1 @chapter Filtergraph description
2 @c man begin FILTERGRAPH DESCRIPTION
4 A filtergraph is a directed graph of connected filters. It can contain
5 cycles, and there can be multiple links between a pair of
6 filters. Each link has one input pad on one side connecting it to one
7 filter from which it takes its input, and one output pad on the other
8 side connecting it to one filter accepting its output.
10 Each filter in a filtergraph is an instance of a filter class
11 registered in the application, which defines the features and the
12 number of input and output pads of the filter.
14 A filter with no input pads is called a "source", and a filter with no
15 output pads is called a "sink".
17 @anchor{Filtergraph syntax}
18 @section Filtergraph syntax
20 A filtergraph has a textual representation, which is
21 recognized by the @option{-filter}/@option{-vf} and @option{-filter_complex}
22 options in @command{avconv} and @option{-vf} in @command{avplay}, and by the
23 @code{avfilter_graph_parse()}/@code{avfilter_graph_parse2()} functions defined in
24 @file{libavfilter/avfilter.h}.
26 A filterchain consists of a sequence of connected filters, each one
27 connected to the previous one in the sequence. A filterchain is
28 represented by a list of ","-separated filter descriptions.
30 A filtergraph consists of a sequence of filterchains. A sequence of
31 filterchains is represented by a list of ";"-separated filterchain
34 A filter is represented by a string of the form:
35 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
37 @var{filter_name} is the name of the filter class of which the
38 described filter is an instance of, and has to be the name of one of
39 the filter classes registered in the program.
40 The name of the filter class is optionally followed by a string
43 @var{arguments} is a string which contains the parameters used to
44 initialize the filter instance. It may have one of two forms:
48 A ':'-separated list of @var{key=value} pairs.
51 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
52 the option names in the order they are declared. E.g. the @code{fade} filter
53 declares three options in this order -- @option{type}, @option{start_frame} and
54 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
55 @var{in} is assigned to the option @option{type}, @var{0} to
56 @option{start_frame} and @var{30} to @option{nb_frames}.
60 If the option value itself is a list of items (e.g. the @code{format} filter
61 takes a list of pixel formats), the items in the list are usually separated by
64 The list of arguments can be quoted using the character "'" as initial
65 and ending mark, and the character '\' for escaping the characters
66 within the quoted text; otherwise the argument string is considered
67 terminated when the next special character (belonging to the set
68 "[]=;,") is encountered.
70 The name and arguments of the filter are optionally preceded and
71 followed by a list of link labels.
72 A link label allows to name a link and associate it to a filter output
73 or input pad. The preceding labels @var{in_link_1}
74 ... @var{in_link_N}, are associated to the filter input pads,
75 the following labels @var{out_link_1} ... @var{out_link_M}, are
76 associated to the output pads.
78 When two link labels with the same name are found in the
79 filtergraph, a link between the corresponding input and output pad is
82 If an output pad is not labelled, it is linked by default to the first
83 unlabelled input pad of the next filter in the filterchain.
84 For example in the filterchain
86 nullsrc, split[L1], [L2]overlay, nullsink
88 the split filter instance has two output pads, and the overlay filter
89 instance two input pads. The first output pad of split is labelled
90 "L1", the first input pad of overlay is labelled "L2", and the second
91 output pad of split is linked to the second input pad of overlay,
92 which are both unlabelled.
94 In a complete filterchain all the unlabelled filter input and output
95 pads must be connected. A filtergraph is considered valid if all the
96 filter input and output pads of all the filterchains are connected.
98 Libavfilter will automatically insert @ref{scale} filters where format
99 conversion is required. It is possible to specify swscale flags
100 for those automatically inserted scalers by prepending
101 @code{sws_flags=@var{flags};}
102 to the filtergraph description.
104 Here is a BNF description of the filtergraph syntax:
106 @var{NAME} ::= sequence of alphanumeric characters and '_'
107 @var{LINKLABEL} ::= "[" @var{NAME} "]"
108 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
109 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
110 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
111 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
112 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
115 @c man end FILTERGRAPH DESCRIPTION
117 @chapter Audio Filters
118 @c man begin AUDIO FILTERS
120 When you configure your Libav build, you can disable any of the
121 existing filters using --disable-filters.
122 The configure output will show the audio filters included in your
125 Below is a description of the currently available audio filters.
129 Convert the input audio to one of the specified formats. The framework will
130 negotiate the most appropriate format to minimize conversions.
132 It accepts the following parameters:
136 A '|'-separated list of requested sample formats.
139 A '|'-separated list of requested sample rates.
141 @item channel_layouts
142 A '|'-separated list of requested channel layouts.
146 If a parameter is omitted, all values are allowed.
148 Force the output to either unsigned 8-bit or signed 16-bit stereo
150 aformat=sample_fmts=u8|s16:channel_layouts=stereo
155 Mixes multiple audio inputs into a single output.
159 avconv -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
161 will mix 3 input audio streams to a single output with the same duration as the
162 first input and a dropout transition time of 3 seconds.
164 It accepts the following parameters:
168 The number of inputs. If unspecified, it defaults to 2.
171 How to determine the end-of-stream.
175 The duration of the longest input. (default)
178 The duration of the shortest input.
181 The duration of the first input.
185 @item dropout_transition
186 The transition time, in seconds, for volume renormalization when an input
187 stream ends. The default value is 2 seconds.
193 Pass the audio source unchanged to the output.
197 Change the PTS (presentation timestamp) of the input audio frames.
199 It accepts the following parameters:
204 The expression which is evaluated for each frame to construct its timestamp.
208 The expression is evaluated through the eval API and can contain the following
213 the presentation timestamp in input
216 These are approximated values for the mathematical constants e
217 (Euler's number), pi (Greek pi), and phi (the golden ratio).
220 The number of audio samples passed through the filter so far, starting at 0.
223 The number of audio samples in the current frame.
226 The audio sample rate.
229 The PTS of the first frame.
232 The previous input PTS.
235 The previous output PTS.
238 The wallclock (RTC) time in microseconds.
241 The wallclock (RTC) time at the start of the movie in microseconds.
248 # Start counting PTS from zero
249 asetpts=expr=PTS-STARTPTS
251 # Generate timestamps by counting samples
254 # Generate timestamps from a "live source" and rebase onto the current timebase
255 asetpts='(RTCTIME - RTCSTART) / (TB * 1000000)"
260 Set the timebase to use for the output frames timestamps.
261 It is mainly useful for testing timebase configuration.
263 This filter accepts the following parameters:
268 The expression which is evaluated into the output timebase.
272 The expression can contain the constants @var{PI}, @var{E}, @var{PHI}, @var{AVTB} (the
273 default timebase), @var{intb} (the input timebase), and @var{sr} (the sample rate,
276 The default value for the input is @var{intb}.
281 # Set the timebase to 1/25:
284 # Set the timebase to 1/10:
287 # Set the timebase to 1001/1000:
290 # Set the timebase to 2*intb:
293 # Set the default timebase value:
296 # Set the timebase to twice the sample rate:
302 Show a line containing various information for each input audio frame.
303 The input audio is not modified.
305 The shown line contains a sequence of key/value pairs of the form
306 @var{key}:@var{value}.
308 It accepts the following parameters:
312 The (sequential) number of the input frame, starting from 0.
315 The presentation timestamp of the input frame, in time base units; the time base
316 depends on the filter input pad, and is usually 1/@var{sample_rate}.
319 The presentation timestamp of the input frame in seconds.
328 The sample rate for the audio frame.
331 The number of samples (per channel) in the frame.
334 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
335 audio, the data is treated as if all the planes were concatenated.
337 @item plane_checksums
338 A list of Adler-32 checksums for each data plane.
343 Split input audio into several identical outputs.
345 It accepts a single parameter, which specifies the number of outputs. If
346 unspecified, it defaults to 2.
350 avconv -i INPUT -filter_complex asplit=5 OUTPUT
352 will create 5 copies of the input audio.
355 Synchronize audio data with timestamps by squeezing/stretching it and/or
356 dropping samples/adding silence when needed.
358 It accepts the following parameters:
362 Enable stretching/squeezing the data to make it match the timestamps. Disabled
363 by default. When disabled, time gaps are covered with silence.
366 The minimum difference between timestamps and audio data (in seconds) to trigger
367 adding/dropping samples. The default value is 0.1. If you get an imperfect
368 sync with this filter, try setting this parameter to 0.
371 The maximum compensation in samples per second. Only relevant with compensate=1.
372 The default value is 500.
375 Assume that the first PTS should be this value. The time base is 1 / sample
376 rate. This allows for padding/trimming at the start of the stream. By default,
377 no assumption is made about the first frame's expected PTS, so no padding or
378 trimming is done. For example, this could be set to 0 to pad the beginning with
379 silence if an audio stream starts after the video stream or to trim any samples
380 with a negative PTS due to encoder delay.
385 Trim the input so that the output contains one continuous subpart of the input.
387 It accepts the following parameters:
390 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
391 sample with the timestamp @var{start} will be the first sample in the output.
394 Timestamp (in seconds) of the first audio sample that will be dropped. I.e. the
395 audio sample immediately preceding the one with the timestamp @var{end} will be
396 the last sample in the output.
399 Same as @var{start}, except this option sets the start timestamp in samples
403 Same as @var{end}, except this option sets the end timestamp in samples instead
407 The maximum duration of the output in seconds.
410 The number of the first sample that should be output.
413 The number of the first sample that should be dropped.
416 Note that the first two sets of the start/end options and the @option{duration}
417 option look at the frame timestamp, while the _sample options simply count the
418 samples that pass through the filter. So start/end_pts and start/end_sample will
419 give different results when the timestamps are wrong, inexact or do not start at
420 zero. Also note that this filter does not modify the timestamps. If you wish
421 to have the output timestamps start at zero, insert the asetpts filter after the
424 If multiple start or end options are set, this filter tries to be greedy and
425 keep all samples that match at least one of the specified constraints. To keep
426 only the part that matches all the constraints at once, chain multiple atrim
429 The defaults are such that all the input is kept. So it is possible to set e.g.
430 just the end values to keep everything before the specified time.
435 Drop everything except the second minute of input:
437 avconv -i INPUT -af atrim=60:120
441 Keep only the first 1000 samples:
443 avconv -i INPUT -af atrim=end_sample=1000
449 Bauer stereo to binaural transformation, which improves headphone listening of
450 stereo audio records.
452 It accepts the following parameters:
456 Pre-defined crossfeed level.
460 Default level (fcut=700, feed=50).
463 Chu Moy circuit (fcut=700, feed=60).
466 Jan Meier circuit (fcut=650, feed=95).
471 Cut frequency (in Hz).
478 @section channelsplit
479 Split each channel from an input audio stream into a separate output stream.
481 It accepts the following parameters:
484 The channel layout of the input stream. The default is "stereo".
487 For example, assuming a stereo input MP3 file,
489 avconv -i in.mp3 -filter_complex channelsplit out.mkv
491 will create an output Matroska file with two audio streams, one containing only
492 the left channel and the other the right channel.
494 Split a 5.1 WAV file into per-channel files:
496 avconv -i in.wav -filter_complex
497 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
498 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
499 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
504 Remap input channels to new locations.
506 It accepts the following parameters:
509 The channel layout of the output stream.
512 Map channels from input to output. The argument is a '|'-separated list of
513 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
514 @var{in_channel} form. @var{in_channel} can be either the name of the input
515 channel (e.g. FL for front left) or its index in the input channel layout.
516 @var{out_channel} is the name of the output channel or its index in the output
517 channel layout. If @var{out_channel} is not given then it is implicitly an
518 index, starting with zero and increasing by one for each mapping.
521 If no mapping is present, the filter will implicitly map input channels to
522 output channels, preserving indices.
524 For example, assuming a 5.1+downmix input MOV file,
526 avconv -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
528 will create an output WAV file tagged as stereo from the downmix channels of
531 To fix a 5.1 WAV improperly encoded in AAC's native channel order
533 avconv -i in.wav -filter 'channelmap=1|2|0|5|3|4:channel_layout=5.1' out.wav
537 Compress or expand the audio's dynamic range.
539 It accepts the following parameters:
545 A list of times in seconds for each channel over which the instantaneous level
546 of the input signal is averaged to determine its volume. @var{attacks} refers to
547 increase of volume and @var{decays} refers to decrease of volume. For most
548 situations, the attack time (response to the audio getting louder) should be
549 shorter than the decay time, because the human ear is more sensitive to sudden
550 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
551 a typical value for decay is 0.8 seconds.
554 A list of points for the transfer function, specified in dB relative to the
555 maximum possible signal amplitude. Each key points list must be defined using
556 the following syntax: @code{x0/y0|x1/y1|x2/y2|....}
558 The input values must be in strictly increasing order but the transfer function
559 does not have to be monotonically rising. The point @code{0/0} is assumed but
560 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
561 function are @code{-70/-70|-60/-20}.
564 Set the curve radius in dB for all joints. It defaults to 0.01.
567 Set the additional gain in dB to be applied at all points on the transfer
568 function. This allows for easy adjustment of the overall gain.
572 Set an initial volume, in dB, to be assumed for each channel when filtering
573 starts. This permits the user to supply a nominal level initially, so that, for
574 example, a very large gain is not applied to initial signal levels before the
575 companding has begun to operate. A typical value for audio which is initially
576 quiet is -90 dB. It defaults to 0.
579 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
580 delayed before being fed to the volume adjuster. Specifying a delay
581 approximately equal to the attack/decay times allows the filter to effectively
582 operate in predictive rather than reactive mode. It defaults to 0.
590 Make music with both quiet and loud passages suitable for listening to in a
593 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
597 A noise gate for when the noise is at a lower level than the signal:
599 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
603 Here is another noise gate, this time for when the noise is at a higher level
604 than the signal (making it, in some ways, similar to squelch):
606 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
611 Join multiple input streams into one multi-channel stream.
613 It accepts the following parameters:
617 The number of input streams. It defaults to 2.
620 The desired output channel layout. It defaults to stereo.
623 Map channels from inputs to output. The argument is a '|'-separated list of
624 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
625 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
626 can be either the name of the input channel (e.g. FL for front left) or its
627 index in the specified input stream. @var{out_channel} is the name of the output
631 The filter will attempt to guess the mappings when they are not specified
632 explicitly. It does so by first trying to find an unused matching input channel
633 and if that fails it picks the first unused input channel.
635 Join 3 inputs (with properly set channel layouts):
637 avconv -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
640 Build a 5.1 output from 6 single-channel streams:
642 avconv -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
643 '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'
648 Convert the audio sample format, sample rate and channel layout. It is
649 not meant to be used directly; it is inserted automatically by libavfilter
650 whenever conversion is needed. Use the @var{aformat} filter to force a specific
655 Adjust the input audio volume.
657 It accepts the following parameters:
661 This expresses how the audio volume will be increased or decreased.
663 Output values are clipped to the maximum value.
665 The output audio volume is given by the relation:
667 @var{output_volume} = @var{volume} * @var{input_volume}
670 The default value for @var{volume} is 1.0.
673 This parameter represents the mathematical precision.
675 It determines which input sample formats will be allowed, which affects the
676 precision of the volume scaling.
680 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
682 32-bit floating-point; this limits input sample format to FLT. (default)
684 64-bit floating-point; this limits input sample format to DBL.
688 Choose the behaviour on encountering ReplayGain side data in input frames.
692 Remove ReplayGain side data, ignoring its contents (the default).
695 Ignore ReplayGain side data, but leave it in the frame.
698 Prefer the track gain, if present.
701 Prefer the album gain, if present.
704 @item replaygain_preamp
705 Pre-amplification gain in dB to apply to the selected replaygain gain.
707 Default value for @var{replaygain_preamp} is 0.0.
709 @item replaygain_noclip
710 Prevent clipping by limiting the gain applied.
712 Default value for @var{replaygain_noclip} is 1.
720 Halve the input audio volume:
724 volume=volume=-6.0206dB
728 Increase input audio power by 6 decibels using fixed-point precision:
730 volume=volume=6dB:precision=fixed
734 @c man end AUDIO FILTERS
736 @chapter Audio Sources
737 @c man begin AUDIO SOURCES
739 Below is a description of the currently available audio sources.
743 The null audio source; it never returns audio frames. It is mainly useful as a
744 template and for use in analysis / debugging tools.
746 It accepts, as an optional parameter, a string of the form
747 @var{sample_rate}:@var{channel_layout}.
749 @var{sample_rate} specifies the sample rate, and defaults to 44100.
751 @var{channel_layout} specifies the channel layout, and can be either an
752 integer or a string representing a channel layout. The default value
753 of @var{channel_layout} is 3, which corresponds to CH_LAYOUT_STEREO.
755 Check the channel_layout_map definition in
756 @file{libavutil/channel_layout.c} for the mapping between strings and
757 channel layout values.
761 # Set the sample rate to 48000 Hz and the channel layout to CH_LAYOUT_MONO
769 Buffer audio frames, and make them available to the filter chain.
771 This source is not intended to be part of user-supplied graph descriptions; it
772 is for insertion by calling programs, through the interface defined in
773 @file{libavfilter/buffersrc.h}.
775 It accepts the following parameters:
779 The timebase which will be used for timestamps of submitted frames. It must be
780 either a floating-point number or in @var{numerator}/@var{denominator} form.
783 The audio sample rate.
786 The name of the sample format, as returned by @code{av_get_sample_fmt_name()}.
789 The channel layout of the audio data, in the form that can be accepted by
790 @code{av_get_channel_layout()}.
793 All the parameters need to be explicitly defined.
795 @c man end AUDIO SOURCES
798 @c man begin AUDIO SINKS
800 Below is a description of the currently available audio sinks.
804 Null audio sink; do absolutely nothing with the input audio. It is
805 mainly useful as a template and for use in analysis / debugging
809 This sink is intended for programmatic use. Frames that arrive on this sink can
810 be retrieved by the calling program, using the interface defined in
811 @file{libavfilter/buffersink.h}.
813 It does not accept any parameters.
815 @c man end AUDIO SINKS
817 @chapter Video Filters
818 @c man begin VIDEO FILTERS
820 When you configure your Libav build, you can disable any of the
821 existing filters using --disable-filters.
822 The configure output will show the video filters included in your
825 Below is a description of the currently available video filters.
829 Detect frames that are (almost) completely black. Can be useful to
830 detect chapter transitions or commercials. Output lines consist of
831 the frame number of the detected frame, the percentage of blackness,
832 the position in the file if known or -1 and the timestamp in seconds.
834 In order to display the output lines, you need to set the loglevel at
835 least to the AV_LOG_INFO value.
837 It accepts the following parameters:
842 The percentage of the pixels that have to be below the threshold; it defaults to
846 The threshold below which a pixel value is considered black; it defaults to 32.
852 Apply a boxblur algorithm to the input video.
854 It accepts the following parameters:
867 The chroma and alpha parameters are optional. If not specified, they default
868 to the corresponding values set for @var{luma_radius} and
871 @var{luma_radius}, @var{chroma_radius}, and @var{alpha_radius} represent
872 the radius in pixels of the box used for blurring the corresponding
873 input plane. They are expressions, and can contain the following
877 The input width and height in pixels.
880 The input chroma image width and height in pixels.
883 The horizontal and vertical chroma subsample values. For example, for the
884 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
887 The radius must be a non-negative number, and must not be greater than
888 the value of the expression @code{min(w,h)/2} for the luma and alpha planes,
889 and of @code{min(cw,ch)/2} for the chroma planes.
891 @var{luma_power}, @var{chroma_power}, and @var{alpha_power} represent
892 how many times the boxblur filter is applied to the corresponding
900 Apply a boxblur filter with the luma, chroma, and alpha radii
903 boxblur=luma_radius=2:luma_power=1
907 Set the luma radius to 2, and alpha and chroma radius to 0:
913 Set the luma and chroma radii to a fraction of the video dimension:
915 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
922 Copy the input source unchanged to the output. This is mainly useful for
927 Crop the input video to given dimensions.
929 It accepts the following parameters:
934 The width of the output video.
937 The height of the output video.
940 The horizontal position, in the input video, of the left edge of the output
944 The vertical position, in the input video, of the top edge of the output video.
948 The parameters are expressions containing the following constants:
952 These are approximated values for the mathematical constants e
953 (Euler's number), pi (Greek pi), and phi (the golden ratio).
956 The computed values for @var{x} and @var{y}. They are evaluated for
960 The input width and height.
963 These are the same as @var{in_w} and @var{in_h}.
966 The output (cropped) width and height.
969 These are the same as @var{out_w} and @var{out_h}.
972 The number of the input frame, starting from 0.
975 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
979 The @var{out_w} and @var{out_h} parameters specify the expressions for
980 the width and height of the output (cropped) video. They are only
981 evaluated during the configuration of the filter.
983 The default value of @var{out_w} is "in_w", and the default value of
984 @var{out_h} is "in_h".
986 The expression for @var{out_w} may depend on the value of @var{out_h},
987 and the expression for @var{out_h} may depend on @var{out_w}, but they
988 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
989 evaluated after @var{out_w} and @var{out_h}.
991 The @var{x} and @var{y} parameters specify the expressions for the
992 position of the top-left corner of the output (non-cropped) area. They
993 are evaluated for each frame. If the evaluated value is not valid, it
994 is approximated to the nearest valid value.
996 The default value of @var{x} is "(in_w-out_w)/2", and the default
997 value for @var{y} is "(in_h-out_h)/2", which set the cropped area at
998 the center of the input image.
1000 The expression for @var{x} may depend on @var{y}, and the expression
1001 for @var{y} may depend on @var{x}.
1005 # Crop the central input area with size 100x100
1006 crop=out_w=100:out_h=100
1008 # Crop the central input area with size 2/3 of the input video
1009 "crop=out_w=2/3*in_w:out_h=2/3*in_h"
1011 # Crop the input video central square
1014 # Delimit the rectangle with the top-left corner placed at position
1015 # 100:100 and the right-bottom corner corresponding to the right-bottom
1016 # corner of the input image
1017 crop=out_w=in_w-100:out_h=in_h-100:x=100:y=100
1019 # Crop 10 pixels from the left and right borders, and 20 pixels from
1020 # the top and bottom borders
1021 "crop=out_w=in_w-2*10:out_h=in_h-2*20"
1023 # Keep only the bottom right quarter of the input image
1024 "crop=out_w=in_w/2:out_h=in_h/2:x=in_w/2:y=in_h/2"
1026 # Crop height for getting Greek harmony
1027 "crop=out_w=in_w:out_h=1/PHI*in_w"
1030 "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)"
1032 # Erratic camera effect depending on timestamp
1033 "crop=out_w=in_w/2:out_h=in_h/2:x=(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):y=(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)"
1035 # Set x depending on the value of y
1036 "crop=in_w/2:in_h/2:y:10+10*sin(n/10)"
1041 Auto-detect the crop size.
1043 It calculates the necessary cropping parameters and prints the
1044 recommended parameters via the logging system. The detected dimensions
1045 correspond to the non-black area of the input video.
1047 It accepts the following parameters:
1052 The threshold, an optional parameter between nothing (0) and
1053 everything (255). It defaults to 24.
1056 The value which the width/height should be divisible by. It defaults to
1057 16. The offset is automatically adjusted to center the video. Use 2 to
1058 get only even dimensions (needed for 4:2:2 video). 16 is best when
1059 encoding to most video codecs.
1062 A counter that determines how many frames cropdetect will reset
1063 the previously detected largest video area after. It will then start over
1064 and detect the current optimal crop area. It defaults to 0.
1066 This can be useful when channel logos distort the video area. 0
1067 indicates 'never reset', and returns the largest area encountered during
1073 Suppress a TV station logo by a simple interpolation of the surrounding
1074 pixels. Just set a rectangle covering the logo and watch it disappear
1075 (and sometimes something even uglier appear - your mileage may vary).
1077 It accepts the following parameters:
1081 Specify the top left corner coordinates of the logo. They must be
1085 Specify the width and height of the logo to clear. They must be
1089 Specify the thickness of the fuzzy edge of the rectangle (added to
1090 @var{w} and @var{h}). The default value is 4.
1093 When set to 1, a green rectangle is drawn on the screen to simplify
1094 finding the right @var{x}, @var{y}, @var{w}, @var{h} parameters, and
1095 @var{band} is set to 4. The default value is 0.
1104 Set a rectangle covering the area with top left corner coordinates 0,0
1105 and size 100x77, and a band of size 10:
1107 delogo=x=0:y=0:w=100:h=77:band=10
1114 Draw a colored box on the input image.
1116 It accepts the following parameters:
1121 Specify the top left corner coordinates of the box. It defaults to 0.
1124 Specify the width and height of the box; if 0 they are interpreted as
1125 the input width and height. It defaults to 0.
1128 Specify the color of the box to write. It can be the name of a color
1129 (case insensitive match) or a 0xRRGGBB[AA] sequence.
1134 # Draw a black box around the edge of the input image
1137 # Draw a box with color red and an opacity of 50%
1138 drawbox=x=10:y=20:width=200:height=60:color=red@@0.5"
1143 Draw a text string or text from a specified file on top of a video, using the
1144 libfreetype library.
1146 To enable compilation of this filter, you need to configure Libav with
1147 @code{--enable-libfreetype}.
1148 To enable default font fallback and the @var{font} option you need to
1149 configure Libav with @code{--enable-libfontconfig}.
1151 The filter also recognizes strftime() sequences in the provided text
1152 and expands them accordingly. Check the documentation of strftime().
1154 It accepts the following parameters:
1159 The font family to be used for drawing text. By default Sans.
1162 The font file to be used for drawing text. The path must be included.
1163 This parameter is mandatory if the fontconfig support is disabled.
1166 The text string to be drawn. The text must be a sequence of UTF-8
1168 This parameter is mandatory if no file is specified with the parameter
1172 A text file containing text to be drawn. The text must be a sequence
1173 of UTF-8 encoded characters.
1175 This parameter is mandatory if no text string is specified with the
1176 parameter @var{text}.
1178 If both text and textfile are specified, an error is thrown.
1181 The offsets where text will be drawn within the video frame.
1182 It is relative to the top/left border of the output image.
1183 They accept expressions similar to the @ref{overlay} filter:
1187 The computed values for @var{x} and @var{y}. They are evaluated for
1190 @item main_w, main_h
1191 The main input width and height.
1194 These are the same as @var{main_w} and @var{main_h}.
1196 @item text_w, text_h
1197 The rendered text's width and height.
1200 These are the same as @var{text_w} and @var{text_h}.
1203 The number of frames processed, starting from 0.
1206 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
1210 The default value of @var{x} and @var{y} is 0.
1213 The font size to be used for drawing text.
1214 The default value of @var{fontsize} is 16.
1217 The color to be used for drawing fonts.
1218 It is either a string (e.g. "red"), or in 0xRRGGBB[AA] format
1219 (e.g. "0xff000033"), possibly followed by an alpha specifier.
1220 The default value of @var{fontcolor} is "black".
1223 The color to be used for drawing box around text.
1224 It is either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
1225 (e.g. "0xff00ff"), possibly followed by an alpha specifier.
1226 The default value of @var{boxcolor} is "white".
1229 Used to draw a box around text using the background color.
1230 The value must be either 1 (enable) or 0 (disable).
1231 The default value of @var{box} is 0.
1233 @item shadowx, shadowy
1234 The x and y offsets for the text shadow position with respect to the
1235 position of the text. They can be either positive or negative
1236 values. The default value for both is "0".
1239 The color to be used for drawing a shadow behind the drawn text. It
1240 can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
1241 form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
1242 The default value of @var{shadowcolor} is "black".
1245 The flags to be used for loading the fonts.
1247 The flags map the corresponding flags supported by libfreetype, and are
1248 a combination of the following values:
1255 @item vertical_layout
1256 @item force_autohint
1259 @item ignore_global_advance_width
1261 @item ignore_transform
1268 Default value is "render".
1270 For more information consult the documentation for the FT_LOAD_*
1274 The size in number of spaces to use for rendering the tab.
1278 If true, check and fix text coords to avoid clipping.
1281 For example the command:
1283 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
1286 will draw "Test Text" with font FreeSerif, using the default values
1287 for the optional parameters.
1291 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
1292 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
1295 will draw 'Test Text' with font FreeSerif of size 24 at position x=100
1296 and y=50 (counting from the top-left corner of the screen), text is
1297 yellow with a red box around it. Both the text and the box have an
1300 Note that the double quotes are not necessary if spaces are not used
1301 within the parameter list.
1303 For more information about libfreetype, check:
1304 @url{http://www.freetype.org/}.
1308 Apply a fade-in/out effect to the input video.
1310 It accepts the following parameters:
1315 The effect type can be either "in" for a fade-in, or "out" for a fade-out
1319 The number of the frame to start applying the fade effect at.
1322 The number of frames that the fade effect lasts. At the end of the
1323 fade-in effect, the output video will have the same intensity as the input video.
1324 At the end of the fade-out transition, the output video will be completely black.
1330 # Fade in the first 30 frames of video
1331 fade=type=in:nb_frames=30
1333 # Fade out the last 45 frames of a 200-frame video
1334 fade=type=out:start_frame=155:nb_frames=45
1336 # Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video
1337 fade=type=in:start_frame=0:nb_frames=25, fade=type=out:start_frame=975:nb_frames=25
1339 # Make the first 5 frames black, then fade in from frame 5-24
1340 fade=type=in:start_frame=5:nb_frames=20
1345 Transform the field order of the input video.
1347 It accepts the following parameters:
1352 The output field order. Valid values are @var{tff} for top field first or @var{bff}
1353 for bottom field first.
1356 The default value is "tff".
1358 The transformation is done by shifting the picture content up or down
1359 by one line, and filling the remaining line with appropriate picture content.
1360 This method is consistent with most broadcast field order converters.
1362 If the input video is not flagged as being interlaced, or it is already
1363 flagged as being of the required output field order, then this filter does
1364 not alter the incoming video.
1366 It is very useful when converting to or from PAL DV material,
1367 which is bottom field first.
1371 ./avconv -i in.vob -vf "fieldorder=order=bff" out.dv
1376 Buffer input images and send them when they are requested.
1378 It is mainly useful when auto-inserted by the libavfilter
1381 It does not take parameters.
1385 Convert the input video to one of the specified pixel formats.
1386 Libavfilter will try to pick one that is suitable as input to
1389 It accepts the following parameters:
1393 A '|'-separated list of pixel format names, such as
1394 "pix_fmts=yuv420p|monow|rgb24".
1400 # Convert the input video to the "yuv420p" format
1401 format=pix_fmts=yuv420p
1403 # Convert the input video to any of the formats in the list
1404 format=pix_fmts=yuv420p|yuv444p|yuv410p
1410 Convert the video to specified constant framerate by duplicating or dropping
1411 frames as necessary.
1413 It accepts the following parameters:
1417 The desired output framerate.
1420 Assume the first PTS should be the given value, in seconds. This allows for
1421 padding/trimming at the start of stream. By default, no assumption is made
1422 about the first frame's expected PTS, so no padding or trimming is done.
1423 For example, this could be set to 0 to pad the beginning with duplicates of
1424 the first frame if a video stream starts after the audio stream or to trim any
1425 frames with a negative PTS.
1431 Pack two different video streams into a stereoscopic video, setting proper
1432 metadata on supported codecs. The two views should have the same size and
1433 framerate and processing will stop when the shorter video ends. Please note
1434 that you may conveniently adjust view properties with the @ref{scale} and
1437 It accepts the following parameters:
1441 The desired packing format. Supported values are:
1446 The views are next to each other (default).
1449 The views are on top of each other.
1452 The views are packed by line.
1455 The views are packed by column.
1458 The views are temporally interleaved.
1467 # Convert left and right views into a frame-sequential video
1468 avconv -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
1470 # Convert views into a side-by-side video with the same output resolution as the input
1471 avconv -i LEFT -i RIGHT -filter_complex [0:v]scale=w=iw/2[left],[1:v]scale=w=iw/2[right],[left][right]framepack=sbs OUTPUT
1477 Apply a frei0r effect to the input video.
1479 To enable the compilation of this filter, you need to install the frei0r
1480 header and configure Libav with --enable-frei0r.
1482 It accepts the following parameters:
1487 The name of the frei0r effect to load. If the environment variable
1488 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
1489 directories specified by the colon-separated list in @env{FREIOR_PATH}.
1490 Otherwise, the standard frei0r paths are searched, in this order:
1491 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
1492 @file{/usr/lib/frei0r-1/}.
1495 A '|'-separated list of parameters to pass to the frei0r effect.
1499 A frei0r effect parameter can be a boolean (its value is either
1500 "y" or "n"), a double, a color (specified as
1501 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
1502 numbers between 0.0 and 1.0, inclusive) or by an @code{av_parse_color()} color
1503 description), a position (specified as @var{X}/@var{Y}, where
1504 @var{X} and @var{Y} are floating point numbers) and/or a string.
1506 The number and types of parameters depend on the loaded effect. If an
1507 effect parameter is not specified, the default value is set.
1511 # Apply the distort0r effect, setting the first two double parameters
1512 frei0r=filter_name=distort0r:filter_params=0.5|0.01
1514 # Apply the colordistance effect, taking a color as the first parameter
1515 frei0r=colordistance:0.2/0.3/0.4
1516 frei0r=colordistance:violet
1517 frei0r=colordistance:0x112233
1519 # Apply the perspective effect, specifying the top left and top right
1521 frei0r=perspective:0.2/0.2|0.8/0.2
1524 For more information, see
1525 @url{http://piksel.org/frei0r}
1529 Fix the banding artifacts that are sometimes introduced into nearly flat
1530 regions by truncation to 8bit colordepth.
1531 Interpolate the gradients that should go where the bands are, and
1534 It is designed for playback only. Do not use it prior to
1535 lossy compression, because compression tends to lose the dither and
1536 bring back the bands.
1538 It accepts the following parameters:
1543 The maximum amount by which the filter will change any one pixel. This is also
1544 the threshold for detecting nearly flat regions. Acceptable values range from
1545 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
1549 The neighborhood to fit the gradient to. A larger radius makes for smoother
1550 gradients, but also prevents the filter from modifying the pixels near detailed
1551 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
1552 values will be clipped to the valid range.
1557 # Default parameters
1558 gradfun=strength=1.2:radius=16
1560 # Omitting the radius
1566 Flip the input video horizontally.
1568 For example, to horizontally flip the input video with @command{avconv}:
1570 avconv -i in.avi -vf "hflip" out.avi
1575 This is a high precision/quality 3d denoise filter. It aims to reduce
1576 image noise, producing smooth images and making still images really
1577 still. It should enhance compressibility.
1579 It accepts the following optional parameters:
1583 A non-negative floating point number which specifies spatial luma strength.
1586 @item chroma_spatial
1587 A non-negative floating point number which specifies spatial chroma strength.
1588 It defaults to 3.0*@var{luma_spatial}/4.0.
1591 A floating point number which specifies luma temporal strength. It defaults to
1592 6.0*@var{luma_spatial}/4.0.
1595 A floating point number which specifies chroma temporal strength. It defaults to
1596 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
1601 Simple interlacing filter from progressive contents. This interleaves upper (or
1602 lower) lines from odd frames with lower (or upper) lines from even frames,
1603 halving the frame rate and preserving image height.
1606 Original Original New Frame
1607 Frame 'j' Frame 'j+1' (tff)
1608 ========== =========== ==================
1609 Line 0 --------------------> Frame 'j' Line 0
1610 Line 1 Line 1 ----> Frame 'j+1' Line 1
1611 Line 2 ---------------------> Frame 'j' Line 2
1612 Line 3 Line 3 ----> Frame 'j+1' Line 3
1614 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
1617 It accepts the following optional parameters:
1621 This determines whether the interlaced frame is taken from the even
1622 (tff - default) or odd (bff) lines of the progressive frame.
1625 Enable (default) or disable the vertical lowpass filter to avoid twitter
1626 interlacing and reduce moire patterns.
1629 @section lut, lutrgb, lutyuv
1631 Compute a look-up table for binding each pixel component input value
1632 to an output value, and apply it to the input video.
1634 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
1635 to an RGB input video.
1637 These filters accept the following parameters:
1639 @item @var{c0} (first pixel component)
1640 @item @var{c1} (second pixel component)
1641 @item @var{c2} (third pixel component)
1642 @item @var{c3} (fourth pixel component, corresponds to the alpha component)
1644 @item @var{r} (red component)
1645 @item @var{g} (green component)
1646 @item @var{b} (blue component)
1647 @item @var{a} (alpha component)
1649 @item @var{y} (Y/luminance component)
1650 @item @var{u} (U/Cb component)
1651 @item @var{v} (V/Cr component)
1654 Each of them specifies the expression to use for computing the lookup table for
1655 the corresponding pixel component values.
1657 The exact component associated to each of the @var{c*} options depends on the
1660 The @var{lut} filter requires either YUV or RGB pixel formats in input,
1661 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
1663 The expressions can contain the following constants and functions:
1667 These are approximated values for the mathematical constants e
1668 (Euler's number), pi (Greek pi), and phi (the golden ratio).
1671 The input width and height.
1674 The input value for the pixel component.
1677 The input value, clipped to the @var{minval}-@var{maxval} range.
1680 The maximum value for the pixel component.
1683 The minimum value for the pixel component.
1686 The negated value for the pixel component value, clipped to the
1687 @var{minval}-@var{maxval} range; it corresponds to the expression
1688 "maxval-clipval+minval".
1691 The computed value in @var{val}, clipped to the
1692 @var{minval}-@var{maxval} range.
1694 @item gammaval(gamma)
1695 The computed gamma correction value of the pixel component value,
1696 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
1698 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
1702 All expressions default to "val".
1706 # Negate input video
1707 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
1708 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
1710 # The above is the same as
1711 lutrgb="r=negval:g=negval:b=negval"
1712 lutyuv="y=negval:u=negval:v=negval"
1717 # Remove chroma components, turning the video into a graytone image
1718 lutyuv="u=128:v=128"
1720 # Apply a luma burning effect
1723 # Remove green and blue components
1726 # Set a constant alpha channel value on input
1727 format=rgba,lutrgb=a="maxval-minval/2"
1729 # Correct luminance gamma by a factor of 0.5
1730 lutyuv=y=gammaval(0.5)
1737 It accepts an integer in input; if non-zero it negates the
1738 alpha component (if available). The default value in input is 0.
1742 Force libavfilter not to use any of the specified pixel formats for the
1743 input to the next filter.
1745 It accepts the following parameters:
1749 A '|'-separated list of pixel format names, such as
1750 apix_fmts=yuv420p|monow|rgb24".
1756 # Force libavfilter to use a format different from "yuv420p" for the
1757 # input to the vflip filter
1758 noformat=pix_fmts=yuv420p,vflip
1760 # Convert the input video to any of the formats not contained in the list
1761 noformat=yuv420p|yuv444p|yuv410p
1766 Pass the video source unchanged to the output.
1770 Apply a video transform using libopencv.
1772 To enable this filter, install the libopencv library and headers and
1773 configure Libav with --enable-libopencv.
1775 It accepts the following parameters:
1780 The name of the libopencv filter to apply.
1783 The parameters to pass to the libopencv filter. If not specified, the default
1788 Refer to the official libopencv documentation for more precise
1790 @url{http://opencv.willowgarage.com/documentation/c/image_filtering.html}
1792 Several libopencv filters are supported; see the following subsections.
1797 Dilate an image by using a specific structuring element.
1798 It corresponds to the libopencv function @code{cvDilate}.
1800 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
1802 @var{struct_el} represents a structuring element, and has the syntax:
1803 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
1805 @var{cols} and @var{rows} represent the number of columns and rows of
1806 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
1807 point, and @var{shape} the shape for the structuring element. @var{shape}
1808 must be "rect", "cross", "ellipse", or "custom".
1810 If the value for @var{shape} is "custom", it must be followed by a
1811 string of the form "=@var{filename}". The file with name
1812 @var{filename} is assumed to represent a binary image, with each
1813 printable character corresponding to a bright pixel. When a custom
1814 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
1815 or columns and rows of the read file are assumed instead.
1817 The default value for @var{struct_el} is "3x3+0x0/rect".
1819 @var{nb_iterations} specifies the number of times the transform is
1820 applied to the image, and defaults to 1.
1824 # Use the default values
1827 # Dilate using a structuring element with a 5x5 cross, iterating two times
1828 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
1830 # Read the shape from the file diamond.shape, iterating two times.
1831 # The file diamond.shape may contain a pattern of characters like this
1837 # The specified columns and rows are ignored
1838 # but the anchor point coordinates are not
1839 ocv=dilate:0x0+2x2/custom=diamond.shape|2
1844 Erode an image by using a specific structuring element.
1845 It corresponds to the libopencv function @code{cvErode}.
1847 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
1848 with the same syntax and semantics as the @ref{dilate} filter.
1852 Smooth the input video.
1854 The filter takes the following parameters:
1855 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
1857 @var{type} is the type of smooth filter to apply, and must be one of
1858 the following values: "blur", "blur_no_scale", "median", "gaussian",
1859 or "bilateral". The default value is "gaussian".
1861 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
1862 depend on the smooth type. @var{param1} and
1863 @var{param2} accept integer positive values or 0. @var{param3} and
1864 @var{param4} accept floating point values.
1866 The default value for @var{param1} is 3. The default value for the
1867 other parameters is 0.
1869 These parameters correspond to the parameters assigned to the
1870 libopencv function @code{cvSmooth}.
1875 Overlay one video on top of another.
1877 It takes two inputs and has one output. The first input is the "main"
1878 video on which the second input is overlayed.
1880 It accepts the following parameters:
1885 The horizontal position of the left edge of the overlaid video on the main video.
1888 The vertical position of the top edge of the overlaid video on the main video.
1892 The parameters are expressions containing the following parameters:
1895 @item main_w, main_h
1896 The main input width and height.
1899 These are the same as @var{main_w} and @var{main_h}.
1901 @item overlay_w, overlay_h
1902 The overlay input width and height.
1905 These are the same as @var{overlay_w} and @var{overlay_h}.
1908 The action to take when EOF is encountered on the secondary input; it accepts
1909 one of the following values:
1913 Repeat the last frame (the default).
1917 Pass the main input through.
1922 Be aware that frames are taken from each input video in timestamp
1923 order, hence, if their initial timestamps differ, it is a a good idea
1924 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
1925 have them begin in the same zero timestamp, as the example for
1926 the @var{movie} filter does.
1930 # Draw the overlay at 10 pixels from the bottom right
1931 # corner of the main video
1932 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
1934 # Insert a transparent PNG logo in the bottom left corner of the input
1935 avconv -i input -i logo -filter_complex 'overlay=x=10:y=main_h-overlay_h-10' output
1937 # Insert 2 different transparent PNG logos (second logo on bottom
1939 avconv -i input -i logo1 -i logo2 -filter_complex
1940 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
1942 # Add a transparent color layer on top of the main video;
1943 # WxH specifies the size of the main input to the overlay filter
1944 color=red@.3:WxH [over]; [in][over] overlay [out]
1946 # Mask 10-20 seconds of a video by applying the delogo filter to a section
1947 avconv -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
1948 -vf '[in]split[split_main][split_delogo];[split_delogo]trim=start=360:end=371,delogo=0:0:640:480[delogoed];[split_main][delogoed]overlay=eof_action=pass[out]'
1952 You can chain together more overlays but the efficiency of such
1953 approach is yet to be tested.
1957 Add paddings to the input image, and place the original input at the
1958 provided @var{x}, @var{y} coordinates.
1960 It accepts the following parameters:
1965 Specify the size of the output image with the paddings added. If the
1966 value for @var{width} or @var{height} is 0, the corresponding input size
1967 is used for the output.
1969 The @var{width} expression can reference the value set by the
1970 @var{height} expression, and vice versa.
1972 The default value of @var{width} and @var{height} is 0.
1976 Specify the offsets to place the input image at within the padded area,
1977 with respect to the top/left border of the output image.
1979 The @var{x} expression can reference the value set by the @var{y}
1980 expression, and vice versa.
1982 The default value of @var{x} and @var{y} is 0.
1986 Specify the color of the padded area. It can be the name of a color
1987 (case insensitive match) or an 0xRRGGBB[AA] sequence.
1989 The default value of @var{color} is "black".
1993 The parameters @var{width}, @var{height}, @var{x}, and @var{y} are
1994 expressions containing the following constants:
1998 These are approximated values for the mathematical constants e
1999 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2002 The input video width and height.
2005 These are the same as @var{in_w} and @var{in_h}.
2008 The output width and height (the size of the padded area), as
2009 specified by the @var{width} and @var{height} expressions.
2012 These are the same as @var{out_w} and @var{out_h}.
2015 The x and y offsets as specified by the @var{x} and @var{y}
2016 expressions, or NAN if not yet specified.
2019 The input display aspect ratio, same as @var{iw} / @var{ih}.
2022 The horizontal and vertical chroma subsample values. For example for the
2023 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2029 # Add paddings with the color "violet" to the input video. The output video
2030 # size is 640x480, and the top-left corner of the input video is placed at
2032 pad=width=640:height=480:x=0:y=40:color=violet
2034 # Pad the input to get an output with dimensions increased by 3/2,
2035 # and put the input video at the center of the padded area
2036 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
2038 # Pad the input to get a squared output with size equal to the maximum
2039 # value between the input width and height, and put the input video at
2040 # the center of the padded area
2041 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
2043 # Pad the input to get a final w/h ratio of 16:9
2044 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
2046 # Double the output size and put the input video in the bottom-right
2047 # corner of the output padded area
2048 pad="2*iw:2*ih:ow-iw:oh-ih"
2051 @section pixdesctest
2053 Pixel format descriptor test filter, mainly useful for internal
2054 testing. The output video should be equal to the input video.
2058 format=monow, pixdesctest
2061 can be used to test the monowhite pixel format descriptor definition.
2066 Scale the input video and/or convert the image format.
2068 It accepts the following parameters:
2073 The output video width.
2076 The output video height.
2080 The parameters @var{w} and @var{h} are expressions containing
2081 the following constants:
2085 These are approximated values for the mathematical constants e
2086 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2089 The input width and height.
2092 These are the same as @var{in_w} and @var{in_h}.
2095 The output (cropped) width and height.
2098 These are the same as @var{out_w} and @var{out_h}.
2101 This is the same as @var{iw} / @var{ih}.
2104 input sample aspect ratio
2107 The input display aspect ratio; it is the same as
2108 (@var{iw} / @var{ih}) * @var{sar}.
2111 The horizontal and vertical chroma subsample values. For example, for the
2112 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2115 If the input image format is different from the format requested by
2116 the next filter, the scale filter will convert the input to the
2119 If the value for @var{w} or @var{h} is 0, the respective input
2120 size is used for the output.
2122 If the value for @var{w} or @var{h} is -1, the scale filter will use, for the
2123 respective output size, a value that maintains the aspect ratio of the input
2126 The default value of @var{w} and @var{h} is 0.
2130 # Scale the input video to a size of 200x100
2133 # Scale the input to 2x
2135 # The above is the same as
2138 # Scale the input to half the original size
2141 # Increase the width, and set the height to the same size
2144 # Seek Greek harmony
2148 # Increase the height, and set the width to 3/2 of the height
2149 scale=w=3/2*oh:h=3/5*ih
2151 # Increase the size, making the size a multiple of the chroma
2152 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
2154 # Increase the width to a maximum of 500 pixels,
2155 # keeping the same aspect ratio as the input
2156 scale=w='min(500\, iw*3/2):h=-1'
2160 Select frames to pass in output.
2162 It accepts the following parameters:
2167 An expression, which is evaluated for each input frame. If the expression is
2168 evaluated to a non-zero value, the frame is selected and passed to the output,
2169 otherwise it is discarded.
2173 The expression can contain the following constants:
2177 These are approximated values for the mathematical constants e
2178 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2181 The (sequential) number of the filtered frame, starting from 0.
2184 The (sequential) number of the selected frame, starting from 0.
2186 @item prev_selected_n
2187 The sequential number of the last selected frame. It's NAN if undefined.
2190 The timebase of the input timestamps.
2193 The PTS (Presentation TimeStamp) of the filtered video frame,
2194 expressed in @var{TB} units. It's NAN if undefined.
2197 The PTS of the filtered video frame,
2198 expressed in seconds. It's NAN if undefined.
2201 The PTS of the previously filtered video frame. It's NAN if undefined.
2203 @item prev_selected_pts
2204 The PTS of the last previously filtered video frame. It's NAN if undefined.
2206 @item prev_selected_t
2207 The PTS of the last previously selected video frame. It's NAN if undefined.
2210 The PTS of the first video frame in the video. It's NAN if undefined.
2213 The time of the first video frame in the video. It's NAN if undefined.
2216 The type of the filtered frame. It can assume one of the following
2228 @item interlace_type
2229 The frame interlace type. It can assume one of the following values:
2232 The frame is progressive (not interlaced).
2234 The frame is top-field-first.
2236 The frame is bottom-field-first.
2240 This is 1 if the filtered frame is a key-frame, 0 otherwise.
2244 The default value of the select expression is "1".
2249 # Select all the frames in input
2252 # The above is the same as
2258 # Select only I-frames
2259 select='expr=eq(pict_type\,I)'
2261 # Select one frame per 100
2262 select='not(mod(n\,100))'
2264 # Select only frames contained in the 10-20 time interval
2265 select='gte(t\,10)*lte(t\,20)'
2267 # Select only I frames contained in the 10-20 time interval
2268 select='gte(t\,10)*lte(t\,20)*eq(pict_type\,I)'
2270 # Select frames with a minimum distance of 10 seconds
2271 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
2277 Set the Display Aspect Ratio for the filter output video.
2279 This is done by changing the specified Sample (aka Pixel) Aspect
2280 Ratio, according to the following equation:
2281 @math{DAR = HORIZONTAL_RESOLUTION / VERTICAL_RESOLUTION * SAR}
2283 Keep in mind that this filter does not modify the pixel dimensions of
2284 the video frame. Also, the display aspect ratio set by this filter may
2285 be changed by later filters in the filterchain, e.g. in case of
2286 scaling or if another "setdar" or a "setsar" filter is applied.
2288 It accepts the following parameters:
2293 The output display aspect ratio.
2297 The parameter @var{dar} is an expression containing
2298 the following constants:
2302 These are approximated values for the mathematical constants e
2303 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2306 The input width and height.
2309 This is the same as @var{w} / @var{h}.
2312 The input sample aspect ratio.
2315 The input display aspect ratio. It is the same as
2316 (@var{w} / @var{h}) * @var{sar}.
2319 The horizontal and vertical chroma subsample values. For example, for the
2320 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2323 To change the display aspect ratio to 16:9, specify:
2326 # The above is equivalent to
2330 Also see the the @ref{setsar} filter documentation.
2334 Change the PTS (presentation timestamp) of the input video frames.
2336 It accepts the following parameters:
2341 The expression which is evaluated for each frame to construct its timestamp.
2345 The expression is evaluated through the eval API and can contain the following
2350 The presentation timestamp in input.
2353 These are approximated values for the mathematical constants e
2354 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2357 The count of the input frame, starting from 0.
2360 The PTS of the first video frame.
2363 State whether the current frame is interlaced.
2366 The previous input PTS.
2369 The previous output PTS.
2372 The wallclock (RTC) time in microseconds.
2375 The wallclock (RTC) time at the start of the movie in microseconds.
2378 The timebase of the input timestamps.
2385 # Start counting the PTS from zero
2386 setpts=expr=PTS-STARTPTS
2397 # Fixed rate 25 fps with some jitter
2398 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
2400 # Generate timestamps from a "live source" and rebase onto the current timebase
2401 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)"
2407 Set the Sample (aka Pixel) Aspect Ratio for the filter output video.
2409 Note that as a consequence of the application of this filter, the
2410 output display aspect ratio will change according to the following
2412 @math{DAR = HORIZONTAL_RESOLUTION / VERTICAL_RESOLUTION * SAR}
2414 Keep in mind that the sample aspect ratio set by this filter may be
2415 changed by later filters in the filterchain, e.g. if another "setsar"
2416 or a "setdar" filter is applied.
2418 It accepts the following parameters:
2423 The output sample aspect ratio.
2427 The parameter @var{sar} is an expression containing
2428 the following constants:
2432 These are approximated values for the mathematical constants e
2433 (Euler's number), pi (Greek pi), and phi (the golden ratio).
2436 The input width and height.
2439 These are the same as @var{w} / @var{h}.
2442 The input sample aspect ratio.
2445 The input display aspect ratio. It is the same as
2446 (@var{w} / @var{h}) * @var{sar}.
2449 Horizontal and vertical chroma subsample values. For example, for the
2450 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2453 To change the sample aspect ratio to 10:11, specify:
2460 Set the timebase to use for the output frames timestamps.
2461 It is mainly useful for testing timebase configuration.
2463 It accepts the following parameters:
2468 The expression which is evaluated into the output timebase.
2472 The expression can contain the constants "PI", "E", "PHI", "AVTB" (the
2473 default timebase), and "intb" (the input timebase).
2475 The default value for the input is "intb".
2480 # Set the timebase to 1/25
2483 # Set the timebase to 1/10
2486 # Set the timebase to 1001/1000
2489 #Set the timebase to 2*intb
2492 #Set the default timebase value
2498 Show a line containing various information for each input video frame.
2499 The input video is not modified.
2501 The shown line contains a sequence of key/value pairs of the form
2502 @var{key}:@var{value}.
2504 It accepts the following parameters:
2508 The (sequential) number of the input frame, starting from 0.
2511 The Presentation TimeStamp of the input frame, expressed as a number of
2512 time base units. The time base unit depends on the filter input pad.
2515 The Presentation TimeStamp of the input frame, expressed as a number of
2519 The position of the frame in the input stream, or -1 if this information is
2520 unavailable and/or meaningless (for example in case of synthetic video).
2523 The pixel format name.
2526 The sample aspect ratio of the input frame, expressed in the form
2527 @var{num}/@var{den}.
2530 The size of the input frame, expressed in the form
2531 @var{width}x@var{height}.
2534 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
2535 for bottom field first).
2538 This is 1 if the frame is a key frame, 0 otherwise.
2541 The picture type of the input frame ("I" for an I-frame, "P" for a
2542 P-frame, "B" for a B-frame, or "?" for an unknown type).
2543 Also refer to the documentation of the @code{AVPictureType} enum and of
2544 the @code{av_get_picture_type_char} function defined in
2545 @file{libavutil/avutil.h}.
2548 The Adler-32 checksum of all the planes of the input frame.
2550 @item plane_checksum
2551 The Adler-32 checksum of each plane of the input frame, expressed in the form
2552 "[@var{c0} @var{c1} @var{c2} @var{c3}]".
2555 @section shuffleplanes
2557 Reorder and/or duplicate video planes.
2559 It accepts the following parameters:
2564 The index of the input plane to be used as the first output plane.
2567 The index of the input plane to be used as the second output plane.
2570 The index of the input plane to be used as the third output plane.
2573 The index of the input plane to be used as the fourth output plane.
2577 The first plane has the index 0. The default is to keep the input unchanged.
2579 Swap the second and third planes of the input:
2581 avconv -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
2586 Split input video into several identical outputs.
2588 It accepts a single parameter, which specifies the number of outputs. If
2589 unspecified, it defaults to 2.
2591 Create 5 copies of the input video:
2593 avconv -i INPUT -filter_complex split=5 OUTPUT
2598 Transpose rows with columns in the input video and optionally flip it.
2600 It accepts the following parameters:
2605 The direction of the transpose.
2609 The direction can assume the following values:
2613 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
2621 Rotate by 90 degrees clockwise, that is:
2629 Rotate by 90 degrees counterclockwise, that is:
2637 Rotate by 90 degrees clockwise and vertically flip, that is:
2646 Trim the input so that the output contains one continuous subpart of the input.
2648 It accepts the following parameters:
2651 The timestamp (in seconds) of the start of the kept section. The frame with the
2652 timestamp @var{start} will be the first frame in the output.
2655 The timestamp (in seconds) of the first frame that will be dropped. The frame
2656 immediately preceding the one with the timestamp @var{end} will be the last
2657 frame in the output.
2660 This is the same as @var{start}, except this option sets the start timestamp
2661 in timebase units instead of seconds.
2664 This is the same as @var{end}, except this option sets the end timestamp
2665 in timebase units instead of seconds.
2668 The maximum duration of the output in seconds.
2671 The number of the first frame that should be passed to the output.
2674 The number of the first frame that should be dropped.
2677 Note that the first two sets of the start/end options and the @option{duration}
2678 option look at the frame timestamp, while the _frame variants simply count the
2679 frames that pass through the filter. Also note that this filter does not modify
2680 the timestamps. If you wish for the output timestamps to start at zero, insert a
2681 setpts filter after the trim filter.
2683 If multiple start or end options are set, this filter tries to be greedy and
2684 keep all the frames that match at least one of the specified constraints. To keep
2685 only the part that matches all the constraints at once, chain multiple trim
2688 The defaults are such that all the input is kept. So it is possible to set e.g.
2689 just the end values to keep everything before the specified time.
2694 Drop everything except the second minute of input:
2696 avconv -i INPUT -vf trim=60:120
2700 Keep only the first second:
2702 avconv -i INPUT -vf trim=duration=1
2708 Sharpen or blur the input video.
2710 It accepts the following parameters:
2715 Set the luma matrix horizontal size. It must be an integer between 3
2716 and 13. The default value is 5.
2719 Set the luma matrix vertical size. It must be an integer between 3
2720 and 13. The default value is 5.
2723 Set the luma effect strength. It must be a floating point number between -2.0
2724 and 5.0. The default value is 1.0.
2726 @item chroma_msize_x
2727 Set the chroma matrix horizontal size. It must be an integer between 3
2728 and 13. The default value is 5.
2730 @item chroma_msize_y
2731 Set the chroma matrix vertical size. It must be an integer between 3
2732 and 13. The default value is 5.
2735 Set the chroma effect strength. It must be a floating point number between -2.0
2736 and 5.0. The default value is 0.0.
2740 Negative values for the amount will blur the input video, while positive
2741 values will sharpen. All parameters are optional and default to the
2742 equivalent of the string '5:5:1.0:5:5:0.0'.
2745 # Strong luma sharpen effect parameters
2746 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
2748 # A strong blur of both luma and chroma parameters
2749 unsharp=7:7:-2:7:7:-2
2751 # Use the default values with @command{avconv}
2752 ./avconv -i in.avi -vf "unsharp" out.mp4
2757 Flip the input video vertically.
2760 ./avconv -i in.avi -vf "vflip" out.avi
2765 Deinterlace the input video ("yadif" means "yet another deinterlacing
2768 It accepts the following parameters:
2773 The interlacing mode to adopt. It accepts one of the following values:
2777 Output one frame for each frame.
2779 Output one frame for each field.
2781 Like 0, but it skips the spatial interlacing check.
2783 Like 1, but it skips the spatial interlacing check.
2786 The default value is 0.
2789 The picture field parity assumed for the input interlaced video. It accepts one
2790 of the following values:
2794 Assume the top field is first.
2796 Assume the bottom field is first.
2798 Enable automatic detection of field parity.
2801 The default value is -1.
2802 If the interlacing is unknown or the decoder does not export this information,
2803 top field first will be assumed.
2806 Whether the deinterlacer should trust the interlaced flag and only deinterlace
2807 frames marked as interlaced.
2811 Deinterlace all frames.
2813 Only deinterlace frames marked as interlaced.
2816 The default value is 0.
2820 @c man end VIDEO FILTERS
2822 @chapter Video Sources
2823 @c man begin VIDEO SOURCES
2825 Below is a description of the currently available video sources.
2829 Buffer video frames, and make them available to the filter chain.
2831 This source is mainly intended for a programmatic use, in particular
2832 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
2834 It accepts the following parameters:
2839 The input video width.
2842 The input video height.
2845 The name of the input video pixel format.
2848 The time base used for input timestamps.
2851 The sample (pixel) aspect ratio of the input video.
2857 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
2860 will instruct the source to accept video frames with size 320x240 and
2861 with format "yuv410p", assuming 1/24 as the timestamps timebase and
2862 square pixels (1:1 sample aspect ratio).
2866 Provide an uniformly colored input.
2868 It accepts the following parameters:
2873 Specify the color of the source. It can be the name of a color (case
2874 insensitive match) or a 0xRRGGBB[AA] sequence, possibly followed by an
2875 alpha specifier. The default value is "black".
2878 Specify the size of the sourced video, it may be a string of the form
2879 @var{width}x@var{height}, or the name of a size abbreviation. The
2880 default value is "320x240".
2883 Specify the frame rate of the sourced video, as the number of frames
2884 generated per second. It has to be a string in the format
2885 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
2886 number or a valid video frame rate abbreviation. The default value is
2891 The following graph description will generate a red source
2892 with an opacity of 0.2, with size "qcif" and a frame rate of 10
2893 frames per second, which will be overlayed over the source connected
2894 to the pad with identifier "in":
2897 "color=red@@0.2:qcif:10 [color]; [in][color] overlay [out]"
2902 Read a video stream from a movie container.
2904 Note that this source is a hack that bypasses the standard input path. It can be
2905 useful in applications that do not support arbitrary filter graphs, but its use
2906 is discouraged in those that do. It should never be used with
2907 @command{avconv}; the @option{-filter_complex} option fully replaces it.
2909 It accepts the following parameters:
2914 The name of the resource to read (not necessarily a file; it can also be a
2915 device or a stream accessed through some protocol).
2917 @item format_name, f
2918 Specifies the format assumed for the movie to read, and can be either
2919 the name of a container or an input device. If not specified, the
2920 format is guessed from @var{movie_name} or by probing.
2922 @item seek_point, sp
2923 Specifies the seek point in seconds. The frames will be output
2924 starting from this seek point. The parameter is evaluated with
2925 @code{av_strtod}, so the numerical value may be suffixed by an IS
2926 postfix. The default value is "0".
2928 @item stream_index, si
2929 Specifies the index of the video stream to read. If the value is -1,
2930 the most suitable video stream will be automatically selected. The default
2935 It allows overlaying a second video on top of the main input of
2936 a filtergraph, as shown in this graph:
2938 input -----------> deltapts0 --> overlay --> output
2941 movie --> scale--> deltapts1 -------+
2946 # Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
2947 # on top of the input labelled "in"
2948 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [movie];
2949 [in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
2951 # Read from a video4linux2 device, and overlay it on top of the input
2953 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [movie];
2954 [in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
2960 Null video source: never return images. It is mainly useful as a
2961 template and to be employed in analysis / debugging tools.
2963 It accepts a string of the form
2964 @var{width}:@var{height}:@var{timebase} as an optional parameter.
2966 @var{width} and @var{height} specify the size of the configured
2967 source. The default values of @var{width} and @var{height} are
2968 respectively 352 and 288 (corresponding to the CIF size format).
2970 @var{timebase} specifies an arithmetic expression representing a
2971 timebase. The expression can contain the constants "PI", "E", "PHI", and
2972 "AVTB" (the default timebase), and defaults to the value "AVTB".
2976 Provide a frei0r source.
2978 To enable compilation of this filter you need to install the frei0r
2979 header and configure Libav with --enable-frei0r.
2981 This source accepts the following parameters:
2986 The size of the video to generate. It may be a string of the form
2987 @var{width}x@var{height} or a frame size abbreviation.
2990 The framerate of the generated video. It may be a string of the form
2991 @var{num}/@var{den} or a frame rate abbreviation.
2994 The name to the frei0r source to load. For more information regarding frei0r and
2995 how to set the parameters, read the @ref{frei0r} section in the video filters
2999 A '|'-separated list of parameters to pass to the frei0r source.
3005 # Generate a frei0r partik0l source with size 200x200 and framerate 10
3006 # which is overlayed on the overlay filter main input
3007 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
3010 @section rgbtestsrc, testsrc
3012 The @code{rgbtestsrc} source generates an RGB test pattern useful for
3013 detecting RGB vs BGR issues. You should see a red, green and blue
3014 stripe from top to bottom.
3016 The @code{testsrc} source generates a test video pattern, showing a
3017 color pattern, a scrolling gradient and a timestamp. This is mainly
3018 intended for testing purposes.
3020 The sources accept the following parameters:
3025 Specify the size of the sourced video, it may be a string of the form
3026 @var{width}x@var{height}, or the name of a size abbreviation. The
3027 default value is "320x240".
3030 Specify the frame rate of the sourced video, as the number of frames
3031 generated per second. It has to be a string in the format
3032 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
3033 number or a valid video frame rate abbreviation. The default value is
3037 Set the sample aspect ratio of the sourced video.
3040 Set the video duration of the sourced video. The accepted syntax is:
3042 [-]HH[:MM[:SS[.m...]]]
3045 Also see the the @code{av_parse_time()} function.
3047 If not specified, or the expressed duration is negative, the video is
3048 supposed to be generated forever.
3051 For example the following:
3053 testsrc=duration=5.3:size=qcif:rate=10
3056 will generate a video with a duration of 5.3 seconds, with size
3057 176x144 and a framerate of 10 frames per second.
3059 @c man end VIDEO SOURCES
3061 @chapter Video Sinks
3062 @c man begin VIDEO SINKS
3064 Below is a description of the currently available video sinks.
3068 Buffer video frames, and make them available to the end of the filter
3071 This sink is intended for programmatic use through the interface defined in
3072 @file{libavfilter/buffersink.h}.
3076 Null video sink: do absolutely nothing with the input video. It is
3077 mainly useful as a template and for use in analysis / debugging
3080 @c man end VIDEO SINKS