lavfi: add compand audio filter
[libav.git] / doc / filters.texi
CommitLineData
1ebe5c4c
SS
1@chapter Filtergraph description
2@c man begin FILTERGRAPH DESCRIPTION
3
4A filtergraph is a directed graph of connected filters. It can contain
5cycles, and there can be multiple links between a pair of
6filters. Each link has one input pad on one side connecting it to one
7filter from which it takes its input, and one output pad on the other
8side connecting it to the one filter accepting its output.
9
10Each filter in a filtergraph is an instance of a filter class
11registered in the application, which defines the features and the
12number of input and output pads of the filter.
13
14A filter with no input pads is called a "source", a filter with no
15output pads is called a "sink".
16
3b266da3 17@anchor{Filtergraph syntax}
1ebe5c4c
SS
18@section Filtergraph syntax
19
7ce118ba
AK
20A filtergraph can be represented using a textual representation, which is
21recognized by the @option{-filter}/@option{-vf} and @option{-filter_complex}
22options in @command{avconv} and @option{-vf} in @command{avplay}, and by the
23@code{avfilter_graph_parse()}/@code{avfilter_graph_parse2()} function defined in
38f0c078 24@file{libavfilter/avfilter.h}.
1ebe5c4c
SS
25
26A filterchain consists of a sequence of connected filters, each one
27connected to the previous one in the sequence. A filterchain is
28represented by a list of ","-separated filter descriptions.
29
30A filtergraph consists of a sequence of filterchains. A sequence of
31filterchains is represented by a list of ";"-separated filterchain
32descriptions.
33
34A 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}]
36
37@var{filter_name} is the name of the filter class of which the
38described filter is an instance of, and has to be the name of one of
39the filter classes registered in the program.
40The name of the filter class is optionally followed by a string
41"=@var{arguments}".
42
43@var{arguments} is a string which contains the parameters used to
b439c992
AK
44initialize the filter instance. It may have one of the two allowed forms:
45@itemize
46
47@item
48A ':'-separated list of @var{key=value} pairs.
49
50@item
51A ':'-separated list of @var{value}. In this case, the keys are assumed to be
52the option names in the order they are declared. E.g. the @code{fade} filter
53declares 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}.
57
58@end itemize
59
60If the option value itself is a list of items (e.g. the @code{format} filter
61takes a list of pixel formats), the items in the list are usually separated by
62'|'.
1ebe5c4c
SS
63
64The list of arguments can be quoted using the character "'" as initial
65and ending mark, and the character '\' for escaping the characters
66within the quoted text; otherwise the argument string is considered
67terminated when the next special character (belonging to the set
68"[]=;,") is encountered.
69
70The name and arguments of the filter are optionally preceded and
71followed by a list of link labels.
72A link label allows to name a link and associate it to a filter output
73or input pad. The preceding labels @var{in_link_1}
74... @var{in_link_N}, are associated to the filter input pads,
75the following labels @var{out_link_1} ... @var{out_link_M}, are
76associated to the output pads.
77
78When two link labels with the same name are found in the
79filtergraph, a link between the corresponding input and output pad is
80created.
81
82If an output pad is not labelled, it is linked by default to the first
83unlabelled input pad of the next filter in the filterchain.
84For example in the filterchain:
85@example
86nullsrc, split[L1], [L2]overlay, nullsink
87@end example
88the split filter instance has two output pads, and the overlay filter
89instance 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
91output pad of split is linked to the second input pad of overlay,
92which are both unlabelled.
93
94In a complete filterchain all the unlabelled filter input and output
95pads must be connected. A filtergraph is considered valid if all the
96filter input and output pads of all the filterchains are connected.
97
dd249245 98Libavfilter will automatically insert @ref{scale} filters where format
12e7e1d0
AK
99conversion is required. It is possible to specify swscale flags
100for those automatically inserted scalers by prepending
101@code{sws_flags=@var{flags};}
102to the filtergraph description.
103
1ebe5c4c
SS
104Follows a BNF description for the filtergraph syntax:
105@example
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 (eventually quoted)
b5ad422b 110@var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
1ebe5c4c 111@var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
12e7e1d0 112@var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
1ebe5c4c
SS
113@end example
114
115@c man end FILTERGRAPH DESCRIPTION
116
f59e9eaf
SS
117@chapter Audio Filters
118@c man begin AUDIO FILTERS
119
f8a45fa1 120When you configure your Libav build, you can disable any of the
f59e9eaf
SS
121existing filters using --disable-filters.
122The configure output will show the audio filters included in your
123build.
124
125Below is a description of the currently available audio filters.
126
fb604ae8
AK
127@section aformat
128
129Convert the input audio to one of the specified formats. The framework will
130negotiate the most appropriate format to minimize conversions.
131
132The filter accepts the following named parameters:
133@table @option
134
135@item sample_fmts
0af7fe1f 136A '|'-separated list of requested sample formats.
fb604ae8
AK
137
138@item sample_rates
0af7fe1f 139A '|'-separated list of requested sample rates.
fb604ae8
AK
140
141@item channel_layouts
0af7fe1f 142A '|'-separated list of requested channel layouts.
fb604ae8
AK
143
144@end table
145
146If a parameter is omitted, all values are allowed.
147
148For example to force the output to either unsigned 8-bit or signed 16-bit stereo:
149@example
0af7fe1f 150aformat=sample_fmts=u8|s16:channel_layouts=stereo
fb604ae8
AK
151@end example
152
c7448c18
JR
153@section amix
154
155Mixes multiple audio inputs into a single output.
156
157For example
158@example
159avconv -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
160@end example
161will mix 3 input audio streams to a single output with the same duration as the
162first input and a dropout transition time of 3 seconds.
163
164The filter accepts the following named parameters:
165@table @option
166
167@item inputs
168Number of inputs. If unspecified, it defaults to 2.
169
170@item duration
171How to determine the end-of-stream.
172@table @option
173
174@item longest
175Duration of longest input. (default)
176
177@item shortest
178Duration of shortest input.
179
180@item first
181Duration of first input.
182
183@end table
184
185@item dropout_transition
186Transition time, in seconds, for volume renormalization when an input
187stream ends. The default value is 2 seconds.
188
189@end table
190
99046339
HM
191@section anull
192
193Pass the audio source unchanged to the output.
194
b4729382
AK
195@section asetpts
196
197Change the PTS (presentation timestamp) of the input audio frames.
198
199This filter accepts the following options:
200
201@table @option
202
203@item expr
204The expression which is evaluated for each frame to construct its timestamp.
205
206@end table
207
208The expression is evaluated through the eval API and can contain the following
209constants:
210
211@table @option
212@item PTS
213the presentation timestamp in input
214
215@item PI
216Greek PI
217
218@item PHI
219golden ratio
220
221@item E
222Euler number
223
224@item N
225Number of the audio samples pass through the filter so far, starting at 0.
226
227@item S
228Number of the audio samples in the current frame.
229
230@item SR
231Audio sample rate.
232
233@item STARTPTS
234the PTS of the first frame
235
236@item PREV_INPTS
237previous input PTS
238
239@item PREV_OUTPTS
240previous output PTS
241
242@item RTCTIME
243wallclock (RTC) time in microseconds
244
245@item RTCSTART
246wallclock (RTC) time at the start of the movie in microseconds
247
248@end table
249
250Some examples follow:
251
252@example
253# start counting PTS from zero
254asetpts=expr=PTS-STARTPTS
255
256#generate timestamps by counting samples
257asetpts=expr=N/SR/TB
258
259# generate timestamps from a "live source" and rebase onto the current timebase
260asetpts='(RTCTIME - RTCSTART) / (TB * 1000000)"
261@end example
262
263
20dd41af
AK
264@section ashowinfo
265
266Show a line containing various information for each input audio frame.
267The input audio is not modified.
268
269The shown line contains a sequence of key/value pairs of the form
270@var{key}:@var{value}.
271
272A description of each shown parameter follows:
273
274@table @option
275@item n
276sequential number of the input frame, starting from 0
277
278@item pts
279Presentation timestamp of the input frame, in time base units; the time base
280depends on the filter input pad, and is usually 1/@var{sample_rate}.
281
282@item pts_time
283presentation timestamp of the input frame in seconds
284
285@item fmt
286sample format
287
288@item chlayout
289channel layout
290
291@item rate
292sample rate for the audio frame
293
294@item nb_samples
295number of samples (per channel) in the frame
296
297@item checksum
298Adler-32 checksum (printed in hexadecimal) of the audio data. For planar audio
299the data is treated as if all the planes were concatenated.
300
301@item plane_checksums
302A list of Adler-32 checksums for each data plane.
303@end table
304
afeb3590
JR
305@section asplit
306
307Split input audio into several identical outputs.
308
309The filter accepts a single parameter which specifies the number of outputs. If
310unspecified, it defaults to 2.
311
312For example
313@example
314avconv -i INPUT -filter_complex asplit=5 OUTPUT
315@end example
316will create 5 copies of the input audio.
317
9f26421b
AK
318@section asyncts
319Synchronize audio data with timestamps by squeezing/stretching it and/or
320dropping samples/adding silence when needed.
321
322The filter accepts the following named parameters:
323@table @option
324
325@item compensate
04438790
AU
326Enable stretching/squeezing the data to make it match the timestamps. Disabled
327by default. When disabled, time gaps are covered with silence.
9f26421b
AK
328
329@item min_delta
330Minimum difference between timestamps and audio data (in seconds) to trigger
04438790
AU
331adding/dropping samples. Default value is 0.1. If you get non-perfect sync with
332this filter, try setting this parameter to 0.
9f26421b
AK
333
334@item max_comp
04438790
AU
335Maximum compensation in samples per second. Relevant only with compensate=1.
336Default value 500.
9f26421b 337
70d71b58 338@item first_pts
b35e5d98 339Assume the first pts should be this value. The time base is 1 / sample rate.
70d71b58
JR
340This allows for padding/trimming at the start of stream. By default, no
341assumption is made about the first frame's expected pts, so no padding or
342trimming is done. For example, this could be set to 0 to pad the beginning with
b35e5d98
JR
343silence if an audio stream starts after the video stream or to trim any samples
344with a negative pts due to encoder delay.
70d71b58 345
9f26421b
AK
346@end table
347
a1e05b04
AK
348@section atrim
349Trim the input so that the output contains one continuous subpart of the input.
350
351This filter accepts the following options:
352@table @option
353@item start
354Timestamp (in seconds) of the start of the kept section. I.e. the audio sample
355with the timestamp @var{start} will be the first sample in the output.
356
357@item end
358Timestamp (in seconds) of the first audio sample that will be dropped. I.e. the
359audio sample immediately preceding the one with the timestamp @var{end} will be
360the last sample in the output.
361
362@item start_pts
363Same as @var{start}, except this option sets the start timestamp in samples
364instead of seconds.
365
366@item end_pts
367Same as @var{end}, except this option sets the end timestamp in samples instead
368of seconds.
369
370@item duration
371Maximum duration of the output in seconds.
372
373@item start_sample
374Number of the first sample that should be passed to output.
375
376@item end_sample
377Number of the first sample that should be dropped.
378@end table
379
380Note that the first two sets of the start/end options and the @option{duration}
381option look at the frame timestamp, while the _sample options simply count the
382samples that pass through the filter. So start/end_pts and start/end_sample will
383give different results when the timestamps are wrong, inexact or do not start at
384zero. Also note that this filter does not modify the timestamps. If you wish
385that the output timestamps start at zero, insert the asetpts filter after the
386atrim filter.
387
388If multiple start or end options are set, this filter tries to be greedy and
389keep all samples that match at least one of the specified constraints. To keep
390only the part that matches all the constraints at once, chain multiple atrim
391filters.
392
393The defaults are such that all the input is kept. So it is possible to set e.g.
394just the end values to keep everything before the specified time.
395
396Examples:
397@itemize
398@item
399drop everything except the second minute of input
400@example
401avconv -i INPUT -af atrim=60:120
402@end example
403
404@item
405keep only the first 1000 samples
406@example
407avconv -i INPUT -af atrim=end_sample=1000
408@end example
409
410@end itemize
411
d6251368
AK
412@section channelsplit
413Split each channel in input audio stream into a separate output stream.
414
415This filter accepts the following named parameters:
416@table @option
417@item channel_layout
418Channel layout of the input stream. Default is "stereo".
419@end table
420
421For example, assuming a stereo input MP3 file
422@example
423avconv -i in.mp3 -filter_complex channelsplit out.mkv
424@end example
425will create an output Matroska file with two audio streams, one containing only
426the left channel and the other the right channel.
427
428To split a 5.1 WAV file into per-channel files
429@example
430avconv -i in.wav -filter_complex
431'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
432-map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
433front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
434side_right.wav
435@end example
436
41e637e4
AC
437@section channelmap
438Remap input channels to new locations.
439
440This filter accepts the following named parameters:
441@table @option
442@item channel_layout
443Channel layout of the output stream.
444
445@item map
ba8efac9 446Map channels from input to output. The argument is a '|'-separated list of
41e637e4
AC
447mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
448@var{in_channel} form. @var{in_channel} can be either the name of the input
449channel (e.g. FL for front left) or its index in the input channel layout.
450@var{out_channel} is the name of the output channel or its index in the output
451channel layout. If @var{out_channel} is not given then it is implicitly an
452index, starting with zero and increasing by one for each mapping.
453@end table
454
455If no mapping is present, the filter will implicitly map input channels to
456output channels preserving index.
457
458For example, assuming a 5.1+downmix input MOV file
459@example
ba8efac9 460avconv -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
41e637e4
AC
461@end example
462will create an output WAV file tagged as stereo from the downmix channels of
463the input.
464
465To fix a 5.1 WAV improperly encoded in AAC's native channel order
466@example
ba8efac9 467avconv -i in.wav -filter 'channelmap=1|2|0|5|3|4:channel_layout=5.1' out.wav
41e637e4
AC
468@end example
469
738f8358
AK
470@section compand
471Compress or expand audio dynamic range.
472
473A description of the accepted options follows.
474
475@table @option
476
477@item attacks
478@item decays
479Set list of times in seconds for each channel over which the instantaneous level
480of the input signal is averaged to determine its volume. @var{attacks} refers to
481increase of volume and @var{decays} refers to decrease of volume. For most
482situations, the attack time (response to the audio getting louder) should be
483shorter than the decay time because the human ear is more sensitive to sudden
484loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
485a typical value for decay is 0.8 seconds.
486
487@item points
488Set list of points for the transfer function, specified in dB relative to the
489maximum possible signal amplitude. Each key points list must be defined using
490the following syntax: @code{x0/y0|x1/y1|x2/y2|....}
491
492The input values must be in strictly increasing order but the transfer function
493does not have to be monotonically rising. The point @code{0/0} is assumed but
494may be overridden (by @code{0/out-dBn}). Typical values for the transfer
495function are @code{-70/-70|-60/-20}.
496
497@item soft-knee
498Set the curve radius in dB for all joints. Defaults to 0.01.
499
500@item gain
501Set additional gain in dB to be applied at all points on the transfer function.
502This allows easy adjustment of the overall gain. Defaults to 0.
503
504@item volume
505Set initial volume in dB to be assumed for each channel when filtering starts.
506This permits the user to supply a nominal level initially, so that, for
507example, a very large gain is not applied to initial signal levels before the
508companding has begun to operate. A typical value for audio which is initially
509quiet is -90 dB. Defaults to 0.
510
511@item delay
512Set delay in seconds. The input audio is analyzed immediately, but audio is
513delayed before being fed to the volume adjuster. Specifying a delay
514approximately equal to the attack/decay times allows the filter to effectively
515operate in predictive rather than reactive mode. Defaults to 0.
516
517@end table
518
519@subsection Examples
520
521@itemize
522@item
523Make music with both quiet and loud passages suitable for listening in a noisy
524environment:
525@example
526compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
527@end example
528
529@item
530Noise gate for when the noise is at a lower level than the signal:
531@example
532compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
533@end example
534
535@item
536Here is another noise gate, this time for when the noise is at a higher level
537than the signal (making it, in some ways, similar to squelch):
538@example
539compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
540@end example
541@end itemize
542
dc07fb6f
AK
543@section join
544Join multiple input streams into one multi-channel stream.
545
546The filter accepts the following named parameters:
547@table @option
548
549@item inputs
550Number of input streams. Defaults to 2.
551
552@item channel_layout
553Desired output channel layout. Defaults to stereo.
554
555@item map
dd7fc37c 556Map channels from inputs to output. The argument is a '|'-separated list of
dc07fb6f
AK
557mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
558form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
7a8059eb 559can be either the name of the input channel (e.g. FL for front left) or its
dc07fb6f
AK
560index in the specified input stream. @var{out_channel} is the name of the output
561channel.
562@end table
563
564The filter will attempt to guess the mappings when those are not specified
565explicitly. It does so by first trying to find an unused matching input channel
566and if that fails it picks the first unused input channel.
567
568E.g. to join 3 inputs (with properly set channel layouts)
569@example
570avconv -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
571@end example
572
573To build a 5.1 output from 6 single-channel streams:
574@example
575avconv -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
dd7fc37c 576'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'
dc07fb6f
AK
577out
578@end example
579
d371e7b9
AK
580@section resample
581Convert the audio sample format, sample rate and channel layout. This filter is
582not meant to be used directly, it is inserted automatically by libavfilter
583whenever conversion is needed. Use the @var{aformat} filter to force a specific
584conversion.
585
b384e031
JR
586@section volume
587
588Adjust the input audio volume.
589
590The filter accepts the following named parameters:
591@table @option
592
593@item volume
594Expresses how the audio volume will be increased or decreased.
595
596Output values are clipped to the maximum value.
597
598The output audio volume is given by the relation:
599@example
600@var{output_volume} = @var{volume} * @var{input_volume}
601@end example
602
603Default value for @var{volume} is 1.0.
604
605@item precision
606Mathematical precision.
607
608This determines which input sample formats will be allowed, which affects the
609precision of the volume scaling.
610
611@table @option
612@item fixed
6138-bit fixed-point; limits input sample format to U8, S16, and S32.
614@item float
61532-bit floating-point; limits input sample format to FLT. (default)
616@item double
61764-bit floating-point; limits input sample format to DBL.
618@end table
619@end table
620
621@subsection Examples
622
623@itemize
624@item
625Halve the input audio volume:
626@example
627volume=volume=0.5
628volume=volume=1/2
629volume=volume=-6.0206dB
630@end example
631
632@item
633Increase input audio power by 6 decibels using fixed-point precision:
634@example
635volume=volume=6dB:precision=fixed
636@end example
637@end itemize
638
f59e9eaf
SS
639@c man end AUDIO FILTERS
640
1ee410f3
SS
641@chapter Audio Sources
642@c man begin AUDIO SOURCES
643
644Below is a description of the currently available audio sources.
645
646@section anullsrc
647
648Null audio source, never return audio frames. It is mainly useful as a
649template and to be employed in analysis / debugging tools.
650
651It accepts as optional parameter a string of the form
652@var{sample_rate}:@var{channel_layout}.
653
654@var{sample_rate} specify the sample rate, and defaults to 44100.
655
656@var{channel_layout} specify the channel layout, and can be either an
657integer or a string representing a channel layout. The default value
658of @var{channel_layout} is 3, which corresponds to CH_LAYOUT_STEREO.
659
660Check the channel_layout_map definition in
a903f8f0 661@file{libavutil/channel_layout.c} for the mapping between strings and
6ef93402 662channel layout values.
1ee410f3
SS
663
664Follow some examples:
665@example
666# set the sample rate to 48000 Hz and the channel layout to CH_LAYOUT_MONO.
667anullsrc=48000:4
668
669# same as
670anullsrc=48000:mono
671@end example
672
4c66c407
AK
673@section abuffer
674Buffer audio frames, and make them available to the filter chain.
675
676This source is not intended to be part of user-supplied graph descriptions but
677for insertion by calling programs through the interface defined in
678@file{libavfilter/buffersrc.h}.
679
680It accepts the following named parameters:
681@table @option
682
683@item time_base
684Timebase which will be used for timestamps of submitted frames. It must be
685either a floating-point number or in @var{numerator}/@var{denominator} form.
686
687@item sample_rate
688Audio sample rate.
689
690@item sample_fmt
691Name of the sample format, as returned by @code{av_get_sample_fmt_name()}.
692
693@item channel_layout
694Channel layout of the audio data, in the form that can be accepted by
695@code{av_get_channel_layout()}.
696@end table
697
698All the parameters need to be explicitly defined.
699
1ee410f3
SS
700@c man end AUDIO SOURCES
701
f0a55438
HM
702@chapter Audio Sinks
703@c man begin AUDIO SINKS
704
705Below is a description of the currently available audio sinks.
706
707@section anullsink
708
709Null audio sink, do absolutely nothing with the input audio. It is
710mainly useful as a template and to be employed in analysis / debugging
711tools.
712
a2cd9be2
AK
713@section abuffersink
714This sink is intended for programmatic use. Frames that arrive on this sink can
715be retrieved by the calling program using the interface defined in
716@file{libavfilter/buffersink.h}.
717
718This filter accepts no parameters.
719
f0a55438
HM
720@c man end AUDIO SINKS
721
3275ac6a
SS
722@chapter Video Filters
723@c man begin VIDEO FILTERS
724
f8a45fa1 725When you configure your Libav build, you can disable any of the
3275ac6a
SS
726existing filters using --disable-filters.
727The configure output will show the video filters included in your
728build.
729
730Below is a description of the currently available video filters.
731
13fabd7a
SS
732@section blackframe
733
734Detect frames that are (almost) completely black. Can be useful to
735detect chapter transitions or commercials. Output lines consist of
736the frame number of the detected frame, the percentage of blackness,
737the position in the file if known or -1 and the timestamp in seconds.
738
739In order to display the output lines, you need to set the loglevel at
740least to the AV_LOG_INFO value.
741
62dcdb02
AK
742The filter accepts the following options:
743
744@table @option
745
746@item amount
747The percentage of the pixels that have to be below the threshold, defaults to
74898.
13fabd7a 749
62dcdb02
AK
750@item threshold
751Threshold below which a pixel value is considered black, defaults to 32.
13fabd7a 752
62dcdb02 753@end table
13fabd7a 754
ce6b6ef6
SS
755@section boxblur
756
757Apply boxblur algorithm to the input video.
758
51def31d
AK
759This filter accepts the following options:
760
761@table @option
762
763@item luma_radius
764@item luma_power
765@item chroma_radius
766@item chroma_power
767@item alpha_radius
768@item alpha_power
769
770@end table
ce6b6ef6
SS
771
772Chroma and alpha parameters are optional, if not specified they default
773to the corresponding values set for @var{luma_radius} and
774@var{luma_power}.
775
776@var{luma_radius}, @var{chroma_radius}, and @var{alpha_radius} represent
777the radius in pixels of the box used for blurring the corresponding
778input plane. They are expressions, and can contain the following
779constants:
780@table @option
781@item w, h
da9cea77 782the input width and height in pixels
ce6b6ef6
SS
783
784@item cw, ch
785the input chroma image width and height in pixels
786
787@item hsub, vsub
788horizontal and vertical chroma subsample values. For example for the
789pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
790@end table
791
792The radius must be a non-negative number, and must not be greater than
793the value of the expression @code{min(w,h)/2} for the luma and alpha planes,
794and of @code{min(cw,ch)/2} for the chroma planes.
795
796@var{luma_power}, @var{chroma_power}, and @var{alpha_power} represent
797how many times the boxblur filter is applied to the corresponding
798plane.
799
800Some examples follow:
801
802@itemize
803
804@item
805Apply a boxblur filter with luma, chroma, and alpha radius
806set to 2:
807@example
51def31d 808boxblur=luma_radius=2:luma_power=1
ce6b6ef6
SS
809@end example
810
811@item
812Set luma radius to 2, alpha and chroma radius to 0
813@example
814boxblur=2:1:0:0:0:0
815@end example
816
817@item
818Set luma and chroma radius to a fraction of the video dimension
819@example
51def31d 820boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
ce6b6ef6
SS
821@end example
822
823@end itemize
824
b5670209
SS
825@section copy
826
827Copy the input source unchanged to the output. Mainly useful for
828testing purposes.
829
3275ac6a
SS
830@section crop
831
fba0156a
AK
832Crop the input video to given dimensions.
833
834This filter accepts the following options:
835
836@table @option
837
838@item out_w
839Width of the output video.
840
841@item out_h
842Height of the output video.
843
844@item x
845Horizontal position, in the input video, of the left edge of the output video.
846
847@item y
848Vertical position, in the input video, of the top edge of the output video.
849
850@end table
3275ac6a 851
75b67a8a
SS
852The parameters are expressions containing the following constants:
853
854@table @option
855@item E, PI, PHI
856the corresponding mathematical approximated values for e
857(euler number), pi (greek PI), PHI (golden ratio)
858
859@item x, y
860the computed values for @var{x} and @var{y}. They are evaluated for
861each new frame.
862
863@item in_w, in_h
e83c2dde 864the input width and height
75b67a8a
SS
865
866@item iw, ih
867same as @var{in_w} and @var{in_h}
868
869@item out_w, out_h
e83c2dde 870the output (cropped) width and height
75b67a8a
SS
871
872@item ow, oh
873same as @var{out_w} and @var{out_h}
874
875@item n
876the number of input frame, starting from 0
877
75b67a8a
SS
878@item t
879timestamp expressed in seconds, NAN if the input timestamp is unknown
3275ac6a 880
75b67a8a 881@end table
3275ac6a 882
75b67a8a
SS
883The @var{out_w} and @var{out_h} parameters specify the expressions for
884the width and height of the output (cropped) video. They are
885evaluated just at the configuration of the filter.
3275ac6a 886
75b67a8a
SS
887The default value of @var{out_w} is "in_w", and the default value of
888@var{out_h} is "in_h".
2bc05d35 889
75b67a8a
SS
890The expression for @var{out_w} may depend on the value of @var{out_h},
891and the expression for @var{out_h} may depend on @var{out_w}, but they
892cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
893evaluated after @var{out_w} and @var{out_h}.
2bc05d35 894
75b67a8a
SS
895The @var{x} and @var{y} parameters specify the expressions for the
896position of the top-left corner of the output (non-cropped) area. They
897are evaluated for each frame. If the evaluated value is not valid, it
898is approximated to the nearest valid value.
3275ac6a 899
75b67a8a
SS
900The default value of @var{x} is "(in_w-out_w)/2", and the default
901value for @var{y} is "(in_h-out_h)/2", which set the cropped area at
902the center of the input image.
903
904The expression for @var{x} may depend on @var{y}, and the expression
905for @var{y} may depend on @var{x}.
906
907Follow some examples:
3275ac6a 908@example
75b67a8a 909# crop the central input area with size 100x100
fba0156a 910crop=out_w=100:out_h=100
75b67a8a
SS
911
912# crop the central input area with size 2/3 of the input video
fba0156a 913"crop=out_w=2/3*in_w:out_h=2/3*in_h"
75b67a8a
SS
914
915# crop the input video central square
fba0156a 916crop=out_w=in_h
75b67a8a
SS
917
918# delimit the rectangle with the top-left corner placed at position
919# 100:100 and the right-bottom corner corresponding to the right-bottom
920# corner of the input image.
fba0156a 921crop=out_w=in_w-100:out_h=in_h-100:x=100:y=100
3275ac6a 922
e6dba1d8 923# crop 10 pixels from the left and right borders, and 20 pixels from
75b67a8a 924# the top and bottom borders
fba0156a 925"crop=out_w=in_w-2*10:out_h=in_h-2*20"
3275ac6a 926
75b67a8a 927# keep only the bottom right quarter of the input image
fba0156a 928"crop=out_w=in_w/2:out_h=in_h/2:x=in_w/2:y=in_h/2"
75b67a8a
SS
929
930# crop height for getting Greek harmony
fba0156a 931"crop=out_w=in_w:out_h=1/PHI*in_w"
75b67a8a
SS
932
933# trembling effect
934"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)"
935
e6dba1d8 936# erratic camera effect depending on timestamp
fba0156a 937"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)"
75b67a8a
SS
938
939# set x depending on the value of y
940"crop=in_w/2:in_h/2:y:10+10*sin(n/10)"
941@end example
3275ac6a 942
68b79bfc
SS
943@section cropdetect
944
945Auto-detect crop size.
946
947Calculate necessary cropping parameters and prints the recommended
948parameters through the logging system. The detected dimensions
949correspond to the non-black area of the input video.
950
460e7b4f 951This filter accepts the following options:
68b79bfc
SS
952
953@table @option
954
955@item limit
956Threshold, which can be optionally specified from nothing (0) to
957everything (255), defaults to 24.
958
959@item round
960Value which the width/height should be divisible by, defaults to
96116. The offset is automatically adjusted to center the video. Use 2 to
962get only even dimensions (needed for 4:2:2 video). 16 is best when
963encoding to most video codecs.
964
965@item reset
966Counter that determines after how many frames cropdetect will reset
967the previously detected largest video area and start over to detect
968the current optimal crop area. Defaults to 0.
969
970This can be useful when channel logos distort the video area. 0
971indicates never reset and return the largest area encountered during
972playback.
973@end table
974
b157be1f
SS
975@section delogo
976
977Suppress a TV station logo by a simple interpolation of the surrounding
978pixels. Just set a rectangle covering the logo and watch it disappear
979(and sometimes something even uglier appear - your mileage may vary).
980
63e58c55 981This filter accepts the following options:
b157be1f
SS
982@table @option
983
984@item x, y
985Specify the top left corner coordinates of the logo. They must be
986specified.
987
988@item w, h
989Specify the width and height of the logo to clear. They must be
990specified.
991
992@item band, t
993Specify the thickness of the fuzzy edge of the rectangle (added to
994@var{w} and @var{h}). The default value is 4.
995
996@item show
997When set to 1, a green rectangle is drawn on the screen to simplify
998finding the right @var{x}, @var{y}, @var{w}, @var{h} parameters, and
999@var{band} is set to 4. The default value is 0.
1000
1001@end table
1002
1003Some examples follow.
1004
1005@itemize
1006
1007@item
1008Set a rectangle covering the area with top left corner coordinates 0,0
1009and size 100x77, setting a band of size 10:
1010@example
b157be1f
SS
1011delogo=x=0:y=0:w=100:h=77:band=10
1012@end example
1013
1014@end itemize
1015
e40032e2
SS
1016@section drawbox
1017
1018Draw a colored box on the input image.
1019
335c3129 1020This filter accepts the following options:
e40032e2
SS
1021
1022@table @option
1023
1024@item x, y
1025Specify the top left corner coordinates of the box. Default to 0.
1026
1027@item width, height
1028Specify the width and height of the box, if 0 they are interpreted as
1029the input width and height. Default to 0.
1030
1031@item color
1032Specify the color of the box to write, it can be the name of a color
1033(case insensitive match) or a 0xRRGGBB[AA] sequence.
1034@end table
1035
1036Follow some examples:
1037@example
1038# draw a black box around the edge of the input image
1039drawbox
1040
1041# draw a box with color red and an opacity of 50%
335c3129 1042drawbox=x=10:y=20:width=200:height=60:color=red@@0.5"
e40032e2
SS
1043@end example
1044
a5b64584
SS
1045@section drawtext
1046
1047Draw text string or text from specified file on top of video using the
1048libfreetype library.
1049
0c50edb7 1050To enable compilation of this filter you need to configure Libav with
a5b64584
SS
1051@code{--enable-libfreetype}.
1052
1053The filter also recognizes strftime() sequences in the provided text
1054and expands them accordingly. Check the documentation of strftime().
1055
a5b64584
SS
1056The description of the accepted parameters follows.
1057
1058@table @option
1059
1060@item fontfile
1061The font file to be used for drawing text. Path must be included.
1062This parameter is mandatory.
1063
1064@item text
1065The text string to be drawn. The text must be a sequence of UTF-8
1066encoded characters.
1067This parameter is mandatory if no file is specified with the parameter
1068@var{textfile}.
1069
1070@item textfile
1071A text file containing text to be drawn. The text must be a sequence
1072of UTF-8 encoded characters.
1073
1074This parameter is mandatory if no text string is specified with the
1075parameter @var{text}.
1076
1077If both text and textfile are specified, an error is thrown.
1078
1079@item x, y
1080The offsets where text will be drawn within the video frame.
1081Relative to the top/left border of the output image.
2cf74eca
LB
1082They accept expressions similar to the @ref{overlay} filter:
1083@table @option
1084
1085@item x, y
1086the computed values for @var{x} and @var{y}. They are evaluated for
1087each new frame.
1088
1089@item main_w, main_h
1090main input width and height
1091
1092@item W, H
1093same as @var{main_w} and @var{main_h}
1094
1095@item text_w, text_h
1096rendered text width and height
1097
1098@item w, h
1099same as @var{text_w} and @var{text_h}
1100
1101@item n
1102the number of frames processed, starting from 0
1103
1104@item t
1105timestamp expressed in seconds, NAN if the input timestamp is unknown
1106
1107@end table
a5b64584
SS
1108
1109The default value of @var{x} and @var{y} is 0.
1110
1111@item fontsize
1112The font size to be used for drawing text.
1113The default value of @var{fontsize} is 16.
1114
1115@item fontcolor
1116The color to be used for drawing fonts.
1117Either a string (e.g. "red") or in 0xRRGGBB[AA] format
1118(e.g. "0xff000033"), possibly followed by an alpha specifier.
1119The default value of @var{fontcolor} is "black".
1120
1121@item boxcolor
1122The color to be used for drawing box around text.
1123Either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
1124(e.g. "0xff00ff"), possibly followed by an alpha specifier.
1125The default value of @var{boxcolor} is "white".
1126
1127@item box
1128Used to draw a box around text using background color.
1129Value should be either 1 (enable) or 0 (disable).
1130The default value of @var{box} is 0.
1131
994de197
SS
1132@item shadowx, shadowy
1133The x and y offsets for the text shadow position with respect to the
1134position of the text. They can be either positive or negative
1135values. Default value for both is "0".
1136
1137@item shadowcolor
1138The color to be used for drawing a shadow behind the drawn text. It
1139can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
1140form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
1141The default value of @var{shadowcolor} is "black".
1142
a5b64584
SS
1143@item ft_load_flags
1144Flags to be used for loading the fonts.
1145
1146The flags map the corresponding flags supported by libfreetype, and are
1147a combination of the following values:
1148@table @var
1149@item default
1150@item no_scale
1151@item no_hinting
1152@item render
1153@item no_bitmap
1154@item vertical_layout
1155@item force_autohint
1156@item crop_bitmap
1157@item pedantic
1158@item ignore_global_advance_width
1159@item no_recurse
1160@item ignore_transform
1161@item monochrome
1162@item linear_design
1163@item no_autohint
1164@item end table
1165@end table
1166
1167Default value is "render".
1168
1169For more information consult the documentation for the FT_LOAD_*
1170libfreetype flags.
1171
1172@item tabsize
1173The size in number of spaces to use for rendering the tab.
1174Default value is 4.
e496c45d
AU
1175
1176@item fix_bounds
1177If true, check and fix text coords to avoid clipping.
a5b64584
SS
1178@end table
1179
1180For example the command:
1181@example
1182drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
1183@end example
1184
1185will draw "Test Text" with font FreeSerif, using the default values
1186for the optional parameters.
1187
1188The command:
1189@example
1190drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
1191 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
1192@end example
1193
1194will draw 'Test Text' with font FreeSerif of size 24 at position x=100
1195and y=50 (counting from the top-left corner of the screen), text is
1196yellow with a red box around it. Both the text and the box have an
1197opacity of 20%.
1198
1199Note that the double quotes are not necessary if spaces are not used
1200within the parameter list.
1201
1202For more information about libfreetype, check:
1203@url{http://www.freetype.org/}.
1204
aadfc9ee
BM
1205@section fade
1206
1207Apply fade-in/out effect to input video.
1208
b9dfee9f
AK
1209This filter accepts the following options:
1210
1211@table @option
aadfc9ee 1212
b9dfee9f
AK
1213@item type
1214The effect type -- can be either "in" for fade-in, or "out" for a fade-out
1215effect.
aadfc9ee 1216
b9dfee9f
AK
1217@item start_frame
1218The number of the start frame for starting to apply the fade effect.
aadfc9ee 1219
b9dfee9f
AK
1220@item nb_frames
1221The number of frames for which the fade effect has to last. At the end of the
1222fade-in effect the output video will have the same intensity as the input video,
1223at the end of the fade-out transition the output video will be completely black.
1224
1225@end table
aadfc9ee
BM
1226
1227A few usage examples follow, usable too as test scenarios.
1228@example
1229# fade in first 30 frames of video
b9dfee9f 1230fade=type=in:nb_frames=30
aadfc9ee
BM
1231
1232# fade out last 45 frames of a 200-frame video
b9dfee9f 1233fade=type=out:start_frame=155:nb_frames=45
aadfc9ee
BM
1234
1235# fade in first 25 frames and fade out last 25 frames of a 1000-frame video
b9dfee9f 1236fade=type=in:start_frame=0:nb_frames=25, fade=type=out:start_frame=975:nb_frames=25
aadfc9ee
BM
1237
1238# make first 5 frames black, then fade in from frame 5-24
b9dfee9f 1239fade=type=in:start_frame=5:nb_frames=20
aadfc9ee
BM
1240@end example
1241
2f84bb42
MH
1242@section fieldorder
1243
1244Transform the field order of the input video.
1245
a39c1540 1246This filter accepts the following options:
2f84bb42
MH
1247
1248@table @option
a39c1540
AK
1249
1250@item order
1251Output field order. Valid values are @var{tff} for top field first or @var{bff}
1252for bottom field first.
2f84bb42
MH
1253@end table
1254
1255Default value is "tff".
1256
1257Transformation is achieved by shifting the picture content up or down
1258by one line, and filling the remaining line with appropriate picture content.
1259This method is consistent with most broadcast field order converters.
1260
1261If the input video is not flagged as being interlaced, or it is already
1262flagged as being of the required output field order then this filter does
1263not alter the incoming video.
1264
1265This filter is very useful when converting to or from PAL DV material,
1266which is bottom field first.
1267
1268For example:
1269@example
a39c1540 1270./avconv -i in.vob -vf "fieldorder=order=bff" out.dv
2f84bb42
MH
1271@end example
1272
7f1af825
SS
1273@section fifo
1274
1275Buffer input images and send them when they are requested.
1276
1277This filter is mainly useful when auto-inserted by the libavfilter
1278framework.
1279
1280The filter does not take parameters.
1281
3275ac6a
SS
1282@section format
1283
1284Convert the input video to one of the specified pixel formats.
1285Libavfilter will try to pick one that is supported for the input to
1286the next filter.
1287
e67a87ea
AK
1288This filter accepts the following parameters:
1289@table @option
1290
1291@item pix_fmts
1292A '|'-separated list of pixel format names, for example
1293"pix_fmts=yuv420p|monow|rgb24".
1294
1295@end table
3275ac6a 1296
f150e4dc 1297Some examples follow:
3275ac6a 1298@example
f150e4dc 1299# convert the input video to the format "yuv420p"
e67a87ea 1300format=pix_fmts=yuv420p
3275ac6a 1301
f150e4dc 1302# convert the input video to any of the formats in the list
e67a87ea 1303format=pix_fmts=yuv420p|yuv444p|yuv410p
f150e4dc 1304@end example
3275ac6a 1305
dd249245 1306@anchor{fps}
54c5dd89
AK
1307@section fps
1308
1309Convert the video to specified constant framerate by duplicating or dropping
1310frames as necessary.
1311
1312This filter accepts the following named parameters:
1313@table @option
1314
1315@item fps
1316Desired output framerate.
1317
545a0b80
JR
1318@item start_time
1319Assume the first PTS should be the given value, in seconds. This allows for
1320padding/trimming at the start of stream. By default, no assumption is made
1321about the first frame's expected PTS, so no padding or trimming is done.
1322For example, this could be set to 0 to pad the beginning with duplicates of
1323the first frame if a video stream starts after the audio stream or to trim any
1324frames with a negative PTS.
1325
54c5dd89
AK
1326@end table
1327
a7d0e7ea
VG
1328@section framepack
1329
1330Pack two different video streams into a stereoscopic video, setting proper
1331metadata on supported codecs. The two views should have the same size and
1332framerate and processing will stop when the shorter video ends. Please note
1333that you may conveniently adjust view properties with the @ref{scale} and
1334@ref{fps} filters.
1335
1336This filter accepts the following named parameters:
1337@table @option
1338
1339@item format
1340Desired packing format. Supported values are:
1341
1342@table @option
1343
1344@item sbs
1345Views are next to each other (default).
1346
1347@item tab
1348Views are on top of each other.
1349
1350@item lines
1351Views are packed by line.
1352
1353@item columns
1354Views are eacked by column.
1355
1356@item frameseq
1357Views are temporally interleaved.
1358
1359@end table
1360
1361@end table
1362
1363Some examples follow:
1364
1365@example
1366# Convert left and right views into a frame sequential video.
1367avconv -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
1368
1369# Convert views into a side-by-side video with the same output resolution as the input.
1370avconv -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
1371@end example
1372
f8608dca 1373@anchor{frei0r}
47941088
SS
1374@section frei0r
1375
1376Apply a frei0r effect to the input video.
1377
1378To enable compilation of this filter you need to install the frei0r
f8a45fa1 1379header and configure Libav with --enable-frei0r.
47941088 1380
5aa1a668
AK
1381This filter accepts the following options:
1382
1383@table @option
1384
1385@item filter_name
1386The name to the frei0r effect to load. If the environment variable
1387@env{FREI0R_PATH} is defined, the frei0r effect is searched in each one of the
1388directories specified by the colon separated list in @env{FREIOR_PATH},
1389otherwise in the standard frei0r paths, which are in this order:
1390@file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
1391@file{/usr/lib/frei0r-1/}.
47941088 1392
5aa1a668
AK
1393@item filter_params
1394A '|'-separated list of parameters to pass to the frei0r effect.
47941088 1395
5aa1a668 1396@end table
47941088
SS
1397
1398A frei0r effect parameter can be a boolean (whose values are specified
1399with "y" and "n"), a double, a color (specified by the syntax
1400@var{R}/@var{G}/@var{B}, @var{R}, @var{G}, and @var{B} being float
1401numbers from 0.0 to 1.0) or by an @code{av_parse_color()} color
1402description), a position (specified by the syntax @var{X}/@var{Y},
1403@var{X} and @var{Y} being float numbers) and a string.
1404
1405The number and kind of parameters depend on the loaded effect. If an
1406effect parameter is not specified the default value is set.
1407
1408Some examples follow:
1409@example
1410# apply the distort0r effect, set the first two double parameters
5aa1a668 1411frei0r=filter_name=distort0r:filter_params=0.5|0.01
47941088
SS
1412
1413# apply the colordistance effect, takes a color as first parameter
1414frei0r=colordistance:0.2/0.3/0.4
1415frei0r=colordistance:violet
1416frei0r=colordistance:0x112233
1417
1418# apply the perspective effect, specify the top left and top right
1419# image positions
5aa1a668 1420frei0r=perspective:0.2/0.2|0.8/0.2
47941088
SS
1421@end example
1422
1423For more information see:
1424@url{http://piksel.org/frei0r}
1425
d5f187fd
N
1426@section gradfun
1427
1428Fix the banding artifacts that are sometimes introduced into nearly flat
1429regions by truncation to 8bit colordepth.
1430Interpolate the gradients that should go where the bands are, and
1431dither them.
1432
3eccfaa0
SS
1433This filter is designed for playback only. Do not use it prior to
1434lossy compression, because compression tends to lose the dither and
1435bring back the bands.
1436
7ed833d7
AK
1437This filter accepts the following options:
1438
1439@table @option
d5f187fd 1440
7ed833d7
AK
1441@item strength
1442The maximum amount by which the filter will change any one pixel. Also the
1443threshold for detecting nearly flat regions. Acceptable values range from .51 to
144464, default value is 1.2, out-of-range values will be clipped to the valid
1445range.
d5f187fd 1446
7ed833d7
AK
1447@item radius
1448The neighborhood to fit the gradient to. A larger radius makes for smoother
1449gradients, but also prevents the filter from modifying the pixels near detailed
1450regions. Acceptable values are 8-32, default value is 16, out-of-range values
1451will be clipped to the valid range.
1452
1453@end table
d5f187fd
N
1454
1455@example
1456# default parameters
7ed833d7 1457gradfun=strength=1.2:radius=16
d5f187fd
N
1458
1459# omitting radius
1460gradfun=1.2
1461@end example
1462
a1e171df
SS
1463@section hflip
1464
1465Flip the input video horizontally.
1466
9270b8a3 1467For example to horizontally flip the input video with @command{avconv}:
a1e171df 1468@example
9270b8a3 1469avconv -i in.avi -vf "hflip" out.avi
a1e171df
SS
1470@end example
1471
a4dc7aa5
BC
1472@section hqdn3d
1473
1474High precision/quality 3d denoise filter. This filter aims to reduce
1475image noise producing smooth images and making still images really
1476still. It should enhance compressibility.
1477
1478It accepts the following optional parameters:
a4dc7aa5
BC
1479
1480@table @option
1481@item luma_spatial
1482a non-negative float number which specifies spatial luma strength,
1483defaults to 4.0
1484
1485@item chroma_spatial
1486a non-negative float number which specifies spatial chroma strength,
1487defaults to 3.0*@var{luma_spatial}/4.0
1488
1489@item luma_tmp
1490a float number which specifies luma temporal strength, defaults to
14916.0*@var{luma_spatial}/4.0
1492
1493@item chroma_tmp
1494a float number which specifies chroma temporal strength, defaults to
1495@var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}
1496@end table
1497
3fce1367
VG
1498@section interlace
1499
1500Simple interlacing filter from progressive contents. This interleaves upper (or
1501lower) lines from odd frames with lower (or upper) lines from even frames,
35b05c51 1502halving the frame rate and preserving image height. A vertical lowpass filter
c91488ab 1503is always applied in order to avoid twitter effects and reduce moiré patterns.
3fce1367
VG
1504
1505@example
1506 Original Original New Frame
1507 Frame 'j' Frame 'j+1' (tff)
1508 ========== =========== ==================
1509 Line 0 --------------------> Frame 'j' Line 0
1510 Line 1 Line 1 ----> Frame 'j+1' Line 1
1511 Line 2 ---------------------> Frame 'j' Line 2
1512 Line 3 Line 3 ----> Frame 'j+1' Line 3
1513 ... ... ...
1514New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
1515@end example
1516
1517It accepts the following optional parameters:
1518
1519@table @option
1520@item scan
1521determines whether the interlaced frame is taken from the even (tff - default)
1522or odd (bff) lines of the progressive frame.
3fce1367
VG
1523@end table
1524
8fe0c527
SS
1525@section lut, lutrgb, lutyuv
1526
1527Compute a look-up table for binding each pixel component input value
1528to an output value, and apply it to input video.
1529
1530@var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
1531to an RGB input video.
1532
20b46f8f 1533These filters accept the following options:
8fe0c527 1534@table @option
91f5f875
LB
1535@item @var{c0} (first pixel component)
1536@item @var{c1} (second pixel component)
1537@item @var{c2} (third pixel component)
1538@item @var{c3} (fourth pixel component, corresponds to the alpha component)
8fe0c527 1539
91f5f875
LB
1540@item @var{r} (red component)
1541@item @var{g} (green component)
1542@item @var{b} (blue component)
1543@item @var{a} (alpha component)
8fe0c527 1544
91f5f875
LB
1545@item @var{y} (Y/luminance component)
1546@item @var{u} (U/Cb component)
1547@item @var{v} (V/Cr component)
8fe0c527
SS
1548@end table
1549
20b46f8f
AK
1550Each of them specifies the expression to use for computing the lookup table for
1551the corresponding pixel component values.
1552
1553The exact component associated to each of the @var{c*} options depends on the
1554format in input.
1555
1556The @var{lut} filter requires either YUV or RGB pixel formats in input,
1557@var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
1558
8fe0c527
SS
1559The expressions can contain the following constants and functions:
1560
1561@table @option
1562@item E, PI, PHI
1563the corresponding mathematical approximated values for e
1564(euler number), pi (greek PI), PHI (golden ratio)
1565
1566@item w, h
da9cea77 1567the input width and height
8fe0c527
SS
1568
1569@item val
1570input value for the pixel component
1571
1572@item clipval
1573the input value clipped in the @var{minval}-@var{maxval} range
1574
1575@item maxval
1576maximum value for the pixel component
1577
1578@item minval
1579minimum value for the pixel component
1580
1581@item negval
1582the negated value for the pixel component value clipped in the
1583@var{minval}-@var{maxval} range , it corresponds to the expression
1584"maxval-clipval+minval"
1585
1586@item clip(val)
1587the computed value in @var{val} clipped in the
1588@var{minval}-@var{maxval} range
1589
1590@item gammaval(gamma)
1591the computed gamma correction value of the pixel component value
1592clipped in the @var{minval}-@var{maxval} range, corresponds to the
1593expression
1594"pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
1595
1596@end table
1597
1598All expressions default to "val".
1599
1600Some examples follow:
1601@example
1602# negate input video
1603lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
1604lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
1605
1606# the above is the same as
1607lutrgb="r=negval:g=negval:b=negval"
1608lutyuv="y=negval:u=negval:v=negval"
1609
1610# negate luminance
1611lutyuv=negval
1612
1613# remove chroma components, turns the video into a graytone image
1614lutyuv="u=128:v=128"
1615
1616# apply a luma burning effect
1617lutyuv="y=2*val"
1618
1619# remove green and blue components
1620lutrgb="g=0:b=0"
1621
1622# set a constant alpha channel value on input
1623format=rgba,lutrgb=a="maxval-minval/2"
1624
1625# correct luminance gamma by a 0.5 factor
1626lutyuv=y=gammaval(0.5)
1627@end example
1628
171868e2
SS
1629@section negate
1630
1631Negate input video.
1632
1633This filter accepts an integer in input, if non-zero it negates the
1634alpha component (if available). The default value in input is 0.
3275ac6a 1635
ef4d34aa
AK
1636@section noformat
1637
3275ac6a
SS
1638Force libavfilter not to use any of the specified pixel formats for the
1639input to the next filter.
1640
e67a87ea
AK
1641This filter accepts the following parameters:
1642@table @option
1643
1644@item pix_fmts
1645A '|'-separated list of pixel format names, for example
1646"pix_fmts=yuv420p|monow|rgb24".
1647
1648@end table
3275ac6a 1649
f150e4dc 1650Some examples follow:
3275ac6a 1651@example
f150e4dc
SS
1652# force libavfilter to use a format different from "yuv420p" for the
1653# input to the vflip filter
e67a87ea 1654noformat=pix_fmts=yuv420p,vflip
3275ac6a 1655
f150e4dc 1656# convert the input video to any of the formats not contained in the list
e67a87ea 1657noformat=yuv420p|yuv444p|yuv410p
f150e4dc 1658@end example
3275ac6a
SS
1659
1660@section null
1661
99046339 1662Pass the video source unchanged to the output.
3275ac6a 1663
cf69ad35 1664@section ocv
6ebf0bfc 1665
cf69ad35 1666Apply video transform using libopencv.
6ebf0bfc
SS
1667
1668To enable this filter install libopencv library and headers and
f8a45fa1 1669configure Libav with --enable-libopencv.
6ebf0bfc 1670
ee0e8d4b
AK
1671This filter accepts the following parameters:
1672
1673@table @option
cf69ad35 1674
ee0e8d4b
AK
1675@item filter_name
1676The name of the libopencv filter to apply.
cf69ad35 1677
ee0e8d4b
AK
1678@item filter_params
1679The parameters to pass to the libopencv filter. If not specified the default
1680values are assumed.
1681
1682@end table
cf69ad35
SS
1683
1684Refer to the official libopencv documentation for more precise
da9cea77 1685information:
cf69ad35
SS
1686@url{http://opencv.willowgarage.com/documentation/c/image_filtering.html}
1687
1688Follows the list of supported libopencv filters.
1689
17fc9493 1690@anchor{dilate}
91cbb6ba
SS
1691@subsection dilate
1692
1693Dilate an image by using a specific structuring element.
1694This filter corresponds to the libopencv function @code{cvDilate}.
1695
ee0e8d4b 1696It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
91cbb6ba
SS
1697
1698@var{struct_el} represents a structuring element, and has the syntax:
1699@var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
1700
da9cea77 1701@var{cols} and @var{rows} represent the number of columns and rows of
91cbb6ba
SS
1702the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
1703point, and @var{shape} the shape for the structuring element, and
1704can be one of the values "rect", "cross", "ellipse", "custom".
1705
1706If the value for @var{shape} is "custom", it must be followed by a
1707string of the form "=@var{filename}". The file with name
1708@var{filename} is assumed to represent a binary image, with each
1709printable character corresponding to a bright pixel. When a custom
1710@var{shape} is used, @var{cols} and @var{rows} are ignored, the number
1711or columns and rows of the read file are assumed instead.
1712
1713The default value for @var{struct_el} is "3x3+0x0/rect".
1714
1715@var{nb_iterations} specifies the number of times the transform is
1716applied to the image, and defaults to 1.
1717
1718Follow some example:
1719@example
1720# use the default values
1721ocv=dilate
1722
1723# dilate using a structuring element with a 5x5 cross, iterate two times
ee0e8d4b 1724ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
91cbb6ba
SS
1725
1726# read the shape from the file diamond.shape, iterate two times
1727# the file diamond.shape may contain a pattern of characters like this:
1728# *
1729# ***
1730# *****
1731# ***
1732# *
1733# the specified cols and rows are ignored (but not the anchor point coordinates)
ee0e8d4b 1734ocv=dilate:0x0+2x2/custom=diamond.shape|2
91cbb6ba
SS
1735@end example
1736
17fc9493
SS
1737@subsection erode
1738
1739Erode an image by using a specific structuring element.
1740This filter corresponds to the libopencv function @code{cvErode}.
1741
1742The filter accepts the parameters: @var{struct_el}:@var{nb_iterations},
4c989761 1743with the same syntax and semantics as the @ref{dilate} filter.
17fc9493 1744
cf69ad35
SS
1745@subsection smooth
1746
1747Smooth the input video.
1748
1749The filter takes the following parameters:
ee0e8d4b 1750@var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
6ebf0bfc
SS
1751
1752@var{type} is the type of smooth filter to apply, and can be one of
58d94364 1753the following values: "blur", "blur_no_scale", "median", "gaussian",
6ebf0bfc
SS
1754"bilateral". The default value is "gaussian".
1755
1756@var{param1}, @var{param2}, @var{param3}, and @var{param4} are
1757parameters whose meanings depend on smooth type. @var{param1} and
1758@var{param2} accept integer positive values or 0, @var{param3} and
1759@var{param4} accept float values.
1760
1761The default value for @var{param1} is 3, the default value for the
1762other parameters is 0.
1763
58d94364 1764These parameters correspond to the parameters assigned to the
cf69ad35 1765libopencv function @code{cvSmooth}.
6ebf0bfc 1766
2cf74eca 1767@anchor{overlay}
58935b25
SS
1768@section overlay
1769
1770Overlay one video on top of another.
1771
1772It takes two inputs and one output, the first input is the "main"
1773video on which the second input is overlayed.
1774
9087eaf1
AK
1775This filter accepts the following parameters:
1776
1777@table @option
1778
1779@item x
1780The horizontal position of the left edge of the overlaid video on the main video.
1781
1782@item y
1783The vertical position of the top edge of the overlaid video on the main video.
1784
1785@end table
58935b25 1786
9087eaf1 1787The parameters are expressions containing the following parameters:
58935b25
SS
1788
1789@table @option
1790@item main_w, main_h
1791main input width and height
1792
1793@item W, H
1794same as @var{main_w} and @var{main_h}
1795
1796@item overlay_w, overlay_h
1797overlay input width and height
1798
1799@item w, h
1800same as @var{overlay_w} and @var{overlay_h}
de203abd
KL
1801
1802@item eof_action
1803The action to take when EOF is encountered on the secondary input, accepts one
1804of the following values:
1805
1806@table @option
1807@item repeat
1808repeat the last frame (the default)
1809@item endall
1810end both streams
1811@item pass
1812pass through the main input
1813@end table
1814
58935b25
SS
1815@end table
1816
1817Be aware that frames are taken from each input video in timestamp
1818order, hence, if their initial timestamps differ, it is a a good idea
1819to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
1820have them begin in the same zero timestamp, as it does the example for
1821the @var{movie} filter.
1822
1823Follow some examples:
1824@example
1825# draw the overlay at 10 pixels from the bottom right
1826# corner of the main video.
9087eaf1 1827overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
58935b25
SS
1828
1829# insert a transparent PNG logo in the bottom left corner of the input
9087eaf1 1830avconv -i input -i logo -filter_complex 'overlay=x=10:y=main_h-overlay_h-10' output
58935b25
SS
1831
1832# insert 2 different transparent PNG logos (second logo on bottom
1833# right corner):
7ce118ba 1834avconv -i input -i logo1 -i logo2 -filter_complex
9087eaf1 1835'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
58935b25
SS
1836
1837# add a transparent color layer on top of the main video,
1838# WxH specifies the size of the main input to the overlay filter
1839color=red@.3:WxH [over]; [in][over] overlay [out]
de203abd
KL
1840
1841# mask 10-20 seconds of a video by applying the delogo filter to a section
1842avconv -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
1843-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]'
1844masked.avi
58935b25
SS
1845@end example
1846
b0641ab7 1847You can chain together more overlays but the efficiency of such
58935b25
SS
1848approach is yet to be tested.
1849
3275ac6a
SS
1850@section pad
1851
1852Add paddings to the input image, and places the original input at the
1853given coordinates @var{x}, @var{y}.
1854
40c885c5
AK
1855This filter accepts the following parameters:
1856
1857@table @option
1858@item width, height
1859
1860Specify the size of the output image with the paddings added. If the
1861value for @var{width} or @var{height} is 0, the corresponding input size
1862is used for the output.
1863
1864The @var{width} expression can reference the value set by the
1865@var{height} expression, and vice versa.
1866
1867The default value of @var{width} and @var{height} is 0.
1868
1869@item x, y
1870
1871Specify the offsets where to place the input image in the padded area
1872with respect to the top/left border of the output image.
1873
1874The @var{x} expression can reference the value set by the @var{y}
1875expression, and vice versa.
1876
1877The default value of @var{x} and @var{y} is 0.
1878
1879@item color
1880
1881Specify the color of the padded area, it can be the name of a color
1882(case insensitive match) or a 0xRRGGBB[AA] sequence.
1883
1884The default value of @var{color} is "black".
1885
1886@end table
3275ac6a 1887
73a4f7c2
SS
1888The parameters @var{width}, @var{height}, @var{x}, and @var{y} are
1889expressions containing the following constants:
1890
1891@table @option
1892@item E, PI, PHI
1893the corresponding mathematical approximated values for e
1894(euler number), pi (greek PI), phi (golden ratio)
1895
1896@item in_w, in_h
e83c2dde 1897the input video width and height
73a4f7c2
SS
1898
1899@item iw, ih
1900same as @var{in_w} and @var{in_h}
1901
1902@item out_w, out_h
e83c2dde 1903the output width and height, that is the size of the padded area as
73a4f7c2
SS
1904specified by the @var{width} and @var{height} expressions
1905
1906@item ow, oh
1907same as @var{out_w} and @var{out_h}
1908
1909@item x, y
1910x and y offsets as specified by the @var{x} and @var{y}
1911expressions, or NAN if not yet specified
1912
1913@item a
1914input display aspect ratio, same as @var{iw} / @var{ih}
1915
1916@item hsub, vsub
1917horizontal and vertical chroma subsample values. For example for the
1918pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
1919@end table
1920
73a4f7c2 1921Some examples follow:
3d17f4b9
SS
1922
1923@example
1924# Add paddings with color "violet" to the input video. Output video
1925# size is 640x480, the top-left corner of the input video is placed at
aeefbf61 1926# column 0, row 40.
40c885c5 1927pad=width=640:height=480:x=0:y=40:color=violet
73a4f7c2
SS
1928
1929# pad the input to get an output with dimensions increased bt 3/2,
1930# and put the input video at the center of the padded area
1931pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
1932
1933# pad the input to get a squared output with size equal to the maximum
1934# value between the input width and height, and put the input video at
1935# the center of the padded area
1936pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
1937
1938# pad the input to get a final w/h ratio of 16:9
1939pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
1940
1941# double output size and put the input video in the bottom-right
1942# corner of the output padded area
1943pad="2*iw:2*ih:ow-iw:oh-ih"
3d17f4b9
SS
1944@end example
1945
ce2e4ae3
SS
1946@section pixdesctest
1947
1948Pixel format descriptor test filter, mainly useful for internal
1949testing. The output video should be equal to the input video.
1950
1951For example:
1952@example
1953format=monow, pixdesctest
1954@end example
1955
1956can be used to test the monowhite pixel format descriptor definition.
1957
dd249245 1958@anchor{scale}
3275ac6a
SS
1959@section scale
1960
c334c113 1961Scale the input video and/or convert the image format.
3275ac6a 1962
c334c113
AK
1963This filter accepts the following options:
1964
1965@table @option
1966
1967@item w
1968Output video width.
1969
1970@item h
1971Output video height.
1972
1973@end table
1974
1975The parameters @var{w} and @var{h} are expressions containing
68e23c08 1976the following constants:
3275ac6a 1977
68e23c08
SS
1978@table @option
1979@item E, PI, PHI
1980the corresponding mathematical approximated values for e
1981(euler number), pi (greek PI), phi (golden ratio)
1982
1983@item in_w, in_h
e83c2dde 1984the input width and height
68e23c08
SS
1985
1986@item iw, ih
1987same as @var{in_w} and @var{in_h}
1988
1989@item out_w, out_h
e83c2dde 1990the output (cropped) width and height
68e23c08
SS
1991
1992@item ow, oh
1993same as @var{out_w} and @var{out_h}
3275ac6a 1994
3a14b160
RP
1995@item a
1996same as @var{iw} / @var{ih}
68e23c08 1997
46b29397
SS
1998@item sar
1999input sample aspect ratio
2000
3a14b160
RP
2001@item dar
2002input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
2003
68e23c08
SS
2004@item hsub, vsub
2005horizontal and vertical chroma subsample values. For example for the
2006pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2007@end table
3275ac6a
SS
2008
2009If the input image format is different from the format requested by
2010the next filter, the scale filter will convert the input to the
2011requested format.
2012
c334c113 2013If the value for @var{w} or @var{h} is 0, the respective input
3275ac6a
SS
2014size is used for the output.
2015
c334c113
AK
2016If the value for @var{w} or @var{h} is -1, the scale filter will use, for the
2017respective output size, a value that maintains the aspect ratio of the input
2018image.
3275ac6a 2019
c334c113 2020The default value of @var{w} and @var{h} is 0.
3275ac6a 2021
68e23c08
SS
2022Some examples follow:
2023@example
2024# scale the input video to a size of 200x100.
c334c113 2025scale=w=200:h=100
68e23c08
SS
2026
2027# scale the input to 2x
c334c113 2028scale=w=2*iw:h=2*ih
68e23c08
SS
2029# the above is the same as
2030scale=2*in_w:2*in_h
2031
2032# scale the input to half size
c334c113 2033scale=w=iw/2:h=ih/2
68e23c08
SS
2034
2035# increase the width, and set the height to the same size
2036scale=3/2*iw:ow
2037
2038# seek for Greek harmony
2039scale=iw:1/PHI*iw
2040scale=ih*PHI:ih
2041
2042# increase the height, and set the width to 3/2 of the height
c334c113 2043scale=w=3/2*oh:h=3/5*ih
68e23c08
SS
2044
2045# increase the size, but make the size a multiple of the chroma
2046scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
2047
2048# increase the width to a maximum of 500 pixels, keep the same input aspect ratio
c334c113 2049scale=w='min(500\, iw*3/2):h=-1'
68e23c08
SS
2050@end example
2051
d763fb7d
SS
2052@section select
2053Select frames to pass in output.
2054
95f1f56a
AK
2055This filter accepts the following options:
2056
2057@table @option
2058
2059@item expr
2060An expression, which is evaluated for each input frame. If the expression is
2061evaluated to a non-zero value, the frame is selected and passed to the output,
2062otherwise it is discarded.
2063
2064@end table
d763fb7d
SS
2065
2066The expression can contain the following constants:
2067
2068@table @option
2069@item PI
2070Greek PI
2071
2072@item PHI
2073golden ratio
2074
2075@item E
2076Euler number
2077
2078@item n
2079the sequential number of the filtered frame, starting from 0
2080
2081@item selected_n
2082the sequential number of the selected frame, starting from 0
2083
2084@item prev_selected_n
2085the sequential number of the last selected frame, NAN if undefined
2086
2087@item TB
2088timebase of the input timestamps
2089
2090@item pts
2091the PTS (Presentation TimeStamp) of the filtered video frame,
2092expressed in @var{TB} units, NAN if undefined
2093
2094@item t
2095the PTS (Presentation TimeStamp) of the filtered video frame,
2096expressed in seconds, NAN if undefined
2097
2098@item prev_pts
2099the PTS of the previously filtered video frame, NAN if undefined
2100
2101@item prev_selected_pts
2102the PTS of the last previously filtered video frame, NAN if undefined
2103
2104@item prev_selected_t
2105the PTS of the last previously selected video frame, NAN if undefined
2106
2107@item start_pts
2108the PTS of the first video frame in the video, NAN if undefined
2109
2110@item start_t
2111the time of the first video frame in the video, NAN if undefined
2112
2113@item pict_type
2114the type of the filtered frame, can assume one of the following
2115values:
2116@table @option
2117@item I
2118@item P
2119@item B
2120@item S
2121@item SI
2122@item SP
2123@item BI
2124@end table
2125
2126@item interlace_type
2127the frame interlace type, can assume one of the following values:
2128@table @option
2129@item PROGRESSIVE
2130the frame is progressive (not interlaced)
2131@item TOPFIRST
2132the frame is top-field-first
2133@item BOTTOMFIRST
2134the frame is bottom-field-first
2135@end table
2136
2137@item key
21381 if the filtered frame is a key-frame, 0 otherwise
2139
d763fb7d
SS
2140@end table
2141
2142The default value of the select expression is "1".
2143
2144Some examples follow:
2145
2146@example
2147# select all frames in input
2148select
2149
2150# the above is the same as:
95f1f56a 2151select=expr=1
d763fb7d
SS
2152
2153# skip all frames:
95f1f56a 2154select=expr=0
d763fb7d
SS
2155
2156# select only I-frames
95f1f56a 2157select='expr=eq(pict_type\,I)'
d763fb7d
SS
2158
2159# select one frame every 100
2160select='not(mod(n\,100))'
2161
2162# select only frames contained in the 10-20 time interval
2163select='gte(t\,10)*lte(t\,20)'
2164
2165# select only I frames contained in the 10-20 time interval
2166select='gte(t\,10)*lte(t\,20)*eq(pict_type\,I)'
2167
2168# select frames with a minimum distance of 10 seconds
2169select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
2170@end example
2171
2fd8756b
SS
2172@anchor{setdar}
2173@section setdar
2174
2175Set the Display Aspect Ratio for the filter output video.
2176
2177This is done by changing the specified Sample (aka Pixel) Aspect
2178Ratio, according to the following equation:
2179@math{DAR = HORIZONTAL_RESOLUTION / VERTICAL_RESOLUTION * SAR}
2180
2181Keep in mind that this filter does not modify the pixel dimensions of
2182the video frame. Also the display aspect ratio set by this filter may
2183be changed by later filters in the filterchain, e.g. in case of
2184scaling or if another "setdar" or a "setsar" filter is applied.
2185
2831b307
AK
2186This filter accepts the following options:
2187
2188@table @option
2189
2190@item dar
34a0ac41 2191Output display aspect ratio.
2831b307
AK
2192
2193@end table
2fd8756b 2194
34a0ac41
RP
2195The parameter @var{dar} is an expression containing
2196the following constants:
2197
2198@table @option
2199@item E, PI, PHI
2200the corresponding mathematical approximated values for e
2201(euler number), pi (greek PI), phi (golden ratio)
2202
2203@item w, h
2204the input width and height
2205
2206@item a
2207same as @var{w} / @var{h}
2208
2209@item sar
2210input sample aspect ratio
2211
2212@item dar
2213input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
2214
2215@item hsub, vsub
2216horizontal and vertical chroma subsample values. For example for the
2217pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2218@end table
2219
2fd8756b
SS
2220For example to change the display aspect ratio to 16:9, specify:
2221@example
2831b307 2222setdar=dar=16/9
2fd8756b 2223# the above is equivalent to
2831b307 2224setdar=dar=1.77777
2fd8756b
SS
2225@end example
2226
4c989761 2227See also the @ref{setsar} filter documentation.
2fd8756b 2228
a532bb39
SS
2229@section setpts
2230
2231Change the PTS (presentation timestamp) of the input video frames.
2232
33b97faa
AK
2233This filter accepts the following options:
2234
2235@table @option
2236
2237@item expr
2238The expression which is evaluated for each frame to construct its timestamp.
2239
2240@end table
2241
2242The expression is evaluated through the eval API and can contain the following
2243constants:
a532bb39
SS
2244
2245@table @option
2246@item PTS
2247the presentation timestamp in input
2248
2249@item PI
2250Greek PI
2251
2252@item PHI
2253golden ratio
2254
2255@item E
2256Euler number
2257
2258@item N
2259the count of the input frame, starting from 0.
2260
2261@item STARTPTS
2262the PTS of the first video frame
2263
2264@item INTERLACED
2265tell if the current frame is interlaced
2266
a532bb39
SS
2267@item PREV_INPTS
2268previous input PTS
2269
2270@item PREV_OUTPTS
2271previous output PTS
2272
0b55b16a
VP
2273@item RTCTIME
2274wallclock (RTC) time in microseconds
2275
2276@item RTCSTART
2277wallclock (RTC) time at the start of the movie in microseconds
2278
0ebfdae0
CW
2279@item TB
2280timebase of the input timestamps
2281
a532bb39
SS
2282@end table
2283
2284Some examples follow:
2285
2286@example
2287# start counting PTS from zero
33b97faa 2288setpts=expr=PTS-STARTPTS
a532bb39
SS
2289
2290# fast motion
33b97faa 2291setpts=expr=0.5*PTS
a532bb39
SS
2292
2293# slow motion
2294setpts=2.0*PTS
2295
2296# fixed rate 25 fps
2297setpts=N/(25*TB)
2298
2299# fixed rate 25 fps with some jitter
2300setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
0b55b16a
VP
2301
2302# generate timestamps from a "live source" and rebase onto the current timebase
2303setpts='(RTCTIME - RTCSTART) / (TB * 1000000)"
a532bb39
SS
2304@end example
2305
2fd8756b
SS
2306@anchor{setsar}
2307@section setsar
2308
2309Set the Sample (aka Pixel) Aspect Ratio for the filter output video.
2310
2311Note that as a consequence of the application of this filter, the
2312output display aspect ratio will change according to the following
2313equation:
2314@math{DAR = HORIZONTAL_RESOLUTION / VERTICAL_RESOLUTION * SAR}
2315
2316Keep in mind that the sample aspect ratio set by this filter may be
2317changed by later filters in the filterchain, e.g. if another "setsar"
2318or a "setdar" filter is applied.
2319
2831b307
AK
2320This filter accepts the following options:
2321
2322@table @option
2323
2324@item sar
34a0ac41 2325Output sample aspect ratio.
2831b307
AK
2326
2327@end table
2fd8756b 2328
34a0ac41
RP
2329The parameter @var{sar} is an expression containing
2330the following constants:
2331
2332@table @option
2333@item E, PI, PHI
2334the corresponding mathematical approximated values for e
2335(euler number), pi (greek PI), phi (golden ratio)
2336
2337@item w, h
2338the input width and height
2339
2340@item a
2341same as @var{w} / @var{h}
2342
2343@item sar
2344input sample aspect ratio
2345
2346@item dar
2347input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
2348
2349@item hsub, vsub
2350horizontal and vertical chroma subsample values. For example for the
2351pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
2352@end table
2353
2fd8756b
SS
2354For example to change the sample aspect ratio to 10:11, specify:
2355@example
2831b307 2356setsar=sar=10/11
2fd8756b
SS
2357@end example
2358
d89e3b36
SS
2359@section settb
2360
2361Set the timebase to use for the output frames timestamps.
2362It is mainly useful for testing timebase configuration.
2363
ffea3b00
AK
2364This filter accepts the following options:
2365
2366@table @option
2367
2368@item expr
2369The expression which is evaluated into the output timebase.
2370
2371@end table
2372
d89e3b36
SS
2373The expression can contain the constants "PI", "E", "PHI", "AVTB" (the
2374default timebase), and "intb" (the input timebase).
2375
2376The default value for the input is "intb".
2377
2378Follow some examples.
2379
2380@example
2381# set the timebase to 1/25
ffea3b00 2382settb=expr=1/25
d89e3b36
SS
2383
2384# set the timebase to 1/10
ffea3b00 2385settb=expr=0.1
d89e3b36
SS
2386
2387#set the timebase to 1001/1000
2388settb=1+0.001
2389
2390#set the timebase to 2*intb
2391settb=2*intb
2392
2393#set the default timebase value
2394settb=AVTB
2395@end example
2396
ee42716b
SS
2397@section showinfo
2398
2399Show a line containing various information for each input video frame.
2400The input video is not modified.
2401
2402The shown line contains a sequence of key/value pairs of the form
2403@var{key}:@var{value}.
2404
2405A description of each shown parameter follows:
2406
2407@table @option
2408@item n
2409sequential number of the input frame, starting from 0
2410
2411@item pts
2412Presentation TimeStamp of the input frame, expressed as a number of
2413time base units. The time base unit depends on the filter input pad.
2414
2415@item pts_time
2416Presentation TimeStamp of the input frame, expressed as a number of
2417seconds
2418
2419@item pos
2420position of the frame in the input stream, -1 if this information in
da9cea77 2421unavailable and/or meaningless (for example in case of synthetic video)
ee42716b
SS
2422
2423@item fmt
2424pixel format name
2425
2426@item sar
2427sample aspect ratio of the input frame, expressed in the form
2428@var{num}/@var{den}
2429
2430@item s
2431size of the input frame, expressed in the form
2432@var{width}x@var{height}
2433
2434@item i
2435interlaced mode ("P" for "progressive", "T" for top field first, "B"
2436for bottom field first)
2437
2438@item iskey
24391 if the frame is a key frame, 0 otherwise
2440
2441@item type
2442picture type of the input frame ("I" for an I-frame, "P" for a
2443P-frame, "B" for a B-frame, "?" for unknown type).
2444Check also the documentation of the @code{AVPictureType} enum and of
2445the @code{av_get_picture_type_char} function defined in
2446@file{libavutil/avutil.h}.
2447
2448@item checksum
2449Adler-32 checksum of all the planes of the input frame
2450
2451@item plane_checksum
2452Adler-32 checksum of each plane of the input frame, expressed in the form
2453"[@var{c0} @var{c1} @var{c2} @var{c3}]"
2454@end table
2455
fd18ee0f
AK
2456@section split
2457
2458Split input video into several identical outputs.
2459
2460The filter accepts a single parameter which specifies the number of outputs. If
2461unspecified, it defaults to 2.
2462
2463For example
2464@example
2465avconv -i INPUT -filter_complex split=5 OUTPUT
2466@end example
2467will create 5 copies of the input video.
2468
43945b27
SS
2469@section transpose
2470
2471Transpose rows with columns in the input video and optionally flip it.
2472
0c2466de
AK
2473This filter accepts the following options:
2474
2475@table @option
2476
2477@item dir
2478The direction of the transpose.
2479
2480@end table
2481
2482The direction can assume the following values:
43945b27
SS
2483
2484@table @samp
0c2466de 2485@item cclock_flip
43945b27
SS
2486Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
2487@example
2488L.R L.l
2489. . -> . .
2490l.r R.r
2491@end example
2492
0c2466de 2493@item clock
43945b27
SS
2494Rotate by 90 degrees clockwise, that is:
2495@example
2496L.R l.L
2497. . -> . .
2498l.r r.R
2499@end example
2500
0c2466de 2501@item cclock
43945b27
SS
2502Rotate by 90 degrees counterclockwise, that is:
2503@example
2504L.R R.r
2505. . -> . .
2506l.r L.l
2507@end example
2508
0c2466de 2509@item clock_flip
43945b27
SS
2510Rotate by 90 degrees clockwise and vertically flip, that is:
2511@example
2512L.R r.R
2513. . -> . .
2514l.r l.L
2515@end example
2516@end table
2517
a1e05b04
AK
2518@section trim
2519Trim the input so that the output contains one continuous subpart of the input.
2520
2521This filter accepts the following options:
2522@table @option
2523@item start
2524Timestamp (in seconds) of the start of the kept section. I.e. the frame with the
2525timestamp @var{start} will be the first frame in the output.
2526
2527@item end
2528Timestamp (in seconds) of the first frame that will be dropped. I.e. the frame
2529immediately preceding the one with the timestamp @var{end} will be the last
2530frame in the output.
2531
2532@item start_pts
2533Same as @var{start}, except this option sets the start timestamp in timebase
2534units instead of seconds.
2535
2536@item end_pts
2537Same as @var{end}, except this option sets the end timestamp in timebase units
2538instead of seconds.
2539
2540@item duration
2541Maximum duration of the output in seconds.
2542
2543@item start_frame
2544Number of the first frame that should be passed to output.
2545
2546@item end_frame
2547Number of the first frame that should be dropped.
2548@end table
2549
2550Note that the first two sets of the start/end options and the @option{duration}
2551option look at the frame timestamp, while the _frame variants simply count the
2552frames that pass through the filter. Also note that this filter does not modify
2553the timestamps. If you wish that the output timestamps start at zero, insert a
2554setpts filter after the trim filter.
2555
2556If multiple start or end options are set, this filter tries to be greedy and
2557keep all the frames that match at least one of the specified constraints. To keep
2558only the part that matches all the constraints at once, chain multiple trim
2559filters.
2560
2561The defaults are such that all the input is kept. So it is possible to set e.g.
2562just the end values to keep everything before the specified time.
2563
2564Examples:
2565@itemize
2566@item
2567drop everything except the second minute of input
2568@example
2569avconv -i INPUT -vf trim=60:120
2570@end example
2571
2572@item
2573keep only the first second
2574@example
2575avconv -i INPUT -vf trim=duration=1
2576@end example
2577
2578@end itemize
3275ac6a
SS
2579@section unsharp
2580
843b5fd0
SS
2581Sharpen or blur the input video.
2582
2583It accepts the following parameters:
3275ac6a
SS
2584
2585@table @option
2586
2587@item luma_msize_x
2588Set the luma matrix horizontal size. It can be an integer between 3
2589and 13, default value is 5.
2590
2591@item luma_msize_y
2592Set the luma matrix vertical size. It can be an integer between 3
2593and 13, default value is 5.
2594
2595@item luma_amount
2596Set the luma effect strength. It can be a float number between -2.0
2597and 5.0, default value is 1.0.
2598
2599@item chroma_msize_x
2600Set the chroma matrix horizontal size. It can be an integer between 3
1ee20141 2601and 13, default value is 5.
3275ac6a
SS
2602
2603@item chroma_msize_y
2604Set the chroma matrix vertical size. It can be an integer between 3
1ee20141 2605and 13, default value is 5.
3275ac6a 2606
2725f2d4 2607@item chroma_amount
3275ac6a
SS
2608Set the chroma effect strength. It can be a float number between -2.0
2609and 5.0, default value is 0.0.
2610
2611@end table
2612
b83e9efc
AK
2613Negative values for the amount will blur the input video, while positive
2614values will sharpen. All parameters are optional and default to the
2615equivalent of the string '5:5:1.0:5:5:0.0'.
2616
3275ac6a
SS
2617@example
2618# Strong luma sharpen effect parameters
b83e9efc 2619unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
3275ac6a
SS
2620
2621# Strong blur of both luma and chroma parameters
2622unsharp=7:7:-2:7:7:-2
2623
9270b8a3
LB
2624# Use the default values with @command{avconv}
2625./avconv -i in.avi -vf "unsharp" out.mp4
3275ac6a
SS
2626@end example
2627
2628@section vflip
2629
2630Flip the input video vertically.
2631
2632@example
9270b8a3 2633./avconv -i in.avi -vf "vflip" out.avi
3275ac6a
SS
2634@end example
2635
acbac567
MN
2636@section yadif
2637
1653027a
SS
2638Deinterlace the input video ("yadif" means "yet another deinterlacing
2639filter").
acbac567 2640
7536c671
AK
2641This filter accepts the following options:
2642
2643@table @option
acbac567 2644
7536c671
AK
2645@item mode
2646The interlacing mode to adopt, accepts one of the following values:
acbac567 2647
1653027a
SS
2648@table @option
2649@item 0
2650output 1 frame for each frame
2651@item 1
2652output 1 frame for each field
2653@item 2
2654like 0 but skips spatial interlacing check
2655@item 3
2656like 1 but skips spatial interlacing check
2657@end table
acbac567
MN
2658
2659Default value is 0.
2660
7536c671
AK
2661@item parity
2662The picture field parity assumed for the input interlaced video, accepts one of
2663the following values:
acbac567 2664
1653027a
SS
2665@table @option
2666@item 0
1653027a 2667assume top field first
4703a7b5
SS
2668@item 1
2669assume bottom field first
1653027a
SS
2670@item -1
2671enable automatic detection
acbac567
MN
2672@end table
2673
1653027a 2674Default value is -1.
a51c71bb
BC
2675If interlacing is unknown or decoder does not export this information,
2676top field first will be assumed.
1653027a 2677
7536c671
AK
2678@item auto
2679Whether deinterlacer should trust the interlaced flag and only deinterlace
2680frames marked as interlaced
ab09df9d
JP
2681
2682@table @option
2683@item 0
2684deinterlace all frames
2685@item 1
2686only deinterlace frames marked as interlaced
2687@end table
2688
2689Default value is 0.
2690
7536c671
AK
2691@end table
2692
3275ac6a
SS
2693@c man end VIDEO FILTERS
2694
2695@chapter Video Sources
2696@c man begin VIDEO SOURCES
2697
2698Below is a description of the currently available video sources.
2699
24413399
SS
2700@section buffer
2701
2702Buffer video frames, and make them available to the filter chain.
2703
2704This source is mainly intended for a programmatic use, in particular
ac1a31ab 2705through the interface defined in @file{libavfilter/vsrc_buffer.h}.
24413399 2706
d28cb849 2707This filter accepts the following parameters:
24413399 2708
d28cb849 2709@table @option
24413399 2710
d28cb849
AK
2711@item width
2712Input video width.
24413399 2713
d28cb849
AK
2714@item height
2715Input video height.
24413399 2716
d28cb849
AK
2717@item pix_fmt
2718Name of the input video pixel format.
24413399 2719
d28cb849
AK
2720@item time_base
2721The time base used for input timestamps.
24413399 2722
d28cb849
AK
2723@item sar
2724Sample (pixel) aspect ratio of the input video.
7a11c82f 2725
24413399
SS
2726@end table
2727
2728For example:
2729@example
d28cb849 2730buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
24413399
SS
2731@end example
2732
2733will instruct the source to accept video frames with size 320x240 and
7a11c82f
MN
2734with format "yuv410p", assuming 1/24 as the timestamps timebase and
2735square pixels (1:1 sample aspect ratio).
24413399 2736
23ccf3c7
SS
2737@section color
2738
2739Provide an uniformly colored input.
2740
2741It accepts the following parameters:
23ccf3c7
SS
2742
2743@table @option
2744
2745@item color
2746Specify the color of the source. It can be the name of a color (case
2747insensitive match) or a 0xRRGGBB[AA] sequence, possibly followed by an
2748alpha specifier. The default value is "black".
2749
7bc1a883 2750@item size
23ccf3c7 2751Specify the size of the sourced video, it may be a string of the form
e83c2dde 2752@var{width}x@var{height}, or the name of a size abbreviation. The
23ccf3c7
SS
2753default value is "320x240".
2754
7bc1a883 2755@item framerate
23ccf3c7
SS
2756Specify the frame rate of the sourced video, as the number of frames
2757generated per second. It has to be a string in the format
2758@var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
2759number or a valid video frame rate abbreviation. The default value is
2760"25".
2761
2762@end table
2763
2764For example the following graph description will generate a red source
2765with an opacity of 0.2, with size "qcif" and a frame rate of 10
2766frames per second, which will be overlayed over the source connected
2767to the pad with identifier "in".
2768
2769@example
2770"color=red@@0.2:qcif:10 [color]; [in][color] overlay [out]"
2771@end example
2772
9409c381
SS
2773@section movie
2774
2775Read a video stream from a movie container.
2776
7ca14c73
AK
2777Note that this source is a hack that bypasses the standard input path. It can be
2778useful in applications that do not support arbitrary filter graphs, but its use
2779is discouraged in those that do. Specifically in @command{avconv} this filter
2780should never be used, the @option{-filter_complex} option fully replaces it.
2781
a42d6e6c 2782This filter accepts the following options:
9409c381
SS
2783
2784@table @option
2785
a42d6e6c
AK
2786@item filename
2787The name of the resource to read (not necessarily a file but also a device or a
2788stream accessed through some protocol).
2789
9409c381
SS
2790@item format_name, f
2791Specifies the format assumed for the movie to read, and can be either
2792the name of a container or an input device. If not specified the
2793format is guessed from @var{movie_name} or by probing.
2794
2795@item seek_point, sp
2796Specifies the seek point in seconds, the frames will be output
2797starting from this seek point, the parameter is evaluated with
2798@code{av_strtod} so the numerical value may be suffixed by an IS
2799postfix. Default value is "0".
2800
2801@item stream_index, si
2802Specifies the index of the video stream to read. If the value is -1,
2803the best suited video stream will be automatically selected. Default
2804value is "-1".
2805
2806@end table
2807
2808This filter allows to overlay a second video on top of main input of
2809a filtergraph as shown in this graph:
2810@example
2811input -----------> deltapts0 --> overlay --> output
2812 ^
2813 |
2814movie --> scale--> deltapts1 -------+
2815@end example
2816
2817Some examples follow:
2818@example
2819# skip 3.2 seconds from the start of the avi file in.avi, and overlay it
2820# on top of the input labelled as "in".
2821movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [movie];
2822[in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
2823
2824# read from a video4linux2 device, and overlay it on top of the input
2825# labelled as "in"
2826movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [movie];
2827[in] setpts=PTS-STARTPTS, [movie] overlay=16:16 [out]
2828
2829@end example
2830
3275ac6a
SS
2831@section nullsrc
2832
2833Null video source, never return images. It is mainly useful as a
2834template and to be employed in analysis / debugging tools.
2835
2836It accepts as optional parameter a string of the form
16134b7c 2837@var{width}:@var{height}:@var{timebase}.
3275ac6a 2838
16134b7c
SS
2839@var{width} and @var{height} specify the size of the configured
2840source. The default values of @var{width} and @var{height} are
2841respectively 352 and 288 (corresponding to the CIF size format).
2842
2843@var{timebase} specifies an arithmetic expression representing a
2844timebase. The expression can contain the constants "PI", "E", "PHI",
2845"AVTB" (the default timebase), and defaults to the value "AVTB".
3275ac6a 2846
f8608dca
SS
2847@section frei0r_src
2848
2849Provide a frei0r source.
2850
2851To enable compilation of this filter you need to install the frei0r
f8a45fa1 2852header and configure Libav with --enable-frei0r.
f8608dca 2853
5aa1a668
AK
2854This source accepts the following options:
2855
2856@table @option
2857
2858@item size
2859The size of the video to generate, may be a string of the form
2860@var{width}x@var{height} or a frame size abbreviation.
f8608dca 2861
5aa1a668
AK
2862@item framerate
2863Framerate of the generated video, may be a string of the form
2864@var{num}/@var{den} or a frame rate abbreviation.
2865
2866@item filter_name
2867The name to the frei0r source to load. For more information regarding frei0r and
2868how to set the parameters read the section @ref{frei0r} in the description of
2869the video filters.
2870
2871@item filter_params
2872A '|'-separated list of parameters to pass to the frei0r source.
2873
2874@end table
f8608dca
SS
2875
2876Some examples follow:
2877@example
2878# generate a frei0r partik0l source with size 200x200 and framerate 10
2879# which is overlayed on the overlay filter main input
5aa1a668 2880frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
f8608dca
SS
2881@end example
2882
ec2ac927 2883@section rgbtestsrc, testsrc
0244879f 2884
ec2ac927
SS
2885The @code{rgbtestsrc} source generates an RGB test pattern useful for
2886detecting RGB vs BGR issues. You should see a red, green and blue
2887stripe from top to bottom.
0244879f 2888
ec2ac927
SS
2889The @code{testsrc} source generates a test video pattern, showing a
2890color pattern, a scrolling gradient and a timestamp. This is mainly
2891intended for testing purposes.
2892
7b3eb745 2893The sources accept the following options:
0244879f
SS
2894
2895@table @option
2896
2897@item size, s
2898Specify the size of the sourced video, it may be a string of the form
da9cea77 2899@var{width}x@var{height}, or the name of a size abbreviation. The
0244879f
SS
2900default value is "320x240".
2901
2902@item rate, r
2903Specify the frame rate of the sourced video, as the number of frames
2904generated per second. It has to be a string in the format
2905@var{frame_rate_num}/@var{frame_rate_den}, an integer number, a float
2906number or a valid video frame rate abbreviation. The default value is
2907"25".
2908
2909@item sar
2910Set the sample aspect ratio of the sourced video.
2911
2912@item duration
2913Set the video duration of the sourced video. The accepted syntax is:
2914@example
2915[-]HH[:MM[:SS[.m...]]]
2916[-]S+[.m...]
2917@end example
2918See also the function @code{av_parse_time()}.
2919
2920If not specified, or the expressed duration is negative, the video is
2921supposed to be generated forever.
2922@end table
2923
2924For example the following:
2925@example
2926testsrc=duration=5.3:size=qcif:rate=10
2927@end example
2928
2929will generate a video with a duration of 5.3 seconds, with size
2930176x144 and a framerate of 10 frames per second.
2931
3275ac6a
SS
2932@c man end VIDEO SOURCES
2933
2934@chapter Video Sinks
2935@c man begin VIDEO SINKS
2936
2937Below is a description of the currently available video sinks.
2938
ac712309
AK
2939@section buffersink
2940
2941Buffer video frames, and make them available to the end of the filter
2942graph.
2943
2944This sink is intended for a programmatic use through the interface defined in
2945@file{libavfilter/buffersink.h}.
2946
3275ac6a
SS
2947@section nullsink
2948
2949Null video sink, do absolutely nothing with the input video. It is
2950mainly useful as a template and to be employed in analysis / debugging
2951tools.
2952
2953@c man end VIDEO SINKS