1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, it is possible for filters to have multiple inputs and
8 To illustrate the sorts of things that are possible, we can
9 use a complex filter graph. For example, the following one:
12 input --> split ---------------------> overlay --> output
15 +-----> crop --> vflip -------+
18 splits the stream in two streams, sends one stream through the crop filter
19 and the vflip filter before merging it back with the other stream by
20 overlaying it on top. You can use the following command to achieve this:
23 ffmpeg -i input -vf "[in] split [T1], [T2] overlay=0:H/2 [out]; [T1] crop=iw:ih/2:0:ih/2, vflip [T2]" output
26 The result will be that in output the top half of the video is mirrored
29 Filters are loaded using the @var{-vf} or @var{-af} option passed to
30 @command{ffmpeg} or to @command{ffplay}. Filters in the same linear
31 chain are separated by commas. In our example, @var{split,
32 overlay} are in one linear chain, and @var{crop, vflip} are in
33 another. The points where the linear chains join are labeled by names
34 enclosed in square brackets. In our example, that is @var{[T1]} and
35 @var{[T2]}. The special labels @var{[in]} and @var{[out]} are the points
36 where video is input and output.
38 Some filters take in input a list of parameters: they are specified
39 after the filter name and an equal sign, and are separated from each other
42 There exist so-called @var{source filters} that do not have an
43 audio/video input, and @var{sink filters} that will not have audio/video
46 @c man end FILTERING INTRODUCTION
49 @c man begin GRAPH2DOT
51 The @file{graph2dot} program included in the FFmpeg @file{tools}
52 directory can be used to parse a filter graph description and issue a
53 corresponding textual representation in the dot language.
60 to see how to use @file{graph2dot}.
62 You can then pass the dot description to the @file{dot} program (from
63 the graphviz suite of programs) and obtain a graphical representation
66 For example the sequence of commands:
68 echo @var{GRAPH_DESCRIPTION} | \
69 tools/graph2dot -o graph.tmp && \
70 dot -Tpng graph.tmp -o graph.png && \
74 can be used to create and display an image representing the graph
75 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
76 a complete self-contained graph, with its inputs and outputs explicitly defined.
77 For example if your command line is of the form:
79 ffmpeg -i infile -vf scale=640:360 outfile
81 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
83 nullsrc,scale=640:360,nullsink
85 you may also need to set the @var{nullsrc} parameters and add a @var{format}
86 filter in order to simulate a specific input file.
90 @chapter Filtergraph description
91 @c man begin FILTERGRAPH DESCRIPTION
93 A filtergraph is a directed graph of connected filters. It can contain
94 cycles, and there can be multiple links between a pair of
95 filters. Each link has one input pad on one side connecting it to one
96 filter from which it takes its input, and one output pad on the other
97 side connecting it to the one filter accepting its output.
99 Each filter in a filtergraph is an instance of a filter class
100 registered in the application, which defines the features and the
101 number of input and output pads of the filter.
103 A filter with no input pads is called a "source", a filter with no
104 output pads is called a "sink".
106 @anchor{Filtergraph syntax}
107 @section Filtergraph syntax
109 A filtergraph can be represented using a textual representation, which is
110 recognized by the @option{-filter}/@option{-vf} and @option{-filter_complex}
111 options in @command{ffmpeg} and @option{-vf} in @command{ffplay}, and by the
112 @code{avfilter_graph_parse()}/@code{avfilter_graph_parse2()} function defined in
113 @file{libavfilter/avfiltergraph.h}.
115 A filterchain consists of a sequence of connected filters, each one
116 connected to the previous one in the sequence. A filterchain is
117 represented by a list of ","-separated filter descriptions.
119 A filtergraph consists of a sequence of filterchains. A sequence of
120 filterchains is represented by a list of ";"-separated filterchain
123 A filter is represented by a string of the form:
124 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
126 @var{filter_name} is the name of the filter class of which the
127 described filter is an instance of, and has to be the name of one of
128 the filter classes registered in the program.
129 The name of the filter class is optionally followed by a string
132 @var{arguments} is a string which contains the parameters used to
133 initialize the filter instance, and are described in the filter
136 The list of arguments can be quoted using the character "'" as initial
137 and ending mark, and the character '\' for escaping the characters
138 within the quoted text; otherwise the argument string is considered
139 terminated when the next special character (belonging to the set
140 "[]=;,") is encountered.
142 The name and arguments of the filter are optionally preceded and
143 followed by a list of link labels.
144 A link label allows to name a link and associate it to a filter output
145 or input pad. The preceding labels @var{in_link_1}
146 ... @var{in_link_N}, are associated to the filter input pads,
147 the following labels @var{out_link_1} ... @var{out_link_M}, are
148 associated to the output pads.
150 When two link labels with the same name are found in the
151 filtergraph, a link between the corresponding input and output pad is
154 If an output pad is not labelled, it is linked by default to the first
155 unlabelled input pad of the next filter in the filterchain.
156 For example in the filterchain:
158 nullsrc, split[L1], [L2]overlay, nullsink
160 the split filter instance has two output pads, and the overlay filter
161 instance two input pads. The first output pad of split is labelled
162 "L1", the first input pad of overlay is labelled "L2", and the second
163 output pad of split is linked to the second input pad of overlay,
164 which are both unlabelled.
166 In a complete filterchain all the unlabelled filter input and output
167 pads must be connected. A filtergraph is considered valid if all the
168 filter input and output pads of all the filterchains are connected.
170 Libavfilter will automatically insert scale filters where format
171 conversion is required. It is possible to specify swscale flags
172 for those automatically inserted scalers by prepending
173 @code{sws_flags=@var{flags};}
174 to the filtergraph description.
176 Follows a BNF description for the filtergraph syntax:
178 @var{NAME} ::= sequence of alphanumeric characters and '_'
179 @var{LINKLABEL} ::= "[" @var{NAME} "]"
180 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
181 @var{FILTER_ARGUMENTS} ::= sequence of chars (eventually quoted)
182 @var{FILTER} ::= [@var{LINKNAMES}] @var{NAME} ["=" @var{ARGUMENTS}] [@var{LINKNAMES}]
183 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
184 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
187 @section Notes on filtergraph escaping
189 Some filter arguments require the use of special characters, typically
190 @code{:} to separate key=value pairs in a named options list. In this
191 case the user should perform a first level escaping when specifying
192 the filter arguments. For example, consider the following literal
193 string to be embedded in the @ref{drawtext} filter arguments:
195 this is a 'string': may contain one, or more, special characters
198 Since @code{:} is special for the filter arguments syntax, it needs to
199 be escaped, so you get:
201 text=this is a \'string\'\: may contain one, or more, special characters
204 A second level of escaping is required when embedding the filter
205 arguments in a filtergraph description, in order to escape all the
206 filtergraph special characters. Thus the example above becomes:
208 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
211 Finally an additional level of escaping may be needed when writing the
212 filtergraph description in a shell command, which depends on the
213 escaping rules of the adopted shell. For example, assuming that
214 @code{\} is special and needs to be escaped with another @code{\}, the
215 previous string will finally result in:
217 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
220 Sometimes, it might be more convenient to employ quoting in place of
221 escaping. For example the string:
223 Caesar: tu quoque, Brute, fili mi
226 Can be quoted in the filter arguments as:
228 text='Caesar: tu quoque, Brute, fili mi'
231 And finally inserted in a filtergraph like:
233 drawtext=text=\'Caesar: tu quoque\, Brute\, fili mi\'
236 See the ``Quoting and escaping'' section in the ffmpeg-utils manual
237 for more information about the escaping and quoting rules adopted by
240 @c man end FILTERGRAPH DESCRIPTION
242 @chapter Audio Filters
243 @c man begin AUDIO FILTERS
245 When you configure your FFmpeg build, you can disable any of the
246 existing filters using @code{--disable-filters}.
247 The configure output will show the audio filters included in your
250 Below is a description of the currently available audio filters.
254 Convert the input audio format to the specified formats.
256 The filter accepts a string of the form:
257 "@var{sample_format}:@var{channel_layout}".
259 @var{sample_format} specifies the sample format, and can be a string or the
260 corresponding numeric value defined in @file{libavutil/samplefmt.h}. Use 'p'
261 suffix for a planar sample format.
263 @var{channel_layout} specifies the channel layout, and can be a string
264 or the corresponding number value defined in @file{libavutil/channel_layout.h}.
266 The special parameter "auto", signifies that the filter will
267 automatically select the output format depending on the output filter.
269 Some examples follow.
273 Convert input to float, planar, stereo:
279 Convert input to unsigned 8-bit, automatically select out channel layout:
287 Set output format constraints for the input audio. The framework will
288 negotiate the most appropriate format to minimize conversions.
290 The filter accepts the following named parameters:
294 A comma-separated list of requested sample formats.
297 A comma-separated list of requested sample rates.
299 @item channel_layouts
300 A comma-separated list of requested channel layouts.
304 If a parameter is omitted, all values are allowed.
306 For example to force the output to either unsigned 8-bit or signed 16-bit stereo:
308 aformat='sample_fmts=u8,s16:channel_layouts=stereo'
313 Merge two or more audio streams into a single multi-channel stream.
315 The filter accepts the following named options:
320 Set the number of inputs. Default is 2.
324 If the channel layouts of the inputs are disjoint, and therefore compatible,
325 the channel layout of the output will be set accordingly and the channels
326 will be reordered as necessary. If the channel layouts of the inputs are not
327 disjoint, the output will have all the channels of the first input then all
328 the channels of the second input, in that order, and the channel layout of
329 the output will be the default value corresponding to the total number of
332 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
333 is FC+BL+BR, then the output will be in 5.1, with the channels in the
334 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
335 first input, b1 is the first channel of the second input).
337 On the other hand, if both input are in stereo, the output channels will be
338 in the default order: a1, a2, b1, b2, and the channel layout will be
339 arbitrarily set to 4.0, which may or may not be the expected value.
341 All inputs must have the same sample rate, and format.
343 If inputs do not have the same duration, the output will stop with the
346 Example: merge two mono files into a stereo stream:
348 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
351 Example: multiple merges:
354 amovie=input.mkv:si=0 [a0];
355 amovie=input.mkv:si=1 [a1];
356 amovie=input.mkv:si=2 [a2];
357 amovie=input.mkv:si=3 [a3];
358 amovie=input.mkv:si=4 [a4];
359 amovie=input.mkv:si=5 [a5];
360 [a0][a1][a2][a3][a4][a5] amerge=inputs=6" -c:a pcm_s16le output.mkv
365 Mixes multiple audio inputs into a single output.
369 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
371 will mix 3 input audio streams to a single output with the same duration as the
372 first input and a dropout transition time of 3 seconds.
374 The filter accepts the following named parameters:
378 Number of inputs. If unspecified, it defaults to 2.
381 How to determine the end-of-stream.
385 Duration of longest input. (default)
388 Duration of shortest input.
391 Duration of first input.
395 @item dropout_transition
396 Transition time, in seconds, for volume renormalization when an input
397 stream ends. The default value is 2 seconds.
403 Pass the audio source unchanged to the output.
407 Pad the end of a audio stream with silence, this can be used together with
408 -shortest to extend audio streams to the same length as the video stream.
413 Resample the input audio to the specified parameters, using the
414 libswresample library. If none are specified then the filter will
415 automatically convert between its input and output.
417 This filter is also able to stretch/squeeze the audio data to make it match
418 the timestamps or to inject silence / cut out audio to make it match the
419 timestamps, do a combination of both or do neither.
421 The filter accepts the syntax
422 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
423 expresses a sample rate and @var{resampler_options} is a list of
424 @var{key}=@var{value} pairs, separated by ":". See the
425 ffmpeg-resampler manual for the complete list of supported options.
427 For example, to resample the input audio to 44100Hz:
432 To stretch/squeeze samples to the given timestamps, with a maximum of 1000
433 samples per second compensation:
438 @section asetnsamples
440 Set the number of samples per each output audio frame.
442 The last output packet may contain a different number of samples, as
443 the filter will flush all the remaining samples when the input audio
446 The filter accepts parameters as a list of @var{key}=@var{value} pairs,
451 @item nb_out_samples, n
452 Set the number of frames per each output audio frame. The number is
453 intended as the number of samples @emph{per each channel}.
454 Default value is 1024.
457 If set to 1, the filter will pad the last audio frame with zeroes, so
458 that the last frame will contain the same number of samples as the
459 previous ones. Default value is 1.
462 For example, to set the number of per-frame samples to 1234 and
463 disable padding for the last frame, use:
465 asetnsamples=n=1234:p=0
470 Show a line containing various information for each input audio frame.
471 The input audio is not modified.
473 The shown line contains a sequence of key/value pairs of the form
474 @var{key}:@var{value}.
476 A description of each shown parameter follows:
480 sequential number of the input frame, starting from 0
483 Presentation timestamp of the input frame, in time base units; the time base
484 depends on the filter input pad, and is usually 1/@var{sample_rate}.
487 presentation timestamp of the input frame in seconds
490 position of the frame in the input stream, -1 if this information in
491 unavailable and/or meaningless (for example in case of synthetic audio)
500 sample rate for the audio frame
503 number of samples (per channel) in the frame
506 Adler-32 checksum (printed in hexadecimal) of the audio data. For planar audio
507 the data is treated as if all the planes were concatenated.
509 @item plane_checksums
510 A list of Adler-32 checksums for each data plane.
515 Split input audio into several identical outputs.
517 The filter accepts a single parameter which specifies the number of outputs. If
518 unspecified, it defaults to 2.
522 [in] asplit [out0][out1]
525 will create two separate outputs from the same input.
527 To create 3 or more outputs, you need to specify the number of
530 [in] asplit=3 [out0][out1][out2]
534 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
536 will create 5 copies of the input audio.
541 Forward two audio streams and control the order the buffers are forwarded.
543 The argument to the filter is an expression deciding which stream should be
544 forwarded next: if the result is negative, the first stream is forwarded; if
545 the result is positive or zero, the second stream is forwarded. It can use
546 the following variables:
550 number of buffers forwarded so far on each stream
552 number of samples forwarded so far on each stream
554 current timestamp of each stream
557 The default value is @code{t1-t2}, which means to always forward the stream
558 that has a smaller timestamp.
560 Example: stress-test @code{amerge} by randomly sending buffers on the wrong
561 input, while avoiding too much of a desynchronization:
563 amovie=file.ogg [a] ; amovie=file.mp3 [b] ;
564 [a] [b] astreamsync=(2*random(1))-1+tanh(5*(t1-t2)) [a2] [b2] ;
572 The filter accepts exactly one parameter, the audio tempo. If not
573 specified then the filter will assume nominal 1.0 tempo. Tempo must
574 be in the [0.5, 2.0] range.
576 For example, to slow down audio to 80% tempo:
581 For example, to speed up audio to 125% tempo:
588 Make audio easier to listen to on headphones.
590 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
591 so that when listened to on headphones the stereo image is moved from
592 inside your head (standard for headphones) to outside and in front of
593 the listener (standard for speakers).
599 Mix channels with specific gain levels. The filter accepts the output
600 channel layout followed by a set of channels definitions.
602 This filter is also designed to remap efficiently the channels of an audio
605 The filter accepts parameters of the form:
606 "@var{l}:@var{outdef}:@var{outdef}:..."
610 output channel layout or number of channels
613 output channel specification, of the form:
614 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
617 output channel to define, either a channel name (FL, FR, etc.) or a channel
618 number (c0, c1, etc.)
621 multiplicative coefficient for the channel, 1 leaving the volume unchanged
624 input channel to use, see out_name for details; it is not possible to mix
625 named and numbered input channels
628 If the `=' in a channel specification is replaced by `<', then the gains for
629 that specification will be renormalized so that the total is 1, thus
630 avoiding clipping noise.
632 @subsection Mixing examples
634 For example, if you want to down-mix from stereo to mono, but with a bigger
635 factor for the left channel:
637 pan=1:c0=0.9*c0+0.1*c1
640 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
643 pan=stereo: FL < FL + 0.5*FC + 0.6*BL + 0.6*SL : FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
646 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
647 that should be preferred (see "-ac" option) unless you have very specific
650 @subsection Remapping examples
652 The channel remapping will be effective if, and only if:
655 @item gain coefficients are zeroes or ones,
656 @item only one input per channel output,
659 If all these conditions are satisfied, the filter will notify the user ("Pure
660 channel mapping detected"), and use an optimized and lossless method to do the
663 For example, if you have a 5.1 source and want a stereo audio stream by
664 dropping the extra channels:
666 pan="stereo: c0=FL : c1=FR"
669 Given the same source, you can also switch front left and front right channels
670 and keep the input channel layout:
672 pan="5.1: c0=c1 : c1=c0 : c2=c2 : c3=c3 : c4=c4 : c5=c5"
675 If the input is a stereo audio stream, you can mute the front left channel (and
676 still keep the stereo channel layout) with:
681 Still with a stereo audio stream input, you can copy the right channel in both
682 front left and right:
684 pan="stereo: c0=FR : c1=FR"
687 @section silencedetect
689 Detect silence in an audio stream.
691 This filter logs a message when it detects that the input audio volume is less
692 or equal to a noise tolerance value for a duration greater or equal to the
693 minimum detected noise duration.
695 The printed times and duration are expressed in seconds.
699 Set silence duration until notification (default is 2 seconds).
702 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
703 specified value) or amplitude ratio. Default is -60dB, or 0.001.
706 Detect 5 seconds of silence with -50dB noise tolerance:
708 silencedetect=n=-50dB:d=5
711 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
712 tolerance in @file{silence.mp3}:
714 ffmpeg -f lavfi -i amovie=silence.mp3,silencedetect=noise=0.0001 -f null -
718 Synchronize audio data with timestamps by squeezing/stretching it and/or
719 dropping samples/adding silence when needed.
721 As an alternative, you can use @ref{aresample} to do squeezing/stretching.
723 The filter accepts the following named parameters:
727 Enable stretching/squeezing the data to make it match the timestamps. Disabled
728 by default. When disabled, time gaps are covered with silence.
731 Minimum difference between timestamps and audio data (in seconds) to trigger
732 adding/dropping samples. Default value is 0.1. If you get non-perfect sync with
733 this filter, try setting this parameter to 0.
736 Maximum compensation in samples per second. Relevant only with compensate=1.
740 Assume the first pts should be this value. The time base is 1 / sample rate.
741 This allows for padding/trimming at the start of stream. By default, no
742 assumption is made about the first frame's expected pts, so no padding or
743 trimming is done. For example, this could be set to 0 to pad the beginning with
744 silence if an audio stream starts after the video stream or to trim any samples
745 with a negative pts due to encoder delay.
749 @section channelsplit
750 Split each channel in input audio stream into a separate output stream.
752 This filter accepts the following named parameters:
755 Channel layout of the input stream. Default is "stereo".
758 For example, assuming a stereo input MP3 file
760 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
762 will create an output Matroska file with two audio streams, one containing only
763 the left channel and the other the right channel.
765 To split a 5.1 WAV file into per-channel files
767 ffmpeg -i in.wav -filter_complex
768 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
769 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
770 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
775 Remap input channels to new locations.
777 This filter accepts the following named parameters:
780 Channel layout of the output stream.
783 Map channels from input to output. The argument is a comma-separated list of
784 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
785 @var{in_channel} form. @var{in_channel} can be either the name of the input
786 channel (e.g. FL for front left) or its index in the input channel layout.
787 @var{out_channel} is the name of the output channel or its index in the output
788 channel layout. If @var{out_channel} is not given then it is implicitly an
789 index, starting with zero and increasing by one for each mapping.
792 If no mapping is present, the filter will implicitly map input channels to
793 output channels preserving index.
795 For example, assuming a 5.1+downmix input MOV file
797 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL\,DR-FR' out.wav
799 will create an output WAV file tagged as stereo from the downmix channels of
802 To fix a 5.1 WAV improperly encoded in AAC's native channel order
804 ffmpeg -i in.wav -filter 'channelmap=1\,2\,0\,5\,3\,4:channel_layout=5.1' out.wav
808 Join multiple input streams into one multi-channel stream.
810 The filter accepts the following named parameters:
814 Number of input streams. Defaults to 2.
817 Desired output channel layout. Defaults to stereo.
820 Map channels from inputs to output. The argument is a comma-separated list of
821 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
822 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
823 can be either the name of the input channel (e.g. FL for front left) or its
824 index in the specified input stream. @var{out_channel} is the name of the output
828 The filter will attempt to guess the mappings when those are not specified
829 explicitly. It does so by first trying to find an unused matching input channel
830 and if that fails it picks the first unused input channel.
832 E.g. to join 3 inputs (with properly set channel layouts)
834 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
837 To build a 5.1 output from 6 single-channel streams:
839 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
840 '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'
845 Convert the audio sample format, sample rate and channel layout. This filter is
846 not meant to be used directly.
850 Adjust the input audio volume.
852 The filter accepts the following named parameters. If the key of the
853 first options is omitted, the arguments are interpreted according to
854 the following syntax:
856 volume=@var{volume}:@var{precision}
862 Expresses how the audio volume will be increased or decreased.
864 Output values are clipped to the maximum value.
866 The output audio volume is given by the relation:
868 @var{output_volume} = @var{volume} * @var{input_volume}
871 Default value for @var{volume} is 1.0.
874 Set the mathematical precision.
876 This determines which input sample formats will be allowed, which affects the
877 precision of the volume scaling.
881 8-bit fixed-point; limits input sample format to U8, S16, and S32.
883 32-bit floating-point; limits input sample format to FLT. (default)
885 64-bit floating-point; limits input sample format to DBL.
893 Halve the input audio volume:
897 volume=volume=-6.0206dB
900 In all the above example the named key for @option{volume} can be
901 omitted, for example like in:
907 Increase input audio power by 6 decibels using fixed-point precision:
909 volume=volume=6dB:precision=fixed
913 @section volumedetect
915 Detect the volume of the input video.
917 The filter has no parameters. The input is not modified. Statistics about
918 the volume will be printed in the log when the input stream end is reached.
920 In particular it will show the mean volume (root mean square), maximum
921 volume (on a per-sample basis), and the beginning of an histogram of the
922 registered volume values (from the maximum value to a cumulated 1/1000 of
925 All volumes are in decibels relative to the maximum PCM value.
927 Here is an excerpt of the output:
929 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
930 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
931 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
932 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
933 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
934 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
935 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
936 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
937 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
943 The mean square energy is approximately -27 dB, or 10^-2.7.
945 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
947 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
950 In other words, raising the volume by +4 dB does not cause any clipping,
951 raising it by +5 dB causes clipping for 6 samples, etc.
953 @c man end AUDIO FILTERS
955 @chapter Audio Sources
956 @c man begin AUDIO SOURCES
958 Below is a description of the currently available audio sources.
962 Buffer audio frames, and make them available to the filter chain.
964 This source is mainly intended for a programmatic use, in particular
965 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
967 It accepts the following mandatory parameters:
968 @var{sample_rate}:@var{sample_fmt}:@var{channel_layout}
973 The sample rate of the incoming audio buffers.
976 The sample format of the incoming audio buffers.
977 Either a sample format name or its corresponging integer representation from
978 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
981 The channel layout of the incoming audio buffers.
982 Either a channel layout name from channel_layout_map in
983 @file{libavutil/channel_layout.c} or its corresponding integer representation
984 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
990 abuffer=44100:s16p:stereo
993 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
994 Since the sample format with name "s16p" corresponds to the number
995 6 and the "stereo" channel layout corresponds to the value 0x3, this is
1003 Generate an audio signal specified by an expression.
1005 This source accepts in input one or more expressions (one for each
1006 channel), which are evaluated and used to generate a corresponding
1009 It accepts the syntax: @var{exprs}[::@var{options}].
1010 @var{exprs} is a list of expressions separated by ":", one for each
1011 separate channel. In case the @var{channel_layout} is not
1012 specified, the selected channel layout depends on the number of
1013 provided expressions.
1015 @var{options} is an optional sequence of @var{key}=@var{value} pairs,
1018 The description of the accepted options follows.
1022 @item channel_layout, c
1023 Set the channel layout. The number of channels in the specified layout
1024 must be equal to the number of specified expressions.
1027 Set the minimum duration of the sourced audio. See the function
1028 @code{av_parse_time()} for the accepted format.
1029 Note that the resulting duration may be greater than the specified
1030 duration, as the generated audio is always cut at the end of a
1033 If not specified, or the expressed duration is negative, the audio is
1034 supposed to be generated forever.
1037 Set the number of samples per channel per each output frame,
1040 @item sample_rate, s
1041 Specify the sample rate, default to 44100.
1044 Each expression in @var{exprs} can contain the following constants:
1048 number of the evaluated sample, starting from 0
1051 time of the evaluated sample expressed in seconds, starting from 0
1058 @subsection Examples
1070 Generate a sin signal with frequency of 440 Hz, set sample rate to
1073 aevalsrc="sin(440*2*PI*t)::s=8000"
1077 Generate a two channels signal, specify the channel layout (Front
1078 Center + Back Center) explicitly:
1080 aevalsrc="sin(420*2*PI*t):cos(430*2*PI*t)::c=FC|BC"
1084 Generate white noise:
1086 aevalsrc="-2+random(0)"
1090 Generate an amplitude modulated signal:
1092 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
1096 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
1098 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) : 0.1*sin(2*PI*(360+2.5/2)*t)"
1105 Null audio source, return unprocessed audio frames. It is mainly useful
1106 as a template and to be employed in analysis / debugging tools, or as
1107 the source for filters which ignore the input data (for example the sox
1110 It accepts an optional sequence of @var{key}=@var{value} pairs,
1113 The description of the accepted options follows.
1117 @item sample_rate, s
1118 Specify the sample rate, and defaults to 44100.
1120 @item channel_layout, cl
1122 Specify the channel layout, and can be either an integer or a string
1123 representing a channel layout. The default value of @var{channel_layout}
1126 Check the channel_layout_map definition in
1127 @file{libavutil/channel_layout.c} for the mapping between strings and
1128 channel layout values.
1131 Set the number of samples per requested frames.
1135 Follow some examples:
1137 # set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
1138 anullsrc=r=48000:cl=4
1141 anullsrc=r=48000:cl=mono
1145 Buffer audio frames, and make them available to the filter chain.
1147 This source is not intended to be part of user-supplied graph descriptions but
1148 for insertion by calling programs through the interface defined in
1149 @file{libavfilter/buffersrc.h}.
1151 It accepts the following named parameters:
1155 Timebase which will be used for timestamps of submitted frames. It must be
1156 either a floating-point number or in @var{numerator}/@var{denominator} form.
1162 Name of the sample format, as returned by @code{av_get_sample_fmt_name()}.
1164 @item channel_layout
1165 Channel layout of the audio data, in the form that can be accepted by
1166 @code{av_get_channel_layout()}.
1169 All the parameters need to be explicitly defined.
1173 Synthesize a voice utterance using the libflite library.
1175 To enable compilation of this filter you need to configure FFmpeg with
1176 @code{--enable-libflite}.
1178 Note that the flite library is not thread-safe.
1180 The source accepts parameters as a list of @var{key}=@var{value} pairs,
1183 The description of the accepted parameters follows.
1188 If set to 1, list the names of the available voices and exit
1189 immediately. Default value is 0.
1192 Set the maximum number of samples per frame. Default value is 512.
1195 Set the filename containing the text to speak.
1198 Set the text to speak.
1201 Set the voice to use for the speech synthesis. Default value is
1202 @code{kal}. See also the @var{list_voices} option.
1205 @subsection Examples
1209 Read from file @file{speech.txt}, and synthetize the text using the
1210 standard flite voice:
1212 flite=textfile=speech.txt
1216 Read the specified text selecting the @code{slt} voice:
1218 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
1222 Input text to ffmpeg:
1224 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
1228 Make @file{ffplay} speak the specified text, using @code{flite} and
1229 the @code{lavfi} device:
1231 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
1235 For more information about libflite, check:
1236 @url{http://www.speech.cs.cmu.edu/flite/}
1238 @c man end AUDIO SOURCES
1240 @chapter Audio Sinks
1241 @c man begin AUDIO SINKS
1243 Below is a description of the currently available audio sinks.
1245 @section abuffersink
1247 Buffer audio frames, and make them available to the end of filter chain.
1249 This sink is mainly intended for programmatic use, in particular
1250 through the interface defined in @file{libavfilter/buffersink.h}.
1252 It requires a pointer to an AVABufferSinkContext structure, which
1253 defines the incoming buffers' formats, to be passed as the opaque
1254 parameter to @code{avfilter_init_filter} for initialization.
1258 Null audio sink, do absolutely nothing with the input audio. It is
1259 mainly useful as a template and to be employed in analysis / debugging
1262 @section abuffersink
1263 This sink is intended for programmatic use. Frames that arrive on this sink can
1264 be retrieved by the calling program using the interface defined in
1265 @file{libavfilter/buffersink.h}.
1267 This filter accepts no parameters.
1269 @c man end AUDIO SINKS
1271 @chapter Video Filters
1272 @c man begin VIDEO FILTERS
1274 When you configure your FFmpeg build, you can disable any of the
1275 existing filters using @code{--disable-filters}.
1276 The configure output will show the video filters included in your
1279 Below is a description of the currently available video filters.
1281 @section alphaextract
1283 Extract the alpha component from the input as a grayscale video. This
1284 is especially useful with the @var{alphamerge} filter.
1288 Add or replace the alpha component of the primary input with the
1289 grayscale value of a second input. This is intended for use with
1290 @var{alphaextract} to allow the transmission or storage of frame
1291 sequences that have alpha in a format that doesn't support an alpha
1294 For example, to reconstruct full frames from a normal YUV-encoded video
1295 and a separate video created with @var{alphaextract}, you might use:
1297 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
1300 Since this filter is designed for reconstruction, it operates on frame
1301 sequences without considering timestamps, and terminates when either
1302 input reaches end of stream. This will cause problems if your encoding
1303 pipeline drops frames. If you're trying to apply an image as an
1304 overlay to a video stream, consider the @var{overlay} filter instead.
1308 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
1309 and libavformat to work. On the other hand, it is limited to ASS (Advanced
1310 Substation Alpha) subtitles files.
1314 Compute the bounding box for the non-black pixels in the input frame
1317 This filter computes the bounding box containing all the pixels with a
1318 luminance value greater than the minimum allowed value.
1319 The parameters describing the bounding box are printed on the filter
1322 @section blackdetect
1324 Detect video intervals that are (almost) completely black. Can be
1325 useful to detect chapter transitions, commercials, or invalid
1326 recordings. Output lines contains the time for the start, end and
1327 duration of the detected black interval expressed in seconds.
1329 In order to display the output lines, you need to set the loglevel at
1330 least to the AV_LOG_INFO value.
1332 This filter accepts a list of options in the form of
1333 @var{key}=@var{value} pairs separated by ":". A description of the
1334 accepted options follows.
1337 @item black_min_duration, d
1338 Set the minimum detected black duration expressed in seconds. It must
1339 be a non-negative floating point number.
1341 Default value is 2.0.
1343 @item picture_black_ratio_th, pic_th
1344 Set the threshold for considering a picture "black".
1345 Express the minimum value for the ratio:
1347 @var{nb_black_pixels} / @var{nb_pixels}
1350 for which a picture is considered black.
1351 Default value is 0.98.
1353 @item pixel_black_th, pix_th
1354 Set the threshold for considering a pixel "black".
1356 The threshold expresses the maximum pixel luminance value for which a
1357 pixel is considered "black". The provided value is scaled according to
1358 the following equation:
1360 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
1363 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
1364 the input video format, the range is [0-255] for YUV full-range
1365 formats and [16-235] for YUV non full-range formats.
1367 Default value is 0.10.
1370 The following example sets the maximum pixel threshold to the minimum
1371 value, and detects only black intervals of 2 or more seconds:
1373 blackdetect=d=2:pix_th=0.00
1378 Detect frames that are (almost) completely black. Can be useful to
1379 detect chapter transitions or commercials. Output lines consist of
1380 the frame number of the detected frame, the percentage of blackness,
1381 the position in the file if known or -1 and the timestamp in seconds.
1383 In order to display the output lines, you need to set the loglevel at
1384 least to the AV_LOG_INFO value.
1386 The filter accepts the syntax:
1388 blackframe[=@var{amount}:[@var{threshold}]]
1391 @var{amount} is the percentage of the pixels that have to be below the
1392 threshold, and defaults to 98.
1394 @var{threshold} is the threshold below which a pixel value is
1395 considered black, and defaults to 32.
1399 Apply boxblur algorithm to the input video.
1401 This filter accepts the parameters:
1402 @var{luma_radius}:@var{luma_power}:@var{chroma_radius}:@var{chroma_power}:@var{alpha_radius}:@var{alpha_power}
1404 Chroma and alpha parameters are optional, if not specified they default
1405 to the corresponding values set for @var{luma_radius} and
1408 @var{luma_radius}, @var{chroma_radius}, and @var{alpha_radius} represent
1409 the radius in pixels of the box used for blurring the corresponding
1410 input plane. They are expressions, and can contain the following
1414 the input width and height in pixels
1417 the input chroma image width and height in pixels
1420 horizontal and vertical chroma subsample values. For example for the
1421 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
1424 The radius must be a non-negative number, and must not be greater than
1425 the value of the expression @code{min(w,h)/2} for the luma and alpha planes,
1426 and of @code{min(cw,ch)/2} for the chroma planes.
1428 @var{luma_power}, @var{chroma_power}, and @var{alpha_power} represent
1429 how many times the boxblur filter is applied to the corresponding
1432 Some examples follow:
1437 Apply a boxblur filter with luma, chroma, and alpha radius
1444 Set luma radius to 2, alpha and chroma radius to 0
1450 Set luma and chroma radius to a fraction of the video dimension
1452 boxblur=min(h\,w)/10:1:min(cw\,ch)/10:1
1457 @section colormatrix
1459 The colormatrix filter allows conversion between any of the following color
1460 space: BT.709 (@var{bt709}), BT.601 (@var{bt601}), SMPTE-240M (@var{smpte240m})
1461 and FCC (@var{fcc}).
1463 The syntax of the parameters is @var{source}:@var{destination}:
1466 colormatrix=bt601:smpte240m
1471 Copy the input source unchanged to the output. Mainly useful for
1476 Crop the input video.
1478 This filter accepts a list of @var{key}=@var{value} pairs as argument,
1479 separated by ':'. If the key of the first options is omitted, the
1480 arguments are interpreted according to the syntax
1481 @var{out_w}:@var{out_h}:@var{x}:@var{y}:@var{keep_aspect}.
1483 A description of the accepted options follows:
1486 Set the crop area width. It defaults to @code{iw}.
1487 This expression is evaluated only once during the filter
1491 Set the crop area width. It defaults to @code{ih}.
1492 This expression is evaluated only once during the filter
1496 Set the expression for the x top-left coordinate of the cropped area.
1497 It defaults to @code{(in_w-out_w)/2}.
1498 This expression is evaluated per-frame.
1501 Set the expression for the y top-left coordinate of the cropped area.
1502 It defaults to @code{(in_h-out_h)/2}.
1503 This expression is evaluated per-frame.
1506 If set to 1 will force the output display aspect ratio
1507 to be the same of the input, by changing the output sample aspect
1508 ratio. It defaults to 0.
1511 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
1512 expressions containing the following constants:
1516 the computed values for @var{x} and @var{y}. They are evaluated for
1520 the input width and height
1523 same as @var{in_w} and @var{in_h}
1526 the output (cropped) width and height
1529 same as @var{out_w} and @var{out_h}
1532 same as @var{iw} / @var{ih}
1535 input sample aspect ratio
1538 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
1541 horizontal and vertical chroma subsample values. For example for the
1542 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
1545 the number of input frame, starting from 0
1548 the position in the file of the input frame, NAN if unknown
1551 timestamp expressed in seconds, NAN if the input timestamp is unknown
1555 The expression for @var{out_w} may depend on the value of @var{out_h},
1556 and the expression for @var{out_h} may depend on @var{out_w}, but they
1557 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
1558 evaluated after @var{out_w} and @var{out_h}.
1560 The @var{x} and @var{y} parameters specify the expressions for the
1561 position of the top-left corner of the output (non-cropped) area. They
1562 are evaluated for each frame. If the evaluated value is not valid, it
1563 is approximated to the nearest valid value.
1565 The expression for @var{x} may depend on @var{y}, and the expression
1566 for @var{y} may depend on @var{x}.
1568 @subsection Examples
1571 Crop area with size 100x100 at position (12,34).
1576 Using named options, the example above becomes:
1578 crop=w=100:h=100:x=12:y=34
1582 Crop the central input area with size 100x100:
1588 Crop the central input area with size 2/3 of the input video:
1590 crop=2/3*in_w:2/3*in_h
1594 Crop the input video central square:
1600 Delimit the rectangle with the top-left corner placed at position
1601 100:100 and the right-bottom corner corresponding to the right-bottom
1602 corner of the input image:
1604 crop=in_w-100:in_h-100:100:100
1608 Crop 10 pixels from the left and right borders, and 20 pixels from
1609 the top and bottom borders
1611 crop=in_w-2*10:in_h-2*20
1615 Keep only the bottom right quarter of the input image:
1617 crop=in_w/2:in_h/2:in_w/2:in_h/2
1621 Crop height for getting Greek harmony:
1623 crop=in_w:1/PHI*in_w
1627 Appply trembling effect:
1629 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)
1633 Apply erratic camera effect depending on timestamp:
1635 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)"
1639 Set x depending on the value of y:
1641 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
1647 Auto-detect crop size.
1649 Calculate necessary cropping parameters and prints the recommended
1650 parameters through the logging system. The detected dimensions
1651 correspond to the non-black area of the input video.
1653 It accepts the syntax:
1655 cropdetect[=@var{limit}[:@var{round}[:@var{reset}]]]
1661 Threshold, which can be optionally specified from nothing (0) to
1662 everything (255), defaults to 24.
1665 Value which the width/height should be divisible by, defaults to
1666 16. The offset is automatically adjusted to center the video. Use 2 to
1667 get only even dimensions (needed for 4:2:2 video). 16 is best when
1668 encoding to most video codecs.
1671 Counter that determines after how many frames cropdetect will reset
1672 the previously detected largest video area and start over to detect
1673 the current optimal crop area. Defaults to 0.
1675 This can be useful when channel logos distort the video area. 0
1676 indicates never reset and return the largest area encountered during
1682 This filter drops frames that do not differ greatly from the previous
1683 frame in order to reduce framerate. The main use of this filter is
1684 for very-low-bitrate encoding (e.g. streaming over dialup modem), but
1685 it could in theory be used for fixing movies that were
1686 inverse-telecined incorrectly.
1688 It accepts the following parameters:
1689 @var{max}:@var{hi}:@var{lo}:@var{frac}.
1694 Set the maximum number of consecutive frames which can be dropped (if
1695 positive), or the minimum interval between dropped frames (if
1696 negative). If the value is 0, the frame is dropped unregarding the
1697 number of previous sequentially dropped frames.
1702 Set the dropping threshold values.
1704 Values for @var{hi} and @var{lo} are for 8x8 pixel blocks and
1705 represent actual pixel value differences, so a threshold of 64
1706 corresponds to 1 unit of difference for each pixel, or the same spread
1707 out differently over the block.
1709 A frame is a candidate for dropping if no 8x8 blocks differ by more
1710 than a threshold of @var{hi}, and if no more than @var{frac} blocks (1
1711 meaning the whole image) differ by more than a threshold of @var{lo}.
1713 Default value for @var{hi} is 64*12, default value for @var{lo} is
1714 64*5, and default value for @var{frac} is 0.33.
1719 Suppress a TV station logo by a simple interpolation of the surrounding
1720 pixels. Just set a rectangle covering the logo and watch it disappear
1721 (and sometimes something even uglier appear - your mileage may vary).
1723 The filter accepts parameters as a string of the form
1724 "@var{x}:@var{y}:@var{w}:@var{h}:@var{band}", or as a list of
1725 @var{key}=@var{value} pairs, separated by ":".
1727 The description of the accepted parameters follows.
1732 Specify the top left corner coordinates of the logo. They must be
1736 Specify the width and height of the logo to clear. They must be
1740 Specify the thickness of the fuzzy edge of the rectangle (added to
1741 @var{w} and @var{h}). The default value is 4.
1744 When set to 1, a green rectangle is drawn on the screen to simplify
1745 finding the right @var{x}, @var{y}, @var{w}, @var{h} parameters, and
1746 @var{band} is set to 4. The default value is 0.
1750 Some examples follow.
1755 Set a rectangle covering the area with top left corner coordinates 0,0
1756 and size 100x77, setting a band of size 10:
1758 delogo=0:0:100:77:10
1762 As the previous example, but use named options:
1764 delogo=x=0:y=0:w=100:h=77:band=10
1771 Attempt to fix small changes in horizontal and/or vertical shift. This
1772 filter helps remove camera shake from hand-holding a camera, bumping a
1773 tripod, moving on a vehicle, etc.
1775 The filter accepts parameters as a string of the form
1776 "@var{x}:@var{y}:@var{w}:@var{h}:@var{rx}:@var{ry}:@var{edge}:@var{blocksize}:@var{contrast}:@var{search}:@var{filename}"
1778 A description of the accepted parameters follows.
1783 Specify a rectangular area where to limit the search for motion
1785 If desired the search for motion vectors can be limited to a
1786 rectangular area of the frame defined by its top left corner, width
1787 and height. These parameters have the same meaning as the drawbox
1788 filter which can be used to visualise the position of the bounding
1791 This is useful when simultaneous movement of subjects within the frame
1792 might be confused for camera motion by the motion vector search.
1794 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
1795 then the full frame is used. This allows later options to be set
1796 without specifying the bounding box for the motion vector search.
1798 Default - search the whole frame.
1801 Specify the maximum extent of movement in x and y directions in the
1802 range 0-64 pixels. Default 16.
1805 Specify how to generate pixels to fill blanks at the edge of the
1806 frame. An integer from 0 to 3 as follows:
1809 Fill zeroes at blank locations
1811 Original image at blank locations
1813 Extruded edge value at blank locations
1815 Mirrored edge at blank locations
1818 The default setting is mirror edge at blank locations.
1821 Specify the blocksize to use for motion search. Range 4-128 pixels,
1825 Specify the contrast threshold for blocks. Only blocks with more than
1826 the specified contrast (difference between darkest and lightest
1827 pixels) will be considered. Range 1-255, default 125.
1830 Specify the search strategy 0 = exhaustive search, 1 = less exhaustive
1831 search. Default - exhaustive search.
1834 If set then a detailed log of the motion search is written to the
1841 Draw a colored box on the input image.
1843 The filter accepts parameters as a list of @var{key}=@var{value} pairs,
1846 The description of the accepted parameters follows.
1850 Specify the top left corner coordinates of the box. Default to 0.
1854 Specify the width and height of the box, if 0 they are interpreted as
1855 the input width and height. Default to 0.
1858 Specify the color of the box to write, it can be the name of a color
1859 (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
1860 value @code{invert} is used, the box edge color is the same as the
1861 video with inverted luma.
1864 Set the thickness of the box edge. Default value is @code{4}.
1867 If the key of the first options is omitted, the arguments are
1868 interpreted according to the following syntax:
1870 drawbox=@var{x}:@var{y}:@var{width}:@var{height}:@var{color}:@var{thickness}
1873 Some examples follow:
1876 Draw a black box around the edge of the input image:
1882 Draw a box with color red and an opacity of 50%:
1884 drawbox=10:20:200:60:red@@0.5
1887 The previous example can be specified as:
1889 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
1893 Fill the box with pink color:
1895 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
1902 Draw text string or text from specified file on top of video using the
1903 libfreetype library.
1905 To enable compilation of this filter you need to configure FFmpeg with
1906 @code{--enable-libfreetype}.
1910 The filter accepts parameters as a list of @var{key}=@var{value} pairs,
1913 The description of the accepted parameters follows.
1918 Used to draw a box around text using background color.
1919 Value should be either 1 (enable) or 0 (disable).
1920 The default value of @var{box} is 0.
1923 The color to be used for drawing box around text.
1924 Either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
1925 (e.g. "0xff00ff"), possibly followed by an alpha specifier.
1926 The default value of @var{boxcolor} is "white".
1929 Set an expression which specifies if the text should be drawn. If the
1930 expression evaluates to 0, the text is not drawn. This is useful for
1931 specifying that the text should be drawn only when specific conditions
1934 Default value is "1".
1936 See below for the list of accepted constants and functions.
1939 Select how the @var{text} is expanded. Can be either @code{none},
1940 @code{strftime} (default for compatibity reasons but deprecated) or
1941 @code{normal}. See the @ref{drawtext_expansion, Text expansion} section
1945 If true, check and fix text coords to avoid clipping.
1948 The color to be used for drawing fonts.
1949 Either a string (e.g. "red") or in 0xRRGGBB[AA] format
1950 (e.g. "0xff000033"), possibly followed by an alpha specifier.
1951 The default value of @var{fontcolor} is "black".
1954 The font file to be used for drawing text. Path must be included.
1955 This parameter is mandatory.
1958 The font size to be used for drawing text.
1959 The default value of @var{fontsize} is 16.
1962 Flags to be used for loading the fonts.
1964 The flags map the corresponding flags supported by libfreetype, and are
1965 a combination of the following values:
1972 @item vertical_layout
1973 @item force_autohint
1976 @item ignore_global_advance_width
1978 @item ignore_transform
1985 Default value is "render".
1987 For more information consult the documentation for the FT_LOAD_*
1991 The color to be used for drawing a shadow behind the drawn text. It
1992 can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
1993 form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
1994 The default value of @var{shadowcolor} is "black".
1996 @item shadowx, shadowy
1997 The x and y offsets for the text shadow position with respect to the
1998 position of the text. They can be either positive or negative
1999 values. Default value for both is "0".
2002 The size in number of spaces to use for rendering the tab.
2006 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
2007 format. It can be used with or without text parameter. @var{timecode_rate}
2008 option must be specified.
2010 @item timecode_rate, rate, r
2011 Set the timecode frame rate (timecode only).
2014 The text string to be drawn. The text must be a sequence of UTF-8
2016 This parameter is mandatory if no file is specified with the parameter
2020 A text file containing text to be drawn. The text must be a sequence
2021 of UTF-8 encoded characters.
2023 This parameter is mandatory if no text string is specified with the
2024 parameter @var{text}.
2026 If both @var{text} and @var{textfile} are specified, an error is thrown.
2029 If set to 1, the @var{textfile} will be reloaded before each frame.
2030 Be sure to update it atomically, or it may be read partially, or even fail.
2033 The expressions which specify the offsets where text will be drawn
2034 within the video frame. They are relative to the top/left border of the
2037 The default value of @var{x} and @var{y} is "0".
2039 See below for the list of accepted constants and functions.
2042 The parameters for @var{x} and @var{y} are expressions containing the
2043 following constants and functions:
2047 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
2050 horizontal and vertical chroma subsample values. For example for the
2051 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2054 the height of each text line
2062 @item max_glyph_a, ascent
2063 the maximum distance from the baseline to the highest/upper grid
2064 coordinate used to place a glyph outline point, for all the rendered
2066 It is a positive value, due to the grid's orientation with the Y axis
2069 @item max_glyph_d, descent
2070 the maximum distance from the baseline to the lowest grid coordinate
2071 used to place a glyph outline point, for all the rendered glyphs.
2072 This is a negative value, due to the grid's orientation, with the Y axis
2076 maximum glyph height, that is the maximum height for all the glyphs
2077 contained in the rendered text, it is equivalent to @var{ascent} -
2081 maximum glyph width, that is the maximum width for all the glyphs
2082 contained in the rendered text
2085 the number of input frame, starting from 0
2087 @item rand(min, max)
2088 return a random number included between @var{min} and @var{max}
2091 input sample aspect ratio
2094 timestamp expressed in seconds, NAN if the input timestamp is unknown
2097 the height of the rendered text
2100 the width of the rendered text
2103 the x and y offset coordinates where the text is drawn.
2105 These parameters allow the @var{x} and @var{y} expressions to refer
2106 each other, so you can for example specify @code{y=x/dar}.
2109 If libavfilter was built with @code{--enable-fontconfig}, then
2110 @option{fontfile} can be a fontconfig pattern or omitted.
2112 @anchor{drawtext_expansion}
2113 @subsection Text expansion
2115 If @option{expansion} is set to @code{strftime} (which is the default for
2116 now), the filter recognizes strftime() sequences in the provided text and
2117 expands them accordingly. Check the documentation of strftime(). This
2118 feature is deprecated.
2120 If @option{expansion} is set to @code{none}, the text is printed verbatim.
2122 If @option{expansion} is set to @code{normal} (which will be the default),
2123 the following expansion mechanism is used.
2125 The backslash character '\', followed by any character, always expands to
2126 the second character.
2128 Sequence of the form @code{%@{...@}} are expanded. The text between the
2129 braces is a function name, possibly followed by arguments separated by ':'.
2130 If the arguments contain special characters or delimiters (':' or '@}'),
2131 they should be escaped.
2133 Note that they probably must also be escaped as the value for the
2134 @option{text} option in the filter argument string and as the filter
2135 argument in the filter graph description, and possibly also for the shell,
2136 that makes up to four levels of escaping; using a text file avoids these
2139 The following functions are available:
2144 The expression evaluation result.
2146 It must take one argument specifying the expression to be evaluated,
2147 which accepts the same constants and functions as the @var{x} and
2148 @var{y} values. Note that not all constants should be used, for
2149 example the text size is not known when evaluating the expression, so
2150 the constants @var{text_w} and @var{text_h} will have an undefined
2154 The time at which the filter is running, expressed in UTC.
2155 It can accept an argument: a strftime() format string.
2158 The time at which the filter is running, expressed in the local time zone.
2159 It can accept an argument: a strftime() format string.
2162 The frame number, starting from 0.
2165 The timestamp of the current frame, in seconds, with microsecond accuracy.
2169 @subsection Examples
2171 Some examples follow.
2176 Draw "Test Text" with font FreeSerif, using the default values for the
2177 optional parameters.
2180 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
2184 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
2185 and y=50 (counting from the top-left corner of the screen), text is
2186 yellow with a red box around it. Both the text and the box have an
2190 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
2191 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
2194 Note that the double quotes are not necessary if spaces are not used
2195 within the parameter list.
2198 Show the text at the center of the video frame:
2200 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h-line_h)/2"
2204 Show a text line sliding from right to left in the last row of the video
2205 frame. The file @file{LONG_LINE} is assumed to contain a single line
2208 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
2212 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
2214 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
2218 Draw a single green letter "g", at the center of the input video.
2219 The glyph baseline is placed at half screen height.
2221 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
2225 Show text for 1 second every 3 seconds:
2227 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:draw=lt(mod(t\,3)\,1):text='blink'"
2231 Use fontconfig to set the font. Note that the colons need to be escaped.
2233 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
2237 Print the date of a real-time encoding (see strftime(3)):
2239 drawtext='fontfile=FreeSans.ttf:expansion=normal:text=%@{localtime:%a %b %d %Y@}'
2244 For more information about libfreetype, check:
2245 @url{http://www.freetype.org/}.
2247 For more information about fontconfig, check:
2248 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
2252 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
2254 This filter accepts the following optional named parameters:
2258 Set low and high threshold values used by the Canny thresholding
2261 The high threshold selects the "strong" edge pixels, which are then
2262 connected through 8-connectivity with the "weak" edge pixels selected
2263 by the low threshold.
2265 @var{low} and @var{high} threshold values must be choosen in the range
2266 [0,1], and @var{low} should be lesser or equal to @var{high}.
2268 Default value for @var{low} is @code{20/255}, and default value for @var{high}
2274 edgedetect=low=0.1:high=0.4
2279 Apply fade-in/out effect to input video.
2281 It accepts the parameters:
2282 @var{type}:@var{start_frame}:@var{nb_frames}[:@var{options}]
2284 @var{type} specifies if the effect type, can be either "in" for
2285 fade-in, or "out" for a fade-out effect.
2287 @var{start_frame} specifies the number of the start frame for starting
2288 to apply the fade effect.
2290 @var{nb_frames} specifies the number of frames for which the fade
2291 effect has to last. At the end of the fade-in effect the output video
2292 will have the same intensity as the input video, at the end of the
2293 fade-out transition the output video will be completely black.
2295 @var{options} is an optional sequence of @var{key}=@var{value} pairs,
2296 separated by ":". The description of the accepted options follows.
2303 @item start_frame, s
2304 See @var{start_frame}.
2307 See @var{nb_frames}.
2310 If set to 1, fade only alpha channel, if one exists on the input.
2314 A few usage examples follow, usable too as test scenarios.
2316 # fade in first 30 frames of video
2319 # fade out last 45 frames of a 200-frame video
2322 # fade in first 25 frames and fade out last 25 frames of a 1000-frame video
2323 fade=in:0:25, fade=out:975:25
2325 # make first 5 frames black, then fade in from frame 5-24
2328 # fade in alpha over first 25 frames of video
2329 fade=in:0:25:alpha=1
2334 Extract a single field from an interlaced image using stride
2335 arithmetic to avoid wasting CPU time. The output frames are marked as
2338 This filter accepts the following named options:
2341 Specify whether to extract the top (if the value is @code{0} or
2342 @code{top}) or the bottom field (if the value is @code{1} or
2346 If the option key is not specified, the first value sets the @var{type}
2347 option. For example:
2359 Transform the field order of the input video.
2361 It accepts one parameter which specifies the required field order that
2362 the input interlaced video will be transformed to. The parameter can
2363 assume one of the following values:
2367 output bottom field first
2369 output top field first
2372 Default value is "tff".
2374 Transformation is achieved by shifting the picture content up or down
2375 by one line, and filling the remaining line with appropriate picture content.
2376 This method is consistent with most broadcast field order converters.
2378 If the input video is not flagged as being interlaced, or it is already
2379 flagged as being of the required output field order then this filter does
2380 not alter the incoming video.
2382 This filter is very useful when converting to or from PAL DV material,
2383 which is bottom field first.
2387 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
2392 Buffer input images and send them when they are requested.
2394 This filter is mainly useful when auto-inserted by the libavfilter
2397 The filter does not take parameters.
2401 Convert the input video to one of the specified pixel formats.
2402 Libavfilter will try to pick one that is supported for the input to
2405 The filter accepts a list of pixel format names, separated by ":",
2406 for example "yuv420p:monow:rgb24".
2408 Some examples follow:
2410 # convert the input video to the format "yuv420p"
2413 # convert the input video to any of the formats in the list
2414 format=yuv420p:yuv444p:yuv410p
2419 Convert the video to specified constant framerate by duplicating or dropping
2420 frames as necessary.
2422 This filter accepts the following named parameters:
2426 Desired output framerate. The default is @code{25}.
2431 Possible values are:
2434 zero round towards 0
2438 round towards -infinity
2440 round towards +infinity
2444 The default is @code{near}.
2448 Alternatively, the options can be specified as a flat string:
2449 @var{fps}[:@var{round}].
2451 See also the @ref{setpts} filter.
2455 Select one frame every N.
2457 This filter accepts in input a string representing a positive
2458 integer. Default argument is @code{1}.
2463 Apply a frei0r effect to the input video.
2465 To enable compilation of this filter you need to install the frei0r
2466 header and configure FFmpeg with @code{--enable-frei0r}.
2468 The filter supports the syntax:
2470 @var{filter_name}[@{:|=@}@var{param1}:@var{param2}:...:@var{paramN}]
2473 @var{filter_name} is the name of the frei0r effect to load. If the
2474 environment variable @env{FREI0R_PATH} is defined, the frei0r effect
2475 is searched in each one of the directories specified by the colon (or
2476 semicolon on Windows platforms) separated list in @env{FREIOR_PATH},
2477 otherwise in the standard frei0r paths, which are in this order:
2478 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
2479 @file{/usr/lib/frei0r-1/}.
2481 @var{param1}, @var{param2}, ... , @var{paramN} specify the parameters
2482 for the frei0r effect.
2484 A frei0r effect parameter can be a boolean (whose values are specified
2485 with "y" and "n"), a double, a color (specified by the syntax
2486 @var{R}/@var{G}/@var{B}, @var{R}, @var{G}, and @var{B} being float
2487 numbers from 0.0 to 1.0) or by an @code{av_parse_color()} color
2488 description), a position (specified by the syntax @var{X}/@var{Y},
2489 @var{X} and @var{Y} being float numbers) and a string.
2491 The number and kind of parameters depend on the loaded effect. If an
2492 effect parameter is not specified the default value is set.
2494 Some examples follow:
2498 Apply the distort0r effect, set the first two double parameters:
2500 frei0r=distort0r:0.5:0.01
2504 Apply the colordistance effect, take a color as first parameter:
2506 frei0r=colordistance:0.2/0.3/0.4
2507 frei0r=colordistance:violet
2508 frei0r=colordistance:0x112233
2512 Apply the perspective effect, specify the top left and top right image
2515 frei0r=perspective:0.2/0.2:0.8/0.2
2519 For more information see:
2520 @url{http://frei0r.dyne.org}
2524 The filter takes one, two or three equations as parameter, separated by ':'.
2525 The first equation is mandatory and applies to the luma plane. The two
2526 following are respectively for chroma blue and chroma red planes.
2528 The filter syntax allows named parameters:
2532 the luminance expression
2534 the chrominance blue expression
2536 the chrominance red expression
2539 If one of the chrominance expression is not defined, it falls back on the other
2540 one. If none of them are specified, they will evaluate the luminance
2543 The expressions can use the following variables and functions:
2547 The sequential number of the filtered frame, starting from @code{0}.
2550 The coordinates of the current sample.
2553 The width and height of the image.
2556 Width and height scale depending on the currently filtered plane. It is the
2557 ratio between the corresponding luma plane number of pixels and the current
2558 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
2559 @code{0.5,0.5} for chroma planes.
2562 Time of the current frame, expressed in seconds.
2565 Return the value of the pixel at location (@var{x},@var{y}) of the current
2569 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
2573 Return the value of the pixel at location (@var{x},@var{y}) of the
2574 blue-difference chroma plane.
2577 Return the value of the pixel at location (@var{x},@var{y}) of the
2578 red-difference chroma plane.
2581 For functions, if @var{x} and @var{y} are outside the area, the value will be
2582 automatically clipped to the closer edge.
2584 Some examples follow:
2588 Flip the image horizontally:
2594 Generate a bidimensional sine wave, with angle @code{PI/3} and a
2595 wavelength of 100 pixels:
2597 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
2601 Generate a fancy enigmatic moving light:
2603 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
2609 Fix the banding artifacts that are sometimes introduced into nearly flat
2610 regions by truncation to 8bit color depth.
2611 Interpolate the gradients that should go where the bands are, and
2614 This filter is designed for playback only. Do not use it prior to
2615 lossy compression, because compression tends to lose the dither and
2616 bring back the bands.
2618 The filter accepts a list of options in the form of @var{key}=@var{value} pairs
2619 separated by ":". A description of the accepted options follows.
2624 The maximum amount by which the filter will change
2625 any one pixel. Also the threshold for detecting nearly flat
2626 regions. Acceptable values range from @code{0.51} to @code{64}, default value
2630 The neighborhood to fit the gradient to. A larger
2631 radius makes for smoother gradients, but also prevents the filter from
2632 modifying the pixels near detailed regions. Acceptable values are
2633 @code{8-32}, default value is @code{16}.
2637 Alternatively, the options can be specified as a flat string:
2638 @var{strength}[:@var{radius}]
2640 @subsection Examples
2644 Apply the filter with a @code{3.5} strength and radius of @code{8}:
2650 Specify radius, omitting the strength (which will fall-back to the default
2660 Flip the input video horizontally.
2662 For example to horizontally flip the input video with @command{ffmpeg}:
2664 ffmpeg -i in.avi -vf "hflip" out.avi
2668 This filter applies a global color histogram equalization on a
2671 It can be used to correct video that has a compressed range of pixel
2672 intensities. The filter redistributes the pixel intensities to
2673 equalize their distribution across the intensity range. It may be
2674 viewed as an "automatically adjusting contrast filter". This filter is
2675 useful only for correcting degraded or poorly captured source
2678 The filter accepts parameters as a list of @var{key}=@var{value}
2679 pairs, separated by ":". If the key of the first options is omitted,
2680 the arguments are interpreted according to syntax
2681 @var{strength}:@var{intensity}:@var{antibanding}.
2683 This filter accepts the following named options:
2687 Determine the amount of equalization to be applied. As the strength
2688 is reduced, the distribution of pixel intensities more-and-more
2689 approaches that of the input frame. The value must be a float number
2690 in the range [0,1] and defaults to 0.200.
2693 Set the maximum intensity that can generated and scale the output
2694 values appropriately. The strength should be set as desired and then
2695 the intensity can be limited if needed to avoid washing-out. The value
2696 must be a float number in the range [0,1] and defaults to 0.210.
2699 Set the antibanding level. If enabled the filter will randomly vary
2700 the luminance of output pixels by a small amount to avoid banding of
2701 the histogram. Possible values are @code{none}, @code{weak} or
2702 @code{strong}. It defaults to @code{none}.
2707 High precision/quality 3d denoise filter. This filter aims to reduce
2708 image noise producing smooth images and making still images really
2709 still. It should enhance compressibility.
2711 It accepts the following optional parameters:
2712 @var{luma_spatial}:@var{chroma_spatial}:@var{luma_tmp}:@var{chroma_tmp}
2716 a non-negative float number which specifies spatial luma strength,
2719 @item chroma_spatial
2720 a non-negative float number which specifies spatial chroma strength,
2721 defaults to 3.0*@var{luma_spatial}/4.0
2724 a float number which specifies luma temporal strength, defaults to
2725 6.0*@var{luma_spatial}/4.0
2728 a float number which specifies chroma temporal strength, defaults to
2729 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}
2734 Modify the hue and/or the saturation of the input.
2736 This filter accepts the following optional named options:
2740 Specify the hue angle as a number of degrees. It accepts a float
2741 number or an expression, and defaults to 0.0.
2744 Specify the hue angle as a number of degrees. It accepts a float
2745 number or an expression, and defaults to 0.0.
2748 Specify the saturation in the [-10,10] range. It accepts a float number and
2752 The @var{h}, @var{H} and @var{s} parameters are expressions containing the
2753 following constants:
2757 frame count of the input frame starting from 0
2760 presentation timestamp of the input frame expressed in time base units
2763 frame rate of the input video, NAN if the input frame rate is unknown
2766 timestamp expressed in seconds, NAN if the input timestamp is unknown
2769 time base of the input video
2772 The options can also be set using the syntax: @var{hue}:@var{saturation}
2774 In this case @var{hue} is expressed in degrees.
2776 Some examples follow:
2779 Set the hue to 90 degrees and the saturation to 1.0:
2785 Same command but expressing the hue in radians:
2791 Same command without named options, hue must be expressed in degrees:
2797 Note that "h:s" syntax does not support expressions for the values of
2798 h and s, so the following example will issue an error:
2804 Rotate hue and make the saturation swing between 0
2805 and 2 over a period of 1 second:
2807 hue="H=2*PI*t: s=sin(2*PI*t)+1"
2811 Apply a 3 seconds saturation fade-in effect starting at 0:
2816 The general fade-in expression can be written as:
2818 hue="s=min(0\, max((t-START)/DURATION\, 1))"
2822 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
2824 hue="s=max(0\, min(1\, (8-t)/3))"
2827 The general fade-out expression can be written as:
2829 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
2834 @subsection Commands
2836 This filter supports the following command:
2839 Modify the hue and/or the saturation of the input video.
2840 The command accepts the same named options and syntax than when calling the
2841 filter from the command-line.
2843 If a parameter is omitted, it is kept at its current value.
2848 Interlaceing detect filter. This filter tries to detect if the input is
2849 interlaced or progressive. Top or bottom field first.
2853 Deinterlace input video by applying Donald Graft's adaptive kernel
2854 deinterling. Work on interlaced parts of a video to produce
2857 This filter accepts parameters as a list of @var{key}=@var{value}
2858 pairs, separated by ":". If the key of the first options is omitted,
2859 the arguments are interpreted according to the following syntax:
2860 @var{thresh}:@var{map}:@var{order}:@var{sharp}:@var{twoway}.
2862 The description of the accepted parameters follows.
2866 Set the threshold which affects the filter's tolerance when
2867 determining if a pixel line must be processed. It must be an integer
2868 in the range [0,255] and defaults to 10. A value of 0 will result in
2869 applying the process on every pixels.
2872 Paint pixels exceeding the threshold value to white if set to 1.
2876 Set the fields order. Swap fields if set to 1, leave fields alone if
2880 Enable additional sharpening if set to 1. Default is 0.
2883 Enable twoway sharpening if set to 1. Default is 0.
2886 @subsection Examples
2890 Apply default values:
2892 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
2896 Enable additional sharpening:
2902 Paint processed pixels in white:
2908 @section lut, lutrgb, lutyuv
2910 Compute a look-up table for binding each pixel component input value
2911 to an output value, and apply it to input video.
2913 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
2914 to an RGB input video.
2916 These filters accept in input a ":"-separated list of options, which
2917 specify the expressions used for computing the lookup table for the
2918 corresponding pixel component values.
2920 The @var{lut} filter requires either YUV or RGB pixel formats in
2921 input, and accepts the options:
2923 @item @var{c0} (first pixel component)
2924 @item @var{c1} (second pixel component)
2925 @item @var{c2} (third pixel component)
2926 @item @var{c3} (fourth pixel component, corresponds to the alpha component)
2929 The exact component associated to each option depends on the format in
2932 The @var{lutrgb} filter requires RGB pixel formats in input, and
2933 accepts the options:
2935 @item @var{r} (red component)
2936 @item @var{g} (green component)
2937 @item @var{b} (blue component)
2938 @item @var{a} (alpha component)
2941 The @var{lutyuv} filter requires YUV pixel formats in input, and
2942 accepts the options:
2944 @item @var{y} (Y/luminance component)
2945 @item @var{u} (U/Cb component)
2946 @item @var{v} (V/Cr component)
2947 @item @var{a} (alpha component)
2950 The expressions can contain the following constants and functions:
2954 the input width and height
2957 input value for the pixel component
2960 the input value clipped in the @var{minval}-@var{maxval} range
2963 maximum value for the pixel component
2966 minimum value for the pixel component
2969 the negated value for the pixel component value clipped in the
2970 @var{minval}-@var{maxval} range , it corresponds to the expression
2971 "maxval-clipval+minval"
2974 the computed value in @var{val} clipped in the
2975 @var{minval}-@var{maxval} range
2977 @item gammaval(gamma)
2978 the computed gamma correction value of the pixel component value
2979 clipped in the @var{minval}-@var{maxval} range, corresponds to the
2981 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
2985 All expressions default to "val".
2987 Some examples follow:
2989 # negate input video
2990 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
2991 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
2993 # the above is the same as
2994 lutrgb="r=negval:g=negval:b=negval"
2995 lutyuv="y=negval:u=negval:v=negval"
3000 # remove chroma components, turns the video into a graytone image
3001 lutyuv="u=128:v=128"
3003 # apply a luma burning effect
3006 # remove green and blue components
3009 # set a constant alpha channel value on input
3010 format=rgba,lutrgb=a="maxval-minval/2"
3012 # correct luminance gamma by a 0.5 factor
3013 lutyuv=y=gammaval(0.5)
3018 Apply an MPlayer filter to the input video.
3020 This filter provides a wrapper around most of the filters of
3023 This wrapper is considered experimental. Some of the wrapped filters
3024 may not work properly and we may drop support for them, as they will
3025 be implemented natively into FFmpeg. Thus you should avoid
3026 depending on them when writing portable scripts.
3028 The filters accepts the parameters:
3029 @var{filter_name}[:=]@var{filter_params}
3031 @var{filter_name} is the name of a supported MPlayer filter,
3032 @var{filter_params} is a string containing the parameters accepted by
3035 The list of the currently supported filters follows:
3069 The parameter syntax and behavior for the listed filters are the same
3070 of the corresponding MPlayer filters. For detailed instructions check
3071 the "VIDEO FILTERS" section in the MPlayer manual.
3073 Some examples follow:
3076 Adjust gamma, brightness, contrast:
3082 Add temporal noise to input video:
3088 See also mplayer(1), @url{http://www.mplayerhq.hu/}.
3094 This filter accepts an integer in input, if non-zero it negates the
3095 alpha component (if available). The default value in input is 0.
3099 Force libavfilter not to use any of the specified pixel formats for the
3100 input to the next filter.
3102 The filter accepts a list of pixel format names, separated by ":",
3103 for example "yuv420p:monow:rgb24".
3105 Some examples follow:
3107 # force libavfilter to use a format different from "yuv420p" for the
3108 # input to the vflip filter
3109 noformat=yuv420p,vflip
3111 # convert the input video to any of the formats not contained in the list
3112 noformat=yuv420p:yuv444p:yuv410p
3117 Pass the video source unchanged to the output.
3121 Apply video transform using libopencv.
3123 To enable this filter install libopencv library and headers and
3124 configure FFmpeg with @code{--enable-libopencv}.
3126 The filter takes the parameters: @var{filter_name}@{:=@}@var{filter_params}.
3128 @var{filter_name} is the name of the libopencv filter to apply.
3130 @var{filter_params} specifies the parameters to pass to the libopencv
3131 filter. If not specified the default values are assumed.
3133 Refer to the official libopencv documentation for more precise
3135 @url{http://opencv.willowgarage.com/documentation/c/image_filtering.html}
3137 Follows the list of supported libopencv filters.
3142 Dilate an image by using a specific structuring element.
3143 This filter corresponds to the libopencv function @code{cvDilate}.
3145 It accepts the parameters: @var{struct_el}:@var{nb_iterations}.
3147 @var{struct_el} represents a structuring element, and has the syntax:
3148 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
3150 @var{cols} and @var{rows} represent the number of columns and rows of
3151 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
3152 point, and @var{shape} the shape for the structuring element, and
3153 can be one of the values "rect", "cross", "ellipse", "custom".
3155 If the value for @var{shape} is "custom", it must be followed by a
3156 string of the form "=@var{filename}". The file with name
3157 @var{filename} is assumed to represent a binary image, with each
3158 printable character corresponding to a bright pixel. When a custom
3159 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
3160 or columns and rows of the read file are assumed instead.
3162 The default value for @var{struct_el} is "3x3+0x0/rect".
3164 @var{nb_iterations} specifies the number of times the transform is
3165 applied to the image, and defaults to 1.
3167 Follow some example:
3169 # use the default values
3172 # dilate using a structuring element with a 5x5 cross, iterate two times
3173 ocv=dilate=5x5+2x2/cross:2
3175 # read the shape from the file diamond.shape, iterate two times
3176 # the file diamond.shape may contain a pattern of characters like this:
3182 # the specified cols and rows are ignored (but not the anchor point coordinates)
3183 ocv=0x0+2x2/custom=diamond.shape:2
3188 Erode an image by using a specific structuring element.
3189 This filter corresponds to the libopencv function @code{cvErode}.
3191 The filter accepts the parameters: @var{struct_el}:@var{nb_iterations},
3192 with the same syntax and semantics as the @ref{dilate} filter.
3196 Smooth the input video.
3198 The filter takes the following parameters:
3199 @var{type}:@var{param1}:@var{param2}:@var{param3}:@var{param4}.
3201 @var{type} is the type of smooth filter to apply, and can be one of
3202 the following values: "blur", "blur_no_scale", "median", "gaussian",
3203 "bilateral". The default value is "gaussian".
3205 @var{param1}, @var{param2}, @var{param3}, and @var{param4} are
3206 parameters whose meanings depend on smooth type. @var{param1} and
3207 @var{param2} accept integer positive values or 0, @var{param3} and
3208 @var{param4} accept float values.
3210 The default value for @var{param1} is 3, the default value for the
3211 other parameters is 0.
3213 These parameters correspond to the parameters assigned to the
3214 libopencv function @code{cvSmooth}.
3219 Overlay one video on top of another.
3221 It takes two inputs and one output, the first input is the "main"
3222 video on which the second input is overlayed.
3224 This filter accepts a list of @var{key}=@var{value} pairs as argument,
3225 separated by ":". If the key of the first options is omitted, the
3226 arguments are interpreted according to the syntax @var{x}:@var{y}.
3228 A description of the accepted options follows.
3232 Set the expression for the x and y coordinates of the overlayed video
3233 on the main video. Default value is 0.
3235 The @var{x} and @var{y} expressions can contain the following
3238 @item main_w, main_h
3239 main input width and height
3242 same as @var{main_w} and @var{main_h}
3244 @item overlay_w, overlay_h
3245 overlay input width and height
3248 same as @var{overlay_w} and @var{overlay_h}
3252 If set to 1, force the filter to accept inputs in the RGB
3253 color space. Default value is 0.
3256 Be aware that frames are taken from each input video in timestamp
3257 order, hence, if their initial timestamps differ, it is a a good idea
3258 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
3259 have them begin in the same zero timestamp, as it does the example for
3260 the @var{movie} filter.
3262 You can chain together more overlays but you should test the
3263 efficiency of such approach.
3265 @subsection Examples
3269 Draw the overlay at 10 pixels from the bottom right corner of the main
3272 overlay=main_w-overlay_w-10:main_h-overlay_h-10
3275 Using named options the example above becomes:
3277 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
3281 Insert a transparent PNG logo in the bottom left corner of the input,
3282 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
3284 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
3288 Insert 2 different transparent PNG logos (second logo on bottom
3289 right corner) using the @command{ffmpeg} tool:
3291 ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=10:H-h-10,overlay=W-w-10:H-h-10' output
3295 Add a transparent color layer on top of the main video, WxH specifies
3296 the size of the main input to the overlay filter:
3298 color=red@@.3:WxH [over]; [in][over] overlay [out]
3302 Play an original video and a filtered version (here with the deshake
3303 filter) side by side using the @command{ffplay} tool:
3305 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
3308 The above command is the same as:
3310 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
3314 Chain several overlays in cascade:
3316 nullsrc=s=200x200 [bg];
3317 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
3318 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
3319 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
3320 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
3321 [in3] null, [mid2] overlay=100:100 [out0]
3328 Add paddings to the input image, and places the original input at the
3329 given coordinates @var{x}, @var{y}.
3331 It accepts the following parameters:
3332 @var{width}:@var{height}:@var{x}:@var{y}:@var{color}.
3334 The parameters @var{width}, @var{height}, @var{x}, and @var{y} are
3335 expressions containing the following constants:
3339 the input video width and height
3342 same as @var{in_w} and @var{in_h}
3345 the output width and height, that is the size of the padded area as
3346 specified by the @var{width} and @var{height} expressions
3349 same as @var{out_w} and @var{out_h}
3352 x and y offsets as specified by the @var{x} and @var{y}
3353 expressions, or NAN if not yet specified
3356 same as @var{iw} / @var{ih}
3359 input sample aspect ratio
3362 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
3365 horizontal and vertical chroma subsample values. For example for the
3366 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3369 Follows the description of the accepted parameters.
3374 Specify the size of the output image with the paddings added. If the
3375 value for @var{width} or @var{height} is 0, the corresponding input size
3376 is used for the output.
3378 The @var{width} expression can reference the value set by the
3379 @var{height} expression, and vice versa.
3381 The default value of @var{width} and @var{height} is 0.
3385 Specify the offsets where to place the input image in the padded area
3386 with respect to the top/left border of the output image.
3388 The @var{x} expression can reference the value set by the @var{y}
3389 expression, and vice versa.
3391 The default value of @var{x} and @var{y} is 0.
3395 Specify the color of the padded area, it can be the name of a color
3396 (case insensitive match) or a 0xRRGGBB[AA] sequence.
3398 The default value of @var{color} is "black".
3402 @subsection Examples
3406 Add paddings with color "violet" to the input video. Output video
3407 size is 640x480, the top-left corner of the input video is placed at
3410 pad=640:480:0:40:violet
3414 Pad the input to get an output with dimensions increased by 3/2,
3415 and put the input video at the center of the padded area:
3417 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
3421 Pad the input to get a squared output with size equal to the maximum
3422 value between the input width and height, and put the input video at
3423 the center of the padded area:
3425 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
3429 Pad the input to get a final w/h ratio of 16:9:
3431 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
3435 In case of anamorphic video, in order to set the output display aspect
3436 correctly, it is necessary to use @var{sar} in the expression,
3437 according to the relation:
3439 (ih * X / ih) * sar = output_dar
3440 X = output_dar / sar
3443 Thus the previous example needs to be modified to:
3445 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
3449 Double output size and put the input video in the bottom-right
3450 corner of the output padded area:
3452 pad="2*iw:2*ih:ow-iw:oh-ih"
3456 @section pixdesctest
3458 Pixel format descriptor test filter, mainly useful for internal
3459 testing. The output video should be equal to the input video.
3463 format=monow, pixdesctest
3466 can be used to test the monowhite pixel format descriptor definition.
3470 Enable the specified chain of postprocessing subfilters using libpostproc. This
3471 library should be automatically selected with a GPL build (@code{--enable-gpl}).
3472 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
3473 Each subfilter and some options have a short and a long name that can be used
3474 interchangeably, i.e. dr/dering are the same.
3476 All subfilters share common options to determine their scope:
3480 Honor the quality commands for this subfilter.
3483 Do chrominance filtering, too (default).
3486 Do luminance filtering only (no chrominance).
3489 Do chrominance filtering only (no luminance).
3492 These options can be appended after the subfilter name, separated by a ':'.
3494 Available subfilters are:
3497 @item hb/hdeblock[:difference[:flatness]]
3498 Horizontal deblocking filter
3501 Difference factor where higher values mean more deblocking (default: @code{32}).
3503 Flatness threshold where lower values mean more deblocking (default: @code{39}).
3506 @item vb/vdeblock[:difference[:flatness]]
3507 Vertical deblocking filter
3510 Difference factor where higher values mean more deblocking (default: @code{32}).
3512 Flatness threshold where lower values mean more deblocking (default: @code{39}).
3515 @item ha/hadeblock[:difference[:flatness]]
3516 Accurate horizontal deblocking filter
3519 Difference factor where higher values mean more deblocking (default: @code{32}).
3521 Flatness threshold where lower values mean more deblocking (default: @code{39}).
3524 @item va/vadeblock[:difference[:flatness]]
3525 Accurate vertical deblocking filter
3528 Difference factor where higher values mean more deblocking (default: @code{32}).
3530 Flatness threshold where lower values mean more deblocking (default: @code{39}).
3534 The horizontal and vertical deblocking filters share the difference and
3535 flatness values so you cannot set different horizontal and vertical
3540 Experimental horizontal deblocking filter
3543 Experimental vertical deblocking filter
3548 @item tn/tmpnoise[:threshold1[:threshold2[:threshold3]]], temporal noise reducer
3551 larger -> stronger filtering
3553 larger -> stronger filtering
3555 larger -> stronger filtering
3558 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
3561 Stretch luminance to @code{0-255}.
3564 @item lb/linblenddeint
3565 Linear blend deinterlacing filter that deinterlaces the given block by
3566 filtering all lines with a @code{(1 2 1)} filter.
3568 @item li/linipoldeint
3569 Linear interpolating deinterlacing filter that deinterlaces the given block by
3570 linearly interpolating every second line.
3572 @item ci/cubicipoldeint
3573 Cubic interpolating deinterlacing filter deinterlaces the given block by
3574 cubically interpolating every second line.
3576 @item md/mediandeint
3577 Median deinterlacing filter that deinterlaces the given block by applying a
3578 median filter to every second line.
3580 @item fd/ffmpegdeint
3581 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
3582 second line with a @code{(-1 4 2 4 -1)} filter.
3585 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
3586 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
3588 @item fq/forceQuant[:quantizer]
3589 Overrides the quantizer table from the input with the constant quantizer you
3597 Default pp filter combination (@code{hb:a,vb:a,dr:a})
3600 Fast pp filter combination (@code{h1:a,v1:a,dr:a})
3603 High quality pp filter combination (@code{ha:a:128:7,va:a,dr:a})
3606 @subsection Examples
3610 Apply horizontal and vertical deblocking, deringing and automatic
3611 brightness/contrast:
3617 Apply default filters without brightness/contrast correction:
3623 Apply default filters and temporal denoiser:
3625 pp=default/tmpnoise:1:2:3
3629 Apply deblocking on luminance only, and switch vertical deblocking on or off
3630 automatically depending on available CPU time:
3638 Suppress a TV station logo, using an image file to determine which
3639 pixels comprise the logo. It works by filling in the pixels that
3640 comprise the logo with neighboring pixels.
3642 This filter requires one argument which specifies the filter bitmap
3643 file, which can be any image format supported by libavformat. The
3644 width and height of the image file must match those of the video
3645 stream being processed.
3647 Pixels in the provided bitmap image with a value of zero are not
3648 considered part of the logo, non-zero pixels are considered part of
3649 the logo. If you use white (255) for the logo and black (0) for the
3650 rest, you will be safe. For making the filter bitmap, it is
3651 recommended to take a screen capture of a black frame with the logo
3652 visible, and then using a threshold filter followed by the erode
3653 filter once or twice.
3655 If needed, little splotches can be fixed manually. Remember that if
3656 logo pixels are not covered, the filter quality will be much
3657 reduced. Marking too many pixels as part of the logo does not hurt as
3658 much, but it will increase the amount of blurring needed to cover over
3659 the image and will destroy more information than necessary, and extra
3660 pixels will slow things down on a large logo.
3664 Scale (resize) the input video, using the libswscale library.
3666 The scale filter forces the output display aspect ratio to be the same
3667 of the input, by changing the output sample aspect ratio.
3669 This filter accepts a list of named options in the form of
3670 @var{key}=@var{value} pairs separated by ":". If the key for the first
3671 two options is not specified, the assumed keys for the first two
3672 values are @code{w} and @code{h}. If the first option has no key and
3673 can be interpreted like a video size specification, it will be used
3674 to set the video size.
3676 A description of the accepted options follows.
3680 Set the video width expression, default value is @code{iw}. See below
3681 for the list of accepted constants.
3684 Set the video heiht expression, default value is @code{ih}.
3685 See below for the list of accepted constants.
3688 Set the interlacing. It accepts the following values:
3692 force interlaced aware scaling
3695 do not apply interlaced scaling
3698 select interlaced aware scaling depending on whether the source frames
3699 are flagged as interlaced or not
3702 Default value is @code{0}.
3705 Set libswscale scaling flags. If not explictly specified the filter
3706 applies a bilinear scaling algorithm.
3709 Set the video size, the value must be a valid abbreviation or in the
3710 form @var{width}x@var{height}.
3713 The values of the @var{w} and @var{h} options are expressions
3714 containing the following constants:
3718 the input width and height
3721 same as @var{in_w} and @var{in_h}
3724 the output (cropped) width and height
3727 same as @var{out_w} and @var{out_h}
3730 same as @var{iw} / @var{ih}
3733 input sample aspect ratio
3736 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
3739 horizontal and vertical chroma subsample values. For example for the
3740 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
3743 If the input image format is different from the format requested by
3744 the next filter, the scale filter will convert the input to the
3747 If the value for @var{width} or @var{height} is 0, the respective input
3748 size is used for the output.
3750 If the value for @var{width} or @var{height} is -1, the scale filter will
3751 use, for the respective output size, a value that maintains the aspect
3752 ratio of the input image.
3754 @subsection Examples
3758 Scale the input video to a size of 200x100:
3763 This is equivalent to:
3774 Specify a size abbreviation for the output size:
3779 which can also be written as:
3785 Scale the input to 2x:
3791 The above is the same as:
3797 Scale the input to 2x with forced interlaced scaling:
3799 scale=2*iw:2*ih:interl=1
3803 Scale the input to half size:
3809 Increase the width, and set the height to the same size:
3815 Seek for Greek harmony:
3822 Increase the height, and set the width to 3/2 of the height:
3828 Increase the size, but make the size a multiple of the chroma:
3830 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
3834 Increase the width to a maximum of 500 pixels, keep the same input
3837 scale='min(500\, iw*3/2):-1'
3841 @section setdar, setsar
3843 The @code{setdar} filter sets the Display Aspect Ratio for the filter
3846 This is done by changing the specified Sample (aka Pixel) Aspect
3847 Ratio, according to the following equation:
3849 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
3852 Keep in mind that the @code{setdar} filter does not modify the pixel
3853 dimensions of the video frame. Also the display aspect ratio set by
3854 this filter may be changed by later filters in the filterchain,
3855 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
3858 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
3859 the filter output video.
3861 Note that as a consequence of the application of this filter, the
3862 output display aspect ratio will change according to the equation
3865 Keep in mind that the sample aspect ratio set by the @code{setsar}
3866 filter may be changed by later filters in the filterchain, e.g. if
3867 another "setsar" or a "setdar" filter is applied.
3869 The @code{setdar} and @code{setsar} filters accept a string in the
3870 form @var{num}:@var{den} expressing an aspect ratio, or the following
3871 named options, expressed as a sequence of @var{key}=@var{value} pairs,
3876 Set the maximum integer value to use for expressing numerator and
3877 denominator when reducing the expressed aspect ratio to a rational.
3878 Default value is @code{100}.
3881 Set the aspect ratio used by the filter.
3883 The parameter can be a floating point number string, an expression, or
3884 a string of the form @var{num}:@var{den}, where @var{num} and
3885 @var{den} are the numerator and denominator of the aspect ratio. If
3886 the parameter is not specified, it is assumed the value "0".
3887 In case the form "@var{num}:@var{den}" the @code{:} character should
3891 If the keys are omitted in the named options list, the specifed values
3892 are assumed to be @var{ratio} and @var{max} in that order.
3894 For example to change the display aspect ratio to 16:9, specify:
3899 The example above is equivalent to:
3904 To change the sample aspect ratio to 10:11, specify:
3909 To set a display aspect ratio of 16:9, and specify a maximum integer value of
3910 1000 in the aspect ratio reduction, use the command:
3912 setdar=ratio='16:9':max=1000
3917 Force field for the output video frame.
3919 The @code{setfield} filter marks the interlace type field for the
3920 output frames. It does not change the input frame, but only sets the
3921 corresponding property, which affects how the frame is treated by
3922 following filters (e.g. @code{fieldorder} or @code{yadif}).
3924 This filter accepts a single option @option{mode}, which can be
3925 specified either by setting @code{mode=VALUE} or setting the value
3926 alone. Available values are:
3930 Keep the same field property.
3933 Mark the frame as bottom-field-first.
3936 Mark the frame as top-field-first.
3939 Mark the frame as progressive.
3944 Show a line containing various information for each input video frame.
3945 The input video is not modified.
3947 The shown line contains a sequence of key/value pairs of the form
3948 @var{key}:@var{value}.
3950 A description of each shown parameter follows:
3954 sequential number of the input frame, starting from 0
3957 Presentation TimeStamp of the input frame, expressed as a number of
3958 time base units. The time base unit depends on the filter input pad.
3961 Presentation TimeStamp of the input frame, expressed as a number of
3965 position of the frame in the input stream, -1 if this information in
3966 unavailable and/or meaningless (for example in case of synthetic video)
3972 sample aspect ratio of the input frame, expressed in the form
3976 size of the input frame, expressed in the form
3977 @var{width}x@var{height}
3980 interlaced mode ("P" for "progressive", "T" for top field first, "B"
3981 for bottom field first)
3984 1 if the frame is a key frame, 0 otherwise
3987 picture type of the input frame ("I" for an I-frame, "P" for a
3988 P-frame, "B" for a B-frame, "?" for unknown type).
3989 Check also the documentation of the @code{AVPictureType} enum and of
3990 the @code{av_get_picture_type_char} function defined in
3991 @file{libavutil/avutil.h}.
3994 Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame
3996 @item plane_checksum
3997 Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
3998 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]"
4003 Blur the input video without impacting the outlines.
4005 The filter accepts the following parameters:
4006 @var{luma_radius}:@var{luma_strength}:@var{luma_threshold}[:@var{chroma_radius}:@var{chroma_strength}:@var{chroma_threshold}]
4008 Parameters prefixed by @var{luma} indicate that they work on the
4009 luminance of the pixels whereas parameters prefixed by @var{chroma}
4010 refer to the chrominance of the pixels.
4012 If the chroma parameters are not set, the luma parameters are used for
4013 either the luminance and the chrominance of the pixels.
4015 @var{luma_radius} or @var{chroma_radius} must be a float number in the
4016 range [0.1,5.0] that specifies the variance of the gaussian filter
4017 used to blur the image (slower if larger).
4019 @var{luma_strength} or @var{chroma_strength} must be a float number in
4020 the range [-1.0,1.0] that configures the blurring. A value included in
4021 [0.0,1.0] will blur the image whereas a value included in [-1.0,0.0]
4022 will sharpen the image.
4024 @var{luma_threshold} or @var{chroma_threshold} must be an integer in
4025 the range [-30,30] that is used as a coefficient to determine whether
4026 a pixel should be blurred or not. A value of 0 will filter all the
4027 image, a value included in [0,30] will filter flat areas and a value
4028 included in [-30,0] will filter edges.
4033 Draw subtitles on top of input video using the libass library.
4035 To enable compilation of this filter you need to configure FFmpeg with
4036 @code{--enable-libass}. This filter also requires a build with libavcodec and
4037 libavformat to convert the passed subtitles file to ASS (Advanced Substation
4038 Alpha) subtitles format.
4040 This filter accepts the following named options, expressed as a
4041 sequence of @var{key}=@var{value} pairs, separated by ":".
4045 Set the filename of the subtitle file to read. It must be specified.
4048 Specify the size of the original video, the video for which the ASS file
4049 was composed. Due to a misdesign in ASS aspect ratio arithmetic, this is
4050 necessary to correctly scale the fonts if the aspect ratio has been changed.
4053 If the first key is not specified, it is assumed that the first value
4054 specifies the @option{filename}.
4056 For example, to render the file @file{sub.srt} on top of the input
4057 video, use the command:
4062 which is equivalent to:
4064 subtitles=filename=sub.srt
4069 Split input video into several identical outputs.
4071 The filter accepts a single parameter which specifies the number of outputs. If
4072 unspecified, it defaults to 2.
4076 ffmpeg -i INPUT -filter_complex split=5 OUTPUT
4078 will create 5 copies of the input video.
4082 [in] split [splitout1][splitout2];
4083 [splitout1] crop=100:100:0:0 [cropout];
4084 [splitout2] pad=200:200:100:100 [padout];
4087 will create two separate outputs from the same input, one cropped and
4092 Scale the input by 2x and smooth using the Super2xSaI (Scale and
4093 Interpolate) pixel art scaling algorithm.
4095 Useful for enlarging pixel art images without reducing sharpness.
4101 Select the most representative frame in a given sequence of consecutive frames.
4103 It accepts as argument the frames batch size to analyze (default @var{N}=100);
4104 in a set of @var{N} frames, the filter will pick one of them, and then handle
4105 the next batch of @var{N} frames until the end.
4107 Since the filter keeps track of the whole frames sequence, a bigger @var{N}
4108 value will result in a higher memory usage, so a high value is not recommended.
4110 The following example extract one picture each 50 frames:
4115 Complete example of a thumbnail creation with @command{ffmpeg}:
4117 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
4122 Tile several successive frames together.
4124 It accepts a list of options in the form of @var{key}=@var{value} pairs
4125 separated by ":". A description of the accepted options follows.
4130 Set the grid size (i.e. the number of lines and columns) in the form
4134 Set the outer border margin in pixels.
4137 Set the inner border thickness (i.e. the number of pixels between frames). For
4138 more advanced padding options (such as having different values for the edges),
4139 refer to the pad video filter.
4142 Set the maximum number of frames to render in the given area. It must be less
4143 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
4144 the area will be used.
4148 Alternatively, the options can be specified as a flat string:
4150 @var{layout}[:@var{nb_frames}[:@var{margin}[:@var{padding}]]]
4152 For example, produce 8×8 PNG tiles of all keyframes (@option{-skip_frame
4155 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
4157 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
4158 duplicating each output frame to accomodate the originally detected frame
4161 Another example to display @code{5} pictures in an area of @code{3x2} frames,
4162 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
4163 mixed flat and named options:
4165 tile=3x2:nb_frames=5:padding=7:margin=2
4170 Perform various types of temporal field interlacing.
4172 Frames are counted starting from 1, so the first input frame is
4175 This filter accepts options in the form of @var{key}=@var{value} pairs
4177 Alternatively, the @var{mode} option can be specified as a value alone,
4178 optionally followed by a ":" and further ":" separated @var{key}=@var{value}
4181 A description of the accepted options follows.
4186 Specify the mode of the interlacing. This option can also be specified
4187 as a value alone. See below for a list of values for this option.
4189 Available values are:
4193 Move odd frames into the upper field, even into the lower field,
4194 generating a double height frame at half framerate.
4197 Only output even frames, odd frames are dropped, generating a frame with
4198 unchanged height at half framerate.
4201 Only output odd frames, even frames are dropped, generating a frame with
4202 unchanged height at half framerate.
4205 Expand each frame to full height, but pad alternate lines with black,
4206 generating a frame with double height at the same input framerate.
4208 @item interleave_top, 4
4209 Interleave the upper field from odd frames with the lower field from
4210 even frames, generating a frame with unchanged height at half framerate.
4212 @item interleave_bottom, 5
4213 Interleave the lower field from odd frames with the upper field from
4214 even frames, generating a frame with unchanged height at half framerate.
4216 @item interlacex2, 6
4217 Double frame rate with unchanged height. Frames are inserted each
4218 containing the second temporal field from the previous input frame and
4219 the first temporal field from the next input frame. This mode relies on
4220 the top_field_first flag. Useful for interlaced video displays with no
4221 field synchronisation.
4224 Numeric values are deprecated but are accepted for backward
4225 compatibility reasons.
4227 Default mode is @code{merge}.
4230 Specify flags influencing the filter process.
4232 Available value for @var{flags} is:
4235 @item low_pass_filter, vlfp
4236 Enable vertical low-pass filtering in the filter.
4237 Vertical low-pass filtering is required when creating an interlaced
4238 destination from a progressive source which contains high-frequency
4239 vertical detail. Filtering will reduce interlace 'twitter' and Moire
4242 Vertical low-pass filtering can only be enabled for @option{mode}
4243 @var{interleave_top} and @var{interleave_bottom}.
4250 Transpose rows with columns in the input video and optionally flip it.
4252 The filter accepts parameters as a list of @var{key}=@var{value}
4253 pairs, separated by ':'. If the key of the first options is omitted,
4254 the arguments are interpreted according to the syntax
4255 @var{dir}:@var{passthrough}.
4259 Specify the transposition direction. Can assume the following values:
4263 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
4271 Rotate by 90 degrees clockwise, that is:
4279 Rotate by 90 degrees counterclockwise, that is:
4287 Rotate by 90 degrees clockwise and vertically flip, that is:
4295 For values between 4-7, the transposition is only done if the input
4296 video geometry is portrait and not landscape. These values are
4297 deprecated, the @code{passthrough} option should be used instead.
4300 Do not apply the transposition if the input geometry matches the one
4301 specified by the specified value. It accepts the following values:
4304 Always apply transposition.
4306 Preserve portrait geometry (when @var{height} >= @var{width}).
4308 Preserve landscape geometry (when @var{width} >= @var{height}).
4311 Default value is @code{none}.
4314 For example to rotate by 90 degrees clockwise and preserve portrait
4317 transpose=dir=1:passthrough=portrait
4320 The command above can also be specified as:
4322 transpose=1:portrait
4327 Sharpen or blur the input video.
4329 It accepts the following parameters:
4330 @var{luma_msize_x}:@var{luma_msize_y}:@var{luma_amount}:@var{chroma_msize_x}:@var{chroma_msize_y}:@var{chroma_amount}
4332 Negative values for the amount will blur the input video, while positive
4333 values will sharpen. All parameters are optional and default to the
4334 equivalent of the string '5:5:1.0:5:5:0.0'.
4339 Set the luma matrix horizontal size. It can be an integer between 3
4340 and 13, default value is 5.
4343 Set the luma matrix vertical size. It can be an integer between 3
4344 and 13, default value is 5.
4347 Set the luma effect strength. It can be a float number between -2.0
4348 and 5.0, default value is 1.0.
4350 @item chroma_msize_x
4351 Set the chroma matrix horizontal size. It can be an integer between 3
4352 and 13, default value is 5.
4354 @item chroma_msize_y
4355 Set the chroma matrix vertical size. It can be an integer between 3
4356 and 13, default value is 5.
4359 Set the chroma effect strength. It can be a float number between -2.0
4360 and 5.0, default value is 0.0.
4365 # Strong luma sharpen effect parameters
4368 # Strong blur of both luma and chroma parameters
4369 unsharp=7:7:-2:7:7:-2
4371 # Use the default values with @command{ffmpeg}
4372 ffmpeg -i in.avi -vf "unsharp" out.mp4
4377 Flip the input video vertically.
4380 ffmpeg -i in.avi -vf "vflip" out.avi
4385 Deinterlace the input video ("yadif" means "yet another deinterlacing
4388 The filter accepts parameters as a list of @var{key}=@var{value}
4389 pairs, separated by ":". If the key of the first options is omitted,
4390 the arguments are interpreted according to syntax
4391 @var{mode}:@var{parity}:@var{deint}.
4393 The description of the accepted parameters follows.
4397 Specify the interlacing mode to adopt. Accept one of the following
4402 output 1 frame for each frame
4404 output 1 frame for each field
4405 @item 2, send_frame_nospatial
4406 like @code{send_frame} but skip spatial interlacing check
4407 @item 3, send_field_nospatial
4408 like @code{send_field} but skip spatial interlacing check
4411 Default value is @code{send_frame}.
4414 Specify the picture field parity assumed for the input interlaced
4415 video. Accept one of the following values:
4419 assume top field first
4421 assume bottom field first
4423 enable automatic detection
4426 Default value is @code{auto}.
4427 If interlacing is unknown or decoder does not export this information,
4428 top field first will be assumed.
4431 Specify which frames to deinterlace. Accept one of the following
4436 deinterlace all frames
4438 only deinterlace frames marked as interlaced
4441 Default value is @code{all}.
4444 @c man end VIDEO FILTERS
4446 @chapter Video Sources
4447 @c man begin VIDEO SOURCES
4449 Below is a description of the currently available video sources.
4453 Buffer video frames, and make them available to the filter chain.
4455 This source is mainly intended for a programmatic use, in particular
4456 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
4458 It accepts a list of options in the form of @var{key}=@var{value} pairs
4459 separated by ":". A description of the accepted options follows.
4464 Specify the size (width and height) of the buffered video frames.
4467 A string representing the pixel format of the buffered video frames.
4468 It may be a number corresponding to a pixel format, or a pixel format
4472 Specify the timebase assumed by the timestamps of the buffered frames.
4475 Specify the frame rate expected for the video stream.
4478 Specify the sample aspect ratio assumed by the video frames.
4481 Specify the optional parameters to be used for the scale filter which
4482 is automatically inserted when an input change is detected in the
4483 input size or format.
4488 buffer=size=320x240:pix_fmt=yuv410p:time_base=1/24:pixel_aspect=1/1
4491 will instruct the source to accept video frames with size 320x240 and
4492 with format "yuv410p", assuming 1/24 as the timestamps timebase and
4493 square pixels (1:1 sample aspect ratio).
4494 Since the pixel format with name "yuv410p" corresponds to the number 6
4495 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
4496 this example corresponds to:
4498 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
4501 Alternatively, the options can be specified as a flat string, but this
4502 syntax is deprecated:
4504 @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}]
4508 Create a pattern generated by an elementary cellular automaton.
4510 The initial state of the cellular automaton can be defined through the
4511 @option{filename}, and @option{pattern} options. If such options are
4512 not specified an initial state is created randomly.
4514 At each new frame a new row in the video is filled with the result of
4515 the cellular automaton next generation. The behavior when the whole
4516 frame is filled is defined by the @option{scroll} option.
4518 This source accepts a list of options in the form of
4519 @var{key}=@var{value} pairs separated by ":". A description of the
4520 accepted options follows.
4524 Read the initial cellular automaton state, i.e. the starting row, from
4526 In the file, each non-whitespace character is considered an alive
4527 cell, a newline will terminate the row, and further characters in the
4528 file will be ignored.
4531 Read the initial cellular automaton state, i.e. the starting row, from
4532 the specified string.
4534 Each non-whitespace character in the string is considered an alive
4535 cell, a newline will terminate the row, and further characters in the
4536 string will be ignored.
4539 Set the video rate, that is the number of frames generated per second.
4542 @item random_fill_ratio, ratio
4543 Set the random fill ratio for the initial cellular automaton row. It
4544 is a floating point number value ranging from 0 to 1, defaults to
4547 This option is ignored when a file or a pattern is specified.
4549 @item random_seed, seed
4550 Set the seed for filling randomly the initial row, must be an integer
4551 included between 0 and UINT32_MAX. If not specified, or if explicitly
4552 set to -1, the filter will try to use a good random seed on a best
4556 Set the cellular automaton rule, it is a number ranging from 0 to 255.
4557 Default value is 110.
4560 Set the size of the output video.
4562 If @option{filename} or @option{pattern} is specified, the size is set
4563 by default to the width of the specified initial state row, and the
4564 height is set to @var{width} * PHI.
4566 If @option{size} is set, it must contain the width of the specified
4567 pattern string, and the specified pattern will be centered in the
4570 If a filename or a pattern string is not specified, the size value
4571 defaults to "320x518" (used for a randomly generated initial state).
4574 If set to 1, scroll the output upward when all the rows in the output
4575 have been already filled. If set to 0, the new generated row will be
4576 written over the top row just after the bottom row is filled.
4579 @item start_full, full
4580 If set to 1, completely fill the output with generated rows before
4581 outputting the first frame.
4582 This is the default behavior, for disabling set the value to 0.
4585 If set to 1, stitch the left and right row edges together.
4586 This is the default behavior, for disabling set the value to 0.
4589 @subsection Examples
4593 Read the initial state from @file{pattern}, and specify an output of
4596 cellauto=f=pattern:s=200x400
4600 Generate a random initial row with a width of 200 cells, with a fill
4603 cellauto=ratio=2/3:s=200x200
4607 Create a pattern generated by rule 18 starting by a single alive cell
4608 centered on an initial row with width 100:
4610 cellauto=p=@@:s=100x400:full=0:rule=18
4614 Specify a more elaborated initial pattern:
4616 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
4623 Generate a Mandelbrot set fractal, and progressively zoom towards the
4624 point specified with @var{start_x} and @var{start_y}.
4626 This source accepts a list of options in the form of
4627 @var{key}=@var{value} pairs separated by ":". A description of the
4628 accepted options follows.
4633 Set the terminal pts value. Default value is 400.
4636 Set the terminal scale value.
4637 Must be a floating point value. Default value is 0.3.
4640 Set the inner coloring mode, that is the algorithm used to draw the
4641 Mandelbrot fractal internal region.
4643 It shall assume one of the following values:
4648 Show time until convergence.
4650 Set color based on point closest to the origin of the iterations.
4655 Default value is @var{mincol}.
4658 Set the bailout value. Default value is 10.0.
4661 Set the maximum of iterations performed by the rendering
4662 algorithm. Default value is 7189.
4665 Set outer coloring mode.
4666 It shall assume one of following values:
4668 @item iteration_count
4669 Set iteration cound mode.
4670 @item normalized_iteration_count
4671 set normalized iteration count mode.
4673 Default value is @var{normalized_iteration_count}.
4676 Set frame rate, expressed as number of frames per second. Default
4680 Set frame size. Default value is "640x480".
4683 Set the initial scale value. Default value is 3.0.
4686 Set the initial x position. Must be a floating point value between
4687 -100 and 100. Default value is -0.743643887037158704752191506114774.
4690 Set the initial y position. Must be a floating point value between
4691 -100 and 100. Default value is -0.131825904205311970493132056385139.
4696 Generate various test patterns, as generated by the MPlayer test filter.
4698 The size of the generated video is fixed, and is 256x256.
4699 This source is useful in particular for testing encoding features.
4701 This source accepts an optional sequence of @var{key}=@var{value} pairs,
4702 separated by ":". The description of the accepted options follows.
4707 Specify the frame rate of the sourced video, as the number of frames
4708 generated per second. It has to be a string in the format
4709 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
4710 number or a valid video frame rate abbreviation. The default value is
4714 Set the video duration of the sourced video. The accepted syntax is:
4719 See also the function @code{av_parse_time()}.
4721 If not specified, or the expressed duration is negative, the video is
4722 supposed to be generated forever.
4726 Set the number or the name of the test to perform. Supported tests are:
4741 Default value is "all", which will cycle through the list of all tests.
4744 For example the following:
4749 will generate a "dc_luma" test pattern.
4753 Provide a frei0r source.
4755 To enable compilation of this filter you need to install the frei0r
4756 header and configure FFmpeg with @code{--enable-frei0r}.
4758 The source supports the syntax:
4760 @var{size}:@var{rate}:@var{src_name}[@{=|:@}@var{param1}:@var{param2}:...:@var{paramN}]
4763 @var{size} is the size of the video to generate, may be a string of the
4764 form @var{width}x@var{height} or a frame size abbreviation.
4765 @var{rate} is the rate of the video to generate, may be a string of
4766 the form @var{num}/@var{den} or a frame rate abbreviation.
4767 @var{src_name} is the name to the frei0r source to load. For more
4768 information regarding frei0r and how to set the parameters read the
4769 section @ref{frei0r} in the description of the video filters.
4771 For example, to generate a frei0r partik0l source with size 200x200
4772 and frame rate 10 which is overlayed on the overlay filter main input:
4774 frei0r_src=200x200:10:partik0l=1234 [overlay]; [in][overlay] overlay
4779 Generate a life pattern.
4781 This source is based on a generalization of John Conway's life game.
4783 The sourced input represents a life grid, each pixel represents a cell
4784 which can be in one of two possible states, alive or dead. Every cell
4785 interacts with its eight neighbours, which are the cells that are
4786 horizontally, vertically, or diagonally adjacent.
4788 At each interaction the grid evolves according to the adopted rule,
4789 which specifies the number of neighbor alive cells which will make a
4790 cell stay alive or born. The @option{rule} option allows to specify
4793 This source accepts a list of options in the form of
4794 @var{key}=@var{value} pairs separated by ":". A description of the
4795 accepted options follows.
4799 Set the file from which to read the initial grid state. In the file,
4800 each non-whitespace character is considered an alive cell, and newline
4801 is used to delimit the end of each row.
4803 If this option is not specified, the initial grid is generated
4807 Set the video rate, that is the number of frames generated per second.
4810 @item random_fill_ratio, ratio
4811 Set the random fill ratio for the initial random grid. It is a
4812 floating point number value ranging from 0 to 1, defaults to 1/PHI.
4813 It is ignored when a file is specified.
4815 @item random_seed, seed
4816 Set the seed for filling the initial random grid, must be an integer
4817 included between 0 and UINT32_MAX. If not specified, or if explicitly
4818 set to -1, the filter will try to use a good random seed on a best
4824 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
4825 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
4826 @var{NS} specifies the number of alive neighbor cells which make a
4827 live cell stay alive, and @var{NB} the number of alive neighbor cells
4828 which make a dead cell to become alive (i.e. to "born").
4829 "s" and "b" can be used in place of "S" and "B", respectively.
4831 Alternatively a rule can be specified by an 18-bits integer. The 9
4832 high order bits are used to encode the next cell state if it is alive
4833 for each number of neighbor alive cells, the low order bits specify
4834 the rule for "borning" new cells. Higher order bits encode for an
4835 higher number of neighbor cells.
4836 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
4837 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
4839 Default value is "S23/B3", which is the original Conway's game of life
4840 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
4841 cells, and will born a new cell if there are three alive cells around
4845 Set the size of the output video.
4847 If @option{filename} is specified, the size is set by default to the
4848 same size of the input file. If @option{size} is set, it must contain
4849 the size specified in the input file, and the initial grid defined in
4850 that file is centered in the larger resulting area.
4852 If a filename is not specified, the size value defaults to "320x240"
4853 (used for a randomly generated initial grid).
4856 If set to 1, stitch the left and right grid edges together, and the
4857 top and bottom edges also. Defaults to 1.
4860 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
4861 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
4862 value from 0 to 255.
4865 Set the color of living (or new born) cells.
4868 Set the color of dead cells. If @option{mold} is set, this is the first color
4869 used to represent a dead cell.
4872 Set mold color, for definitely dead and moldy cells.
4875 @subsection Examples
4879 Read a grid from @file{pattern}, and center it on a grid of size
4882 life=f=pattern:s=300x300
4886 Generate a random grid of size 200x200, with a fill ratio of 2/3:
4888 life=ratio=2/3:s=200x200
4892 Specify a custom rule for evolving a randomly generated grid:
4898 Full example with slow death effect (mold) using @command{ffplay}:
4900 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
4904 @section color, nullsrc, rgbtestsrc, smptebars, testsrc
4906 The @code{color} source provides an uniformly colored input.
4908 The @code{nullsrc} source returns unprocessed video frames. It is
4909 mainly useful to be employed in analysis / debugging tools, or as the
4910 source for filters which ignore the input data.
4912 The @code{rgbtestsrc} source generates an RGB test pattern useful for
4913 detecting RGB vs BGR issues. You should see a red, green and blue
4914 stripe from top to bottom.
4916 The @code{smptebars} source generates a color bars pattern, based on
4917 the SMPTE Engineering Guideline EG 1-1990.
4919 The @code{testsrc} source generates a test video pattern, showing a
4920 color pattern, a scrolling gradient and a timestamp. This is mainly
4921 intended for testing purposes.
4923 These sources accept an optional sequence of @var{key}=@var{value} pairs,
4924 separated by ":". The description of the accepted options follows.
4929 Specify the color of the source, only used in the @code{color}
4930 source. It can be the name of a color (case insensitive match) or a
4931 0xRRGGBB[AA] sequence, possibly followed by an alpha specifier. The
4932 default value is "black".
4935 Specify the size of the sourced video, it may be a string of the form
4936 @var{width}x@var{height}, or the name of a size abbreviation. The
4937 default value is "320x240".
4940 Specify the frame rate of the sourced video, as the number of frames
4941 generated per second. It has to be a string in the format
4942 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
4943 number or a valid video frame rate abbreviation. The default value is
4947 Set the sample aspect ratio of the sourced video.
4950 Set the video duration of the sourced video. The accepted syntax is:
4952 [-]HH[:MM[:SS[.m...]]]
4955 See also the function @code{av_parse_time()}.
4957 If not specified, or the expressed duration is negative, the video is
4958 supposed to be generated forever.
4961 Set the number of decimals to show in the timestamp, only used in the
4962 @code{testsrc} source.
4964 The displayed timestamp value will correspond to the original
4965 timestamp value multiplied by the power of 10 of the specified
4966 value. Default value is 0.
4969 For example the following:
4971 testsrc=duration=5.3:size=qcif:rate=10
4974 will generate a video with a duration of 5.3 seconds, with size
4975 176x144 and a frame rate of 10 frames per second.
4977 The following graph description will generate a red source
4978 with an opacity of 0.2, with size "qcif" and a frame rate of 10
4981 color=c=red@@0.2:s=qcif:r=10
4984 If the input content is to be ignored, @code{nullsrc} can be used. The
4985 following command generates noise in the luminance plane by employing
4986 the @code{geq} filter:
4988 nullsrc=s=256x256, geq=random(1)*255:128:128
4991 @c man end VIDEO SOURCES
4993 @chapter Video Sinks
4994 @c man begin VIDEO SINKS
4996 Below is a description of the currently available video sinks.
5000 Buffer video frames, and make them available to the end of the filter
5003 This sink is mainly intended for a programmatic use, in particular
5004 through the interface defined in @file{libavfilter/buffersink.h}.
5006 It does not require a string parameter in input, but you need to
5007 specify a pointer to a list of supported pixel formats terminated by
5008 -1 in the opaque parameter provided to @code{avfilter_init_filter}
5009 when initializing this sink.
5013 Null video sink, do absolutely nothing with the input video. It is
5014 mainly useful as a template and to be employed in analysis / debugging
5017 @c man end VIDEO SINKS
5019 @chapter Multimedia Filters
5020 @c man begin MULTIMEDIA FILTERS
5022 Below is a description of the currently available multimedia filters.
5024 @section aselect, select
5025 Select frames to pass in output.
5027 These filters accept a single option @option{expr} or @option{e}
5028 specifying the select expression, which can be specified either by
5029 specyfing @code{expr=VALUE} or specifying the expression
5032 The select expression is evaluated for each input frame. If the
5033 evaluation result is a non-zero value, the frame is selected and
5034 passed to the output, otherwise it is discarded.
5036 The expression can contain the following constants:
5040 the sequential number of the filtered frame, starting from 0
5043 the sequential number of the selected frame, starting from 0
5045 @item prev_selected_n
5046 the sequential number of the last selected frame, NAN if undefined
5049 timebase of the input timestamps
5052 the PTS (Presentation TimeStamp) of the filtered video frame,
5053 expressed in @var{TB} units, NAN if undefined
5056 the PTS (Presentation TimeStamp) of the filtered video frame,
5057 expressed in seconds, NAN if undefined
5060 the PTS of the previously filtered video frame, NAN if undefined
5062 @item prev_selected_pts
5063 the PTS of the last previously filtered video frame, NAN if undefined
5065 @item prev_selected_t
5066 the PTS of the last previously selected video frame, NAN if undefined
5069 the PTS of the first video frame in the video, NAN if undefined
5072 the time of the first video frame in the video, NAN if undefined
5074 @item pict_type @emph{(video only)}
5075 the type of the filtered frame, can assume one of the following
5087 @item interlace_type @emph{(video only)}
5088 the frame interlace type, can assume one of the following values:
5091 the frame is progressive (not interlaced)
5093 the frame is top-field-first
5095 the frame is bottom-field-first
5098 @item consumed_sample_n @emph{(audio only)}
5099 the number of selected samples before the current frame
5101 @item samples_n @emph{(audio only)}
5102 the number of samples in the current frame
5104 @item sample_rate @emph{(audio only)}
5105 the input sample rate
5108 1 if the filtered frame is a key-frame, 0 otherwise
5111 the position in the file of the filtered frame, -1 if the information
5112 is not available (e.g. for synthetic video)
5114 @item scene @emph{(video only)}
5115 value between 0 and 1 to indicate a new scene; a low value reflects a low
5116 probability for the current frame to introduce a new scene, while a higher
5117 value means the current frame is more likely to be one (see the example below)
5121 The default value of the select expression is "1".
5123 @subsection Examples
5127 Select all frames in input:
5132 The example above is the same as:
5144 Select only I-frames:
5146 select='eq(pict_type\,I)'
5150 Select one frame every 100:
5152 select='not(mod(n\,100))'
5156 Select only frames contained in the 10-20 time interval:
5158 select='gte(t\,10)*lte(t\,20)'
5162 Select only I frames contained in the 10-20 time interval:
5164 select='gte(t\,10)*lte(t\,20)*eq(pict_type\,I)'
5168 Select frames with a minimum distance of 10 seconds:
5170 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
5174 Use aselect to select only audio frames with samples number > 100:
5176 aselect='gt(samples_n\,100)'
5180 Create a mosaic of the first scenes:
5182 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
5185 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
5189 @section asendcmd, sendcmd
5191 Send commands to filters in the filtergraph.
5193 These filters read commands to be sent to other filters in the
5196 @code{asendcmd} must be inserted between two audio filters,
5197 @code{sendcmd} must be inserted between two video filters, but apart
5198 from that they act the same way.
5200 The specification of commands can be provided in the filter arguments
5201 with the @var{commands} option, or in a file specified by the
5202 @var{filename} option.
5204 These filters accept the following options:
5207 Set the commands to be read and sent to the other filters.
5209 Set the filename of the commands to be read and sent to the other
5213 @subsection Commands syntax
5215 A commands description consists of a sequence of interval
5216 specifications, comprising a list of commands to be executed when a
5217 particular event related to that interval occurs. The occurring event
5218 is typically the current frame time entering or leaving a given time
5221 An interval is specified by the following syntax:
5223 @var{START}[-@var{END}] @var{COMMANDS};
5226 The time interval is specified by the @var{START} and @var{END} times.
5227 @var{END} is optional and defaults to the maximum time.
5229 The current frame time is considered within the specified interval if
5230 it is included in the interval [@var{START}, @var{END}), that is when
5231 the time is greater or equal to @var{START} and is lesser than
5234 @var{COMMANDS} consists of a sequence of one or more command
5235 specifications, separated by ",", relating to that interval. The
5236 syntax of a command specification is given by:
5238 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
5241 @var{FLAGS} is optional and specifies the type of events relating to
5242 the time interval which enable sending the specified command, and must
5243 be a non-null sequence of identifier flags separated by "+" or "|" and
5244 enclosed between "[" and "]".
5246 The following flags are recognized:
5249 The command is sent when the current frame timestamp enters the
5250 specified interval. In other words, the command is sent when the
5251 previous frame timestamp was not in the given interval, and the
5255 The command is sent when the current frame timestamp leaves the
5256 specified interval. In other words, the command is sent when the
5257 previous frame timestamp was in the given interval, and the
5261 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
5264 @var{TARGET} specifies the target of the command, usually the name of
5265 the filter class or a specific filter instance name.
5267 @var{COMMAND} specifies the name of the command for the target filter.
5269 @var{ARG} is optional and specifies the optional list of argument for
5270 the given @var{COMMAND}.
5272 Between one interval specification and another, whitespaces, or
5273 sequences of characters starting with @code{#} until the end of line,
5274 are ignored and can be used to annotate comments.
5276 A simplified BNF description of the commands specification syntax
5279 @var{COMMAND_FLAG} ::= "enter" | "leave"
5280 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
5281 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
5282 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
5283 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
5284 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
5287 @subsection Examples
5291 Specify audio tempo change at second 4:
5293 asendcmd=c='4.0 atempo tempo 1.5',atempo
5297 Specify a list of drawtext and hue commands in a file.
5299 # show text in the interval 5-10
5300 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
5301 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
5303 # desaturate the image in the interval 15-20
5304 15.0-20.0 [enter] hue reinit s=0,
5305 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
5306 [leave] hue reinit s=1,
5307 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
5309 # apply an exponential saturation fade-out effect, starting from time 25
5310 25 [enter] hue s=exp(t-25)
5313 A filtergraph allowing to read and process the above command list
5314 stored in a file @file{test.cmd}, can be specified with:
5316 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
5321 @section asetpts, setpts
5323 Change the PTS (presentation timestamp) of the input frames.
5325 @code{asetpts} works on audio frames, @code{setpts} on video frames.
5327 Accept in input an expression evaluated through the eval API, which
5328 can contain the following constants:
5332 frame rate, only defined for constant frame-rate video
5335 the presentation timestamp in input
5338 the count of the input frame, starting from 0.
5340 @item NB_CONSUMED_SAMPLES
5341 the number of consumed samples, not including the current frame (only
5345 the number of samples in the current frame (only audio)
5351 the PTS of the first frame
5354 the time in seconds of the first frame
5357 tell if the current frame is interlaced
5360 the time in seconds of the current frame
5366 original position in the file of the frame, or undefined if undefined
5367 for the current frame
5373 previous input time in seconds
5379 previous output time in seconds
5382 @subsection Examples
5386 Start counting PTS from zero
5392 Apply fast motion effect:
5398 Apply slow motion effect:
5404 Set fixed rate of 25 frames per second:
5410 Set fixed rate 25 fps with some jitter:
5412 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
5416 Apply an offset of 10 seconds to the input PTS:
5424 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
5425 it unchanged. By default, it logs a message at a frequency of 10Hz with the
5426 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
5427 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
5429 The filter also has a video output (see the @var{video} option) with a real
5430 time graph to observe the loudness evolution. The graphic contains the logged
5431 message mentioned above, so it is not printed anymore when this option is set,
5432 unless the verbose logging is set. The main graphing area contains the
5433 short-term loudness (3 seconds of analysis), and the gauge on the right is for
5434 the momentary loudness (400 milliseconds).
5436 More information about the Loudness Recommendation EBU R128 on
5437 @url{http://tech.ebu.ch/loudness}.
5439 The filter accepts the following named parameters:
5444 Activate the video output. The audio stream is passed unchanged whether this
5445 option is set or no. The video stream will be the first output stream if
5446 activated. Default is @code{0}.
5449 Set the video size. This option is for video only. Default and minimum
5450 resolution is @code{640x480}.
5453 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
5454 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
5455 other integer value between this range is allowed.
5459 Example of real-time graph using @command{ffplay}, with a EBU scale meter +18:
5461 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
5464 Run an analysis with @command{ffmpeg}:
5466 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
5469 @section settb, asettb
5471 Set the timebase to use for the output frames timestamps.
5472 It is mainly useful for testing timebase configuration.
5474 It accepts in input an arithmetic expression representing a rational.
5475 The expression can contain the constants "AVTB" (the
5476 default timebase), "intb" (the input timebase) and "sr" (the sample rate,
5479 The default value for the input is "intb".
5481 @subsection Examples
5485 Set the timebase to 1/25:
5491 Set the timebase to 1/10:
5497 Set the timebase to 1001/1000:
5503 Set the timebase to 2*intb:
5509 Set the default timebase value:
5517 Concatenate audio and video streams, joining them together one after the
5520 The filter works on segments of synchronized video and audio streams. All
5521 segments must have the same number of streams of each type, and that will
5522 also be the number of streams at output.
5524 The filter accepts the following named parameters:
5528 Set the number of segments. Default is 2.
5531 Set the number of output video streams, that is also the number of video
5532 streams in each segment. Default is 1.
5535 Set the number of output audio streams, that is also the number of video
5536 streams in each segment. Default is 0.
5539 Activate unsafe mode: do not fail if segments have a different format.
5543 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
5544 @var{a} audio outputs.
5546 There are @var{n}×(@var{v}+@var{a}) inputs: first the inputs for the first
5547 segment, in the same order as the outputs, then the inputs for the second
5550 Related streams do not always have exactly the same duration, for various
5551 reasons including codec frame size or sloppy authoring. For that reason,
5552 related synchronized streams (e.g. a video and its audio track) should be
5553 concatenated at once. The concat filter will use the duration of the longest
5554 stream in each segment (except the last one), and if necessary pad shorter
5555 audio streams with silence.
5557 For this filter to work correctly, all segments must start at timestamp 0.
5559 All corresponding streams must have the same parameters in all segments; the
5560 filtering system will automatically select a common pixel format for video
5561 streams, and a common sample format, sample rate and channel layout for
5562 audio streams, but other settings, such as resolution, must be converted
5563 explicitly by the user.
5565 Different frame rates are acceptable but will result in variable frame rate
5566 at output; be sure to configure the output file to handle it.
5571 Concatenate an opening, an episode and an ending, all in bilingual version
5572 (video in stream 0, audio in streams 1 and 2):
5574 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
5575 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
5576 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
5577 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
5581 Concatenate two parts, handling audio and video separately, using the
5582 (a)movie sources, and adjusting the resolution:
5584 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
5585 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
5586 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
5588 Note that a desync will happen at the stitch if the audio and video streams
5589 do not have exactly the same duration in the first file.
5593 @section showspectrum
5595 Convert input audio to a video output, representing the audio frequency
5598 The filter accepts the following named parameters:
5601 Specify the video size for the output. Default value is @code{640x480}.
5603 Specify if the spectrum should slide along the window. Default value is
5607 The usage is very similar to the showwaves filter; see the examples in that
5612 Convert input audio to a video output, representing the samples waves.
5614 The filter accepts the following named parameters:
5618 Set the number of samples which are printed on the same column. A
5619 larger value will decrease the frame rate. Must be a positive
5620 integer. This option can be set only if the value for @var{rate}
5621 is not explicitly specified.
5624 Set the (approximate) output frame rate. This is done by setting the
5625 option @var{n}. Default value is "25".
5628 Specify the video size for the output. Default value is "600x240".
5631 Some examples follow.
5634 Output the input file audio and the corresponding video representation
5637 amovie=a.mp3,asplit[out0],showwaves[out1]
5641 Create a synthetic signal and show it with showwaves, forcing a
5642 framerate of 30 frames per second:
5644 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
5648 @c man end MULTIMEDIA FILTERS
5650 @chapter Multimedia Sources
5651 @c man begin MULTIMEDIA SOURCES
5653 Below is a description of the currently available multimedia sources.
5657 This is the same as @ref{movie} source, except it selects an audio
5663 Read audio and/or video stream(s) from a movie container.
5665 It accepts the syntax: @var{movie_name}[:@var{options}] where
5666 @var{movie_name} is the name of the resource to read (not necessarily
5667 a file but also a device or a stream accessed through some protocol),
5668 and @var{options} is an optional sequence of @var{key}=@var{value}
5669 pairs, separated by ":".
5671 The description of the accepted options follows.
5675 @item format_name, f
5676 Specifies the format assumed for the movie to read, and can be either
5677 the name of a container or an input device. If not specified the
5678 format is guessed from @var{movie_name} or by probing.
5680 @item seek_point, sp
5681 Specifies the seek point in seconds, the frames will be output
5682 starting from this seek point, the parameter is evaluated with
5683 @code{av_strtod} so the numerical value may be suffixed by an IS
5684 postfix. Default value is "0".
5687 Specifies the streams to read. Several streams can be specified,
5688 separated by "+". The source will then have as many outputs, in the
5689 same order. The syntax is explained in the ``Stream specifiers''
5690 section in the ffmpeg manual. Two special names, "dv" and "da" specify
5691 respectively the default (best suited) video and audio stream. Default
5692 is "dv", or "da" if the filter is called as "amovie".
5694 @item stream_index, si
5695 Specifies the index of the video stream to read. If the value is -1,
5696 the best suited video stream will be automatically selected. Default
5697 value is "-1". Deprecated. If the filter is called "amovie", it will select
5698 audio instead of video.
5701 Specifies how many times to read the stream in sequence.
5702 If the value is less than 1, the stream will be read again and again.
5703 Default value is "1".
5705 Note that when the movie is looped the source timestamps are not
5706 changed, so it will generate non monotonically increasing timestamps.
5709 This filter allows to overlay a second video on top of main input of
5710 a filtergraph as shown in this graph:
5712 input -----------> deltapts0 --> overlay --> output
5715 movie --> scale--> deltapts1 -------+
5718 Some examples follow.
5722 Skip 3.2 seconds from the start of the avi file in.avi, and overlay it
5723 on top of the input labelled as "in":
5725 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [movie];
5726 [in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
5730 Read from a video4linux2 device, and overlay it on top of the input
5733 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [movie];
5734 [in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
5738 Read the first video stream and the audio stream with id 0x81 from
5739 dvd.vob; the video is connected to the pad named "video" and the audio is
5740 connected to the pad named "audio":
5742 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
5746 @c man end MULTIMEDIA SOURCES