite parity in the current frame
4012 @item @var{n} matches the field of the opposite parity in the next frame
4017 The @var{u} and @var{b} matching are a bit special in the sense that they match
4018 from the opposite parity flag. In the following examples, we assume that we are
4019 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
4020 'x' is placed above and below each matched fields.
4022 With bottom matching (@option{field}=@var{bottom}):
4027 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
4028 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
4036 With top matching (@option{field}=@var{top}):
4041 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
4042 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
4050 @subsection Examples
4052 Simple IVTC of a top field first telecined stream:
4054 fieldmatch=order=tff:combmatch=none, decimate
4057 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
4059 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
4064 Transform the field order of the input video.
4066 This filter accepts the following options:
4071 Output field order. Valid values are @var{tff} for top field first or @var{bff}
4072 for bottom field first.
4075 Default value is @samp{tff}.
4077 Transformation is achieved by shifting the picture content up or down
4078 by one line, and filling the remaining line with appropriate picture content.
4079 This method is consistent with most broadcast field order converters.
4081 If the input video is not flagged as being interlaced, or it is already
4082 flagged as being of the required output field order then this filter does
4083 not alter the incoming video.
4085 This filter is very useful when converting to or from PAL DV material,
4086 which is bottom field first.
4090 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
4095 Buffer input images and send them when they are requested.
4097 This filter is mainly useful when auto-inserted by the libavfilter
4100 The filter does not take parameters.
4105 Convert the input video to one of the specified pixel formats.
4106 Libavfilter will try to pick one that is supported for the input to
4109 This filter accepts the following parameters:
4113 A '|'-separated list of pixel format names, for example
4114 "pix_fmts=yuv420p|monow|rgb24".
4118 @subsection Examples
4122 Convert the input video to the format @var{yuv420p}
4124 format=pix_fmts=yuv420p
4127 Convert the input video to any of the formats in the list
4129 format=pix_fmts=yuv420p|yuv444p|yuv410p
4135 Convert the video to specified constant frame rate by duplicating or dropping
4136 frames as necessary.
4138 This filter accepts the following named parameters:
4142 Desired output frame rate. The default is @code{25}.
4147 Possible values are:
4150 zero round towards 0
4154 round towards -infinity
4156 round towards +infinity
4160 The default is @code{near}.
4164 Alternatively, the options can be specified as a flat string:
4165 @var{fps}[:@var{round}].
4167 See also the @ref{setpts} filter.
4169 @subsection Examples
4173 A typical usage in order to set the fps to 25:
4179 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
4181 fps=fps=film:round=near
4187 Select one frame every N-th frame.
4189 This filter accepts the following option:
4192 Select frame after every @code{step} frames.
4193 Allowed values are positive integers higher than 0. Default value is @code{1}.
4199 Apply a frei0r effect to the input video.
4201 To enable compilation of this filter you need to install the frei0r
4202 header and configure FFmpeg with @code{--enable-frei0r}.
4204 This filter accepts the following options:
4209 The name to the frei0r effect to load. If the environment variable
4210 @env{FREI0R_PATH} is defined, the frei0r effect is searched in each one of the
4211 directories specified by the colon separated list in @env{FREIOR_PATH},
4212 otherwise in the standard frei0r paths, which are in this order:
4213 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
4214 @file{/usr/lib/frei0r-1/}.
4217 A '|'-separated list of parameters to pass to the frei0r effect.
4221 A frei0r effect parameter can be a boolean (whose values are specified
4222 with "y" and "n"), a double, a color (specified by the syntax
4223 @var{R}/@var{G}/@var{B}, @var{R}, @var{G}, and @var{B} being float
4224 numbers from 0.0 to 1.0) or by an @code{av_parse_color()} color
4225 description), a position (specified by the syntax @var{X}/@var{Y},
4226 @var{X} and @var{Y} being float numbers) and a string.
4228 The number and kind of parameters depend on the loaded effect. If an
4229 effect parameter is not specified the default value is set.
4231 @subsection Examples
4235 Apply the distort0r effect, set the first two double parameters:
4237 frei0r=filter_name=distort0r:filter_params=0.5|0.01
4241 Apply the colordistance effect, take a color as first parameter:
4243 frei0r=colordistance:0.2/0.3/0.4
4244 frei0r=colordistance:violet
4245 frei0r=colordistance:0x112233
4249 Apply the perspective effect, specify the top left and top right image
4252 frei0r=perspective:0.2/0.2|0.8/0.2
4256 For more information see:
4257 @url{http://frei0r.dyne.org}
4261 The filter accepts the following options:
4265 Set the luminance expression.
4267 Set the chrominance blue expression.
4269 Set the chrominance red expression.
4271 Set the alpha expression.
4273 Set the red expression.
4275 Set the green expression.
4277 Set the blue expression.
4280 The colorspace is selected according to the specified options. If one
4281 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
4282 options is specified, the filter will automatically select a YCbCr
4283 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
4284 @option{blue_expr} options is specified, it will select an RGB
4287 If one of the chrominance expression is not defined, it falls back on the other
4288 one. If no alpha expression is specified it will evaluate to opaque value.
4289 If none of chrominance expressions are specified, they will evaluate
4290 to the luminance expression.
4292 The expressions can use the following variables and functions:
4296 The sequential number of the filtered frame, starting from @code{0}.
4300 The coordinates of the current sample.
4304 The width and height of the image.
4308 Width and height scale depending on the currently filtered plane. It is the
4309 ratio between the corresponding luma plane number of pixels and the current
4310 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4311 @code{0.5,0.5} for chroma planes.
4314 Time of the current frame, expressed in seconds.
4317 Return the value of the pixel at location (@var{x},@var{y}) of the current
4321 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
4325 Return the value of the pixel at location (@var{x},@var{y}) of the
4326 blue-difference chroma plane. Return 0 if there is no such plane.
4329 Return the value of the pixel at location (@var{x},@var{y}) of the
4330 red-difference chroma plane. Return 0 if there is no such plane.
4335 Return the value of the pixel at location (@var{x},@var{y}) of the
4336 red/green/blue component. Return 0 if there is no such component.
4339 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
4340 plane. Return 0 if there is no such plane.
4343 For functions, if @var{x} and @var{y} are outside the area, the value will be
4344 automatically clipped to the closer edge.
4346 @subsection Examples
4350 Flip the image horizontally:
4356 Generate a bidimensional sine wave, with angle @code{PI/3} and a
4357 wavelength of 100 pixels:
4359 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
4363 Generate a fancy enigmatic moving light:
4365 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
4369 Generate a quick emboss effect:
4371 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
4375 Modify RGB components depending on pixel position:
4377 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
4383 Fix the banding artifacts that are sometimes introduced into nearly flat
4384 regions by truncation to 8bit color depth.
4385 Interpolate the gradients that should go where the bands are, and
4388 This filter is designed for playback only. Do not use it prior to
4389 lossy compression, because compression tends to lose the dither and
4390 bring back the bands.
4392 This filter accepts the following options:
4397 The maximum amount by which the filter will change any one pixel. Also the
4398 threshold for detecting nearly flat regions. Acceptable values range from .51 to
4399 64, default value is 1.2, out-of-range values will be clipped to the valid
4403 The neighborhood to fit the gradient to. A larger radius makes for smoother
4404 gradients, but also prevents the filter from modifying the pixels near detailed
4405 regions. Acceptable values are 8-32, default value is 16, out-of-range values
4406 will be clipped to the valid range.
4410 Alternatively, the options can be specified as a flat string:
4411 @var{strength}[:@var{radius}]
4413 @subsection Examples
4417 Apply the filter with a @code{3.5} strength and radius of @code{8}:
4423 Specify radius, omitting the strength (which will fall-back to the default
4434 Apply a Hald CLUT to a video stream.
4436 First input is the video stream to process, and second one is the Hald CLUT.
4437 The Hald CLUT input can be a simple picture or a complete video stream.
4439 The filter accepts the following options:
4443 Force termination when the shortest input terminates. Default is @code{0}.
4445 Continue applying the last CLUT after the end of the stream. A value of
4446 @code{0} disable the filter after the last frame of the CLUT is reached.
4447 Default is @code{1}.
4450 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
4451 filters share the same internals).
4453 More information about the Hald CLUT can be found on Eskil Steenberg's website
4454 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
4456 @subsection Workflow examples
4458 @subsubsection Hald CLUT video stream
4460 Generate an identity Hald CLUT stream altered with various effects:
4462 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
4465 Note: make sure you use a lossless codec.
4467 Then use it with @code{haldclut} to apply it on some random stream:
4469 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
4472 The Hald CLUT will be applied to the 10 first seconds (duration of
4473 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
4474 to the remaining frames of the @code{mandelbrot} stream.
4476 @subsubsection Hald CLUT with preview
4478 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
4479 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
4480 biggest possible square starting at the top left of the picture. The remaining
4481 padding pixels (bottom or right) will be ignored. This area can be used to add
4482 a preview of the Hald CLUT.
4484 Typically, the following generated Hald CLUT will be supported by the
4485 @code{haldclut} filter:
4488 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
4489 pad=iw+320 [padded_clut];
4490 smptebars=s=320x256, split [a][b];
4491 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
4492 [main][b] overlay=W-320" -frames:v 1 clut.png
4495 It contains the original and a preview of the effect of the CLUT: SMPTE color
4496 bars are displayed on the right-top, and below the same color bars processed by
4499 Then, the effect of this Hald CLUT can be visualized with:
4501 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
4506 Flip the input video horizontally.
4508 For example to horizontally flip the input video with @command{ffmpeg}:
4510 ffmpeg -i in.avi -vf "hflip" out.avi
4514 This filter applies a global color histogram equalization on a
4517 It can be used to correct video that has a compressed range of pixel
4518 intensities. The filter redistributes the pixel intensities to
4519 equalize their distribution across the intensity range. It may be
4520 viewed as an "automatically adjusting contrast filter". This filter is
4521 useful only for correcting degraded or poorly captured source
4524 The filter accepts the following options:
4528 Determine the amount of equalization to be applied. As the strength
4529 is reduced, the distribution of pixel intensities more-and-more
4530 approaches that of the input frame. The value must be a float number
4531 in the range [0,1] and defaults to 0.200.
4534 Set the maximum intensity that can generated and scale the output
4535 values appropriately. The strength should be set as desired and then
4536 the intensity can be limited if needed to avoid washing-out. The value
4537 must be a float number in the range [0,1] and defaults to 0.210.
4540 Set the antibanding level. If enabled the filter will randomly vary
4541 the luminance of output pixels by a small amount to avoid banding of
4542 the histogram. Possible values are @code{none}, @code{weak} or
4543 @code{strong}. It defaults to @code{none}.
4548 Compute and draw a color distribution histogram for the input video.
4550 The computed histogram is a representation of distribution of color components
4553 The filter accepts the following options:
4559 It accepts the following values:
4562 standard histogram that display color components distribution in an image.
4563 Displays color graph for each color component. Shows distribution
4564 of the Y, U, V, A or G, B, R components, depending on input format,
4565 in current frame. Bellow each graph is color component scale meter.
4568 chroma values in vectorscope, if brighter more such chroma values are
4569 distributed in an image.
4570 Displays chroma values (U/V color placement) in two dimensional graph
4571 (which is called a vectorscope). It can be used to read of the hue and
4572 saturation of the current frame. At a same time it is a histogram.
4573 The whiter a pixel in the vectorscope, the more pixels of the input frame
4574 correspond to that pixel (that is the more pixels have this chroma value).
4575 The V component is displayed on the horizontal (X) axis, with the leftmost
4576 side being V = 0 and the rightmost side being V = 255.
4577 The U component is displayed on the vertical (Y) axis, with the top
4578 representing U = 0 and the bottom representing U = 255.
4580 The position of a white pixel in the graph corresponds to the chroma value
4581 of a pixel of the input clip. So the graph can be used to read of the
4582 hue (color flavor) and the saturation (the dominance of the hue in the color).
4583 As the hue of a color changes, it moves around the square. At the center of
4584 the square, the saturation is zero, which means that the corresponding pixel
4585 has no color. If you increase the amount of a specific color, while leaving
4586 the other colors unchanged, the saturation increases, and you move towards
4587 the edge of the square.
4590 chroma values in vectorscope, similar as @code{color} but actual chroma values
4594 per row/column color component graph. In row mode graph in the left side represents
4595 color component value 0 and right side represents value = 255. In column mode top
4596 side represents color component value = 0 and bottom side represents value = 255.
4598 Default value is @code{levels}.
4601 Set height of level in @code{levels}. Default value is @code{200}.
4602 Allowed range is [50, 2048].
4605 Set height of color scale in @code{levels}. Default value is @code{12}.
4606 Allowed range is [0, 40].
4609 Set step for @code{waveform} mode. Smaller values are useful to find out how much
4610 of same luminance values across input rows/columns are distributed.
4611 Default value is @code{10}. Allowed range is [1, 255].
4614 Set mode for @code{waveform}. Can be either @code{row}, or @code{column}.
4615 Default is @code{row}.
4618 Set display mode for @code{waveform} and @code{levels}.
4619 It accepts the following values:
4622 Display separate graph for the color components side by side in
4623 @code{row} waveform mode or one below other in @code{column} waveform mode
4624 for @code{waveform} histogram mode. For @code{levels} histogram mode
4625 per color component graphs are placed one bellow other.
4627 This display mode in @code{waveform} histogram mode makes it easy to spot
4628 color casts in the highlights and shadows of an image, by comparing the
4629 contours of the top and the bottom of each waveform.
4630 Since whites, grays, and blacks are characterized by
4631 exactly equal amounts of red, green, and blue, neutral areas of the
4632 picture should display three waveforms of roughly equal width/height.
4633 If not, the correction is easy to make by making adjustments to level the
4637 Presents information that's identical to that in the @code{parade}, except
4638 that the graphs representing color components are superimposed directly
4641 This display mode in @code{waveform} histogram mode can make it easier to spot
4642 the relative differences or similarities in overlapping areas of the color
4643 components that are supposed to be identical, such as neutral whites, grays,
4646 Default is @code{parade}.
4649 Set mode for @code{levels}. Can be either @code{linear}, or @code{logarithmic}.
4650 Default is @code{linear}.
4653 @subsection Examples
4658 Calculate and draw histogram:
4660 ffplay -i input -vf histogram
4668 High precision/quality 3d denoise filter. This filter aims to reduce
4669 image noise producing smooth images and making still images really
4670 still. It should enhance compressibility.
4672 It accepts the following optional parameters:
4676 a non-negative float number which specifies spatial luma strength,
4679 @item chroma_spatial
4680 a non-negative float number which specifies spatial chroma strength,
4681 defaults to 3.0*@var{luma_spatial}/4.0
4684 a float number which specifies luma temporal strength, defaults to
4685 6.0*@var{luma_spatial}/4.0
4688 a float number which specifies chroma temporal strength, defaults to
4689 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}
4694 Modify the hue and/or the saturation of the input.
4696 This filter accepts the following options:
4700 Specify the hue angle as a number of degrees. It accepts an expression,
4701 and defaults to "0".
4704 Specify the saturation in the [-10,10] range. It accepts an expression and
4708 Specify the hue angle as a number of radians. It accepts an
4709 expression, and defaults to "0".
4712 @option{h} and @option{H} are mutually exclusive, and can't be
4713 specified at the same time.
4715 The @option{h}, @option{H} and @option{s} option values are
4716 expressions containing the following constants:
4720 frame count of the input frame starting from 0
4723 presentation timestamp of the input frame expressed in time base units
4726 frame rate of the input video, NAN if the input frame rate is unknown
4729 timestamp expressed in seconds, NAN if the input timestamp is unknown
4732 time base of the input video
4735 @subsection Examples
4739 Set the hue to 90 degrees and the saturation to 1.0:
4745 Same command but expressing the hue in radians:
4751 Rotate hue and make the saturation swing between 0
4752 and 2 over a period of 1 second:
4754 hue="H=2*PI*t: s=sin(2*PI*t)+1"
4758 Apply a 3 seconds saturation fade-in effect starting at 0:
4763 The general fade-in expression can be written as:
4765 hue="s=min(0\, max((t-START)/DURATION\, 1))"
4769 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
4771 hue="s=max(0\, min(1\, (8-t)/3))"
4774 The general fade-out expression can be written as:
4776 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
4781 @subsection Commands
4783 This filter supports the following commands:
4788 Modify the hue and/or the saturation of the input video.
4789 The command accepts the same syntax of the corresponding option.
4791 If the specified expression is not valid, it is kept at its current
4797 Detect video interlacing type.
4799 This filter tries to detect if the input is interlaced or progressive,
4800 top or bottom field first.
4802 The filter accepts the following options:
4806 Set interlacing threshold.
4808 Set progressive threshold.
4813 Deinterleave or interleave fields.
4815 This filter allows to process interlaced images fields without
4816 deinterlacing them. Deinterleaving splits the input frame into 2
4817 fields (so called half pictures). Odd lines are moved to the top
4818 half of the output image, even lines to the bottom half.
4819 You can process (filter) them independently and then re-interleave them.
4821 The filter accepts the following options:
4825 @item chroma_mode, c
4827 Available values for @var{luma_mode}, @var{chroma_mode} and
4828 @var{alpha_mode} are:
4834 @item deinterleave, d
4835 Deinterleave fields, placing one above the other.
4838 Interleave fields. Reverse the effect of deinterleaving.
4840 Default value is @code{none}.
4843 @item chroma_swap, cs
4844 @item alpha_swap, as
4845 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
4850 Simple interlacing filter from progressive contents. This interleaves upper (or
4851 lower) lines from odd frames with lower (or upper) lines from even frames,
4852 halving the frame rate and preserving image height.
4855 Original Original New Frame
4856 Frame 'j' Frame 'j+1' (tff)
4857 ========== =========== ==================
4858 Line 0 --------------------> Frame 'j' Line 0
4859 Line 1 Line 1 ----> Frame 'j+1' Line 1
4860 Line 2 ---------------------> Frame 'j' Line 2
4861 Line 3 Line 3 ----> Frame 'j+1' Line 3
4863 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
4866 It accepts the following optional parameters:
4870 determines whether the interlaced frame is taken from the even (tff - default)
4871 or odd (bff) lines of the progressive frame.
4874 Enable (default) or disable the vertical lowpass filter to avoid twitter
4875 interlacing and reduce moire patterns.
4880 Deinterlace input video by applying Donald Graft's adaptive kernel
4881 deinterling. Work on interlaced parts of a video to produce
4884 The description of the accepted parameters follows.
4888 Set the threshold which affects the filter's tolerance when
4889 determining if a pixel line must be processed. It must be an integer
4890 in the range [0,255] and defaults to 10. A value of 0 will result in
4891 applying the process on every pixels.
4894 Paint pixels exceeding the threshold value to white if set to 1.
4898 Set the fields order. Swap fields if set to 1, leave fields alone if
4902 Enable additional sharpening if set to 1. Default is 0.
4905 Enable twoway sharpening if set to 1. Default is 0.
4908 @subsection Examples
4912 Apply default values:
4914 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
4918 Enable additional sharpening:
4924 Paint processed pixels in white:
4933 Apply a 3D LUT to an input video.
4935 The filter accepts the following options:
4939 Set the 3D LUT file name.
4941 Currently supported formats:
4953 Select interpolation mode.
4955 Available values are:
4959 Use values from the nearest defined point.
4961 Interpolate values using the 8 points defining a cube.
4963 Interpolate values using a tetrahedron.
4967 @section lut, lutrgb, lutyuv
4969 Compute a look-up table for binding each pixel component input value
4970 to an output value, and apply it to input video.
4972 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
4973 to an RGB input video.
4975 These filters accept the following options:
4978 set first pixel component expression
4980 set second pixel component expression
4982 set third pixel component expression
4984 set fourth pixel component expression, corresponds to the alpha component
4987 set red component expression
4989 set green component expression
4991 set blue component expression
4993 alpha component expression
4996 set Y/luminance component expression
4998 set U/Cb component expression
5000 set V/Cr component expression
5003 Each of them specifies the expression to use for computing the lookup table for
5004 the corresponding pixel component values.
5006 The exact component associated to each of the @var{c*} options depends on the
5009 The @var{lut} filter requires either YUV or RGB pixel formats in input,
5010 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
5012 The expressions can contain the following constants and functions:
5017 the input width and height
5020 input value for the pixel component
5023 the input value clipped in the @var{minval}-@var{maxval} range
5026 maximum value for the pixel component
5029 minimum value for the pixel component
5032 the negated value for the pixel component value clipped in the
5033 @var{minval}-@var{maxval} range , it corresponds to the expression
5034 "maxval-clipval+minval"
5037 the computed value in @var{val} clipped in the
5038 @var{minval}-@var{maxval} range
5040 @item gammaval(gamma)
5041 the computed gamma correction value of the pixel component value
5042 clipped in the @var{minval}-@var{maxval} range, corresponds to the
5044 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
5048 All expressions default to "val".
5050 @subsection Examples
5056 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
5057 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
5060 The above is the same as:
5062 lutrgb="r=negval:g=negval:b=negval"
5063 lutyuv="y=negval:u=negval:v=negval"
5073 Remove chroma components, turns the video into a graytone image:
5075 lutyuv="u=128:v=128"
5079 Apply a luma burning effect:
5085 Remove green and blue components:
5091 Set a constant alpha channel value on input:
5093 format=rgba,lutrgb=a="maxval-minval/2"
5097 Correct luminance gamma by a 0.5 factor:
5099 lutyuv=y=gammaval(0.5)
5103 Discard least significant bits of luma:
5105 lutyuv=y='bitand(val, 128+64+32)'
5111 Apply motion-compensation deinterlacing.
5113 It needs one field per frame as input and must thus be used together
5114 with yadif=1/3 or equivalent.
5116 This filter accepts the following options:
5119 Set the deinterlacing mode.
5121 It accepts one of the following values:
5126 use iterative motion estimation
5128 like @samp{slow}, but use multiple reference frames.
5130 Default value is @samp{fast}.
5133 Set the picture field parity assumed for the input video. It must be
5134 one of the following values:
5138 assume top field first
5140 assume bottom field first
5143 Default value is @samp{bff}.
5146 Set per-block quantization parameter (QP) used by the internal
5149 Higher values should result in a smoother motion vector field but less
5150 optimal individual vectors. Default value is 1.
5155 Apply an MPlayer filter to the input video.
5157 This filter provides a wrapper around some of the filters of
5160 This wrapper is considered experimental. Some of the wrapped filters
5161 may not work properly and we may drop support for them, as they will
5162 be implemented natively into FFmpeg. Thus you should avoid
5163 depending on them when writing portable scripts.
5165 The filter accepts the parameters:
5166 @var{filter_name}[:=]@var{filter_params}
5168 @var{filter_name} is the name of a supported MPlayer filter,
5169 @var{filter_params} is a string containing the parameters accepted by
5172 The list of the currently supported filters follows:
5188 The parameter syntax and behavior for the listed filters are the same
5189 of the corresponding MPlayer filters. For detailed instructions check
5190 the "VIDEO FILTERS" section in the MPlayer manual.
5192 @subsection Examples
5196 Adjust gamma, brightness, contrast:
5202 See also mplayer(1), @url{http://www.mplayerhq.hu/}.
5206 Drop frames that do not differ greatly from the previous frame in
5207 order to reduce frame rate.
5209 The main use of this filter is for very-low-bitrate encoding
5210 (e.g. streaming over dialup modem), but it could in theory be used for
5211 fixing movies that were inverse-telecined incorrectly.
5213 A description of the accepted options follows.
5217 Set the maximum number of consecutive frames which can be dropped (if
5218 positive), or the minimum interval between dropped frames (if
5219 negative). If the value is 0, the frame is dropped unregarding the
5220 number of previous sequentially dropped frames.
5227 Set the dropping threshold values.
5229 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
5230 represent actual pixel value differences, so a threshold of 64
5231 corresponds to 1 unit of difference for each pixel, or the same spread
5232 out differently over the block.
5234 A frame is a candidate for dropping if no 8x8 blocks differ by more
5235 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
5236 meaning the whole image) differ by more than a threshold of @option{lo}.
5238 Default value for @option{hi} is 64*12, default value for @option{lo} is
5239 64*5, and default value for @option{frac} is 0.33.
5247 This filter accepts an integer in input, if non-zero it negates the
5248 alpha component (if available). The default value in input is 0.
5252 Force libavfilter not to use any of the specified pixel formats for the
5253 input to the next filter.
5255 This filter accepts the following parameters:
5259 A '|'-separated list of pixel format names, for example
5260 "pix_fmts=yuv420p|monow|rgb24".
5264 @subsection Examples
5268 Force libavfilter to use a format different from @var{yuv420p} for the
5269 input to the vflip filter:
5271 noformat=pix_fmts=yuv420p,vflip
5275 Convert the input video to any of the formats not contained in the list:
5277 noformat=yuv420p|yuv444p|yuv410p
5283 Add noise on video input frame.
5285 The filter accepts the following options:
5293 Set noise seed for specific pixel component or all pixel components in case
5294 of @var{all_seed}. Default value is @code{123457}.
5296 @item all_strength, alls
5297 @item c0_strength, c0s
5298 @item c1_strength, c1s
5299 @item c2_strength, c2s
5300 @item c3_strength, c3s
5301 Set noise strength for specific pixel component or all pixel components in case
5302 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
5304 @item all_flags, allf
5309 Set pixel component flags or set flags for all components if @var{all_flags}.
5310 Available values for component flags are:
5313 averaged temporal noise (smoother)
5315 mix random noise with a (semi)regular pattern
5317 temporal noise (noise pattern changes between frames)
5319 uniform noise (gaussian otherwise)
5323 @subsection Examples
5325 Add temporal and uniform noise to input video:
5327 noise=alls=20:allf=t+u
5332 Pass the video source unchanged to the output.
5336 Apply video transform using libopencv.
5338 To enable this filter install libopencv library and headers and
5339 configure FFmpeg with @code{--enable-libopencv}.
5341 This filter accepts the following parameters:
5346 The name of the libopencv filter to apply.
5349 The parameters to pass to the libopencv filter. If not specified the default
5354 Refer to the official libopencv documentation for more precise
5356 @url{http://opencv.willowgarage.com/documentation/c/image_filtering.html}
5358 Follows the list of supported libopencv filters.
5363 Dilate an image by using a specific structuring element.
5364 This filter corresponds to the libopencv function @code{cvDilate}.
5366 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
5368 @var{struct_el} represents a structuring element, and has the syntax:
5369 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
5371 @var{cols} and @var{rows} represent the number of columns and rows of
5372 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
5373 point, and @var{shape} the shape for the structuring element, and
5374 can be one of the values "rect", "cross", "ellipse", "custom".
5376 If the value for @var{shape} is "custom", it must be followed by a
5377 string of the form "=@var{filename}". The file with name
5378 @var{filename} is assumed to represent a binary image, with each
5379 printable character corresponding to a bright pixel. When a custom
5380 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
5381 or columns and rows of the read file are assumed instead.
5383 The default value for @var{struct_el} is "3x3+0x0/rect".
5385 @var{nb_iterations} specifies the number of times the transform is
5386 applied to the image, and defaults to 1.
5388 Follow some example:
5390 # use the default values
5393 # dilate using a structuring element with a 5x5 cross, iterate two times
5394 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
5396 # read the shape from the file diamond.shape, iterate two times
5397 # the file diamond.shape may contain a pattern of characters like this:
5403 # the specified cols and rows are ignored (but not the anchor point coordinates)
5404 ocv=dilate:0x0+2x2/custom=diamond.shape|2
5409 Erode an image by using a specific structuring element.
5410 This filter corresponds to the libopencv function @code{cvErode}.
5412 The filter accepts the parameters: @var{struct_el}:@var{nb_iterations},
5413 with the same syntax and semantics as the @ref{dilate} filter.
5417 Smooth the input video.
5419 The filter takes the following parameters:
5420 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
5422 @var{type} is the type of smooth filter to apply, and can be one of
5423 the following values: "blur", "blur_no_scale", "median", "gaussian",
5424 "bilateral". The default value is "gaussian".
5426 @var{param1}, @var{param2}, @var{param3}, and @var{param4} are
5427 parameters whose meanings depend on smooth type. @var{param1} and
5428 @var{param2} accept integer positive values or 0, @var{param3} and
5429 @var{param4} accept float values.
5431 The default value for @var{param1} is 3, the default value for the
5432 other parameters is 0.
5434 These parameters correspond to the parameters assigned to the
5435 libopencv function @code{cvSmooth}.
5440 Overlay one video on top of another.
5442 It takes two inputs and one output, the first input is the "main"
5443 video on which the second input is overlayed.
5445 This filter accepts the following parameters:
5447 A description of the accepted options follows.
5452 Set the expression for the x and y coordinates of the overlayed video
5453 on the main video. Default value is "0" for both expressions. In case
5454 the expression is invalid, it is set to a huge value (meaning that the
5455 overlay will not be displayed within the output visible area).
5458 Set when the expressions for @option{x}, and @option{y} are evaluated.
5460 It accepts the following values:
5463 only evaluate expressions once during the filter initialization or
5464 when a command is processed
5467 evaluate expressions for each incoming frame
5470 Default value is @samp{frame}.
5473 If set to 1, force the output to terminate when the shortest input
5474 terminates. Default value is 0.
5477 Set the format for the output video.
5479 It accepts the following values:
5491 Default value is @samp{yuv420}.
5493 @item rgb @emph{(deprecated)}
5494 If set to 1, force the filter to accept inputs in the RGB
5495 color space. Default value is 0. This option is deprecated, use
5496 @option{format} instead.
5499 If set to 1, force the filter to draw the last overlay frame over the
5500 main input until the end of the stream. A value of 0 disables this
5501 behavior. Default value is 1.
5504 The @option{x}, and @option{y} expressions can contain the following
5510 main input width and height
5514 overlay input width and height
5518 the computed values for @var{x} and @var{y}. They are evaluated for
5523 horizontal and vertical chroma subsample values of the output
5524 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
5528 the number of input frame, starting from 0
5531 the position in the file of the input frame, NAN if unknown
5534 timestamp expressed in seconds, NAN if the input timestamp is unknown
5537 Note that the @var{n}, @var{pos}, @var{t} variables are available only
5538 when evaluation is done @emph{per frame}, and will evaluate to NAN
5539 when @option{eval} is set to @samp{init}.
5541 Be aware that frames are taken from each input video in timestamp
5542 order, hence, if their initial timestamps differ, it is a good idea
5543 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
5544 have them begin in the same zero timestamp, as it does the example for
5545 the @var{movie} filter.
5547 You can chain together more overlays but you should test the
5548 efficiency of such approach.
5550 @subsection Commands
5552 This filter supports the following commands:
5556 Modify the x and y of the overlay input.
5557 The command accepts the same syntax of the corresponding option.
5559 If the specified expression is not valid, it is kept at its current
5563 @subsection Examples
5567 Draw the overlay at 10 pixels from the bottom right corner of the main
5570 overlay=main_w-overlay_w-10:main_h-overlay_h-10
5573 Using named options the example above becomes:
5575 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
5579 Insert a transparent PNG logo in the bottom left corner of the input,
5580 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
5582 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
5586 Insert 2 different transparent PNG logos (second logo on bottom
5587 right corner) using the @command{ffmpeg} tool:
5589 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
5593 Add a transparent color layer on top of the main video, @code{WxH}
5594 must specify the size of the main input to the overlay filter:
5596 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
5600 Play an original video and a filtered version (here with the deshake
5601 filter) side by side using the @command{ffplay} tool:
5603 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
5606 The above command is the same as:
5608 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
5612 Make a sliding overlay appearing from the left to the right top part of the
5613 screen starting since time 2:
5615 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
5619 Compose output by putting two input videos side to side:
5621 ffmpeg -i left.avi -i right.avi -filter_complex "
5622 nullsrc=size=200x100 [background];
5623 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
5624 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
5625 [background][left] overlay=shortest=1 [background+left];
5626 [background+left][right] overlay=shortest=1:x=100 [left+right]
5631 Chain several overlays in cascade:
5633 nullsrc=s=200x200 [bg];
5634 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
5635 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
5636 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
5637 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
5638 [in3] null, [mid2] overlay=100:100 [out0]
5645 Apply Overcomplete Wavelet denoiser.
5647 The filter accepts the following options:
5653 Larger depth values will denoise lower frequency components more, but
5654 slow down filtering.
5656 Must be an int in the range 8-16, default is @code{8}.
5658 @item luma_strength, ls
5661 Must be a double value in the range 0-1000, default is @code{1.0}.
5663 @item chroma_strength, cs
5664 Set chroma strength.
5666 Must be a double value in the range 0-1000, default is @code{1.0}.
5671 Add paddings to the input image, and place the original input at the
5672 given coordinates @var{x}, @var{y}.
5674 This filter accepts the following parameters:
5679 Specify an expression for the size of the output image with the
5680 paddings added. If the value for @var{width} or @var{height} is 0, the
5681 corresponding input size is used for the output.
5683 The @var{width} expression can reference the value set by the
5684 @var{height} expression, and vice versa.
5686 The default value of @var{width} and @var{height} is 0.
5690 Specify an expression for the offsets where to place the input image
5691 in the padded area with respect to the top/left border of the output
5694 The @var{x} expression can reference the value set by the @var{y}
5695 expression, and vice versa.
5697 The default value of @var{x} and @var{y} is 0.
5700 Specify the color of the padded area, it can be the name of a color
5701 (case insensitive match) or a 0xRRGGBB[AA] sequence.
5703 The default value of @var{color} is "black".
5706 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
5707 options are expressions containing the following constants:
5712 the input video width and height
5716 same as @var{in_w} and @var{in_h}
5720 the output width and height, that is the size of the padded area as
5721 specified by the @var{width} and @var{height} expressions
5725 same as @var{out_w} and @var{out_h}
5729 x and y offsets as specified by the @var{x} and @var{y}
5730 expressions, or NAN if not yet specified
5733 same as @var{iw} / @var{ih}
5736 input sample aspect ratio
5739 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5743 horizontal and vertical chroma subsample values. For example for the
5744 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5747 @subsection Examples
5751 Add paddings with color "violet" to the input video. Output video
5752 size is 640x480, the top-left corner of the input video is placed at
5755 pad=640:480:0:40:violet
5758 The example above is equivalent to the following command:
5760 pad=width=640:height=480:x=0:y=40:color=violet
5764 Pad the input to get an output with dimensions increased by 3/2,
5765 and put the input video at the center of the padded area:
5767 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
5771 Pad the input to get a squared output with size equal to the maximum
5772 value between the input width and height, and put the input video at
5773 the center of the padded area:
5775 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
5779 Pad the input to get a final w/h ratio of 16:9:
5781 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
5785 In case of anamorphic video, in order to set the output display aspect
5786 correctly, it is necessary to use @var{sar} in the expression,
5787 according to the relation:
5789 (ih * X / ih) * sar = output_dar
5790 X = output_dar / sar
5793 Thus the previous example needs to be modified to:
5795 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
5799 Double output size and put the input video in the bottom-right
5800 corner of the output padded area:
5802 pad="2*iw:2*ih:ow-iw:oh-ih"
5806 @section perspective
5808 Correct perspective of video not recorded perpendicular to the screen.
5810 A description of the accepted parameters follows.
5821 Set coordinates expression for top left, top right, bottom left and bottom right corners.
5822 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
5824 The expressions can use the following variables:
5829 the width and height of video frame.
5833 Set interpolation for perspective correction.
5835 It accepts the following values:
5841 Default value is @samp{linear}.
5844 @section pixdesctest
5846 Pixel format descriptor test filter, mainly useful for internal
5847 testing. The output video should be equal to the input video.
5851 format=monow, pixdesctest
5854 can be used to test the monowhite pixel format descriptor definition.
5858 Enable the specified chain of postprocessing subfilters using libpostproc. This
5859 library should be automatically selected with a GPL build (@code{--enable-gpl}).
5860 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
5861 Each subfilter and some options have a short and a long name that can be used
5862 interchangeably, i.e. dr/dering are the same.
5864 The filters accept the following options:
5868 Set postprocessing subfilters string.
5871 All subfilters share common options to determine their scope:
5875 Honor the quality commands for this subfilter.
5878 Do chrominance filtering, too (default).
5881 Do luminance filtering only (no chrominance).
5884 Do chrominance filtering only (no luminance).
5887 These options can be appended after the subfilter name, separated by a '|'.
5889 Available subfilters are:
5892 @item hb/hdeblock[|difference[|flatness]]
5893 Horizontal deblocking filter
5896 Difference factor where higher values mean more deblocking (default: @code{32}).
5898 Flatness threshold where lower values mean more deblocking (default: @code{39}).
5901 @item vb/vdeblock[|difference[|flatness]]
5902 Vertical deblocking filter
5905 Difference factor where higher values mean more deblocking (default: @code{32}).
5907 Flatness threshold where lower values mean more deblocking (default: @code{39}).
5910 @item ha/hadeblock[|difference[|flatness]]
5911 Accurate horizontal deblocking filter
5914 Difference factor where higher values mean more deblocking (default: @code{32}).
5916 Flatness threshold where lower values mean more deblocking (default: @code{39}).
5919 @item va/vadeblock[|difference[|flatness]]
5920 Accurate vertical deblocking filter
5923 Difference factor where higher values mean more deblocking (default: @code{32}).
5925 Flatness threshold where lower values mean more deblocking (default: @code{39}).
5929 The horizontal and vertical deblocking filters share the difference and
5930 flatness values so you cannot set different horizontal and vertical
5935 Experimental horizontal deblocking filter
5938 Experimental vertical deblocking filter
5943 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
5946 larger -> stronger filtering
5948 larger -> stronger filtering
5950 larger -> stronger filtering
5953 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
5956 Stretch luminance to @code{0-255}.
5959 @item lb/linblenddeint
5960 Linear blend deinterlacing filter that deinterlaces the given block by
5961 filtering all lines with a @code{(1 2 1)} filter.
5963 @item li/linipoldeint
5964 Linear interpolating deinterlacing filter that deinterlaces the given block by
5965 linearly interpolating every second line.
5967 @item ci/cubicipoldeint
5968 Cubic interpolating deinterlacing filter deinterlaces the given block by
5969 cubically interpolating every second line.
5971 @item md/mediandeint
5972 Median deinterlacing filter that deinterlaces the given block by applying a
5973 median filter to every second line.
5975 @item fd/ffmpegdeint
5976 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
5977 second line with a @code{(-1 4 2 4 -1)} filter.
5980 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
5981 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
5983 @item fq/forceQuant[|quantizer]
5984 Overrides the quantizer table from the input with the constant quantizer you
5992 Default pp filter combination (@code{hb|a,vb|a,dr|a})
5995 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
5998 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
6001 @subsection Examples
6005 Apply horizontal and vertical deblocking, deringing and automatic
6006 brightness/contrast:
6012 Apply default filters without brightness/contrast correction:
6018 Apply default filters and temporal denoiser:
6020 pp=default/tmpnoise|1|2|3
6024 Apply deblocking on luminance only, and switch vertical deblocking on or off
6025 automatically depending on available CPU time:
6033 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
6034 Ratio) between two input videos.
6036 This filter takes in input two input videos, the first input is
6037 considered the "main" source and is passed unchanged to the
6038 output. The second input is used as a "reference" video for computing
6041 Both video inputs must have the same resolution and pixel format for
6042 this filter to work correctly. Also it assumes that both inputs
6043 have the same number of frames, which are compared one by one.
6045 The obtained average PSNR is printed through the logging system.
6047 The filter stores the accumulated MSE (mean squared error) of each
6048 frame, and at the end of the processing it is averaged across all frames
6049 equally, and the following formula is applied to obtain the PSNR:
6052 PSNR = 10*log10(MAX^2/MSE)
6055 Where MAX is the average of the maximum values of each component of the
6058 The description of the accepted parameters follows.
6062 If specified the filter will use the named file to save the PSNR of
6063 each individual frame.
6066 The file printed if @var{stats_file} is selected, contains a sequence of
6067 key/value pairs of the form @var{key}:@var{value} for each compared
6070 A description of each shown parameter follows:
6074 sequential number of the input frame, starting from 1
6077 Mean Square Error pixel-by-pixel average difference of the compared
6078 frames, averaged over all the image components.
6080 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
6081 Mean Square Error pixel-by-pixel average difference of the compared
6082 frames for the component specified by the suffix.
6084 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
6085 Peak Signal to Noise ratio of the compared frames for the component
6086 specified by the suffix.
6091 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
6092 [main][ref] psnr="stats_file=stats.log" [out]
6095 On this example the input file being processed is compared with the
6096 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
6097 is stored in @file{stats.log}.
6101 Suppress a TV station logo, using an image file to determine which
6102 pixels comprise the logo. It works by filling in the pixels that
6103 comprise the logo with neighboring pixels.
6105 The filter accepts the following options:
6109 Set the filter bitmap file, which can be any image format supported by
6110 libavformat. The width and height of the image file must match those of the
6111 video stream being processed.
6114 Pixels in the provided bitmap image with a value of zero are not
6115 considered part of the logo, non-zero pixels are considered part of
6116 the logo. If you use white (255) for the logo and black (0) for the
6117 rest, you will be safe. For making the filter bitmap, it is
6118 recommended to take a screen capture of a black frame with the logo
6119 visible, and then using a threshold filter followed by the erode
6120 filter once or twice.
6122 If needed, little splotches can be fixed manually. Remember that if
6123 logo pixels are not covered, the filter quality will be much
6124 reduced. Marking too many pixels as part of the logo does not hurt as
6125 much, but it will increase the amount of blurring needed to cover over
6126 the image and will destroy more information than necessary, and extra
6127 pixels will slow things down on a large logo.
6131 Rotate video by an arbitrary angle expressed in radians.
6133 The filter accepts the following options:
6135 A description of the optional parameters follows.
6138 Set an expression for the angle by which to rotate the input video
6139 clockwise, expressed as a number of radians. A negative value will
6140 result in a counter-clockwise rotation. By default it is set to "0".
6142 This expression is evaluated for each frame.
6145 Set the output width expression, default value is "iw".
6146 This expression is evaluated just once during configuration.
6149 Set the output height expression, default value is "ih".
6150 This expression is evaluated just once during configuration.
6153 Enable bilinear interpolation if set to 1, a value of 0 disables
6154 it. Default value is 1.
6157 Set the color used to fill the output area not covered by the rotated
6158 image. If the special value "none" is selected then no background is
6159 printed (useful for example if the background is never shown). Default
6163 The expressions for the angle and the output size can contain the
6164 following constants and functions:
6168 sequential number of the input frame, starting from 0. It is always NAN
6169 before the first frame is filtered.
6172 time in seconds of the input frame, it is set to 0 when the filter is
6173 configured. It is always NAN before the first frame is filtered.
6177 horizontal and vertical chroma subsample values. For example for the
6178 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6182 the input video width and heigth
6186 the output width and heigth, that is the size of the padded area as
6187 specified by the @var{width} and @var{height} expressions
6191 the minimal width/height required for completely containing the input
6192 video rotated by @var{a} radians.
6194 These are only available when computing the @option{out_w} and
6195 @option{out_h} expressions.
6198 @subsection Examples
6202 Rotate the input by PI/6 radians clockwise:
6208 Rotate the input by PI/6 radians counter-clockwise:
6214 Apply a constant rotation with period T, starting from an angle of PI/3:
6216 rotate=PI/3+2*PI*t/T
6220 Make the input video rotation oscillating with a period of T
6221 seconds and an amplitude of A radians:
6223 rotate=A*sin(2*PI/T*t)
6227 Rotate the video, output size is choosen so that the whole rotating
6228 input video is always completely contained in the output:
6230 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
6234 Rotate the video, reduce the output size so that no background is ever
6237 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
6241 @subsection Commands
6243 The filter supports the following commands:
6247 Set the angle expression.
6248 The command accepts the same syntax of the corresponding option.
6250 If the specified expression is not valid, it is kept at its current
6256 Apply Shape Adaptive Blur.
6258 The filter accepts the following options:
6261 @item luma_radius, lr
6262 Set luma blur filter strength, must be a value in range 0.1-4.0, default
6263 value is 1.0. A greater value will result in a more blurred image, and
6264 in slower processing.
6266 @item luma_pre_filter_radius, lpfr
6267 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
6270 @item luma_strength, ls
6271 Set luma maximum difference between pixels to still be considered, must
6272 be a value in the 0.1-100.0 range, default value is 1.0.
6274 @item chroma_radius, cr
6275 Set chroma blur filter strength, must be a value in range 0.1-4.0. A
6276 greater value will result in a more blurred image, and in slower
6279 @item chroma_pre_filter_radius, cpfr
6280 Set chroma pre-filter radius, must be a value in the 0.1-2.0 range.
6282 @item chroma_strength, cs
6283 Set chroma maximum difference between pixels to still be considered,
6284 must be a value in the 0.1-100.0 range.
6287 Each chroma option value, if not explicitly specified, is set to the
6288 corresponding luma option value.
6292 Scale (resize) the input video, using the libswscale library.
6294 The scale filter forces the output display aspect ratio to be the same
6295 of the input, by changing the output sample aspect ratio.
6297 If the input image format is different from the format requested by
6298 the next filter, the scale filter will convert the input to the
6302 The filter accepts the following options:
6307 Set the output video dimension expression. Default value is the input
6310 If the value is 0, the input width is used for the output.
6312 If one of the values is -1, the scale filter will use a value that
6313 maintains the aspect ratio of the input image, calculated from the
6314 other specified dimension. If both of them are -1, the input size is
6317 See below for the list of accepted constants for use in the dimension
6321 Set the interlacing mode. It accepts the following values:
6325 Force interlaced aware scaling.
6328 Do not apply interlaced scaling.
6331 Select interlaced aware scaling depending on whether the source frames
6332 are flagged as interlaced or not.
6335 Default value is @samp{0}.
6338 Set libswscale scaling flags. If not explictly specified the filter
6339 applies a bilinear scaling algorithm.
6342 Set the video size, the value must be a valid abbreviation or in the
6343 form @var{width}x@var{height}.
6345 @item in_color_matrix
6346 @item out_color_matrix
6347 Set in/output YCbCr color space type.
6349 This allows the autodetected value to be overridden as well as allows forcing
6350 a specific value used for the output and encoder.
6352 If not specified, the color space type depends on the pixel format.
6358 Choose automatically.
6361 Format conforming to International Telecommunication Union (ITU)
6362 Recommendation BT.709.
6365 Set color space conforming to the United States Federal Communications
6366 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
6369 Set color space conforming to:
6373 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
6376 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
6379 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
6384 Set color space conforming to SMPTE ST 240:1999.
6389 Set in/output YCbCr sample range.
6391 This allows the autodetected value to be overridden as well as allows forcing
6392 a specific value used for the output and encoder. If not specified, the
6393 range depends on the pixel format. Possible values:
6397 Choose automatically.
6400 Set full range (0-255 in case of 8-bit luma).
6403 Set "MPEG" range (16-235 in case of 8-bit luma).
6407 Set the dithering algorithm
6411 Choose automatically.
6420 error diffusion dither
6423 @item force_original_aspect_ratio
6424 Enable decreasing or increasing output video width or height if necessary to
6425 keep the original aspect ratio. Possible values:
6429 Scale the video as specified and disable this feature.
6432 The output video dimensions will automatically be decreased if needed.
6435 The output video dimensions will automatically be increased if needed.
6439 One useful instance of this option is that when you know a specific device's
6440 maximum allowed resolution, you can use this to limit the output video to
6441 that, while retaining the aspect ratio. For example, device A allows
6442 1280x720 playback, and your video is 1920x800. Using this option (set it to
6443 decrease) and specifying 1280x720 to the command line makes the output
6446 Please note that this is a different thing than specifying -1 for @option{w}
6447 or @option{h}, you still need to specify the output resolution for this option
6452 The values of the @option{w} and @option{h} options are expressions
6453 containing the following constants:
6458 the input width and height
6462 same as @var{in_w} and @var{in_h}
6466 the output (scaled) width and height
6470 same as @var{out_w} and @var{out_h}
6473 same as @var{iw} / @var{ih}
6476 input sample aspect ratio
6479 input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
6483 horizontal and vertical chroma subsample values. For example for the
6484 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6487 @subsection Examples
6491 Scale the input video to a size of 200x100:
6496 This is equivalent to:
6507 Specify a size abbreviation for the output size:
6512 which can also be written as:
6518 Scale the input to 2x:
6524 The above is the same as:
6530 Scale the input to 2x with forced interlaced scaling:
6532 scale=2*iw:2*ih:interl=1
6536 Scale the input to half size:
6542 Increase the width, and set the height to the same size:
6548 Seek for Greek harmony:
6555 Increase the height, and set the width to 3/2 of the height:
6557 scale=w=3/2*oh:h=3/5*ih
6561 Increase the size, but make the size a multiple of the chroma
6564 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
6568 Increase the width to a maximum of 500 pixels, keep the same input
6571 scale=w='min(500\, iw*3/2):h=-1'
6575 @section separatefields
6577 The @code{separatefields} takes a frame-based video input and splits
6578 each frame into its components fields, producing a new half height clip
6579 with twice the frame rate and twice the frame count.
6581 This filter use field-dominance information in frame to decide which
6582 of each pair of fields to place first in the output.
6583 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
6585 @section setdar, setsar
6587 The @code{setdar} filter sets the Display Aspect Ratio for the filter
6590 This is done by changing the specified Sample (aka Pixel) Aspect
6591 Ratio, according to the following equation:
6593 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
6596 Keep in mind that the @code{setdar} filter does not modify the pixel
6597 dimensions of the video frame. Also the display aspect ratio set by
6598 this filter may be changed by later filters in the filterchain,
6599 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
6602 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
6603 the filter output video.
6605 Note that as a consequence of the application of this filter, the
6606 output display aspect ratio will change according to the equation
6609 Keep in mind that the sample aspect ratio set by the @code{setsar}
6610 filter may be changed by later filters in the filterchain, e.g. if
6611 another "setsar" or a "setdar" filter is applied.
6613 The filters accept the following options:
6616 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
6617 Set the aspect ratio used by the filter.
6619 The parameter can be a floating point number string, an expression, or
6620 a string of the form @var{num}:@var{den}, where @var{num} and
6621 @var{den} are the numerator and denominator of the aspect ratio. If
6622 the parameter is not specified, it is assumed the value "0".
6623 In case the form "@var{num}:@var{den}" is used, the @code{:} character
6627 Set the maximum integer value to use for expressing numerator and
6628 denominator when reducing the expressed aspect ratio to a rational.
6629 Default value is @code{100}.
6633 @subsection Examples
6638 To change the display aspect ratio to 16:9, specify one of the following:
6646 To change the sample aspect ratio to 10:11, specify:
6652 To set a display aspect ratio of 16:9, and specify a maximum integer value of
6653 1000 in the aspect ratio reduction, use the command:
6655 setdar=ratio=16/9:max=1000
6663 Force field for the output video frame.
6665 The @code{setfield} filter marks the interlace type field for the
6666 output frames. It does not change the input frame, but only sets the
6667 corresponding property, which affects how the frame is treated by
6668 following filters (e.g. @code{fieldorder} or @code{yadif}).
6670 The filter accepts the following options:
6675 Available values are:
6679 Keep the same field property.
6682 Mark the frame as bottom-field-first.
6685 Mark the frame as top-field-first.
6688 Mark the frame as progressive.
6694 Show a line containing various information for each input video frame.
6695 The input video is not modified.
6697 The shown line contains a sequence of key/value pairs of the form
6698 @var{key}:@var{value}.
6700 A description of each shown parameter follows:
6704 sequential number of the input frame, starting from 0
6707 Presentation TimeStamp of the input frame, expressed as a number of
6708 time base units. The time base unit depends on the filter input pad.
6711 Presentation TimeStamp of the input frame, expressed as a number of
6715 position of the frame in the input stream, -1 if this information in
6716 unavailable and/or meaningless (for example in case of synthetic video)
6722 sample aspect ratio of the input frame, expressed in the form
6726 size of the input frame, expressed in the form
6727 @var{width}x@var{height}
6730 interlaced mode ("P" for "progressive", "T" for top field first, "B"
6731 for bottom field first)
6734 1 if the frame is a key frame, 0 otherwise
6737 picture type of the input frame ("I" for an I-frame, "P" for a
6738 P-frame, "B" for a B-frame, "?" for unknown type).
6739 Check also the documentation of the @code{AVPictureType} enum and of
6740 the @code{av_get_picture_type_char} function defined in
6741 @file{libavutil/avutil.h}.
6744 Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame
6746 @item plane_checksum
6747 Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
6748 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]"
6754 Blur the input video without impacting the outlines.
6756 The filter accepts the following options:
6759 @item luma_radius, lr
6760 Set the luma radius. The option value must be a float number in
6761 the range [0.1,5.0] that specifies the variance of the gaussian filter
6762 used to blur the image (slower if larger). Default value is 1.0.
6764 @item luma_strength, ls
6765 Set the luma strength. The option value must be a float number
6766 in the range [-1.0,1.0] that configures the blurring. A value included
6767 in [0.0,1.0] will blur the image whereas a value included in
6768 [-1.0,0.0] will sharpen the image. Default value is 1.0.
6770 @item luma_threshold, lt
6771 Set the luma threshold used as a coefficient to determine
6772 whether a pixel should be blurred or not. The option value must be an
6773 integer in the range [-30,30]. A value of 0 will filter all the image,
6774 a value included in [0,30] will filter flat areas and a value included
6775 in [-30,0] will filter edges. Default value is 0.
6777 @item chroma_radius, cr
6778 Set the chroma radius. The option value must be a float number in
6779 the range [0.1,5.0] that specifies the variance of the gaussian filter
6780 used to blur the image (slower if larger). Default value is 1.0.
6782 @item chroma_strength, cs
6783 Set the chroma strength. The option value must be a float number
6784 in the range [-1.0,1.0] that configures the blurring. A value included
6785 in [0.0,1.0] will blur the image whereas a value included in
6786 [-1.0,0.0] will sharpen the image. Default value is 1.0.
6788 @item chroma_threshold, ct
6789 Set the chroma threshold used as a coefficient to determine
6790 whether a pixel should be blurred or not. The option value must be an
6791 integer in the range [-30,30]. A value of 0 will filter all the image,
6792 a value included in [0,30] will filter flat areas and a value included
6793 in [-30,0] will filter edges. Default value is 0.
6796 If a chroma option is not explicitly set, the corresponding luma value
6801 Convert between different stereoscopic image formats.
6803 The filters accept the following options:
6807 Set stereoscopic image format of input.
6809 Available values for input image formats are:
6812 side by side parallel (left eye left, right eye right)
6815 side by side crosseye (right eye left, left eye right)
6818 side by side parallel with half width resolution
6819 (left eye left, right eye right)
6822 side by side crosseye with half width resolution
6823 (right eye left, left eye right)
6826 above-below (left eye above, right eye below)
6829 above-below (right eye above, left eye below)
6832 above-below with half height resolution
6833 (left eye above, right eye below)
6836 above-below with half height resolution
6837 (right eye above, left eye below)
6840 alternating frames (left eye first, right eye second)
6843 alternating frames (right eye first, left eye second)
6845 Default value is @samp{sbsl}.
6849 Set stereoscopic image format of output.
6851 Available values for output image formats are all the input formats as well as:
6854 anaglyph red/blue gray
6855 (red filter on left eye, blue filter on right eye)
6858 anaglyph red/green gray
6859 (red filter on left eye, green filter on right eye)
6862 anaglyph red/cyan gray
6863 (red filter on left eye, cyan filter on right eye)
6866 anaglyph red/cyan half colored
6867 (red filter on left eye, cyan filter on right eye)
6870 anaglyph red/cyan color
6871 (red filter on left eye, cyan filter on right eye)
6874 anaglyph red/cyan color optimized with the least squares projection of dubois
6875 (red filter on left eye, cyan filter on right eye)
6878 anaglyph green/magenta gray
6879 (green filter on left eye, magenta filter on right eye)
6882 anaglyph green/magenta half colored
6883 (green filter on left eye, magenta filter on right eye)
6886 anaglyph green/magenta colored
6887 (green filter on left eye, magenta filter on right eye)
6890 anaglyph green/magenta color optimized with the least squares projection of dubois
6891 (green filter on left eye, magenta filter on right eye)
6894 anaglyph yellow/blue gray
6895 (yellow filter on left eye, blue filter on right eye)
6898 anaglyph yellow/blue half colored
6899 (yellow filter on left eye, blue filter on right eye)
6902 anaglyph yellow/blue colored
6903 (yellow filter on left eye, blue filter on right eye)
6906 anaglyph yellow/blue color optimized with the least squares projection of dubois
6907 (yellow filter on left eye, blue filter on right eye)
6910 interleaved rows (left eye has top row, right eye starts on next row)
6913 interleaved rows (right eye has top row, left eye starts on next row)
6916 mono output (left eye only)
6919 mono output (right eye only)
6922 Default value is @samp{arcd}.
6925 @subsection Examples
6929 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
6935 Convert input video from above bellow (left eye above, right eye below) to side by side crosseye.
6943 Apply a simple postprocessing filter that compresses and decompresses the image
6944 at several (or - in the case of @option{quality} level @code{6} - all) shifts
6945 and average the results.
6947 The filter accepts the following options:
6951 Set quality. This option defines the number of levels for averaging. It accepts
6952 an integer in the range 0-6. If set to @code{0}, the filter will have no
6953 effect. A value of @code{6} means the higher quality. For each increment of
6954 that value the speed drops by a factor of approximately 2. Default value is
6958 Force a constant quantization parameter. If not set, the filter will use the QP
6959 from the video stream (if available).
6962 Set thresholding mode. Available modes are:
6966 Set hard thresholding (default).
6968 Set soft thresholding (better de-ringing effect, but likely blurrier).
6972 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
6973 option may cause flicker since the B-Frames have often larger QP. Default is
6974 @code{0} (not enabled).
6980 Draw subtitles on top of input video using the libass library.
6982 To enable compilation of this filter you need to configure FFmpeg with
6983 @code{--enable-libass}. This filter also requires a build with libavcodec and
6984 libavformat to convert the passed subtitles file to ASS (Advanced Substation
6985 Alpha) subtitles format.
6987 The filter accepts the following options:
6991 Set the filename of the subtitle file to read. It must be specified.
6994 Specify the size of the original video, the video for which the ASS file
6995 was composed. Due to a misdesign in ASS aspect ratio arithmetic, this is
6996 necessary to correctly scale the fonts if the aspect ratio has been changed.
6999 Set subtitles input character encoding. @code{subtitles} filter only. Only
7000 useful if not UTF-8.
7003 If the first key is not specified, it is assumed that the first value
7004 specifies the @option{filename}.
7006 For example, to render the file @file{sub.srt} on top of the input
7007 video, use the command:
7012 which is equivalent to:
7014 subtitles=filename=sub.srt
7019 Scale the input by 2x and smooth using the Super2xSaI (Scale and
7020 Interpolate) pixel art scaling algorithm.
7022 Useful for enlarging pixel art images without reducing sharpness.
7029 Apply telecine process to the video.
7031 This filter accepts the following options:
7040 The default value is @code{top}.
7044 A string of numbers representing the pulldown pattern you wish to apply.
7045 The default value is @code{23}.
7049 Some typical patterns:
7054 24p: 2332 (preferred)
7061 24p: 222222222223 ("Euro pulldown")
7067 Select the most representative frame in a given sequence of consecutive frames.
7069 The filter accepts the following options:
7073 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
7074 will pick one of them, and then handle the next batch of @var{n} frames until
7075 the end. Default is @code{100}.
7078 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
7079 value will result in a higher memory usage, so a high value is not recommended.
7081 @subsection Examples
7085 Extract one picture each 50 frames:
7091 Complete example of a thumbnail creation with @command{ffmpeg}:
7093 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
7099 Tile several successive frames together.
7101 The filter accepts the following options:
7106 Set the grid size (i.e. the number of lines and columns) in the form
7110 Set the maximum number of frames to render in the given area. It must be less
7111 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
7112 the area will be used.
7115 Set the outer border margin in pixels.
7118 Set the inner border thickness (i.e. the number of pixels between frames). For
7119 more advanced padding options (such as having different values for the edges),
7120 refer to the pad video filter.
7123 Specify the color of the unused area, it can be the name of a color
7124 (case insensitive match) or a 0xRRGGBB[AA] sequence.
7125 The default value of @var{color} is "black".
7128 @subsection Examples
7132 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
7134 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
7136 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
7137 duplicating each output frame to accomodate the originally detected frame
7141 Display @code{5} pictures in an area of @code{3x2} frames,
7142 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
7143 mixed flat and named options:
7145 tile=3x2:nb_frames=5:padding=7:margin=2
7151 Perform various types of temporal field interlacing.
7153 Frames are counted starting from 1, so the first input frame is
7156 The filter accepts the following options:
7161 Specify the mode of the interlacing. This option can also be specified
7162 as a value alone. See below for a list of values for this option.
7164 Available values are:
7168 Move odd frames into the upper field, even into the lower field,
7169 generating a double height frame at half frame rate.
7172 Only output even frames, odd frames are dropped, generating a frame with
7173 unchanged height at half frame rate.
7176 Only output odd frames, even frames are dropped, generating a frame with
7177 unchanged height at half frame rate.
7180 Expand each frame to full height, but pad alternate lines with black,
7181 generating a frame with double height at the same input frame rate.
7183 @item interleave_top, 4
7184 Interleave the upper field from odd frames with the lower field from
7185 even frames, generating a frame with unchanged height at half frame rate.
7187 @item interleave_bottom, 5
7188 Interleave the lower field from odd frames with the upper field from
7189 even frames, generating a frame with unchanged height at half frame rate.
7191 @item interlacex2, 6
7192 Double frame rate with unchanged height. Frames are inserted each
7193 containing the second temporal field from the previous input frame and
7194 the first temporal field from the next input frame. This mode relies on
7195 the top_field_first flag. Useful for interlaced video displays with no
7196 field synchronisation.
7199 Numeric values are deprecated but are accepted for backward
7200 compatibility reasons.
7202 Default mode is @code{merge}.
7205 Specify flags influencing the filter process.
7207 Available value for @var{flags} is:
7210 @item low_pass_filter, vlfp
7211 Enable vertical low-pass filtering in the filter.
7212 Vertical low-pass filtering is required when creating an interlaced
7213 destination from a progressive source which contains high-frequency
7214 vertical detail. Filtering will reduce interlace 'twitter' and Moire
7217 Vertical low-pass filtering can only be enabled for @option{mode}
7218 @var{interleave_top} and @var{interleave_bottom}.
7225 Transpose rows with columns in the input video and optionally flip it.
7227 This filter accepts the following options:
7232 Specify the transposition direction.
7234 Can assume the following values:
7236 @item 0, 4, cclock_flip
7237 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
7245 Rotate by 90 degrees clockwise, that is:
7253 Rotate by 90 degrees counterclockwise, that is:
7260 @item 3, 7, clock_flip
7261 Rotate by 90 degrees clockwise and vertically flip, that is:
7269 For values between 4-7, the transposition is only done if the input
7270 video geometry is portrait and not landscape. These values are
7271 deprecated, the @code{passthrough} option should be used instead.
7273 Numerical values are deprecated, and should be dropped in favor of
7277 Do not apply the transposition if the input geometry matches the one
7278 specified by the specified value. It accepts the following values:
7281 Always apply transposition.
7283 Preserve portrait geometry (when @var{height} >= @var{width}).
7285 Preserve landscape geometry (when @var{width} >= @var{height}).
7288 Default value is @code{none}.
7291 For example to rotate by 90 degrees clockwise and preserve portrait
7294 transpose=dir=1:passthrough=portrait
7297 The command above can also be specified as:
7299 transpose=1:portrait
7303 Trim the input so that the output contains one continuous subpart of the input.
7305 This filter accepts the following options:
7308 Specify time of the start of the kept section, i.e. the frame with the
7309 timestamp @var{start} will be the first frame in the output.
7312 Specify time of the first frame that will be dropped, i.e. the frame
7313 immediately preceding the one with the timestamp @var{end} will be the last
7314 frame in the output.
7317 Same as @var{start}, except this option sets the start timestamp in timebase
7318 units instead of seconds.
7321 Same as @var{end}, except this option sets the end timestamp in timebase units
7325 Specify maximum duration of the output.
7328 Number of the first frame that should be passed to output.
7331 Number of the first frame that should be dropped.
7334 @option{start}, @option{end}, @option{duration} are expressed as time
7335 duration specifications, check the "Time duration" section in the
7336 ffmpeg-utils manual.
7338 Note that the first two sets of the start/end options and the @option{duration}
7339 option look at the frame timestamp, while the _frame variants simply count the
7340 frames that pass through the filter. Also note that this filter does not modify
7341 the timestamps. If you wish that the output timestamps start at zero, insert a
7342 setpts filter after the trim filter.
7344 If multiple start or end options are set, this filter tries to be greedy and
7345 keep all the frames that match at least one of the specified constraints. To keep
7346 only the part that matches all the constraints at once, chain multiple trim
7349 The defaults are such that all the input is kept. So it is possible to set e.g.
7350 just the end values to keep everything before the specified time.
7355 drop everything except the second minute of input
7357 ffmpeg -i INPUT -vf trim=60:120
7361 keep only the first second
7363 ffmpeg -i INPUT -vf trim=duration=1
7371 Sharpen or blur the input video.
7373 It accepts the following parameters:
7376 @item luma_msize_x, lx
7377 Set the luma matrix horizontal size. It must be an odd integer between
7378 3 and 63, default value is 5.
7380 @item luma_msize_y, ly
7381 Set the luma matrix vertical size. It must be an odd integer between 3
7382 and 63, default value is 5.
7384 @item luma_amount, la
7385 Set the luma effect strength. It can be a float number, reasonable
7386 values lay between -1.5 and 1.5.
7388 Negative values will blur the input video, while positive values will
7389 sharpen it, a value of zero will disable the effect.
7391 Default value is 1.0.
7393 @item chroma_msize_x, cx
7394 Set the chroma matrix horizontal size. It must be an odd integer
7395 between 3 and 63, default value is 5.
7397 @item chroma_msize_y, cy
7398 Set the chroma matrix vertical size. It must be an odd integer
7399 between 3 and 63, default value is 5.
7401 @item chroma_amount, ca
7402 Set the chroma effect strength. It can be a float number, reasonable
7403 values lay between -1.5 and 1.5.
7405 Negative values will blur the input video, while positive values will
7406 sharpen it, a value of zero will disable the effect.
7408 Default value is 0.0.
7411 If set to 1, specify using OpenCL capabilities, only available if
7412 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
7416 All parameters are optional and default to the equivalent of the
7417 string '5:5:1.0:5:5:0.0'.
7419 @subsection Examples
7423 Apply strong luma sharpen effect:
7425 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
7429 Apply strong blur of both luma and chroma parameters:
7431 unsharp=7:7:-2:7:7:-2
7435 @anchor{vidstabdetect}
7436 @section vidstabdetect
7438 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
7439 @ref{vidstabtransform} for pass 2.
7441 This filter generates a file with relative translation and rotation
7442 transform information about subsequent frames, which is then used by
7443 the @ref{vidstabtransform} filter.
7445 To enable compilation of this filter you need to configure FFmpeg with
7446 @code{--enable-libvidstab}.
7448 This filter accepts the following options:
7452 Set the path to the file used to write the transforms information.
7453 Default value is @file{transforms.trf}.
7456 Set how shaky the video is and how quick the camera is. It accepts an
7457 integer in the range 1-10, a value of 1 means little shakiness, a
7458 value of 10 means strong shakiness. Default value is 5.
7461 Set the accuracy of the detection process. It must be a value in the
7462 range 1-15. A value of 1 means low accuracy, a value of 15 means high
7463 accuracy. Default value is 9.
7466 Set stepsize of the search process. The region around minimum is
7467 scanned with 1 pixel resolution. Default value is 6.
7470 Set minimum contrast. Below this value a local measurement field is
7471 discarded. Must be a floating point value in the range 0-1. Default
7475 Set reference frame number for tripod mode.
7477 If enabled, the motion of the frames is compared to a reference frame
7478 in the filtered stream, identified by the specified number. The idea
7479 is to compensate all movements in a more-or-less static scene and keep
7480 the camera view absolutely still.
7482 If set to 0, it is disabled. The frames are counted starting from 1.
7485 Show fields and transforms in the resulting frames. It accepts an
7486 integer in the range 0-2. Default value is 0, which disables any
7490 @subsection Examples
7500 Analyze strongly shaky movie and put the results in file
7501 @file{mytransforms.trf}:
7503 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
7507 Visualize the result of internal transformations in the resulting
7510 vidstabdetect=show=1
7514 Analyze a video with medium shakiness using @command{ffmpeg}:
7516 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
7520 @anchor{vidstabtransform}
7521 @section vidstabtransform
7523 Video stabilization/deshaking: pass 2 of 2,
7524 see @ref{vidstabdetect} for pass 1.
7526 Read a file with transform information for each frame and
7527 apply/compensate them. Together with the @ref{vidstabdetect}
7528 filter this can be used to deshake videos. See also
7529 @url{http://public.hronopik.de/vid.stab}. It is important to also use
7530 the unsharp filter, see below.
7532 To enable compilation of this filter you need to configure FFmpeg with
7533 @code{--enable-libvidstab}.
7535 This filter accepts the following options:
7540 path to the file used to read the transforms (default: @file{transforms.trf})
7543 number of frames (value*2 + 1) used for lowpass filtering the camera movements
7544 (default: 10). For example a number of 10 means that 21 frames are used
7545 (10 in the past and 10 in the future) to smoothen the motion in the
7546 video. A larger values leads to a smoother video, but limits the
7547 acceleration of the camera (pan/tilt movements).
7550 maximal number of pixels to translate frames (default: -1 no limit)
7553 maximal angle in radians (degree*PI/180) to rotate frames (default: -1
7557 How to deal with borders that may be visible due to movement
7558 compensation. Available values are:
7562 keep image information from previous frame (default)
7564 fill the border black
7570 keep transforms normal (default)
7577 consider transforms as
7582 relative to previous frame (default)
7587 percentage to zoom (default: 0)
7596 if 1 then optimal zoom value is determined (default).
7597 Optimal zoom means no (or only little) border should be visible.
7598 Note that the value given at zoom is added to the one calculated
7602 type of interpolation
7604 Available values are:
7609 linear only horizontal
7611 linear in both directions (default)
7613 cubic in both directions (slow)
7617 virtual tripod mode means that the video is stabilized such that the
7618 camera stays stationary. Use also @code{tripod} option of
7619 @ref{vidstabdetect}.
7624 virtual tripod mode: equivalent to @code{relative=0:smoothing=0}
7629 @subsection Examples
7633 typical call with default default values:
7634 (note the unsharp filter which is always recommended)
7636 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
7640 zoom in a bit more and load transform data from a given file
7642 vidstabtransform=zoom=5:input="mytransforms.trf"
7646 smoothen the video even more
7648 vidstabtransform=smoothing=30
7655 Flip the input video vertically.
7657 For example, to vertically flip a video with @command{ffmpeg}:
7659 ffmpeg -i in.avi -vf "vflip" out.avi
7664 Make or reverse a natural vignetting effect.
7666 The filter accepts the following options:
7670 Set lens angle expression as a number of radians.
7672 The value is clipped in the @code{[0,PI/2]} range.
7674 Default value: @code{"PI/5"}
7678 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
7682 Set forward/backward mode.
7684 Available modes are:
7687 The larger the distance from the central point, the darker the image becomes.
7690 The larger the distance from the central point, the brighter the image becomes.
7691 This can be used to reverse a vignette effect, though there is no automatic
7692 detection to extract the lens @option{angle} and other settings (yet). It can
7693 also be used to create a burning effect.
7696 Default value is @samp{forward}.
7699 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
7701 It accepts the following values:
7704 Evaluate expressions only once during the filter initialization.
7707 Evaluate expressions for each incoming frame. This is way slower than the
7708 @samp{init} mode since it requires all the scalers to be re-computed, but it
7709 allows advanced dynamic expressions.
7712 Default value is @samp{init}.
7715 Set dithering to reduce the circular banding effects. Default is @code{1}
7719 Set vignette aspect. This setting allows to adjust the shape of the vignette.
7720 Setting this value to the SAR of the input will make a rectangular vignetting
7721 following the dimensions of the video.
7723 Default is @code{1/1}.
7726 @subsection Expressions
7728 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
7729 following parameters.
7734 input width and height
7737 the number of input frame, starting from 0
7740 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
7741 @var{TB} units, NAN if undefined
7744 frame rate of the input video, NAN if the input frame rate is unknown
7747 the PTS (Presentation TimeStamp) of the filtered video frame,
7748 expressed in seconds, NAN if undefined
7751 time base of the input video
7755 @subsection Examples
7759 Apply simple strong vignetting effect:
7765 Make a flickering vignetting:
7767 vignette='PI/4+random(1)*PI/50':eval=frame
7775 Deinterlace the input video ("yadif" means "yet another deinterlacing
7778 This filter accepts the following options:
7784 The interlacing mode to adopt, accepts one of the following values:
7788 output 1 frame for each frame
7790 output 1 frame for each field
7791 @item 2, send_frame_nospatial
7792 like @code{send_frame} but skip spatial interlacing check
7793 @item 3, send_field_nospatial
7794 like @code{send_field} but skip spatial interlacing check
7797 Default value is @code{send_frame}.
7800 The picture field parity assumed for the input interlaced video, accepts one of
7801 the following values:
7805 assume top field first
7807 assume bottom field first
7809 enable automatic detection
7812 Default value is @code{auto}.
7813 If interlacing is unknown or decoder does not export this information,
7814 top field first will be assumed.
7817 Specify which frames to deinterlace. Accept one of the following
7822 deinterlace all frames
7824 only deinterlace frames marked as interlaced
7827 Default value is @code{all}.
7830 @c man end VIDEO FILTERS
7832 @chapter Video Sources
7833 @c man begin VIDEO SOURCES
7835 Below is a description of the currently available video sources.
7839 Buffer video frames, and make them available to the filter chain.
7841 This source is mainly intended for a programmatic use, in particular
7842 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
7844 This source accepts the following options:
7849 Specify the size (width and height) of the buffered video frames.
7858 A string representing the pixel format of the buffered video frames.
7859 It may be a number corresponding to a pixel format, or a pixel format
7863 Specify the timebase assumed by the timestamps of the buffered frames.
7866 Specify the frame rate expected for the video stream.
7868 @item pixel_aspect, sar
7869 Specify the sample aspect ratio assumed by the video frames.
7872 Specify the optional parameters to be used for the scale filter which
7873 is automatically inserted when an input change is detected in the
7874 input size or format.
7879 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
7882 will instruct the source to accept video frames with size 320x240 and
7883 with format "yuv410p", assuming 1/24 as the timestamps timebase and
7884 square pixels (1:1 sample aspect ratio).
7885 Since the pixel format with name "yuv410p" corresponds to the number 6
7886 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
7887 this example corresponds to:
7889 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
7892 Alternatively, the options can be specified as a flat string, but this
7893 syntax is deprecated:
7895 @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}]
7899 Create a pattern generated by an elementary cellular automaton.
7901 The initial state of the cellular automaton can be defined through the
7902 @option{filename}, and @option{pattern} options. If such options are
7903 not specified an initial state is created randomly.
7905 At each new frame a new row in the video is filled with the result of
7906 the cellular automaton next generation. The behavior when the whole
7907 frame is filled is defined by the @option{scroll} option.
7909 This source accepts the following options:
7913 Read the initial cellular automaton state, i.e. the starting row, from
7915 In the file, each non-whitespace character is considered an alive
7916 cell, a newline will terminate the row, and further characters in the
7917 file will be ignored.
7920 Read the initial cellular automaton state, i.e. the starting row, from
7921 the specified string.
7923 Each non-whitespace character in the string is considered an alive
7924 cell, a newline will terminate the row, and further characters in the
7925 string will be ignored.
7928 Set the video rate, that is the number of frames generated per second.
7931 @item random_fill_ratio, ratio
7932 Set the random fill ratio for the initial cellular automaton row. It
7933 is a floating point number value ranging from 0 to 1, defaults to
7936 This option is ignored when a file or a pattern is specified.
7938 @item random_seed, seed
7939 Set the seed for filling randomly the initial row, must be an integer
7940 included between 0 and UINT32_MAX. If not specified, or if explicitly
7941 set to -1, the filter will try to use a good random seed on a best
7945 Set the cellular automaton rule, it is a number ranging from 0 to 255.
7946 Default value is 110.
7949 Set the size of the output video.
7951 If @option{filename} or @option{pattern} is specified, the size is set
7952 by default to the width of the specified initial state row, and the
7953 height is set to @var{width} * PHI.
7955 If @option{size} is set, it must contain the width of the specified
7956 pattern string, and the specified pattern will be centered in the
7959 If a filename or a pattern string is not specified, the size value
7960 defaults to "320x518" (used for a randomly generated initial state).
7963 If set to 1, scroll the output upward when all the rows in the output
7964 have been already filled. If set to 0, the new generated row will be
7965 written over the top row just after the bottom row is filled.
7968 @item start_full, full
7969 If set to 1, completely fill the output with generated rows before
7970 outputting the first frame.
7971 This is the default behavior, for disabling set the value to 0.
7974 If set to 1, stitch the left and right row edges together.
7975 This is the default behavior, for disabling set the value to 0.
7978 @subsection Examples
7982 Read the initial state from @file{pattern}, and specify an output of
7985 cellauto=f=pattern:s=200x400
7989 Generate a random initial row with a width of 200 cells, with a fill
7992 cellauto=ratio=2/3:s=200x200
7996 Create a pattern generated by rule 18 starting by a single alive cell
7997 centered on an initial row with width 100:
7999 cellauto=p=@@:s=100x400:full=0:rule=18
8003 Specify a more elaborated initial pattern:
8005 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
8012 Generate a Mandelbrot set fractal, and progressively zoom towards the
8013 point specified with @var{start_x} and @var{start_y}.
8015 This source accepts the following options:
8020 Set the terminal pts value. Default value is 400.
8023 Set the terminal scale value.
8024 Must be a floating point value. Default value is 0.3.
8027 Set the inner coloring mode, that is the algorithm used to draw the
8028 Mandelbrot fractal internal region.
8030 It shall assume one of the following values:
8035 Show time until convergence.
8037 Set color based on point closest to the origin of the iterations.
8042 Default value is @var{mincol}.
8045 Set the bailout value. Default value is 10.0.
8048 Set the maximum of iterations performed by the rendering
8049 algorithm. Default value is 7189.
8052 Set outer coloring mode.
8053 It shall assume one of following values:
8055 @item iteration_count
8056 Set iteration cound mode.
8057 @item normalized_iteration_count
8058 set normalized iteration count mode.
8060 Default value is @var{normalized_iteration_count}.
8063 Set frame rate, expressed as number of frames per second. Default
8067 Set frame size. Default value is "640x480".
8070 Set the initial scale value. Default value is 3.0.
8073 Set the initial x position. Must be a floating point value between
8074 -100 and 100. Default value is -0.743643887037158704752191506114774.
8077 Set the initial y position. Must be a floating point value between
8078 -100 and 100. Default value is -0.131825904205311970493132056385139.
8083 Generate various test patterns, as generated by the MPlayer test filter.
8085 The size of the generated video is fixed, and is 256x256.
8086 This source is useful in particular for testing encoding features.
8088 This source accepts the following options:
8093 Specify the frame rate of the sourced video, as the number of frames
8094 generated per second. It has to be a string in the format
8095 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
8096 number or a valid video frame rate abbreviation. The default value is
8100 Set the video duration of the sourced video. The accepted syntax is:
8105 See also the function @code{av_parse_time()}.
8107 If not specified, or the expressed duration is negative, the video is
8108 supposed to be generated forever.
8112 Set the number or the name of the test to perform. Supported tests are:
8127 Default value is "all", which will cycle through the list of all tests.
8130 For example the following:
8135 will generate a "dc_luma" test pattern.
8139 Provide a frei0r source.
8141 To enable compilation of this filter you need to install the frei0r
8142 header and configure FFmpeg with @code{--enable-frei0r}.
8144 This source accepts the following options:
8149 The size of the video to generate, may be a string of the form
8150 @var{width}x@var{height} or a frame size abbreviation.
8153 Framerate of the generated video, may be a string of the form
8154 @var{num}/@var{den} or a frame rate abbreviation.
8157 The name to the frei0r source to load. For more information regarding frei0r and
8158 how to set the parameters read the section @ref{frei0r} in the description of
8162 A '|'-separated list of parameters to pass to the frei0r source.
8166 For example, to generate a frei0r partik0l source with size 200x200
8167 and frame rate 10 which is overlayed on the overlay filter main input:
8169 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
8174 Generate a life pattern.
8176 This source is based on a generalization of John Conway's life game.
8178 The sourced input represents a life grid, each pixel represents a cell
8179 which can be in one of two possible states, alive or dead. Every cell
8180 interacts with its eight neighbours, which are the cells that are
8181 horizontally, vertically, or diagonally adjacent.
8183 At each interaction the grid evolves according to the adopted rule,
8184 which specifies the number of neighbor alive cells which will make a
8185 cell stay alive or born. The @option{rule} option allows to specify
8188 This source accepts the following options:
8192 Set the file from which to read the initial grid state. In the file,
8193 each non-whitespace character is considered an alive cell, and newline
8194 is used to delimit the end of each row.
8196 If this option is not specified, the initial grid is generated
8200 Set the video rate, that is the number of frames generated per second.
8203 @item random_fill_ratio, ratio
8204 Set the random fill ratio for the initial random grid. It is a
8205 floating point number value ranging from 0 to 1, defaults to 1/PHI.
8206 It is ignored when a file is specified.
8208 @item random_seed, seed
8209 Set the seed for filling the initial random grid, must be an integer
8210 included between 0 and UINT32_MAX. If not specified, or if explicitly
8211 set to -1, the filter will try to use a good random seed on a best
8217 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
8218 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
8219 @var{NS} specifies the number of alive neighbor cells which make a
8220 live cell stay alive, and @var{NB} the number of alive neighbor cells
8221 which make a dead cell to become alive (i.e. to "born").
8222 "s" and "b" can be used in place of "S" and "B", respectively.
8224 Alternatively a rule can be specified by an 18-bits integer. The 9
8225 high order bits are used to encode the next cell state if it is alive
8226 for each number of neighbor alive cells, the low order bits specify
8227 the rule for "borning" new cells. Higher order bits encode for an
8228 higher number of neighbor cells.
8229 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
8230 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
8232 Default value is "S23/B3", which is the original Conway's game of life
8233 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
8234 cells, and will born a new cell if there are three alive cells around
8238 Set the size of the output video.
8240 If @option{filename} is specified, the size is set by default to the
8241 same size of the input file. If @option{size} is set, it must contain
8242 the size specified in the input file, and the initial grid defined in
8243 that file is centered in the larger resulting area.
8245 If a filename is not specified, the size value defaults to "320x240"
8246 (used for a randomly generated initial grid).
8249 If set to 1, stitch the left and right grid edges together, and the
8250 top and bottom edges also. Defaults to 1.
8253 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
8254 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
8255 value from 0 to 255.
8258 Set the color of living (or new born) cells.
8261 Set the color of dead cells. If @option{mold} is set, this is the first color
8262 used to represent a dead cell.
8265 Set mold color, for definitely dead and moldy cells.
8268 @subsection Examples
8272 Read a grid from @file{pattern}, and center it on a grid of size
8275 life=f=pattern:s=300x300
8279 Generate a random grid of size 200x200, with a fill ratio of 2/3:
8281 life=ratio=2/3:s=200x200
8285 Specify a custom rule for evolving a randomly generated grid:
8291 Full example with slow death effect (mold) using @command{ffplay}:
8293 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
8298 @anchor{haldclutsrc}
8302 @anchor{smptehdbars}
8304 @section color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc
8306 The @code{color} source provides an uniformly colored input.
8308 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
8309 @ref{haldclut} filter.
8311 The @code{nullsrc} source returns unprocessed video frames. It is
8312 mainly useful to be employed in analysis / debugging tools, or as the
8313 source for filters which ignore the input data.
8315 The @code{rgbtestsrc} source generates an RGB test pattern useful for
8316 detecting RGB vs BGR issues. You should see a red, green and blue
8317 stripe from top to bottom.
8319 The @code{smptebars} source generates a color bars pattern, based on
8320 the SMPTE Engineering Guideline EG 1-1990.
8322 The @code{smptehdbars} source generates a color bars pattern, based on
8323 the SMPTE RP 219-2002.
8325 The @code{testsrc} source generates a test video pattern, showing a
8326 color pattern, a scrolling gradient and a timestamp. This is mainly
8327 intended for testing purposes.
8329 The sources accept the following options:
8334 Specify the color of the source, only available in the @code{color}
8335 source. It can be the name of a color (case insensitive match) or a
8336 0xRRGGBB[AA] sequence, possibly followed by an alpha specifier. The
8337 default value is "black".
8340 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
8341 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
8342 pixels to be used as identity matrix for 3D lookup tables. Each component is
8343 coded on a @code{1/(N*N)} scale.
8346 Specify the size of the sourced video, it may be a string of the form
8347 @var{width}x@var{height}, or the name of a size abbreviation. The
8348 default value is "320x240".
8350 This option is not available with the @code{haldclutsrc} filter.
8353 Specify the frame rate of the sourced video, as the number of frames
8354 generated per second. It has to be a string in the format
8355 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
8356 number or a valid video frame rate abbreviation. The default value is
8360 Set the sample aspect ratio of the sourced video.
8363 Set the video duration of the sourced video. The accepted syntax is:
8365 [-]HH[:MM[:SS[.m...]]]
8368 See also the function @code{av_parse_time()}.
8370 If not specified, or the expressed duration is negative, the video is
8371 supposed to be generated forever.
8374 Set the number of decimals to show in the timestamp, only available in the
8375 @code{testsrc} source.
8377 The displayed timestamp value will correspond to the original
8378 timestamp value multiplied by the power of 10 of the specified
8379 value. Default value is 0.
8382 For example the following:
8384 testsrc=duration=5.3:size=qcif:rate=10
8387 will generate a video with a duration of 5.3 seconds, with size
8388 176x144 and a frame rate of 10 frames per second.
8390 The following graph description will generate a red source
8391 with an opacity of 0.2, with size "qcif" and a frame rate of 10
8394 color=c=red@@0.2:s=qcif:r=10
8397 If the input content is to be ignored, @code{nullsrc} can be used. The
8398 following command generates noise in the luminance plane by employing
8399 the @code{geq} filter:
8401 nullsrc=s=256x256, geq=random(1)*255:128:128
8404 @subsection Commands
8406 The @code{color} source supports the following commands:
8410 Set the color of the created image. Accepts the same syntax of the
8411 corresponding @option{color} option.
8414 @c man end VIDEO SOURCES
8416 @chapter Video Sinks
8417 @c man begin VIDEO SINKS
8419 Below is a description of the currently available video sinks.
8423 Buffer video frames, and make them available to the end of the filter
8426 This sink is mainly intended for a programmatic use, in particular
8427 through the interface defined in @file{libavfilter/buffersink.h}
8428 or the options system.
8430 It accepts a pointer to an AVBufferSinkContext structure, which
8431 defines the incoming buffers' formats, to be passed as the opaque
8432 parameter to @code{avfilter_init_filter} for initialization.
8436 Null video sink, do absolutely nothing with the input video. It is
8437 mainly useful as a template and to be employed in analysis / debugging
8440 @c man end VIDEO SINKS
8442 @chapter Multimedia Filters
8443 @c man begin MULTIMEDIA FILTERS
8445 Below is a description of the currently available multimedia filters.
8447 @section avectorscope
8449 Convert input audio to a video output, representing the audio vector
8452 The filter is used to measure the difference between channels of stereo
8453 audio stream. A monoaural signal, consisting of identical left and right
8454 signal, results in straight vertical line. Any stereo separation is visible
8455 as a deviation from this line, creating a Lissajous figure.
8456 If the straight (or deviation from it) but horizontal line appears this
8457 indicates that the left and right channels are out of phase.
8459 The filter accepts the following options:
8463 Set the vectorscope mode.
8465 Available values are:
8468 Lissajous rotated by 45 degrees.
8471 Same as above but not rotated.
8474 Default value is @samp{lissajous}.
8477 Set the video size for the output. Default value is @code{400x400}.
8480 Set the output frame rate. Default value is @code{25}.
8485 Specify the red, green and blue contrast. Default values are @code{40}, @code{160} and @code{80}.
8486 Allowed range is @code{[0, 255]}.
8491 Specify the red, green and blue fade. Default values are @code{15}, @code{10} and @code{5}.
8492 Allowed range is @code{[0, 255]}.
8495 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
8498 @subsection Examples
8502 Complete example using @command{ffplay}:
8504 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
8505 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
8511 Concatenate audio and video streams, joining them together one after the
8514 The filter works on segments of synchronized video and audio streams. All
8515 segments must have the same number of streams of each type, and that will
8516 also be the number of streams at output.
8518 The filter accepts the following options:
8523 Set the number of segments. Default is 2.
8526 Set the number of output video streams, that is also the number of video
8527 streams in each segment. Default is 1.
8530 Set the number of output audio streams, that is also the number of video
8531 streams in each segment. Default is 0.
8534 Activate unsafe mode: do not fail if segments have a different format.
8538 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
8539 @var{a} audio outputs.
8541 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
8542 segment, in the same order as the outputs, then the inputs for the second
8545 Related streams do not always have exactly the same duration, for various
8546 reasons including codec frame size or sloppy authoring. For that reason,
8547 related synchronized streams (e.g. a video and its audio track) should be
8548 concatenated at once. The concat filter will use the duration of the longest
8549 stream in each segment (except the last one), and if necessary pad shorter
8550 audio streams with silence.
8552 For this filter to work correctly, all segments must start at timestamp 0.
8554 All corresponding streams must have the same parameters in all segments; the
8555 filtering system will automatically select a common pixel format for video
8556 streams, and a common sample format, sample rate and channel layout for
8557 audio streams, but other settings, such as resolution, must be converted
8558 explicitly by the user.
8560 Different frame rates are acceptable but will result in variable frame rate
8561 at output; be sure to configure the output file to handle it.
8563 @subsection Examples
8567 Concatenate an opening, an episode and an ending, all in bilingual version
8568 (video in stream 0, audio in streams 1 and 2):
8570 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
8571 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
8572 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
8573 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
8577 Concatenate two parts, handling audio and video separately, using the
8578 (a)movie sources, and adjusting the resolution:
8580 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
8581 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
8582 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
8584 Note that a desync will happen at the stitch if the audio and video streams
8585 do not have exactly the same duration in the first file.
8591 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
8592 it unchanged. By default, it logs a message at a frequency of 10Hz with the
8593 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
8594 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
8596 The filter also has a video output (see the @var{video} option) with a real
8597 time graph to observe the loudness evolution. The graphic contains the logged
8598 message mentioned above, so it is not printed anymore when this option is set,
8599 unless the verbose logging is set. The main graphing area contains the
8600 short-term loudness (3 seconds of analysis), and the gauge on the right is for
8601 the momentary loudness (400 milliseconds).
8603 More information about the Loudness Recommendation EBU R128 on
8604 @url{http://tech.ebu.ch/loudness}.
8606 The filter accepts the following options:
8611 Activate the video output. The audio stream is passed unchanged whether this
8612 option is set or no. The video stream will be the first output stream if
8613 activated. Default is @code{0}.
8616 Set the video size. This option is for video only. Default and minimum
8617 resolution is @code{640x480}.
8620 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
8621 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
8622 other integer value between this range is allowed.
8625 Set metadata injection. If set to @code{1}, the audio input will be segmented
8626 into 100ms output frames, each of them containing various loudness information
8627 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
8629 Default is @code{0}.
8632 Force the frame logging level.
8634 Available values are:
8637 information logging level
8639 verbose logging level
8642 By default, the logging level is set to @var{info}. If the @option{video} or
8643 the @option{metadata} options are set, it switches to @var{verbose}.
8646 @subsection Examples
8650 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
8652 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
8656 Run an analysis with @command{ffmpeg}:
8658 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
8662 @section interleave, ainterleave
8664 Temporally interleave frames from several inputs.
8666 @code{interleave} works with video inputs, @code{ainterleave} with audio.
8668 These filters read frames from several inputs and send the oldest
8669 queued frame to the output.
8671 Input streams must have a well defined, monotonically increasing frame
8674 In order to submit one frame to output, these filters need to enqueue
8675 at least one frame for each input, so they cannot work in case one
8676 input is not yet terminated and will not receive incoming frames.
8678 For example consider the case when one input is a @code{select} filter
8679 which always drop input frames. The @code{interleave} filter will keep
8680 reading from that input, but it will never be able to send new frames
8681 to output until the input will send an end-of-stream signal.
8683 Also, depending on inputs synchronization, the filters will drop
8684 frames in case one input receives more frames than the other ones, and
8685 the queue is already filled.
8687 These filters accept the following options:
8691 Set the number of different inputs, it is 2 by default.
8694 @subsection Examples
8698 Interleave frames belonging to different streams using @command{ffmpeg}:
8700 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
8704 Add flickering blur effect:
8706 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
8710 @section perms, aperms
8712 Set read/write permissions for the output frames.
8714 These filters are mainly aimed at developers to test direct path in the
8715 following filter in the filtergraph.
8717 The filters accept the following options:
8721 Select the permissions mode.
8723 It accepts the following values:
8726 Do nothing. This is the default.
8728 Set all the output frames read-only.
8730 Set all the output frames directly writable.
8732 Make the frame read-only if writable, and writable if read-only.
8734 Set each output frame read-only or writable randomly.
8738 Set the seed for the @var{random} mode, must be an integer included between
8739 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
8740 @code{-1}, the filter will try to use a good random seed on a best effort
8744 Note: in case of auto-inserted filter between the permission filter and the
8745 following one, the permission might not be received as expected in that
8746 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
8747 perms/aperms filter can avoid this problem.
8749 @section select, aselect
8751 Select frames to pass in output.
8753 This filter accepts the following options:
8758 Set expression, which is evaluated for each input frame.
8760 If the expression is evaluated to zero, the frame is discarded.
8762 If the evaluation result is negative or NaN, the frame is sent to the
8763 first output; otherwise it is sent to the output with index
8764 @code{ceil(val)-1}, assuming that the input index starts from 0.
8766 For example a value of @code{1.2} corresponds to the output with index
8767 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
8770 Set the number of outputs. The output to which to send the selected
8771 frame is based on the result of the evaluation. Default value is 1.
8774 The expression can contain the following constants:
8778 the sequential number of the filtered frame, starting from 0
8781 the sequential number of the selected frame, starting from 0
8783 @item prev_selected_n
8784 the sequential number of the last selected frame, NAN if undefined
8787 timebase of the input timestamps
8790 the PTS (Presentation TimeStamp) of the filtered video frame,
8791 expressed in @var{TB} units, NAN if undefined
8794 the PTS (Presentation TimeStamp) of the filtered video frame,
8795 expressed in seconds, NAN if undefined
8798 the PTS of the previously filtered video frame, NAN if undefined
8800 @item prev_selected_pts
8801 the PTS of the last previously filtered video frame, NAN if undefined
8803 @item prev_selected_t
8804 the PTS of the last previously selected video frame, NAN if undefined
8807 the PTS of the first video frame in the video, NAN if undefined
8810 the time of the first video frame in the video, NAN if undefined
8812 @item pict_type @emph{(video only)}
8813 the type of the filtered frame, can assume one of the following
8825 @item interlace_type @emph{(video only)}
8826 the frame interlace type, can assume one of the following values:
8829 the frame is progressive (not interlaced)
8831 the frame is top-field-first
8833 the frame is bottom-field-first
8836 @item consumed_sample_n @emph{(audio only)}
8837 the number of selected samples before the current frame
8839 @item samples_n @emph{(audio only)}
8840 the number of samples in the current frame
8842 @item sample_rate @emph{(audio only)}
8843 the input sample rate
8846 1 if the filtered frame is a key-frame, 0 otherwise
8849 the position in the file of the filtered frame, -1 if the information
8850 is not available (e.g. for synthetic video)
8852 @item scene @emph{(video only)}
8853 value between 0 and 1 to indicate a new scene; a low value reflects a low
8854 probability for the current frame to introduce a new scene, while a higher
8855 value means the current frame is more likely to be one (see the example below)
8859 The default value of the select expression is "1".
8861 @subsection Examples
8865 Select all frames in input:
8870 The example above is the same as:
8882 Select only I-frames:
8884 select='eq(pict_type\,I)'
8888 Select one frame every 100:
8890 select='not(mod(n\,100))'
8894 Select only frames contained in the 10-20 time interval:
8896 select='gte(t\,10)*lte(t\,20)'
8900 Select only I frames contained in the 10-20 time interval:
8902 select='gte(t\,10)*lte(t\,20)*eq(pict_type\,I)'
8906 Select frames with a minimum distance of 10 seconds:
8908 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
8912 Use aselect to select only audio frames with samples number > 100:
8914 aselect='gt(samples_n\,100)'
8918 Create a mosaic of the first scenes:
8920 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
8923 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
8927 Send even and odd frames to separate outputs, and compose them:
8929 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
8933 @section sendcmd, asendcmd
8935 Send commands to filters in the filtergraph.
8937 These filters read commands to be sent to other filters in the
8940 @code{sendcmd} must be inserted between two video filters,
8941 @code{asendcmd} must be inserted between two audio filters, but apart
8942 from that they act the same way.
8944 The specification of commands can be provided in the filter arguments
8945 with the @var{commands} option, or in a file specified by the
8946 @var{filename} option.
8948 These filters accept the following options:
8951 Set the commands to be read and sent to the other filters.
8953 Set the filename of the commands to be read and sent to the other
8957 @subsection Commands syntax
8959 A commands description consists of a sequence of interval
8960 specifications, comprising a list of commands to be executed when a
8961 particular event related to that interval occurs. The occurring event
8962 is typically the current frame time entering or leaving a given time
8965 An interval is specified by the following syntax:
8967 @var{START}[-@var{END}] @var{COMMANDS};
8970 The time interval is specified by the @var{START} and @var{END} times.
8971 @var{END} is optional and defaults to the maximum time.
8973 The current frame time is considered within the specified interval if
8974 it is included in the interval [@var{START}, @var{END}), that is when
8975 the time is greater or equal to @var{START} and is lesser than
8978 @var{COMMANDS} consists of a sequence of one or more command
8979 specifications, separated by ",", relating to that interval. The
8980 syntax of a command specification is given by:
8982 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
8985 @var{FLAGS} is optional and specifies the type of events relating to
8986 the time interval which enable sending the specified command, and must
8987 be a non-null sequence of identifier flags separated by "+" or "|" and
8988 enclosed between "[" and "]".
8990 The following flags are recognized:
8993 The command is sent when the current frame timestamp enters the
8994 specified interval. In other words, the command is sent when the
8995 previous frame timestamp was not in the given interval, and the
8999 The command is sent when the current frame timestamp leaves the
9000 specified interval. In other words, the command is sent when the
9001 previous frame timestamp was in the given interval, and the
9005 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
9008 @var{TARGET} specifies the target of the command, usually the name of
9009 the filter class or a specific filter instance name.
9011 @var{COMMAND} specifies the name of the command for the target filter.
9013 @var{ARG} is optional and specifies the optional list of argument for
9014 the given @var{COMMAND}.
9016 Between one interval specification and another, whitespaces, or
9017 sequences of characters starting with @code{#} until the end of line,
9018 are ignored and can be used to annotate comments.
9020 A simplified BNF description of the commands specification syntax
9023 @var{COMMAND_FLAG} ::= "enter" | "leave"
9024 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
9025 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
9026 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
9027 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
9028 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
9031 @subsection Examples
9035 Specify audio tempo change at second 4:
9037 asendcmd=c='4.0 atempo tempo 1.5',atempo
9041 Specify a list of drawtext and hue commands in a file.
9043 # show text in the interval 5-10
9044 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
9045 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
9047 # desaturate the image in the interval 15-20
9048 15.0-20.0 [enter] hue s 0,
9049 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
9051 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
9053 # apply an exponential saturation fade-out effect, starting from time 25
9054 25 [enter] hue s exp(25-t)
9057 A filtergraph allowing to read and process the above command list
9058 stored in a file @file{test.cmd}, can be specified with:
9060 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
9065 @section setpts, asetpts
9067 Change the PTS (presentation timestamp) of the input frames.
9069 @code{setpts} works on video frames, @code{asetpts} on audio frames.
9071 This filter accepts the following options:
9076 The expression which is evaluated for each frame to construct its timestamp.
9080 The expression is evaluated through the eval API and can contain the following
9085 frame rate, only defined for constant frame-rate video
9088 the presentation timestamp in input
9091 the count of the input frame for video or the number of consumed samples,
9092 not including the current frame for audio, starting from 0.
9094 @item NB_CONSUMED_SAMPLES
9095 the number of consumed samples, not including the current frame (only
9099 the number of samples in the current frame (only audio)
9101 @item SAMPLE_RATE, SR
9105 the PTS of the first frame
9108 the time in seconds of the first frame
9111 tell if the current frame is interlaced
9114 the time in seconds of the current frame
9120 original position in the file of the frame, or undefined if undefined
9121 for the current frame
9127 previous input time in seconds
9133 previous output time in seconds
9136 wallclock (RTC) time in microseconds. This is deprecated, use time(0)
9140 wallclock (RTC) time at the start of the movie in microseconds
9143 @subsection Examples
9147 Start counting PTS from zero
9153 Apply fast motion effect:
9159 Apply slow motion effect:
9165 Set fixed rate of 25 frames per second:
9171 Set fixed rate 25 fps with some jitter:
9173 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
9177 Apply an offset of 10 seconds to the input PTS:
9183 Generate timestamps from a "live source" and rebase onto the current timebase:
9185 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
9189 Generate timestamps by counting samples:
9196 @section settb, asettb
9198 Set the timebase to use for the output frames timestamps.
9199 It is mainly useful for testing timebase configuration.
9201 This filter accepts the following options:
9206 The expression which is evaluated into the output timebase.
9210 The value for @option{tb} is an arithmetic expression representing a
9211 rational. The expression can contain the constants "AVTB" (the default
9212 timebase), "intb" (the input timebase) and "sr" (the sample rate,
9213 audio only). Default value is "intb".
9215 @subsection Examples
9219 Set the timebase to 1/25:
9225 Set the timebase to 1/10:
9231 Set the timebase to 1001/1000:
9237 Set the timebase to 2*intb:
9243 Set the default timebase value:
9249 @section showspectrum
9251 Convert input audio to a video output, representing the audio frequency
9254 The filter accepts the following options:
9258 Specify the video size for the output. Default value is @code{640x512}.
9261 Specify if the spectrum should slide along the window. Default value is
9265 Specify display mode.
9267 It accepts the following values:
9270 all channels are displayed in the same row
9272 all channels are displayed in separate rows
9275 Default value is @samp{combined}.
9278 Specify display color mode.
9280 It accepts the following values:
9283 each channel is displayed in a separate color
9285 each channel is is displayed using the same color scheme
9288 Default value is @samp{channel}.
9291 Specify scale used for calculating intensity color values.
9293 It accepts the following values:
9298 square root, default
9305 Default value is @samp{sqrt}.
9308 Set saturation modifier for displayed colors. Negative values provide
9309 alternative color scheme. @code{0} is no saturation at all.
9310 Saturation must be in [-10.0, 10.0] range.
9311 Default value is @code{1}.
9314 The usage is very similar to the showwaves filter; see the examples in that
9317 @subsection Examples
9321 Large window with logarithmic color scaling:
9323 showspectrum=s=1280x480:scale=log
9327 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
9329 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
9330 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
9336 Convert input audio to a video output, representing the samples waves.
9338 The filter accepts the following options:
9342 Specify the video size for the output. Default value is "600x240".
9347 Available values are:
9350 Draw a point for each sample.
9353 Draw a vertical line for each sample.
9356 Default value is @code{point}.
9359 Set the number of samples which are printed on the same column. A
9360 larger value will decrease the frame rate. Must be a positive
9361 integer. This option can be set only if the value for @var{rate}
9362 is not explicitly specified.
9365 Set the (approximate) output frame rate. This is done by setting the
9366 option @var{n}. Default value is "25".
9370 @subsection Examples
9374 Output the input file audio and the corresponding video representation
9377 amovie=a.mp3,asplit[out0],showwaves[out1]
9381 Create a synthetic signal and show it with showwaves, forcing a
9382 frame rate of 30 frames per second:
9384 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
9388 @section split, asplit
9390 Split input into several identical outputs.
9392 @code{asplit} works with audio input, @code{split} with video.
9394 The filter accepts a single parameter which specifies the number of outputs. If
9395 unspecified, it defaults to 2.
9397 @subsection Examples
9401 Create two separate outputs from the same input:
9403 [in] split [out0][out1]
9407 To create 3 or more outputs, you need to specify the number of
9410 [in] asplit=3 [out0][out1][out2]
9414 Create two separate outputs from the same input, one cropped and
9417 [in] split [splitout1][splitout2];
9418 [splitout1] crop=100:100:0:0 [cropout];
9419 [splitout2] pad=200:200:100:100 [padout];
9423 Create 5 copies of the input audio with @command{ffmpeg}:
9425 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
9431 Receive commands sent through a libzmq client, and forward them to
9432 filters in the filtergraph.
9434 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
9435 must be inserted between two video filters, @code{azmq} between two
9438 To enable these filters you need to install the libzmq library and
9439 headers and configure FFmpeg with @code{--enable-libzmq}.
9441 For more information about libzmq see:
9442 @url{http://www.zeromq.org/}
9444 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
9445 receives messages sent through a network interface defined by the
9446 @option{bind_address} option.
9448 The received message must be in the form:
9450 @var{TARGET} @var{COMMAND} [@var{ARG}]
9453 @var{TARGET} specifies the target of the command, usually the name of
9454 the filter class or a specific filter instance name.
9456 @var{COMMAND} specifies the name of the command for the target filter.
9458 @var{ARG} is optional and specifies the optional argument list for the
9459 given @var{COMMAND}.
9461 Upon reception, the message is processed and the corresponding command
9462 is injected into the filtergraph. Depending on the result, the filter
9463 will send a reply to the client, adopting the format:
9465 @var{ERROR_CODE} @var{ERROR_REASON}
9469 @var{MESSAGE} is optional.
9471 @subsection Examples
9473 Look at @file{tools/zmqsend} for an example of a zmq client which can
9474 be used to send commands processed by these filters.
9476 Consider the following filtergraph generated by @command{ffplay}
9478 ffplay -dumpgraph 1 -f lavfi "
9479 color=s=100x100:c=red [l];
9480 color=s=100x100:c=blue [r];
9481 nullsrc=s=200x100, zmq [bg];
9482 [bg][l] overlay [bg+l];
9483 [bg+l][r] overlay=x=100 "
9486 To change the color of the left side of the video, the following
9487 command can be used:
9489 echo Parsed_color_0 c yellow | tools/zmqsend
9492 To change the right side:
9494 echo Parsed_color_1 c pink | tools/zmqsend
9497 @c man end MULTIMEDIA FILTERS
9499 @chapter Multimedia Sources
9500 @c man begin MULTIMEDIA SOURCES
9502 Below is a description of the currently available multimedia sources.
9506 This is the same as @ref{movie} source, except it selects an audio
9512 Read audio and/or video stream(s) from a movie container.
9514 This filter accepts the following options:
9518 The name of the resource to read (not necessarily a file but also a device or a
9519 stream accessed through some protocol).
9521 @item format_name, f
9522 Specifies the format assumed for the movie to read, and can be either
9523 the name of a container or an input device. If not specified the
9524 format is guessed from @var{movie_name} or by probing.
9526 @item seek_point, sp
9527 Specifies the seek point in seconds, the frames will be output
9528 starting from this seek point, the parameter is evaluated with
9529 @code{av_strtod} so the numerical value may be suffixed by an IS
9530 postfix. Default value is "0".
9533 Specifies the streams to read. Several streams can be specified,
9534 separated by "+". The source will then have as many outputs, in the
9535 same order. The syntax is explained in the ``Stream specifiers''
9536 section in the ffmpeg manual. Two special names, "dv" and "da" specify
9537 respectively the default (best suited) video and audio stream. Default
9538 is "dv", or "da" if the filter is called as "amovie".
9540 @item stream_index, si
9541 Specifies the index of the video stream to read. If the value is -1,
9542 the best suited video stream will be automatically selected. Default
9543 value is "-1". Deprecated. If the filter is called "amovie", it will select
9544 audio instead of video.
9547 Specifies how many times to read the stream in sequence.
9548 If the value is less than 1, the stream will be read again and again.
9549 Default value is "1".
9551 Note that when the movie is looped the source timestamps are not
9552 changed, so it will generate non monotonically increasing timestamps.
9555 This filter allows to overlay a second video on top of main input of
9556 a filtergraph as shown in this graph:
9558 input -----------> deltapts0 --> overlay --> output
9561 movie --> scale--> deltapts1 -------+
9564 @subsection Examples
9568 Skip 3.2 seconds from the start of the avi file in.avi, and overlay it
9569 on top of the input labelled as "in":
9571 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
9572 [in] setpts=PTS-STARTPTS [main];
9573 [main][over] overlay=16:16 [out]
9577 Read from a video4linux2 device, and overlay it on top of the input
9580 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
9581 [in] setpts=PTS-STARTPTS [main];
9582 [main][over] overlay=16:16 [out]
9586 Read the first video stream and the audio stream with id 0x81 from
9587 dvd.vob; the video is connected to the pad named "video" and the audio is
9588 connected to the pad named "audio":
9590 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
9594 @c man end MULTIMEDIA SOURCES