1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, sends one
20 stream through the crop filter and the vflip filter before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that in output the top half of the video is mirrored
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to the one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph can be represented using a textual representation, which is
118 recognized by the @option{-filter}/@option{-vf} and @option{-filter_complex}
119 options in @command{ffmpeg} and @option{-vf} in @command{ffplay}, and by the
120 @code{avfilter_graph_parse()}/@code{avfilter_graph_parse2()} function defined in
121 @file{libavfilter/avfilter.h}.
123 A filterchain consists of a sequence of connected filters, each one
124 connected to the previous one in the sequence. A filterchain is
125 represented by a list of ","-separated filter descriptions.
127 A filtergraph consists of a sequence of filterchains. A sequence of
128 filterchains is represented by a list of ";"-separated filterchain
131 A filter is represented by a string of the form:
132 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
134 @var{filter_name} is the name of the filter class of which the
135 described filter is an instance of, and has to be the name of one of
136 the filter classes registered in the program.
137 The name of the filter class is optionally followed by a string
140 @var{arguments} is a string which contains the parameters used to
141 initialize the filter instance. It may have one of the following forms:
145 A ':'-separated list of @var{key=value} pairs.
148 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
149 the option names in the order they are declared. E.g. the @code{fade} filter
150 declares three options in this order -- @option{type}, @option{start_frame} and
151 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
152 @var{in} is assigned to the option @option{type}, @var{0} to
153 @option{start_frame} and @var{30} to @option{nb_frames}.
156 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
157 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
158 follow the same constraints order of the previous point. The following
159 @var{key=value} pairs can be set in any preferred order.
163 If the option value itself is a list of items (e.g. the @code{format} filter
164 takes a list of pixel formats), the items in the list are usually separated by
167 The list of arguments can be quoted using the character "'" as initial
168 and ending mark, and the character '\' for escaping the characters
169 within the quoted text; otherwise the argument string is considered
170 terminated when the next special character (belonging to the set
171 "[]=;,") is encountered.
173 The name and arguments of the filter are optionally preceded and
174 followed by a list of link labels.
175 A link label allows to name a link and associate it to a filter output
176 or input pad. The preceding labels @var{in_link_1}
177 ... @var{in_link_N}, are associated to the filter input pads,
178 the following labels @var{out_link_1} ... @var{out_link_M}, are
179 associated to the output pads.
181 When two link labels with the same name are found in the
182 filtergraph, a link between the corresponding input and output pad is
185 If an output pad is not labelled, it is linked by default to the first
186 unlabelled input pad of the next filter in the filterchain.
187 For example in the filterchain:
189 nullsrc, split[L1], [L2]overlay, nullsink
191 the split filter instance has two output pads, and the overlay filter
192 instance two input pads. The first output pad of split is labelled
193 "L1", the first input pad of overlay is labelled "L2", and the second
194 output pad of split is linked to the second input pad of overlay,
195 which are both unlabelled.
197 In a complete filterchain all the unlabelled filter input and output
198 pads must be connected. A filtergraph is considered valid if all the
199 filter input and output pads of all the filterchains are connected.
201 Libavfilter will automatically insert scale filters where format
202 conversion is required. It is possible to specify swscale flags
203 for those automatically inserted scalers by prepending
204 @code{sws_flags=@var{flags};}
205 to the filtergraph description.
207 Follows a BNF description for the filtergraph syntax:
209 @var{NAME} ::= sequence of alphanumeric characters and '_'
210 @var{LINKLABEL} ::= "[" @var{NAME} "]"
211 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
212 @var{FILTER_ARGUMENTS} ::= sequence of chars (eventually quoted)
213 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
214 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
215 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
218 @section Notes on filtergraph escaping
220 Some filter arguments require the use of special characters, typically
221 @code{:} to separate key=value pairs in a named options list. In this
222 case the user should perform a first level escaping when specifying
223 the filter arguments. For example, consider the following literal
224 string to be embedded in the @ref{drawtext} filter arguments:
226 this is a 'string': may contain one, or more, special characters
229 Since @code{:} is special for the filter arguments syntax, it needs to
230 be escaped, so you get:
232 text=this is a \'string\'\: may contain one, or more, special characters
235 A second level of escaping is required when embedding the filter
236 arguments in a filtergraph description, in order to escape all the
237 filtergraph special characters. Thus the example above becomes:
239 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
242 Finally an additional level of escaping may be needed when writing the
243 filtergraph description in a shell command, which depends on the
244 escaping rules of the adopted shell. For example, assuming that
245 @code{\} is special and needs to be escaped with another @code{\}, the
246 previous string will finally result in:
248 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
251 Sometimes, it might be more convenient to employ quoting in place of
252 escaping. For example the string:
254 Caesar: tu quoque, Brute, fili mi
257 Can be quoted in the filter arguments as:
259 text='Caesar: tu quoque, Brute, fili mi'
262 And finally inserted in a filtergraph like:
264 drawtext=text=\'Caesar: tu quoque\, Brute\, fili mi\'
267 See the ``Quoting and escaping'' section in the ffmpeg-utils manual
268 for more information about the escaping and quoting rules adopted by
271 @chapter Timeline editing
273 Some filters support a generic @option{enable} option. For the filters
274 supporting timeline editing, this option can be set to an expression which is
275 evaluated before sending a frame to the filter. If the evaluation is non-zero,
276 the filter will be enabled, otherwise the frame will be sent unchanged to the
277 next filter in the filtergraph.
279 The expression accepts the following values:
282 timestamp expressed in seconds, NAN if the input timestamp is unknown
285 sequential number of the input frame, starting from 0
288 the position in the file of the input frame, NAN if unknown
291 Additionally, these filters support an @option{enable} command that can be used
292 to re-define the expression.
294 Like any other filtering option, the @option{enable} option follows the same
297 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
298 minutes, and a @ref{curves} filter starting at 3 seconds:
300 smartblur = enable='between(t,10,3*60)',
301 curves = enable='gte(t,3)' : preset=cross_process
304 @c man end FILTERGRAPH DESCRIPTION
306 @chapter Audio Filters
307 @c man begin AUDIO FILTERS
309 When you configure your FFmpeg build, you can disable any of the
310 existing filters using @code{--disable-filters}.
311 The configure output will show the audio filters included in your
314 Below is a description of the currently available audio filters.
318 Convert the input audio format to the specified formats.
320 @emph{This filter is deprecated. Use @ref{aformat} instead.}
322 The filter accepts a string of the form:
323 "@var{sample_format}:@var{channel_layout}".
325 @var{sample_format} specifies the sample format, and can be a string or the
326 corresponding numeric value defined in @file{libavutil/samplefmt.h}. Use 'p'
327 suffix for a planar sample format.
329 @var{channel_layout} specifies the channel layout, and can be a string
330 or the corresponding number value defined in @file{libavutil/channel_layout.h}.
332 The special parameter "auto", signifies that the filter will
333 automatically select the output format depending on the output filter.
339 Convert input to float, planar, stereo:
345 Convert input to unsigned 8-bit, automatically select out channel layout:
353 Delay one or more audio channels.
355 Samples in delayed channel are filled with silence.
357 The filter accepts the following option:
361 Set list of delays in milliseconds for each channel separated by '|'.
362 At least one delay greater than 0 should be provided.
363 Unused delays will be silently ignored. If number of given delays is
364 smaller than number of channels all remaining channels will not be delayed.
371 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
372 the second channel (and any other channels that may be present) unchanged.
380 Apply echoing to the input audio.
382 Echoes are reflected sound and can occur naturally amongst mountains
383 (and sometimes large buildings) when talking or shouting; digital echo
384 effects emulate this behaviour and are often used to help fill out the
385 sound of a single instrument or vocal. The time difference between the
386 original signal and the reflection is the @code{delay}, and the
387 loudness of the reflected signal is the @code{decay}.
388 Multiple echoes can have different delays and decays.
390 A description of the accepted parameters follows.
394 Set input gain of reflected signal. Default is @code{0.6}.
397 Set output gain of reflected signal. Default is @code{0.3}.
400 Set list of time intervals in milliseconds between original signal and reflections
401 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
402 Default is @code{1000}.
405 Set list of loudnesses of reflected signals separated by '|'.
406 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
407 Default is @code{0.5}.
414 Make it sound as if there are twice as many instruments as are actually playing:
416 aecho=0.8:0.88:60:0.4
420 If delay is very short, then it sound like a (metallic) robot playing music:
426 A longer delay will sound like an open air concert in the mountains:
428 aecho=0.8:0.9:1000:0.3
432 Same as above but with one more mountain:
434 aecho=0.8:0.9:1000|1800:0.3|0.25
440 Apply fade-in/out effect to input audio.
442 A description of the accepted parameters follows.
446 Specify the effect type, can be either @code{in} for fade-in, or
447 @code{out} for a fade-out effect. Default is @code{in}.
449 @item start_sample, ss
450 Specify the number of the start sample for starting to apply the fade
451 effect. Default is 0.
454 Specify the number of samples for which the fade effect has to last. At
455 the end of the fade-in effect the output audio will have the same
456 volume as the input audio, at the end of the fade-out transition
457 the output audio will be silence. Default is 44100.
460 Specify time for starting to apply the fade effect. Default is 0.
461 The accepted syntax is:
463 [-]HH[:MM[:SS[.m...]]]
466 See also the function @code{av_parse_time()}.
467 If set this option is used instead of @var{start_sample} one.
470 Specify the duration for which the fade effect has to last. Default is 0.
471 The accepted syntax is:
473 [-]HH[:MM[:SS[.m...]]]
476 See also the function @code{av_parse_time()}.
477 At the end of the fade-in effect the output audio will have the same
478 volume as the input audio, at the end of the fade-out transition
479 the output audio will be silence.
480 If set this option is used instead of @var{nb_samples} one.
483 Set curve for fade transition.
485 It accepts the following values:
488 select triangular, linear slope (default)
490 select quarter of sine wave
492 select half of sine wave
494 select exponential sine wave
498 select inverted parabola
514 Fade in first 15 seconds of audio:
520 Fade out last 25 seconds of a 900 seconds audio:
522 afade=t=out:st=875:d=25
529 Set output format constraints for the input audio. The framework will
530 negotiate the most appropriate format to minimize conversions.
532 The filter accepts the following named parameters:
536 A '|'-separated list of requested sample formats.
539 A '|'-separated list of requested sample rates.
541 @item channel_layouts
542 A '|'-separated list of requested channel layouts.
546 If a parameter is omitted, all values are allowed.
548 For example to force the output to either unsigned 8-bit or signed 16-bit stereo:
550 aformat=sample_fmts=u8|s16:channel_layouts=stereo
555 Apply a two-pole all-pass filter with central frequency (in Hz)
556 @var{frequency}, and filter-width @var{width}.
557 An all-pass filter changes the audio's frequency to phase relationship
558 without changing its frequency to amplitude relationship.
560 The filter accepts the following options:
567 Set method to specify band-width of filter.
580 Specify the band-width of a filter in width_type units.
585 Merge two or more audio streams into a single multi-channel stream.
587 The filter accepts the following options:
592 Set the number of inputs. Default is 2.
596 If the channel layouts of the inputs are disjoint, and therefore compatible,
597 the channel layout of the output will be set accordingly and the channels
598 will be reordered as necessary. If the channel layouts of the inputs are not
599 disjoint, the output will have all the channels of the first input then all
600 the channels of the second input, in that order, and the channel layout of
601 the output will be the default value corresponding to the total number of
604 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
605 is FC+BL+BR, then the output will be in 5.1, with the channels in the
606 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
607 first input, b1 is the first channel of the second input).
609 On the other hand, if both input are in stereo, the output channels will be
610 in the default order: a1, a2, b1, b2, and the channel layout will be
611 arbitrarily set to 4.0, which may or may not be the expected value.
613 All inputs must have the same sample rate, and format.
615 If inputs do not have the same duration, the output will stop with the
622 Merge two mono files into a stereo stream:
624 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
628 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
630 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
636 Mixes multiple audio inputs into a single output.
640 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
642 will mix 3 input audio streams to a single output with the same duration as the
643 first input and a dropout transition time of 3 seconds.
645 The filter accepts the following named parameters:
649 Number of inputs. If unspecified, it defaults to 2.
652 How to determine the end-of-stream.
656 Duration of longest input. (default)
659 Duration of shortest input.
662 Duration of first input.
666 @item dropout_transition
667 Transition time, in seconds, for volume renormalization when an input
668 stream ends. The default value is 2 seconds.
674 Pass the audio source unchanged to the output.
678 Pad the end of a audio stream with silence, this can be used together with
679 -shortest to extend audio streams to the same length as the video stream.
682 Add a phasing effect to the input audio.
684 A phaser filter creates series of peaks and troughs in the frequency spectrum.
685 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
687 A description of the accepted parameters follows.
691 Set input gain. Default is 0.4.
694 Set output gain. Default is 0.74
697 Set delay in milliseconds. Default is 3.0.
700 Set decay. Default is 0.4.
703 Set modulation speed in Hz. Default is 0.5.
706 Set modulation type. Default is triangular.
708 It accepts the following values:
718 Resample the input audio to the specified parameters, using the
719 libswresample library. If none are specified then the filter will
720 automatically convert between its input and output.
722 This filter is also able to stretch/squeeze the audio data to make it match
723 the timestamps or to inject silence / cut out audio to make it match the
724 timestamps, do a combination of both or do neither.
726 The filter accepts the syntax
727 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
728 expresses a sample rate and @var{resampler_options} is a list of
729 @var{key}=@var{value} pairs, separated by ":". See the
730 ffmpeg-resampler manual for the complete list of supported options.
736 Resample the input audio to 44100Hz:
742 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
743 samples per second compensation:
749 @section asetnsamples
751 Set the number of samples per each output audio frame.
753 The last output packet may contain a different number of samples, as
754 the filter will flush all the remaining samples when the input audio
757 The filter accepts the following options:
761 @item nb_out_samples, n
762 Set the number of frames per each output audio frame. The number is
763 intended as the number of samples @emph{per each channel}.
764 Default value is 1024.
767 If set to 1, the filter will pad the last audio frame with zeroes, so
768 that the last frame will contain the same number of samples as the
769 previous ones. Default value is 1.
772 For example, to set the number of per-frame samples to 1234 and
773 disable padding for the last frame, use:
775 asetnsamples=n=1234:p=0
780 Set the sample rate without altering the PCM data.
781 This will result in a change of speed and pitch.
783 The filter accepts the following options:
787 Set the output sample rate. Default is 44100 Hz.
792 Show a line containing various information for each input audio frame.
793 The input audio is not modified.
795 The shown line contains a sequence of key/value pairs of the form
796 @var{key}:@var{value}.
798 A description of each shown parameter follows:
802 sequential number of the input frame, starting from 0
805 Presentation timestamp of the input frame, in time base units; the time base
806 depends on the filter input pad, and is usually 1/@var{sample_rate}.
809 presentation timestamp of the input frame in seconds
812 position of the frame in the input stream, -1 if this information in
813 unavailable and/or meaningless (for example in case of synthetic audio)
822 sample rate for the audio frame
825 number of samples (per channel) in the frame
828 Adler-32 checksum (printed in hexadecimal) of the audio data. For planar audio
829 the data is treated as if all the planes were concatenated.
831 @item plane_checksums
832 A list of Adler-32 checksums for each data plane.
837 Display time domain statistical information about the audio channels.
838 Statistics are calculated and displayed for each audio channel and,
839 where applicable, an overall figure is also given.
841 The filter accepts the following option:
844 Short window length in seconds, used for peak and trough RMS measurement.
845 Default is @code{0.05} (50 miliseconds). Allowed range is @code{[0.1 - 10]}.
848 A description of each shown parameter follows:
852 Mean amplitude displacement from zero.
855 Minimal sample level.
858 Maximal sample level.
862 Standard peak and RMS level measured in dBFS.
866 Peak and trough values for RMS level measured over a short window.
869 Standard ratio of peak to RMS level (note: not in dB).
872 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
873 (i.e. either @var{Min level} or @var{Max level}).
876 Number of occasions (not the number of samples) that the signal attained either
877 @var{Min level} or @var{Max level}.
882 Forward two audio streams and control the order the buffers are forwarded.
884 The filter accepts the following options:
888 Set the expression deciding which stream should be
889 forwarded next: if the result is negative, the first stream is forwarded; if
890 the result is positive or zero, the second stream is forwarded. It can use
891 the following variables:
895 number of buffers forwarded so far on each stream
897 number of samples forwarded so far on each stream
899 current timestamp of each stream
902 The default value is @code{t1-t2}, which means to always forward the stream
903 that has a smaller timestamp.
908 Stress-test @code{amerge} by randomly sending buffers on the wrong
909 input, while avoiding too much of a desynchronization:
911 amovie=file.ogg [a] ; amovie=file.mp3 [b] ;
912 [a] [b] astreamsync=(2*random(1))-1+tanh(5*(t1-t2)) [a2] [b2] ;
918 Synchronize audio data with timestamps by squeezing/stretching it and/or
919 dropping samples/adding silence when needed.
921 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
923 The filter accepts the following named parameters:
927 Enable stretching/squeezing the data to make it match the timestamps. Disabled
928 by default. When disabled, time gaps are covered with silence.
931 Minimum difference between timestamps and audio data (in seconds) to trigger
932 adding/dropping samples. Default value is 0.1. If you get non-perfect sync with
933 this filter, try setting this parameter to 0.
936 Maximum compensation in samples per second. Relevant only with compensate=1.
940 Assume the first pts should be this value. The time base is 1 / sample rate.
941 This allows for padding/trimming at the start of stream. By default, no
942 assumption is made about the first frame's expected pts, so no padding or
943 trimming is done. For example, this could be set to 0 to pad the beginning with
944 silence if an audio stream starts after the video stream or to trim any samples
945 with a negative pts due to encoder delay.
953 The filter accepts exactly one parameter, the audio tempo. If not
954 specified then the filter will assume nominal 1.0 tempo. Tempo must
955 be in the [0.5, 2.0] range.
961 Slow down audio to 80% tempo:
967 To speed up audio to 125% tempo:
975 Trim the input so that the output contains one continuous subpart of the input.
977 This filter accepts the following options:
980 Specify time of the start of the kept section, i.e. the audio sample
981 with the timestamp @var{start} will be the first sample in the output.
984 Specify time of the first audio sample that will be dropped, i.e. the
985 audio sample immediately preceding the one with the timestamp @var{end} will be
986 the last sample in the output.
989 Same as @var{start}, except this option sets the start timestamp in samples
993 Same as @var{end}, except this option sets the end timestamp in samples instead
997 Specify maximum duration of the output.
1000 Number of the first sample that should be passed to output.
1003 Number of the first sample that should be dropped.
1006 @option{start}, @option{end}, @option{duration} are expressed as time
1007 duration specifications, check the "Time duration" section in the
1008 ffmpeg-utils manual.
1010 Note that the first two sets of the start/end options and the @option{duration}
1011 option look at the frame timestamp, while the _sample options simply count the
1012 samples that pass through the filter. So start/end_pts and start/end_sample will
1013 give different results when the timestamps are wrong, inexact or do not start at
1014 zero. Also note that this filter does not modify the timestamps. If you wish
1015 that the output timestamps start at zero, insert the asetpts filter after the
1018 If multiple start or end options are set, this filter tries to be greedy and
1019 keep all samples that match at least one of the specified constraints. To keep
1020 only the part that matches all the constraints at once, chain multiple atrim
1023 The defaults are such that all the input is kept. So it is possible to set e.g.
1024 just the end values to keep everything before the specified time.
1029 drop everything except the second minute of input
1031 ffmpeg -i INPUT -af atrim=60:120
1035 keep only the first 1000 samples
1037 ffmpeg -i INPUT -af atrim=end_sample=1000
1044 Apply a two-pole Butterworth band-pass filter with central
1045 frequency @var{frequency}, and (3dB-point) band-width width.
1046 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1047 instead of the default: constant 0dB peak gain.
1048 The filter roll off at 6dB per octave (20dB per decade).
1050 The filter accepts the following options:
1054 Set the filter's central frequency. Default is @code{3000}.
1057 Constant skirt gain if set to 1. Defaults to 0.
1060 Set method to specify band-width of filter.
1073 Specify the band-width of a filter in width_type units.
1078 Apply a two-pole Butterworth band-reject filter with central
1079 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1080 The filter roll off at 6dB per octave (20dB per decade).
1082 The filter accepts the following options:
1086 Set the filter's central frequency. Default is @code{3000}.
1089 Set method to specify band-width of filter.
1102 Specify the band-width of a filter in width_type units.
1107 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1108 shelving filter with a response similar to that of a standard
1109 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1111 The filter accepts the following options:
1115 Give the gain at 0 Hz. Its useful range is about -20
1116 (for a large cut) to +20 (for a large boost).
1117 Beware of clipping when using a positive gain.
1120 Set the filter's central frequency and so can be used
1121 to extend or reduce the frequency range to be boosted or cut.
1122 The default value is @code{100} Hz.
1125 Set method to specify band-width of filter.
1138 Determine how steep is the filter's shelf transition.
1143 Apply a biquad IIR filter with the given coefficients.
1144 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1145 are the numerator and denominator coefficients respectively.
1149 Remap input channels to new locations.
1151 This filter accepts the following named parameters:
1153 @item channel_layout
1154 Channel layout of the output stream.
1157 Map channels from input to output. The argument is a '|'-separated list of
1158 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1159 @var{in_channel} form. @var{in_channel} can be either the name of the input
1160 channel (e.g. FL for front left) or its index in the input channel layout.
1161 @var{out_channel} is the name of the output channel or its index in the output
1162 channel layout. If @var{out_channel} is not given then it is implicitly an
1163 index, starting with zero and increasing by one for each mapping.
1166 If no mapping is present, the filter will implicitly map input channels to
1167 output channels preserving index.
1169 For example, assuming a 5.1+downmix input MOV file
1171 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1173 will create an output WAV file tagged as stereo from the downmix channels of
1176 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1178 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:channel_layout=5.1' out.wav
1181 @section channelsplit
1183 Split each channel in input audio stream into a separate output stream.
1185 This filter accepts the following named parameters:
1187 @item channel_layout
1188 Channel layout of the input stream. Default is "stereo".
1191 For example, assuming a stereo input MP3 file
1193 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1195 will create an output Matroska file with two audio streams, one containing only
1196 the left channel and the other the right channel.
1198 To split a 5.1 WAV file into per-channel files
1200 ffmpeg -i in.wav -filter_complex
1201 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1202 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1203 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1209 Compress or expand audio dynamic range.
1211 A description of the accepted options follows.
1216 Set list of times in seconds for each channel over which the instantaneous
1217 level of the input signal is averaged to determine its volume.
1218 @option{attacks} refers to increase of volume and @option{decays} refers
1219 to decrease of volume.
1220 For most situations, the attack time (response to the audio getting louder)
1221 should be shorter than the decay time because the human ear is more sensitive
1222 to sudden loud audio than sudden soft audio.
1223 Typical value for attack is @code{0.3} seconds and for decay @code{0.8}
1227 Set list of points for transfer function, specified in dB relative to maximum
1228 possible signal amplitude.
1229 Each key points list need to be defined using the following syntax:
1230 @code{x0/y0 x1/y1 x2/y2 ...}.
1232 The input values must be in strictly increasing order but the transfer
1233 function does not have to be monotonically rising.
1234 The point @code{0/0} is assumed but may be overridden (by @code{0/out-dBn}).
1235 Typical values for the transfer function are @code{-70/-70 -60/-20}.
1238 Set amount for which the points at where adjacent line segments on the
1239 transfer function meet will be rounded. Defaults is @code{0.01}.
1242 Set additional gain in dB to be applied at all points on the transfer function
1243 and allows easy adjustment of the overall gain.
1244 Default is @code{0}.
1247 Set initial volume in dB to be assumed for each channel when filtering starts.
1248 This permits the user to supply a nominal level initially, so that,
1249 for example, a very large gain is not applied to initial signal levels before
1250 the companding has begun to operate. A typical value for audio which is
1251 initially quiet is -90 dB. Default is @code{0}.
1254 Set delay in seconds. Default is @code{0}. The input audio
1255 is analysed immediately, but audio is delayed before being fed to the
1256 volume adjuster. Specifying a delay approximately equal to the attack/decay
1257 times allows the filter to effectively operate in predictive rather than
1261 @subsection Examples
1264 Make music with both quiet and loud passages suitable for listening
1265 in a noisy environment:
1267 compand=.3 .3:1 1:-90/-60 -60/-40 -40/-30 -20/-20:6:0:-90:0.2
1271 Noise-gate for when the noise is at a lower level than the signal:
1273 compand=.1 .1:.2 .2:-900/-900 -50.1/-900 -50/-50:.01:0:-90:.1
1277 Here is another noise-gate, this time for when the noise is at a higher level
1278 than the signal (making it, in some ways, similar to squelch):
1280 compand=.1 .1:.1 .1:-45.1/-45.1 -45/-900 0/-900:.01:45:-90:.1
1286 Make audio easier to listen to on headphones.
1288 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
1289 so that when listened to on headphones the stereo image is moved from
1290 inside your head (standard for headphones) to outside and in front of
1291 the listener (standard for speakers).
1297 Apply a two-pole peaking equalisation (EQ) filter. With this
1298 filter, the signal-level at and around a selected frequency can
1299 be increased or decreased, whilst (unlike bandpass and bandreject
1300 filters) that at all other frequencies is unchanged.
1302 In order to produce complex equalisation curves, this filter can
1303 be given several times, each with a different central frequency.
1305 The filter accepts the following options:
1309 Set the filter's central frequency in Hz.
1312 Set method to specify band-width of filter.
1325 Specify the band-width of a filter in width_type units.
1328 Set the required gain or attenuation in dB.
1329 Beware of clipping when using a positive gain.
1334 Apply a high-pass filter with 3dB point frequency.
1335 The filter can be either single-pole, or double-pole (the default).
1336 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
1338 The filter accepts the following options:
1342 Set frequency in Hz. Default is 3000.
1345 Set number of poles. Default is 2.
1348 Set method to specify band-width of filter.
1361 Specify the band-width of a filter in width_type units.
1362 Applies only to double-pole filter.
1363 The default is 0.707q and gives a Butterworth response.
1368 Join multiple input streams into one multi-channel stream.
1370 The filter accepts the following named parameters:
1374 Number of input streams. Defaults to 2.
1376 @item channel_layout
1377 Desired output channel layout. Defaults to stereo.
1380 Map channels from inputs to output. The argument is a '|'-separated list of
1381 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
1382 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
1383 can be either the name of the input channel (e.g. FL for front left) or its
1384 index in the specified input stream. @var{out_channel} is the name of the output
1388 The filter will attempt to guess the mappings when those are not specified
1389 explicitly. It does so by first trying to find an unused matching input channel
1390 and if that fails it picks the first unused input channel.
1392 E.g. to join 3 inputs (with properly set channel layouts)
1394 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
1397 To build a 5.1 output from 6 single-channel streams:
1399 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
1400 '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'
1406 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
1408 To enable compilation of this filter you need to configure FFmpeg with
1409 @code{--enable-ladspa}.
1413 Specifies the name of LADSPA plugin library to load. If the environment
1414 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
1415 each one of the directories specified by the colon separated list in
1416 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
1417 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
1418 @file{/usr/lib/ladspa/}.
1421 Specifies the plugin within the library. Some libraries contain only
1422 one plugin, but others contain many of them. If this is not set filter
1423 will list all available plugins within the specified library.
1426 Set the '|' separated list of controls which are zero or more floating point
1427 values that determine the behavior of the loaded plugin (for example delay,
1429 Controls need to be defined using the following syntax:
1430 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
1431 @var{valuei} is the value set on the @var{i}-th control.
1432 If @option{controls} is set to @code{help}, all available controls and
1433 their valid ranges are printed.
1435 @item sample_rate, s
1436 Specify the sample rate, default to 44100. Only used if plugin have
1440 Set the number of samples per channel per each output frame, default
1441 is 1024. Only used if plugin have zero inputs.
1444 Set the minimum duration of the sourced audio. See the function
1445 @code{av_parse_time()} for the accepted format, also check the "Time duration"
1446 section in the ffmpeg-utils manual.
1447 Note that the resulting duration may be greater than the specified duration,
1448 as the generated audio is always cut at the end of a complete frame.
1449 If not specified, or the expressed duration is negative, the audio is
1450 supposed to be generated forever.
1451 Only used if plugin have zero inputs.
1455 @subsection Examples
1459 List all available plugins within amp (LADSPA example plugin) library:
1465 List all available controls and their valid ranges for @code{vcf_notch}
1466 plugin from @code{VCF} library:
1468 ladspa=f=vcf:p=vcf_notch:c=help
1472 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
1475 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
1479 Add reverberation to the audio using TAP-plugins
1480 (Tom's Audio Processing plugins):
1482 ladspa=file=tap_reverb:tap_reverb
1486 Generate white noise, with 0.2 amplitude:
1488 ladspa=file=cmt:noise_source_white:c=c0=.2
1492 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
1493 @code{C* Audio Plugin Suite} (CAPS) library:
1495 ladspa=file=caps:Click:c=c1=20'
1499 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
1501 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
1505 @subsection Commands
1507 This filter supports the following commands:
1510 Modify the @var{N}-th control value.
1512 If the specified value is not valid, it is ignored and prior one is kept.
1517 Apply a low-pass filter with 3dB point frequency.
1518 The filter can be either single-pole or double-pole (the default).
1519 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
1521 The filter accepts the following options:
1525 Set frequency in Hz. Default is 500.
1528 Set number of poles. Default is 2.
1531 Set method to specify band-width of filter.
1544 Specify the band-width of a filter in width_type units.
1545 Applies only to double-pole filter.
1546 The default is 0.707q and gives a Butterworth response.
1551 Mix channels with specific gain levels. The filter accepts the output
1552 channel layout followed by a set of channels definitions.
1554 This filter is also designed to remap efficiently the channels of an audio
1557 The filter accepts parameters of the form:
1558 "@var{l}:@var{outdef}:@var{outdef}:..."
1562 output channel layout or number of channels
1565 output channel specification, of the form:
1566 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
1569 output channel to define, either a channel name (FL, FR, etc.) or a channel
1570 number (c0, c1, etc.)
1573 multiplicative coefficient for the channel, 1 leaving the volume unchanged
1576 input channel to use, see out_name for details; it is not possible to mix
1577 named and numbered input channels
1580 If the `=' in a channel specification is replaced by `<', then the gains for
1581 that specification will be renormalized so that the total is 1, thus
1582 avoiding clipping noise.
1584 @subsection Mixing examples
1586 For example, if you want to down-mix from stereo to mono, but with a bigger
1587 factor for the left channel:
1589 pan=1:c0=0.9*c0+0.1*c1
1592 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
1593 7-channels surround:
1595 pan=stereo: FL < FL + 0.5*FC + 0.6*BL + 0.6*SL : FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
1598 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
1599 that should be preferred (see "-ac" option) unless you have very specific
1602 @subsection Remapping examples
1604 The channel remapping will be effective if, and only if:
1607 @item gain coefficients are zeroes or ones,
1608 @item only one input per channel output,
1611 If all these conditions are satisfied, the filter will notify the user ("Pure
1612 channel mapping detected"), and use an optimized and lossless method to do the
1615 For example, if you have a 5.1 source and want a stereo audio stream by
1616 dropping the extra channels:
1618 pan="stereo: c0=FL : c1=FR"
1621 Given the same source, you can also switch front left and front right channels
1622 and keep the input channel layout:
1624 pan="5.1: c0=c1 : c1=c0 : c2=c2 : c3=c3 : c4=c4 : c5=c5"
1627 If the input is a stereo audio stream, you can mute the front left channel (and
1628 still keep the stereo channel layout) with:
1633 Still with a stereo audio stream input, you can copy the right channel in both
1634 front left and right:
1636 pan="stereo: c0=FR : c1=FR"
1641 Convert the audio sample format, sample rate and channel layout. This filter is
1642 not meant to be used directly.
1644 @section silencedetect
1646 Detect silence in an audio stream.
1648 This filter logs a message when it detects that the input audio volume is less
1649 or equal to a noise tolerance value for a duration greater or equal to the
1650 minimum detected noise duration.
1652 The printed times and duration are expressed in seconds.
1654 The filter accepts the following options:
1658 Set silence duration until notification (default is 2 seconds).
1661 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
1662 specified value) or amplitude ratio. Default is -60dB, or 0.001.
1665 @subsection Examples
1669 Detect 5 seconds of silence with -50dB noise tolerance:
1671 silencedetect=n=-50dB:d=5
1675 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
1676 tolerance in @file{silence.mp3}:
1678 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
1684 Boost or cut treble (upper) frequencies of the audio using a two-pole
1685 shelving filter with a response similar to that of a standard
1686 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1688 The filter accepts the following options:
1692 Give the gain at whichever is the lower of ~22 kHz and the
1693 Nyquist frequency. Its useful range is about -20 (for a large cut)
1694 to +20 (for a large boost). Beware of clipping when using a positive gain.
1697 Set the filter's central frequency and so can be used
1698 to extend or reduce the frequency range to be boosted or cut.
1699 The default value is @code{3000} Hz.
1702 Set method to specify band-width of filter.
1715 Determine how steep is the filter's shelf transition.
1720 Adjust the input audio volume.
1722 The filter accepts the following options:
1727 Expresses how the audio volume will be increased or decreased.
1729 Output values are clipped to the maximum value.
1731 The output audio volume is given by the relation:
1733 @var{output_volume} = @var{volume} * @var{input_volume}
1736 Default value for @var{volume} is 1.0.
1739 Set the mathematical precision.
1741 This determines which input sample formats will be allowed, which affects the
1742 precision of the volume scaling.
1746 8-bit fixed-point; limits input sample format to U8, S16, and S32.
1748 32-bit floating-point; limits input sample format to FLT. (default)
1750 64-bit floating-point; limits input sample format to DBL.
1754 @subsection Examples
1758 Halve the input audio volume:
1762 volume=volume=-6.0206dB
1765 In all the above example the named key for @option{volume} can be
1766 omitted, for example like in:
1772 Increase input audio power by 6 decibels using fixed-point precision:
1774 volume=volume=6dB:precision=fixed
1778 @section volumedetect
1780 Detect the volume of the input video.
1782 The filter has no parameters. The input is not modified. Statistics about
1783 the volume will be printed in the log when the input stream end is reached.
1785 In particular it will show the mean volume (root mean square), maximum
1786 volume (on a per-sample basis), and the beginning of a histogram of the
1787 registered volume values (from the maximum value to a cumulated 1/1000 of
1790 All volumes are in decibels relative to the maximum PCM value.
1792 @subsection Examples
1794 Here is an excerpt of the output:
1796 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
1797 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
1798 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
1799 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
1800 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
1801 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
1802 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
1803 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
1804 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
1810 The mean square energy is approximately -27 dB, or 10^-2.7.
1812 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
1814 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
1817 In other words, raising the volume by +4 dB does not cause any clipping,
1818 raising it by +5 dB causes clipping for 6 samples, etc.
1820 @c man end AUDIO FILTERS
1822 @chapter Audio Sources
1823 @c man begin AUDIO SOURCES
1825 Below is a description of the currently available audio sources.
1829 Buffer audio frames, and make them available to the filter chain.
1831 This source is mainly intended for a programmatic use, in particular
1832 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
1834 It accepts the following named parameters:
1839 Timebase which will be used for timestamps of submitted frames. It must be
1840 either a floating-point number or in @var{numerator}/@var{denominator} form.
1843 The sample rate of the incoming audio buffers.
1846 The sample format of the incoming audio buffers.
1847 Either a sample format name or its corresponging integer representation from
1848 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
1850 @item channel_layout
1851 The channel layout of the incoming audio buffers.
1852 Either a channel layout name from channel_layout_map in
1853 @file{libavutil/channel_layout.c} or its corresponding integer representation
1854 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
1857 The number of channels of the incoming audio buffers.
1858 If both @var{channels} and @var{channel_layout} are specified, then they
1863 @subsection Examples
1866 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
1869 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
1870 Since the sample format with name "s16p" corresponds to the number
1871 6 and the "stereo" channel layout corresponds to the value 0x3, this is
1874 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
1879 Generate an audio signal specified by an expression.
1881 This source accepts in input one or more expressions (one for each
1882 channel), which are evaluated and used to generate a corresponding
1885 This source accepts the following options:
1889 Set the '|'-separated expressions list for each separate channel. In case the
1890 @option{channel_layout} option is not specified, the selected channel layout
1891 depends on the number of provided expressions.
1893 @item channel_layout, c
1894 Set the channel layout. The number of channels in the specified layout
1895 must be equal to the number of specified expressions.
1898 Set the minimum duration of the sourced audio. See the function
1899 @code{av_parse_time()} for the accepted format.
1900 Note that the resulting duration may be greater than the specified
1901 duration, as the generated audio is always cut at the end of a
1904 If not specified, or the expressed duration is negative, the audio is
1905 supposed to be generated forever.
1908 Set the number of samples per channel per each output frame,
1911 @item sample_rate, s
1912 Specify the sample rate, default to 44100.
1915 Each expression in @var{exprs} can contain the following constants:
1919 number of the evaluated sample, starting from 0
1922 time of the evaluated sample expressed in seconds, starting from 0
1929 @subsection Examples
1939 Generate a sin signal with frequency of 440 Hz, set sample rate to
1942 aevalsrc="sin(440*2*PI*t):s=8000"
1946 Generate a two channels signal, specify the channel layout (Front
1947 Center + Back Center) explicitly:
1949 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
1953 Generate white noise:
1955 aevalsrc="-2+random(0)"
1959 Generate an amplitude modulated signal:
1961 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
1965 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
1967 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
1974 Null audio source, return unprocessed audio frames. It is mainly useful
1975 as a template and to be employed in analysis / debugging tools, or as
1976 the source for filters which ignore the input data (for example the sox
1979 This source accepts the following options:
1983 @item channel_layout, cl
1985 Specify the channel layout, and can be either an integer or a string
1986 representing a channel layout. The default value of @var{channel_layout}
1989 Check the channel_layout_map definition in
1990 @file{libavutil/channel_layout.c} for the mapping between strings and
1991 channel layout values.
1993 @item sample_rate, r
1994 Specify the sample rate, and defaults to 44100.
1997 Set the number of samples per requested frames.
2001 @subsection Examples
2005 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
2007 anullsrc=r=48000:cl=4
2011 Do the same operation with a more obvious syntax:
2013 anullsrc=r=48000:cl=mono
2017 All the parameters need to be explicitly defined.
2021 Synthesize a voice utterance using the libflite library.
2023 To enable compilation of this filter you need to configure FFmpeg with
2024 @code{--enable-libflite}.
2026 Note that the flite library is not thread-safe.
2028 The filter accepts the following options:
2033 If set to 1, list the names of the available voices and exit
2034 immediately. Default value is 0.
2037 Set the maximum number of samples per frame. Default value is 512.
2040 Set the filename containing the text to speak.
2043 Set the text to speak.
2046 Set the voice to use for the speech synthesis. Default value is
2047 @code{kal}. See also the @var{list_voices} option.
2050 @subsection Examples
2054 Read from file @file{speech.txt}, and synthetize the text using the
2055 standard flite voice:
2057 flite=textfile=speech.txt
2061 Read the specified text selecting the @code{slt} voice:
2063 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
2067 Input text to ffmpeg:
2069 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
2073 Make @file{ffplay} speak the specified text, using @code{flite} and
2074 the @code{lavfi} device:
2076 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
2080 For more information about libflite, check:
2081 @url{http://www.speech.cs.cmu.edu/flite/}
2085 Generate an audio signal made of a sine wave with amplitude 1/8.
2087 The audio signal is bit-exact.
2089 The filter accepts the following options:
2094 Set the carrier frequency. Default is 440 Hz.
2096 @item beep_factor, b
2097 Enable a periodic beep every second with frequency @var{beep_factor} times
2098 the carrier frequency. Default is 0, meaning the beep is disabled.
2100 @item sample_rate, r
2101 Specify the sample rate, default is 44100.
2104 Specify the duration of the generated audio stream.
2106 @item samples_per_frame
2107 Set the number of samples per output frame, default is 1024.
2110 @subsection Examples
2115 Generate a simple 440 Hz sine wave:
2121 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
2125 sine=frequency=220:beep_factor=4:duration=5
2130 @c man end AUDIO SOURCES
2132 @chapter Audio Sinks
2133 @c man begin AUDIO SINKS
2135 Below is a description of the currently available audio sinks.
2137 @section abuffersink
2139 Buffer audio frames, and make them available to the end of filter chain.
2141 This sink is mainly intended for programmatic use, in particular
2142 through the interface defined in @file{libavfilter/buffersink.h}
2143 or the options system.
2145 It accepts a pointer to an AVABufferSinkContext structure, which
2146 defines the incoming buffers' formats, to be passed as the opaque
2147 parameter to @code{avfilter_init_filter} for initialization.
2151 Null audio sink, do absolutely nothing with the input audio. It is
2152 mainly useful as a template and to be employed in analysis / debugging
2155 @c man end AUDIO SINKS
2157 @chapter Video Filters
2158 @c man begin VIDEO FILTERS
2160 When you configure your FFmpeg build, you can disable any of the
2161 existing filters using @code{--disable-filters}.
2162 The configure output will show the video filters included in your
2165 Below is a description of the currently available video filters.
2167 @section alphaextract
2169 Extract the alpha component from the input as a grayscale video. This
2170 is especially useful with the @var{alphamerge} filter.
2174 Add or replace the alpha component of the primary input with the
2175 grayscale value of a second input. This is intended for use with
2176 @var{alphaextract} to allow the transmission or storage of frame
2177 sequences that have alpha in a format that doesn't support an alpha
2180 For example, to reconstruct full frames from a normal YUV-encoded video
2181 and a separate video created with @var{alphaextract}, you might use:
2183 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
2186 Since this filter is designed for reconstruction, it operates on frame
2187 sequences without considering timestamps, and terminates when either
2188 input reaches end of stream. This will cause problems if your encoding
2189 pipeline drops frames. If you're trying to apply an image as an
2190 overlay to a video stream, consider the @var{overlay} filter instead.
2194 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
2195 and libavformat to work. On the other hand, it is limited to ASS (Advanced
2196 Substation Alpha) subtitles files.
2200 Compute the bounding box for the non-black pixels in the input frame
2203 This filter computes the bounding box containing all the pixels with a
2204 luminance value greater than the minimum allowed value.
2205 The parameters describing the bounding box are printed on the filter
2208 The filter accepts the following option:
2212 Set the minimal luminance value. Default is @code{16}.
2215 @section blackdetect
2217 Detect video intervals that are (almost) completely black. Can be
2218 useful to detect chapter transitions, commercials, or invalid
2219 recordings. Output lines contains the time for the start, end and
2220 duration of the detected black interval expressed in seconds.
2222 In order to display the output lines, you need to set the loglevel at
2223 least to the AV_LOG_INFO value.
2225 The filter accepts the following options:
2228 @item black_min_duration, d
2229 Set the minimum detected black duration expressed in seconds. It must
2230 be a non-negative floating point number.
2232 Default value is 2.0.
2234 @item picture_black_ratio_th, pic_th
2235 Set the threshold for considering a picture "black".
2236 Express the minimum value for the ratio:
2238 @var{nb_black_pixels} / @var{nb_pixels}
2241 for which a picture is considered black.
2242 Default value is 0.98.
2244 @item pixel_black_th, pix_th
2245 Set the threshold for considering a pixel "black".
2247 The threshold expresses the maximum pixel luminance value for which a
2248 pixel is considered "black". The provided value is scaled according to
2249 the following equation:
2251 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
2254 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
2255 the input video format, the range is [0-255] for YUV full-range
2256 formats and [16-235] for YUV non full-range formats.
2258 Default value is 0.10.
2261 The following example sets the maximum pixel threshold to the minimum
2262 value, and detects only black intervals of 2 or more seconds:
2264 blackdetect=d=2:pix_th=0.00
2269 Detect frames that are (almost) completely black. Can be useful to
2270 detect chapter transitions or commercials. Output lines consist of
2271 the frame number of the detected frame, the percentage of blackness,
2272 the position in the file if known or -1 and the timestamp in seconds.
2274 In order to display the output lines, you need to set the loglevel at
2275 least to the AV_LOG_INFO value.
2277 The filter accepts the following options:
2282 Set the percentage of the pixels that have to be below the threshold, defaults
2285 @item threshold, thresh
2286 Set the threshold below which a pixel value is considered black, defaults to
2293 Blend two video frames into each other.
2295 It takes two input streams and outputs one stream, the first input is the
2296 "top" layer and second input is "bottom" layer.
2297 Output terminates when shortest input terminates.
2299 A description of the accepted options follows.
2307 Set blend mode for specific pixel component or all pixel components in case
2308 of @var{all_mode}. Default value is @code{normal}.
2310 Available values for component modes are:
2343 Set blend opacity for specific pixel component or all pixel components in case
2344 of @var{all_opacity}. Only used in combination with pixel component blend modes.
2351 Set blend expression for specific pixel component or all pixel components in case
2352 of @var{all_expr}. Note that related mode options will be ignored if those are set.
2354 The expressions can use the following variables:
2358 The sequential number of the filtered frame, starting from @code{0}.
2362 the coordinates of the current sample
2366 the width and height of currently filtered plane
2370 Width and height scale depending on the currently filtered plane. It is the
2371 ratio between the corresponding luma plane number of pixels and the current
2372 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
2373 @code{0.5,0.5} for chroma planes.
2376 Time of the current frame, expressed in seconds.
2379 Value of pixel component at current location for first video frame (top layer).
2382 Value of pixel component at current location for second video frame (bottom layer).
2386 Force termination when the shortest input terminates. Default is @code{0}.
2388 Continue applying the last bottom frame after the end of the stream. A value of
2389 @code{0} disable the filter after the last frame of the bottom layer is reached.
2390 Default is @code{1}.
2393 @subsection Examples
2397 Apply transition from bottom layer to top layer in first 10 seconds:
2399 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
2403 Apply 1x1 checkerboard effect:
2405 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
2411 Apply boxblur algorithm to the input video.
2413 The filter accepts the following options:
2417 @item luma_radius, lr
2418 @item luma_power, lp
2419 @item chroma_radius, cr
2420 @item chroma_power, cp
2421 @item alpha_radius, ar
2422 @item alpha_power, ap
2426 A description of the accepted options follows.
2429 @item luma_radius, lr
2430 @item chroma_radius, cr
2431 @item alpha_radius, ar
2432 Set an expression for the box radius in pixels used for blurring the
2433 corresponding input plane.
2435 The radius value must be a non-negative number, and must not be
2436 greater than the value of the expression @code{min(w,h)/2} for the
2437 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
2440 Default value for @option{luma_radius} is "2". If not specified,
2441 @option{chroma_radius} and @option{alpha_radius} default to the
2442 corresponding value set for @option{luma_radius}.
2444 The expressions can contain the following constants:
2448 the input width and height in pixels
2452 the input chroma image width and height in pixels
2456 horizontal and vertical chroma subsample values. For example for the
2457 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2460 @item luma_power, lp
2461 @item chroma_power, cp
2462 @item alpha_power, ap
2463 Specify how many times the boxblur filter is applied to the
2464 corresponding plane.
2466 Default value for @option{luma_power} is 2. If not specified,
2467 @option{chroma_power} and @option{alpha_power} default to the
2468 corresponding value set for @option{luma_power}.
2470 A value of 0 will disable the effect.
2473 @subsection Examples
2477 Apply a boxblur filter with luma, chroma, and alpha radius
2480 boxblur=luma_radius=2:luma_power=1
2485 Set luma radius to 2, alpha and chroma radius to 0:
2487 boxblur=2:1:cr=0:ar=0
2491 Set luma and chroma radius to a fraction of the video dimension:
2493 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
2497 @section colorbalance
2498 Modify intensity of primary colors (red, green and blue) of input frames.
2500 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
2501 regions for the red-cyan, green-magenta or blue-yellow balance.
2503 A positive adjustment value shifts the balance towards the primary color, a negative
2504 value towards the complementary color.
2506 The filter accepts the following options:
2512 Adjust red, green and blue shadows (darkest pixels).
2517 Adjust red, green and blue midtones (medium pixels).
2522 Adjust red, green and blue highlights (brightest pixels).
2524 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
2527 @subsection Examples
2531 Add red color cast to shadows:
2537 @section colorchannelmixer
2539 Adjust video input frames by re-mixing color channels.
2541 This filter modifies a color channel by adding the values associated to
2542 the other channels of the same pixels. For example if the value to
2543 modify is red, the output value will be:
2545 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
2548 The filter accepts the following options:
2555 Adjust contribution of input red, green, blue and alpha channels for output red channel.
2556 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
2562 Adjust contribution of input red, green, blue and alpha channels for output green channel.
2563 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
2569 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
2570 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
2576 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
2577 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
2579 Allowed ranges for options are @code{[-2.0, 2.0]}.
2582 @subsection Examples
2586 Convert source to grayscale:
2588 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
2591 Simulate sepia tones:
2593 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
2597 @section colormatrix
2599 Convert color matrix.
2601 The filter accepts the following options:
2606 Specify the source and destination color matrix. Both values must be
2609 The accepted values are:
2625 For example to convert from BT.601 to SMPTE-240M, use the command:
2627 colormatrix=bt601:smpte240m
2632 Copy the input source unchanged to the output. Mainly useful for
2637 Crop the input video to given dimensions.
2639 The filter accepts the following options:
2643 Width of the output video. It defaults to @code{iw}.
2644 This expression is evaluated only once during the filter
2648 Height of the output video. It defaults to @code{ih}.
2649 This expression is evaluated only once during the filter
2653 Horizontal position, in the input video, of the left edge of the output video.
2654 It defaults to @code{(in_w-out_w)/2}.
2655 This expression is evaluated per-frame.
2658 Vertical position, in the input video, of the top edge of the output video.
2659 It defaults to @code{(in_h-out_h)/2}.
2660 This expression is evaluated per-frame.
2663 If set to 1 will force the output display aspect ratio
2664 to be the same of the input, by changing the output sample aspect
2665 ratio. It defaults to 0.
2668 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
2669 expressions containing the following constants:
2674 the computed values for @var{x} and @var{y}. They are evaluated for
2679 the input width and height
2683 same as @var{in_w} and @var{in_h}
2687 the output (cropped) width and height
2691 same as @var{out_w} and @var{out_h}
2694 same as @var{iw} / @var{ih}
2697 input sample aspect ratio
2700 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
2704 horizontal and vertical chroma subsample values. For example for the
2705 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2708 the number of input frame, starting from 0
2711 the position in the file of the input frame, NAN if unknown
2714 timestamp expressed in seconds, NAN if the input timestamp is unknown
2718 The expression for @var{out_w} may depend on the value of @var{out_h},
2719 and the expression for @var{out_h} may depend on @var{out_w}, but they
2720 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
2721 evaluated after @var{out_w} and @var{out_h}.
2723 The @var{x} and @var{y} parameters specify the expressions for the
2724 position of the top-left corner of the output (non-cropped) area. They
2725 are evaluated for each frame. If the evaluated value is not valid, it
2726 is approximated to the nearest valid value.
2728 The expression for @var{x} may depend on @var{y}, and the expression
2729 for @var{y} may depend on @var{x}.
2731 @subsection Examples
2735 Crop area with size 100x100 at position (12,34).
2740 Using named options, the example above becomes:
2742 crop=w=100:h=100:x=12:y=34
2746 Crop the central input area with size 100x100:
2752 Crop the central input area with size 2/3 of the input video:
2754 crop=2/3*in_w:2/3*in_h
2758 Crop the input video central square:
2765 Delimit the rectangle with the top-left corner placed at position
2766 100:100 and the right-bottom corner corresponding to the right-bottom
2767 corner of the input image:
2769 crop=in_w-100:in_h-100:100:100
2773 Crop 10 pixels from the left and right borders, and 20 pixels from
2774 the top and bottom borders
2776 crop=in_w-2*10:in_h-2*20
2780 Keep only the bottom right quarter of the input image:
2782 crop=in_w/2:in_h/2:in_w/2:in_h/2
2786 Crop height for getting Greek harmony:
2788 crop=in_w:1/PHI*in_w
2792 Appply trembling effect:
2794 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)
2798 Apply erratic camera effect depending on timestamp:
2800 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)"
2804 Set x depending on the value of y:
2806 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
2812 Auto-detect crop size.
2814 Calculate necessary cropping parameters and prints the recommended
2815 parameters through the logging system. The detected dimensions
2816 correspond to the non-black area of the input video.
2818 The filter accepts the following options:
2823 Set higher black value threshold, which can be optionally specified
2824 from nothing (0) to everything (255). An intensity value greater
2825 to the set value is considered non-black. Default value is 24.
2828 Set the value for which the width/height should be divisible by. The
2829 offset is automatically adjusted to center the video. Use 2 to get
2830 only even dimensions (needed for 4:2:2 video). 16 is best when
2831 encoding to most video codecs. Default value is 16.
2833 @item reset_count, reset
2834 Set the counter that determines after how many frames cropdetect will
2835 reset the previously detected largest video area and start over to
2836 detect the current optimal crop area. Default value is 0.
2838 This can be useful when channel logos distort the video area. 0
2839 indicates never reset and return the largest area encountered during
2846 Apply color adjustments using curves.
2848 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
2849 component (red, green and blue) has its values defined by @var{N} key points
2850 tied from each other using a smooth curve. The x-axis represents the pixel
2851 values from the input frame, and the y-axis the new pixel values to be set for
2854 By default, a component curve is defined by the two points @var{(0;0)} and
2855 @var{(1;1)}. This creates a straight line where each original pixel value is
2856 "adjusted" to its own value, which means no change to the image.
2858 The filter allows you to redefine these two points and add some more. A new
2859 curve (using a natural cubic spline interpolation) will be define to pass
2860 smoothly through all these new coordinates. The new defined points needs to be
2861 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
2862 be in the @var{[0;1]} interval. If the computed curves happened to go outside
2863 the vector spaces, the values will be clipped accordingly.
2865 If there is no key point defined in @code{x=0}, the filter will automatically
2866 insert a @var{(0;0)} point. In the same way, if there is no key point defined
2867 in @code{x=1}, the filter will automatically insert a @var{(1;1)} point.
2869 The filter accepts the following options:
2873 Select one of the available color presets. This option can be used in addition
2874 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
2875 options takes priority on the preset values.
2876 Available presets are:
2879 @item color_negative
2882 @item increase_contrast
2884 @item linear_contrast
2885 @item medium_contrast
2887 @item strong_contrast
2890 Default is @code{none}.
2892 Set the master key points. These points will define a second pass mapping. It
2893 is sometimes called a "luminance" or "value" mapping. It can be used with
2894 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
2895 post-processing LUT.
2897 Set the key points for the red component.
2899 Set the key points for the green component.
2901 Set the key points for the blue component.
2903 Set the key points for all components (not including master).
2904 Can be used in addition to the other key points component
2905 options. In this case, the unset component(s) will fallback on this
2906 @option{all} setting.
2908 Specify a Photoshop curves file (@code{.asv}) to import the settings from.
2911 To avoid some filtergraph syntax conflicts, each key points list need to be
2912 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
2914 @subsection Examples
2918 Increase slightly the middle level of blue:
2920 curves=blue='0.5/0.58'
2926 curves=r='0/0.11 .42/.51 1/0.95':g='0.50/0.48':b='0/0.22 .49/.44 1/0.8'
2928 Here we obtain the following coordinates for each components:
2931 @code{(0;0.11) (0.42;0.51) (1;0.95)}
2933 @code{(0;0) (0.50;0.48) (1;1)}
2935 @code{(0;0.22) (0.49;0.44) (1;0.80)}
2939 The previous example can also be achieved with the associated built-in preset:
2941 curves=preset=vintage
2951 Use a Photoshop preset and redefine the points of the green component:
2953 curves=psfile='MyCurvesPresets/purple.asv':green='0.45/0.53'
2959 Denoise frames using 2D DCT (frequency domain filtering).
2961 This filter is not designed for real time and can be extremely slow.
2963 The filter accepts the following options:
2967 Set the noise sigma constant.
2969 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
2970 coefficient (absolute value) below this threshold with be dropped.
2972 If you need a more advanced filtering, see @option{expr}.
2974 Default is @code{0}.
2977 Set number overlapping pixels for each block. Each block is of size
2978 @code{16x16}. Since the filter can be slow, you may want to reduce this value,
2979 at the cost of a less effective filter and the risk of various artefacts.
2981 If the overlapping value doesn't allow to process the whole input width or
2982 height, a warning will be displayed and according borders won't be denoised.
2984 Default value is @code{15}.
2987 Set the coefficient factor expression.
2989 For each coefficient of a DCT block, this expression will be evaluated as a
2990 multiplier value for the coefficient.
2992 If this is option is set, the @option{sigma} option will be ignored.
2994 The absolute value of the coefficient can be accessed through the @var{c}
2998 @subsection Examples
3000 Apply a denoise with a @option{sigma} of @code{4.5}:
3005 The same operation can be achieved using the expression system:
3007 dctdnoiz=e='gte(c, 4.5*3)'
3013 Drop duplicated frames at regular intervals.
3015 The filter accepts the following options:
3019 Set the number of frames from which one will be dropped. Setting this to
3020 @var{N} means one frame in every batch of @var{N} frames will be dropped.
3021 Default is @code{5}.
3024 Set the threshold for duplicate detection. If the difference metric for a frame
3025 is less than or equal to this value, then it is declared as duplicate. Default
3029 Set scene change threshold. Default is @code{15}.
3033 Set the size of the x and y-axis blocks used during metric calculations.
3034 Larger blocks give better noise suppression, but also give worse detection of
3035 small movements. Must be a power of two. Default is @code{32}.
3038 Mark main input as a pre-processed input and activate clean source input
3039 stream. This allows the input to be pre-processed with various filters to help
3040 the metrics calculation while keeping the frame selection lossless. When set to
3041 @code{1}, the first stream is for the pre-processed input, and the second
3042 stream is the clean source from where the kept frames are chosen. Default is
3046 Set whether or not chroma is considered in the metric calculations. Default is
3052 Suppress a TV station logo by a simple interpolation of the surrounding
3053 pixels. Just set a rectangle covering the logo and watch it disappear
3054 (and sometimes something even uglier appear - your mileage may vary).
3056 This filter accepts the following options:
3061 Specify the top left corner coordinates of the logo. They must be
3066 Specify the width and height of the logo to clear. They must be
3070 Specify the thickness of the fuzzy edge of the rectangle (added to
3071 @var{w} and @var{h}). The default value is 4.
3074 When set to 1, a green rectangle is drawn on the screen to simplify
3075 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
3076 The default value is 0.
3078 The rectangle is drawn on the outermost pixels which will be (partly)
3079 replaced with interpolated values. The values of the next pixels
3080 immediately outside this rectangle in each direction will be used to
3081 compute the interpolated pixel values inside the rectangle.
3085 @subsection Examples
3089 Set a rectangle covering the area with top left corner coordinates 0,0
3090 and size 100x77, setting a band of size 10:
3092 delogo=x=0:y=0:w=100:h=77:band=10
3099 Attempt to fix small changes in horizontal and/or vertical shift. This
3100 filter helps remove camera shake from hand-holding a camera, bumping a
3101 tripod, moving on a vehicle, etc.
3103 The filter accepts the following options:
3111 Specify a rectangular area where to limit the search for motion
3113 If desired the search for motion vectors can be limited to a
3114 rectangular area of the frame defined by its top left corner, width
3115 and height. These parameters have the same meaning as the drawbox
3116 filter which can be used to visualise the position of the bounding
3119 This is useful when simultaneous movement of subjects within the frame
3120 might be confused for camera motion by the motion vector search.
3122 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
3123 then the full frame is used. This allows later options to be set
3124 without specifying the bounding box for the motion vector search.
3126 Default - search the whole frame.
3130 Specify the maximum extent of movement in x and y directions in the
3131 range 0-64 pixels. Default 16.
3134 Specify how to generate pixels to fill blanks at the edge of the
3135 frame. Available values are:
3138 Fill zeroes at blank locations
3140 Original image at blank locations
3142 Extruded edge value at blank locations
3144 Mirrored edge at blank locations
3146 Default value is @samp{mirror}.
3149 Specify the blocksize to use for motion search. Range 4-128 pixels,
3153 Specify the contrast threshold for blocks. Only blocks with more than
3154 the specified contrast (difference between darkest and lightest
3155 pixels) will be considered. Range 1-255, default 125.
3158 Specify the search strategy. Available values are:
3161 Set exhaustive search
3163 Set less exhaustive search.
3165 Default value is @samp{exhaustive}.
3168 If set then a detailed log of the motion search is written to the
3172 If set to 1, specify using OpenCL capabilities, only available if
3173 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
3179 Draw a colored box on the input image.
3181 This filter accepts the following options:
3186 The expressions which specify the top left corner coordinates of the box. Default to 0.
3190 The expressions which specify the width and height of the box, if 0 they are interpreted as
3191 the input width and height. Default to 0.
3194 Specify the color of the box to write, it can be the name of a color
3195 (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
3196 value @code{invert} is used, the box edge color is the same as the
3197 video with inverted luma.
3200 The expression which sets the thickness of the box edge. Default value is @code{3}.
3202 See below for the list of accepted constants.
3205 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
3206 following constants:
3210 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
3214 horizontal and vertical chroma subsample values. For example for the
3215 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3219 The input width and height.
3222 The input sample aspect ratio.
3226 The x and y offset coordinates where the box is drawn.
3230 The width and height of the drawn box.
3233 The thickness of the drawn box.
3235 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
3236 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
3240 @subsection Examples
3244 Draw a black box around the edge of the input image:
3250 Draw a box with color red and an opacity of 50%:
3252 drawbox=10:20:200:60:red@@0.5
3255 The previous example can be specified as:
3257 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
3261 Fill the box with pink color:
3263 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
3267 Draw a 2-pixel red 2.40:1 mask:
3269 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
3275 Draw a grid on the input image.
3277 This filter accepts the following options:
3282 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
3286 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
3287 input width and height, respectively, minus @code{thickness}, so image gets
3288 framed. Default to 0.
3291 Specify the color of the grid, it can be the name of a color
3292 (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
3293 value @code{invert} is used, the grid color is the same as the
3294 video with inverted luma.
3295 Note that you can append opacity value (in range of 0.0 - 1.0)
3296 to color name after @@ sign.
3299 The expression which sets the thickness of the grid line. Default value is @code{1}.
3301 See below for the list of accepted constants.
3304 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
3305 following constants:
3309 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
3313 horizontal and vertical chroma subsample values. For example for the
3314 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3318 The input grid cell width and height.
3321 The input sample aspect ratio.
3325 The x and y coordinates of some point of grid intersection (meant to configure offset).
3329 The width and height of the drawn cell.
3332 The thickness of the drawn cell.
3334 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
3335 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
3339 @subsection Examples
3343 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
3345 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
3349 Draw a white 3x3 grid with an opacity of 50%:
3351 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
3358 Draw text string or text from specified file on top of video using the
3359 libfreetype library.
3361 To enable compilation of this filter you need to configure FFmpeg with
3362 @code{--enable-libfreetype}.
3366 The description of the accepted parameters follows.
3371 Used to draw a box around text using background color.
3372 Value should be either 1 (enable) or 0 (disable).
3373 The default value of @var{box} is 0.
3376 The color to be used for drawing box around text.
3377 Either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
3378 (e.g. "0xff00ff"), possibly followed by an alpha specifier.
3379 The default value of @var{boxcolor} is "white".
3382 Select how the @var{text} is expanded. Can be either @code{none},
3383 @code{strftime} (deprecated) or
3384 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
3388 If true, check and fix text coords to avoid clipping.
3391 The color to be used for drawing fonts.
3392 Either a string (e.g. "red") or in 0xRRGGBB[AA] format
3393 (e.g. "0xff000033"), possibly followed by an alpha specifier.
3394 The default value of @var{fontcolor} is "black".
3397 The font file to be used for drawing text. Path must be included.
3398 This parameter is mandatory.
3401 The font size to be used for drawing text.
3402 The default value of @var{fontsize} is 16.
3405 Flags to be used for loading the fonts.
3407 The flags map the corresponding flags supported by libfreetype, and are
3408 a combination of the following values:
3415 @item vertical_layout
3416 @item force_autohint
3419 @item ignore_global_advance_width
3421 @item ignore_transform
3427 Default value is "render".
3429 For more information consult the documentation for the FT_LOAD_*
3433 The color to be used for drawing a shadow behind the drawn text. It
3434 can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
3435 form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
3436 The default value of @var{shadowcolor} is "black".
3440 The x and y offsets for the text shadow position with respect to the
3441 position of the text. They can be either positive or negative
3442 values. Default value for both is "0".
3445 The starting frame number for the n/frame_num variable. The default value
3449 The size in number of spaces to use for rendering the tab.
3453 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
3454 format. It can be used with or without text parameter. @var{timecode_rate}
3455 option must be specified.
3457 @item timecode_rate, rate, r
3458 Set the timecode frame rate (timecode only).
3461 The text string to be drawn. The text must be a sequence of UTF-8
3463 This parameter is mandatory if no file is specified with the parameter
3467 A text file containing text to be drawn. The text must be a sequence
3468 of UTF-8 encoded characters.
3470 This parameter is mandatory if no text string is specified with the
3471 parameter @var{text}.
3473 If both @var{text} and @var{textfile} are specified, an error is thrown.
3476 If set to 1, the @var{textfile} will be reloaded before each frame.
3477 Be sure to update it atomically, or it may be read partially, or even fail.
3481 The expressions which specify the offsets where text will be drawn
3482 within the video frame. They are relative to the top/left border of the
3485 The default value of @var{x} and @var{y} is "0".
3487 See below for the list of accepted constants and functions.
3490 The parameters for @var{x} and @var{y} are expressions containing the
3491 following constants and functions:
3495 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
3499 horizontal and vertical chroma subsample values. For example for the
3500 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3503 the height of each text line
3511 @item max_glyph_a, ascent
3512 the maximum distance from the baseline to the highest/upper grid
3513 coordinate used to place a glyph outline point, for all the rendered
3515 It is a positive value, due to the grid's orientation with the Y axis
3518 @item max_glyph_d, descent
3519 the maximum distance from the baseline to the lowest grid coordinate
3520 used to place a glyph outline point, for all the rendered glyphs.
3521 This is a negative value, due to the grid's orientation, with the Y axis
3525 maximum glyph height, that is the maximum height for all the glyphs
3526 contained in the rendered text, it is equivalent to @var{ascent} -
3530 maximum glyph width, that is the maximum width for all the glyphs
3531 contained in the rendered text
3534 the number of input frame, starting from 0
3536 @item rand(min, max)
3537 return a random number included between @var{min} and @var{max}
3540 input sample aspect ratio
3543 timestamp expressed in seconds, NAN if the input timestamp is unknown
3546 the height of the rendered text
3549 the width of the rendered text
3553 the x and y offset coordinates where the text is drawn.
3555 These parameters allow the @var{x} and @var{y} expressions to refer
3556 each other, so you can for example specify @code{y=x/dar}.
3559 If libavfilter was built with @code{--enable-fontconfig}, then
3560 @option{fontfile} can be a fontconfig pattern or omitted.
3562 @anchor{drawtext_expansion}
3563 @subsection Text expansion
3565 If @option{expansion} is set to @code{strftime},
3566 the filter recognizes strftime() sequences in the provided text and
3567 expands them accordingly. Check the documentation of strftime(). This
3568 feature is deprecated.
3570 If @option{expansion} is set to @code{none}, the text is printed verbatim.
3572 If @option{expansion} is set to @code{normal} (which is the default),
3573 the following expansion mechanism is used.
3575 The backslash character '\', followed by any character, always expands to
3576 the second character.
3578 Sequence of the form @code{%@{...@}} are expanded. The text between the
3579 braces is a function name, possibly followed by arguments separated by ':'.
3580 If the arguments contain special characters or delimiters (':' or '@}'),
3581 they should be escaped.
3583 Note that they probably must also be escaped as the value for the
3584 @option{text} option in the filter argument string and as the filter
3585 argument in the filtergraph description, and possibly also for the shell,
3586 that makes up to four levels of escaping; using a text file avoids these
3589 The following functions are available:
3594 The expression evaluation result.
3596 It must take one argument specifying the expression to be evaluated,
3597 which accepts the same constants and functions as the @var{x} and
3598 @var{y} values. Note that not all constants should be used, for
3599 example the text size is not known when evaluating the expression, so
3600 the constants @var{text_w} and @var{text_h} will have an undefined
3604 The time at which the filter is running, expressed in UTC.
3605 It can accept an argument: a strftime() format string.
3608 The time at which the filter is running, expressed in the local time zone.
3609 It can accept an argument: a strftime() format string.
3612 Frame metadata. It must take one argument specifying metadata key.
3615 The frame number, starting from 0.
3618 A 1 character description of the current picture type.
3621 The timestamp of the current frame, in seconds, with microsecond accuracy.
3625 @subsection Examples
3629 Draw "Test Text" with font FreeSerif, using the default values for the
3630 optional parameters.
3633 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
3637 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
3638 and y=50 (counting from the top-left corner of the screen), text is
3639 yellow with a red box around it. Both the text and the box have an
3643 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
3644 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
3647 Note that the double quotes are not necessary if spaces are not used
3648 within the parameter list.
3651 Show the text at the center of the video frame:
3653 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h-line_h)/2"
3657 Show a text line sliding from right to left in the last row of the video
3658 frame. The file @file{LONG_LINE} is assumed to contain a single line
3661 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
3665 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
3667 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
3671 Draw a single green letter "g", at the center of the input video.
3672 The glyph baseline is placed at half screen height.
3674 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
3678 Show text for 1 second every 3 seconds:
3680 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
3684 Use fontconfig to set the font. Note that the colons need to be escaped.
3686 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
3690 Print the date of a real-time encoding (see strftime(3)):
3692 drawtext='fontfile=FreeSans.ttf:text=%@{localtime:%a %b %d %Y@}'
3697 For more information about libfreetype, check:
3698 @url{http://www.freetype.org/}.
3700 For more information about fontconfig, check:
3701 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
3705 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
3707 The filter accepts the following options:
3712 Set low and high threshold values used by the Canny thresholding
3715 The high threshold selects the "strong" edge pixels, which are then
3716 connected through 8-connectivity with the "weak" edge pixels selected
3717 by the low threshold.
3719 @var{low} and @var{high} threshold values must be choosen in the range
3720 [0,1], and @var{low} should be lesser or equal to @var{high}.
3722 Default value for @var{low} is @code{20/255}, and default value for @var{high}
3728 edgedetect=low=0.1:high=0.4
3731 @section extractplanes
3733 Extract color channel components from input video stream into
3734 separate grayscale video streams.
3736 The filter accepts the following option:
3740 Set plane(s) to extract.
3742 Available values for planes are:
3753 Choosing planes not available in the input will result in an error.
3754 That means you cannot select @code{r}, @code{g}, @code{b} planes
3755 with @code{y}, @code{u}, @code{v} planes at same time.
3758 @subsection Examples
3762 Extract luma, u and v color channel component from input video frame
3763 into 3 grayscale outputs:
3765 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
3771 Apply fade-in/out effect to input video.
3773 This filter accepts the following options:
3777 The effect type -- can be either "in" for fade-in, or "out" for a fade-out
3779 Default is @code{in}.
3781 @item start_frame, s
3782 Specify the number of the start frame for starting to apply the fade
3783 effect. Default is 0.
3786 The number of frames for which the fade effect has to last. At the end of the
3787 fade-in effect the output video will have the same intensity as the input video,
3788 at the end of the fade-out transition the output video will be completely black.
3792 If set to 1, fade only alpha channel, if one exists on the input.
3795 @item start_time, st
3796 Specify the timestamp (in seconds) of the frame to start to apply the fade
3797 effect. If both start_frame and start_time are specified, the fade will start at
3798 whichever comes last. Default is 0.
3801 The number of seconds for which the fade effect has to last. At the end of the
3802 fade-in effect the output video will have the same intensity as the input video,
3803 at the end of the fade-out transition the output video will be completely black.
3804 If both duration and nb_frames are specified, duration is used. Default is 0.
3807 @subsection Examples
3811 Fade in first 30 frames of video:
3816 The command above is equivalent to:
3822 Fade out last 45 frames of a 200-frame video:
3825 fade=type=out:start_frame=155:nb_frames=45
3829 Fade in first 25 frames and fade out last 25 frames of a 1000-frame video:
3831 fade=in:0:25, fade=out:975:25
3835 Make first 5 frames black, then fade in from frame 5-24:
3841 Fade in alpha over first 25 frames of video:
3843 fade=in:0:25:alpha=1
3847 Make first 5.5 seconds black, then fade in for 0.5 seconds:
3849 fade=t=in:st=5.5:d=0.5
3856 Extract a single field from an interlaced image using stride
3857 arithmetic to avoid wasting CPU time. The output frames are marked as
3860 The filter accepts the following options:
3864 Specify whether to extract the top (if the value is @code{0} or
3865 @code{top}) or the bottom field (if the value is @code{1} or
3871 Field matching filter for inverse telecine. It is meant to reconstruct the
3872 progressive frames from a telecined stream. The filter does not drop duplicated
3873 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
3874 followed by a decimation filter such as @ref{decimate} in the filtergraph.
3876 The separation of the field matching and the decimation is notably motivated by
3877 the possibility of inserting a de-interlacing filter fallback between the two.
3878 If the source has mixed telecined and real interlaced content,
3879 @code{fieldmatch} will not be able to match fields for the interlaced parts.
3880 But these remaining combed frames will be marked as interlaced, and thus can be
3881 de-interlaced by a later filter such as @ref{yadif} before decimation.
3883 In addition to the various configuration options, @code{fieldmatch} can take an
3884 optional second stream, activated through the @option{ppsrc} option. If
3885 enabled, the frames reconstruction will be based on the fields and frames from
3886 this second stream. This allows the first input to be pre-processed in order to
3887 help the various algorithms of the filter, while keeping the output lossless
3888 (assuming the fields are matched properly). Typically, a field-aware denoiser,
3889 or brightness/contrast adjustments can help.
3891 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
3892 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
3893 which @code{fieldmatch} is based on. While the semantic and usage are very
3894 close, some behaviour and options names can differ.
3896 The filter accepts the following options:
3900 Specify the assumed field order of the input stream. Available values are:
3904 Auto detect parity (use FFmpeg's internal parity value).
3906 Assume bottom field first.
3908 Assume top field first.
3911 Note that it is sometimes recommended not to trust the parity announced by the
3914 Default value is @var{auto}.
3917 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
3918 sense that it won't risk creating jerkiness due to duplicate frames when
3919 possible, but if there are bad edits or blended fields it will end up
3920 outputting combed frames when a good match might actually exist. On the other
3921 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
3922 but will almost always find a good frame if there is one. The other values are
3923 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
3924 jerkiness and creating duplicate frames versus finding good matches in sections
3925 with bad edits, orphaned fields, blended fields, etc.
3927 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
3929 Available values are:
3933 2-way matching (p/c)
3935 2-way matching, and trying 3rd match if still combed (p/c + n)
3937 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
3939 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
3940 still combed (p/c + n + u/b)
3942 3-way matching (p/c/n)
3944 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
3945 detected as combed (p/c/n + u/b)
3948 The parenthesis at the end indicate the matches that would be used for that
3949 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
3952 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
3955 Default value is @var{pc_n}.
3958 Mark the main input stream as a pre-processed input, and enable the secondary
3959 input stream as the clean source to pick the fields from. See the filter
3960 introduction for more details. It is similar to the @option{clip2} feature from
3963 Default value is @code{0} (disabled).
3966 Set the field to match from. It is recommended to set this to the same value as
3967 @option{order} unless you experience matching failures with that setting. In
3968 certain circumstances changing the field that is used to match from can have a
3969 large impact on matching performance. Available values are:
3973 Automatic (same value as @option{order}).
3975 Match from the bottom field.
3977 Match from the top field.
3980 Default value is @var{auto}.
3983 Set whether or not chroma is included during the match comparisons. In most
3984 cases it is recommended to leave this enabled. You should set this to @code{0}
3985 only if your clip has bad chroma problems such as heavy rainbowing or other
3986 artifacts. Setting this to @code{0} could also be used to speed things up at
3987 the cost of some accuracy.
3989 Default value is @code{1}.
3993 These define an exclusion band which excludes the lines between @option{y0} and
3994 @option{y1} from being included in the field matching decision. An exclusion
3995 band can be used to ignore subtitles, a logo, or other things that may
3996 interfere with the matching. @option{y0} sets the starting scan line and
3997 @option{y1} sets the ending line; all lines in between @option{y0} and
3998 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
3999 @option{y0} and @option{y1} to the same value will disable the feature.
4000 @option{y0} and @option{y1} defaults to @code{0}.
4003 Set the scene change detection threshold as a percentage of maximum change on
4004 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
4005 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
4006 @option{scthresh} is @code{[0.0, 100.0]}.
4008 Default value is @code{12.0}.
4011 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
4012 account the combed scores of matches when deciding what match to use as the
4013 final match. Available values are:
4017 No final matching based on combed scores.
4019 Combed scores are only used when a scene change is detected.
4021 Use combed scores all the time.
4024 Default is @var{sc}.
4027 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
4028 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
4029 Available values are:
4033 No forced calculation.
4035 Force p/c/n calculations.
4037 Force p/c/n/u/b calculations.
4040 Default value is @var{none}.
4043 This is the area combing threshold used for combed frame detection. This
4044 essentially controls how "strong" or "visible" combing must be to be detected.
4045 Larger values mean combing must be more visible and smaller values mean combing
4046 can be less visible or strong and still be detected. Valid settings are from
4047 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
4048 be detected as combed). This is basically a pixel difference value. A good
4049 range is @code{[8, 12]}.
4051 Default value is @code{9}.
4054 Sets whether or not chroma is considered in the combed frame decision. Only
4055 disable this if your source has chroma problems (rainbowing, etc.) that are
4056 causing problems for the combed frame detection with chroma enabled. Actually,
4057 using @option{chroma}=@var{0} is usually more reliable, except for the case
4058 where there is chroma only combing in the source.
4060 Default value is @code{0}.
4064 Respectively set the x-axis and y-axis size of the window used during combed
4065 frame detection. This has to do with the size of the area in which
4066 @option{combpel} pixels are required to be detected as combed for a frame to be
4067 declared combed. See the @option{combpel} parameter description for more info.
4068 Possible values are any number that is a power of 2 starting at 4 and going up
4071 Default value is @code{16}.
4074 The number of combed pixels inside any of the @option{blocky} by
4075 @option{blockx} size blocks on the frame for the frame to be detected as
4076 combed. While @option{cthresh} controls how "visible" the combing must be, this
4077 setting controls "how much" combing there must be in any localized area (a
4078 window defined by the @option{blockx} and @option{blocky} settings) on the
4079 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
4080 which point no frames will ever be detected as combed). This setting is known
4081 as @option{MI} in TFM/VFM vocabulary.
4083 Default value is @code{80}.
4086 @anchor{p/c/n/u/b meaning}
4087 @subsection p/c/n/u/b meaning
4089 @subsubsection p/c/n
4091 We assume the following telecined stream:
4094 Top fields: 1 2 2 3 4
4095 Bottom fields: 1 2 3 4 4
4098 The numbers correspond to the progressive frame the fields relate to. Here, the
4099 first two frames are progressive, the 3rd and 4th are combed, and so on.
4101 When @code{fieldmatch} is configured to run a matching from bottom
4102 (@option{field}=@var{bottom}) this is how this input stream get transformed:
4107 B 1 2 3 4 4 <-- matching reference
4116 As a result of the field matching, we can see that some frames get duplicated.
4117 To perform a complete inverse telecine, you need to rely on a decimation filter
4118 after this operation. See for instance the @ref{decimate} filter.
4120 The same operation now matching from top fields (@option{field}=@var{top})
4125 T 1 2 2 3 4 <-- matching reference
4135 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
4136 basically, they refer to the frame and field of the opposite parity:
4139 @item @var{p} matches the field of the opposite parity in the previous frame
4140 @item @var{c} matches the field of the opposite parity in the current frame
4141 @item @var{n} matches the field of the opposite parity in the next frame
4146 The @var{u} and @var{b} matching are a bit special in the sense that they match
4147 from the opposite parity flag. In the following examples, we assume that we are
4148 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
4149 'x' is placed above and below each matched fields.
4151 With bottom matching (@option{field}=@var{bottom}):
4156 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
4157 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
4165 With top matching (@option{field}=@var{top}):
4170 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
4171 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
4179 @subsection Examples
4181 Simple IVTC of a top field first telecined stream:
4183 fieldmatch=order=tff:combmatch=none, decimate
4186 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
4188 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
4193 Transform the field order of the input video.
4195 This filter accepts the following options:
4200 Output field order. Valid values are @var{tff} for top field first or @var{bff}
4201 for bottom field first.
4204 Default value is @samp{tff}.
4206 Transformation is achieved by shifting the picture content up or down
4207 by one line, and filling the remaining line with appropriate picture content.
4208 This method is consistent with most broadcast field order converters.
4210 If the input video is not flagged as being interlaced, or it is already
4211 flagged as being of the required output field order then this filter does
4212 not alter the incoming video.
4214 This filter is very useful when converting to or from PAL DV material,
4215 which is bottom field first.
4219 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
4224 Buffer input images and send them when they are requested.
4226 This filter is mainly useful when auto-inserted by the libavfilter
4229 The filter does not take parameters.
4234 Convert the input video to one of the specified pixel formats.
4235 Libavfilter will try to pick one that is supported for the input to
4238 This filter accepts the following parameters:
4242 A '|'-separated list of pixel format names, for example
4243 "pix_fmts=yuv420p|monow|rgb24".
4247 @subsection Examples
4251 Convert the input video to the format @var{yuv420p}
4253 format=pix_fmts=yuv420p
4256 Convert the input video to any of the formats in the list
4258 format=pix_fmts=yuv420p|yuv444p|yuv410p
4264 Convert the video to specified constant frame rate by duplicating or dropping
4265 frames as necessary.
4267 This filter accepts the following named parameters:
4271 Desired output frame rate. The default is @code{25}.
4276 Possible values are:
4279 zero round towards 0
4283 round towards -infinity
4285 round towards +infinity
4289 The default is @code{near}.
4292 Assume the first PTS should be the given value, in seconds. This allows for
4293 padding/trimming at the start of stream. By default, no assumption is made
4294 about the first frame's expected PTS, so no padding or trimming is done.
4295 For example, this could be set to 0 to pad the beginning with duplicates of
4296 the first frame if a video stream starts after the audio stream or to trim any
4297 frames with a negative PTS.
4301 Alternatively, the options can be specified as a flat string:
4302 @var{fps}[:@var{round}].
4304 See also the @ref{setpts} filter.
4306 @subsection Examples
4310 A typical usage in order to set the fps to 25:
4316 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
4318 fps=fps=film:round=near
4324 Select one frame every N-th frame.
4326 This filter accepts the following option:
4329 Select frame after every @code{step} frames.
4330 Allowed values are positive integers higher than 0. Default value is @code{1}.
4336 Apply a frei0r effect to the input video.
4338 To enable compilation of this filter you need to install the frei0r
4339 header and configure FFmpeg with @code{--enable-frei0r}.
4341 This filter accepts the following options:
4346 The name to the frei0r effect to load. If the environment variable
4347 @env{FREI0R_PATH} is defined, the frei0r effect is searched in each one of the
4348 directories specified by the colon separated list in @env{FREIOR_PATH},
4349 otherwise in the standard frei0r paths, which are in this order:
4350 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
4351 @file{/usr/lib/frei0r-1/}.
4354 A '|'-separated list of parameters to pass to the frei0r effect.
4358 A frei0r effect parameter can be a boolean (whose values are specified
4359 with "y" and "n"), a double, a color (specified by the syntax
4360 @var{R}/@var{G}/@var{B}, @var{R}, @var{G}, and @var{B} being float
4361 numbers from 0.0 to 1.0) or by an @code{av_parse_color()} color
4362 description), a position (specified by the syntax @var{X}/@var{Y},
4363 @var{X} and @var{Y} being float numbers) and a string.
4365 The number and kind of parameters depend on the loaded effect. If an
4366 effect parameter is not specified the default value is set.
4368 @subsection Examples
4372 Apply the distort0r effect, set the first two double parameters:
4374 frei0r=filter_name=distort0r:filter_params=0.5|0.01
4378 Apply the colordistance effect, take a color as first parameter:
4380 frei0r=colordistance:0.2/0.3/0.4
4381 frei0r=colordistance:violet
4382 frei0r=colordistance:0x112233
4386 Apply the perspective effect, specify the top left and top right image
4389 frei0r=perspective:0.2/0.2|0.8/0.2
4393 For more information see:
4394 @url{http://frei0r.dyne.org}
4398 The filter accepts the following options:
4402 Set the luminance expression.
4404 Set the chrominance blue expression.
4406 Set the chrominance red expression.
4408 Set the alpha expression.
4410 Set the red expression.
4412 Set the green expression.
4414 Set the blue expression.
4417 The colorspace is selected according to the specified options. If one
4418 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
4419 options is specified, the filter will automatically select a YCbCr
4420 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
4421 @option{blue_expr} options is specified, it will select an RGB
4424 If one of the chrominance expression is not defined, it falls back on the other
4425 one. If no alpha expression is specified it will evaluate to opaque value.
4426 If none of chrominance expressions are specified, they will evaluate
4427 to the luminance expression.
4429 The expressions can use the following variables and functions:
4433 The sequential number of the filtered frame, starting from @code{0}.
4437 The coordinates of the current sample.
4441 The width and height of the image.
4445 Width and height scale depending on the currently filtered plane. It is the
4446 ratio between the corresponding luma plane number of pixels and the current
4447 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4448 @code{0.5,0.5} for chroma planes.
4451 Time of the current frame, expressed in seconds.
4454 Return the value of the pixel at location (@var{x},@var{y}) of the current
4458 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
4462 Return the value of the pixel at location (@var{x},@var{y}) of the
4463 blue-difference chroma plane. Return 0 if there is no such plane.
4466 Return the value of the pixel at location (@var{x},@var{y}) of the
4467 red-difference chroma plane. Return 0 if there is no such plane.
4472 Return the value of the pixel at location (@var{x},@var{y}) of the
4473 red/green/blue component. Return 0 if there is no such component.
4476 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
4477 plane. Return 0 if there is no such plane.
4480 For functions, if @var{x} and @var{y} are outside the area, the value will be
4481 automatically clipped to the closer edge.
4483 @subsection Examples
4487 Flip the image horizontally:
4493 Generate a bidimensional sine wave, with angle @code{PI/3} and a
4494 wavelength of 100 pixels:
4496 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
4500 Generate a fancy enigmatic moving light:
4502 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
4506 Generate a quick emboss effect:
4508 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
4512 Modify RGB components depending on pixel position:
4514 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
4520 Fix the banding artifacts that are sometimes introduced into nearly flat
4521 regions by truncation to 8bit color depth.
4522 Interpolate the gradients that should go where the bands are, and
4525 This filter is designed for playback only. Do not use it prior to
4526 lossy compression, because compression tends to lose the dither and
4527 bring back the bands.
4529 This filter accepts the following options:
4534 The maximum amount by which the filter will change any one pixel. Also the
4535 threshold for detecting nearly flat regions. Acceptable values range from .51 to
4536 64, default value is 1.2, out-of-range values will be clipped to the valid
4540 The neighborhood to fit the gradient to. A larger radius makes for smoother
4541 gradients, but also prevents the filter from modifying the pixels near detailed
4542 regions. Acceptable values are 8-32, default value is 16, out-of-range values
4543 will be clipped to the valid range.
4547 Alternatively, the options can be specified as a flat string:
4548 @var{strength}[:@var{radius}]
4550 @subsection Examples
4554 Apply the filter with a @code{3.5} strength and radius of @code{8}:
4560 Specify radius, omitting the strength (which will fall-back to the default
4571 Apply a Hald CLUT to a video stream.
4573 First input is the video stream to process, and second one is the Hald CLUT.
4574 The Hald CLUT input can be a simple picture or a complete video stream.
4576 The filter accepts the following options:
4580 Force termination when the shortest input terminates. Default is @code{0}.
4582 Continue applying the last CLUT after the end of the stream. A value of
4583 @code{0} disable the filter after the last frame of the CLUT is reached.
4584 Default is @code{1}.
4587 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
4588 filters share the same internals).
4590 More information about the Hald CLUT can be found on Eskil Steenberg's website
4591 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
4593 @subsection Workflow examples
4595 @subsubsection Hald CLUT video stream
4597 Generate an identity Hald CLUT stream altered with various effects:
4599 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
4602 Note: make sure you use a lossless codec.
4604 Then use it with @code{haldclut} to apply it on some random stream:
4606 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
4609 The Hald CLUT will be applied to the 10 first seconds (duration of
4610 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
4611 to the remaining frames of the @code{mandelbrot} stream.
4613 @subsubsection Hald CLUT with preview
4615 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
4616 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
4617 biggest possible square starting at the top left of the picture. The remaining
4618 padding pixels (bottom or right) will be ignored. This area can be used to add
4619 a preview of the Hald CLUT.
4621 Typically, the following generated Hald CLUT will be supported by the
4622 @code{haldclut} filter:
4625 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
4626 pad=iw+320 [padded_clut];
4627 smptebars=s=320x256, split [a][b];
4628 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
4629 [main][b] overlay=W-320" -frames:v 1 clut.png
4632 It contains the original and a preview of the effect of the CLUT: SMPTE color
4633 bars are displayed on the right-top, and below the same color bars processed by
4636 Then, the effect of this Hald CLUT can be visualized with:
4638 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
4643 Flip the input video horizontally.
4645 For example to horizontally flip the input video with @command{ffmpeg}:
4647 ffmpeg -i in.avi -vf "hflip" out.avi
4651 This filter applies a global color histogram equalization on a
4654 It can be used to correct video that has a compressed range of pixel
4655 intensities. The filter redistributes the pixel intensities to
4656 equalize their distribution across the intensity range. It may be
4657 viewed as an "automatically adjusting contrast filter". This filter is
4658 useful only for correcting degraded or poorly captured source
4661 The filter accepts the following options:
4665 Determine the amount of equalization to be applied. As the strength
4666 is reduced, the distribution of pixel intensities more-and-more
4667 approaches that of the input frame. The value must be a float number
4668 in the range [0,1] and defaults to 0.200.
4671 Set the maximum intensity that can generated and scale the output
4672 values appropriately. The strength should be set as desired and then
4673 the intensity can be limited if needed to avoid washing-out. The value
4674 must be a float number in the range [0,1] and defaults to 0.210.
4677 Set the antibanding level. If enabled the filter will randomly vary
4678 the luminance of output pixels by a small amount to avoid banding of
4679 the histogram. Possible values are @code{none}, @code{weak} or
4680 @code{strong}. It defaults to @code{none}.
4685 Compute and draw a color distribution histogram for the input video.
4687 The computed histogram is a representation of distribution of color components
4690 The filter accepts the following options:
4696 It accepts the following values:
4699 standard histogram that display color components distribution in an image.
4700 Displays color graph for each color component. Shows distribution
4701 of the Y, U, V, A or R, G, B components, depending on input format,
4702 in current frame. Bellow each graph is color component scale meter.
4705 chroma values in vectorscope, if brighter more such chroma values are
4706 distributed in an image.
4707 Displays chroma values (U/V color placement) in two dimensional graph
4708 (which is called a vectorscope). It can be used to read of the hue and
4709 saturation of the current frame. At a same time it is a histogram.
4710 The whiter a pixel in the vectorscope, the more pixels of the input frame
4711 correspond to that pixel (that is the more pixels have this chroma value).
4712 The V component is displayed on the horizontal (X) axis, with the leftmost
4713 side being V = 0 and the rightmost side being V = 255.
4714 The U component is displayed on the vertical (Y) axis, with the top
4715 representing U = 0 and the bottom representing U = 255.
4717 The position of a white pixel in the graph corresponds to the chroma value
4718 of a pixel of the input clip. So the graph can be used to read of the
4719 hue (color flavor) and the saturation (the dominance of the hue in the color).
4720 As the hue of a color changes, it moves around the square. At the center of
4721 the square, the saturation is zero, which means that the corresponding pixel
4722 has no color. If you increase the amount of a specific color, while leaving
4723 the other colors unchanged, the saturation increases, and you move towards
4724 the edge of the square.
4727 chroma values in vectorscope, similar as @code{color} but actual chroma values
4731 per row/column color component graph. In row mode graph in the left side represents
4732 color component value 0 and right side represents value = 255. In column mode top
4733 side represents color component value = 0 and bottom side represents value = 255.
4735 Default value is @code{levels}.
4738 Set height of level in @code{levels}. Default value is @code{200}.
4739 Allowed range is [50, 2048].
4742 Set height of color scale in @code{levels}. Default value is @code{12}.
4743 Allowed range is [0, 40].
4746 Set step for @code{waveform} mode. Smaller values are useful to find out how much
4747 of same luminance values across input rows/columns are distributed.
4748 Default value is @code{10}. Allowed range is [1, 255].
4751 Set mode for @code{waveform}. Can be either @code{row}, or @code{column}.
4752 Default is @code{row}.
4754 @item waveform_mirror
4755 Set mirroring mode for @code{waveform}. @code{0} means unmirrored, @code{1}
4756 means mirrored. In mirrored mode, higher values will be represented on the left
4757 side for @code{row} mode and at the top for @code{column} mode. Default is
4758 @code{0} (unmirrored).
4761 Set display mode for @code{waveform} and @code{levels}.
4762 It accepts the following values:
4765 Display separate graph for the color components side by side in
4766 @code{row} waveform mode or one below other in @code{column} waveform mode
4767 for @code{waveform} histogram mode. For @code{levels} histogram mode
4768 per color component graphs are placed one bellow other.
4770 This display mode in @code{waveform} histogram mode makes it easy to spot
4771 color casts in the highlights and shadows of an image, by comparing the
4772 contours of the top and the bottom of each waveform.
4773 Since whites, grays, and blacks are characterized by
4774 exactly equal amounts of red, green, and blue, neutral areas of the
4775 picture should display three waveforms of roughly equal width/height.
4776 If not, the correction is easy to make by making adjustments to level the
4780 Presents information that's identical to that in the @code{parade}, except
4781 that the graphs representing color components are superimposed directly
4784 This display mode in @code{waveform} histogram mode can make it easier to spot
4785 the relative differences or similarities in overlapping areas of the color
4786 components that are supposed to be identical, such as neutral whites, grays,
4789 Default is @code{parade}.
4792 Set mode for @code{levels}. Can be either @code{linear}, or @code{logarithmic}.
4793 Default is @code{linear}.
4796 @subsection Examples
4801 Calculate and draw histogram:
4803 ffplay -i input -vf histogram
4811 High precision/quality 3d denoise filter. This filter aims to reduce
4812 image noise producing smooth images and making still images really
4813 still. It should enhance compressibility.
4815 It accepts the following optional parameters:
4819 a non-negative float number which specifies spatial luma strength,
4822 @item chroma_spatial
4823 a non-negative float number which specifies spatial chroma strength,
4824 defaults to 3.0*@var{luma_spatial}/4.0
4827 a float number which specifies luma temporal strength, defaults to
4828 6.0*@var{luma_spatial}/4.0
4831 a float number which specifies chroma temporal strength, defaults to
4832 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}
4837 Modify the hue and/or the saturation of the input.
4839 This filter accepts the following options:
4843 Specify the hue angle as a number of degrees. It accepts an expression,
4844 and defaults to "0".
4847 Specify the saturation in the [-10,10] range. It accepts an expression and
4851 Specify the hue angle as a number of radians. It accepts an
4852 expression, and defaults to "0".
4855 Specify the brightness in the [-10,10] range. It accepts an expression and
4859 @option{h} and @option{H} are mutually exclusive, and can't be
4860 specified at the same time.
4862 The @option{b}, @option{h}, @option{H} and @option{s} option values are
4863 expressions containing the following constants:
4867 frame count of the input frame starting from 0
4870 presentation timestamp of the input frame expressed in time base units
4873 frame rate of the input video, NAN if the input frame rate is unknown
4876 timestamp expressed in seconds, NAN if the input timestamp is unknown
4879 time base of the input video
4882 @subsection Examples
4886 Set the hue to 90 degrees and the saturation to 1.0:
4892 Same command but expressing the hue in radians:
4898 Rotate hue and make the saturation swing between 0
4899 and 2 over a period of 1 second:
4901 hue="H=2*PI*t: s=sin(2*PI*t)+1"
4905 Apply a 3 seconds saturation fade-in effect starting at 0:
4910 The general fade-in expression can be written as:
4912 hue="s=min(0\, max((t-START)/DURATION\, 1))"
4916 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
4918 hue="s=max(0\, min(1\, (8-t)/3))"
4921 The general fade-out expression can be written as:
4923 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
4928 @subsection Commands
4930 This filter supports the following commands:
4936 Modify the hue and/or the saturation and/or brightness of the input video.
4937 The command accepts the same syntax of the corresponding option.
4939 If the specified expression is not valid, it is kept at its current
4945 Detect video interlacing type.
4947 This filter tries to detect if the input is interlaced or progressive,
4948 top or bottom field first.
4950 The filter accepts the following options:
4954 Set interlacing threshold.
4956 Set progressive threshold.
4961 Deinterleave or interleave fields.
4963 This filter allows to process interlaced images fields without
4964 deinterlacing them. Deinterleaving splits the input frame into 2
4965 fields (so called half pictures). Odd lines are moved to the top
4966 half of the output image, even lines to the bottom half.
4967 You can process (filter) them independently and then re-interleave them.
4969 The filter accepts the following options:
4973 @item chroma_mode, c
4975 Available values for @var{luma_mode}, @var{chroma_mode} and
4976 @var{alpha_mode} are:
4982 @item deinterleave, d
4983 Deinterleave fields, placing one above the other.
4986 Interleave fields. Reverse the effect of deinterleaving.
4988 Default value is @code{none}.
4991 @item chroma_swap, cs
4992 @item alpha_swap, as
4993 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
4998 Simple interlacing filter from progressive contents. This interleaves upper (or
4999 lower) lines from odd frames with lower (or upper) lines from even frames,
5000 halving the frame rate and preserving image height.
5003 Original Original New Frame
5004 Frame 'j' Frame 'j+1' (tff)
5005 ========== =========== ==================
5006 Line 0 --------------------> Frame 'j' Line 0
5007 Line 1 Line 1 ----> Frame 'j+1' Line 1
5008 Line 2 ---------------------> Frame 'j' Line 2
5009 Line 3 Line 3 ----> Frame 'j+1' Line 3
5011 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
5014 It accepts the following optional parameters:
5018 determines whether the interlaced frame is taken from the even (tff - default)
5019 or odd (bff) lines of the progressive frame.
5022 Enable (default) or disable the vertical lowpass filter to avoid twitter
5023 interlacing and reduce moire patterns.
5028 Deinterlace input video by applying Donald Graft's adaptive kernel
5029 deinterling. Work on interlaced parts of a video to produce
5032 The description of the accepted parameters follows.
5036 Set the threshold which affects the filter's tolerance when
5037 determining if a pixel line must be processed. It must be an integer
5038 in the range [0,255] and defaults to 10. A value of 0 will result in
5039 applying the process on every pixels.
5042 Paint pixels exceeding the threshold value to white if set to 1.
5046 Set the fields order. Swap fields if set to 1, leave fields alone if
5050 Enable additional sharpening if set to 1. Default is 0.
5053 Enable twoway sharpening if set to 1. Default is 0.
5056 @subsection Examples
5060 Apply default values:
5062 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
5066 Enable additional sharpening:
5072 Paint processed pixels in white:
5081 Apply a 3D LUT to an input video.
5083 The filter accepts the following options:
5087 Set the 3D LUT file name.
5089 Currently supported formats:
5101 Select interpolation mode.
5103 Available values are:
5107 Use values from the nearest defined point.
5109 Interpolate values using the 8 points defining a cube.
5111 Interpolate values using a tetrahedron.
5115 @section lut, lutrgb, lutyuv
5117 Compute a look-up table for binding each pixel component input value
5118 to an output value, and apply it to input video.
5120 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
5121 to an RGB input video.
5123 These filters accept the following options:
5126 set first pixel component expression
5128 set second pixel component expression
5130 set third pixel component expression
5132 set fourth pixel component expression, corresponds to the alpha component
5135 set red component expression
5137 set green component expression
5139 set blue component expression
5141 alpha component expression
5144 set Y/luminance component expression
5146 set U/Cb component expression
5148 set V/Cr component expression
5151 Each of them specifies the expression to use for computing the lookup table for
5152 the corresponding pixel component values.
5154 The exact component associated to each of the @var{c*} options depends on the
5157 The @var{lut} filter requires either YUV or RGB pixel formats in input,
5158 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
5160 The expressions can contain the following constants and functions:
5165 the input width and height
5168 input value for the pixel component
5171 the input value clipped in the @var{minval}-@var{maxval} range
5174 maximum value for the pixel component
5177 minimum value for the pixel component
5180 the negated value for the pixel component value clipped in the
5181 @var{minval}-@var{maxval} range , it corresponds to the expression
5182 "maxval-clipval+minval"
5185 the computed value in @var{val} clipped in the
5186 @var{minval}-@var{maxval} range
5188 @item gammaval(gamma)
5189 the computed gamma correction value of the pixel component value
5190 clipped in the @var{minval}-@var{maxval} range, corresponds to the
5192 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
5196 All expressions default to "val".
5198 @subsection Examples
5204 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
5205 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
5208 The above is the same as:
5210 lutrgb="r=negval:g=negval:b=negval"
5211 lutyuv="y=negval:u=negval:v=negval"
5221 Remove chroma components, turns the video into a graytone image:
5223 lutyuv="u=128:v=128"
5227 Apply a luma burning effect:
5233 Remove green and blue components:
5239 Set a constant alpha channel value on input:
5241 format=rgba,lutrgb=a="maxval-minval/2"
5245 Correct luminance gamma by a 0.5 factor:
5247 lutyuv=y=gammaval(0.5)
5251 Discard least significant bits of luma:
5253 lutyuv=y='bitand(val, 128+64+32)'
5259 Apply motion-compensation deinterlacing.
5261 It needs one field per frame as input and must thus be used together
5262 with yadif=1/3 or equivalent.
5264 This filter accepts the following options:
5267 Set the deinterlacing mode.
5269 It accepts one of the following values:
5274 use iterative motion estimation
5276 like @samp{slow}, but use multiple reference frames.
5278 Default value is @samp{fast}.
5281 Set the picture field parity assumed for the input video. It must be
5282 one of the following values:
5286 assume top field first
5288 assume bottom field first
5291 Default value is @samp{bff}.
5294 Set per-block quantization parameter (QP) used by the internal
5297 Higher values should result in a smoother motion vector field but less
5298 optimal individual vectors. Default value is 1.
5303 Apply an MPlayer filter to the input video.
5305 This filter provides a wrapper around some of the filters of
5308 This wrapper is considered experimental. Some of the wrapped filters
5309 may not work properly and we may drop support for them, as they will
5310 be implemented natively into FFmpeg. Thus you should avoid
5311 depending on them when writing portable scripts.
5313 The filter accepts the parameters:
5314 @var{filter_name}[:=]@var{filter_params}
5316 @var{filter_name} is the name of a supported MPlayer filter,
5317 @var{filter_params} is a string containing the parameters accepted by
5320 The list of the currently supported filters follows:
5331 The parameter syntax and behavior for the listed filters are the same
5332 of the corresponding MPlayer filters. For detailed instructions check
5333 the "VIDEO FILTERS" section in the MPlayer manual.
5335 @subsection Examples
5339 Adjust gamma, brightness, contrast:
5345 See also mplayer(1), @url{http://www.mplayerhq.hu/}.
5349 Drop frames that do not differ greatly from the previous frame in
5350 order to reduce frame rate.
5352 The main use of this filter is for very-low-bitrate encoding
5353 (e.g. streaming over dialup modem), but it could in theory be used for
5354 fixing movies that were inverse-telecined incorrectly.
5356 A description of the accepted options follows.
5360 Set the maximum number of consecutive frames which can be dropped (if
5361 positive), or the minimum interval between dropped frames (if
5362 negative). If the value is 0, the frame is dropped unregarding the
5363 number of previous sequentially dropped frames.
5370 Set the dropping threshold values.
5372 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
5373 represent actual pixel value differences, so a threshold of 64
5374 corresponds to 1 unit of difference for each pixel, or the same spread
5375 out differently over the block.
5377 A frame is a candidate for dropping if no 8x8 blocks differ by more
5378 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
5379 meaning the whole image) differ by more than a threshold of @option{lo}.
5381 Default value for @option{hi} is 64*12, default value for @option{lo} is
5382 64*5, and default value for @option{frac} is 0.33.
5390 This filter accepts an integer in input, if non-zero it negates the
5391 alpha component (if available). The default value in input is 0.
5395 Force libavfilter not to use any of the specified pixel formats for the
5396 input to the next filter.
5398 This filter accepts the following parameters:
5402 A '|'-separated list of pixel format names, for example
5403 "pix_fmts=yuv420p|monow|rgb24".
5407 @subsection Examples
5411 Force libavfilter to use a format different from @var{yuv420p} for the
5412 input to the vflip filter:
5414 noformat=pix_fmts=yuv420p,vflip
5418 Convert the input video to any of the formats not contained in the list:
5420 noformat=yuv420p|yuv444p|yuv410p
5426 Add noise on video input frame.
5428 The filter accepts the following options:
5436 Set noise seed for specific pixel component or all pixel components in case
5437 of @var{all_seed}. Default value is @code{123457}.
5439 @item all_strength, alls
5440 @item c0_strength, c0s
5441 @item c1_strength, c1s
5442 @item c2_strength, c2s
5443 @item c3_strength, c3s
5444 Set noise strength for specific pixel component or all pixel components in case
5445 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
5447 @item all_flags, allf
5452 Set pixel component flags or set flags for all components if @var{all_flags}.
5453 Available values for component flags are:
5456 averaged temporal noise (smoother)
5458 mix random noise with a (semi)regular pattern
5460 temporal noise (noise pattern changes between frames)
5462 uniform noise (gaussian otherwise)
5466 @subsection Examples
5468 Add temporal and uniform noise to input video:
5470 noise=alls=20:allf=t+u
5475 Pass the video source unchanged to the output.
5479 Apply video transform using libopencv.
5481 To enable this filter install libopencv library and headers and
5482 configure FFmpeg with @code{--enable-libopencv}.
5484 This filter accepts the following parameters:
5489 The name of the libopencv filter to apply.
5492 The parameters to pass to the libopencv filter. If not specified the default
5497 Refer to the official libopencv documentation for more precise
5499 @url{http://opencv.willowgarage.com/documentation/c/image_filtering.html}
5501 Follows the list of supported libopencv filters.
5506 Dilate an image by using a specific structuring element.
5507 This filter corresponds to the libopencv function @code{cvDilate}.
5509 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
5511 @var{struct_el} represents a structuring element, and has the syntax:
5512 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
5514 @var{cols} and @var{rows} represent the number of columns and rows of
5515 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
5516 point, and @var{shape} the shape for the structuring element, and
5517 can be one of the values "rect", "cross", "ellipse", "custom".
5519 If the value for @var{shape} is "custom", it must be followed by a
5520 string of the form "=@var{filename}". The file with name
5521 @var{filename} is assumed to represent a binary image, with each
5522 printable character corresponding to a bright pixel. When a custom
5523 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
5524 or columns and rows of the read file are assumed instead.
5526 The default value for @var{struct_el} is "3x3+0x0/rect".
5528 @var{nb_iterations} specifies the number of times the transform is
5529 applied to the image, and defaults to 1.
5531 Follow some example:
5533 # use the default values
5536 # dilate using a structuring element with a 5x5 cross, iterate two times
5537 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
5539 # read the shape from the file diamond.shape, iterate two times
5540 # the file diamond.shape may contain a pattern of characters like this:
5546 # the specified cols and rows are ignored (but not the anchor point coordinates)
5547 ocv=dilate:0x0+2x2/custom=diamond.shape|2
5552 Erode an image by using a specific structuring element.
5553 This filter corresponds to the libopencv function @code{cvErode}.
5555 The filter accepts the parameters: @var{struct_el}:@var{nb_iterations},
5556 with the same syntax and semantics as the @ref{dilate} filter.
5560 Smooth the input video.
5562 The filter takes the following parameters:
5563 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
5565 @var{type} is the type of smooth filter to apply, and can be one of
5566 the following values: "blur", "blur_no_scale", "median", "gaussian",
5567 "bilateral". The default value is "gaussian".
5569 @var{param1}, @var{param2}, @var{param3}, and @var{param4} are
5570 parameters whose meanings depend on smooth type. @var{param1} and
5571 @var{param2} accept integer positive values or 0, @var{param3} and
5572 @var{param4} accept float values.
5574 The default value for @var{param1} is 3, the default value for the
5575 other parameters is 0.
5577 These parameters correspond to the parameters assigned to the
5578 libopencv function @code{cvSmooth}.
5583 Overlay one video on top of another.
5585 It takes two inputs and one output, the first input is the "main"
5586 video on which the second input is overlayed.
5588 This filter accepts the following parameters:
5590 A description of the accepted options follows.
5595 Set the expression for the x and y coordinates of the overlayed video
5596 on the main video. Default value is "0" for both expressions. In case
5597 the expression is invalid, it is set to a huge value (meaning that the
5598 overlay will not be displayed within the output visible area).
5601 Set when the expressions for @option{x}, and @option{y} are evaluated.
5603 It accepts the following values:
5606 only evaluate expressions once during the filter initialization or
5607 when a command is processed
5610 evaluate expressions for each incoming frame
5613 Default value is @samp{frame}.
5616 If set to 1, force the output to terminate when the shortest input
5617 terminates. Default value is 0.
5620 Set the format for the output video.
5622 It accepts the following values:
5634 Default value is @samp{yuv420}.
5636 @item rgb @emph{(deprecated)}
5637 If set to 1, force the filter to accept inputs in the RGB
5638 color space. Default value is 0. This option is deprecated, use
5639 @option{format} instead.
5642 If set to 1, force the filter to draw the last overlay frame over the
5643 main input until the end of the stream. A value of 0 disables this
5644 behavior. Default value is 1.
5647 The @option{x}, and @option{y} expressions can contain the following
5653 main input width and height
5657 overlay input width and height
5661 the computed values for @var{x} and @var{y}. They are evaluated for
5666 horizontal and vertical chroma subsample values of the output
5667 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
5671 the number of input frame, starting from 0
5674 the position in the file of the input frame, NAN if unknown
5677 timestamp expressed in seconds, NAN if the input timestamp is unknown
5680 Note that the @var{n}, @var{pos}, @var{t} variables are available only
5681 when evaluation is done @emph{per frame}, and will evaluate to NAN
5682 when @option{eval} is set to @samp{init}.
5684 Be aware that frames are taken from each input video in timestamp
5685 order, hence, if their initial timestamps differ, it is a good idea
5686 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
5687 have them begin in the same zero timestamp, as it does the example for
5688 the @var{movie} filter.
5690 You can chain together more overlays but you should test the
5691 efficiency of such approach.
5693 @subsection Commands
5695 This filter supports the following commands:
5699 Modify the x and y of the overlay input.
5700 The command accepts the same syntax of the corresponding option.
5702 If the specified expression is not valid, it is kept at its current
5706 @subsection Examples
5710 Draw the overlay at 10 pixels from the bottom right corner of the main
5713 overlay=main_w-overlay_w-10:main_h-overlay_h-10
5716 Using named options the example above becomes:
5718 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
5722 Insert a transparent PNG logo in the bottom left corner of the input,
5723 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
5725 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
5729 Insert 2 different transparent PNG logos (second logo on bottom
5730 right corner) using the @command{ffmpeg} tool:
5732 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
5736 Add a transparent color layer on top of the main video, @code{WxH}
5737 must specify the size of the main input to the overlay filter:
5739 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
5743 Play an original video and a filtered version (here with the deshake
5744 filter) side by side using the @command{ffplay} tool:
5746 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
5749 The above command is the same as:
5751 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
5755 Make a sliding overlay appearing from the left to the right top part of the
5756 screen starting since time 2:
5758 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
5762 Compose output by putting two input videos side to side:
5764 ffmpeg -i left.avi -i right.avi -filter_complex "
5765 nullsrc=size=200x100 [background];
5766 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
5767 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
5768 [background][left] overlay=shortest=1 [background+left];
5769 [background+left][right] overlay=shortest=1:x=100 [left+right]
5774 Chain several overlays in cascade:
5776 nullsrc=s=200x200 [bg];
5777 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
5778 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
5779 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
5780 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
5781 [in3] null, [mid2] overlay=100:100 [out0]
5788 Apply Overcomplete Wavelet denoiser.
5790 The filter accepts the following options:
5796 Larger depth values will denoise lower frequency components more, but
5797 slow down filtering.
5799 Must be an int in the range 8-16, default is @code{8}.
5801 @item luma_strength, ls
5804 Must be a double value in the range 0-1000, default is @code{1.0}.
5806 @item chroma_strength, cs
5807 Set chroma strength.
5809 Must be a double value in the range 0-1000, default is @code{1.0}.
5814 Add paddings to the input image, and place the original input at the
5815 given coordinates @var{x}, @var{y}.
5817 This filter accepts the following parameters:
5822 Specify an expression for the size of the output image with the
5823 paddings added. If the value for @var{width} or @var{height} is 0, the
5824 corresponding input size is used for the output.
5826 The @var{width} expression can reference the value set by the
5827 @var{height} expression, and vice versa.
5829 The default value of @var{width} and @var{height} is 0.
5833 Specify an expression for the offsets where to place the input image
5834 in the padded area with respect to the top/left border of the output
5837 The @var{x} expression can reference the value set by the @var{y}
5838 expression, and vice versa.
5840 The default value of @var{x} and @var{y} is 0.
5843 Specify the color of the padded area, it can be the name of a color
5844 (case insensitive match) or a 0xRRGGBB[AA] sequence.
5846 The default value of @var{color} is "black".
5849 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
5850 options are expressions containing the following constants:
5855 the input video width and height
5859 same as @var{in_w} and @var{in_h}
5863 the output width and height, that is the size of the padded area as
5864 specified by the @var{width} and @var{height} expressions
5868 same as @var{out_w} and @var{out_h}
5872 x and y offsets as specified by the @var{x} and @var{y}
5873 expressions, or NAN if not yet specified
5876 same as @var{iw} / @var{ih}
5879 input sample aspect ratio
5882 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5886 horizontal and vertical chroma subsample values. For example for the
5887 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5890 @subsection Examples
5894 Add paddings with color "violet" to the input video. Output video
5895 size is 640x480, the top-left corner of the input video is placed at
5898 pad=640:480:0:40:violet
5901 The example above is equivalent to the following command:
5903 pad=width=640:height=480:x=0:y=40:color=violet
5907 Pad the input to get an output with dimensions increased by 3/2,
5908 and put the input video at the center of the padded area:
5910 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
5914 Pad the input to get a squared output with size equal to the maximum
5915 value between the input width and height, and put the input video at
5916 the center of the padded area:
5918 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
5922 Pad the input to get a final w/h ratio of 16:9:
5924 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
5928 In case of anamorphic video, in order to set the output display aspect
5929 correctly, it is necessary to use @var{sar} in the expression,
5930 according to the relation:
5932 (ih * X / ih) * sar = output_dar
5933 X = output_dar / sar
5936 Thus the previous example needs to be modified to:
5938 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
5942 Double output size and put the input video in the bottom-right
5943 corner of the output padded area:
5945 pad="2*iw:2*ih:ow-iw:oh-ih"
5949 @section perspective
5951 Correct perspective of video not recorded perpendicular to the screen.
5953 A description of the accepted parameters follows.
5964 Set coordinates expression for top left, top right, bottom left and bottom right corners.
5965 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
5967 The expressions can use the following variables:
5972 the width and height of video frame.
5976 Set interpolation for perspective correction.
5978 It accepts the following values:
5984 Default value is @samp{linear}.
5989 Delay interlaced video by one field time so that the field order changes.
5991 The intended use is to fix PAL movies that have been captured with the
5992 opposite field order to the film-to-video transfer.
5994 A description of the accepted parameters follows.
6000 It accepts the following values:
6003 Capture field order top-first, transfer bottom-first.
6004 Filter will delay the bottom field.
6007 Capture field order bottom-first, transfer top-first.
6008 Filter will delay the top field.
6011 Capture and transfer with the same field order. This mode only exists
6012 for the documentation of the other options to refer to, but if you
6013 actually select it, the filter will faithfully do nothing.
6016 Capture field order determined automatically by field flags, transfer
6018 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
6019 basis using field flags. If no field information is available,
6020 then this works just like @samp{u}.
6023 Capture unknown or varying, transfer opposite.
6024 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
6025 analyzing the images and selecting the alternative that produces best
6026 match between the fields.
6029 Capture top-first, transfer unknown or varying.
6030 Filter selects among @samp{t} and @samp{p} using image analysis.
6033 Capture bottom-first, transfer unknown or varying.
6034 Filter selects among @samp{b} and @samp{p} using image analysis.
6037 Capture determined by field flags, transfer unknown or varying.
6038 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
6039 image analysis. If no field information is available, then this works just
6040 like @samp{U}. This is the default mode.
6043 Both capture and transfer unknown or varying.
6044 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
6048 @section pixdesctest
6050 Pixel format descriptor test filter, mainly useful for internal
6051 testing. The output video should be equal to the input video.
6055 format=monow, pixdesctest
6058 can be used to test the monowhite pixel format descriptor definition.
6062 Enable the specified chain of postprocessing subfilters using libpostproc. This
6063 library should be automatically selected with a GPL build (@code{--enable-gpl}).
6064 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
6065 Each subfilter and some options have a short and a long name that can be used
6066 interchangeably, i.e. dr/dering are the same.
6068 The filters accept the following options:
6072 Set postprocessing subfilters string.
6075 All subfilters share common options to determine their scope:
6079 Honor the quality commands for this subfilter.
6082 Do chrominance filtering, too (default).
6085 Do luminance filtering only (no chrominance).
6088 Do chrominance filtering only (no luminance).
6091 These options can be appended after the subfilter name, separated by a '|'.
6093 Available subfilters are:
6096 @item hb/hdeblock[|difference[|flatness]]
6097 Horizontal deblocking filter
6100 Difference factor where higher values mean more deblocking (default: @code{32}).
6102 Flatness threshold where lower values mean more deblocking (default: @code{39}).
6105 @item vb/vdeblock[|difference[|flatness]]
6106 Vertical deblocking filter
6109 Difference factor where higher values mean more deblocking (default: @code{32}).
6111 Flatness threshold where lower values mean more deblocking (default: @code{39}).
6114 @item ha/hadeblock[|difference[|flatness]]
6115 Accurate horizontal deblocking filter
6118 Difference factor where higher values mean more deblocking (default: @code{32}).
6120 Flatness threshold where lower values mean more deblocking (default: @code{39}).
6123 @item va/vadeblock[|difference[|flatness]]
6124 Accurate vertical deblocking filter
6127 Difference factor where higher values mean more deblocking (default: @code{32}).
6129 Flatness threshold where lower values mean more deblocking (default: @code{39}).
6133 The horizontal and vertical deblocking filters share the difference and
6134 flatness values so you cannot set different horizontal and vertical
6139 Experimental horizontal deblocking filter
6142 Experimental vertical deblocking filter
6147 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
6150 larger -> stronger filtering
6152 larger -> stronger filtering
6154 larger -> stronger filtering
6157 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
6160 Stretch luminance to @code{0-255}.
6163 @item lb/linblenddeint
6164 Linear blend deinterlacing filter that deinterlaces the given block by
6165 filtering all lines with a @code{(1 2 1)} filter.
6167 @item li/linipoldeint
6168 Linear interpolating deinterlacing filter that deinterlaces the given block by
6169 linearly interpolating every second line.
6171 @item ci/cubicipoldeint
6172 Cubic interpolating deinterlacing filter deinterlaces the given block by
6173 cubically interpolating every second line.
6175 @item md/mediandeint
6176 Median deinterlacing filter that deinterlaces the given block by applying a
6177 median filter to every second line.
6179 @item fd/ffmpegdeint
6180 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
6181 second line with a @code{(-1 4 2 4 -1)} filter.
6184 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
6185 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
6187 @item fq/forceQuant[|quantizer]
6188 Overrides the quantizer table from the input with the constant quantizer you
6196 Default pp filter combination (@code{hb|a,vb|a,dr|a})
6199 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
6202 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
6205 @subsection Examples
6209 Apply horizontal and vertical deblocking, deringing and automatic
6210 brightness/contrast:
6216 Apply default filters without brightness/contrast correction:
6222 Apply default filters and temporal denoiser:
6224 pp=default/tmpnoise|1|2|3
6228 Apply deblocking on luminance only, and switch vertical deblocking on or off
6229 automatically depending on available CPU time:
6237 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
6238 Ratio) between two input videos.
6240 This filter takes in input two input videos, the first input is
6241 considered the "main" source and is passed unchanged to the
6242 output. The second input is used as a "reference" video for computing
6245 Both video inputs must have the same resolution and pixel format for
6246 this filter to work correctly. Also it assumes that both inputs
6247 have the same number of frames, which are compared one by one.
6249 The obtained average PSNR is printed through the logging system.
6251 The filter stores the accumulated MSE (mean squared error) of each
6252 frame, and at the end of the processing it is averaged across all frames
6253 equally, and the following formula is applied to obtain the PSNR:
6256 PSNR = 10*log10(MAX^2/MSE)
6259 Where MAX is the average of the maximum values of each component of the
6262 The description of the accepted parameters follows.
6266 If specified the filter will use the named file to save the PSNR of
6267 each individual frame.
6270 The file printed if @var{stats_file} is selected, contains a sequence of
6271 key/value pairs of the form @var{key}:@var{value} for each compared
6274 A description of each shown parameter follows:
6278 sequential number of the input frame, starting from 1
6281 Mean Square Error pixel-by-pixel average difference of the compared
6282 frames, averaged over all the image components.
6284 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
6285 Mean Square Error pixel-by-pixel average difference of the compared
6286 frames for the component specified by the suffix.
6288 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
6289 Peak Signal to Noise ratio of the compared frames for the component
6290 specified by the suffix.
6295 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
6296 [main][ref] psnr="stats_file=stats.log" [out]
6299 On this example the input file being processed is compared with the
6300 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
6301 is stored in @file{stats.log}.
6305 Pulldown reversal (inverse telecine) filter, capable of handling mixed
6306 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
6309 The pullup filter is designed to take advantage of future context in making
6310 its decisions. This filter is stateless in the sense that it does not lock
6311 onto a pattern to follow, but it instead looks forward to the following
6312 fields in order to identify matches and rebuild progressive frames.
6314 To produce content with an even framerate, insert the fps filter after
6315 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
6316 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
6318 The filter accepts the following options:
6325 These options set the amount of "junk" to ignore at the left, right, top, and
6326 bottom of the image, respectively. Left and right are in units of 8 pixels,
6327 while top and bottom are in units of 2 lines.
6328 The default is 8 pixels on each side.
6331 Set the strict breaks. Setting this option to 1 will reduce the chances of
6332 filter generating an occasional mismatched frame, but it may also cause an
6333 excessive number of frames to be dropped during high motion sequences.
6334 Conversely, setting it to -1 will make filter match fields more easily.
6335 This may help processing of video where there is slight blurring between
6336 the fields, but may also cause there to be interlaced frames in the output.
6337 Default value is @code{0}.
6340 Set the metric plane to use. It accepts the following values:
6346 Use chroma blue plane.
6349 Use chroma red plane.
6352 This option may be set to use chroma plane instead of the default luma plane
6353 for doing filter's computations. This may improve accuracy on very clean
6354 source material, but more likely will decrease accuracy, especially if there
6355 is chroma noise (rainbow effect) or any grayscale video.
6356 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
6357 load and make pullup usable in realtime on slow machines.
6360 For example to inverse telecined NTSC input:
6362 pullup,fps=24000/1001
6367 Suppress a TV station logo, using an image file to determine which
6368 pixels comprise the logo. It works by filling in the pixels that
6369 comprise the logo with neighboring pixels.
6371 The filter accepts the following options:
6375 Set the filter bitmap file, which can be any image format supported by
6376 libavformat. The width and height of the image file must match those of the
6377 video stream being processed.
6380 Pixels in the provided bitmap image with a value of zero are not
6381 considered part of the logo, non-zero pixels are considered part of
6382 the logo. If you use white (255) for the logo and black (0) for the
6383 rest, you will be safe. For making the filter bitmap, it is
6384 recommended to take a screen capture of a black frame with the logo
6385 visible, and then using a threshold filter followed by the erode
6386 filter once or twice.
6388 If needed, little splotches can be fixed manually. Remember that if
6389 logo pixels are not covered, the filter quality will be much
6390 reduced. Marking too many pixels as part of the logo does not hurt as
6391 much, but it will increase the amount of blurring needed to cover over
6392 the image and will destroy more information than necessary, and extra
6393 pixels will slow things down on a large logo.
6397 Rotate video by an arbitrary angle expressed in radians.
6399 The filter accepts the following options:
6401 A description of the optional parameters follows.
6404 Set an expression for the angle by which to rotate the input video
6405 clockwise, expressed as a number of radians. A negative value will
6406 result in a counter-clockwise rotation. By default it is set to "0".
6408 This expression is evaluated for each frame.
6411 Set the output width expression, default value is "iw".
6412 This expression is evaluated just once during configuration.
6415 Set the output height expression, default value is "ih".
6416 This expression is evaluated just once during configuration.
6419 Enable bilinear interpolation if set to 1, a value of 0 disables
6420 it. Default value is 1.
6423 Set the color used to fill the output area not covered by the rotated
6424 image. If the special value "none" is selected then no background is
6425 printed (useful for example if the background is never shown). Default
6429 The expressions for the angle and the output size can contain the
6430 following constants and functions:
6434 sequential number of the input frame, starting from 0. It is always NAN
6435 before the first frame is filtered.
6438 time in seconds of the input frame, it is set to 0 when the filter is
6439 configured. It is always NAN before the first frame is filtered.
6443 horizontal and vertical chroma subsample values. For example for the
6444 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6448 the input video width and heigth
6452 the output width and heigth, that is the size of the padded area as
6453 specified by the @var{width} and @var{height} expressions
6457 the minimal width/height required for completely containing the input
6458 video rotated by @var{a} radians.
6460 These are only available when computing the @option{out_w} and
6461 @option{out_h} expressions.
6464 @subsection Examples
6468 Rotate the input by PI/6 radians clockwise:
6474 Rotate the input by PI/6 radians counter-clockwise:
6480 Apply a constant rotation with period T, starting from an angle of PI/3:
6482 rotate=PI/3+2*PI*t/T
6486 Make the input video rotation oscillating with a period of T
6487 seconds and an amplitude of A radians:
6489 rotate=A*sin(2*PI/T*t)
6493 Rotate the video, output size is choosen so that the whole rotating
6494 input video is always completely contained in the output:
6496 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
6500 Rotate the video, reduce the output size so that no background is ever
6503 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
6507 @subsection Commands
6509 The filter supports the following commands:
6513 Set the angle expression.
6514 The command accepts the same syntax of the corresponding option.
6516 If the specified expression is not valid, it is kept at its current
6522 Apply Shape Adaptive Blur.
6524 The filter accepts the following options:
6527 @item luma_radius, lr
6528 Set luma blur filter strength, must be a value in range 0.1-4.0, default
6529 value is 1.0. A greater value will result in a more blurred image, and
6530 in slower processing.
6532 @item luma_pre_filter_radius, lpfr
6533 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
6536 @item luma_strength, ls
6537 Set luma maximum difference between pixels to still be considered, must
6538 be a value in the 0.1-100.0 range, default value is 1.0.
6540 @item chroma_radius, cr
6541 Set chroma blur filter strength, must be a value in range 0.1-4.0. A
6542 greater value will result in a more blurred image, and in slower
6545 @item chroma_pre_filter_radius, cpfr
6546 Set chroma pre-filter radius, must be a value in the 0.1-2.0 range.
6548 @item chroma_strength, cs
6549 Set chroma maximum difference between pixels to still be considered,
6550 must be a value in the 0.1-100.0 range.
6553 Each chroma option value, if not explicitly specified, is set to the
6554 corresponding luma option value.
6558 Scale (resize) the input video, using the libswscale library.
6560 The scale filter forces the output display aspect ratio to be the same
6561 of the input, by changing the output sample aspect ratio.
6563 If the input image format is different from the format requested by
6564 the next filter, the scale filter will convert the input to the
6568 The filter accepts the following options, or any of the options
6569 supported by the libswscale scaler.
6571 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
6572 the complete list of scaler options.
6577 Set the output video dimension expression. Default value is the input
6580 If the value is 0, the input width is used for the output.
6582 If one of the values is -1, the scale filter will use a value that
6583 maintains the aspect ratio of the input image, calculated from the
6584 other specified dimension. If both of them are -1, the input size is
6587 See below for the list of accepted constants for use in the dimension
6591 Set the interlacing mode. It accepts the following values:
6595 Force interlaced aware scaling.
6598 Do not apply interlaced scaling.
6601 Select interlaced aware scaling depending on whether the source frames
6602 are flagged as interlaced or not.
6605 Default value is @samp{0}.
6608 Set libswscale scaling flags. If not explictly specified the filter
6609 applies a bicubic scaling algorithm.
6610 See @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for
6611 the complete list of values.
6614 Set the video size, the value must be a valid abbreviation or in the
6615 form @var{width}x@var{height}.
6617 @item in_color_matrix
6618 @item out_color_matrix
6619 Set in/output YCbCr color space type.
6621 This allows the autodetected value to be overridden as well as allows forcing
6622 a specific value used for the output and encoder.
6624 If not specified, the color space type depends on the pixel format.
6630 Choose automatically.
6633 Format conforming to International Telecommunication Union (ITU)
6634 Recommendation BT.709.
6637 Set color space conforming to the United States Federal Communications
6638 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
6641 Set color space conforming to:
6645 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
6648 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
6651 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
6656 Set color space conforming to SMPTE ST 240:1999.
6661 Set in/output YCbCr sample range.
6663 This allows the autodetected value to be overridden as well as allows forcing
6664 a specific value used for the output and encoder. If not specified, the
6665 range depends on the pixel format. Possible values:
6669 Choose automatically.
6672 Set full range (0-255 in case of 8-bit luma).
6675 Set "MPEG" range (16-235 in case of 8-bit luma).
6678 @item force_original_aspect_ratio
6679 Enable decreasing or increasing output video width or height if necessary to
6680 keep the original aspect ratio. Possible values:
6684 Scale the video as specified and disable this feature.
6687 The output video dimensions will automatically be decreased if needed.
6690 The output video dimensions will automatically be increased if needed.
6694 One useful instance of this option is that when you know a specific device's
6695 maximum allowed resolution, you can use this to limit the output video to
6696 that, while retaining the aspect ratio. For example, device A allows
6697 1280x720 playback, and your video is 1920x800. Using this option (set it to
6698 decrease) and specifying 1280x720 to the command line makes the output
6701 Please note that this is a different thing than specifying -1 for @option{w}
6702 or @option{h}, you still need to specify the output resolution for this option
6707 The values of the @option{w} and @option{h} options are expressions
6708 containing the following constants:
6713 the input width and height
6717 same as @var{in_w} and @var{in_h}
6721 the output (scaled) width and height
6725 same as @var{out_w} and @var{out_h}
6728 same as @var{iw} / @var{ih}
6731 input sample aspect ratio
6734 input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
6738 horizontal and vertical chroma subsample values. For example for the
6739 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6742 @subsection Examples
6746 Scale the input video to a size of 200x100:
6751 This is equivalent to:
6762 Specify a size abbreviation for the output size:
6767 which can also be written as:
6773 Scale the input to 2x:
6779 The above is the same as:
6785 Scale the input to 2x with forced interlaced scaling:
6787 scale=2*iw:2*ih:interl=1
6791 Scale the input to half size:
6797 Increase the width, and set the height to the same size:
6803 Seek for Greek harmony:
6810 Increase the height, and set the width to 3/2 of the height:
6812 scale=w=3/2*oh:h=3/5*ih
6816 Increase the size, but make the size a multiple of the chroma
6819 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
6823 Increase the width to a maximum of 500 pixels, keep the same input
6826 scale=w='min(500\, iw*3/2):h=-1'
6830 @section separatefields
6832 The @code{separatefields} takes a frame-based video input and splits
6833 each frame into its components fields, producing a new half height clip
6834 with twice the frame rate and twice the frame count.
6836 This filter use field-dominance information in frame to decide which
6837 of each pair of fields to place first in the output.
6838 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
6840 @section setdar, setsar
6842 The @code{setdar} filter sets the Display Aspect Ratio for the filter
6845 This is done by changing the specified Sample (aka Pixel) Aspect
6846 Ratio, according to the following equation:
6848 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
6851 Keep in mind that the @code{setdar} filter does not modify the pixel
6852 dimensions of the video frame. Also the display aspect ratio set by
6853 this filter may be changed by later filters in the filterchain,
6854 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
6857 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
6858 the filter output video.
6860 Note that as a consequence of the application of this filter, the
6861 output display aspect ratio will change according to the equation
6864 Keep in mind that the sample aspect ratio set by the @code{setsar}
6865 filter may be changed by later filters in the filterchain, e.g. if
6866 another "setsar" or a "setdar" filter is applied.
6868 The filters accept the following options:
6871 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
6872 Set the aspect ratio used by the filter.
6874 The parameter can be a floating point number string, an expression, or
6875 a string of the form @var{num}:@var{den}, where @var{num} and
6876 @var{den} are the numerator and denominator of the aspect ratio. If
6877 the parameter is not specified, it is assumed the value "0".
6878 In case the form "@var{num}:@var{den}" is used, the @code{:} character
6882 Set the maximum integer value to use for expressing numerator and
6883 denominator when reducing the expressed aspect ratio to a rational.
6884 Default value is @code{100}.
6888 @subsection Examples
6893 To change the display aspect ratio to 16:9, specify one of the following:
6901 To change the sample aspect ratio to 10:11, specify:
6907 To set a display aspect ratio of 16:9, and specify a maximum integer value of
6908 1000 in the aspect ratio reduction, use the command:
6910 setdar=ratio=16/9:max=1000
6918 Force field for the output video frame.
6920 The @code{setfield} filter marks the interlace type field for the
6921 output frames. It does not change the input frame, but only sets the
6922 corresponding property, which affects how the frame is treated by
6923 following filters (e.g. @code{fieldorder} or @code{yadif}).
6925 The filter accepts the following options:
6930 Available values are:
6934 Keep the same field property.
6937 Mark the frame as bottom-field-first.
6940 Mark the frame as top-field-first.
6943 Mark the frame as progressive.
6949 Show a line containing various information for each input video frame.
6950 The input video is not modified.
6952 The shown line contains a sequence of key/value pairs of the form
6953 @var{key}:@var{value}.
6955 A description of each shown parameter follows:
6959 sequential number of the input frame, starting from 0
6962 Presentation TimeStamp of the input frame, expressed as a number of
6963 time base units. The time base unit depends on the filter input pad.
6966 Presentation TimeStamp of the input frame, expressed as a number of
6970 position of the frame in the input stream, -1 if this information in
6971 unavailable and/or meaningless (for example in case of synthetic video)
6977 sample aspect ratio of the input frame, expressed in the form
6981 size of the input frame, expressed in the form
6982 @var{width}x@var{height}
6985 interlaced mode ("P" for "progressive", "T" for top field first, "B"
6986 for bottom field first)
6989 1 if the frame is a key frame, 0 otherwise
6992 picture type of the input frame ("I" for an I-frame, "P" for a
6993 P-frame, "B" for a B-frame, "?" for unknown type).
6994 Check also the documentation of the @code{AVPictureType} enum and of
6995 the @code{av_get_picture_type_char} function defined in
6996 @file{libavutil/avutil.h}.
6999 Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame
7001 @item plane_checksum
7002 Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
7003 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]"
7009 Blur the input video without impacting the outlines.
7011 The filter accepts the following options:
7014 @item luma_radius, lr
7015 Set the luma radius. The option value must be a float number in
7016 the range [0.1,5.0] that specifies the variance of the gaussian filter
7017 used to blur the image (slower if larger). Default value is 1.0.
7019 @item luma_strength, ls
7020 Set the luma strength. The option value must be a float number
7021 in the range [-1.0,1.0] that configures the blurring. A value included
7022 in [0.0,1.0] will blur the image whereas a value included in
7023 [-1.0,0.0] will sharpen the image. Default value is 1.0.
7025 @item luma_threshold, lt
7026 Set the luma threshold used as a coefficient to determine
7027 whether a pixel should be blurred or not. The option value must be an
7028 integer in the range [-30,30]. A value of 0 will filter all the image,
7029 a value included in [0,30] will filter flat areas and a value included
7030 in [-30,0] will filter edges. Default value is 0.
7032 @item chroma_radius, cr
7033 Set the chroma radius. The option value must be a float number in
7034 the range [0.1,5.0] that specifies the variance of the gaussian filter
7035 used to blur the image (slower if larger). Default value is 1.0.
7037 @item chroma_strength, cs
7038 Set the chroma strength. The option value must be a float number
7039 in the range [-1.0,1.0] that configures the blurring. A value included
7040 in [0.0,1.0] will blur the image whereas a value included in
7041 [-1.0,0.0] will sharpen the image. Default value is 1.0.
7043 @item chroma_threshold, ct
7044 Set the chroma threshold used as a coefficient to determine
7045 whether a pixel should be blurred or not. The option value must be an
7046 integer in the range [-30,30]. A value of 0 will filter all the image,
7047 a value included in [0,30] will filter flat areas and a value included
7048 in [-30,0] will filter edges. Default value is 0.
7051 If a chroma option is not explicitly set, the corresponding luma value
7056 Convert between different stereoscopic image formats.
7058 The filters accept the following options:
7062 Set stereoscopic image format of input.
7064 Available values for input image formats are:
7067 side by side parallel (left eye left, right eye right)
7070 side by side crosseye (right eye left, left eye right)
7073 side by side parallel with half width resolution
7074 (left eye left, right eye right)
7077 side by side crosseye with half width resolution
7078 (right eye left, left eye right)
7081 above-below (left eye above, right eye below)
7084 above-below (right eye above, left eye below)
7087 above-below with half height resolution
7088 (left eye above, right eye below)
7091 above-below with half height resolution
7092 (right eye above, left eye below)
7095 alternating frames (left eye first, right eye second)
7098 alternating frames (right eye first, left eye second)
7100 Default value is @samp{sbsl}.
7104 Set stereoscopic image format of output.
7106 Available values for output image formats are all the input formats as well as:
7109 anaglyph red/blue gray
7110 (red filter on left eye, blue filter on right eye)
7113 anaglyph red/green gray
7114 (red filter on left eye, green filter on right eye)
7117 anaglyph red/cyan gray
7118 (red filter on left eye, cyan filter on right eye)
7121 anaglyph red/cyan half colored
7122 (red filter on left eye, cyan filter on right eye)
7125 anaglyph red/cyan color
7126 (red filter on left eye, cyan filter on right eye)
7129 anaglyph red/cyan color optimized with the least squares projection of dubois
7130 (red filter on left eye, cyan filter on right eye)
7133 anaglyph green/magenta gray
7134 (green filter on left eye, magenta filter on right eye)
7137 anaglyph green/magenta half colored
7138 (green filter on left eye, magenta filter on right eye)
7141 anaglyph green/magenta colored
7142 (green filter on left eye, magenta filter on right eye)
7145 anaglyph green/magenta color optimized with the least squares projection of dubois
7146 (green filter on left eye, magenta filter on right eye)
7149 anaglyph yellow/blue gray
7150 (yellow filter on left eye, blue filter on right eye)
7153 anaglyph yellow/blue half colored
7154 (yellow filter on left eye, blue filter on right eye)
7157 anaglyph yellow/blue colored
7158 (yellow filter on left eye, blue filter on right eye)
7161 anaglyph yellow/blue color optimized with the least squares projection of dubois
7162 (yellow filter on left eye, blue filter on right eye)
7165 interleaved rows (left eye has top row, right eye starts on next row)
7168 interleaved rows (right eye has top row, left eye starts on next row)
7171 mono output (left eye only)
7174 mono output (right eye only)
7177 Default value is @samp{arcd}.
7180 @subsection Examples
7184 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
7190 Convert input video from above bellow (left eye above, right eye below) to side by side crosseye.
7198 Apply a simple postprocessing filter that compresses and decompresses the image
7199 at several (or - in the case of @option{quality} level @code{6} - all) shifts
7200 and average the results.
7202 The filter accepts the following options:
7206 Set quality. This option defines the number of levels for averaging. It accepts
7207 an integer in the range 0-6. If set to @code{0}, the filter will have no
7208 effect. A value of @code{6} means the higher quality. For each increment of
7209 that value the speed drops by a factor of approximately 2. Default value is
7213 Force a constant quantization parameter. If not set, the filter will use the QP
7214 from the video stream (if available).
7217 Set thresholding mode. Available modes are:
7221 Set hard thresholding (default).
7223 Set soft thresholding (better de-ringing effect, but likely blurrier).
7227 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
7228 option may cause flicker since the B-Frames have often larger QP. Default is
7229 @code{0} (not enabled).
7235 Draw subtitles on top of input video using the libass library.
7237 To enable compilation of this filter you need to configure FFmpeg with
7238 @code{--enable-libass}. This filter also requires a build with libavcodec and
7239 libavformat to convert the passed subtitles file to ASS (Advanced Substation
7240 Alpha) subtitles format.
7242 The filter accepts the following options:
7246 Set the filename of the subtitle file to read. It must be specified.
7249 Specify the size of the original video, the video for which the ASS file
7250 was composed. Due to a misdesign in ASS aspect ratio arithmetic, this is
7251 necessary to correctly scale the fonts if the aspect ratio has been changed.
7254 Set subtitles input character encoding. @code{subtitles} filter only. Only
7255 useful if not UTF-8.
7258 If the first key is not specified, it is assumed that the first value
7259 specifies the @option{filename}.
7261 For example, to render the file @file{sub.srt} on top of the input
7262 video, use the command:
7267 which is equivalent to:
7269 subtitles=filename=sub.srt
7274 Scale the input by 2x and smooth using the Super2xSaI (Scale and
7275 Interpolate) pixel art scaling algorithm.
7277 Useful for enlarging pixel art images without reducing sharpness.
7284 Apply telecine process to the video.
7286 This filter accepts the following options:
7295 The default value is @code{top}.
7299 A string of numbers representing the pulldown pattern you wish to apply.
7300 The default value is @code{23}.
7304 Some typical patterns:
7309 24p: 2332 (preferred)
7316 24p: 222222222223 ("Euro pulldown")
7322 Select the most representative frame in a given sequence of consecutive frames.
7324 The filter accepts the following options:
7328 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
7329 will pick one of them, and then handle the next batch of @var{n} frames until
7330 the end. Default is @code{100}.
7333 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
7334 value will result in a higher memory usage, so a high value is not recommended.
7336 @subsection Examples
7340 Extract one picture each 50 frames:
7346 Complete example of a thumbnail creation with @command{ffmpeg}:
7348 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
7354 Tile several successive frames together.
7356 The filter accepts the following options:
7361 Set the grid size (i.e. the number of lines and columns) in the form
7365 Set the maximum number of frames to render in the given area. It must be less
7366 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
7367 the area will be used.
7370 Set the outer border margin in pixels.
7373 Set the inner border thickness (i.e. the number of pixels between frames). For
7374 more advanced padding options (such as having different values for the edges),
7375 refer to the pad video filter.
7378 Specify the color of the unused area, it can be the name of a color
7379 (case insensitive match) or a 0xRRGGBB[AA] sequence.
7380 The default value of @var{color} is "black".
7383 @subsection Examples
7387 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
7389 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
7391 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
7392 duplicating each output frame to accomodate the originally detected frame
7396 Display @code{5} pictures in an area of @code{3x2} frames,
7397 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
7398 mixed flat and named options:
7400 tile=3x2:nb_frames=5:padding=7:margin=2
7406 Perform various types of temporal field interlacing.
7408 Frames are counted starting from 1, so the first input frame is
7411 The filter accepts the following options:
7416 Specify the mode of the interlacing. This option can also be specified
7417 as a value alone. See below for a list of values for this option.
7419 Available values are:
7423 Move odd frames into the upper field, even into the lower field,
7424 generating a double height frame at half frame rate.
7427 Only output even frames, odd frames are dropped, generating a frame with
7428 unchanged height at half frame rate.
7431 Only output odd frames, even frames are dropped, generating a frame with
7432 unchanged height at half frame rate.
7435 Expand each frame to full height, but pad alternate lines with black,
7436 generating a frame with double height at the same input frame rate.
7438 @item interleave_top, 4
7439 Interleave the upper field from odd frames with the lower field from
7440 even frames, generating a frame with unchanged height at half frame rate.
7442 @item interleave_bottom, 5
7443 Interleave the lower field from odd frames with the upper field from
7444 even frames, generating a frame with unchanged height at half frame rate.
7446 @item interlacex2, 6
7447 Double frame rate with unchanged height. Frames are inserted each
7448 containing the second temporal field from the previous input frame and
7449 the first temporal field from the next input frame. This mode relies on
7450 the top_field_first flag. Useful for interlaced video displays with no
7451 field synchronisation.
7454 Numeric values are deprecated but are accepted for backward
7455 compatibility reasons.
7457 Default mode is @code{merge}.
7460 Specify flags influencing the filter process.
7462 Available value for @var{flags} is:
7465 @item low_pass_filter, vlfp
7466 Enable vertical low-pass filtering in the filter.
7467 Vertical low-pass filtering is required when creating an interlaced
7468 destination from a progressive source which contains high-frequency
7469 vertical detail. Filtering will reduce interlace 'twitter' and Moire
7472 Vertical low-pass filtering can only be enabled for @option{mode}
7473 @var{interleave_top} and @var{interleave_bottom}.
7480 Transpose rows with columns in the input video and optionally flip it.
7482 This filter accepts the following options:
7487 Specify the transposition direction.
7489 Can assume the following values:
7491 @item 0, 4, cclock_flip
7492 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
7500 Rotate by 90 degrees clockwise, that is:
7508 Rotate by 90 degrees counterclockwise, that is:
7515 @item 3, 7, clock_flip
7516 Rotate by 90 degrees clockwise and vertically flip, that is:
7524 For values between 4-7, the transposition is only done if the input
7525 video geometry is portrait and not landscape. These values are
7526 deprecated, the @code{passthrough} option should be used instead.
7528 Numerical values are deprecated, and should be dropped in favor of
7532 Do not apply the transposition if the input geometry matches the one
7533 specified by the specified value. It accepts the following values:
7536 Always apply transposition.
7538 Preserve portrait geometry (when @var{height} >= @var{width}).
7540 Preserve landscape geometry (when @var{width} >= @var{height}).
7543 Default value is @code{none}.
7546 For example to rotate by 90 degrees clockwise and preserve portrait
7549 transpose=dir=1:passthrough=portrait
7552 The command above can also be specified as:
7554 transpose=1:portrait
7558 Trim the input so that the output contains one continuous subpart of the input.
7560 This filter accepts the following options:
7563 Specify time of the start of the kept section, i.e. the frame with the
7564 timestamp @var{start} will be the first frame in the output.
7567 Specify time of the first frame that will be dropped, i.e. the frame
7568 immediately preceding the one with the timestamp @var{end} will be the last
7569 frame in the output.
7572 Same as @var{start}, except this option sets the start timestamp in timebase
7573 units instead of seconds.
7576 Same as @var{end}, except this option sets the end timestamp in timebase units
7580 Specify maximum duration of the output.
7583 Number of the first frame that should be passed to output.
7586 Number of the first frame that should be dropped.
7589 @option{start}, @option{end}, @option{duration} are expressed as time
7590 duration specifications, check the "Time duration" section in the
7591 ffmpeg-utils manual.
7593 Note that the first two sets of the start/end options and the @option{duration}
7594 option look at the frame timestamp, while the _frame variants simply count the
7595 frames that pass through the filter. Also note that this filter does not modify
7596 the timestamps. If you wish that the output timestamps start at zero, insert a
7597 setpts filter after the trim filter.
7599 If multiple start or end options are set, this filter tries to be greedy and
7600 keep all the frames that match at least one of the specified constraints. To keep
7601 only the part that matches all the constraints at once, chain multiple trim
7604 The defaults are such that all the input is kept. So it is possible to set e.g.
7605 just the end values to keep everything before the specified time.
7610 drop everything except the second minute of input
7612 ffmpeg -i INPUT -vf trim=60:120
7616 keep only the first second
7618 ffmpeg -i INPUT -vf trim=duration=1
7626 Sharpen or blur the input video.
7628 It accepts the following parameters:
7631 @item luma_msize_x, lx
7632 Set the luma matrix horizontal size. It must be an odd integer between
7633 3 and 63, default value is 5.
7635 @item luma_msize_y, ly
7636 Set the luma matrix vertical size. It must be an odd integer between 3
7637 and 63, default value is 5.
7639 @item luma_amount, la
7640 Set the luma effect strength. It can be a float number, reasonable
7641 values lay between -1.5 and 1.5.
7643 Negative values will blur the input video, while positive values will
7644 sharpen it, a value of zero will disable the effect.
7646 Default value is 1.0.
7648 @item chroma_msize_x, cx
7649 Set the chroma matrix horizontal size. It must be an odd integer
7650 between 3 and 63, default value is 5.
7652 @item chroma_msize_y, cy
7653 Set the chroma matrix vertical size. It must be an odd integer
7654 between 3 and 63, default value is 5.
7656 @item chroma_amount, ca
7657 Set the chroma effect strength. It can be a float number, reasonable
7658 values lay between -1.5 and 1.5.
7660 Negative values will blur the input video, while positive values will
7661 sharpen it, a value of zero will disable the effect.
7663 Default value is 0.0.
7666 If set to 1, specify using OpenCL capabilities, only available if
7667 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
7671 All parameters are optional and default to the equivalent of the
7672 string '5:5:1.0:5:5:0.0'.
7674 @subsection Examples
7678 Apply strong luma sharpen effect:
7680 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
7684 Apply strong blur of both luma and chroma parameters:
7686 unsharp=7:7:-2:7:7:-2
7690 @anchor{vidstabdetect}
7691 @section vidstabdetect
7693 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
7694 @ref{vidstabtransform} for pass 2.
7696 This filter generates a file with relative translation and rotation
7697 transform information about subsequent frames, which is then used by
7698 the @ref{vidstabtransform} filter.
7700 To enable compilation of this filter you need to configure FFmpeg with
7701 @code{--enable-libvidstab}.
7703 This filter accepts the following options:
7707 Set the path to the file used to write the transforms information.
7708 Default value is @file{transforms.trf}.
7711 Set how shaky the video is and how quick the camera is. It accepts an
7712 integer in the range 1-10, a value of 1 means little shakiness, a
7713 value of 10 means strong shakiness. Default value is 5.
7716 Set the accuracy of the detection process. It must be a value in the
7717 range 1-15. A value of 1 means low accuracy, a value of 15 means high
7718 accuracy. Default value is 9.
7721 Set stepsize of the search process. The region around minimum is
7722 scanned with 1 pixel resolution. Default value is 6.
7725 Set minimum contrast. Below this value a local measurement field is
7726 discarded. Must be a floating point value in the range 0-1. Default
7730 Set reference frame number for tripod mode.
7732 If enabled, the motion of the frames is compared to a reference frame
7733 in the filtered stream, identified by the specified number. The idea
7734 is to compensate all movements in a more-or-less static scene and keep
7735 the camera view absolutely still.
7737 If set to 0, it is disabled. The frames are counted starting from 1.
7740 Show fields and transforms in the resulting frames. It accepts an
7741 integer in the range 0-2. Default value is 0, which disables any
7745 @subsection Examples
7755 Analyze strongly shaky movie and put the results in file
7756 @file{mytransforms.trf}:
7758 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
7762 Visualize the result of internal transformations in the resulting
7765 vidstabdetect=show=1
7769 Analyze a video with medium shakiness using @command{ffmpeg}:
7771 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
7775 @anchor{vidstabtransform}
7776 @section vidstabtransform
7778 Video stabilization/deshaking: pass 2 of 2,
7779 see @ref{vidstabdetect} for pass 1.
7781 Read a file with transform information for each frame and
7782 apply/compensate them. Together with the @ref{vidstabdetect}
7783 filter this can be used to deshake videos. See also
7784 @url{http://public.hronopik.de/vid.stab}. It is important to also use
7785 the unsharp filter, see below.
7787 To enable compilation of this filter you need to configure FFmpeg with
7788 @code{--enable-libvidstab}.
7790 This filter accepts the following options:
7795 path to the file used to read the transforms (default: @file{transforms.trf})
7798 number of frames (value*2 + 1) used for lowpass filtering the camera movements
7799 (default: 10). For example a number of 10 means that 21 frames are used
7800 (10 in the past and 10 in the future) to smoothen the motion in the
7801 video. A larger values leads to a smoother video, but limits the
7802 acceleration of the camera (pan/tilt movements).
7805 maximal number of pixels to translate frames (default: -1 no limit)
7808 maximal angle in radians (degree*PI/180) to rotate frames (default: -1
7812 How to deal with borders that may be visible due to movement
7813 compensation. Available values are:
7817 keep image information from previous frame (default)
7819 fill the border black
7825 keep transforms normal (default)
7832 consider transforms as
7837 relative to previous frame (default)
7842 percentage to zoom (default: 0)
7851 set optimal zooming to avoid borders
7856 optimal static zoom value is determined (only very strong movements will lead to visible borders) (default)
7858 optimal adaptive zoom value is determined (no borders will be visible)
7860 Note that the value given at zoom is added to the one calculated
7864 type of interpolation
7866 Available values are:
7871 linear only horizontal
7873 linear in both directions (default)
7875 cubic in both directions (slow)
7879 virtual tripod mode means that the video is stabilized such that the
7880 camera stays stationary. Use also @code{tripod} option of
7881 @ref{vidstabdetect}.
7886 virtual tripod mode: equivalent to @code{relative=0:smoothing=0}
7891 @subsection Examples
7895 typical call with default default values:
7896 (note the unsharp filter which is always recommended)
7898 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
7902 zoom in a bit more and load transform data from a given file
7904 vidstabtransform=zoom=5:input="mytransforms.trf"
7908 smoothen the video even more
7910 vidstabtransform=smoothing=30
7917 Flip the input video vertically.
7919 For example, to vertically flip a video with @command{ffmpeg}:
7921 ffmpeg -i in.avi -vf "vflip" out.avi
7926 Make or reverse a natural vignetting effect.
7928 The filter accepts the following options:
7932 Set lens angle expression as a number of radians.
7934 The value is clipped in the @code{[0,PI/2]} range.
7936 Default value: @code{"PI/5"}
7940 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
7944 Set forward/backward mode.
7946 Available modes are:
7949 The larger the distance from the central point, the darker the image becomes.
7952 The larger the distance from the central point, the brighter the image becomes.
7953 This can be used to reverse a vignette effect, though there is no automatic
7954 detection to extract the lens @option{angle} and other settings (yet). It can
7955 also be used to create a burning effect.
7958 Default value is @samp{forward}.
7961 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
7963 It accepts the following values:
7966 Evaluate expressions only once during the filter initialization.
7969 Evaluate expressions for each incoming frame. This is way slower than the
7970 @samp{init} mode since it requires all the scalers to be re-computed, but it
7971 allows advanced dynamic expressions.
7974 Default value is @samp{init}.
7977 Set dithering to reduce the circular banding effects. Default is @code{1}
7981 Set vignette aspect. This setting allows to adjust the shape of the vignette.
7982 Setting this value to the SAR of the input will make a rectangular vignetting
7983 following the dimensions of the video.
7985 Default is @code{1/1}.
7988 @subsection Expressions
7990 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
7991 following parameters.
7996 input width and height
7999 the number of input frame, starting from 0
8002 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
8003 @var{TB} units, NAN if undefined
8006 frame rate of the input video, NAN if the input frame rate is unknown
8009 the PTS (Presentation TimeStamp) of the filtered video frame,
8010 expressed in seconds, NAN if undefined
8013 time base of the input video
8017 @subsection Examples
8021 Apply simple strong vignetting effect:
8027 Make a flickering vignetting:
8029 vignette='PI/4+random(1)*PI/50':eval=frame
8036 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
8037 Deinterlacing Filter").
8039 Based on the process described by Martin Weston for BBC R&D, and
8040 implemented based on the de-interlace algorithm written by Jim
8041 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
8042 uses filter coefficients calculated by BBC R&D.
8044 There are two sets of filter coefficients, so called "simple":
8045 and "complex". Which set of filter coefficients is used can
8046 be set by passing an optional parameter:
8050 Set the interlacing filter coefficients. Accepts one of the following values:
8054 Simple filter coefficient set.
8056 More-complex filter coefficient set.
8058 Default value is @samp{complex}.
8061 Specify which frames to deinterlace. Accept one of the following values:
8065 Deinterlace all frames,
8067 Only deinterlace frames marked as interlaced.
8070 Default value is @samp{all}.
8076 Deinterlace the input video ("yadif" means "yet another deinterlacing
8079 This filter accepts the following options:
8085 The interlacing mode to adopt, accepts one of the following values:
8089 output 1 frame for each frame
8091 output 1 frame for each field
8092 @item 2, send_frame_nospatial
8093 like @code{send_frame} but skip spatial interlacing check
8094 @item 3, send_field_nospatial
8095 like @code{send_field} but skip spatial interlacing check
8098 Default value is @code{send_frame}.
8101 The picture field parity assumed for the input interlaced video, accepts one of
8102 the following values:
8106 assume top field first
8108 assume bottom field first
8110 enable automatic detection
8113 Default value is @code{auto}.
8114 If interlacing is unknown or decoder does not export this information,
8115 top field first will be assumed.
8118 Specify which frames to deinterlace. Accept one of the following
8123 deinterlace all frames
8125 only deinterlace frames marked as interlaced
8128 Default value is @code{all}.
8131 @c man end VIDEO FILTERS
8133 @chapter Video Sources
8134 @c man begin VIDEO SOURCES
8136 Below is a description of the currently available video sources.
8140 Buffer video frames, and make them available to the filter chain.
8142 This source is mainly intended for a programmatic use, in particular
8143 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
8145 This source accepts the following options:
8150 Specify the size (width and height) of the buffered video frames.
8159 A string representing the pixel format of the buffered video frames.
8160 It may be a number corresponding to a pixel format, or a pixel format
8164 Specify the timebase assumed by the timestamps of the buffered frames.
8167 Specify the frame rate expected for the video stream.
8169 @item pixel_aspect, sar
8170 Specify the sample aspect ratio assumed by the video frames.
8173 Specify the optional parameters to be used for the scale filter which
8174 is automatically inserted when an input change is detected in the
8175 input size or format.
8180 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
8183 will instruct the source to accept video frames with size 320x240 and
8184 with format "yuv410p", assuming 1/24 as the timestamps timebase and
8185 square pixels (1:1 sample aspect ratio).
8186 Since the pixel format with name "yuv410p" corresponds to the number 6
8187 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
8188 this example corresponds to:
8190 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
8193 Alternatively, the options can be specified as a flat string, but this
8194 syntax is deprecated:
8196 @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}]
8200 Create a pattern generated by an elementary cellular automaton.
8202 The initial state of the cellular automaton can be defined through the
8203 @option{filename}, and @option{pattern} options. If such options are
8204 not specified an initial state is created randomly.
8206 At each new frame a new row in the video is filled with the result of
8207 the cellular automaton next generation. The behavior when the whole
8208 frame is filled is defined by the @option{scroll} option.
8210 This source accepts the following options:
8214 Read the initial cellular automaton state, i.e. the starting row, from
8216 In the file, each non-whitespace character is considered an alive
8217 cell, a newline will terminate the row, and further characters in the
8218 file will be ignored.
8221 Read the initial cellular automaton state, i.e. the starting row, from
8222 the specified string.
8224 Each non-whitespace character in the string is considered an alive
8225 cell, a newline will terminate the row, and further characters in the
8226 string will be ignored.
8229 Set the video rate, that is the number of frames generated per second.
8232 @item random_fill_ratio, ratio
8233 Set the random fill ratio for the initial cellular automaton row. It
8234 is a floating point number value ranging from 0 to 1, defaults to
8237 This option is ignored when a file or a pattern is specified.
8239 @item random_seed, seed
8240 Set the seed for filling randomly the initial row, must be an integer
8241 included between 0 and UINT32_MAX. If not specified, or if explicitly
8242 set to -1, the filter will try to use a good random seed on a best
8246 Set the cellular automaton rule, it is a number ranging from 0 to 255.
8247 Default value is 110.
8250 Set the size of the output video.
8252 If @option{filename} or @option{pattern} is specified, the size is set
8253 by default to the width of the specified initial state row, and the
8254 height is set to @var{width} * PHI.
8256 If @option{size} is set, it must contain the width of the specified
8257 pattern string, and the specified pattern will be centered in the
8260 If a filename or a pattern string is not specified, the size value
8261 defaults to "320x518" (used for a randomly generated initial state).
8264 If set to 1, scroll the output upward when all the rows in the output
8265 have been already filled. If set to 0, the new generated row will be
8266 written over the top row just after the bottom row is filled.
8269 @item start_full, full
8270 If set to 1, completely fill the output with generated rows before
8271 outputting the first frame.
8272 This is the default behavior, for disabling set the value to 0.
8275 If set to 1, stitch the left and right row edges together.
8276 This is the default behavior, for disabling set the value to 0.
8279 @subsection Examples
8283 Read the initial state from @file{pattern}, and specify an output of
8286 cellauto=f=pattern:s=200x400
8290 Generate a random initial row with a width of 200 cells, with a fill
8293 cellauto=ratio=2/3:s=200x200
8297 Create a pattern generated by rule 18 starting by a single alive cell
8298 centered on an initial row with width 100:
8300 cellauto=p=@@:s=100x400:full=0:rule=18
8304 Specify a more elaborated initial pattern:
8306 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
8313 Generate a Mandelbrot set fractal, and progressively zoom towards the
8314 point specified with @var{start_x} and @var{start_y}.
8316 This source accepts the following options:
8321 Set the terminal pts value. Default value is 400.
8324 Set the terminal scale value.
8325 Must be a floating point value. Default value is 0.3.
8328 Set the inner coloring mode, that is the algorithm used to draw the
8329 Mandelbrot fractal internal region.
8331 It shall assume one of the following values:
8336 Show time until convergence.
8338 Set color based on point closest to the origin of the iterations.
8343 Default value is @var{mincol}.
8346 Set the bailout value. Default value is 10.0.
8349 Set the maximum of iterations performed by the rendering
8350 algorithm. Default value is 7189.
8353 Set outer coloring mode.
8354 It shall assume one of following values:
8356 @item iteration_count
8357 Set iteration cound mode.
8358 @item normalized_iteration_count
8359 set normalized iteration count mode.
8361 Default value is @var{normalized_iteration_count}.
8364 Set frame rate, expressed as number of frames per second. Default
8368 Set frame size. Default value is "640x480".
8371 Set the initial scale value. Default value is 3.0.
8374 Set the initial x position. Must be a floating point value between
8375 -100 and 100. Default value is -0.743643887037158704752191506114774.
8378 Set the initial y position. Must be a floating point value between
8379 -100 and 100. Default value is -0.131825904205311970493132056385139.
8384 Generate various test patterns, as generated by the MPlayer test filter.
8386 The size of the generated video is fixed, and is 256x256.
8387 This source is useful in particular for testing encoding features.
8389 This source accepts the following options:
8394 Specify the frame rate of the sourced video, as the number of frames
8395 generated per second. It has to be a string in the format
8396 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
8397 number or a valid video frame rate abbreviation. The default value is
8401 Set the video duration of the sourced video. The accepted syntax is:
8406 See also the function @code{av_parse_time()}.
8408 If not specified, or the expressed duration is negative, the video is
8409 supposed to be generated forever.
8413 Set the number or the name of the test to perform. Supported tests are:
8428 Default value is "all", which will cycle through the list of all tests.
8431 For example the following:
8436 will generate a "dc_luma" test pattern.
8440 Provide a frei0r source.
8442 To enable compilation of this filter you need to install the frei0r
8443 header and configure FFmpeg with @code{--enable-frei0r}.
8445 This source accepts the following options:
8450 The size of the video to generate, may be a string of the form
8451 @var{width}x@var{height} or a frame size abbreviation.
8454 Framerate of the generated video, may be a string of the form
8455 @var{num}/@var{den} or a frame rate abbreviation.
8458 The name to the frei0r source to load. For more information regarding frei0r and
8459 how to set the parameters read the section @ref{frei0r} in the description of
8463 A '|'-separated list of parameters to pass to the frei0r source.
8467 For example, to generate a frei0r partik0l source with size 200x200
8468 and frame rate 10 which is overlayed on the overlay filter main input:
8470 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
8475 Generate a life pattern.
8477 This source is based on a generalization of John Conway's life game.
8479 The sourced input represents a life grid, each pixel represents a cell
8480 which can be in one of two possible states, alive or dead. Every cell
8481 interacts with its eight neighbours, which are the cells that are
8482 horizontally, vertically, or diagonally adjacent.
8484 At each interaction the grid evolves according to the adopted rule,
8485 which specifies the number of neighbor alive cells which will make a
8486 cell stay alive or born. The @option{rule} option allows to specify
8489 This source accepts the following options:
8493 Set the file from which to read the initial grid state. In the file,
8494 each non-whitespace character is considered an alive cell, and newline
8495 is used to delimit the end of each row.
8497 If this option is not specified, the initial grid is generated
8501 Set the video rate, that is the number of frames generated per second.
8504 @item random_fill_ratio, ratio
8505 Set the random fill ratio for the initial random grid. It is a
8506 floating point number value ranging from 0 to 1, defaults to 1/PHI.
8507 It is ignored when a file is specified.
8509 @item random_seed, seed
8510 Set the seed for filling the initial random grid, must be an integer
8511 included between 0 and UINT32_MAX. If not specified, or if explicitly
8512 set to -1, the filter will try to use a good random seed on a best
8518 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
8519 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
8520 @var{NS} specifies the number of alive neighbor cells which make a
8521 live cell stay alive, and @var{NB} the number of alive neighbor cells
8522 which make a dead cell to become alive (i.e. to "born").
8523 "s" and "b" can be used in place of "S" and "B", respectively.
8525 Alternatively a rule can be specified by an 18-bits integer. The 9
8526 high order bits are used to encode the next cell state if it is alive
8527 for each number of neighbor alive cells, the low order bits specify
8528 the rule for "borning" new cells. Higher order bits encode for an
8529 higher number of neighbor cells.
8530 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
8531 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
8533 Default value is "S23/B3", which is the original Conway's game of life
8534 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
8535 cells, and will born a new cell if there are three alive cells around
8539 Set the size of the output video.
8541 If @option{filename} is specified, the size is set by default to the
8542 same size of the input file. If @option{size} is set, it must contain
8543 the size specified in the input file, and the initial grid defined in
8544 that file is centered in the larger resulting area.
8546 If a filename is not specified, the size value defaults to "320x240"
8547 (used for a randomly generated initial grid).
8550 If set to 1, stitch the left and right grid edges together, and the
8551 top and bottom edges also. Defaults to 1.
8554 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
8555 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
8556 value from 0 to 255.
8559 Set the color of living (or new born) cells.
8562 Set the color of dead cells. If @option{mold} is set, this is the first color
8563 used to represent a dead cell.
8566 Set mold color, for definitely dead and moldy cells.
8569 @subsection Examples
8573 Read a grid from @file{pattern}, and center it on a grid of size
8576 life=f=pattern:s=300x300
8580 Generate a random grid of size 200x200, with a fill ratio of 2/3:
8582 life=ratio=2/3:s=200x200
8586 Specify a custom rule for evolving a randomly generated grid:
8592 Full example with slow death effect (mold) using @command{ffplay}:
8594 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
8599 @anchor{haldclutsrc}
8603 @anchor{smptehdbars}
8605 @section color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc
8607 The @code{color} source provides an uniformly colored input.
8609 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
8610 @ref{haldclut} filter.
8612 The @code{nullsrc} source returns unprocessed video frames. It is
8613 mainly useful to be employed in analysis / debugging tools, or as the
8614 source for filters which ignore the input data.
8616 The @code{rgbtestsrc} source generates an RGB test pattern useful for
8617 detecting RGB vs BGR issues. You should see a red, green and blue
8618 stripe from top to bottom.
8620 The @code{smptebars} source generates a color bars pattern, based on
8621 the SMPTE Engineering Guideline EG 1-1990.
8623 The @code{smptehdbars} source generates a color bars pattern, based on
8624 the SMPTE RP 219-2002.
8626 The @code{testsrc} source generates a test video pattern, showing a
8627 color pattern, a scrolling gradient and a timestamp. This is mainly
8628 intended for testing purposes.
8630 The sources accept the following options:
8635 Specify the color of the source, only available in the @code{color}
8636 source. It can be the name of a color (case insensitive match) or a
8637 0xRRGGBB[AA] sequence, possibly followed by an alpha specifier. The
8638 default value is "black".
8641 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
8642 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
8643 pixels to be used as identity matrix for 3D lookup tables. Each component is
8644 coded on a @code{1/(N*N)} scale.
8647 Specify the size of the sourced video, it may be a string of the form
8648 @var{width}x@var{height}, or the name of a size abbreviation. The
8649 default value is "320x240".
8651 This option is not available with the @code{haldclutsrc} filter.
8654 Specify the frame rate of the sourced video, as the number of frames
8655 generated per second. It has to be a string in the format
8656 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
8657 number or a valid video frame rate abbreviation. The default value is
8661 Set the sample aspect ratio of the sourced video.
8664 Set the video duration of the sourced video. The accepted syntax is:
8666 [-]HH[:MM[:SS[.m...]]]
8669 See also the function @code{av_parse_time()}.
8671 If not specified, or the expressed duration is negative, the video is
8672 supposed to be generated forever.
8675 Set the number of decimals to show in the timestamp, only available in the
8676 @code{testsrc} source.
8678 The displayed timestamp value will correspond to the original
8679 timestamp value multiplied by the power of 10 of the specified
8680 value. Default value is 0.
8683 For example the following:
8685 testsrc=duration=5.3:size=qcif:rate=10
8688 will generate a video with a duration of 5.3 seconds, with size
8689 176x144 and a frame rate of 10 frames per second.
8691 The following graph description will generate a red source
8692 with an opacity of 0.2, with size "qcif" and a frame rate of 10
8695 color=c=red@@0.2:s=qcif:r=10
8698 If the input content is to be ignored, @code{nullsrc} can be used. The
8699 following command generates noise in the luminance plane by employing
8700 the @code{geq} filter:
8702 nullsrc=s=256x256, geq=random(1)*255:128:128
8705 @subsection Commands
8707 The @code{color} source supports the following commands:
8711 Set the color of the created image. Accepts the same syntax of the
8712 corresponding @option{color} option.
8715 @c man end VIDEO SOURCES
8717 @chapter Video Sinks
8718 @c man begin VIDEO SINKS
8720 Below is a description of the currently available video sinks.
8724 Buffer video frames, and make them available to the end of the filter
8727 This sink is mainly intended for a programmatic use, in particular
8728 through the interface defined in @file{libavfilter/buffersink.h}
8729 or the options system.
8731 It accepts a pointer to an AVBufferSinkContext structure, which
8732 defines the incoming buffers' formats, to be passed as the opaque
8733 parameter to @code{avfilter_init_filter} for initialization.
8737 Null video sink, do absolutely nothing with the input video. It is
8738 mainly useful as a template and to be employed in analysis / debugging
8741 @c man end VIDEO SINKS
8743 @chapter Multimedia Filters
8744 @c man begin MULTIMEDIA FILTERS
8746 Below is a description of the currently available multimedia filters.
8748 @section avectorscope
8750 Convert input audio to a video output, representing the audio vector
8753 The filter is used to measure the difference between channels of stereo
8754 audio stream. A monoaural signal, consisting of identical left and right
8755 signal, results in straight vertical line. Any stereo separation is visible
8756 as a deviation from this line, creating a Lissajous figure.
8757 If the straight (or deviation from it) but horizontal line appears this
8758 indicates that the left and right channels are out of phase.
8760 The filter accepts the following options:
8764 Set the vectorscope mode.
8766 Available values are:
8769 Lissajous rotated by 45 degrees.
8772 Same as above but not rotated.
8775 Default value is @samp{lissajous}.
8778 Set the video size for the output. Default value is @code{400x400}.
8781 Set the output frame rate. Default value is @code{25}.
8786 Specify the red, green and blue contrast. Default values are @code{40}, @code{160} and @code{80}.
8787 Allowed range is @code{[0, 255]}.
8792 Specify the red, green and blue fade. Default values are @code{15}, @code{10} and @code{5}.
8793 Allowed range is @code{[0, 255]}.
8796 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
8799 @subsection Examples
8803 Complete example using @command{ffplay}:
8805 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
8806 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
8812 Concatenate audio and video streams, joining them together one after the
8815 The filter works on segments of synchronized video and audio streams. All
8816 segments must have the same number of streams of each type, and that will
8817 also be the number of streams at output.
8819 The filter accepts the following options:
8824 Set the number of segments. Default is 2.
8827 Set the number of output video streams, that is also the number of video
8828 streams in each segment. Default is 1.
8831 Set the number of output audio streams, that is also the number of video
8832 streams in each segment. Default is 0.
8835 Activate unsafe mode: do not fail if segments have a different format.
8839 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
8840 @var{a} audio outputs.
8842 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
8843 segment, in the same order as the outputs, then the inputs for the second
8846 Related streams do not always have exactly the same duration, for various
8847 reasons including codec frame size or sloppy authoring. For that reason,
8848 related synchronized streams (e.g. a video and its audio track) should be
8849 concatenated at once. The concat filter will use the duration of the longest
8850 stream in each segment (except the last one), and if necessary pad shorter
8851 audio streams with silence.
8853 For this filter to work correctly, all segments must start at timestamp 0.
8855 All corresponding streams must have the same parameters in all segments; the
8856 filtering system will automatically select a common pixel format for video
8857 streams, and a common sample format, sample rate and channel layout for
8858 audio streams, but other settings, such as resolution, must be converted
8859 explicitly by the user.
8861 Different frame rates are acceptable but will result in variable frame rate
8862 at output; be sure to configure the output file to handle it.
8864 @subsection Examples
8868 Concatenate an opening, an episode and an ending, all in bilingual version
8869 (video in stream 0, audio in streams 1 and 2):
8871 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
8872 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
8873 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
8874 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
8878 Concatenate two parts, handling audio and video separately, using the
8879 (a)movie sources, and adjusting the resolution:
8881 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
8882 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
8883 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
8885 Note that a desync will happen at the stitch if the audio and video streams
8886 do not have exactly the same duration in the first file.
8892 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
8893 it unchanged. By default, it logs a message at a frequency of 10Hz with the
8894 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
8895 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
8897 The filter also has a video output (see the @var{video} option) with a real
8898 time graph to observe the loudness evolution. The graphic contains the logged
8899 message mentioned above, so it is not printed anymore when this option is set,
8900 unless the verbose logging is set. The main graphing area contains the
8901 short-term loudness (3 seconds of analysis), and the gauge on the right is for
8902 the momentary loudness (400 milliseconds).
8904 More information about the Loudness Recommendation EBU R128 on
8905 @url{http://tech.ebu.ch/loudness}.
8907 The filter accepts the following options:
8912 Activate the video output. The audio stream is passed unchanged whether this
8913 option is set or no. The video stream will be the first output stream if
8914 activated. Default is @code{0}.
8917 Set the video size. This option is for video only. Default and minimum
8918 resolution is @code{640x480}.
8921 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
8922 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
8923 other integer value between this range is allowed.
8926 Set metadata injection. If set to @code{1}, the audio input will be segmented
8927 into 100ms output frames, each of them containing various loudness information
8928 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
8930 Default is @code{0}.
8933 Force the frame logging level.
8935 Available values are:
8938 information logging level
8940 verbose logging level
8943 By default, the logging level is set to @var{info}. If the @option{video} or
8944 the @option{metadata} options are set, it switches to @var{verbose}.
8947 @subsection Examples
8951 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
8953 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
8957 Run an analysis with @command{ffmpeg}:
8959 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
8963 @section interleave, ainterleave
8965 Temporally interleave frames from several inputs.
8967 @code{interleave} works with video inputs, @code{ainterleave} with audio.
8969 These filters read frames from several inputs and send the oldest
8970 queued frame to the output.
8972 Input streams must have a well defined, monotonically increasing frame
8975 In order to submit one frame to output, these filters need to enqueue
8976 at least one frame for each input, so they cannot work in case one
8977 input is not yet terminated and will not receive incoming frames.
8979 For example consider the case when one input is a @code{select} filter
8980 which always drop input frames. The @code{interleave} filter will keep
8981 reading from that input, but it will never be able to send new frames
8982 to output until the input will send an end-of-stream signal.
8984 Also, depending on inputs synchronization, the filters will drop
8985 frames in case one input receives more frames than the other ones, and
8986 the queue is already filled.
8988 These filters accept the following options:
8992 Set the number of different inputs, it is 2 by default.
8995 @subsection Examples
8999 Interleave frames belonging to different streams using @command{ffmpeg}:
9001 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
9005 Add flickering blur effect:
9007 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
9011 @section perms, aperms
9013 Set read/write permissions for the output frames.
9015 These filters are mainly aimed at developers to test direct path in the
9016 following filter in the filtergraph.
9018 The filters accept the following options:
9022 Select the permissions mode.
9024 It accepts the following values:
9027 Do nothing. This is the default.
9029 Set all the output frames read-only.
9031 Set all the output frames directly writable.
9033 Make the frame read-only if writable, and writable if read-only.
9035 Set each output frame read-only or writable randomly.
9039 Set the seed for the @var{random} mode, must be an integer included between
9040 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
9041 @code{-1}, the filter will try to use a good random seed on a best effort
9045 Note: in case of auto-inserted filter between the permission filter and the
9046 following one, the permission might not be received as expected in that
9047 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
9048 perms/aperms filter can avoid this problem.
9050 @section select, aselect
9052 Select frames to pass in output.
9054 This filter accepts the following options:
9059 Set expression, which is evaluated for each input frame.
9061 If the expression is evaluated to zero, the frame is discarded.
9063 If the evaluation result is negative or NaN, the frame is sent to the
9064 first output; otherwise it is sent to the output with index
9065 @code{ceil(val)-1}, assuming that the input index starts from 0.
9067 For example a value of @code{1.2} corresponds to the output with index
9068 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
9071 Set the number of outputs. The output to which to send the selected
9072 frame is based on the result of the evaluation. Default value is 1.
9075 The expression can contain the following constants:
9079 the sequential number of the filtered frame, starting from 0
9082 the sequential number of the selected frame, starting from 0
9084 @item prev_selected_n
9085 the sequential number of the last selected frame, NAN if undefined
9088 timebase of the input timestamps
9091 the PTS (Presentation TimeStamp) of the filtered video frame,
9092 expressed in @var{TB} units, NAN if undefined
9095 the PTS (Presentation TimeStamp) of the filtered video frame,
9096 expressed in seconds, NAN if undefined
9099 the PTS of the previously filtered video frame, NAN if undefined
9101 @item prev_selected_pts
9102 the PTS of the last previously filtered video frame, NAN if undefined
9104 @item prev_selected_t
9105 the PTS of the last previously selected video frame, NAN if undefined
9108 the PTS of the first video frame in the video, NAN if undefined
9111 the time of the first video frame in the video, NAN if undefined
9113 @item pict_type @emph{(video only)}
9114 the type of the filtered frame, can assume one of the following
9126 @item interlace_type @emph{(video only)}
9127 the frame interlace type, can assume one of the following values:
9130 the frame is progressive (not interlaced)
9132 the frame is top-field-first
9134 the frame is bottom-field-first
9137 @item consumed_sample_n @emph{(audio only)}
9138 the number of selected samples before the current frame
9140 @item samples_n @emph{(audio only)}
9141 the number of samples in the current frame
9143 @item sample_rate @emph{(audio only)}
9144 the input sample rate
9147 1 if the filtered frame is a key-frame, 0 otherwise
9150 the position in the file of the filtered frame, -1 if the information
9151 is not available (e.g. for synthetic video)
9153 @item scene @emph{(video only)}
9154 value between 0 and 1 to indicate a new scene; a low value reflects a low
9155 probability for the current frame to introduce a new scene, while a higher
9156 value means the current frame is more likely to be one (see the example below)
9160 The default value of the select expression is "1".
9162 @subsection Examples
9166 Select all frames in input:
9171 The example above is the same as:
9183 Select only I-frames:
9185 select='eq(pict_type\,I)'
9189 Select one frame every 100:
9191 select='not(mod(n\,100))'
9195 Select only frames contained in the 10-20 time interval:
9197 select=between(t\,10\,20)
9201 Select only I frames contained in the 10-20 time interval:
9203 select=between(t\,10\,20)*eq(pict_type\,I)
9207 Select frames with a minimum distance of 10 seconds:
9209 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
9213 Use aselect to select only audio frames with samples number > 100:
9215 aselect='gt(samples_n\,100)'
9219 Create a mosaic of the first scenes:
9221 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
9224 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
9228 Send even and odd frames to separate outputs, and compose them:
9230 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
9234 @section sendcmd, asendcmd
9236 Send commands to filters in the filtergraph.
9238 These filters read commands to be sent to other filters in the
9241 @code{sendcmd} must be inserted between two video filters,
9242 @code{asendcmd} must be inserted between two audio filters, but apart
9243 from that they act the same way.
9245 The specification of commands can be provided in the filter arguments
9246 with the @var{commands} option, or in a file specified by the
9247 @var{filename} option.
9249 These filters accept the following options:
9252 Set the commands to be read and sent to the other filters.
9254 Set the filename of the commands to be read and sent to the other
9258 @subsection Commands syntax
9260 A commands description consists of a sequence of interval
9261 specifications, comprising a list of commands to be executed when a
9262 particular event related to that interval occurs. The occurring event
9263 is typically the current frame time entering or leaving a given time
9266 An interval is specified by the following syntax:
9268 @var{START}[-@var{END}] @var{COMMANDS};
9271 The time interval is specified by the @var{START} and @var{END} times.
9272 @var{END} is optional and defaults to the maximum time.
9274 The current frame time is considered within the specified interval if
9275 it is included in the interval [@var{START}, @var{END}), that is when
9276 the time is greater or equal to @var{START} and is lesser than
9279 @var{COMMANDS} consists of a sequence of one or more command
9280 specifications, separated by ",", relating to that interval. The
9281 syntax of a command specification is given by:
9283 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
9286 @var{FLAGS} is optional and specifies the type of events relating to
9287 the time interval which enable sending the specified command, and must
9288 be a non-null sequence of identifier flags separated by "+" or "|" and
9289 enclosed between "[" and "]".
9291 The following flags are recognized:
9294 The command is sent when the current frame timestamp enters the
9295 specified interval. In other words, the command is sent when the
9296 previous frame timestamp was not in the given interval, and the
9300 The command is sent when the current frame timestamp leaves the
9301 specified interval. In other words, the command is sent when the
9302 previous frame timestamp was in the given interval, and the
9306 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
9309 @var{TARGET} specifies the target of the command, usually the name of
9310 the filter class or a specific filter instance name.
9312 @var{COMMAND} specifies the name of the command for the target filter.
9314 @var{ARG} is optional and specifies the optional list of argument for
9315 the given @var{COMMAND}.
9317 Between one interval specification and another, whitespaces, or
9318 sequences of characters starting with @code{#} until the end of line,
9319 are ignored and can be used to annotate comments.
9321 A simplified BNF description of the commands specification syntax
9324 @var{COMMAND_FLAG} ::= "enter" | "leave"
9325 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
9326 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
9327 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
9328 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
9329 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
9332 @subsection Examples
9336 Specify audio tempo change at second 4:
9338 asendcmd=c='4.0 atempo tempo 1.5',atempo
9342 Specify a list of drawtext and hue commands in a file.
9344 # show text in the interval 5-10
9345 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
9346 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
9348 # desaturate the image in the interval 15-20
9349 15.0-20.0 [enter] hue s 0,
9350 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
9352 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
9354 # apply an exponential saturation fade-out effect, starting from time 25
9355 25 [enter] hue s exp(25-t)
9358 A filtergraph allowing to read and process the above command list
9359 stored in a file @file{test.cmd}, can be specified with:
9361 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
9366 @section setpts, asetpts
9368 Change the PTS (presentation timestamp) of the input frames.
9370 @code{setpts} works on video frames, @code{asetpts} on audio frames.
9372 This filter accepts the following options:
9377 The expression which is evaluated for each frame to construct its timestamp.
9381 The expression is evaluated through the eval API and can contain the following
9386 frame rate, only defined for constant frame-rate video
9389 the presentation timestamp in input
9392 the count of the input frame for video or the number of consumed samples,
9393 not including the current frame for audio, starting from 0.
9395 @item NB_CONSUMED_SAMPLES
9396 the number of consumed samples, not including the current frame (only
9400 the number of samples in the current frame (only audio)
9402 @item SAMPLE_RATE, SR
9406 the PTS of the first frame
9409 the time in seconds of the first frame
9412 tell if the current frame is interlaced
9415 the time in seconds of the current frame
9418 original position in the file of the frame, or undefined if undefined
9419 for the current frame
9425 previous input time in seconds
9431 previous output time in seconds
9434 wallclock (RTC) time in microseconds. This is deprecated, use time(0)
9438 wallclock (RTC) time at the start of the movie in microseconds
9441 timebase of the input timestamps
9445 @subsection Examples
9449 Start counting PTS from zero
9455 Apply fast motion effect:
9461 Apply slow motion effect:
9467 Set fixed rate of 25 frames per second:
9473 Set fixed rate 25 fps with some jitter:
9475 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
9479 Apply an offset of 10 seconds to the input PTS:
9485 Generate timestamps from a "live source" and rebase onto the current timebase:
9487 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
9491 Generate timestamps by counting samples:
9498 @section settb, asettb
9500 Set the timebase to use for the output frames timestamps.
9501 It is mainly useful for testing timebase configuration.
9503 This filter accepts the following options:
9508 The expression which is evaluated into the output timebase.
9512 The value for @option{tb} is an arithmetic expression representing a
9513 rational. The expression can contain the constants "AVTB" (the default
9514 timebase), "intb" (the input timebase) and "sr" (the sample rate,
9515 audio only). Default value is "intb".
9517 @subsection Examples
9521 Set the timebase to 1/25:
9527 Set the timebase to 1/10:
9533 Set the timebase to 1001/1000:
9539 Set the timebase to 2*intb:
9545 Set the default timebase value:
9551 @section showspectrum
9553 Convert input audio to a video output, representing the audio frequency
9556 The filter accepts the following options:
9560 Specify the video size for the output. Default value is @code{640x512}.
9563 Specify if the spectrum should slide along the window. Default value is
9567 Specify display mode.
9569 It accepts the following values:
9572 all channels are displayed in the same row
9574 all channels are displayed in separate rows
9577 Default value is @samp{combined}.
9580 Specify display color mode.
9582 It accepts the following values:
9585 each channel is displayed in a separate color
9587 each channel is is displayed using the same color scheme
9590 Default value is @samp{channel}.
9593 Specify scale used for calculating intensity color values.
9595 It accepts the following values:
9600 square root, default
9607 Default value is @samp{sqrt}.
9610 Set saturation modifier for displayed colors. Negative values provide
9611 alternative color scheme. @code{0} is no saturation at all.
9612 Saturation must be in [-10.0, 10.0] range.
9613 Default value is @code{1}.
9616 The usage is very similar to the showwaves filter; see the examples in that
9619 @subsection Examples
9623 Large window with logarithmic color scaling:
9625 showspectrum=s=1280x480:scale=log
9629 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
9631 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
9632 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
9638 Convert input audio to a video output, representing the samples waves.
9640 The filter accepts the following options:
9644 Specify the video size for the output. Default value is "600x240".
9649 Available values are:
9652 Draw a point for each sample.
9655 Draw a vertical line for each sample.
9658 Default value is @code{point}.
9661 Set the number of samples which are printed on the same column. A
9662 larger value will decrease the frame rate. Must be a positive
9663 integer. This option can be set only if the value for @var{rate}
9664 is not explicitly specified.
9667 Set the (approximate) output frame rate. This is done by setting the
9668 option @var{n}. Default value is "25".
9672 @subsection Examples
9676 Output the input file audio and the corresponding video representation
9679 amovie=a.mp3,asplit[out0],showwaves[out1]
9683 Create a synthetic signal and show it with showwaves, forcing a
9684 frame rate of 30 frames per second:
9686 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
9690 @section split, asplit
9692 Split input into several identical outputs.
9694 @code{asplit} works with audio input, @code{split} with video.
9696 The filter accepts a single parameter which specifies the number of outputs. If
9697 unspecified, it defaults to 2.
9699 @subsection Examples
9703 Create two separate outputs from the same input:
9705 [in] split [out0][out1]
9709 To create 3 or more outputs, you need to specify the number of
9712 [in] asplit=3 [out0][out1][out2]
9716 Create two separate outputs from the same input, one cropped and
9719 [in] split [splitout1][splitout2];
9720 [splitout1] crop=100:100:0:0 [cropout];
9721 [splitout2] pad=200:200:100:100 [padout];
9725 Create 5 copies of the input audio with @command{ffmpeg}:
9727 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
9733 Receive commands sent through a libzmq client, and forward them to
9734 filters in the filtergraph.
9736 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
9737 must be inserted between two video filters, @code{azmq} between two
9740 To enable these filters you need to install the libzmq library and
9741 headers and configure FFmpeg with @code{--enable-libzmq}.
9743 For more information about libzmq see:
9744 @url{http://www.zeromq.org/}
9746 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
9747 receives messages sent through a network interface defined by the
9748 @option{bind_address} option.
9750 The received message must be in the form:
9752 @var{TARGET} @var{COMMAND} [@var{ARG}]
9755 @var{TARGET} specifies the target of the command, usually the name of
9756 the filter class or a specific filter instance name.
9758 @var{COMMAND} specifies the name of the command for the target filter.
9760 @var{ARG} is optional and specifies the optional argument list for the
9761 given @var{COMMAND}.
9763 Upon reception, the message is processed and the corresponding command
9764 is injected into the filtergraph. Depending on the result, the filter
9765 will send a reply to the client, adopting the format:
9767 @var{ERROR_CODE} @var{ERROR_REASON}
9771 @var{MESSAGE} is optional.
9773 @subsection Examples
9775 Look at @file{tools/zmqsend} for an example of a zmq client which can
9776 be used to send commands processed by these filters.
9778 Consider the following filtergraph generated by @command{ffplay}
9780 ffplay -dumpgraph 1 -f lavfi "
9781 color=s=100x100:c=red [l];
9782 color=s=100x100:c=blue [r];
9783 nullsrc=s=200x100, zmq [bg];
9784 [bg][l] overlay [bg+l];
9785 [bg+l][r] overlay=x=100 "
9788 To change the color of the left side of the video, the following
9789 command can be used:
9791 echo Parsed_color_0 c yellow | tools/zmqsend
9794 To change the right side:
9796 echo Parsed_color_1 c pink | tools/zmqsend
9799 @c man end MULTIMEDIA FILTERS
9801 @chapter Multimedia Sources
9802 @c man begin MULTIMEDIA SOURCES
9804 Below is a description of the currently available multimedia sources.
9808 This is the same as @ref{movie} source, except it selects an audio
9814 Read audio and/or video stream(s) from a movie container.
9816 This filter accepts the following options:
9820 The name of the resource to read (not necessarily a file but also a device or a
9821 stream accessed through some protocol).
9823 @item format_name, f
9824 Specifies the format assumed for the movie to read, and can be either
9825 the name of a container or an input device. If not specified the
9826 format is guessed from @var{movie_name} or by probing.
9828 @item seek_point, sp
9829 Specifies the seek point in seconds, the frames will be output
9830 starting from this seek point, the parameter is evaluated with
9831 @code{av_strtod} so the numerical value may be suffixed by an IS
9832 postfix. Default value is "0".
9835 Specifies the streams to read. Several streams can be specified,
9836 separated by "+". The source will then have as many outputs, in the
9837 same order. The syntax is explained in the ``Stream specifiers''
9838 section in the ffmpeg manual. Two special names, "dv" and "da" specify
9839 respectively the default (best suited) video and audio stream. Default
9840 is "dv", or "da" if the filter is called as "amovie".
9842 @item stream_index, si
9843 Specifies the index of the video stream to read. If the value is -1,
9844 the best suited video stream will be automatically selected. Default
9845 value is "-1". Deprecated. If the filter is called "amovie", it will select
9846 audio instead of video.
9849 Specifies how many times to read the stream in sequence.
9850 If the value is less than 1, the stream will be read again and again.
9851 Default value is "1".
9853 Note that when the movie is looped the source timestamps are not
9854 changed, so it will generate non monotonically increasing timestamps.
9857 This filter allows to overlay a second video on top of main input of
9858 a filtergraph as shown in this graph:
9860 input -----------> deltapts0 --> overlay --> output
9863 movie --> scale--> deltapts1 -------+
9866 @subsection Examples
9870 Skip 3.2 seconds from the start of the avi file in.avi, and overlay it
9871 on top of the input labelled as "in":
9873 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
9874 [in] setpts=PTS-STARTPTS [main];
9875 [main][over] overlay=16:16 [out]
9879 Read from a video4linux2 device, and overlay it on top of the input
9882 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
9883 [in] setpts=PTS-STARTPTS [main];
9884 [main][over] overlay=16:16 [out]
9888 Read the first video stream and the audio stream with id 0x81 from
9889 dvd.vob; the video is connected to the pad named "video" and the audio is
9890 connected to the pad named "audio":
9892 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
9896 @c man end MULTIMEDIA SOURCES