yading@10: =head1 NAME
yading@10: 
yading@10: ffmpeg-filters - FFmpeg filters
yading@10: 
yading@10: =head1 DESCRIPTION
yading@10: 
yading@10: 
yading@10: This document describes filters, sources, and sinks provided by the
yading@10: libavfilter library.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 FILTERING INTRODUCTION
yading@10: 
yading@10: 
yading@10: Filtering in FFmpeg is enabled through the libavfilter library.
yading@10: 
yading@10: In libavfilter, a filter can have multiple inputs and multiple
yading@10: outputs.
yading@10: To illustrate the sorts of things that are possible, we consider the
yading@10: following filtergraph.
yading@10: 
yading@10: 	
yading@10: 	input --> split ---------------------> overlay --> output
yading@10: 	            |                             ^
yading@10: 	            |                             |
yading@10: 	            +-----> crop --> vflip -------+
yading@10: 
yading@10: 
yading@10: This filtergraph splits the input stream in two streams, sends one
yading@10: stream through the crop filter and the vflip filter before merging it
yading@10: back with the other stream by overlaying it on top. You can use the
yading@10: following command to achieve this:
yading@10: 
yading@10: 	
yading@10: 	ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
yading@10: 
yading@10: 
yading@10: The result will be that in output the top half of the video is mirrored
yading@10: onto the bottom half.
yading@10: 
yading@10: Filters in the same linear chain are separated by commas, and distinct
yading@10: linear chains of filters are separated by semicolons. In our example,
yading@10: I<crop,vflip> are in one linear chain, I<split> and
yading@10: I<overlay> are separately in another. The points where the linear
yading@10: chains join are labelled by names enclosed in square brackets. In the
yading@10: example, the split filter generates two outputs that are associated to
yading@10: the labels I<[main]> and I<[tmp]>.
yading@10: 
yading@10: The stream sent to the second output of I<split>, labelled as
yading@10: I<[tmp]>, is processed through the I<crop> filter, which crops
yading@10: away the lower half part of the video, and then vertically flipped. The
yading@10: I<overlay> filter takes in input the first unchanged output of the
yading@10: split filter (which was labelled as I<[main]>), and overlay on its
yading@10: lower half the output generated by the I<crop,vflip> filterchain.
yading@10: 
yading@10: Some filters take in input a list of parameters: they are specified
yading@10: after the filter name and an equal sign, and are separated from each other
yading@10: by a colon.
yading@10: 
yading@10: There exist so-called I<source filters> that do not have an
yading@10: audio/video input, and I<sink filters> that will not have audio/video
yading@10: output.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 GRAPH
yading@10: 
yading@10: 
yading@10: The F<graph2dot> program included in the FFmpeg F<tools>
yading@10: directory can be used to parse a filtergraph description and issue a
yading@10: corresponding textual representation in the dot language.
yading@10: 
yading@10: Invoke the command:
yading@10: 	
yading@10: 	graph2dot -h
yading@10: 
yading@10: 
yading@10: to see how to use F<graph2dot>.
yading@10: 
yading@10: You can then pass the dot description to the F<dot> program (from
yading@10: the graphviz suite of programs) and obtain a graphical representation
yading@10: of the filtergraph.
yading@10: 
yading@10: For example the sequence of commands:
yading@10: 	
yading@10: 	echo <GRAPH_DESCRIPTION> | \
yading@10: 	tools/graph2dot -o graph.tmp && \
yading@10: 	dot -Tpng graph.tmp -o graph.png && \
yading@10: 	display graph.png
yading@10: 
yading@10: 
yading@10: can be used to create and display an image representing the graph
yading@10: described by the I<GRAPH_DESCRIPTION> string. Note that this string must be
yading@10: a complete self-contained graph, with its inputs and outputs explicitly defined.
yading@10: For example if your command line is of the form:
yading@10: 	
yading@10: 	ffmpeg -i infile -vf scale=640:360 outfile
yading@10: 
yading@10: your I<GRAPH_DESCRIPTION> string will need to be of the form:
yading@10: 	
yading@10: 	nullsrc,scale=640:360,nullsink
yading@10: 
yading@10: you may also need to set the I<nullsrc> parameters and add a I<format>
yading@10: filter in order to simulate a specific input file.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 FILTERGRAPH DESCRIPTION
yading@10: 
yading@10: 
yading@10: A filtergraph is a directed graph of connected filters. It can contain
yading@10: cycles, and there can be multiple links between a pair of
yading@10: filters. Each link has one input pad on one side connecting it to one
yading@10: filter from which it takes its input, and one output pad on the other
yading@10: side connecting it to the one filter accepting its output.
yading@10: 
yading@10: Each filter in a filtergraph is an instance of a filter class
yading@10: registered in the application, which defines the features and the
yading@10: number of input and output pads of the filter.
yading@10: 
yading@10: A filter with no input pads is called a "source", a filter with no
yading@10: output pads is called a "sink".
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 Filtergraph syntax
yading@10: 
yading@10: 
yading@10: A filtergraph can be represented using a textual representation, which is
yading@10: recognized by the B<-filter>/B<-vf> and B<-filter_complex>
yading@10: options in B<ffmpeg> and B<-vf> in B<ffplay>, and by the
yading@10: C<avfilter_graph_parse()>/C<avfilter_graph_parse2()> function defined in
yading@10: F<libavfilter/avfilter.h>.
yading@10: 
yading@10: A filterchain consists of a sequence of connected filters, each one
yading@10: connected to the previous one in the sequence. A filterchain is
yading@10: represented by a list of ","-separated filter descriptions.
yading@10: 
yading@10: A filtergraph consists of a sequence of filterchains. A sequence of
yading@10: filterchains is represented by a list of ";"-separated filterchain
yading@10: descriptions.
yading@10: 
yading@10: A filter is represented by a string of the form:
yading@10: [I<in_link_1>]...[I<in_link_N>]I<filter_name>=I<arguments>[I<out_link_1>]...[I<out_link_M>]
yading@10: 
yading@10: I<filter_name> is the name of the filter class of which the
yading@10: described filter is an instance of, and has to be the name of one of
yading@10: the filter classes registered in the program.
yading@10: The name of the filter class is optionally followed by a string
yading@10: "=I<arguments>".
yading@10: 
yading@10: I<arguments> is a string which contains the parameters used to
yading@10: initialize the filter instance. It may have one of the following forms:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: A ':'-separated list of I<key=value> pairs.
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: A ':'-separated list of I<value>. In this case, the keys are assumed to be
yading@10: the option names in the order they are declared. E.g. the C<fade> filter
yading@10: declares three options in this order -- B<type>, B<start_frame> and
yading@10: B<nb_frames>. Then the parameter list I<in:0:30> means that the value
yading@10: I<in> is assigned to the option B<type>, I<0> to
yading@10: B<start_frame> and I<30> to B<nb_frames>.
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: A ':'-separated list of mixed direct I<value> and long I<key=value>
yading@10: pairs. The direct I<value> must precede the I<key=value> pairs, and
yading@10: follow the same constraints order of the previous point. The following
yading@10: I<key=value> pairs can be set in any preferred order.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If the option value itself is a list of items (e.g. the C<format> filter
yading@10: takes a list of pixel formats), the items in the list are usually separated by
yading@10: '|'.
yading@10: 
yading@10: The list of arguments can be quoted using the character "'" as initial
yading@10: and ending mark, and the character '\' for escaping the characters
yading@10: within the quoted text; otherwise the argument string is considered
yading@10: terminated when the next special character (belonging to the set
yading@10: "[]=;,") is encountered.
yading@10: 
yading@10: The name and arguments of the filter are optionally preceded and
yading@10: followed by a list of link labels.
yading@10: A link label allows to name a link and associate it to a filter output
yading@10: or input pad. The preceding labels I<in_link_1>
yading@10: ... I<in_link_N>, are associated to the filter input pads,
yading@10: the following labels I<out_link_1> ... I<out_link_M>, are
yading@10: associated to the output pads.
yading@10: 
yading@10: When two link labels with the same name are found in the
yading@10: filtergraph, a link between the corresponding input and output pad is
yading@10: created.
yading@10: 
yading@10: If an output pad is not labelled, it is linked by default to the first
yading@10: unlabelled input pad of the next filter in the filterchain.
yading@10: For example in the filterchain:
yading@10: 	
yading@10: 	nullsrc, split[L1], [L2]overlay, nullsink
yading@10: 
yading@10: the split filter instance has two output pads, and the overlay filter
yading@10: instance two input pads. The first output pad of split is labelled
yading@10: "L1", the first input pad of overlay is labelled "L2", and the second
yading@10: output pad of split is linked to the second input pad of overlay,
yading@10: which are both unlabelled.
yading@10: 
yading@10: In a complete filterchain all the unlabelled filter input and output
yading@10: pads must be connected. A filtergraph is considered valid if all the
yading@10: filter input and output pads of all the filterchains are connected.
yading@10: 
yading@10: Libavfilter will automatically insert scale filters where format
yading@10: conversion is required. It is possible to specify swscale flags
yading@10: for those automatically inserted scalers by prepending
yading@10: C<sws_flags=I<flags>;>
yading@10: to the filtergraph description.
yading@10: 
yading@10: Follows a BNF description for the filtergraph syntax:
yading@10: 	
yading@10: 	<NAME>             ::= sequence of alphanumeric characters and '_'
yading@10: 	<LINKLABEL>        ::= "[" <NAME> "]"
yading@10: 	<LINKLABELS>       ::= <LINKLABEL> [<LINKLABELS>]
yading@10: 	<FILTER_ARGUMENTS> ::= sequence of chars (eventually quoted)
yading@10: 	<FILTER>           ::= [<LINKLABELS>] <NAME> ["=" <FILTER_ARGUMENTS>] [<LINKLABELS>]
yading@10: 	<FILTERCHAIN>      ::= <FILTER> [,<FILTERCHAIN>]
yading@10: 	<FILTERGRAPH>      ::= [sws_flags=<flags>;] <FILTERCHAIN> [;<FILTERGRAPH>]
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 Notes on filtergraph escaping
yading@10: 
yading@10: 
yading@10: Some filter arguments require the use of special characters, typically
yading@10: C<:> to separate key=value pairs in a named options list. In this
yading@10: case the user should perform a first level escaping when specifying
yading@10: the filter arguments. For example, consider the following literal
yading@10: string to be embedded in the drawtext filter arguments:
yading@10: 	
yading@10: 	this is a 'string': may contain one, or more, special characters
yading@10: 
yading@10: 
yading@10: Since C<:> is special for the filter arguments syntax, it needs to
yading@10: be escaped, so you get:
yading@10: 	
yading@10: 	text=this is a \'string\'\: may contain one, or more, special characters
yading@10: 
yading@10: 
yading@10: A second level of escaping is required when embedding the filter
yading@10: arguments in a filtergraph description, in order to escape all the
yading@10: filtergraph special characters. Thus the example above becomes:
yading@10: 	
yading@10: 	drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
yading@10: 
yading@10: 
yading@10: Finally an additional level of escaping may be needed when writing the
yading@10: filtergraph description in a shell command, which depends on the
yading@10: escaping rules of the adopted shell. For example, assuming that
yading@10: C<\> is special and needs to be escaped with another C<\>, the
yading@10: previous string will finally result in:
yading@10: 	
yading@10: 	-vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
yading@10: 
yading@10: 
yading@10: Sometimes, it might be more convenient to employ quoting in place of
yading@10: escaping. For example the string:
yading@10: 	
yading@10: 	Caesar: tu quoque, Brute, fili mi
yading@10: 
yading@10: 
yading@10: Can be quoted in the filter arguments as:
yading@10: 	
yading@10: 	text='Caesar: tu quoque, Brute, fili mi'
yading@10: 
yading@10: 
yading@10: And finally inserted in a filtergraph like:
yading@10: 	
yading@10: 	drawtext=text=\'Caesar: tu quoque\, Brute\, fili mi\'
yading@10: 
yading@10: 
yading@10: See the ``Quoting and escaping'' section in the ffmpeg-utils manual
yading@10: for more information about the escaping and quoting rules adopted by
yading@10: FFmpeg.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 AUDIO FILTERS
yading@10: 
yading@10: 
yading@10: When you configure your FFmpeg build, you can disable any of the
yading@10: existing filters using C<--disable-filters>.
yading@10: The configure output will show the audio filters included in your
yading@10: build.
yading@10: 
yading@10: Below is a description of the currently available audio filters.
yading@10: 
yading@10: 
yading@10: =head2 aconvert
yading@10: 
yading@10: 
yading@10: Convert the input audio format to the specified formats.
yading@10: 
yading@10: I<This filter is deprecated. Use aformat> instead.
yading@10: 
yading@10: The filter accepts a string of the form:
yading@10: "I<sample_format>:I<channel_layout>".
yading@10: 
yading@10: I<sample_format> specifies the sample format, and can be a string or the
yading@10: corresponding numeric value defined in F<libavutil/samplefmt.h>. Use 'p'
yading@10: suffix for a planar sample format.
yading@10: 
yading@10: I<channel_layout> specifies the channel layout, and can be a string
yading@10: or the corresponding number value defined in F<libavutil/channel_layout.h>.
yading@10: 
yading@10: The special parameter "auto", signifies that the filter will
yading@10: automatically select the output format depending on the output filter.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Convert input to float, planar, stereo:
yading@10: 	
yading@10: 	aconvert=fltp:stereo
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Convert input to unsigned 8-bit, automatically select out channel layout:
yading@10: 	
yading@10: 	aconvert=u8:auto
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 allpass
yading@10: 
yading@10: 
yading@10: Apply a two-pole all-pass filter with central frequency (in Hz)
yading@10: I<frequency>, and filter-width I<width>.
yading@10: An all-pass filter changes the audio's frequency to phase relationship
yading@10: without changing its frequency to amplitude relationship.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set frequency in Hz.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 highpass
yading@10: 
yading@10: 
yading@10: Apply a high-pass filter with 3dB point frequency.
yading@10: The filter can be either single-pole, or double-pole (the default).
yading@10: The filter roll off at 6dB per pole per octave (20dB per pole per decade).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set frequency in Hz. Default is 3000.
yading@10: 
yading@10: 
yading@10: =item B<poles, p>
yading@10: 
yading@10: Set number of poles. Default is 2.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: Applies only to double-pole filter.
yading@10: The default is 0.707q and gives a Butterworth response.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 lowpass
yading@10: 
yading@10: 
yading@10: Apply a low-pass filter with 3dB point frequency.
yading@10: The filter can be either single-pole or double-pole (the default).
yading@10: The filter roll off at 6dB per pole per octave (20dB per pole per decade).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set frequency in Hz. Default is 500.
yading@10: 
yading@10: 
yading@10: =item B<poles, p>
yading@10: 
yading@10: Set number of poles. Default is 2.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: Applies only to double-pole filter.
yading@10: The default is 0.707q and gives a Butterworth response.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 bass
yading@10: 
yading@10: 
yading@10: Boost or cut the bass (lower) frequencies of the audio using a two-pole
yading@10: shelving filter with a response similar to that of a standard
yading@10: hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<gain, g>
yading@10: 
yading@10: Give the gain at 0 Hz. Its useful range is about -20
yading@10: (for a large cut) to +20 (for a large boost).
yading@10: Beware of clipping when using a positive gain.
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the filter's central frequency and so can be used
yading@10: to extend or reduce the frequency range to be boosted or cut.
yading@10: The default value is C<100> Hz.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Determine how steep is the filter's shelf transition.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 treble
yading@10: 
yading@10: 
yading@10: Boost or cut treble (upper) frequencies of the audio using a two-pole
yading@10: shelving filter with a response similar to that of a standard
yading@10: hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<gain, g>
yading@10: 
yading@10: Give the gain at whichever is the lower of ~22 kHz and the
yading@10: Nyquist frequency. Its useful range is about -20 (for a large cut)
yading@10: to +20 (for a large boost). Beware of clipping when using a positive gain.
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the filter's central frequency and so can be used
yading@10: to extend or reduce the frequency range to be boosted or cut.
yading@10: The default value is C<3000> Hz.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Determine how steep is the filter's shelf transition.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 bandpass
yading@10: 
yading@10: 
yading@10: Apply a two-pole Butterworth band-pass filter with central
yading@10: frequency I<frequency>, and (3dB-point) band-width width.
yading@10: The I<csg> option selects a constant skirt gain (peak gain = Q)
yading@10: instead of the default: constant 0dB peak gain.
yading@10: The filter roll off at 6dB per octave (20dB per decade).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the filter's central frequency. Default is C<3000>.
yading@10: 
yading@10: 
yading@10: =item B<csg>
yading@10: 
yading@10: Constant skirt gain if set to 1. Defaults to 0.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 bandreject
yading@10: 
yading@10: 
yading@10: Apply a two-pole Butterworth band-reject filter with central
yading@10: frequency I<frequency>, and (3dB-point) band-width I<width>.
yading@10: The filter roll off at 6dB per octave (20dB per decade).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the filter's central frequency. Default is C<3000>.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 biquad
yading@10: 
yading@10: 
yading@10: Apply a biquad IIR filter with the given coefficients.
yading@10: Where I<b0>, I<b1>, I<b2> and I<a0>, I<a1>, I<a2>
yading@10: are the numerator and denominator coefficients respectively.
yading@10: 
yading@10: 
yading@10: =head2 equalizer
yading@10: 
yading@10: 
yading@10: Apply a two-pole peaking equalisation (EQ) filter. With this
yading@10: filter, the signal-level at and around a selected frequency can
yading@10: be increased or decreased, whilst (unlike bandpass and bandreject
yading@10: filters) that at all other frequencies is unchanged.
yading@10: 
yading@10: In order to produce complex equalisation curves, this filter can
yading@10: be given several times, each with a different central frequency.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the filter's central frequency in Hz.
yading@10: 
yading@10: 
yading@10: =item B<width_type>
yading@10: 
yading@10: Set method to specify band-width of filter.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Hz
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: Q-Factor
yading@10: 
yading@10: =item B<o>
yading@10: 
yading@10: octave
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: slope
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Specify the band-width of a filter in width_type units.
yading@10: 
yading@10: 
yading@10: =item B<gain, g>
yading@10: 
yading@10: Set the required gain or attenuation in dB.
yading@10: Beware of clipping when using a positive gain.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 afade
yading@10: 
yading@10: 
yading@10: Apply fade-in/out effect to input audio.
yading@10: 
yading@10: A description of the accepted parameters follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<type, t>
yading@10: 
yading@10: Specify the effect type, can be either C<in> for fade-in, or
yading@10: C<out> for a fade-out effect. Default is C<in>.
yading@10: 
yading@10: 
yading@10: =item B<start_sample, ss>
yading@10: 
yading@10: Specify the number of the start sample for starting to apply the fade
yading@10: effect. Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<nb_samples, ns>
yading@10: 
yading@10: Specify the number of samples for which the fade effect has to last. At
yading@10: the end of the fade-in effect the output audio will have the same
yading@10: volume as the input audio, at the end of the fade-out transition
yading@10: the output audio will be silence. Default is 44100.
yading@10: 
yading@10: 
yading@10: =item B<start_time, st>
yading@10: 
yading@10: Specify time for starting to apply the fade effect. Default is 0.
yading@10: The accepted syntax is:
yading@10: 	
yading@10: 	[-]HH[:MM[:SS[.m...]]]
yading@10: 	[-]S+[.m...]
yading@10: 
yading@10: See also the function C<av_parse_time()>.
yading@10: If set this option is used instead of I<start_sample> one.
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Specify the duration for which the fade effect has to last. Default is 0.
yading@10: The accepted syntax is:
yading@10: 	
yading@10: 	[-]HH[:MM[:SS[.m...]]]
yading@10: 	[-]S+[.m...]
yading@10: 
yading@10: See also the function C<av_parse_time()>.
yading@10: At the end of the fade-in effect the output audio will have the same
yading@10: volume as the input audio, at the end of the fade-out transition
yading@10: the output audio will be silence.
yading@10: If set this option is used instead of I<nb_samples> one.
yading@10: 
yading@10: 
yading@10: =item B<curve>
yading@10: 
yading@10: Set curve for fade transition.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<tri>
yading@10: 
yading@10: select triangular, linear slope (default)
yading@10: 
yading@10: =item B<qsin>
yading@10: 
yading@10: select quarter of sine wave
yading@10: 
yading@10: =item B<hsin>
yading@10: 
yading@10: select half of sine wave
yading@10: 
yading@10: =item B<esin>
yading@10: 
yading@10: select exponential sine wave
yading@10: 
yading@10: =item B<log>
yading@10: 
yading@10: select logarithmic
yading@10: 
yading@10: =item B<par>
yading@10: 
yading@10: select inverted parabola
yading@10: 
yading@10: =item B<qua>
yading@10: 
yading@10: select quadratic
yading@10: 
yading@10: =item B<cub>
yading@10: 
yading@10: select cubic
yading@10: 
yading@10: =item B<squ>
yading@10: 
yading@10: select square root
yading@10: 
yading@10: =item B<cbr>
yading@10: 
yading@10: select cubic root
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade in first 15 seconds of audio:
yading@10: 	
yading@10: 	afade=t=in:ss=0:d=15
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade out last 25 seconds of a 900 seconds audio:
yading@10: 	
yading@10: 	afade=t=out:ss=875:d=25
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 aformat
yading@10: 
yading@10: 
yading@10: Set output format constraints for the input audio. The framework will
yading@10: negotiate the most appropriate format to minimize conversions.
yading@10: 
yading@10: The filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<sample_fmts>
yading@10: 
yading@10: A '|'-separated list of requested sample formats.
yading@10: 
yading@10: 
yading@10: =item B<sample_rates>
yading@10: 
yading@10: A '|'-separated list of requested sample rates.
yading@10: 
yading@10: 
yading@10: =item B<channel_layouts>
yading@10: 
yading@10: A '|'-separated list of requested channel layouts.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If a parameter is omitted, all values are allowed.
yading@10: 
yading@10: For example to force the output to either unsigned 8-bit or signed 16-bit stereo:
yading@10: 	
yading@10: 	aformat=sample_fmts=u8|s16:channel_layouts=stereo
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 amerge
yading@10: 
yading@10: 
yading@10: Merge two or more audio streams into a single multi-channel stream.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<inputs>
yading@10: 
yading@10: Set the number of inputs. Default is 2.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If the channel layouts of the inputs are disjoint, and therefore compatible,
yading@10: the channel layout of the output will be set accordingly and the channels
yading@10: will be reordered as necessary. If the channel layouts of the inputs are not
yading@10: disjoint, the output will have all the channels of the first input then all
yading@10: the channels of the second input, in that order, and the channel layout of
yading@10: the output will be the default value corresponding to the total number of
yading@10: channels.
yading@10: 
yading@10: For example, if the first input is in 2.1 (FL+FR+LF) and the second input
yading@10: is FC+BL+BR, then the output will be in 5.1, with the channels in the
yading@10: following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
yading@10: first input, b1 is the first channel of the second input).
yading@10: 
yading@10: On the other hand, if both input are in stereo, the output channels will be
yading@10: in the default order: a1, a2, b1, b2, and the channel layout will be
yading@10: arbitrarily set to 4.0, which may or may not be the expected value.
yading@10: 
yading@10: All inputs must have the same sample rate, and format.
yading@10: 
yading@10: If inputs do not have the same duration, the output will stop with the
yading@10: shortest.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Merge two mono files into a stereo stream:
yading@10: 	
yading@10: 	amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Multiple merges assuming 1 video stream and 6 audio streams in F<input.mkv>:
yading@10: 	
yading@10: 	ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 amix
yading@10: 
yading@10: 
yading@10: Mixes multiple audio inputs into a single output.
yading@10: 
yading@10: For example
yading@10: 	
yading@10: 	ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
yading@10: 
yading@10: will mix 3 input audio streams to a single output with the same duration as the
yading@10: first input and a dropout transition time of 3 seconds.
yading@10: 
yading@10: The filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<inputs>
yading@10: 
yading@10: Number of inputs. If unspecified, it defaults to 2.
yading@10: 
yading@10: 
yading@10: =item B<duration>
yading@10: 
yading@10: How to determine the end-of-stream.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<longest>
yading@10: 
yading@10: Duration of longest input. (default)
yading@10: 
yading@10: 
yading@10: =item B<shortest>
yading@10: 
yading@10: Duration of shortest input.
yading@10: 
yading@10: 
yading@10: =item B<first>
yading@10: 
yading@10: Duration of first input.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<dropout_transition>
yading@10: 
yading@10: Transition time, in seconds, for volume renormalization when an input
yading@10: stream ends. The default value is 2 seconds.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 anull
yading@10: 
yading@10: 
yading@10: Pass the audio source unchanged to the output.
yading@10: 
yading@10: 
yading@10: =head2 apad
yading@10: 
yading@10: 
yading@10: Pad the end of a audio stream with silence, this can be used together with
yading@10: -shortest to extend audio streams to the same length as the video stream.
yading@10: 
yading@10: 
yading@10: =head2 aphaser
yading@10: 
yading@10: Add a phasing effect to the input audio.
yading@10: 
yading@10: A phaser filter creates series of peaks and troughs in the frequency spectrum.
yading@10: The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
yading@10: 
yading@10: A description of the accepted parameters follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<in_gain>
yading@10: 
yading@10: Set input gain. Default is 0.4.
yading@10: 
yading@10: 
yading@10: =item B<out_gain>
yading@10: 
yading@10: Set output gain. Default is 0.74
yading@10: 
yading@10: 
yading@10: =item B<delay>
yading@10: 
yading@10: Set delay in milliseconds. Default is 3.0.
yading@10: 
yading@10: 
yading@10: =item B<decay>
yading@10: 
yading@10: Set decay. Default is 0.4.
yading@10: 
yading@10: 
yading@10: =item B<speed>
yading@10: 
yading@10: Set modulation speed in Hz. Default is 0.5.
yading@10: 
yading@10: 
yading@10: =item B<type>
yading@10: 
yading@10: Set modulation type. Default is triangular.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<triangular, t>
yading@10: 
yading@10: 
yading@10: =item B<sinusoidal, s>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 aresample
yading@10: 
yading@10: 
yading@10: Resample the input audio to the specified parameters, using the
yading@10: libswresample library. If none are specified then the filter will
yading@10: automatically convert between its input and output.
yading@10: 
yading@10: This filter is also able to stretch/squeeze the audio data to make it match
yading@10: the timestamps or to inject silence / cut out audio to make it match the
yading@10: timestamps, do a combination of both or do neither.
yading@10: 
yading@10: The filter accepts the syntax
yading@10: [I<sample_rate>:]I<resampler_options>, where I<sample_rate>
yading@10: expresses a sample rate and I<resampler_options> is a list of
yading@10: I<key>=I<value> pairs, separated by ":". See the
yading@10: ffmpeg-resampler manual for the complete list of supported options.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Resample the input audio to 44100Hz:
yading@10: 	
yading@10: 	aresample=44100
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Stretch/squeeze samples to the given timestamps, with a maximum of 1000
yading@10: samples per second compensation:
yading@10: 	
yading@10: 	aresample=async=1000
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 asetnsamples
yading@10: 
yading@10: 
yading@10: Set the number of samples per each output audio frame.
yading@10: 
yading@10: The last output packet may contain a different number of samples, as
yading@10: the filter will flush all the remaining samples when the input audio
yading@10: signal its end.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<nb_out_samples, n>
yading@10: 
yading@10: Set the number of frames per each output audio frame. The number is
yading@10: intended as the number of samples I<per each channel>.
yading@10: Default value is 1024.
yading@10: 
yading@10: 
yading@10: =item B<pad, p>
yading@10: 
yading@10: If set to 1, the filter will pad the last audio frame with zeroes, so
yading@10: that the last frame will contain the same number of samples as the
yading@10: previous ones. Default value is 1.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example, to set the number of per-frame samples to 1234 and
yading@10: disable padding for the last frame, use:
yading@10: 	
yading@10: 	asetnsamples=n=1234:p=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 ashowinfo
yading@10: 
yading@10: 
yading@10: Show a line containing various information for each input audio frame.
yading@10: The input audio is not modified.
yading@10: 
yading@10: The shown line contains a sequence of key/value pairs of the form
yading@10: I<key>:I<value>.
yading@10: 
yading@10: A description of each shown parameter follows:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: sequential number of the input frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<pts>
yading@10: 
yading@10: Presentation timestamp of the input frame, in time base units; the time base
yading@10: depends on the filter input pad, and is usually 1/I<sample_rate>.
yading@10: 
yading@10: 
yading@10: =item B<pts_time>
yading@10: 
yading@10: presentation timestamp of the input frame in seconds
yading@10: 
yading@10: 
yading@10: =item B<pos>
yading@10: 
yading@10: position of the frame in the input stream, -1 if this information in
yading@10: unavailable and/or meaningless (for example in case of synthetic audio)
yading@10: 
yading@10: 
yading@10: =item B<fmt>
yading@10: 
yading@10: sample format
yading@10: 
yading@10: 
yading@10: =item B<chlayout>
yading@10: 
yading@10: channel layout
yading@10: 
yading@10: 
yading@10: =item B<rate>
yading@10: 
yading@10: sample rate for the audio frame
yading@10: 
yading@10: 
yading@10: =item B<nb_samples>
yading@10: 
yading@10: number of samples (per channel) in the frame
yading@10: 
yading@10: 
yading@10: =item B<checksum>
yading@10: 
yading@10: Adler-32 checksum (printed in hexadecimal) of the audio data. For planar audio
yading@10: the data is treated as if all the planes were concatenated.
yading@10: 
yading@10: 
yading@10: =item B<plane_checksums>
yading@10: 
yading@10: A list of Adler-32 checksums for each data plane.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 astreamsync
yading@10: 
yading@10: 
yading@10: Forward two audio streams and control the order the buffers are forwarded.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<expr, e>
yading@10: 
yading@10: Set the expression deciding which stream should be
yading@10: forwarded next: if the result is negative, the first stream is forwarded; if
yading@10: the result is positive or zero, the second stream is forwarded. It can use
yading@10: the following variables:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item I<b1 b2>
yading@10: 
yading@10: number of buffers forwarded so far on each stream
yading@10: 
yading@10: =item I<s1 s2>
yading@10: 
yading@10: number of samples forwarded so far on each stream
yading@10: 
yading@10: =item I<t1 t2>
yading@10: 
yading@10: current timestamp of each stream
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The default value is C<t1-t2>, which means to always forward the stream
yading@10: that has a smaller timestamp.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: Stress-test C<amerge> by randomly sending buffers on the wrong
yading@10: input, while avoiding too much of a desynchronization:
yading@10: 	
yading@10: 	amovie=file.ogg [a] ; amovie=file.mp3 [b] ;
yading@10: 	[a] [b] astreamsync=(2*random(1))-1+tanh(5*(t1-t2)) [a2] [b2] ;
yading@10: 	[a2] [b2] amerge
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 atempo
yading@10: 
yading@10: 
yading@10: Adjust audio tempo.
yading@10: 
yading@10: The filter accepts exactly one parameter, the audio tempo. If not
yading@10: specified then the filter will assume nominal 1.0 tempo. Tempo must
yading@10: be in the [0.5, 2.0] range.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Slow down audio to 80% tempo:
yading@10: 	
yading@10: 	atempo=0.8
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: To speed up audio to 125% tempo:
yading@10: 	
yading@10: 	atempo=1.25
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 earwax
yading@10: 
yading@10: 
yading@10: Make audio easier to listen to on headphones.
yading@10: 
yading@10: This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
yading@10: so that when listened to on headphones the stereo image is moved from
yading@10: inside your head (standard for headphones) to outside and in front of
yading@10: the listener (standard for speakers).
yading@10: 
yading@10: Ported from SoX.
yading@10: 
yading@10: 
yading@10: =head2 pan
yading@10: 
yading@10: 
yading@10: Mix channels with specific gain levels. The filter accepts the output
yading@10: channel layout followed by a set of channels definitions.
yading@10: 
yading@10: This filter is also designed to remap efficiently the channels of an audio
yading@10: stream.
yading@10: 
yading@10: The filter accepts parameters of the form:
yading@10: "I<l>:I<outdef>:I<outdef>:..."
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<l>
yading@10: 
yading@10: output channel layout or number of channels
yading@10: 
yading@10: 
yading@10: =item B<outdef>
yading@10: 
yading@10: output channel specification, of the form:
yading@10: "I<out_name>=[I<gain>*]I<in_name>[+[I<gain>*]I<in_name>...]"
yading@10: 
yading@10: 
yading@10: =item B<out_name>
yading@10: 
yading@10: output channel to define, either a channel name (FL, FR, etc.) or a channel
yading@10: number (c0, c1, etc.)
yading@10: 
yading@10: 
yading@10: =item B<gain>
yading@10: 
yading@10: multiplicative coefficient for the channel, 1 leaving the volume unchanged
yading@10: 
yading@10: 
yading@10: =item B<in_name>
yading@10: 
yading@10: input channel to use, see out_name for details; it is not possible to mix
yading@10: named and numbered input channels
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If the `=' in a channel specification is replaced by `E<lt>', then the gains for
yading@10: that specification will be renormalized so that the total is 1, thus
yading@10: avoiding clipping noise.
yading@10: 
yading@10: 
yading@10: =head3 Mixing examples
yading@10: 
yading@10: 
yading@10: For example, if you want to down-mix from stereo to mono, but with a bigger
yading@10: factor for the left channel:
yading@10: 	
yading@10: 	pan=1:c0=0.9*c0+0.1*c1
yading@10: 
yading@10: 
yading@10: A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
yading@10: 7-channels surround:
yading@10: 	
yading@10: 	pan=stereo: FL < FL + 0.5*FC + 0.6*BL + 0.6*SL : FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
yading@10: 
yading@10: 
yading@10: Note that B<ffmpeg> integrates a default down-mix (and up-mix) system
yading@10: that should be preferred (see "-ac" option) unless you have very specific
yading@10: needs.
yading@10: 
yading@10: 
yading@10: =head3 Remapping examples
yading@10: 
yading@10: 
yading@10: The channel remapping will be effective if, and only if:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *<gain coefficients are zeroes or ones,>
yading@10: 
yading@10: 
yading@10: =item *<only one input per channel output,>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If all these conditions are satisfied, the filter will notify the user ("Pure
yading@10: channel mapping detected"), and use an optimized and lossless method to do the
yading@10: remapping.
yading@10: 
yading@10: For example, if you have a 5.1 source and want a stereo audio stream by
yading@10: dropping the extra channels:
yading@10: 	
yading@10: 	pan="stereo: c0=FL : c1=FR"
yading@10: 
yading@10: 
yading@10: Given the same source, you can also switch front left and front right channels
yading@10: and keep the input channel layout:
yading@10: 	
yading@10: 	pan="5.1: c0=c1 : c1=c0 : c2=c2 : c3=c3 : c4=c4 : c5=c5"
yading@10: 
yading@10: 
yading@10: If the input is a stereo audio stream, you can mute the front left channel (and
yading@10: still keep the stereo channel layout) with:
yading@10: 	
yading@10: 	pan="stereo:c1=c1"
yading@10: 
yading@10: 
yading@10: Still with a stereo audio stream input, you can copy the right channel in both
yading@10: front left and right:
yading@10: 	
yading@10: 	pan="stereo: c0=FR : c1=FR"
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 silencedetect
yading@10: 
yading@10: 
yading@10: Detect silence in an audio stream.
yading@10: 
yading@10: This filter logs a message when it detects that the input audio volume is less
yading@10: or equal to a noise tolerance value for a duration greater or equal to the
yading@10: minimum detected noise duration.
yading@10: 
yading@10: The printed times and duration are expressed in seconds.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Set silence duration until notification (default is 2 seconds).
yading@10: 
yading@10: 
yading@10: =item B<noise, n>
yading@10: 
yading@10: Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
yading@10: specified value) or amplitude ratio. Default is -60dB, or 0.001.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Detect 5 seconds of silence with -50dB noise tolerance:
yading@10: 	
yading@10: 	silencedetect=n=-50dB:d=5
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Complete example with B<ffmpeg> to detect silence with 0.0001 noise
yading@10: tolerance in F<silence.mp3>:
yading@10: 	
yading@10: 	ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 asyncts
yading@10: 
yading@10: Synchronize audio data with timestamps by squeezing/stretching it and/or
yading@10: dropping samples/adding silence when needed.
yading@10: 
yading@10: This filter is not built by default, please use aresample to do squeezing/stretching.
yading@10: 
yading@10: The filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<compensate>
yading@10: 
yading@10: Enable stretching/squeezing the data to make it match the timestamps. Disabled
yading@10: by default. When disabled, time gaps are covered with silence.
yading@10: 
yading@10: 
yading@10: =item B<min_delta>
yading@10: 
yading@10: Minimum difference between timestamps and audio data (in seconds) to trigger
yading@10: adding/dropping samples. Default value is 0.1. If you get non-perfect sync with
yading@10: this filter, try setting this parameter to 0.
yading@10: 
yading@10: 
yading@10: =item B<max_comp>
yading@10: 
yading@10: Maximum compensation in samples per second. Relevant only with compensate=1.
yading@10: Default value 500.
yading@10: 
yading@10: 
yading@10: =item B<first_pts>
yading@10: 
yading@10: Assume the first pts should be this value. The time base is 1 / sample rate.
yading@10: This allows for padding/trimming at the start of stream. By default, no
yading@10: assumption is made about the first frame's expected pts, so no padding or
yading@10: trimming is done. For example, this could be set to 0 to pad the beginning with
yading@10: silence if an audio stream starts after the video stream or to trim any samples
yading@10: with a negative pts due to encoder delay.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 channelsplit
yading@10: 
yading@10: Split each channel in input audio stream into a separate output stream.
yading@10: 
yading@10: This filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<channel_layout>
yading@10: 
yading@10: Channel layout of the input stream. Default is "stereo".
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example, assuming a stereo input MP3 file
yading@10: 	
yading@10: 	ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
yading@10: 
yading@10: will create an output Matroska file with two audio streams, one containing only
yading@10: the left channel and the other the right channel.
yading@10: 
yading@10: To split a 5.1 WAV file into per-channel files
yading@10: 	
yading@10: 	ffmpeg -i in.wav -filter_complex
yading@10: 	'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
yading@10: 	-map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
yading@10: 	front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
yading@10: 	side_right.wav
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 channelmap
yading@10: 
yading@10: Remap input channels to new locations.
yading@10: 
yading@10: This filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<channel_layout>
yading@10: 
yading@10: Channel layout of the output stream.
yading@10: 
yading@10: 
yading@10: =item B<map>
yading@10: 
yading@10: Map channels from input to output. The argument is a '|'-separated list of
yading@10: mappings, each in the C<I<in_channel>-I<out_channel>> or
yading@10: I<in_channel> form. I<in_channel> can be either the name of the input
yading@10: channel (e.g. FL for front left) or its index in the input channel layout.
yading@10: I<out_channel> is the name of the output channel or its index in the output
yading@10: channel layout. If I<out_channel> is not given then it is implicitly an
yading@10: index, starting with zero and increasing by one for each mapping.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If no mapping is present, the filter will implicitly map input channels to
yading@10: output channels preserving index.
yading@10: 
yading@10: For example, assuming a 5.1+downmix input MOV file
yading@10: 	
yading@10: 	ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
yading@10: 
yading@10: will create an output WAV file tagged as stereo from the downmix channels of
yading@10: the input.
yading@10: 
yading@10: To fix a 5.1 WAV improperly encoded in AAC's native channel order
yading@10: 	
yading@10: 	ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:channel_layout=5.1' out.wav
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 join
yading@10: 
yading@10: Join multiple input streams into one multi-channel stream.
yading@10: 
yading@10: The filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<inputs>
yading@10: 
yading@10: Number of input streams. Defaults to 2.
yading@10: 
yading@10: 
yading@10: =item B<channel_layout>
yading@10: 
yading@10: Desired output channel layout. Defaults to stereo.
yading@10: 
yading@10: 
yading@10: =item B<map>
yading@10: 
yading@10: Map channels from inputs to output. The argument is a '|'-separated list of
yading@10: mappings, each in the C<I<input_idx>.I<in_channel>-I<out_channel>>
yading@10: form. I<input_idx> is the 0-based index of the input stream. I<in_channel>
yading@10: can be either the name of the input channel (e.g. FL for front left) or its
yading@10: index in the specified input stream. I<out_channel> is the name of the output
yading@10: channel.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The filter will attempt to guess the mappings when those are not specified
yading@10: explicitly. It does so by first trying to find an unused matching input channel
yading@10: and if that fails it picks the first unused input channel.
yading@10: 
yading@10: E.g. to join 3 inputs (with properly set channel layouts)
yading@10: 	
yading@10: 	ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
yading@10: 
yading@10: 
yading@10: To build a 5.1 output from 6 single-channel streams:
yading@10: 	
yading@10: 	ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
yading@10: 	'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'
yading@10: 	out
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 resample
yading@10: 
yading@10: Convert the audio sample format, sample rate and channel layout. This filter is
yading@10: not meant to be used directly.
yading@10: 
yading@10: 
yading@10: =head2 volume
yading@10: 
yading@10: 
yading@10: Adjust the input audio volume.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<volume>
yading@10: 
yading@10: Expresses how the audio volume will be increased or decreased.
yading@10: 
yading@10: Output values are clipped to the maximum value.
yading@10: 
yading@10: The output audio volume is given by the relation:
yading@10: 	
yading@10: 	<output_volume> = <volume> * <input_volume>
yading@10: 
yading@10: 
yading@10: Default value for I<volume> is 1.0.
yading@10: 
yading@10: 
yading@10: =item B<precision>
yading@10: 
yading@10: Set the mathematical precision.
yading@10: 
yading@10: This determines which input sample formats will be allowed, which affects the
yading@10: precision of the volume scaling.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<fixed>
yading@10: 
yading@10: 8-bit fixed-point; limits input sample format to U8, S16, and S32.
yading@10: 
yading@10: =item B<float>
yading@10: 
yading@10: 32-bit floating-point; limits input sample format to FLT. (default)
yading@10: 
yading@10: =item B<double>
yading@10: 
yading@10: 64-bit floating-point; limits input sample format to DBL.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Halve the input audio volume:
yading@10: 	
yading@10: 	volume=volume=0.5
yading@10: 	volume=volume=1/2
yading@10: 	volume=volume=-6.0206dB
yading@10: 
yading@10: 
yading@10: In all the above example the named key for B<volume> can be
yading@10: omitted, for example like in:
yading@10: 	
yading@10: 	volume=0.5
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase input audio power by 6 decibels using fixed-point precision:
yading@10: 	
yading@10: 	volume=volume=6dB:precision=fixed
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 volumedetect
yading@10: 
yading@10: 
yading@10: Detect the volume of the input video.
yading@10: 
yading@10: The filter has no parameters. The input is not modified. Statistics about
yading@10: the volume will be printed in the log when the input stream end is reached.
yading@10: 
yading@10: In particular it will show the mean volume (root mean square), maximum
yading@10: volume (on a per-sample basis), and the beginning of an histogram of the
yading@10: registered volume values (from the maximum value to a cumulated 1/1000 of
yading@10: the samples).
yading@10: 
yading@10: All volumes are in decibels relative to the maximum PCM value.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: Here is an excerpt of the output:
yading@10: 	
yading@10: 	[Parsed_volumedetect_0  0xa23120] mean_volume: -27 dB
yading@10: 	[Parsed_volumedetect_0  0xa23120] max_volume: -4 dB
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_4db: 6
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_5db: 62
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_6db: 286
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_7db: 1042
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_8db: 2551
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_9db: 4609
yading@10: 	[Parsed_volumedetect_0  0xa23120] histogram_10db: 8409
yading@10: 
yading@10: 
yading@10: It means that:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: The mean square energy is approximately -27 dB, or 10^-2.7.
yading@10: 
yading@10: =item *
yading@10: 
yading@10: The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
yading@10: 
yading@10: =item *
yading@10: 
yading@10: There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: In other words, raising the volume by +4 dB does not cause any clipping,
yading@10: raising it by +5 dB causes clipping for 6 samples, etc.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 AUDIO SOURCES
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available audio sources.
yading@10: 
yading@10: 
yading@10: =head2 abuffer
yading@10: 
yading@10: 
yading@10: Buffer audio frames, and make them available to the filter chain.
yading@10: 
yading@10: This source is mainly intended for a programmatic use, in particular
yading@10: through the interface defined in F<libavfilter/asrc_abuffer.h>.
yading@10: 
yading@10: It accepts the following named parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<time_base>
yading@10: 
yading@10: Timebase which will be used for timestamps of submitted frames. It must be
yading@10: either a floating-point number or in I<numerator>/I<denominator> form.
yading@10: 
yading@10: 
yading@10: =item B<sample_rate>
yading@10: 
yading@10: The sample rate of the incoming audio buffers.
yading@10: 
yading@10: 
yading@10: =item B<sample_fmt>
yading@10: 
yading@10: The sample format of the incoming audio buffers.
yading@10: Either a sample format name or its corresponging integer representation from
yading@10: the enum AVSampleFormat in F<libavutil/samplefmt.h>
yading@10: 
yading@10: 
yading@10: =item B<channel_layout>
yading@10: 
yading@10: The channel layout of the incoming audio buffers.
yading@10: Either a channel layout name from channel_layout_map in
yading@10: F<libavutil/channel_layout.c> or its corresponding integer representation
yading@10: from the AV_CH_LAYOUT_* macros in F<libavutil/channel_layout.h>
yading@10: 
yading@10: 
yading@10: =item B<channels>
yading@10: 
yading@10: The number of channels of the incoming audio buffers.
yading@10: If both I<channels> and I<channel_layout> are specified, then they
yading@10: must be consistent.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 	
yading@10: 	abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
yading@10: 
yading@10: 
yading@10: will instruct the source to accept planar 16bit signed stereo at 44100Hz.
yading@10: Since the sample format with name "s16p" corresponds to the number
yading@10: 6 and the "stereo" channel layout corresponds to the value 0x3, this is
yading@10: equivalent to:
yading@10: 	
yading@10: 	abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 aevalsrc
yading@10: 
yading@10: 
yading@10: Generate an audio signal specified by an expression.
yading@10: 
yading@10: This source accepts in input one or more expressions (one for each
yading@10: channel), which are evaluated and used to generate a corresponding
yading@10: audio signal.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<exprs>
yading@10: 
yading@10: Set the '|'-separated expressions list for each separate channel. In case the
yading@10: B<channel_layout> option is not specified, the selected channel layout
yading@10: depends on the number of provided expressions.
yading@10: 
yading@10: 
yading@10: =item B<channel_layout, c>
yading@10: 
yading@10: Set the channel layout. The number of channels in the specified layout
yading@10: must be equal to the number of specified expressions.
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Set the minimum duration of the sourced audio. See the function
yading@10: C<av_parse_time()> for the accepted format.
yading@10: Note that the resulting duration may be greater than the specified
yading@10: duration, as the generated audio is always cut at the end of a
yading@10: complete frame.
yading@10: 
yading@10: If not specified, or the expressed duration is negative, the audio is
yading@10: supposed to be generated forever.
yading@10: 
yading@10: 
yading@10: =item B<nb_samples, n>
yading@10: 
yading@10: Set the number of samples per channel per each output frame,
yading@10: default to 1024.
yading@10: 
yading@10: 
yading@10: =item B<sample_rate, s>
yading@10: 
yading@10: Specify the sample rate, default to 44100.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Each expression in I<exprs> can contain the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: number of the evaluated sample, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: time of the evaluated sample expressed in seconds, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: sample rate
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate silence:
yading@10: 	
yading@10: 	aevalsrc=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a sin signal with frequency of 440 Hz, set sample rate to
yading@10: 8000 Hz:
yading@10: 	
yading@10: 	aevalsrc="sin(440*2*PI*t):s=8000"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a two channels signal, specify the channel layout (Front
yading@10: Center + Back Center) explicitly:
yading@10: 	
yading@10: 	aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate white noise:
yading@10: 	
yading@10: 	aevalsrc="-2+random(0)"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate an amplitude modulated signal:
yading@10: 	
yading@10: 	aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate 2.5 Hz binaural beats on a 360 Hz carrier:
yading@10: 	
yading@10: 	aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 anullsrc
yading@10: 
yading@10: 
yading@10: Null audio source, return unprocessed audio frames. It is mainly useful
yading@10: as a template and to be employed in analysis / debugging tools, or as
yading@10: the source for filters which ignore the input data (for example the sox
yading@10: synth filter).
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<channel_layout, cl>
yading@10: 
yading@10: 
yading@10: Specify the channel layout, and can be either an integer or a string
yading@10: representing a channel layout. The default value of I<channel_layout>
yading@10: is "stereo".
yading@10: 
yading@10: Check the channel_layout_map definition in
yading@10: F<libavutil/channel_layout.c> for the mapping between strings and
yading@10: channel layout values.
yading@10: 
yading@10: 
yading@10: =item B<sample_rate, r>
yading@10: 
yading@10: Specify the sample rate, and defaults to 44100.
yading@10: 
yading@10: 
yading@10: =item B<nb_samples, n>
yading@10: 
yading@10: Set the number of samples per requested frames.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
yading@10: 	
yading@10: 	anullsrc=r=48000:cl=4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Do the same operation with a more obvious syntax:
yading@10: 	
yading@10: 	anullsrc=r=48000:cl=mono
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 abuffer
yading@10: 
yading@10: Buffer audio frames, and make them available to the filter chain.
yading@10: 
yading@10: This source is not intended to be part of user-supplied graph descriptions but
yading@10: for insertion by calling programs through the interface defined in
yading@10: F<libavfilter/buffersrc.h>.
yading@10: 
yading@10: It accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<time_base>
yading@10: 
yading@10: Timebase which will be used for timestamps of submitted frames. It must be
yading@10: either a floating-point number or in I<numerator>/I<denominator> form.
yading@10: 
yading@10: 
yading@10: =item B<sample_rate>
yading@10: 
yading@10: Audio sample rate.
yading@10: 
yading@10: 
yading@10: =item B<sample_fmt>
yading@10: 
yading@10: Name of the sample format, as returned by C<av_get_sample_fmt_name()>.
yading@10: 
yading@10: 
yading@10: =item B<channel_layout>
yading@10: 
yading@10: Channel layout of the audio data, in the form that can be accepted by
yading@10: C<av_get_channel_layout()>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: All the parameters need to be explicitly defined.
yading@10: 
yading@10: 
yading@10: =head2 flite
yading@10: 
yading@10: 
yading@10: Synthesize a voice utterance using the libflite library.
yading@10: 
yading@10: To enable compilation of this filter you need to configure FFmpeg with
yading@10: C<--enable-libflite>.
yading@10: 
yading@10: Note that the flite library is not thread-safe.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<list_voices>
yading@10: 
yading@10: If set to 1, list the names of the available voices and exit
yading@10: immediately. Default value is 0.
yading@10: 
yading@10: 
yading@10: =item B<nb_samples, n>
yading@10: 
yading@10: Set the maximum number of samples per frame. Default value is 512.
yading@10: 
yading@10: 
yading@10: =item B<textfile>
yading@10: 
yading@10: Set the filename containing the text to speak.
yading@10: 
yading@10: 
yading@10: =item B<text>
yading@10: 
yading@10: Set the text to speak.
yading@10: 
yading@10: 
yading@10: =item B<voice, v>
yading@10: 
yading@10: Set the voice to use for the speech synthesis. Default value is
yading@10: C<kal>. See also the I<list_voices> option.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read from file F<speech.txt>, and synthetize the text using the
yading@10: standard flite voice:
yading@10: 	
yading@10: 	flite=textfile=speech.txt
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read the specified text selecting the C<slt> voice:
yading@10: 	
yading@10: 	flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Input text to ffmpeg:
yading@10: 	
yading@10: 	ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Make F<ffplay> speak the specified text, using C<flite> and
yading@10: the C<lavfi> device:
yading@10: 	
yading@10: 	ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For more information about libflite, check:
yading@10: E<lt>B<http://www.speech.cs.cmu.edu/flite/>E<gt>
yading@10: 
yading@10: 
yading@10: =head2 sine
yading@10: 
yading@10: 
yading@10: Generate an audio signal made of a sine wave with amplitude 1/8.
yading@10: 
yading@10: The audio signal is bit-exact.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<frequency, f>
yading@10: 
yading@10: Set the carrier frequency. Default is 440 Hz.
yading@10: 
yading@10: 
yading@10: =item B<beep_factor, b>
yading@10: 
yading@10: Enable a periodic beep every second with frequency I<beep_factor> times
yading@10: the carrier frequency. Default is 0, meaning the beep is disabled.
yading@10: 
yading@10: 
yading@10: =item B<sample_rate, s>
yading@10: 
yading@10: Specify the sample rate, default is 44100.
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Specify the duration of the generated audio stream.
yading@10: 
yading@10: 
yading@10: =item B<samples_per_frame>
yading@10: 
yading@10: Set the number of samples per output frame, default is 1024.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a simple 440 Hz sine wave:
yading@10: 	
yading@10: 	sine
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
yading@10: 	
yading@10: 	sine=220:4:d=5
yading@10: 	sine=f=220:b=4:d=5
yading@10: 	sine=frequency=220:beep_factor=4:duration=5
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 AUDIO SINKS
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available audio sinks.
yading@10: 
yading@10: 
yading@10: =head2 abuffersink
yading@10: 
yading@10: 
yading@10: Buffer audio frames, and make them available to the end of filter chain.
yading@10: 
yading@10: This sink is mainly intended for programmatic use, in particular
yading@10: through the interface defined in F<libavfilter/buffersink.h>
yading@10: or the options system.
yading@10: 
yading@10: It accepts a pointer to an AVABufferSinkContext structure, which
yading@10: defines the incoming buffers' formats, to be passed as the opaque
yading@10: parameter to C<avfilter_init_filter> for initialization.
yading@10: 
yading@10: 
yading@10: =head2 anullsink
yading@10: 
yading@10: 
yading@10: Null audio sink, do absolutely nothing with the input audio. It is
yading@10: mainly useful as a template and to be employed in analysis / debugging
yading@10: tools.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 VIDEO FILTERS
yading@10: 
yading@10: 
yading@10: When you configure your FFmpeg build, you can disable any of the
yading@10: existing filters using C<--disable-filters>.
yading@10: The configure output will show the video filters included in your
yading@10: build.
yading@10: 
yading@10: Below is a description of the currently available video filters.
yading@10: 
yading@10: 
yading@10: =head2 alphaextract
yading@10: 
yading@10: 
yading@10: Extract the alpha component from the input as a grayscale video. This
yading@10: is especially useful with the I<alphamerge> filter.
yading@10: 
yading@10: 
yading@10: =head2 alphamerge
yading@10: 
yading@10: 
yading@10: Add or replace the alpha component of the primary input with the
yading@10: grayscale value of a second input. This is intended for use with
yading@10: I<alphaextract> to allow the transmission or storage of frame
yading@10: sequences that have alpha in a format that doesn't support an alpha
yading@10: channel.
yading@10: 
yading@10: For example, to reconstruct full frames from a normal YUV-encoded video
yading@10: and a separate video created with I<alphaextract>, you might use:
yading@10: 	
yading@10: 	movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
yading@10: 
yading@10: 
yading@10: Since this filter is designed for reconstruction, it operates on frame
yading@10: sequences without considering timestamps, and terminates when either
yading@10: input reaches end of stream. This will cause problems if your encoding
yading@10: pipeline drops frames. If you're trying to apply an image as an
yading@10: overlay to a video stream, consider the I<overlay> filter instead.
yading@10: 
yading@10: 
yading@10: =head2 ass
yading@10: 
yading@10: 
yading@10: Same as the subtitles filter, except that it doesn't require libavcodec
yading@10: and libavformat to work. On the other hand, it is limited to ASS (Advanced
yading@10: Substation Alpha) subtitles files.
yading@10: 
yading@10: 
yading@10: =head2 bbox
yading@10: 
yading@10: 
yading@10: Compute the bounding box for the non-black pixels in the input frame
yading@10: luminance plane.
yading@10: 
yading@10: This filter computes the bounding box containing all the pixels with a
yading@10: luminance value greater than the minimum allowed value.
yading@10: The parameters describing the bounding box are printed on the filter
yading@10: log.
yading@10: 
yading@10: 
yading@10: =head2 blackdetect
yading@10: 
yading@10: 
yading@10: Detect video intervals that are (almost) completely black. Can be
yading@10: useful to detect chapter transitions, commercials, or invalid
yading@10: recordings. Output lines contains the time for the start, end and
yading@10: duration of the detected black interval expressed in seconds.
yading@10: 
yading@10: In order to display the output lines, you need to set the loglevel at
yading@10: least to the AV_LOG_INFO value.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<black_min_duration, d>
yading@10: 
yading@10: Set the minimum detected black duration expressed in seconds. It must
yading@10: be a non-negative floating point number.
yading@10: 
yading@10: Default value is 2.0.
yading@10: 
yading@10: 
yading@10: =item B<picture_black_ratio_th, pic_th>
yading@10: 
yading@10: Set the threshold for considering a picture "black".
yading@10: Express the minimum value for the ratio:
yading@10: 	
yading@10: 	<nb_black_pixels> / <nb_pixels>
yading@10: 
yading@10: 
yading@10: for which a picture is considered black.
yading@10: Default value is 0.98.
yading@10: 
yading@10: 
yading@10: =item B<pixel_black_th, pix_th>
yading@10: 
yading@10: Set the threshold for considering a pixel "black".
yading@10: 
yading@10: The threshold expresses the maximum pixel luminance value for which a
yading@10: pixel is considered "black". The provided value is scaled according to
yading@10: the following equation:
yading@10: 	
yading@10: 	<absolute_threshold> = <luminance_minimum_value> + <pixel_black_th> * <luminance_range_size>
yading@10: 
yading@10: 
yading@10: I<luminance_range_size> and I<luminance_minimum_value> depend on
yading@10: the input video format, the range is [0-255] for YUV full-range
yading@10: formats and [16-235] for YUV non full-range formats.
yading@10: 
yading@10: Default value is 0.10.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The following example sets the maximum pixel threshold to the minimum
yading@10: value, and detects only black intervals of 2 or more seconds:
yading@10: 	
yading@10: 	blackdetect=d=2:pix_th=0.00
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 blackframe
yading@10: 
yading@10: 
yading@10: Detect frames that are (almost) completely black. Can be useful to
yading@10: detect chapter transitions or commercials. Output lines consist of
yading@10: the frame number of the detected frame, the percentage of blackness,
yading@10: the position in the file if known or -1 and the timestamp in seconds.
yading@10: 
yading@10: In order to display the output lines, you need to set the loglevel at
yading@10: least to the AV_LOG_INFO value.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<amount>
yading@10: 
yading@10: Set the percentage of the pixels that have to be below the threshold, defaults
yading@10: to C<98>.
yading@10: 
yading@10: 
yading@10: =item B<threshold, thresh>
yading@10: 
yading@10: Set the threshold below which a pixel value is considered black, defaults to
yading@10: C<32>.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 blend
yading@10: 
yading@10: 
yading@10: Blend two video frames into each other.
yading@10: 
yading@10: It takes two input streams and outputs one stream, the first input is the
yading@10: "top" layer and second input is "bottom" layer.
yading@10: Output terminates when shortest input terminates.
yading@10: 
yading@10: A description of the accepted options follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<c0_mode>
yading@10: 
yading@10: 
yading@10: =item B<c1_mode>
yading@10: 
yading@10: 
yading@10: =item B<c2_mode>
yading@10: 
yading@10: 
yading@10: =item B<c3_mode>
yading@10: 
yading@10: 
yading@10: =item B<all_mode>
yading@10: 
yading@10: Set blend mode for specific pixel component or all pixel components in case
yading@10: of I<all_mode>. Default value is C<normal>.
yading@10: 
yading@10: Available values for component modes are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<addition>
yading@10: 
yading@10: 
yading@10: =item B<and>
yading@10: 
yading@10: 
yading@10: =item B<average>
yading@10: 
yading@10: 
yading@10: =item B<burn>
yading@10: 
yading@10: 
yading@10: =item B<darken>
yading@10: 
yading@10: 
yading@10: =item B<difference>
yading@10: 
yading@10: 
yading@10: =item B<divide>
yading@10: 
yading@10: 
yading@10: =item B<dodge>
yading@10: 
yading@10: 
yading@10: =item B<exclusion>
yading@10: 
yading@10: 
yading@10: =item B<hardlight>
yading@10: 
yading@10: 
yading@10: =item B<lighten>
yading@10: 
yading@10: 
yading@10: =item B<multiply>
yading@10: 
yading@10: 
yading@10: =item B<negation>
yading@10: 
yading@10: 
yading@10: =item B<normal>
yading@10: 
yading@10: 
yading@10: =item B<or>
yading@10: 
yading@10: 
yading@10: =item B<overlay>
yading@10: 
yading@10: 
yading@10: =item B<phoenix>
yading@10: 
yading@10: 
yading@10: =item B<pinlight>
yading@10: 
yading@10: 
yading@10: =item B<reflect>
yading@10: 
yading@10: 
yading@10: =item B<screen>
yading@10: 
yading@10: 
yading@10: =item B<softlight>
yading@10: 
yading@10: 
yading@10: =item B<subtract>
yading@10: 
yading@10: 
yading@10: =item B<vividlight>
yading@10: 
yading@10: 
yading@10: =item B<xor>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<c0_opacity>
yading@10: 
yading@10: 
yading@10: =item B<c1_opacity>
yading@10: 
yading@10: 
yading@10: =item B<c2_opacity>
yading@10: 
yading@10: 
yading@10: =item B<c3_opacity>
yading@10: 
yading@10: 
yading@10: =item B<all_opacity>
yading@10: 
yading@10: Set blend opacity for specific pixel component or all pixel components in case
yading@10: of I<all_opacity>. Only used in combination with pixel component blend modes.
yading@10: 
yading@10: 
yading@10: =item B<c0_expr>
yading@10: 
yading@10: 
yading@10: =item B<c1_expr>
yading@10: 
yading@10: 
yading@10: =item B<c2_expr>
yading@10: 
yading@10: 
yading@10: =item B<c3_expr>
yading@10: 
yading@10: 
yading@10: =item B<all_expr>
yading@10: 
yading@10: Set blend expression for specific pixel component or all pixel components in case
yading@10: of I<all_expr>. Note that related mode options will be ignored if those are set.
yading@10: 
yading@10: The expressions can use the following variables:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<N>
yading@10: 
yading@10: The sequential number of the filtered frame, starting from C<0>.
yading@10: 
yading@10: 
yading@10: =item B<X>
yading@10: 
yading@10: 
yading@10: =item B<Y>
yading@10: 
yading@10: the coordinates of the current sample
yading@10: 
yading@10: 
yading@10: =item B<W>
yading@10: 
yading@10: 
yading@10: =item B<H>
yading@10: 
yading@10: the width and height of currently filtered plane
yading@10: 
yading@10: 
yading@10: =item B<SW>
yading@10: 
yading@10: 
yading@10: =item B<SH>
yading@10: 
yading@10: Width and height scale depending on the currently filtered plane. It is the
yading@10: ratio between the corresponding luma plane number of pixels and the current
yading@10: plane ones. E.g. for YUV4:2:0 the values are C<1,1> for the luma plane, and
yading@10: C<0.5,0.5> for chroma planes.
yading@10: 
yading@10: 
yading@10: =item B<T>
yading@10: 
yading@10: Time of the current frame, expressed in seconds.
yading@10: 
yading@10: 
yading@10: =item B<TOP, A>
yading@10: 
yading@10: Value of pixel component at current location for first video frame (top layer).
yading@10: 
yading@10: 
yading@10: =item B<BOTTOM, B>
yading@10: 
yading@10: Value of pixel component at current location for second video frame (bottom layer).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply transition from bottom layer to top layer in first 10 seconds:
yading@10: 	
yading@10: 	blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply 1x1 checkerboard effect:
yading@10: 	
yading@10: 	blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 boxblur
yading@10: 
yading@10: 
yading@10: Apply boxblur algorithm to the input video.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<luma_radius, lr>
yading@10: 
yading@10: 
yading@10: =item B<luma_power, lp>
yading@10: 
yading@10: 
yading@10: =item B<chroma_radius, cr>
yading@10: 
yading@10: 
yading@10: =item B<chroma_power, cp>
yading@10: 
yading@10: 
yading@10: =item B<alpha_radius, ar>
yading@10: 
yading@10: 
yading@10: =item B<alpha_power, ap>
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: A description of the accepted options follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<luma_radius, lr>
yading@10: 
yading@10: 
yading@10: =item B<chroma_radius, cr>
yading@10: 
yading@10: 
yading@10: =item B<alpha_radius, ar>
yading@10: 
yading@10: Set an expression for the box radius in pixels used for blurring the
yading@10: corresponding input plane.
yading@10: 
yading@10: The radius value must be a non-negative number, and must not be
yading@10: greater than the value of the expression C<min(w,h)/2> for the
yading@10: luma and alpha planes, and of C<min(cw,ch)/2> for the chroma
yading@10: planes.
yading@10: 
yading@10: Default value for B<luma_radius> is "2". If not specified,
yading@10: B<chroma_radius> and B<alpha_radius> default to the
yading@10: corresponding value set for B<luma_radius>.
yading@10: 
yading@10: The expressions can contain the following constants:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<w, h>
yading@10: 
yading@10: the input width and height in pixels
yading@10: 
yading@10: 
yading@10: =item B<cw, ch>
yading@10: 
yading@10: the input chroma image width and height in pixels
yading@10: 
yading@10: 
yading@10: =item B<hsub, vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values. For example for the
yading@10: pixel format "yuv422p" I<hsub> is 2 and I<vsub> is 1.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<luma_power, lp>
yading@10: 
yading@10: 
yading@10: =item B<chroma_power, cp>
yading@10: 
yading@10: 
yading@10: =item B<alpha_power, ap>
yading@10: 
yading@10: Specify how many times the boxblur filter is applied to the
yading@10: corresponding plane.
yading@10: 
yading@10: Default value for B<luma_power> is 2. If not specified,
yading@10: B<chroma_power> and B<alpha_power> default to the
yading@10: corresponding value set for B<luma_power>.
yading@10: 
yading@10: A value of 0 will disable the effect.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply a boxblur filter with luma, chroma, and alpha radius
yading@10: set to 2:
yading@10: 	
yading@10: 	boxblur=luma_radius=2:luma_power=1
yading@10: 	boxblur=2:1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set luma radius to 2, alpha and chroma radius to 0:
yading@10: 	
yading@10: 	boxblur=2:1:cr=0:ar=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set luma and chroma radius to a fraction of the video dimension:
yading@10: 	
yading@10: 	boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 colorbalance
yading@10: 
yading@10: Modify intensity of primary colors (red, green and blue) of input frames.
yading@10: 
yading@10: The filter allows an input frame to be adjusted in the shadows, midtones or highlights
yading@10: regions for the red-cyan, green-magenta or blue-yellow balance.
yading@10: 
yading@10: A positive adjustment value shifts the balance towards the primary color, a negative
yading@10: value towards the complementary color.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<rs>
yading@10: 
yading@10: 
yading@10: =item B<gs>
yading@10: 
yading@10: 
yading@10: =item B<bs>
yading@10: 
yading@10: Adjust red, green and blue shadows (darkest pixels).
yading@10: 
yading@10: 
yading@10: =item B<rm>
yading@10: 
yading@10: 
yading@10: =item B<gm>
yading@10: 
yading@10: 
yading@10: =item B<bm>
yading@10: 
yading@10: Adjust red, green and blue midtones (medium pixels).
yading@10: 
yading@10: 
yading@10: =item B<rh>
yading@10: 
yading@10: 
yading@10: =item B<gh>
yading@10: 
yading@10: 
yading@10: =item B<bh>
yading@10: 
yading@10: Adjust red, green and blue highlights (brightest pixels).
yading@10: 
yading@10: Allowed ranges for options are C<[-1.0, 1.0]>. Defaults are C<0>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Add red color cast to shadows:
yading@10: 	
yading@10: 	colorbalance=rs=.3
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 colorchannelmixer
yading@10: 
yading@10: 
yading@10: Adjust video input frames by re-mixing color channels.
yading@10: 
yading@10: This filter modifies a color channel by adding the values associated to
yading@10: the other channels of the same pixels. For example if the value to
yading@10: modify is red, the output value will be:
yading@10: 	
yading@10: 	<red>=<red>*<rr> + <blue>*<rb> + <green>*<rg> + <alpha>*<ra>
yading@10: 
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<rr>
yading@10: 
yading@10: 
yading@10: =item B<rg>
yading@10: 
yading@10: 
yading@10: =item B<rb>
yading@10: 
yading@10: 
yading@10: =item B<ra>
yading@10: 
yading@10: Adjust contribution of input red, green, blue and alpha channels for output red channel.
yading@10: Default is C<1> for I<rr>, and C<0> for I<rg>, I<rb> and I<ra>.
yading@10: 
yading@10: 
yading@10: =item B<gr>
yading@10: 
yading@10: 
yading@10: =item B<gg>
yading@10: 
yading@10: 
yading@10: =item B<gb>
yading@10: 
yading@10: 
yading@10: =item B<ga>
yading@10: 
yading@10: Adjust contribution of input red, green, blue and alpha channels for output green channel.
yading@10: Default is C<1> for I<gg>, and C<0> for I<gr>, I<gb> and I<ga>.
yading@10: 
yading@10: 
yading@10: =item B<br>
yading@10: 
yading@10: 
yading@10: =item B<bg>
yading@10: 
yading@10: 
yading@10: =item B<bb>
yading@10: 
yading@10: 
yading@10: =item B<ba>
yading@10: 
yading@10: Adjust contribution of input red, green, blue and alpha channels for output blue channel.
yading@10: Default is C<1> for I<bb>, and C<0> for I<br>, I<bg> and I<ba>.
yading@10: 
yading@10: 
yading@10: =item B<ar>
yading@10: 
yading@10: 
yading@10: =item B<ag>
yading@10: 
yading@10: 
yading@10: =item B<ab>
yading@10: 
yading@10: 
yading@10: =item B<aa>
yading@10: 
yading@10: Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
yading@10: Default is C<1> for I<aa>, and C<0> for I<ar>, I<ag> and I<ab>.
yading@10: 
yading@10: Allowed ranges for options are C<[-2.0, 2.0]>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Convert source to grayscale:
yading@10: 	
yading@10: 	colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 colormatrix
yading@10: 
yading@10: 
yading@10: Convert color matrix.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<src>
yading@10: 
yading@10: 
yading@10: =item B<dst>
yading@10: 
yading@10: Specify the source and destination color matrix. Both values must be
yading@10: specified.
yading@10: 
yading@10: The accepted values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<bt709>
yading@10: 
yading@10: BT.709
yading@10: 
yading@10: 
yading@10: =item B<bt601>
yading@10: 
yading@10: BT.601
yading@10: 
yading@10: 
yading@10: =item B<smpte240m>
yading@10: 
yading@10: SMPTE-240M
yading@10: 
yading@10: 
yading@10: =item B<fcc>
yading@10: 
yading@10: FCC
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example to convert from BT.601 to SMPTE-240M, use the command:
yading@10: 	
yading@10: 	colormatrix=bt601:smpte240m
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 copy
yading@10: 
yading@10: 
yading@10: Copy the input source unchanged to the output. Mainly useful for
yading@10: testing purposes.
yading@10: 
yading@10: 
yading@10: =head2 crop
yading@10: 
yading@10: 
yading@10: Crop the input video to given dimensions.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<w, out_w>
yading@10: 
yading@10: Width of the output video. It defaults to C<iw>.
yading@10: This expression is evaluated only once during the filter
yading@10: configuration.
yading@10: 
yading@10: 
yading@10: =item B<h, out_h>
yading@10: 
yading@10: Height of the output video. It defaults to C<ih>.
yading@10: This expression is evaluated only once during the filter
yading@10: configuration.
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: Horizontal position, in the input video, of the left edge of the output video.
yading@10: It defaults to C<(in_w-out_w)/2>.
yading@10: This expression is evaluated per-frame.
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: Vertical position, in the input video, of the top edge of the output video.
yading@10: It defaults to C<(in_h-out_h)/2>.
yading@10: This expression is evaluated per-frame.
yading@10: 
yading@10: 
yading@10: =item B<keep_aspect>
yading@10: 
yading@10: If set to 1 will force the output display aspect ratio
yading@10: to be the same of the input, by changing the output sample aspect
yading@10: ratio. It defaults to 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The I<out_w>, I<out_h>, I<x>, I<y> parameters are
yading@10: expressions containing the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: the computed values for I<x> and I<y>. They are evaluated for
yading@10: each new frame.
yading@10: 
yading@10: 
yading@10: =item B<in_w, in_h>
yading@10: 
yading@10: the input width and height
yading@10: 
yading@10: 
yading@10: =item B<iw, ih>
yading@10: 
yading@10: same as I<in_w> and I<in_h>
yading@10: 
yading@10: 
yading@10: =item B<out_w, out_h>
yading@10: 
yading@10: the output (cropped) width and height
yading@10: 
yading@10: 
yading@10: =item B<ow, oh>
yading@10: 
yading@10: same as I<out_w> and I<out_h>
yading@10: 
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: same as I<iw> / I<ih>
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: input sample aspect ratio
yading@10: 
yading@10: 
yading@10: =item B<dar>
yading@10: 
yading@10: input display aspect ratio, it is the same as (I<iw> / I<ih>) * I<sar>
yading@10: 
yading@10: 
yading@10: =item B<hsub, vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values. For example for the
yading@10: pixel format "yuv422p" I<hsub> is 2 and I<vsub> is 1.
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: the number of input frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<pos>
yading@10: 
yading@10: the position in the file of the input frame, NAN if unknown
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: timestamp expressed in seconds, NAN if the input timestamp is unknown
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The expression for I<out_w> may depend on the value of I<out_h>,
yading@10: and the expression for I<out_h> may depend on I<out_w>, but they
yading@10: cannot depend on I<x> and I<y>, as I<x> and I<y> are
yading@10: evaluated after I<out_w> and I<out_h>.
yading@10: 
yading@10: The I<x> and I<y> parameters specify the expressions for the
yading@10: position of the top-left corner of the output (non-cropped) area. They
yading@10: are evaluated for each frame. If the evaluated value is not valid, it
yading@10: is approximated to the nearest valid value.
yading@10: 
yading@10: The expression for I<x> may depend on I<y>, and the expression
yading@10: for I<y> may depend on I<x>.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop area with size 100x100 at position (12,34).
yading@10: 	
yading@10: 	crop=100:100:12:34
yading@10: 
yading@10: 
yading@10: Using named options, the example above becomes:
yading@10: 	
yading@10: 	crop=w=100:h=100:x=12:y=34
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop the central input area with size 100x100:
yading@10: 	
yading@10: 	crop=100:100
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop the central input area with size 2/3 of the input video:
yading@10: 	
yading@10: 	crop=2/3*in_w:2/3*in_h
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop the input video central square:
yading@10: 	
yading@10: 	crop=out_w=in_h
yading@10: 	crop=in_h
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Delimit the rectangle with the top-left corner placed at position
yading@10: 100:100 and the right-bottom corner corresponding to the right-bottom
yading@10: corner of the input image:
yading@10: 	
yading@10: 	crop=in_w-100:in_h-100:100:100
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop 10 pixels from the left and right borders, and 20 pixels from
yading@10: the top and bottom borders
yading@10: 	
yading@10: 	crop=in_w-2*10:in_h-2*20
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Keep only the bottom right quarter of the input image:
yading@10: 	
yading@10: 	crop=in_w/2:in_h/2:in_w/2:in_h/2
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Crop height for getting Greek harmony:
yading@10: 	
yading@10: 	crop=in_w:1/PHI*in_w
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Appply trembling effect:
yading@10: 	
yading@10: 	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)
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply erratic camera effect depending on timestamp:
yading@10: 	
yading@10: 	crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set x depending on the value of y:
yading@10: 	
yading@10: 	crop=in_w/2:in_h/2:y:10+10*sin(n/10)
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 cropdetect
yading@10: 
yading@10: 
yading@10: Auto-detect crop size.
yading@10: 
yading@10: Calculate necessary cropping parameters and prints the recommended
yading@10: parameters through the logging system. The detected dimensions
yading@10: correspond to the non-black area of the input video.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<limit>
yading@10: 
yading@10: Set higher black value threshold, which can be optionally specified
yading@10: from nothing (0) to everything (255). An intensity value greater
yading@10: to the set value is considered non-black. Default value is 24.
yading@10: 
yading@10: 
yading@10: =item B<round>
yading@10: 
yading@10: Set the value for which the width/height should be divisible by. The
yading@10: offset is automatically adjusted to center the video. Use 2 to get
yading@10: only even dimensions (needed for 4:2:2 video). 16 is best when
yading@10: encoding to most video codecs. Default value is 16.
yading@10: 
yading@10: 
yading@10: =item B<reset_count, reset>
yading@10: 
yading@10: Set the counter that determines after how many frames cropdetect will
yading@10: reset the previously detected largest video area and start over to
yading@10: detect the current optimal crop area. Default value is 0.
yading@10: 
yading@10: This can be useful when channel logos distort the video area. 0
yading@10: indicates never reset and return the largest area encountered during
yading@10: playback.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 curves
yading@10: 
yading@10: 
yading@10: Apply color adjustments using curves.
yading@10: 
yading@10: This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
yading@10: component (red, green and blue) has its values defined by I<N> key points
yading@10: tied from each other using a smooth curve. The x-axis represents the pixel
yading@10: values from the input frame, and the y-axis the new pixel values to be set for
yading@10: the output frame.
yading@10: 
yading@10: By default, a component curve is defined by the two points I<(0;0)> and
yading@10: I<(1;1)>. This creates a straight line where each original pixel value is
yading@10: "adjusted" to its own value, which means no change to the image.
yading@10: 
yading@10: The filter allows you to redefine these two points and add some more. A new
yading@10: curve (using a natural cubic spline interpolation) will be define to pass
yading@10: smoothly through all these new coordinates. The new defined points needs to be
yading@10: strictly increasing over the x-axis, and their I<x> and I<y> values must
yading@10: be in the I<[0;1]> interval.  If the computed curves happened to go outside
yading@10: the vector spaces, the values will be clipped accordingly.
yading@10: 
yading@10: If there is no key point defined in C<x=0>, the filter will automatically
yading@10: insert a I<(0;0)> point. In the same way, if there is no key point defined
yading@10: in C<x=1>, the filter will automatically insert a I<(1;1)> point.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<preset>
yading@10: 
yading@10: Select one of the available color presets. This option can be used in addition
yading@10: to the B<r>, B<g>, B<b> parameters; in this case, the later
yading@10: options takes priority on the preset values.
yading@10: Available presets are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: 
yading@10: =item B<color_negative>
yading@10: 
yading@10: 
yading@10: =item B<cross_process>
yading@10: 
yading@10: 
yading@10: =item B<darker>
yading@10: 
yading@10: 
yading@10: =item B<increase_contrast>
yading@10: 
yading@10: 
yading@10: =item B<lighter>
yading@10: 
yading@10: 
yading@10: =item B<linear_contrast>
yading@10: 
yading@10: 
yading@10: =item B<medium_contrast>
yading@10: 
yading@10: 
yading@10: =item B<negative>
yading@10: 
yading@10: 
yading@10: =item B<strong_contrast>
yading@10: 
yading@10: 
yading@10: =item B<vintage>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default is C<none>.
yading@10: 
yading@10: =item B<master, m>
yading@10: 
yading@10: Set the master key points. These points will define a second pass mapping. It
yading@10: is sometimes called a "luminance" or "value" mapping. It can be used with
yading@10: B<r>, B<g>, B<b> or B<all> since it acts like a
yading@10: post-processing LUT.
yading@10: 
yading@10: =item B<red, r>
yading@10: 
yading@10: Set the key points for the red component.
yading@10: 
yading@10: =item B<green, g>
yading@10: 
yading@10: Set the key points for the green component.
yading@10: 
yading@10: =item B<blue, b>
yading@10: 
yading@10: Set the key points for the blue component.
yading@10: 
yading@10: =item B<all>
yading@10: 
yading@10: Set the key points for all components (not including master).
yading@10: Can be used in addition to the other key points component
yading@10: options. In this case, the unset component(s) will fallback on this
yading@10: B<all> setting.
yading@10: 
yading@10: =item B<psfile>
yading@10: 
yading@10: Specify a Photoshop curves file (C<.asv>) to import the settings from.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: To avoid some filtergraph syntax conflicts, each key points list need to be
yading@10: defined using the following syntax: C<x0/y0 x1/y1 x2/y2 ...>.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase slightly the middle level of blue:
yading@10: 	
yading@10: 	curves=blue='0.5/0.58'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Vintage effect:
yading@10: 	
yading@10: 	curves=r='0/0.11 .42/.51 1/0.95':g='0.50/0.48':b='0/0.22 .49/.44 1/0.8'
yading@10: 
yading@10: Here we obtain the following coordinates for each components:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item I<red>
yading@10: 
yading@10: C<(0;0.11) (0.42;0.51) (1;0.95)>
yading@10: 
yading@10: =item I<green>
yading@10: 
yading@10: C<(0;0) (0.50;0.48) (1;1)>
yading@10: 
yading@10: =item I<blue>
yading@10: 
yading@10: C<(0;0.22) (0.49;0.44) (1;0.80)>
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: The previous example can also be achieved with the associated built-in preset:
yading@10: 	
yading@10: 	curves=preset=vintage
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Or simply:
yading@10: 	
yading@10: 	curves=vintage
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Use a Photoshop preset and redefine the points of the green component:
yading@10: 	
yading@10: 	curves=psfile='MyCurvesPresets/purple.asv':green='0.45/0.53'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 decimate
yading@10: 
yading@10: 
yading@10: Drop duplicated frames at regular intervals.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<cycle>
yading@10: 
yading@10: Set the number of frames from which one will be dropped. Setting this to
yading@10: I<N> means one frame in every batch of I<N> frames will be dropped.
yading@10: Default is C<5>.
yading@10: 
yading@10: 
yading@10: =item B<dupthresh>
yading@10: 
yading@10: Set the threshold for duplicate detection. If the difference metric for a frame
yading@10: is less than or equal to this value, then it is declared as duplicate. Default
yading@10: is C<1.1>
yading@10: 
yading@10: 
yading@10: =item B<scthresh>
yading@10: 
yading@10: Set scene change threshold. Default is C<15>.
yading@10: 
yading@10: 
yading@10: =item B<blockx>
yading@10: 
yading@10: 
yading@10: =item B<blocky>
yading@10: 
yading@10: Set the size of the x and y-axis blocks used during metric calculations.
yading@10: Larger blocks give better noise suppression, but also give worse detection of
yading@10: small movements. Must be a power of two. Default is C<32>.
yading@10: 
yading@10: 
yading@10: =item B<ppsrc>
yading@10: 
yading@10: Mark main input as a pre-processed input and activate clean source input
yading@10: stream. This allows the input to be pre-processed with various filters to help
yading@10: the metrics calculation while keeping the frame selection lossless. When set to
yading@10: C<1>, the first stream is for the pre-processed input, and the second
yading@10: stream is the clean source from where the kept frames are chosen. Default is
yading@10: C<0>.
yading@10: 
yading@10: 
yading@10: =item B<chroma>
yading@10: 
yading@10: Set whether or not chroma is considered in the metric calculations. Default is
yading@10: C<1>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 delogo
yading@10: 
yading@10: 
yading@10: Suppress a TV station logo by a simple interpolation of the surrounding
yading@10: pixels. Just set a rectangle covering the logo and watch it disappear
yading@10: (and sometimes something even uglier appear - your mileage may vary).
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: Specify the top left corner coordinates of the logo. They must be
yading@10: specified.
yading@10: 
yading@10: 
yading@10: =item B<w, h>
yading@10: 
yading@10: Specify the width and height of the logo to clear. They must be
yading@10: specified.
yading@10: 
yading@10: 
yading@10: =item B<band, t>
yading@10: 
yading@10: Specify the thickness of the fuzzy edge of the rectangle (added to
yading@10: I<w> and I<h>). The default value is 4.
yading@10: 
yading@10: 
yading@10: =item B<show>
yading@10: 
yading@10: When set to 1, a green rectangle is drawn on the screen to simplify
yading@10: finding the right I<x>, I<y>, I<w>, I<h> parameters, and
yading@10: I<band> is set to 4. The default value is 0.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set a rectangle covering the area with top left corner coordinates 0,0
yading@10: and size 100x77, setting a band of size 10:
yading@10: 	
yading@10: 	delogo=x=0:y=0:w=100:h=77:band=10
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 deshake
yading@10: 
yading@10: 
yading@10: Attempt to fix small changes in horizontal and/or vertical shift. This
yading@10: filter helps remove camera shake from hand-holding a camera, bumping a
yading@10: tripod, moving on a vehicle, etc.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: 
yading@10: =item B<w>
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Specify a rectangular area where to limit the search for motion
yading@10: vectors.
yading@10: If desired the search for motion vectors can be limited to a
yading@10: rectangular area of the frame defined by its top left corner, width
yading@10: and height. These parameters have the same meaning as the drawbox
yading@10: filter which can be used to visualise the position of the bounding
yading@10: box.
yading@10: 
yading@10: This is useful when simultaneous movement of subjects within the frame
yading@10: might be confused for camera motion by the motion vector search.
yading@10: 
yading@10: If any or all of I<x>, I<y>, I<w> and I<h> are set to -1
yading@10: then the full frame is used. This allows later options to be set
yading@10: without specifying the bounding box for the motion vector search.
yading@10: 
yading@10: Default - search the whole frame.
yading@10: 
yading@10: 
yading@10: =item B<rx>
yading@10: 
yading@10: 
yading@10: =item B<ry>
yading@10: 
yading@10: Specify the maximum extent of movement in x and y directions in the
yading@10: range 0-64 pixels. Default 16.
yading@10: 
yading@10: 
yading@10: =item B<edge>
yading@10: 
yading@10: Specify how to generate pixels to fill blanks at the edge of the
yading@10: frame. Available values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<blank, 0>
yading@10: 
yading@10: Fill zeroes at blank locations
yading@10: 
yading@10: =item B<original, 1>
yading@10: 
yading@10: Original image at blank locations
yading@10: 
yading@10: =item B<clamp, 2>
yading@10: 
yading@10: Extruded edge value at blank locations
yading@10: 
yading@10: =item B<mirror, 3>
yading@10: 
yading@10: Mirrored edge at blank locations
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default value is B<mirror>.
yading@10: 
yading@10: 
yading@10: =item B<blocksize>
yading@10: 
yading@10: Specify the blocksize to use for motion search. Range 4-128 pixels,
yading@10: default 8.
yading@10: 
yading@10: 
yading@10: =item B<contrast>
yading@10: 
yading@10: Specify the contrast threshold for blocks. Only blocks with more than
yading@10: the specified contrast (difference between darkest and lightest
yading@10: pixels) will be considered. Range 1-255, default 125.
yading@10: 
yading@10: 
yading@10: =item B<search>
yading@10: 
yading@10: Specify the search strategy. Available values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<exhaustive, 0>
yading@10: 
yading@10: Set exhaustive search
yading@10: 
yading@10: =item B<less, 1>
yading@10: 
yading@10: Set less exhaustive search.
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default value is B<exhaustive>.
yading@10: 
yading@10: 
yading@10: =item B<filename>
yading@10: 
yading@10: If set then a detailed log of the motion search is written to the
yading@10: specified file.
yading@10: 
yading@10: 
yading@10: =item B<opencl>
yading@10: 
yading@10: If set to 1, specify using OpenCL capabilities, only available if
yading@10: FFmpeg was configured with C<--enable-opencl>. Default value is 0.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 drawbox
yading@10: 
yading@10: 
yading@10: Draw a colored box on the input image.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: Specify the top left corner coordinates of the box. Default to 0.
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: 
yading@10: =item B<height, h>
yading@10: 
yading@10: Specify the width and height of the box, if 0 they are interpreted as
yading@10: the input width and height. Default to 0.
yading@10: 
yading@10: 
yading@10: =item B<color, c>
yading@10: 
yading@10: Specify the color of the box to write, it can be the name of a color
yading@10: (case insensitive match) or a 0xRRGGBB[AA] sequence. If the special
yading@10: value C<invert> is used, the box edge color is the same as the
yading@10: video with inverted luma.
yading@10: 
yading@10: 
yading@10: =item B<thickness, t>
yading@10: 
yading@10: Set the thickness of the box edge. Default value is C<4>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw a black box around the edge of the input image:
yading@10: 	
yading@10: 	drawbox
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw a box with color red and an opacity of 50%:
yading@10: 	
yading@10: 	drawbox=10:20:200:60:red@0.5
yading@10: 
yading@10: 
yading@10: The previous example can be specified as:
yading@10: 	
yading@10: 	drawbox=x=10:y=20:w=200:h=60:color=red@0.5
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fill the box with pink color:
yading@10: 	
yading@10: 	drawbox=x=10:y=10:w=100:h=100:color=pink@0.5:t=max
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 drawtext
yading@10: 
yading@10: 
yading@10: Draw text string or text from specified file on top of video using the
yading@10: libfreetype library.
yading@10: 
yading@10: To enable compilation of this filter you need to configure FFmpeg with
yading@10: C<--enable-libfreetype>.
yading@10: 
yading@10: 
yading@10: =head3 Syntax
yading@10: 
yading@10: 
yading@10: The description of the accepted parameters follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<box>
yading@10: 
yading@10: Used to draw a box around text using background color.
yading@10: Value should be either 1 (enable) or 0 (disable).
yading@10: The default value of I<box> is 0.
yading@10: 
yading@10: 
yading@10: =item B<boxcolor>
yading@10: 
yading@10: The color to be used for drawing box around text.
yading@10: Either a string (e.g. "yellow") or in 0xRRGGBB[AA] format
yading@10: (e.g. "0xff00ff"), possibly followed by an alpha specifier.
yading@10: The default value of I<boxcolor> is "white".
yading@10: 
yading@10: 
yading@10: =item B<draw>
yading@10: 
yading@10: Set an expression which specifies if the text should be drawn. If the
yading@10: expression evaluates to 0, the text is not drawn. This is useful for
yading@10: specifying that the text should be drawn only when specific conditions
yading@10: are met.
yading@10: 
yading@10: Default value is "1".
yading@10: 
yading@10: See below for the list of accepted constants and functions.
yading@10: 
yading@10: 
yading@10: =item B<expansion>
yading@10: 
yading@10: Select how the I<text> is expanded. Can be either C<none>,
yading@10: C<strftime> (deprecated) or
yading@10: C<normal> (default). See the drawtext_expansion, Text expansion section
yading@10: below for details.
yading@10: 
yading@10: 
yading@10: =item B<fix_bounds>
yading@10: 
yading@10: If true, check and fix text coords to avoid clipping.
yading@10: 
yading@10: 
yading@10: =item B<fontcolor>
yading@10: 
yading@10: The color to be used for drawing fonts.
yading@10: Either a string (e.g. "red") or in 0xRRGGBB[AA] format
yading@10: (e.g. "0xff000033"), possibly followed by an alpha specifier.
yading@10: The default value of I<fontcolor> is "black".
yading@10: 
yading@10: 
yading@10: =item B<fontfile>
yading@10: 
yading@10: The font file to be used for drawing text. Path must be included.
yading@10: This parameter is mandatory.
yading@10: 
yading@10: 
yading@10: =item B<fontsize>
yading@10: 
yading@10: The font size to be used for drawing text.
yading@10: The default value of I<fontsize> is 16.
yading@10: 
yading@10: 
yading@10: =item B<ft_load_flags>
yading@10: 
yading@10: Flags to be used for loading the fonts.
yading@10: 
yading@10: The flags map the corresponding flags supported by libfreetype, and are
yading@10: a combination of the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item I<default>
yading@10: 
yading@10: 
yading@10: =item I<no_scale>
yading@10: 
yading@10: 
yading@10: =item I<no_hinting>
yading@10: 
yading@10: 
yading@10: =item I<render>
yading@10: 
yading@10: 
yading@10: =item I<no_bitmap>
yading@10: 
yading@10: 
yading@10: =item I<vertical_layout>
yading@10: 
yading@10: 
yading@10: =item I<force_autohint>
yading@10: 
yading@10: 
yading@10: =item I<crop_bitmap>
yading@10: 
yading@10: 
yading@10: =item I<pedantic>
yading@10: 
yading@10: 
yading@10: =item I<ignore_global_advance_width>
yading@10: 
yading@10: 
yading@10: =item I<no_recurse>
yading@10: 
yading@10: 
yading@10: =item I<ignore_transform>
yading@10: 
yading@10: 
yading@10: =item I<monochrome>
yading@10: 
yading@10: 
yading@10: =item I<linear_design>
yading@10: 
yading@10: 
yading@10: =item I<no_autohint>
yading@10: 
yading@10: 
yading@10: =item I<end table>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is "render".
yading@10: 
yading@10: For more information consult the documentation for the FT_LOAD_*
yading@10: libfreetype flags.
yading@10: 
yading@10: 
yading@10: =item B<shadowcolor>
yading@10: 
yading@10: The color to be used for drawing a shadow behind the drawn text.  It
yading@10: can be a color name (e.g. "yellow") or a string in the 0xRRGGBB[AA]
yading@10: form (e.g. "0xff00ff"), possibly followed by an alpha specifier.
yading@10: The default value of I<shadowcolor> is "black".
yading@10: 
yading@10: 
yading@10: =item B<shadowx, shadowy>
yading@10: 
yading@10: The x and y offsets for the text shadow position with respect to the
yading@10: position of the text. They can be either positive or negative
yading@10: values. Default value for both is "0".
yading@10: 
yading@10: 
yading@10: =item B<tabsize>
yading@10: 
yading@10: The size in number of spaces to use for rendering the tab.
yading@10: Default value is 4.
yading@10: 
yading@10: 
yading@10: =item B<timecode>
yading@10: 
yading@10: Set the initial timecode representation in "hh:mm:ss[:;.]ff"
yading@10: format. It can be used with or without text parameter. I<timecode_rate>
yading@10: option must be specified.
yading@10: 
yading@10: 
yading@10: =item B<timecode_rate, rate, r>
yading@10: 
yading@10: Set the timecode frame rate (timecode only).
yading@10: 
yading@10: 
yading@10: =item B<text>
yading@10: 
yading@10: The text string to be drawn. The text must be a sequence of UTF-8
yading@10: encoded characters.
yading@10: This parameter is mandatory if no file is specified with the parameter
yading@10: I<textfile>.
yading@10: 
yading@10: 
yading@10: =item B<textfile>
yading@10: 
yading@10: A text file containing text to be drawn. The text must be a sequence
yading@10: of UTF-8 encoded characters.
yading@10: 
yading@10: This parameter is mandatory if no text string is specified with the
yading@10: parameter I<text>.
yading@10: 
yading@10: If both I<text> and I<textfile> are specified, an error is thrown.
yading@10: 
yading@10: 
yading@10: =item B<reload>
yading@10: 
yading@10: If set to 1, the I<textfile> will be reloaded before each frame.
yading@10: Be sure to update it atomically, or it may be read partially, or even fail.
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: The expressions which specify the offsets where text will be drawn
yading@10: within the video frame. They are relative to the top/left border of the
yading@10: output image.
yading@10: 
yading@10: The default value of I<x> and I<y> is "0".
yading@10: 
yading@10: See below for the list of accepted constants and functions.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The parameters for I<x> and I<y> are expressions containing the
yading@10: following constants and functions:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<dar>
yading@10: 
yading@10: input display aspect ratio, it is the same as (I<w> / I<h>) * I<sar>
yading@10: 
yading@10: 
yading@10: =item B<hsub, vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values. For example for the
yading@10: pixel format "yuv422p" I<hsub> is 2 and I<vsub> is 1.
yading@10: 
yading@10: 
yading@10: =item B<line_h, lh>
yading@10: 
yading@10: the height of each text line
yading@10: 
yading@10: 
yading@10: =item B<main_h, h, H>
yading@10: 
yading@10: the input height
yading@10: 
yading@10: 
yading@10: =item B<main_w, w, W>
yading@10: 
yading@10: the input width
yading@10: 
yading@10: 
yading@10: =item B<max_glyph_a, ascent>
yading@10: 
yading@10: the maximum distance from the baseline to the highest/upper grid
yading@10: coordinate used to place a glyph outline point, for all the rendered
yading@10: glyphs.
yading@10: It is a positive value, due to the grid's orientation with the Y axis
yading@10: upwards.
yading@10: 
yading@10: 
yading@10: =item B<max_glyph_d, descent>
yading@10: 
yading@10: the maximum distance from the baseline to the lowest grid coordinate
yading@10: used to place a glyph outline point, for all the rendered glyphs.
yading@10: This is a negative value, due to the grid's orientation, with the Y axis
yading@10: upwards.
yading@10: 
yading@10: 
yading@10: =item B<max_glyph_h>
yading@10: 
yading@10: maximum glyph height, that is the maximum height for all the glyphs
yading@10: contained in the rendered text, it is equivalent to I<ascent> -
yading@10: I<descent>.
yading@10: 
yading@10: 
yading@10: =item B<max_glyph_w>
yading@10: 
yading@10: maximum glyph width, that is the maximum width for all the glyphs
yading@10: contained in the rendered text
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: the number of input frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<rand(min, max)>
yading@10: 
yading@10: return a random number included between I<min> and I<max>
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: input sample aspect ratio
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: timestamp expressed in seconds, NAN if the input timestamp is unknown
yading@10: 
yading@10: 
yading@10: =item B<text_h, th>
yading@10: 
yading@10: the height of the rendered text
yading@10: 
yading@10: 
yading@10: =item B<text_w, tw>
yading@10: 
yading@10: the width of the rendered text
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: the x and y offset coordinates where the text is drawn.
yading@10: 
yading@10: These parameters allow the I<x> and I<y> expressions to refer
yading@10: each other, so you can for example specify C<y=x/dar>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If libavfilter was built with C<--enable-fontconfig>, then
yading@10: B<fontfile> can be a fontconfig pattern or omitted.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Text expansion
yading@10: 
yading@10: 
yading@10: If B<expansion> is set to C<strftime>,
yading@10: the filter recognizes strftime() sequences in the provided text and
yading@10: expands them accordingly. Check the documentation of strftime(). This
yading@10: feature is deprecated.
yading@10: 
yading@10: If B<expansion> is set to C<none>, the text is printed verbatim.
yading@10: 
yading@10: If B<expansion> is set to C<normal> (which is the default),
yading@10: the following expansion mechanism is used.
yading@10: 
yading@10: The backslash character '\', followed by any character, always expands to
yading@10: the second character.
yading@10: 
yading@10: Sequence of the form C<%{...}> are expanded. The text between the
yading@10: braces is a function name, possibly followed by arguments separated by ':'.
yading@10: If the arguments contain special characters or delimiters (':' or '}'),
yading@10: they should be escaped.
yading@10: 
yading@10: Note that they probably must also be escaped as the value for the
yading@10: B<text> option in the filter argument string and as the filter
yading@10: argument in the filtergraph description, and possibly also for the shell,
yading@10: that makes up to four levels of escaping; using a text file avoids these
yading@10: problems.
yading@10: 
yading@10: The following functions are available:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<expr, e>
yading@10: 
yading@10: The expression evaluation result.
yading@10: 
yading@10: It must take one argument specifying the expression to be evaluated,
yading@10: which accepts the same constants and functions as the I<x> and
yading@10: I<y> values. Note that not all constants should be used, for
yading@10: example the text size is not known when evaluating the expression, so
yading@10: the constants I<text_w> and I<text_h> will have an undefined
yading@10: value.
yading@10: 
yading@10: 
yading@10: =item B<gmtime>
yading@10: 
yading@10: The time at which the filter is running, expressed in UTC.
yading@10: It can accept an argument: a strftime() format string.
yading@10: 
yading@10: 
yading@10: =item B<localtime>
yading@10: 
yading@10: The time at which the filter is running, expressed in the local time zone.
yading@10: It can accept an argument: a strftime() format string.
yading@10: 
yading@10: 
yading@10: =item B<n, frame_num>
yading@10: 
yading@10: The frame number, starting from 0.
yading@10: 
yading@10: 
yading@10: =item B<pts>
yading@10: 
yading@10: The timestamp of the current frame, in seconds, with microsecond accuracy.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw "Test Text" with font FreeSerif, using the default values for the
yading@10: optional parameters.
yading@10: 
yading@10: 	
yading@10: 	drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw 'Test Text' with font FreeSerif of size 24 at position x=100
yading@10: and y=50 (counting from the top-left corner of the screen), text is
yading@10: yellow with a red box around it. Both the text and the box have an
yading@10: opacity of 20%.
yading@10: 
yading@10: 	
yading@10: 	drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
yading@10: 	          x=100: y=50: fontsize=24: fontcolor=yellow@0.2: box=1: boxcolor=red@0.2"
yading@10: 
yading@10: 
yading@10: Note that the double quotes are not necessary if spaces are not used
yading@10: within the parameter list.
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Show the text at the center of the video frame:
yading@10: 	
yading@10: 	drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h-line_h)/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Show a text line sliding from right to left in the last row of the video
yading@10: frame. The file F<LONG_LINE> is assumed to contain a single line
yading@10: with no newlines.
yading@10: 	
yading@10: 	drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Show the content of file F<CREDITS> off the bottom of the frame and scroll up.
yading@10: 	
yading@10: 	drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw a single green letter "g", at the center of the input video.
yading@10: The glyph baseline is placed at half screen height.
yading@10: 	
yading@10: 	drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Show text for 1 second every 3 seconds:
yading@10: 	
yading@10: 	drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:draw=lt(mod(t\,3)\,1):text='blink'"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Use fontconfig to set the font. Note that the colons need to be escaped.
yading@10: 	
yading@10: 	drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Print the date of a real-time encoding (see strftime(3)):
yading@10: 	
yading@10: 	drawtext='fontfile=FreeSans.ttf:text=%{localtime:%a %b %d %Y}'
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For more information about libfreetype, check:
yading@10: E<lt>B<http://www.freetype.org/>E<gt>.
yading@10: 
yading@10: For more information about fontconfig, check:
yading@10: E<lt>B<http://freedesktop.org/software/fontconfig/fontconfig-user.html>E<gt>.
yading@10: 
yading@10: 
yading@10: =head2 edgedetect
yading@10: 
yading@10: 
yading@10: Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<low, high>
yading@10: 
yading@10: Set low and high threshold values used by the Canny thresholding
yading@10: algorithm.
yading@10: 
yading@10: The high threshold selects the "strong" edge pixels, which are then
yading@10: connected through 8-connectivity with the "weak" edge pixels selected
yading@10: by the low threshold.
yading@10: 
yading@10: I<low> and I<high> threshold values must be choosen in the range
yading@10: [0,1], and I<low> should be lesser or equal to I<high>.
yading@10: 
yading@10: Default value for I<low> is C<20/255>, and default value for I<high>
yading@10: is C<50/255>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Example:
yading@10: 	
yading@10: 	edgedetect=low=0.1:high=0.4
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 fade
yading@10: 
yading@10: 
yading@10: Apply fade-in/out effect to input video.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<type, t>
yading@10: 
yading@10: The effect type -- can be either "in" for fade-in, or "out" for a fade-out
yading@10: effect.
yading@10: Default is C<in>.
yading@10: 
yading@10: 
yading@10: =item B<start_frame, s>
yading@10: 
yading@10: Specify the number of the start frame for starting to apply the fade
yading@10: effect. Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<nb_frames, n>
yading@10: 
yading@10: The number of frames for which the fade effect has to last. At the end of the
yading@10: fade-in effect the output video will have the same intensity as the input video,
yading@10: at the end of the fade-out transition the output video will be completely black.
yading@10: Default is 25.
yading@10: 
yading@10: 
yading@10: =item B<alpha>
yading@10: 
yading@10: If set to 1, fade only alpha channel, if one exists on the input.
yading@10: Default value is 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade in first 30 frames of video:
yading@10: 	
yading@10: 	fade=in:0:30
yading@10: 
yading@10: 
yading@10: The command above is equivalent to:
yading@10: 	
yading@10: 	fade=t=in:s=0:n=30
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade out last 45 frames of a 200-frame video:
yading@10: 	
yading@10: 	fade=out:155:45
yading@10: 	fade=type=out:start_frame=155:nb_frames=45
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade in first 25 frames and fade out last 25 frames of a 1000-frame video:
yading@10: 	
yading@10: 	fade=in:0:25, fade=out:975:25
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Make first 5 frames black, then fade in from frame 5-24:
yading@10: 	
yading@10: 	fade=in:5:20
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Fade in alpha over first 25 frames of video:
yading@10: 	
yading@10: 	fade=in:0:25:alpha=1
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 field
yading@10: 
yading@10: 
yading@10: Extract a single field from an interlaced image using stride
yading@10: arithmetic to avoid wasting CPU time. The output frames are marked as
yading@10: non-interlaced.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<type>
yading@10: 
yading@10: Specify whether to extract the top (if the value is C<0> or
yading@10: C<top>) or the bottom field (if the value is C<1> or
yading@10: C<bottom>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 fieldmatch
yading@10: 
yading@10: 
yading@10: Field matching filter for inverse telecine. It is meant to reconstruct the
yading@10: progressive frames from a telecined stream. The filter does not drop duplicated
yading@10: frames, so to achieve a complete inverse telecine C<fieldmatch> needs to be
yading@10: followed by a decimation filter such as decimate in the filtergraph.
yading@10: 
yading@10: The separation of the field matching and the decimation is notably motivated by
yading@10: the possibility of inserting a de-interlacing filter fallback between the two.
yading@10: If the source has mixed telecined and real interlaced content,
yading@10: C<fieldmatch> will not be able to match fields for the interlaced parts.
yading@10: But these remaining combed frames will be marked as interlaced, and thus can be
yading@10: de-interlaced by a later filter such as yadif before decimation.
yading@10: 
yading@10: In addition to the various configuration options, C<fieldmatch> can take an
yading@10: optional second stream, activated through the B<ppsrc> option. If
yading@10: enabled, the frames reconstruction will be based on the fields and frames from
yading@10: this second stream. This allows the first input to be pre-processed in order to
yading@10: help the various algorithms of the filter, while keeping the output lossless
yading@10: (assuming the fields are matched properly). Typically, a field-aware denoiser,
yading@10: or brightness/contrast adjustments can help.
yading@10: 
yading@10: Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
yading@10: and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
yading@10: which C<fieldmatch> is based on. While the semantic and usage are very
yading@10: close, some behaviour and options names can differ.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<order>
yading@10: 
yading@10: Specify the assumed field order of the input stream. Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<auto>
yading@10: 
yading@10: Auto detect parity (use FFmpeg's internal parity value).
yading@10: 
yading@10: =item B<bff>
yading@10: 
yading@10: Assume bottom field first.
yading@10: 
yading@10: =item B<tff>
yading@10: 
yading@10: Assume top field first.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Note that it is sometimes recommended not to trust the parity announced by the
yading@10: stream.
yading@10: 
yading@10: Default value is I<auto>.
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Set the matching mode or strategy to use. B<pc> mode is the safest in the
yading@10: sense that it wont risk creating jerkiness due to duplicate frames when
yading@10: possible, but if there are bad edits or blended fields it will end up
yading@10: outputting combed frames when a good match might actually exist. On the other
yading@10: hand, B<pcn_ub> mode is the most risky in terms of creating jerkiness,
yading@10: but will almost always find a good frame if there is one. The other values are
yading@10: all somewhere in between B<pc> and B<pcn_ub> in terms of risking
yading@10: jerkiness and creating duplicate frames versus finding good matches in sections
yading@10: with bad edits, orphaned fields, blended fields, etc.
yading@10: 
yading@10: More details about p/c/n/u/b are available in p/c/n/u/b meaning section.
yading@10: 
yading@10: Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<pc>
yading@10: 
yading@10: 2-way matching (p/c)
yading@10: 
yading@10: =item B<pc_n>
yading@10: 
yading@10: 2-way matching, and trying 3rd match if still combed (p/c + n)
yading@10: 
yading@10: =item B<pc_u>
yading@10: 
yading@10: 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
yading@10: 
yading@10: =item B<pc_n_ub>
yading@10: 
yading@10: 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
yading@10: still combed (p/c + n + u/b)
yading@10: 
yading@10: =item B<pcn>
yading@10: 
yading@10: 3-way matching (p/c/n)
yading@10: 
yading@10: =item B<pcn_ub>
yading@10: 
yading@10: 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
yading@10: detected as combed (p/c/n + u/b)
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The parenthesis at the end indicate the matches that would be used for that
yading@10: mode assuming B<order>=I<tff> (and B<field> on I<auto> or
yading@10: I<top>).
yading@10: 
yading@10: In terms of speed B<pc> mode is by far the fastest and B<pcn_ub> is
yading@10: the slowest.
yading@10: 
yading@10: Default value is I<pc_n>.
yading@10: 
yading@10: 
yading@10: =item B<ppsrc>
yading@10: 
yading@10: Mark the main input stream as a pre-processed input, and enable the secondary
yading@10: input stream as the clean source to pick the fields from. See the filter
yading@10: introduction for more details. It is similar to the B<clip2> feature from
yading@10: VFM/TFM.
yading@10: 
yading@10: Default value is C<0> (disabled).
yading@10: 
yading@10: 
yading@10: =item B<field>
yading@10: 
yading@10: Set the field to match from. It is recommended to set this to the same value as
yading@10: B<order> unless you experience matching failures with that setting. In
yading@10: certain circumstances changing the field that is used to match from can have a
yading@10: large impact on matching performance. Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<auto>
yading@10: 
yading@10: Automatic (same value as B<order>).
yading@10: 
yading@10: =item B<bottom>
yading@10: 
yading@10: Match from the bottom field.
yading@10: 
yading@10: =item B<top>
yading@10: 
yading@10: Match from the top field.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is I<auto>.
yading@10: 
yading@10: 
yading@10: =item B<mchroma>
yading@10: 
yading@10: Set whether or not chroma is included during the match comparisons. In most
yading@10: cases it is recommended to leave this enabled. You should set this to C<0>
yading@10: only if your clip has bad chroma problems such as heavy rainbowing or other
yading@10: artifacts. Setting this to C<0> could also be used to speed things up at
yading@10: the cost of some accuracy.
yading@10: 
yading@10: Default value is C<1>.
yading@10: 
yading@10: 
yading@10: =item B<y0>
yading@10: 
yading@10: 
yading@10: =item B<y1>
yading@10: 
yading@10: These define an exclusion band which excludes the lines between B<y0> and
yading@10: B<y1> from being included in the field matching decision. An exclusion
yading@10: band can be used to ignore subtitles, a logo, or other things that may
yading@10: interfere with the matching. B<y0> sets the starting scan line and
yading@10: B<y1> sets the ending line; all lines in between B<y0> and
yading@10: B<y1> (including B<y0> and B<y1>) will be ignored. Setting
yading@10: B<y0> and B<y1> to the same value will disable the feature.
yading@10: B<y0> and B<y1> defaults to C<0>.
yading@10: 
yading@10: 
yading@10: =item B<scthresh>
yading@10: 
yading@10: Set the scene change detection threshold as a percentage of maximum change on
yading@10: the luma plane. Good values are in the C<[8.0, 14.0]> range. Scene change
yading@10: detection is only relevant in case B<combmatch>=I<sc>.  The range for
yading@10: B<scthresh> is C<[0.0, 100.0]>.
yading@10: 
yading@10: Default value is C<12.0>.
yading@10: 
yading@10: 
yading@10: =item B<combmatch>
yading@10: 
yading@10: When B<combatch> is not I<none>, C<fieldmatch> will take into
yading@10: account the combed scores of matches when deciding what match to use as the
yading@10: final match. Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: No final matching based on combed scores.
yading@10: 
yading@10: =item B<sc>
yading@10: 
yading@10: Combed scores are only used when a scene change is detected.
yading@10: 
yading@10: =item B<full>
yading@10: 
yading@10: Use combed scores all the time.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default is I<sc>.
yading@10: 
yading@10: 
yading@10: =item B<combdbg>
yading@10: 
yading@10: Force C<fieldmatch> to calculate the combed metrics for certain matches and
yading@10: print them. This setting is known as B<micout> in TFM/VFM vocabulary.
yading@10: Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: No forced calculation.
yading@10: 
yading@10: =item B<pcn>
yading@10: 
yading@10: Force p/c/n calculations.
yading@10: 
yading@10: =item B<pcnub>
yading@10: 
yading@10: Force p/c/n/u/b calculations.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is I<none>.
yading@10: 
yading@10: 
yading@10: =item B<cthresh>
yading@10: 
yading@10: This is the area combing threshold used for combed frame detection. This
yading@10: essentially controls how "strong" or "visible" combing must be to be detected.
yading@10: Larger values mean combing must be more visible and smaller values mean combing
yading@10: can be less visible or strong and still be detected. Valid settings are from
yading@10: C<-1> (every pixel will be detected as combed) to C<255> (no pixel will
yading@10: be detected as combed). This is basically a pixel difference value. A good
yading@10: range is C<[8, 12]>.
yading@10: 
yading@10: Default value is C<9>.
yading@10: 
yading@10: 
yading@10: =item B<chroma>
yading@10: 
yading@10: Sets whether or not chroma is considered in the combed frame decision.  Only
yading@10: disable this if your source has chroma problems (rainbowing, etc.) that are
yading@10: causing problems for the combed frame detection with chroma enabled. Actually,
yading@10: using B<chroma>=I<0> is usually more reliable, except for the case
yading@10: where there is chroma only combing in the source.
yading@10: 
yading@10: Default value is C<0>.
yading@10: 
yading@10: 
yading@10: =item B<blockx>
yading@10: 
yading@10: 
yading@10: =item B<blocky>
yading@10: 
yading@10: Respectively set the x-axis and y-axis size of the window used during combed
yading@10: frame detection. This has to do with the size of the area in which
yading@10: B<combpel> pixels are required to be detected as combed for a frame to be
yading@10: declared combed. See the B<combpel> parameter description for more info.
yading@10: Possible values are any number that is a power of 2 starting at 4 and going up
yading@10: to 512.
yading@10: 
yading@10: Default value is C<16>.
yading@10: 
yading@10: 
yading@10: =item B<combpel>
yading@10: 
yading@10: The number of combed pixels inside any of the B<blocky> by
yading@10: B<blockx> size blocks on the frame for the frame to be detected as
yading@10: combed. While B<cthresh> controls how "visible" the combing must be, this
yading@10: setting controls "how much" combing there must be in any localized area (a
yading@10: window defined by the B<blockx> and B<blocky> settings) on the
yading@10: frame. Minimum value is C<0> and maximum is C<blocky x blockx> (at
yading@10: which point no frames will ever be detected as combed). This setting is known
yading@10: as B<MI> in TFM/VFM vocabulary.
yading@10: 
yading@10: Default value is C<80>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 p/c/n/u/b meaning
yading@10: 
yading@10: 
yading@10: 
yading@10: =head4 p/c/n
yading@10: 
yading@10: 
yading@10: We assume the following telecined stream:
yading@10: 
yading@10: 	
yading@10: 	Top fields:     1 2 2 3 4
yading@10: 	Bottom fields:  1 2 3 4 4
yading@10: 
yading@10: 
yading@10: The numbers correspond to the progressive frame the fields relate to. Here, the
yading@10: first two frames are progressive, the 3rd and 4th are combed, and so on.
yading@10: 
yading@10: When C<fieldmatch> is configured to run a matching from bottom
yading@10: (B<field>=I<bottom>) this is how this input stream get transformed:
yading@10: 
yading@10: 	
yading@10: 	Input stream:
yading@10: 	                T     1 2 2 3 4
yading@10: 	                B     1 2 3 4 4   <-- matching reference
yading@10: 	
yading@10: 	Matches:              c c n n c
yading@10: 	
yading@10: 	Output stream:
yading@10: 	                T     1 2 3 4 4
yading@10: 	                B     1 2 3 4 4
yading@10: 
yading@10: 
yading@10: As a result of the field matching, we can see that some frames get duplicated.
yading@10: To perform a complete inverse telecine, you need to rely on a decimation filter
yading@10: after this operation. See for instance the decimate filter.
yading@10: 
yading@10: The same operation now matching from top fields (B<field>=I<top>)
yading@10: looks like this:
yading@10: 
yading@10: 	
yading@10: 	Input stream:
yading@10: 	                T     1 2 2 3 4   <-- matching reference
yading@10: 	                B     1 2 3 4 4
yading@10: 	
yading@10: 	Matches:              c c p p c
yading@10: 	
yading@10: 	Output stream:
yading@10: 	                T     1 2 2 3 4
yading@10: 	                B     1 2 2 3 4
yading@10: 
yading@10: 
yading@10: In these examples, we can see what I<p>, I<c> and I<n> mean;
yading@10: basically, they refer to the frame and field of the opposite parity:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *<I<p> matches the field of the opposite parity in the previous frame>
yading@10: 
yading@10: 
yading@10: =item *<I<c> matches the field of the opposite parity in the current frame>
yading@10: 
yading@10: 
yading@10: =item *<I<n> matches the field of the opposite parity in the next frame>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head4 u/b
yading@10: 
yading@10: 
yading@10: The I<u> and I<b> matching are a bit special in the sense that they match
yading@10: from the opposite parity flag. In the following examples, we assume that we are
yading@10: currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
yading@10: 'x' is placed above and below each matched fields.
yading@10: 
yading@10: With bottom matching (B<field>=I<bottom>):
yading@10: 	
yading@10: 	Match:           c         p           n          b          u
yading@10: 	
yading@10: 	                 x       x               x        x          x
yading@10: 	  Top          1 2 2     1 2 2       1 2 2      1 2 2      1 2 2
yading@10: 	  Bottom       1 2 3     1 2 3       1 2 3      1 2 3      1 2 3
yading@10: 	                 x         x           x        x              x
yading@10: 	
yading@10: 	Output frames:
yading@10: 	                 2          1          2          2          2
yading@10: 	                 2          2          2          1          3
yading@10: 
yading@10: 
yading@10: With top matching (B<field>=I<top>):
yading@10: 	
yading@10: 	Match:           c         p           n          b          u
yading@10: 	
yading@10: 	                 x         x           x        x              x
yading@10: 	  Top          1 2 2     1 2 2       1 2 2      1 2 2      1 2 2
yading@10: 	  Bottom       1 2 3     1 2 3       1 2 3      1 2 3      1 2 3
yading@10: 	                 x       x               x        x          x
yading@10: 	
yading@10: 	Output frames:
yading@10: 	                 2          2          2          1          2
yading@10: 	                 2          1          3          2          2
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: Simple IVTC of a top field first telecined stream:
yading@10: 	
yading@10: 	fieldmatch=order=tff:combmatch=none, decimate
yading@10: 
yading@10: 
yading@10: Advanced IVTC, with fallback on yadif for still combed frames:
yading@10: 	
yading@10: 	fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 fieldorder
yading@10: 
yading@10: 
yading@10: Transform the field order of the input video.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<order>
yading@10: 
yading@10: Output field order. Valid values are I<tff> for top field first or I<bff>
yading@10: for bottom field first.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<tff>.
yading@10: 
yading@10: Transformation is achieved by shifting the picture content up or down
yading@10: by one line, and filling the remaining line with appropriate picture content.
yading@10: This method is consistent with most broadcast field order converters.
yading@10: 
yading@10: If the input video is not flagged as being interlaced, or it is already
yading@10: flagged as being of the required output field order then this filter does
yading@10: not alter the incoming video.
yading@10: 
yading@10: This filter is very useful when converting to or from PAL DV material,
yading@10: which is bottom field first.
yading@10: 
yading@10: For example:
yading@10: 	
yading@10: 	ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 fifo
yading@10: 
yading@10: 
yading@10: Buffer input images and send them when they are requested.
yading@10: 
yading@10: This filter is mainly useful when auto-inserted by the libavfilter
yading@10: framework.
yading@10: 
yading@10: The filter does not take parameters.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 format
yading@10: 
yading@10: 
yading@10: Convert the input video to one of the specified pixel formats.
yading@10: Libavfilter will try to pick one that is supported for the input to
yading@10: the next filter.
yading@10: 
yading@10: This filter accepts the following parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<pix_fmts>
yading@10: 
yading@10: A '|'-separated list of pixel format names, for example
yading@10: "pix_fmts=yuv420p|monow|rgb24".
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Convert the input video to the format I<yuv420p>
yading@10: 	
yading@10: 	format=pix_fmts=yuv420p
yading@10: 
yading@10: 
yading@10: Convert the input video to any of the formats in the list
yading@10: 	
yading@10: 	format=pix_fmts=yuv420p|yuv444p|yuv410p
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 fps
yading@10: 
yading@10: 
yading@10: Convert the video to specified constant frame rate by duplicating or dropping
yading@10: frames as necessary.
yading@10: 
yading@10: This filter accepts the following named parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<fps>
yading@10: 
yading@10: Desired output frame rate. The default is C<25>.
yading@10: 
yading@10: 
yading@10: =item B<round>
yading@10: 
yading@10: Rounding method.
yading@10: 
yading@10: Possible values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<zero>
yading@10: 
yading@10: zero round towards 0
yading@10: 
yading@10: =item B<inf>
yading@10: 
yading@10: round away from 0
yading@10: 
yading@10: =item B<down>
yading@10: 
yading@10: round towards -infinity
yading@10: 
yading@10: =item B<up>
yading@10: 
yading@10: round towards +infinity
yading@10: 
yading@10: =item B<near>
yading@10: 
yading@10: round to nearest
yading@10: 
yading@10: =back
yading@10: 
yading@10: The default is C<near>.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Alternatively, the options can be specified as a flat string:
yading@10: I<fps>[:I<round>].
yading@10: 
yading@10: See also the setpts filter.
yading@10: 
yading@10: 
yading@10: =head2 framestep
yading@10: 
yading@10: 
yading@10: Select one frame every N-th frame.
yading@10: 
yading@10: This filter accepts the following option:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<step>
yading@10: 
yading@10: Select frame after every C<step> frames.
yading@10: Allowed values are positive integers higher than 0. Default value is C<1>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 frei0r
yading@10: 
yading@10: 
yading@10: Apply a frei0r effect to the input video.
yading@10: 
yading@10: To enable compilation of this filter you need to install the frei0r
yading@10: header and configure FFmpeg with C<--enable-frei0r>.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<filter_name>
yading@10: 
yading@10: The name to the frei0r effect to load. If the environment variable
yading@10: B<FREI0R_PATH> is defined, the frei0r effect is searched in each one of the
yading@10: directories specified by the colon separated list in B<FREIOR_PATH>,
yading@10: otherwise in the standard frei0r paths, which are in this order:
yading@10: F<HOME/.frei0r-1/lib/>, F</usr/local/lib/frei0r-1/>,
yading@10: F</usr/lib/frei0r-1/>.
yading@10: 
yading@10: 
yading@10: =item B<filter_params>
yading@10: 
yading@10: A '|'-separated list of parameters to pass to the frei0r effect.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: A frei0r effect parameter can be a boolean (whose values are specified
yading@10: with "y" and "n"), a double, a color (specified by the syntax
yading@10: I<R>/I<G>/I<B>, I<R>, I<G>, and I<B> being float
yading@10: numbers from 0.0 to 1.0) or by an C<av_parse_color()> color
yading@10: description), a position (specified by the syntax I<X>/I<Y>,
yading@10: I<X> and I<Y> being float numbers) and a string.
yading@10: 
yading@10: The number and kind of parameters depend on the loaded effect. If an
yading@10: effect parameter is not specified the default value is set.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply the distort0r effect, set the first two double parameters:
yading@10: 	
yading@10: 	frei0r=filter_name=distort0r:filter_params=0.5|0.01
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply the colordistance effect, take a color as first parameter:
yading@10: 	
yading@10: 	frei0r=colordistance:0.2/0.3/0.4
yading@10: 	frei0r=colordistance:violet
yading@10: 	frei0r=colordistance:0x112233
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply the perspective effect, specify the top left and top right image
yading@10: positions:
yading@10: 	
yading@10: 	frei0r=perspective:0.2/0.2|0.8/0.2
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For more information see:
yading@10: E<lt>B<http://frei0r.dyne.org>E<gt>
yading@10: 
yading@10: 
yading@10: =head2 geq
yading@10: 
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<lum_expr>
yading@10: 
yading@10: the luminance expression
yading@10: 
yading@10: =item B<cb_expr>
yading@10: 
yading@10: the chrominance blue expression
yading@10: 
yading@10: =item B<cr_expr>
yading@10: 
yading@10: the chrominance red expression
yading@10: 
yading@10: =item B<alpha_expr>
yading@10: 
yading@10: the alpha expression
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If one of the chrominance expression is not defined, it falls back on the other
yading@10: one. If no alpha expression is specified it will evaluate to opaque value.
yading@10: If none of chrominance expressions are
yading@10: specified, they will evaluate the luminance expression.
yading@10: 
yading@10: The expressions can use the following variables and functions:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<N>
yading@10: 
yading@10: The sequential number of the filtered frame, starting from C<0>.
yading@10: 
yading@10: 
yading@10: =item B<X>
yading@10: 
yading@10: 
yading@10: =item B<Y>
yading@10: 
yading@10: The coordinates of the current sample.
yading@10: 
yading@10: 
yading@10: =item B<W>
yading@10: 
yading@10: 
yading@10: =item B<H>
yading@10: 
yading@10: The width and height of the image.
yading@10: 
yading@10: 
yading@10: =item B<SW>
yading@10: 
yading@10: 
yading@10: =item B<SH>
yading@10: 
yading@10: Width and height scale depending on the currently filtered plane. It is the
yading@10: ratio between the corresponding luma plane number of pixels and the current
yading@10: plane ones. E.g. for YUV4:2:0 the values are C<1,1> for the luma plane, and
yading@10: C<0.5,0.5> for chroma planes.
yading@10: 
yading@10: 
yading@10: =item B<T>
yading@10: 
yading@10: Time of the current frame, expressed in seconds.
yading@10: 
yading@10: 
yading@10: =item B<p(x, y)>
yading@10: 
yading@10: Return the value of the pixel at location (I<x>,I<y>) of the current
yading@10: plane.
yading@10: 
yading@10: 
yading@10: =item B<lum(x, y)>
yading@10: 
yading@10: Return the value of the pixel at location (I<x>,I<y>) of the luminance
yading@10: plane.
yading@10: 
yading@10: 
yading@10: =item B<cb(x, y)>
yading@10: 
yading@10: Return the value of the pixel at location (I<x>,I<y>) of the
yading@10: blue-difference chroma plane. Returns 0 if there is no such plane.
yading@10: 
yading@10: 
yading@10: =item B<cr(x, y)>
yading@10: 
yading@10: Return the value of the pixel at location (I<x>,I<y>) of the
yading@10: red-difference chroma plane. Returns 0 if there is no such plane.
yading@10: 
yading@10: 
yading@10: =item B<alpha(x, y)>
yading@10: 
yading@10: Return the value of the pixel at location (I<x>,I<y>) of the alpha
yading@10: plane. Returns 0 if there is no such plane.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For functions, if I<x> and I<y> are outside the area, the value will be
yading@10: automatically clipped to the closer edge.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Flip the image horizontally:
yading@10: 	
yading@10: 	geq=p(W-X\,Y)
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a bidimensional sine wave, with angle C<PI/3> and a
yading@10: wavelength of 100 pixels:
yading@10: 	
yading@10: 	geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a fancy enigmatic moving light:
yading@10: 	
yading@10: 	nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a quick emboss effect:
yading@10: 	
yading@10: 	format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 gradfun
yading@10: 
yading@10: 
yading@10: Fix the banding artifacts that are sometimes introduced into nearly flat
yading@10: regions by truncation to 8bit color depth.
yading@10: Interpolate the gradients that should go where the bands are, and
yading@10: dither them.
yading@10: 
yading@10: This filter is designed for playback only.  Do not use it prior to
yading@10: lossy compression, because compression tends to lose the dither and
yading@10: bring back the bands.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<strength>
yading@10: 
yading@10: The maximum amount by which the filter will change any one pixel. Also the
yading@10: threshold for detecting nearly flat regions. Acceptable values range from .51 to
yading@10: 64, default value is 1.2, out-of-range values will be clipped to the valid
yading@10: range.
yading@10: 
yading@10: 
yading@10: =item B<radius>
yading@10: 
yading@10: The neighborhood to fit the gradient to. A larger radius makes for smoother
yading@10: gradients, but also prevents the filter from modifying the pixels near detailed
yading@10: regions. Acceptable values are 8-32, default value is 16, out-of-range values
yading@10: will be clipped to the valid range.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Alternatively, the options can be specified as a flat string:
yading@10: I<strength>[:I<radius>]
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply the filter with a C<3.5> strength and radius of C<8>:
yading@10: 	
yading@10: 	gradfun=3.5:8
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify radius, omitting the strength (which will fall-back to the default
yading@10: value):
yading@10: 	
yading@10: 	gradfun=radius=8
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 hflip
yading@10: 
yading@10: 
yading@10: Flip the input video horizontally.
yading@10: 
yading@10: For example to horizontally flip the input video with B<ffmpeg>:
yading@10: 	
yading@10: 	ffmpeg -i in.avi -vf "hflip" out.avi
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 histeq
yading@10: 
yading@10: This filter applies a global color histogram equalization on a
yading@10: per-frame basis.
yading@10: 
yading@10: It can be used to correct video that has a compressed range of pixel
yading@10: intensities.  The filter redistributes the pixel intensities to
yading@10: equalize their distribution across the intensity range. It may be
yading@10: viewed as an "automatically adjusting contrast filter". This filter is
yading@10: useful only for correcting degraded or poorly captured source
yading@10: video.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<strength>
yading@10: 
yading@10: Determine the amount of equalization to be applied.  As the strength
yading@10: is reduced, the distribution of pixel intensities more-and-more
yading@10: approaches that of the input frame. The value must be a float number
yading@10: in the range [0,1] and defaults to 0.200.
yading@10: 
yading@10: 
yading@10: =item B<intensity>
yading@10: 
yading@10: Set the maximum intensity that can generated and scale the output
yading@10: values appropriately.  The strength should be set as desired and then
yading@10: the intensity can be limited if needed to avoid washing-out. The value
yading@10: must be a float number in the range [0,1] and defaults to 0.210.
yading@10: 
yading@10: 
yading@10: =item B<antibanding>
yading@10: 
yading@10: Set the antibanding level. If enabled the filter will randomly vary
yading@10: the luminance of output pixels by a small amount to avoid banding of
yading@10: the histogram. Possible values are C<none>, C<weak> or
yading@10: C<strong>. It defaults to C<none>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 histogram
yading@10: 
yading@10: 
yading@10: Compute and draw a color distribution histogram for the input video.
yading@10: 
yading@10: The computed histogram is a representation of distribution of color components
yading@10: in an image.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Set histogram mode.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<levels>
yading@10: 
yading@10: standard histogram that display color components distribution in an image.
yading@10: Displays color graph for each color component. Shows distribution
yading@10: of the Y, U, V, A or G, B, R components, depending on input format,
yading@10: in current frame. Bellow each graph is color component scale meter.
yading@10: 
yading@10: 
yading@10: =item B<color>
yading@10: 
yading@10: chroma values in vectorscope, if brighter more such chroma values are
yading@10: distributed in an image.
yading@10: Displays chroma values (U/V color placement) in two dimensional graph
yading@10: (which is called a vectorscope). It can be used to read of the hue and
yading@10: saturation of the current frame. At a same time it is a histogram.
yading@10: The whiter a pixel in the vectorscope, the more pixels of the input frame
yading@10: correspond to that pixel (that is the more pixels have this chroma value).
yading@10: The V component is displayed on the horizontal (X) axis, with the leftmost
yading@10: side being V = 0 and the rightmost side being V = 255.
yading@10: The U component is displayed on the vertical (Y) axis, with the top
yading@10: representing U = 0 and the bottom representing U = 255.
yading@10: 
yading@10: The position of a white pixel in the graph corresponds to the chroma value
yading@10: of a pixel of the input clip. So the graph can be used to read of the
yading@10: hue (color flavor) and the saturation (the dominance of the hue in the color).
yading@10: As the hue of a color changes, it moves around the square. At the center of
yading@10: the square, the saturation is zero, which means that the corresponding pixel
yading@10: has no color. If you increase the amount of a specific color, while leaving
yading@10: the other colors unchanged, the saturation increases, and you move towards
yading@10: the edge of the square.
yading@10: 
yading@10: 
yading@10: =item B<color2>
yading@10: 
yading@10: chroma values in vectorscope, similar as C<color> but actual chroma values
yading@10: are displayed.
yading@10: 
yading@10: 
yading@10: =item B<waveform>
yading@10: 
yading@10: per row/column color component graph. In row mode graph in the left side represents
yading@10: color component value 0 and right side represents value = 255. In column mode top
yading@10: side represents color component value = 0 and bottom side represents value = 255.
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default value is C<levels>.
yading@10: 
yading@10: 
yading@10: =item B<level_height>
yading@10: 
yading@10: Set height of level in C<levels>. Default value is C<200>.
yading@10: Allowed range is [50, 2048].
yading@10: 
yading@10: 
yading@10: =item B<scale_height>
yading@10: 
yading@10: Set height of color scale in C<levels>. Default value is C<12>.
yading@10: Allowed range is [0, 40].
yading@10: 
yading@10: 
yading@10: =item B<step>
yading@10: 
yading@10: Set step for C<waveform> mode. Smaller values are useful to find out how much
yading@10: of same luminance values across input rows/columns are distributed.
yading@10: Default value is C<10>. Allowed range is [1, 255].
yading@10: 
yading@10: 
yading@10: =item B<waveform_mode>
yading@10: 
yading@10: Set mode for C<waveform>. Can be either C<row>, or C<column>.
yading@10: Default is C<row>.
yading@10: 
yading@10: 
yading@10: =item B<display_mode>
yading@10: 
yading@10: Set display mode for C<waveform> and C<levels>.
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<parade>
yading@10: 
yading@10: Display separate graph for the color components side by side in
yading@10: C<row> waveform mode or one below other in C<column> waveform mode
yading@10: for C<waveform> histogram mode. For C<levels> histogram mode
yading@10: per color component graphs are placed one bellow other.
yading@10: 
yading@10: This display mode in C<waveform> histogram mode makes it easy to spot
yading@10: color casts in the highlights and shadows of an image, by comparing the
yading@10: contours of the top and the bottom of each waveform.
yading@10: Since whites, grays, and blacks are characterized by
yading@10: exactly equal amounts of red, green, and blue, neutral areas of the
yading@10: picture should display three waveforms of roughly equal width/height.
yading@10: If not, the correction is easy to make by making adjustments to level the
yading@10: three waveforms.
yading@10: 
yading@10: 
yading@10: =item B<overlay>
yading@10: 
yading@10: Presents information that's identical to that in the C<parade>, except
yading@10: that the graphs representing color components are superimposed directly
yading@10: over one another.
yading@10: 
yading@10: This display mode in C<waveform> histogram mode can make it easier to spot
yading@10: the relative differences or similarities in overlapping areas of the color
yading@10: components that are supposed to be identical, such as neutral whites, grays,
yading@10: or blacks.
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default is C<parade>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Calculate and draw histogram:
yading@10: 	
yading@10: 	ffplay -i input -vf histogram
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 hqdn3d
yading@10: 
yading@10: 
yading@10: High precision/quality 3d denoise filter. This filter aims to reduce
yading@10: image noise producing smooth images and making still images really
yading@10: still. It should enhance compressibility.
yading@10: 
yading@10: It accepts the following optional parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<luma_spatial>
yading@10: 
yading@10: a non-negative float number which specifies spatial luma strength,
yading@10: defaults to 4.0
yading@10: 
yading@10: 
yading@10: =item B<chroma_spatial>
yading@10: 
yading@10: a non-negative float number which specifies spatial chroma strength,
yading@10: defaults to 3.0*I<luma_spatial>/4.0
yading@10: 
yading@10: 
yading@10: =item B<luma_tmp>
yading@10: 
yading@10: a float number which specifies luma temporal strength, defaults to
yading@10: 6.0*I<luma_spatial>/4.0
yading@10: 
yading@10: 
yading@10: =item B<chroma_tmp>
yading@10: 
yading@10: a float number which specifies chroma temporal strength, defaults to
yading@10: I<luma_tmp>*I<chroma_spatial>/I<luma_spatial>
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 hue
yading@10: 
yading@10: 
yading@10: Modify the hue and/or the saturation of the input.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: Specify the hue angle as a number of degrees. It accepts an expression,
yading@10: and defaults to "0".
yading@10: 
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: Specify the saturation in the [-10,10] range. It accepts a float number and
yading@10: defaults to "1".
yading@10: 
yading@10: 
yading@10: =item B<H>
yading@10: 
yading@10: Specify the hue angle as a number of radians. It accepts a float
yading@10: number or an expression, and defaults to "0".
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: B<h> and B<H> are mutually exclusive, and can't be
yading@10: specified at the same time.
yading@10: 
yading@10: The B<h>, B<H> and B<s> option values are
yading@10: expressions containing the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: frame count of the input frame starting from 0
yading@10: 
yading@10: 
yading@10: =item B<pts>
yading@10: 
yading@10: presentation timestamp of the input frame expressed in time base units
yading@10: 
yading@10: 
yading@10: =item B<r>
yading@10: 
yading@10: frame rate of the input video, NAN if the input frame rate is unknown
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: timestamp expressed in seconds, NAN if the input timestamp is unknown
yading@10: 
yading@10: 
yading@10: =item B<tb>
yading@10: 
yading@10: time base of the input video
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the hue to 90 degrees and the saturation to 1.0:
yading@10: 	
yading@10: 	hue=h=90:s=1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Same command but expressing the hue in radians:
yading@10: 	
yading@10: 	hue=H=PI/2:s=1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Rotate hue and make the saturation swing between 0
yading@10: and 2 over a period of 1 second:
yading@10: 	
yading@10: 	hue="H=2*PI*t: s=sin(2*PI*t)+1"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply a 3 seconds saturation fade-in effect starting at 0:
yading@10: 	
yading@10: 	hue="s=min(t/3\,1)"
yading@10: 
yading@10: 
yading@10: The general fade-in expression can be written as:
yading@10: 	
yading@10: 	hue="s=min(0\, max((t-START)/DURATION\, 1))"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
yading@10: 	
yading@10: 	hue="s=max(0\, min(1\, (8-t)/3))"
yading@10: 
yading@10: 
yading@10: The general fade-out expression can be written as:
yading@10: 	
yading@10: 	hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Commands
yading@10: 
yading@10: 
yading@10: This filter supports the following commands:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: 
yading@10: =item B<h>
yading@10: 
yading@10: 
yading@10: =item B<H>
yading@10: 
yading@10: Modify the hue and/or the saturation of the input video.
yading@10: The command accepts the same syntax of the corresponding option.
yading@10: 
yading@10: If the specified expression is not valid, it is kept at its current
yading@10: value.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 idet
yading@10: 
yading@10: 
yading@10: Detect video interlacing type.
yading@10: 
yading@10: This filter tries to detect if the input is interlaced or progressive,
yading@10: top or bottom field first.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<intl_thres>
yading@10: 
yading@10: Set interlacing threshold.
yading@10: 
yading@10: =item B<prog_thres>
yading@10: 
yading@10: Set progressive threshold.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 il
yading@10: 
yading@10: 
yading@10: Deinterleave or interleave fields.
yading@10: 
yading@10: This filter allows to process interlaced images fields without
yading@10: deinterlacing them. Deinterleaving splits the input frame into 2
yading@10: fields (so called half pictures). Odd lines are moved to the top
yading@10: half of the output image, even lines to the bottom half.
yading@10: You can process (filter) them independently and then re-interleave them.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<luma_mode, l>
yading@10: 
yading@10: 
yading@10: =item B<chroma_mode, s>
yading@10: 
yading@10: 
yading@10: =item B<alpha_mode, a>
yading@10: 
yading@10: Available values for I<luma_mode>, I<chroma_mode> and
yading@10: I<alpha_mode> are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: Do nothing.
yading@10: 
yading@10: 
yading@10: =item B<deinterleave, d>
yading@10: 
yading@10: Deinterleave fields, placing one above the other.
yading@10: 
yading@10: 
yading@10: =item B<interleave, i>
yading@10: 
yading@10: Interleave fields. Reverse the effect of deinterleaving.
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default value is C<none>.
yading@10: 
yading@10: 
yading@10: =item B<luma_swap, ls>
yading@10: 
yading@10: 
yading@10: =item B<chroma_swap, cs>
yading@10: 
yading@10: 
yading@10: =item B<alpha_swap, as>
yading@10: 
yading@10: Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is C<0>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 interlace
yading@10: 
yading@10: 
yading@10: Simple interlacing filter from progressive contents. This interleaves upper (or
yading@10: lower) lines from odd frames with lower (or upper) lines from even frames,
yading@10: halving the frame rate and preserving image height.
yading@10: 
yading@10: 	
yading@10: 	   Original        Original             New Frame
yading@10: 	   Frame 'j'      Frame 'j+1'             (tff)
yading@10: 	  ==========      ===========       ==================
yading@10: 	    Line 0  -------------------->    Frame 'j' Line 0
yading@10: 	    Line 1          Line 1  ---->   Frame 'j+1' Line 1
yading@10: 	    Line 2 --------------------->    Frame 'j' Line 2
yading@10: 	    Line 3          Line 3  ---->   Frame 'j+1' Line 3
yading@10: 	     ...             ...                   ...
yading@10: 	New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
yading@10: 
yading@10: 
yading@10: It accepts the following optional parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<scan>
yading@10: 
yading@10: determines whether the interlaced frame is taken from the even (tff - default)
yading@10: or odd (bff) lines of the progressive frame.
yading@10: 
yading@10: 
yading@10: =item B<lowpass>
yading@10: 
yading@10: Enable (default) or disable the vertical lowpass filter to avoid twitter
yading@10: interlacing and reduce moire patterns.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 kerndeint
yading@10: 
yading@10: 
yading@10: Deinterlace input video by applying Donald Graft's adaptive kernel
yading@10: deinterling. Work on interlaced parts of a video to produce
yading@10: progressive frames.
yading@10: 
yading@10: The description of the accepted parameters follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<thresh>
yading@10: 
yading@10: Set the threshold which affects the filter's tolerance when
yading@10: determining if a pixel line must be processed. It must be an integer
yading@10: in the range [0,255] and defaults to 10. A value of 0 will result in
yading@10: applying the process on every pixels.
yading@10: 
yading@10: 
yading@10: =item B<map>
yading@10: 
yading@10: Paint pixels exceeding the threshold value to white if set to 1.
yading@10: Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<order>
yading@10: 
yading@10: Set the fields order. Swap fields if set to 1, leave fields alone if
yading@10: 0. Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<sharp>
yading@10: 
yading@10: Enable additional sharpening if set to 1. Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<twoway>
yading@10: 
yading@10: Enable twoway sharpening if set to 1. Default is 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply default values:
yading@10: 	
yading@10: 	kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Enable additional sharpening:
yading@10: 	
yading@10: 	kerndeint=sharp=1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Paint processed pixels in white:
yading@10: 	
yading@10: 	kerndeint=map=1
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 lut, lutrgb, lutyuv
yading@10: 
yading@10: 
yading@10: Compute a look-up table for binding each pixel component input value
yading@10: to an output value, and apply it to input video.
yading@10: 
yading@10: I<lutyuv> applies a lookup table to a YUV input video, I<lutrgb>
yading@10: to an RGB input video.
yading@10: 
yading@10: These filters accept the following options:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<c0>
yading@10: 
yading@10: set first pixel component expression
yading@10: 
yading@10: =item B<c1>
yading@10: 
yading@10: set second pixel component expression
yading@10: 
yading@10: =item B<c2>
yading@10: 
yading@10: set third pixel component expression
yading@10: 
yading@10: =item B<c3>
yading@10: 
yading@10: set fourth pixel component expression, corresponds to the alpha component
yading@10: 
yading@10: 
yading@10: =item B<r>
yading@10: 
yading@10: set red component expression
yading@10: 
yading@10: =item B<g>
yading@10: 
yading@10: set green component expression
yading@10: 
yading@10: =item B<b>
yading@10: 
yading@10: set blue component expression
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: alpha component expression
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: set Y/luminance component expression
yading@10: 
yading@10: =item B<u>
yading@10: 
yading@10: set U/Cb component expression
yading@10: 
yading@10: =item B<v>
yading@10: 
yading@10: set V/Cr component expression
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Each of them specifies the expression to use for computing the lookup table for
yading@10: the corresponding pixel component values.
yading@10: 
yading@10: The exact component associated to each of the I<c*> options depends on the
yading@10: format in input.
yading@10: 
yading@10: The I<lut> filter requires either YUV or RGB pixel formats in input,
yading@10: I<lutrgb> requires RGB pixel formats in input, and I<lutyuv> requires YUV.
yading@10: 
yading@10: The expressions can contain the following constants and functions:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<w, h>
yading@10: 
yading@10: the input width and height
yading@10: 
yading@10: 
yading@10: =item B<val>
yading@10: 
yading@10: input value for the pixel component
yading@10: 
yading@10: 
yading@10: =item B<clipval>
yading@10: 
yading@10: the input value clipped in the I<minval>-I<maxval> range
yading@10: 
yading@10: 
yading@10: =item B<maxval>
yading@10: 
yading@10: maximum value for the pixel component
yading@10: 
yading@10: 
yading@10: =item B<minval>
yading@10: 
yading@10: minimum value for the pixel component
yading@10: 
yading@10: 
yading@10: =item B<negval>
yading@10: 
yading@10: the negated value for the pixel component value clipped in the
yading@10: I<minval>-I<maxval> range , it corresponds to the expression
yading@10: "maxval-clipval+minval"
yading@10: 
yading@10: 
yading@10: =item B<clip(val)>
yading@10: 
yading@10: the computed value in I<val> clipped in the
yading@10: I<minval>-I<maxval> range
yading@10: 
yading@10: 
yading@10: =item B<gammaval(gamma)>
yading@10: 
yading@10: the computed gamma correction value of the pixel component value
yading@10: clipped in the I<minval>-I<maxval> range, corresponds to the
yading@10: expression
yading@10: "pow((clipval-minval)/(maxval-minval)\,I<gamma>)*(maxval-minval)+minval"
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: All expressions default to "val".
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Negate input video:
yading@10: 	
yading@10: 	lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
yading@10: 	lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
yading@10: 
yading@10: 
yading@10: The above is the same as:
yading@10: 	
yading@10: 	lutrgb="r=negval:g=negval:b=negval"
yading@10: 	lutyuv="y=negval:u=negval:v=negval"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Negate luminance:
yading@10: 	
yading@10: 	lutyuv=y=negval
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Remove chroma components, turns the video into a graytone image:
yading@10: 	
yading@10: 	lutyuv="u=128:v=128"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply a luma burning effect:
yading@10: 	
yading@10: 	lutyuv="y=2*val"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Remove green and blue components:
yading@10: 	
yading@10: 	lutrgb="g=0:b=0"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set a constant alpha channel value on input:
yading@10: 	
yading@10: 	format=rgba,lutrgb=a="maxval-minval/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Correct luminance gamma by a 0.5 factor:
yading@10: 	
yading@10: 	lutyuv=y=gammaval(0.5)
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Discard least significant bits of luma:
yading@10: 	
yading@10: 	lutyuv=y='bitand(val, 128+64+32)'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 mp
yading@10: 
yading@10: 
yading@10: Apply an MPlayer filter to the input video.
yading@10: 
yading@10: This filter provides a wrapper around most of the filters of
yading@10: MPlayer/MEncoder.
yading@10: 
yading@10: This wrapper is considered experimental. Some of the wrapped filters
yading@10: may not work properly and we may drop support for them, as they will
yading@10: be implemented natively into FFmpeg. Thus you should avoid
yading@10: depending on them when writing portable scripts.
yading@10: 
yading@10: The filters accepts the parameters:
yading@10: I<filter_name>[:=]I<filter_params>
yading@10: 
yading@10: I<filter_name> is the name of a supported MPlayer filter,
yading@10: I<filter_params> is a string containing the parameters accepted by
yading@10: the named filter.
yading@10: 
yading@10: The list of the currently supported filters follows:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item I<dint>
yading@10: 
yading@10: 
yading@10: =item I<down3dright>
yading@10: 
yading@10: 
yading@10: =item I<eq2>
yading@10: 
yading@10: 
yading@10: =item I<eq>
yading@10: 
yading@10: 
yading@10: =item I<fil>
yading@10: 
yading@10: 
yading@10: =item I<fspp>
yading@10: 
yading@10: 
yading@10: =item I<ilpack>
yading@10: 
yading@10: 
yading@10: =item I<mcdeint>
yading@10: 
yading@10: 
yading@10: =item I<ow>
yading@10: 
yading@10: 
yading@10: =item I<perspective>
yading@10: 
yading@10: 
yading@10: =item I<phase>
yading@10: 
yading@10: 
yading@10: =item I<pp7>
yading@10: 
yading@10: 
yading@10: =item I<pullup>
yading@10: 
yading@10: 
yading@10: =item I<qp>
yading@10: 
yading@10: 
yading@10: =item I<sab>
yading@10: 
yading@10: 
yading@10: =item I<softpulldown>
yading@10: 
yading@10: 
yading@10: =item I<spp>
yading@10: 
yading@10: 
yading@10: =item I<tinterlace>
yading@10: 
yading@10: 
yading@10: =item I<uspp>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The parameter syntax and behavior for the listed filters are the same
yading@10: of the corresponding MPlayer filters. For detailed instructions check
yading@10: the "VIDEO FILTERS" section in the MPlayer manual.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Adjust gamma, brightness, contrast:
yading@10: 	
yading@10: 	mp=eq2=1.0:2:0.5
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: See also mplayer(1), E<lt>B<http://www.mplayerhq.hu/>E<gt>.
yading@10: 
yading@10: 
yading@10: =head2 mpdecimate
yading@10: 
yading@10: 
yading@10: Drop frames that do not differ greatly from the previous frame in
yading@10: order to reduce frame rate.
yading@10: 
yading@10: The main use of this filter is for very-low-bitrate encoding
yading@10: (e.g. streaming over dialup modem), but it could in theory be used for
yading@10: fixing movies that were inverse-telecined incorrectly.
yading@10: 
yading@10: A description of the accepted options follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<max>
yading@10: 
yading@10: Set the maximum number of consecutive frames which can be dropped (if
yading@10: positive), or the minimum interval between dropped frames (if
yading@10: negative). If the value is 0, the frame is dropped unregarding the
yading@10: number of previous sequentially dropped frames.
yading@10: 
yading@10: Default value is 0.
yading@10: 
yading@10: 
yading@10: =item B<hi>
yading@10: 
yading@10: 
yading@10: =item B<lo>
yading@10: 
yading@10: 
yading@10: =item B<frac>
yading@10: 
yading@10: Set the dropping threshold values.
yading@10: 
yading@10: Values for B<hi> and B<lo> are for 8x8 pixel blocks and
yading@10: represent actual pixel value differences, so a threshold of 64
yading@10: corresponds to 1 unit of difference for each pixel, or the same spread
yading@10: out differently over the block.
yading@10: 
yading@10: A frame is a candidate for dropping if no 8x8 blocks differ by more
yading@10: than a threshold of B<hi>, and if no more than B<frac> blocks (1
yading@10: meaning the whole image) differ by more than a threshold of B<lo>.
yading@10: 
yading@10: Default value for B<hi> is 64*12, default value for B<lo> is
yading@10: 64*5, and default value for B<frac> is 0.33.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 negate
yading@10: 
yading@10: 
yading@10: Negate input video.
yading@10: 
yading@10: This filter accepts an integer in input, if non-zero it negates the
yading@10: alpha component (if available). The default value in input is 0.
yading@10: 
yading@10: 
yading@10: =head2 noformat
yading@10: 
yading@10: 
yading@10: Force libavfilter not to use any of the specified pixel formats for the
yading@10: input to the next filter.
yading@10: 
yading@10: This filter accepts the following parameters:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<pix_fmts>
yading@10: 
yading@10: A '|'-separated list of pixel format names, for example
yading@10: "pix_fmts=yuv420p|monow|rgb24".
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Force libavfilter to use a format different from I<yuv420p> for the
yading@10: input to the vflip filter:
yading@10: 	
yading@10: 	noformat=pix_fmts=yuv420p,vflip
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Convert the input video to any of the formats not contained in the list:
yading@10: 	
yading@10: 	noformat=yuv420p|yuv444p|yuv410p
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 noise
yading@10: 
yading@10: 
yading@10: Add noise on video input frame.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<all_seed>
yading@10: 
yading@10: 
yading@10: =item B<c0_seed>
yading@10: 
yading@10: 
yading@10: =item B<c1_seed>
yading@10: 
yading@10: 
yading@10: =item B<c2_seed>
yading@10: 
yading@10: 
yading@10: =item B<c3_seed>
yading@10: 
yading@10: Set noise seed for specific pixel component or all pixel components in case
yading@10: of I<all_seed>. Default value is C<123457>.
yading@10: 
yading@10: 
yading@10: =item B<all_strength, alls>
yading@10: 
yading@10: 
yading@10: =item B<c0_strength, c0s>
yading@10: 
yading@10: 
yading@10: =item B<c1_strength, c1s>
yading@10: 
yading@10: 
yading@10: =item B<c2_strength, c2s>
yading@10: 
yading@10: 
yading@10: =item B<c3_strength, c3s>
yading@10: 
yading@10: Set noise strength for specific pixel component or all pixel components in case
yading@10: I<all_strength>. Default value is C<0>. Allowed range is [0, 100].
yading@10: 
yading@10: 
yading@10: =item B<all_flags, allf>
yading@10: 
yading@10: 
yading@10: =item B<c0_flags, c0f>
yading@10: 
yading@10: 
yading@10: =item B<c1_flags, c1f>
yading@10: 
yading@10: 
yading@10: =item B<c2_flags, c2f>
yading@10: 
yading@10: 
yading@10: =item B<c3_flags, c3f>
yading@10: 
yading@10: Set pixel component flags or set flags for all components if I<all_flags>.
yading@10: Available values for component flags are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: averaged temporal noise (smoother)
yading@10: 
yading@10: =item B<p>
yading@10: 
yading@10: mix random noise with a (semi)regular pattern
yading@10: 
yading@10: =item B<q>
yading@10: 
yading@10: higher quality (slightly better looking, slightly slower)
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: temporal noise (noise pattern changes between frames)
yading@10: 
yading@10: =item B<u>
yading@10: 
yading@10: uniform noise (gaussian otherwise)
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: Add temporal and uniform noise to input video:
yading@10: 	
yading@10: 	noise=alls=20:allf=t+u
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 null
yading@10: 
yading@10: 
yading@10: Pass the video source unchanged to the output.
yading@10: 
yading@10: 
yading@10: =head2 ocv
yading@10: 
yading@10: 
yading@10: Apply video transform using libopencv.
yading@10: 
yading@10: To enable this filter install libopencv library and headers and
yading@10: configure FFmpeg with C<--enable-libopencv>.
yading@10: 
yading@10: This filter accepts the following parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<filter_name>
yading@10: 
yading@10: The name of the libopencv filter to apply.
yading@10: 
yading@10: 
yading@10: =item B<filter_params>
yading@10: 
yading@10: The parameters to pass to the libopencv filter. If not specified the default
yading@10: values are assumed.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Refer to the official libopencv documentation for more precise
yading@10: information:
yading@10: E<lt>B<http://opencv.willowgarage.com/documentation/c/image_filtering.html>E<gt>
yading@10: 
yading@10: Follows the list of supported libopencv filters.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 dilate
yading@10: 
yading@10: 
yading@10: Dilate an image by using a specific structuring element.
yading@10: This filter corresponds to the libopencv function C<cvDilate>.
yading@10: 
yading@10: It accepts the parameters: I<struct_el>|I<nb_iterations>.
yading@10: 
yading@10: I<struct_el> represents a structuring element, and has the syntax:
yading@10: I<cols>xI<rows>+I<anchor_x>xI<anchor_y>/I<shape>
yading@10: 
yading@10: I<cols> and I<rows> represent the number of columns and rows of
yading@10: the structuring element, I<anchor_x> and I<anchor_y> the anchor
yading@10: point, and I<shape> the shape for the structuring element, and
yading@10: can be one of the values "rect", "cross", "ellipse", "custom".
yading@10: 
yading@10: If the value for I<shape> is "custom", it must be followed by a
yading@10: string of the form "=I<filename>". The file with name
yading@10: I<filename> is assumed to represent a binary image, with each
yading@10: printable character corresponding to a bright pixel. When a custom
yading@10: I<shape> is used, I<cols> and I<rows> are ignored, the number
yading@10: or columns and rows of the read file are assumed instead.
yading@10: 
yading@10: The default value for I<struct_el> is "3x3+0x0/rect".
yading@10: 
yading@10: I<nb_iterations> specifies the number of times the transform is
yading@10: applied to the image, and defaults to 1.
yading@10: 
yading@10: Follow some example:
yading@10: 	
yading@10: 	# use the default values
yading@10: 	ocv=dilate
yading@10: 	
yading@10: 	# dilate using a structuring element with a 5x5 cross, iterate two times
yading@10: 	ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
yading@10: 	
yading@10: 	# read the shape from the file diamond.shape, iterate two times
yading@10: 	# the file diamond.shape may contain a pattern of characters like this:
yading@10: 	#   *
yading@10: 	#  ***
yading@10: 	# *****
yading@10: 	#  ***
yading@10: 	#   *
yading@10: 	# the specified cols and rows are ignored (but not the anchor point coordinates)
yading@10: 	ocv=dilate:0x0+2x2/custom=diamond.shape|2
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 erode
yading@10: 
yading@10: 
yading@10: Erode an image by using a specific structuring element.
yading@10: This filter corresponds to the libopencv function C<cvErode>.
yading@10: 
yading@10: The filter accepts the parameters: I<struct_el>:I<nb_iterations>,
yading@10: with the same syntax and semantics as the dilate filter.
yading@10: 
yading@10: 
yading@10: =head3 smooth
yading@10: 
yading@10: 
yading@10: Smooth the input video.
yading@10: 
yading@10: The filter takes the following parameters:
yading@10: I<type>|I<param1>|I<param2>|I<param3>|I<param4>.
yading@10: 
yading@10: I<type> is the type of smooth filter to apply, and can be one of
yading@10: the following values: "blur", "blur_no_scale", "median", "gaussian",
yading@10: "bilateral". The default value is "gaussian".
yading@10: 
yading@10: I<param1>, I<param2>, I<param3>, and I<param4> are
yading@10: parameters whose meanings depend on smooth type. I<param1> and
yading@10: I<param2> accept integer positive values or 0, I<param3> and
yading@10: I<param4> accept float values.
yading@10: 
yading@10: The default value for I<param1> is 3, the default value for the
yading@10: other parameters is 0.
yading@10: 
yading@10: These parameters correspond to the parameters assigned to the
yading@10: libopencv function C<cvSmooth>.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 overlay
yading@10: 
yading@10: 
yading@10: Overlay one video on top of another.
yading@10: 
yading@10: It takes two inputs and one output, the first input is the "main"
yading@10: video on which the second input is overlayed.
yading@10: 
yading@10: This filter accepts the following parameters:
yading@10: 
yading@10: A description of the accepted options follows.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: Set the expression for the x and y coordinates of the overlayed video
yading@10: on the main video. Default value is "0" for both expressions. In case
yading@10: the expression is invalid, it is set to a huge value (meaning that the
yading@10: overlay will not be displayed within the output visible area).
yading@10: 
yading@10: 
yading@10: =item B<enable>
yading@10: 
yading@10: Set the expression which enables the overlay. If the evaluation is
yading@10: different from 0, the overlay is displayed on top of the input
yading@10: frame. By default it is "1".
yading@10: 
yading@10: 
yading@10: =item B<eval>
yading@10: 
yading@10: Set when the expressions for B<x>, B<y>, and
yading@10: B<enable> are evaluated.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<init>
yading@10: 
yading@10: only evaluate expressions once during the filter initialization or
yading@10: when a command is processed
yading@10: 
yading@10: 
yading@10: =item B<frame>
yading@10: 
yading@10: evaluate expressions for each incoming frame
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<frame>.
yading@10: 
yading@10: 
yading@10: =item B<shortest>
yading@10: 
yading@10: If set to 1, force the output to terminate when the shortest input
yading@10: terminates. Default value is 0.
yading@10: 
yading@10: 
yading@10: =item B<format>
yading@10: 
yading@10: Set the format for the output video.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<yuv420>
yading@10: 
yading@10: force YUV420 output
yading@10: 
yading@10: 
yading@10: =item B<yuv444>
yading@10: 
yading@10: force YUV444 output
yading@10: 
yading@10: 
yading@10: =item B<rgb>
yading@10: 
yading@10: force RGB output
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<yuv420>.
yading@10: 
yading@10: 
yading@10: =item B<rgb> I<(deprecated)>
yading@10: 
yading@10: If set to 1, force the filter to accept inputs in the RGB
yading@10: color space. Default value is 0. This option is deprecated, use
yading@10: B<format> instead.
yading@10: 
yading@10: 
yading@10: =item B<repeatlast>
yading@10: 
yading@10: If set to 1, force the filter to draw the last overlay frame over the
yading@10: main input until the end of the stream. A value of 0 disables this
yading@10: behavior, which is enabled by default.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The B<x>, B<y>, and B<enable> expressions can
yading@10: contain the following parameters.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<main_w, W>
yading@10: 
yading@10: 
yading@10: =item B<main_h, H>
yading@10: 
yading@10: main input width and height
yading@10: 
yading@10: 
yading@10: =item B<overlay_w, w>
yading@10: 
yading@10: 
yading@10: =item B<overlay_h, h>
yading@10: 
yading@10: overlay input width and height
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: the computed values for I<x> and I<y>. They are evaluated for
yading@10: each new frame.
yading@10: 
yading@10: 
yading@10: =item B<hsub>
yading@10: 
yading@10: 
yading@10: =item B<vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values of the output
yading@10: format. For example for the pixel format "yuv422p" I<hsub> is 2 and
yading@10: I<vsub> is 1.
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: the number of input frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<pos>
yading@10: 
yading@10: the position in the file of the input frame, NAN if unknown
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: timestamp expressed in seconds, NAN if the input timestamp is unknown
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Note that the I<n>, I<pos>, I<t> variables are available only
yading@10: when evaluation is done I<per frame>, and will evaluate to NAN
yading@10: when B<eval> is set to B<init>.
yading@10: 
yading@10: Be aware that frames are taken from each input video in timestamp
yading@10: order, hence, if their initial timestamps differ, it is a a good idea
yading@10: to pass the two inputs through a I<setpts=PTS-STARTPTS> filter to
yading@10: have them begin in the same zero timestamp, as it does the example for
yading@10: the I<movie> filter.
yading@10: 
yading@10: You can chain together more overlays but you should test the
yading@10: efficiency of such approach.
yading@10: 
yading@10: 
yading@10: =head3 Commands
yading@10: 
yading@10: 
yading@10: This filter supports the following commands:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: 
yading@10: =item B<enable>
yading@10: 
yading@10: Modify the x/y and enable overlay of the overlay input.
yading@10: The command accepts the same syntax of the corresponding option.
yading@10: 
yading@10: If the specified expression is not valid, it is kept at its current
yading@10: value.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Draw the overlay at 10 pixels from the bottom right corner of the main
yading@10: video:
yading@10: 	
yading@10: 	overlay=main_w-overlay_w-10:main_h-overlay_h-10
yading@10: 
yading@10: 
yading@10: Using named options the example above becomes:
yading@10: 	
yading@10: 	overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Insert a transparent PNG logo in the bottom left corner of the input,
yading@10: using the B<ffmpeg> tool with the C<-filter_complex> option:
yading@10: 	
yading@10: 	ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Insert 2 different transparent PNG logos (second logo on bottom
yading@10: right corner) using the B<ffmpeg> tool:
yading@10: 	
yading@10: 	ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Add a transparent color layer on top of the main video, C<WxH>
yading@10: must specify the size of the main input to the overlay filter:
yading@10: 	
yading@10: 	color=color=red@.3:size=WxH [over]; [in][over] overlay [out]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Play an original video and a filtered version (here with the deshake
yading@10: filter) side by side using the B<ffplay> tool:
yading@10: 	
yading@10: 	ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
yading@10: 
yading@10: 
yading@10: The above command is the same as:
yading@10: 	
yading@10: 	ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Make a sliding overlay appearing from the left to the right top part of the
yading@10: screen starting since time 2:
yading@10: 	
yading@10: 	overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Compose output by putting two input videos side to side:
yading@10: 	
yading@10: 	ffmpeg -i left.avi -i right.avi -filter_complex "
yading@10: 	nullsrc=size=200x100 [background];
yading@10: 	[0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
yading@10: 	[1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
yading@10: 	[background][left]       overlay=shortest=1       [background+left];
yading@10: 	[background+left][right] overlay=shortest=1:x=100 [left+right]
yading@10: 	"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Chain several overlays in cascade:
yading@10: 	
yading@10: 	nullsrc=s=200x200 [bg];
yading@10: 	testsrc=s=100x100, split=4 [in0][in1][in2][in3];
yading@10: 	[in0] lutrgb=r=0, [bg]   overlay=0:0     [mid0];
yading@10: 	[in1] lutrgb=g=0, [mid0] overlay=100:0   [mid1];
yading@10: 	[in2] lutrgb=b=0, [mid1] overlay=0:100   [mid2];
yading@10: 	[in3] null,       [mid2] overlay=100:100 [out0]
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 pad
yading@10: 
yading@10: 
yading@10: Add paddings to the input image, and place the original input at the
yading@10: given coordinates I<x>, I<y>.
yading@10: 
yading@10: This filter accepts the following parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: 
yading@10: =item B<height, h>
yading@10: 
yading@10: Specify an expression for the size of the output image with the
yading@10: paddings added. If the value for I<width> or I<height> is 0, the
yading@10: corresponding input size is used for the output.
yading@10: 
yading@10: The I<width> expression can reference the value set by the
yading@10: I<height> expression, and vice versa.
yading@10: 
yading@10: The default value of I<width> and I<height> is 0.
yading@10: 
yading@10: 
yading@10: =item B<x>
yading@10: 
yading@10: 
yading@10: =item B<y>
yading@10: 
yading@10: Specify an expression for the offsets where to place the input image
yading@10: in the padded area with respect to the top/left border of the output
yading@10: image.
yading@10: 
yading@10: The I<x> expression can reference the value set by the I<y>
yading@10: expression, and vice versa.
yading@10: 
yading@10: The default value of I<x> and I<y> is 0.
yading@10: 
yading@10: 
yading@10: =item B<color>
yading@10: 
yading@10: Specify the color of the padded area, it can be the name of a color
yading@10: (case insensitive match) or a 0xRRGGBB[AA] sequence.
yading@10: 
yading@10: The default value of I<color> is "black".
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The value for the I<width>, I<height>, I<x>, and I<y>
yading@10: options are expressions containing the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<in_w, in_h>
yading@10: 
yading@10: the input video width and height
yading@10: 
yading@10: 
yading@10: =item B<iw, ih>
yading@10: 
yading@10: same as I<in_w> and I<in_h>
yading@10: 
yading@10: 
yading@10: =item B<out_w, out_h>
yading@10: 
yading@10: the output width and height, that is the size of the padded area as
yading@10: specified by the I<width> and I<height> expressions
yading@10: 
yading@10: 
yading@10: =item B<ow, oh>
yading@10: 
yading@10: same as I<out_w> and I<out_h>
yading@10: 
yading@10: 
yading@10: =item B<x, y>
yading@10: 
yading@10: x and y offsets as specified by the I<x> and I<y>
yading@10: expressions, or NAN if not yet specified
yading@10: 
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: same as I<iw> / I<ih>
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: input sample aspect ratio
yading@10: 
yading@10: 
yading@10: =item B<dar>
yading@10: 
yading@10: input display aspect ratio, it is the same as (I<iw> / I<ih>) * I<sar>
yading@10: 
yading@10: 
yading@10: =item B<hsub, vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values. For example for the
yading@10: pixel format "yuv422p" I<hsub> is 2 and I<vsub> is 1.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Add paddings with color "violet" to the input video. Output video
yading@10: size is 640x480, the top-left corner of the input video is placed at
yading@10: column 0, row 40:
yading@10: 	
yading@10: 	pad=640:480:0:40:violet
yading@10: 
yading@10: 
yading@10: The example above is equivalent to the following command:
yading@10: 	
yading@10: 	pad=width=640:height=480:x=0:y=40:color=violet
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Pad the input to get an output with dimensions increased by 3/2,
yading@10: and put the input video at the center of the padded area:
yading@10: 	
yading@10: 	pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Pad the input to get a squared output with size equal to the maximum
yading@10: value between the input width and height, and put the input video at
yading@10: the center of the padded area:
yading@10: 	
yading@10: 	pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Pad the input to get a final w/h ratio of 16:9:
yading@10: 	
yading@10: 	pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: In case of anamorphic video, in order to set the output display aspect
yading@10: correctly, it is necessary to use I<sar> in the expression,
yading@10: according to the relation:
yading@10: 	
yading@10: 	(ih * X / ih) * sar = output_dar
yading@10: 	X = output_dar / sar
yading@10: 
yading@10: 
yading@10: Thus the previous example needs to be modified to:
yading@10: 	
yading@10: 	pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Double output size and put the input video in the bottom-right
yading@10: corner of the output padded area:
yading@10: 	
yading@10: 	pad="2*iw:2*ih:ow-iw:oh-ih"
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 pixdesctest
yading@10: 
yading@10: 
yading@10: Pixel format descriptor test filter, mainly useful for internal
yading@10: testing. The output video should be equal to the input video.
yading@10: 
yading@10: For example:
yading@10: 	
yading@10: 	format=monow, pixdesctest
yading@10: 
yading@10: 
yading@10: can be used to test the monowhite pixel format descriptor definition.
yading@10: 
yading@10: 
yading@10: =head2 pp
yading@10: 
yading@10: 
yading@10: Enable the specified chain of postprocessing subfilters using libpostproc. This
yading@10: library should be automatically selected with a GPL build (C<--enable-gpl>).
yading@10: Subfilters must be separated by '/' and can be disabled by prepending a '-'.
yading@10: Each subfilter and some options have a short and a long name that can be used
yading@10: interchangeably, i.e. dr/dering are the same.
yading@10: 
yading@10: The filters accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<subfilters>
yading@10: 
yading@10: Set postprocessing subfilters string.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: All subfilters share common options to determine their scope:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<a/autoq>
yading@10: 
yading@10: Honor the quality commands for this subfilter.
yading@10: 
yading@10: 
yading@10: =item B<c/chrom>
yading@10: 
yading@10: Do chrominance filtering, too (default).
yading@10: 
yading@10: 
yading@10: =item B<y/nochrom>
yading@10: 
yading@10: Do luminance filtering only (no chrominance).
yading@10: 
yading@10: 
yading@10: =item B<n/noluma>
yading@10: 
yading@10: Do chrominance filtering only (no luminance).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: These options can be appended after the subfilter name, separated by a '|'.
yading@10: 
yading@10: Available subfilters are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<hb/hdeblock[|difference[|flatness]]>
yading@10: 
yading@10: Horizontal deblocking filter
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<difference>
yading@10: 
yading@10: Difference factor where higher values mean more deblocking (default: C<32>).
yading@10: 
yading@10: =item B<flatness>
yading@10: 
yading@10: Flatness threshold where lower values mean more deblocking (default: C<39>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<vb/vdeblock[|difference[|flatness]]>
yading@10: 
yading@10: Vertical deblocking filter
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<difference>
yading@10: 
yading@10: Difference factor where higher values mean more deblocking (default: C<32>).
yading@10: 
yading@10: =item B<flatness>
yading@10: 
yading@10: Flatness threshold where lower values mean more deblocking (default: C<39>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<ha/hadeblock[|difference[|flatness]]>
yading@10: 
yading@10: Accurate horizontal deblocking filter
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<difference>
yading@10: 
yading@10: Difference factor where higher values mean more deblocking (default: C<32>).
yading@10: 
yading@10: =item B<flatness>
yading@10: 
yading@10: Flatness threshold where lower values mean more deblocking (default: C<39>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<va/vadeblock[|difference[|flatness]]>
yading@10: 
yading@10: Accurate vertical deblocking filter
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<difference>
yading@10: 
yading@10: Difference factor where higher values mean more deblocking (default: C<32>).
yading@10: 
yading@10: =item B<flatness>
yading@10: 
yading@10: Flatness threshold where lower values mean more deblocking (default: C<39>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The horizontal and vertical deblocking filters share the difference and
yading@10: flatness values so you cannot set different horizontal and vertical
yading@10: thresholds.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<h1/x1hdeblock>
yading@10: 
yading@10: Experimental horizontal deblocking filter
yading@10: 
yading@10: 
yading@10: =item B<v1/x1vdeblock>
yading@10: 
yading@10: Experimental vertical deblocking filter
yading@10: 
yading@10: 
yading@10: =item B<dr/dering>
yading@10: 
yading@10: Deringing filter
yading@10: 
yading@10: 
yading@10: =item B<tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer>
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<threshold1>
yading@10: 
yading@10: larger -E<gt> stronger filtering
yading@10: 
yading@10: =item B<threshold2>
yading@10: 
yading@10: larger -E<gt> stronger filtering
yading@10: 
yading@10: =item B<threshold3>
yading@10: 
yading@10: larger -E<gt> stronger filtering
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<al/autolevels[:f/fullyrange], automatic brightness / contrast correction>
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<f/fullyrange>
yading@10: 
yading@10: Stretch luminance to C<0-255>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<lb/linblenddeint>
yading@10: 
yading@10: Linear blend deinterlacing filter that deinterlaces the given block by
yading@10: filtering all lines with a C<(1 2 1)> filter.
yading@10: 
yading@10: 
yading@10: =item B<li/linipoldeint>
yading@10: 
yading@10: Linear interpolating deinterlacing filter that deinterlaces the given block by
yading@10: linearly interpolating every second line.
yading@10: 
yading@10: 
yading@10: =item B<ci/cubicipoldeint>
yading@10: 
yading@10: Cubic interpolating deinterlacing filter deinterlaces the given block by
yading@10: cubically interpolating every second line.
yading@10: 
yading@10: 
yading@10: =item B<md/mediandeint>
yading@10: 
yading@10: Median deinterlacing filter that deinterlaces the given block by applying a
yading@10: median filter to every second line.
yading@10: 
yading@10: 
yading@10: =item B<fd/ffmpegdeint>
yading@10: 
yading@10: FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
yading@10: second line with a C<(-1 4 2 4 -1)> filter.
yading@10: 
yading@10: 
yading@10: =item B<l5/lowpass5>
yading@10: 
yading@10: Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
yading@10: block by filtering all lines with a C<(-1 2 6 2 -1)> filter.
yading@10: 
yading@10: 
yading@10: =item B<fq/forceQuant[|quantizer]>
yading@10: 
yading@10: Overrides the quantizer table from the input with the constant quantizer you
yading@10: specify.
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<quantizer>
yading@10: 
yading@10: Quantizer to use
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<de/default>
yading@10: 
yading@10: Default pp filter combination (C<hb|a,vb|a,dr|a>)
yading@10: 
yading@10: 
yading@10: =item B<fa/fast>
yading@10: 
yading@10: Fast pp filter combination (C<h1|a,v1|a,dr|a>)
yading@10: 
yading@10: 
yading@10: =item B<ac>
yading@10: 
yading@10: High quality pp filter combination (C<ha|a|128|7,va|a,dr|a>)
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply horizontal and vertical deblocking, deringing and automatic
yading@10: brightness/contrast:
yading@10: 	
yading@10: 	pp=hb/vb/dr/al
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply default filters without brightness/contrast correction:
yading@10: 	
yading@10: 	pp=de/-al
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply default filters and temporal denoiser:
yading@10: 	
yading@10: 	pp=default/tmpnoise|1|2|3
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply deblocking on luminance only, and switch vertical deblocking on or off
yading@10: automatically depending on available CPU time:
yading@10: 	
yading@10: 	pp=hb|y/vb|a
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 removelogo
yading@10: 
yading@10: 
yading@10: Suppress a TV station logo, using an image file to determine which
yading@10: pixels comprise the logo. It works by filling in the pixels that
yading@10: comprise the logo with neighboring pixels.
yading@10: 
yading@10: The filters accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<filename, f>
yading@10: 
yading@10: Set the filter bitmap file, which can be any image format supported by
yading@10: libavformat. The width and height of the image file must match those of the
yading@10: video stream being processed.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Pixels in the provided bitmap image with a value of zero are not
yading@10: considered part of the logo, non-zero pixels are considered part of
yading@10: the logo. If you use white (255) for the logo and black (0) for the
yading@10: rest, you will be safe. For making the filter bitmap, it is
yading@10: recommended to take a screen capture of a black frame with the logo
yading@10: visible, and then using a threshold filter followed by the erode
yading@10: filter once or twice.
yading@10: 
yading@10: If needed, little splotches can be fixed manually. Remember that if
yading@10: logo pixels are not covered, the filter quality will be much
yading@10: reduced. Marking too many pixels as part of the logo does not hurt as
yading@10: much, but it will increase the amount of blurring needed to cover over
yading@10: the image and will destroy more information than necessary, and extra
yading@10: pixels will slow things down on a large logo.
yading@10: 
yading@10: 
yading@10: =head2 scale
yading@10: 
yading@10: 
yading@10: Scale (resize) the input video, using the libswscale library.
yading@10: 
yading@10: The scale filter forces the output display aspect ratio to be the same
yading@10: of the input, by changing the output sample aspect ratio.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<width, w>
yading@10: 
yading@10: Output video width.
yading@10: default value is C<iw>. See below
yading@10: for the list of accepted constants.
yading@10: 
yading@10: 
yading@10: =item B<height, h>
yading@10: 
yading@10: Output video height.
yading@10: default value is C<ih>.
yading@10: See below for the list of accepted constants.
yading@10: 
yading@10: 
yading@10: =item B<interl>
yading@10: 
yading@10: Set the interlacing. It accepts the following values:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<1>
yading@10: 
yading@10: force interlaced aware scaling
yading@10: 
yading@10: 
yading@10: =item B<0>
yading@10: 
yading@10: do not apply interlaced scaling
yading@10: 
yading@10: 
yading@10: =item B<-1>
yading@10: 
yading@10: select interlaced aware scaling depending on whether the source frames
yading@10: are flagged as interlaced or not
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<0>.
yading@10: 
yading@10: 
yading@10: =item B<flags>
yading@10: 
yading@10: Set libswscale scaling flags. If not explictly specified the filter
yading@10: applies a bilinear scaling algorithm.
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Set the video size, the value must be a valid abbreviation or in the
yading@10: form I<width>xI<height>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The values of the I<w> and I<h> options are expressions
yading@10: containing the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<in_w, in_h>
yading@10: 
yading@10: the input width and height
yading@10: 
yading@10: 
yading@10: =item B<iw, ih>
yading@10: 
yading@10: same as I<in_w> and I<in_h>
yading@10: 
yading@10: 
yading@10: =item B<out_w, out_h>
yading@10: 
yading@10: the output (cropped) width and height
yading@10: 
yading@10: 
yading@10: =item B<ow, oh>
yading@10: 
yading@10: same as I<out_w> and I<out_h>
yading@10: 
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: same as I<iw> / I<ih>
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: input sample aspect ratio
yading@10: 
yading@10: 
yading@10: =item B<dar>
yading@10: 
yading@10: input display aspect ratio, it is the same as (I<iw> / I<ih>) * I<sar>
yading@10: 
yading@10: 
yading@10: =item B<hsub, vsub>
yading@10: 
yading@10: horizontal and vertical chroma subsample values. For example for the
yading@10: pixel format "yuv422p" I<hsub> is 2 and I<vsub> is 1.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If the input image format is different from the format requested by
yading@10: the next filter, the scale filter will convert the input to the
yading@10: requested format.
yading@10: 
yading@10: If the value for I<w> or I<h> is 0, the respective input
yading@10: size is used for the output.
yading@10: 
yading@10: If the value for I<w> or I<h> is -1, the scale filter will use, for the
yading@10: respective output size, a value that maintains the aspect ratio of the input
yading@10: image.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Scale the input video to a size of 200x100:
yading@10: 	
yading@10: 	scale=w=200:h=100
yading@10: 
yading@10: 
yading@10: This is equivalent to:
yading@10: 	
yading@10: 	scale=w=200:h=100
yading@10: 
yading@10: 
yading@10: or:
yading@10: 	
yading@10: 	scale=200x100
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify a size abbreviation for the output size:
yading@10: 	
yading@10: 	scale=qcif
yading@10: 
yading@10: 
yading@10: which can also be written as:
yading@10: 	
yading@10: 	scale=size=qcif
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Scale the input to 2x:
yading@10: 	
yading@10: 	scale=w=2*iw:h=2*ih
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: The above is the same as:
yading@10: 	
yading@10: 	scale=2*in_w:2*in_h
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Scale the input to 2x with forced interlaced scaling:
yading@10: 	
yading@10: 	scale=2*iw:2*ih:interl=1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Scale the input to half size:
yading@10: 	
yading@10: 	scale=w=iw/2:h=ih/2
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase the width, and set the height to the same size:
yading@10: 	
yading@10: 	scale=3/2*iw:ow
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Seek for Greek harmony:
yading@10: 	
yading@10: 	scale=iw:1/PHI*iw
yading@10: 	scale=ih*PHI:ih
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase the height, and set the width to 3/2 of the height:
yading@10: 	
yading@10: 	scale=w=3/2*oh:h=3/5*ih
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase the size, but make the size a multiple of the chroma
yading@10: subsample values:
yading@10: 	
yading@10: 	scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Increase the width to a maximum of 500 pixels, keep the same input
yading@10: aspect ratio:
yading@10: 	
yading@10: 	scale=w='min(500\, iw*3/2):h=-1'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 separatefields
yading@10: 
yading@10: 
yading@10: The C<separatefields> takes a frame-based video input and splits
yading@10: each frame into its components fields, producing a new half height clip
yading@10: with twice the frame rate and twice the frame count.
yading@10: 
yading@10: This filter use field-dominance information in frame to decide which
yading@10: of each pair of fields to place first in the output.
yading@10: If it gets it wrong use setfield filter before C<separatefields> filter.
yading@10: 
yading@10: 
yading@10: =head2 setdar, setsar
yading@10: 
yading@10: 
yading@10: The C<setdar> filter sets the Display Aspect Ratio for the filter
yading@10: output video.
yading@10: 
yading@10: This is done by changing the specified Sample (aka Pixel) Aspect
yading@10: Ratio, according to the following equation:
yading@10: 	
yading@10: 	<DAR> = <HORIZONTAL_RESOLUTION> / <VERTICAL_RESOLUTION> * <SAR>
yading@10: 
yading@10: 
yading@10: Keep in mind that the C<setdar> filter does not modify the pixel
yading@10: dimensions of the video frame. Also the display aspect ratio set by
yading@10: this filter may be changed by later filters in the filterchain,
yading@10: e.g. in case of scaling or if another "setdar" or a "setsar" filter is
yading@10: applied.
yading@10: 
yading@10: The C<setsar> filter sets the Sample (aka Pixel) Aspect Ratio for
yading@10: the filter output video.
yading@10: 
yading@10: Note that as a consequence of the application of this filter, the
yading@10: output display aspect ratio will change according to the equation
yading@10: above.
yading@10: 
yading@10: Keep in mind that the sample aspect ratio set by the C<setsar>
yading@10: filter may be changed by later filters in the filterchain, e.g. if
yading@10: another "setsar" or a "setdar" filter is applied.
yading@10: 
yading@10: The filters accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<r, ratio, dar (C<setdar> only), sar (C<setsar> only)>
yading@10: 
yading@10: Set the aspect ratio used by the filter.
yading@10: 
yading@10: The parameter can be a floating point number string, an expression, or
yading@10: a string of the form I<num>:I<den>, where I<num> and
yading@10: I<den> are the numerator and denominator of the aspect ratio. If
yading@10: the parameter is not specified, it is assumed the value "0".
yading@10: In case the form "I<num>:I<den>" is used, the C<:> character
yading@10: should be escaped.
yading@10: 
yading@10: 
yading@10: =item B<max>
yading@10: 
yading@10: Set the maximum integer value to use for expressing numerator and
yading@10: denominator when reducing the expressed aspect ratio to a rational.
yading@10: Default value is C<100>.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: To change the display aspect ratio to 16:9, specify one of the following:
yading@10: 	
yading@10: 	setdar=dar=1.77777
yading@10: 	setdar=dar=16/9
yading@10: 	setdar=dar=1.77777
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: To change the sample aspect ratio to 10:11, specify:
yading@10: 	
yading@10: 	setsar=sar=10/11
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: To set a display aspect ratio of 16:9, and specify a maximum integer value of
yading@10: 1000 in the aspect ratio reduction, use the command:
yading@10: 	
yading@10: 	setdar=ratio=16/9:max=1000
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 setfield
yading@10: 
yading@10: 
yading@10: Force field for the output video frame.
yading@10: 
yading@10: The C<setfield> filter marks the interlace type field for the
yading@10: output frames. It does not change the input frame, but only sets the
yading@10: corresponding property, which affects how the frame is treated by
yading@10: following filters (e.g. C<fieldorder> or C<yadif>).
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<auto>
yading@10: 
yading@10: Keep the same field property.
yading@10: 
yading@10: 
yading@10: =item B<bff>
yading@10: 
yading@10: Mark the frame as bottom-field-first.
yading@10: 
yading@10: 
yading@10: =item B<tff>
yading@10: 
yading@10: Mark the frame as top-field-first.
yading@10: 
yading@10: 
yading@10: =item B<prog>
yading@10: 
yading@10: Mark the frame as progressive.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 showinfo
yading@10: 
yading@10: 
yading@10: Show a line containing various information for each input video frame.
yading@10: The input video is not modified.
yading@10: 
yading@10: The shown line contains a sequence of key/value pairs of the form
yading@10: I<key>:I<value>.
yading@10: 
yading@10: A description of each shown parameter follows:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: sequential number of the input frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<pts>
yading@10: 
yading@10: Presentation TimeStamp of the input frame, expressed as a number of
yading@10: time base units. The time base unit depends on the filter input pad.
yading@10: 
yading@10: 
yading@10: =item B<pts_time>
yading@10: 
yading@10: Presentation TimeStamp of the input frame, expressed as a number of
yading@10: seconds
yading@10: 
yading@10: 
yading@10: =item B<pos>
yading@10: 
yading@10: position of the frame in the input stream, -1 if this information in
yading@10: unavailable and/or meaningless (for example in case of synthetic video)
yading@10: 
yading@10: 
yading@10: =item B<fmt>
yading@10: 
yading@10: pixel format name
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: sample aspect ratio of the input frame, expressed in the form
yading@10: I<num>/I<den>
yading@10: 
yading@10: 
yading@10: =item B<s>
yading@10: 
yading@10: size of the input frame, expressed in the form
yading@10: I<width>xI<height>
yading@10: 
yading@10: 
yading@10: =item B<i>
yading@10: 
yading@10: interlaced mode ("P" for "progressive", "T" for top field first, "B"
yading@10: for bottom field first)
yading@10: 
yading@10: 
yading@10: =item B<iskey>
yading@10: 
yading@10: 1 if the frame is a key frame, 0 otherwise
yading@10: 
yading@10: 
yading@10: =item B<type>
yading@10: 
yading@10: picture type of the input frame ("I" for an I-frame, "P" for a
yading@10: P-frame, "B" for a B-frame, "?" for unknown type).
yading@10: Check also the documentation of the C<AVPictureType> enum and of
yading@10: the C<av_get_picture_type_char> function defined in
yading@10: F<libavutil/avutil.h>.
yading@10: 
yading@10: 
yading@10: =item B<checksum>
yading@10: 
yading@10: Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame
yading@10: 
yading@10: 
yading@10: =item B<plane_checksum>
yading@10: 
yading@10: Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
yading@10: expressed in the form "[I<c0> I<c1> I<c2> I<c3>]"
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 smartblur
yading@10: 
yading@10: 
yading@10: Blur the input video without impacting the outlines.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<luma_radius, lr>
yading@10: 
yading@10: Set the luma radius. The option value must be a float number in
yading@10: the range [0.1,5.0] that specifies the variance of the gaussian filter
yading@10: used to blur the image (slower if larger). Default value is 1.0.
yading@10: 
yading@10: 
yading@10: =item B<luma_strength, ls>
yading@10: 
yading@10: Set the luma strength. The option value must be a float number
yading@10: in the range [-1.0,1.0] that configures the blurring. A value included
yading@10: in [0.0,1.0] will blur the image whereas a value included in
yading@10: [-1.0,0.0] will sharpen the image. Default value is 1.0.
yading@10: 
yading@10: 
yading@10: =item B<luma_threshold, lt>
yading@10: 
yading@10: Set the luma threshold used as a coefficient to determine
yading@10: whether a pixel should be blurred or not. The option value must be an
yading@10: integer in the range [-30,30]. A value of 0 will filter all the image,
yading@10: a value included in [0,30] will filter flat areas and a value included
yading@10: in [-30,0] will filter edges. Default value is 0.
yading@10: 
yading@10: 
yading@10: =item B<chroma_radius, cr>
yading@10: 
yading@10: Set the chroma radius. The option value must be a float number in
yading@10: the range [0.1,5.0] that specifies the variance of the gaussian filter
yading@10: used to blur the image (slower if larger). Default value is 1.0.
yading@10: 
yading@10: 
yading@10: =item B<chroma_strength, cs>
yading@10: 
yading@10: Set the chroma strength. The option value must be a float number
yading@10: in the range [-1.0,1.0] that configures the blurring. A value included
yading@10: in [0.0,1.0] will blur the image whereas a value included in
yading@10: [-1.0,0.0] will sharpen the image. Default value is 1.0.
yading@10: 
yading@10: 
yading@10: =item B<chroma_threshold, ct>
yading@10: 
yading@10: Set the chroma threshold used as a coefficient to determine
yading@10: whether a pixel should be blurred or not. The option value must be an
yading@10: integer in the range [-30,30]. A value of 0 will filter all the image,
yading@10: a value included in [0,30] will filter flat areas and a value included
yading@10: in [-30,0] will filter edges. Default value is 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If a chroma option is not explicitly set, the corresponding luma value
yading@10: is set.
yading@10: 
yading@10: 
yading@10: =head2 stereo3d
yading@10: 
yading@10: 
yading@10: Convert between different stereoscopic image formats.
yading@10: 
yading@10: The filters accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<in>
yading@10: 
yading@10: Set stereoscopic image format of input.
yading@10: 
yading@10: Available values for input image formats are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<sbsl>
yading@10: 
yading@10: side by side parallel (left eye left, right eye right)
yading@10: 
yading@10: 
yading@10: =item B<sbsr>
yading@10: 
yading@10: side by side crosseye (right eye left, left eye right)
yading@10: 
yading@10: 
yading@10: =item B<sbs2l>
yading@10: 
yading@10: side by side parallel with half width resolution
yading@10: (left eye left, right eye right)
yading@10: 
yading@10: 
yading@10: =item B<sbs2r>
yading@10: 
yading@10: side by side crosseye with half width resolution
yading@10: (right eye left, left eye right)
yading@10: 
yading@10: 
yading@10: =item B<abl>
yading@10: 
yading@10: above-below (left eye above, right eye below)
yading@10: 
yading@10: 
yading@10: =item B<abr>
yading@10: 
yading@10: above-below (right eye above, left eye below)
yading@10: 
yading@10: 
yading@10: =item B<ab2l>
yading@10: 
yading@10: above-below with half height resolution
yading@10: (left eye above, right eye below)
yading@10: 
yading@10: 
yading@10: =item B<ab2r>
yading@10: 
yading@10: above-below with half height resolution
yading@10: (right eye above, left eye below)
yading@10: 
yading@10: Default value is B<sbsl>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<out>
yading@10: 
yading@10: Set stereoscopic image format of output.
yading@10: 
yading@10: Available values for output image formats are all the input formats as well as:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<arbg>
yading@10: 
yading@10: anaglyph red/blue gray
yading@10: (red filter on left eye, blue filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<argg>
yading@10: 
yading@10: anaglyph red/green gray
yading@10: (red filter on left eye, green filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<arcg>
yading@10: 
yading@10: anaglyph red/cyan gray
yading@10: (red filter on left eye, cyan filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<arch>
yading@10: 
yading@10: anaglyph red/cyan half colored
yading@10: (red filter on left eye, cyan filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<arcc>
yading@10: 
yading@10: anaglyph red/cyan color
yading@10: (red filter on left eye, cyan filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<arcd>
yading@10: 
yading@10: anaglyph red/cyan color optimized with the least squares projection of dubois
yading@10: (red filter on left eye, cyan filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<agmg>
yading@10: 
yading@10: anaglyph green/magenta gray
yading@10: (green filter on left eye, magenta filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<agmh>
yading@10: 
yading@10: anaglyph green/magenta half colored
yading@10: (green filter on left eye, magenta filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<agmc>
yading@10: 
yading@10: anaglyph green/magenta colored
yading@10: (green filter on left eye, magenta filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<agmd>
yading@10: 
yading@10: anaglyph green/magenta color optimized with the least squares projection of dubois
yading@10: (green filter on left eye, magenta filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<aybg>
yading@10: 
yading@10: anaglyph yellow/blue gray
yading@10: (yellow filter on left eye, blue filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<aybh>
yading@10: 
yading@10: anaglyph yellow/blue half colored
yading@10: (yellow filter on left eye, blue filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<aybc>
yading@10: 
yading@10: anaglyph yellow/blue colored
yading@10: (yellow filter on left eye, blue filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<aybd>
yading@10: 
yading@10: anaglyph yellow/blue color optimized with the least squares projection of dubois
yading@10: (yellow filter on left eye, blue filter on right eye)
yading@10: 
yading@10: 
yading@10: =item B<irl>
yading@10: 
yading@10: interleaved rows (left eye has top row, right eye starts on next row)
yading@10: 
yading@10: 
yading@10: =item B<irr>
yading@10: 
yading@10: interleaved rows (right eye has top row, left eye starts on next row)
yading@10: 
yading@10: 
yading@10: =item B<ml>
yading@10: 
yading@10: mono output (left eye only)
yading@10: 
yading@10: 
yading@10: =item B<mr>
yading@10: 
yading@10: mono output (right eye only)
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<arcd>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 subtitles
yading@10: 
yading@10: 
yading@10: Draw subtitles on top of input video using the libass library.
yading@10: 
yading@10: To enable compilation of this filter you need to configure FFmpeg with
yading@10: C<--enable-libass>. This filter also requires a build with libavcodec and
yading@10: libavformat to convert the passed subtitles file to ASS (Advanced Substation
yading@10: Alpha) subtitles format.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<filename, f>
yading@10: 
yading@10: Set the filename of the subtitle file to read. It must be specified.
yading@10: 
yading@10: 
yading@10: =item B<original_size>
yading@10: 
yading@10: Specify the size of the original video, the video for which the ASS file
yading@10: was composed. Due to a misdesign in ASS aspect ratio arithmetic, this is
yading@10: necessary to correctly scale the fonts if the aspect ratio has been changed.
yading@10: 
yading@10: 
yading@10: =item B<charenc>
yading@10: 
yading@10: Set subtitles input character encoding. C<subtitles> filter only. Only
yading@10: useful if not UTF-8.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If the first key is not specified, it is assumed that the first value
yading@10: specifies the B<filename>.
yading@10: 
yading@10: For example, to render the file F<sub.srt> on top of the input
yading@10: video, use the command:
yading@10: 	
yading@10: 	subtitles=sub.srt
yading@10: 
yading@10: 
yading@10: which is equivalent to:
yading@10: 	
yading@10: 	subtitles=filename=sub.srt
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 super2xsai
yading@10: 
yading@10: 
yading@10: Scale the input by 2x and smooth using the Super2xSaI (Scale and
yading@10: Interpolate) pixel art scaling algorithm.
yading@10: 
yading@10: Useful for enlarging pixel art images without reducing sharpness.
yading@10: 
yading@10: 
yading@10: =head2 swapuv
yading@10: 
yading@10: Swap U & V plane.
yading@10: 
yading@10: 
yading@10: =head2 telecine
yading@10: 
yading@10: 
yading@10: Apply telecine process to the video.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<first_field>
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<top, t>
yading@10: 
yading@10: top field first
yading@10: 
yading@10: =item B<bottom, b>
yading@10: 
yading@10: bottom field first
yading@10: The default value is C<top>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<pattern>
yading@10: 
yading@10: A string of numbers representing the pulldown pattern you wish to apply.
yading@10: The default value is C<23>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 	
yading@10: 	Some typical patterns:
yading@10: 	
yading@10: 	NTSC output (30i):
yading@10: 	27.5p: 32222
yading@10: 	24p: 23 (classic)
yading@10: 	24p: 2332 (preferred)
yading@10: 	20p: 33
yading@10: 	18p: 334
yading@10: 	16p: 3444
yading@10: 	
yading@10: 	PAL output (25i):
yading@10: 	27.5p: 12222
yading@10: 	24p: 222222222223 ("Euro pulldown")
yading@10: 	16.67p: 33
yading@10: 	16p: 33333334
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 thumbnail
yading@10: 
yading@10: Select the most representative frame in a given sequence of consecutive frames.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: Set the frames batch size to analyze; in a set of I<n> frames, the filter
yading@10: will pick one of them, and then handle the next batch of I<n> frames until
yading@10: the end. Default is C<100>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Since the filter keeps track of the whole frames sequence, a bigger I<n>
yading@10: value will result in a higher memory usage, so a high value is not recommended.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Extract one picture each 50 frames:
yading@10: 	
yading@10: 	thumbnail=50
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Complete example of a thumbnail creation with B<ffmpeg>:
yading@10: 	
yading@10: 	ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 tile
yading@10: 
yading@10: 
yading@10: Tile several successive frames together.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<layout>
yading@10: 
yading@10: Set the grid size (i.e. the number of lines and columns) in the form
yading@10: "I<w>xI<h>".
yading@10: 
yading@10: 
yading@10: =item B<nb_frames>
yading@10: 
yading@10: Set the maximum number of frames to render in the given area. It must be less
yading@10: than or equal to I<w>xI<h>. The default value is C<0>, meaning all
yading@10: the area will be used.
yading@10: 
yading@10: 
yading@10: =item B<margin>
yading@10: 
yading@10: Set the outer border margin in pixels.
yading@10: 
yading@10: 
yading@10: =item B<padding>
yading@10: 
yading@10: Set the inner border thickness (i.e. the number of pixels between frames). For
yading@10: more advanced padding options (such as having different values for the edges),
yading@10: refer to the pad video filter.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Produce 8x8 PNG tiles of all keyframes (B<-skip_frame nokey>) in a movie:
yading@10: 	
yading@10: 	ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
yading@10: 
yading@10: The B<-vsync 0> is necessary to prevent B<ffmpeg> from
yading@10: duplicating each output frame to accomodate the originally detected frame
yading@10: rate.
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Display C<5> pictures in an area of C<3x2> frames,
yading@10: with C<7> pixels between them, and C<2> pixels of initial margin, using
yading@10: mixed flat and named options:
yading@10: 	
yading@10: 	tile=3x2:nb_frames=5:padding=7:margin=2
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 tinterlace
yading@10: 
yading@10: 
yading@10: Perform various types of temporal field interlacing.
yading@10: 
yading@10: Frames are counted starting from 1, so the first input frame is
yading@10: considered odd.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Specify the mode of the interlacing. This option can also be specified
yading@10: as a value alone. See below for a list of values for this option.
yading@10: 
yading@10: Available values are:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<merge, 0>
yading@10: 
yading@10: Move odd frames into the upper field, even into the lower field,
yading@10: generating a double height frame at half frame rate.
yading@10: 
yading@10: 
yading@10: =item B<drop_odd, 1>
yading@10: 
yading@10: Only output even frames, odd frames are dropped, generating a frame with
yading@10: unchanged height at half frame rate.
yading@10: 
yading@10: 
yading@10: =item B<drop_even, 2>
yading@10: 
yading@10: Only output odd frames, even frames are dropped, generating a frame with
yading@10: unchanged height at half frame rate.
yading@10: 
yading@10: 
yading@10: =item B<pad, 3>
yading@10: 
yading@10: Expand each frame to full height, but pad alternate lines with black,
yading@10: generating a frame with double height at the same input frame rate.
yading@10: 
yading@10: 
yading@10: =item B<interleave_top, 4>
yading@10: 
yading@10: Interleave the upper field from odd frames with the lower field from
yading@10: even frames, generating a frame with unchanged height at half frame rate.
yading@10: 
yading@10: 
yading@10: =item B<interleave_bottom, 5>
yading@10: 
yading@10: Interleave the lower field from odd frames with the upper field from
yading@10: even frames, generating a frame with unchanged height at half frame rate.
yading@10: 
yading@10: 
yading@10: =item B<interlacex2, 6>
yading@10: 
yading@10: Double frame rate with unchanged height. Frames are inserted each
yading@10: containing the second temporal field from the previous input frame and
yading@10: the first temporal field from the next input frame. This mode relies on
yading@10: the top_field_first flag. Useful for interlaced video displays with no
yading@10: field synchronisation.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Numeric values are deprecated but are accepted for backward
yading@10: compatibility reasons.
yading@10: 
yading@10: Default mode is C<merge>.
yading@10: 
yading@10: 
yading@10: =item B<flags>
yading@10: 
yading@10: Specify flags influencing the filter process.
yading@10: 
yading@10: Available value for I<flags> is:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<low_pass_filter, vlfp>
yading@10: 
yading@10: Enable vertical low-pass filtering in the filter.
yading@10: Vertical low-pass filtering is required when creating an interlaced
yading@10: destination from a progressive source which contains high-frequency
yading@10: vertical detail. Filtering will reduce interlace 'twitter' and Moire
yading@10: patterning.
yading@10: 
yading@10: Vertical low-pass filtering can only be enabled for B<mode>
yading@10: I<interleave_top> and I<interleave_bottom>.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 transpose
yading@10: 
yading@10: 
yading@10: Transpose rows with columns in the input video and optionally flip it.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<dir>
yading@10: 
yading@10: The direction of the transpose.
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<0, 4, cclock_flip>
yading@10: 
yading@10: Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
yading@10: 	
yading@10: 	L.R     L.l
yading@10: 	. . ->  . .
yading@10: 	l.r     R.r
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<1, 5, clock>
yading@10: 
yading@10: Rotate by 90 degrees clockwise, that is:
yading@10: 	
yading@10: 	L.R     l.L
yading@10: 	. . ->  . .
yading@10: 	l.r     r.R
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<2, 6, cclock>
yading@10: 
yading@10: Rotate by 90 degrees counterclockwise, that is:
yading@10: 	
yading@10: 	L.R     R.r
yading@10: 	. . ->  . .
yading@10: 	l.r     L.l
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<3, 7, clock_flip>
yading@10: 
yading@10: Rotate by 90 degrees clockwise and vertically flip, that is:
yading@10: 	
yading@10: 	L.R     r.R
yading@10: 	. . ->  . .
yading@10: 	l.r     l.L
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For values between 4-7, the transposition is only done if the input
yading@10: video geometry is portrait and not landscape. These values are
yading@10: deprecated, the C<passthrough> option should be used instead.
yading@10: 
yading@10: 
yading@10: =item B<passthrough>
yading@10: 
yading@10: Do not apply the transposition if the input geometry matches the one
yading@10: specified by the specified value. It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: Always apply transposition.
yading@10: 
yading@10: =item B<portrait>
yading@10: 
yading@10: Preserve portrait geometry (when I<height> E<gt>= I<width>).
yading@10: 
yading@10: =item B<landscape>
yading@10: 
yading@10: Preserve landscape geometry (when I<width> E<gt>= I<height>).
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<none>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example to rotate by 90 degrees clockwise and preserve portrait
yading@10: layout:
yading@10: 	
yading@10: 	transpose=dir=1:passthrough=portrait
yading@10: 
yading@10: 
yading@10: The command above can also be specified as:
yading@10: 	
yading@10: 	transpose=1:portrait
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 unsharp
yading@10: 
yading@10: 
yading@10: Sharpen or blur the input video.
yading@10: 
yading@10: It accepts the following parameters:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<luma_msize_x, lx>
yading@10: 
yading@10: 
yading@10: =item B<chroma_msize_x, cx>
yading@10: 
yading@10: Set the luma/chroma matrix horizontal size. It must be an odd integer
yading@10: between 3 and 63, default value is 5.
yading@10: 
yading@10: 
yading@10: =item B<luma_msize_y, ly>
yading@10: 
yading@10: 
yading@10: =item B<chroma_msize_y, cy>
yading@10: 
yading@10: Set the luma/chroma matrix vertical size. It must be an odd integer
yading@10: between 3 and 63, default value is 5.
yading@10: 
yading@10: 
yading@10: =item B<luma_amount, la>
yading@10: 
yading@10: 
yading@10: =item B<chroma_amount, ca>
yading@10: 
yading@10: Set the luma/chroma effect strength. It can be a float number,
yading@10: reasonable values lay between -1.5 and 1.5.
yading@10: 
yading@10: Negative values will blur the input video, while positive values will
yading@10: sharpen it, a value of zero will disable the effect.
yading@10: 
yading@10: Default value is 1.0 for B<luma_amount>, 0.0 for
yading@10: B<chroma_amount>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: All parameters are optional and default to the
yading@10: equivalent of the string '5:5:1.0:5:5:0.0'.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply strong luma sharpen effect:
yading@10: 	
yading@10: 	unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply strong blur of both luma and chroma parameters:
yading@10: 	
yading@10: 	unsharp=7:7:-2:7:7:-2
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 vflip
yading@10: 
yading@10: 
yading@10: Flip the input video vertically.
yading@10: 
yading@10: 	
yading@10: 	ffmpeg -i in.avi -vf "vflip" out.avi
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 yadif
yading@10: 
yading@10: 
yading@10: Deinterlace the input video ("yadif" means "yet another deinterlacing
yading@10: filter").
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: The interlacing mode to adopt, accepts one of the following values:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<0, send_frame>
yading@10: 
yading@10: output 1 frame for each frame
yading@10: 
yading@10: =item B<1, send_field>
yading@10: 
yading@10: output 1 frame for each field
yading@10: 
yading@10: =item B<2, send_frame_nospatial>
yading@10: 
yading@10: like C<send_frame> but skip spatial interlacing check
yading@10: 
yading@10: =item B<3, send_field_nospatial>
yading@10: 
yading@10: like C<send_field> but skip spatial interlacing check
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<send_frame>.
yading@10: 
yading@10: 
yading@10: =item B<parity>
yading@10: 
yading@10: The picture field parity assumed for the input interlaced video, accepts one of
yading@10: the following values:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<0, tff>
yading@10: 
yading@10: assume top field first
yading@10: 
yading@10: =item B<1, bff>
yading@10: 
yading@10: assume bottom field first
yading@10: 
yading@10: =item B<-1, auto>
yading@10: 
yading@10: enable automatic detection
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<auto>.
yading@10: If interlacing is unknown or decoder does not export this information,
yading@10: top field first will be assumed.
yading@10: 
yading@10: 
yading@10: =item B<deint>
yading@10: 
yading@10: Specify which frames to deinterlace. Accept one of the following
yading@10: values:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<0, all>
yading@10: 
yading@10: deinterlace all frames
yading@10: 
yading@10: =item B<1, interlaced>
yading@10: 
yading@10: only deinterlace frames marked as interlaced
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<all>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 VIDEO SOURCES
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available video sources.
yading@10: 
yading@10: 
yading@10: =head2 buffer
yading@10: 
yading@10: 
yading@10: Buffer video frames, and make them available to the filter chain.
yading@10: 
yading@10: This source is mainly intended for a programmatic use, in particular
yading@10: through the interface defined in F<libavfilter/vsrc_buffer.h>.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<video_size>
yading@10: 
yading@10: Specify the size (width and height) of the buffered video frames.
yading@10: 
yading@10: 
yading@10: =item B<width>
yading@10: 
yading@10: Input video width.
yading@10: 
yading@10: 
yading@10: =item B<height>
yading@10: 
yading@10: Input video height.
yading@10: 
yading@10: 
yading@10: =item B<pix_fmt>
yading@10: 
yading@10: A string representing the pixel format of the buffered video frames.
yading@10: It may be a number corresponding to a pixel format, or a pixel format
yading@10: name.
yading@10: 
yading@10: 
yading@10: =item B<time_base>
yading@10: 
yading@10: Specify the timebase assumed by the timestamps of the buffered frames.
yading@10: 
yading@10: 
yading@10: =item B<frame_rate>
yading@10: 
yading@10: Specify the frame rate expected for the video stream.
yading@10: 
yading@10: 
yading@10: =item B<pixel_aspect, sar>
yading@10: 
yading@10: Specify the sample aspect ratio assumed by the video frames.
yading@10: 
yading@10: 
yading@10: =item B<sws_param>
yading@10: 
yading@10: Specify the optional parameters to be used for the scale filter which
yading@10: is automatically inserted when an input change is detected in the
yading@10: input size or format.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example:
yading@10: 	
yading@10: 	buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
yading@10: 
yading@10: 
yading@10: will instruct the source to accept video frames with size 320x240 and
yading@10: with format "yuv410p", assuming 1/24 as the timestamps timebase and
yading@10: square pixels (1:1 sample aspect ratio).
yading@10: Since the pixel format with name "yuv410p" corresponds to the number 6
yading@10: (check the enum AVPixelFormat definition in F<libavutil/pixfmt.h>),
yading@10: this example corresponds to:
yading@10: 	
yading@10: 	buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
yading@10: 
yading@10: 
yading@10: Alternatively, the options can be specified as a flat string, but this
yading@10: syntax is deprecated:
yading@10: 
yading@10: I<width>:I<height>:I<pix_fmt>:I<time_base.num>:I<time_base.den>:I<pixel_aspect.num>:I<pixel_aspect.den>[:I<sws_param>]
yading@10: 
yading@10: 
yading@10: =head2 cellauto
yading@10: 
yading@10: 
yading@10: Create a pattern generated by an elementary cellular automaton.
yading@10: 
yading@10: The initial state of the cellular automaton can be defined through the
yading@10: B<filename>, and B<pattern> options. If such options are
yading@10: not specified an initial state is created randomly.
yading@10: 
yading@10: At each new frame a new row in the video is filled with the result of
yading@10: the cellular automaton next generation. The behavior when the whole
yading@10: frame is filled is defined by the B<scroll> option.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<filename, f>
yading@10: 
yading@10: Read the initial cellular automaton state, i.e. the starting row, from
yading@10: the specified file.
yading@10: In the file, each non-whitespace character is considered an alive
yading@10: cell, a newline will terminate the row, and further characters in the
yading@10: file will be ignored.
yading@10: 
yading@10: 
yading@10: =item B<pattern, p>
yading@10: 
yading@10: Read the initial cellular automaton state, i.e. the starting row, from
yading@10: the specified string.
yading@10: 
yading@10: Each non-whitespace character in the string is considered an alive
yading@10: cell, a newline will terminate the row, and further characters in the
yading@10: string will be ignored.
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Set the video rate, that is the number of frames generated per second.
yading@10: Default is 25.
yading@10: 
yading@10: 
yading@10: =item B<random_fill_ratio, ratio>
yading@10: 
yading@10: Set the random fill ratio for the initial cellular automaton row. It
yading@10: is a floating point number value ranging from 0 to 1, defaults to
yading@10: 1/PHI.
yading@10: 
yading@10: This option is ignored when a file or a pattern is specified.
yading@10: 
yading@10: 
yading@10: =item B<random_seed, seed>
yading@10: 
yading@10: Set the seed for filling randomly the initial row, must be an integer
yading@10: included between 0 and UINT32_MAX. If not specified, or if explicitly
yading@10: set to -1, the filter will try to use a good random seed on a best
yading@10: effort basis.
yading@10: 
yading@10: 
yading@10: =item B<rule>
yading@10: 
yading@10: Set the cellular automaton rule, it is a number ranging from 0 to 255.
yading@10: Default value is 110.
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Set the size of the output video.
yading@10: 
yading@10: If B<filename> or B<pattern> is specified, the size is set
yading@10: by default to the width of the specified initial state row, and the
yading@10: height is set to I<width> * PHI.
yading@10: 
yading@10: If B<size> is set, it must contain the width of the specified
yading@10: pattern string, and the specified pattern will be centered in the
yading@10: larger row.
yading@10: 
yading@10: If a filename or a pattern string is not specified, the size value
yading@10: defaults to "320x518" (used for a randomly generated initial state).
yading@10: 
yading@10: 
yading@10: =item B<scroll>
yading@10: 
yading@10: If set to 1, scroll the output upward when all the rows in the output
yading@10: have been already filled. If set to 0, the new generated row will be
yading@10: written over the top row just after the bottom row is filled.
yading@10: Defaults to 1.
yading@10: 
yading@10: 
yading@10: =item B<start_full, full>
yading@10: 
yading@10: If set to 1, completely fill the output with generated rows before
yading@10: outputting the first frame.
yading@10: This is the default behavior, for disabling set the value to 0.
yading@10: 
yading@10: 
yading@10: =item B<stitch>
yading@10: 
yading@10: If set to 1, stitch the left and right row edges together.
yading@10: This is the default behavior, for disabling set the value to 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read the initial state from F<pattern>, and specify an output of
yading@10: size 200x400.
yading@10: 	
yading@10: 	cellauto=f=pattern:s=200x400
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a random initial row with a width of 200 cells, with a fill
yading@10: ratio of 2/3:
yading@10: 	
yading@10: 	cellauto=ratio=2/3:s=200x200
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create a pattern generated by rule 18 starting by a single alive cell
yading@10: centered on an initial row with width 100:
yading@10: 	
yading@10: 	cellauto=p=@s=100x400:full=0:rule=18
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify a more elaborated initial pattern:
yading@10: 	
yading@10: 	cellauto=p='@@ @ @@':s=100x400:full=0:rule=18
yading@10: 
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 mandelbrot
yading@10: 
yading@10: 
yading@10: Generate a Mandelbrot set fractal, and progressively zoom towards the
yading@10: point specified with I<start_x> and I<start_y>.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<end_pts>
yading@10: 
yading@10: Set the terminal pts value. Default value is 400.
yading@10: 
yading@10: 
yading@10: =item B<end_scale>
yading@10: 
yading@10: Set the terminal scale value.
yading@10: Must be a floating point value. Default value is 0.3.
yading@10: 
yading@10: 
yading@10: =item B<inner>
yading@10: 
yading@10: Set the inner coloring mode, that is the algorithm used to draw the
yading@10: Mandelbrot fractal internal region.
yading@10: 
yading@10: It shall assume one of the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<black>
yading@10: 
yading@10: Set black mode.
yading@10: 
yading@10: =item B<convergence>
yading@10: 
yading@10: Show time until convergence.
yading@10: 
yading@10: =item B<mincol>
yading@10: 
yading@10: Set color based on point closest to the origin of the iterations.
yading@10: 
yading@10: =item B<period>
yading@10: 
yading@10: Set period mode.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is I<mincol>.
yading@10: 
yading@10: 
yading@10: =item B<bailout>
yading@10: 
yading@10: Set the bailout value. Default value is 10.0.
yading@10: 
yading@10: 
yading@10: =item B<maxiter>
yading@10: 
yading@10: Set the maximum of iterations performed by the rendering
yading@10: algorithm. Default value is 7189.
yading@10: 
yading@10: 
yading@10: =item B<outer>
yading@10: 
yading@10: Set outer coloring mode.
yading@10: It shall assume one of following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<iteration_count>
yading@10: 
yading@10: Set iteration cound mode.
yading@10: 
yading@10: =item B<normalized_iteration_count>
yading@10: 
yading@10: set normalized iteration count mode.
yading@10: 
yading@10: =back
yading@10: 
yading@10: Default value is I<normalized_iteration_count>.
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Set frame rate, expressed as number of frames per second. Default
yading@10: value is "25".
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Set frame size. Default value is "640x480".
yading@10: 
yading@10: 
yading@10: =item B<start_scale>
yading@10: 
yading@10: Set the initial scale value. Default value is 3.0.
yading@10: 
yading@10: 
yading@10: =item B<start_x>
yading@10: 
yading@10: Set the initial x position. Must be a floating point value between
yading@10: -100 and 100. Default value is -0.743643887037158704752191506114774.
yading@10: 
yading@10: 
yading@10: =item B<start_y>
yading@10: 
yading@10: Set the initial y position. Must be a floating point value between
yading@10: -100 and 100. Default value is -0.131825904205311970493132056385139.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 mptestsrc
yading@10: 
yading@10: 
yading@10: Generate various test patterns, as generated by the MPlayer test filter.
yading@10: 
yading@10: The size of the generated video is fixed, and is 256x256.
yading@10: This source is useful in particular for testing encoding features.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Specify the frame rate of the sourced video, as the number of frames
yading@10: generated per second. It has to be a string in the format
yading@10: I<frame_rate_num>/I<frame_rate_den>, an integer number, a float
yading@10: number or a valid video frame rate abbreviation. The default value is
yading@10: "25".
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Set the video duration of the sourced video. The accepted syntax is:
yading@10: 	
yading@10: 	[-]HH:MM:SS[.m...]
yading@10: 	[-]S+[.m...]
yading@10: 
yading@10: See also the function C<av_parse_time()>.
yading@10: 
yading@10: If not specified, or the expressed duration is negative, the video is
yading@10: supposed to be generated forever.
yading@10: 
yading@10: 
yading@10: =item B<test, t>
yading@10: 
yading@10: 
yading@10: Set the number or the name of the test to perform. Supported tests are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<dc_luma>
yading@10: 
yading@10: 
yading@10: =item B<dc_chroma>
yading@10: 
yading@10: 
yading@10: =item B<freq_luma>
yading@10: 
yading@10: 
yading@10: =item B<freq_chroma>
yading@10: 
yading@10: 
yading@10: =item B<amp_luma>
yading@10: 
yading@10: 
yading@10: =item B<amp_chroma>
yading@10: 
yading@10: 
yading@10: =item B<cbp>
yading@10: 
yading@10: 
yading@10: =item B<mv>
yading@10: 
yading@10: 
yading@10: =item B<ring1>
yading@10: 
yading@10: 
yading@10: =item B<ring2>
yading@10: 
yading@10: 
yading@10: =item B<all>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is "all", which will cycle through the list of all tests.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example the following:
yading@10: 	
yading@10: 	testsrc=t=dc_luma
yading@10: 
yading@10: 
yading@10: will generate a "dc_luma" test pattern.
yading@10: 
yading@10: 
yading@10: =head2 frei0r_src
yading@10: 
yading@10: 
yading@10: Provide a frei0r source.
yading@10: 
yading@10: To enable compilation of this filter you need to install the frei0r
yading@10: header and configure FFmpeg with C<--enable-frei0r>.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<size>
yading@10: 
yading@10: The size of the video to generate, may be a string of the form
yading@10: I<width>xI<height> or a frame size abbreviation.
yading@10: 
yading@10: 
yading@10: =item B<framerate>
yading@10: 
yading@10: Framerate of the generated video, may be a string of the form
yading@10: I<num>/I<den> or a frame rate abbreviation.
yading@10: 
yading@10: 
yading@10: =item B<filter_name>
yading@10: 
yading@10: The name to the frei0r source to load. For more information regarding frei0r and
yading@10: how to set the parameters read the section frei0r in the description of
yading@10: the video filters.
yading@10: 
yading@10: 
yading@10: =item B<filter_params>
yading@10: 
yading@10: A '|'-separated list of parameters to pass to the frei0r source.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example, to generate a frei0r partik0l source with size 200x200
yading@10: and frame rate 10 which is overlayed on the overlay filter main input:
yading@10: 	
yading@10: 	frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 life
yading@10: 
yading@10: 
yading@10: Generate a life pattern.
yading@10: 
yading@10: This source is based on a generalization of John Conway's life game.
yading@10: 
yading@10: The sourced input represents a life grid, each pixel represents a cell
yading@10: which can be in one of two possible states, alive or dead. Every cell
yading@10: interacts with its eight neighbours, which are the cells that are
yading@10: horizontally, vertically, or diagonally adjacent.
yading@10: 
yading@10: At each interaction the grid evolves according to the adopted rule,
yading@10: which specifies the number of neighbor alive cells which will make a
yading@10: cell stay alive or born. The B<rule> option allows to specify
yading@10: the rule to adopt.
yading@10: 
yading@10: This source accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<filename, f>
yading@10: 
yading@10: Set the file from which to read the initial grid state. In the file,
yading@10: each non-whitespace character is considered an alive cell, and newline
yading@10: is used to delimit the end of each row.
yading@10: 
yading@10: If this option is not specified, the initial grid is generated
yading@10: randomly.
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Set the video rate, that is the number of frames generated per second.
yading@10: Default is 25.
yading@10: 
yading@10: 
yading@10: =item B<random_fill_ratio, ratio>
yading@10: 
yading@10: Set the random fill ratio for the initial random grid. It is a
yading@10: floating point number value ranging from 0 to 1, defaults to 1/PHI.
yading@10: It is ignored when a file is specified.
yading@10: 
yading@10: 
yading@10: =item B<random_seed, seed>
yading@10: 
yading@10: Set the seed for filling the initial random grid, must be an integer
yading@10: included between 0 and UINT32_MAX. If not specified, or if explicitly
yading@10: set to -1, the filter will try to use a good random seed on a best
yading@10: effort basis.
yading@10: 
yading@10: 
yading@10: =item B<rule>
yading@10: 
yading@10: Set the life rule.
yading@10: 
yading@10: A rule can be specified with a code of the kind "SI<NS>/BI<NB>",
yading@10: where I<NS> and I<NB> are sequences of numbers in the range 0-8,
yading@10: I<NS> specifies the number of alive neighbor cells which make a
yading@10: live cell stay alive, and I<NB> the number of alive neighbor cells
yading@10: which make a dead cell to become alive (i.e. to "born").
yading@10: "s" and "b" can be used in place of "S" and "B", respectively.
yading@10: 
yading@10: Alternatively a rule can be specified by an 18-bits integer. The 9
yading@10: high order bits are used to encode the next cell state if it is alive
yading@10: for each number of neighbor alive cells, the low order bits specify
yading@10: the rule for "borning" new cells. Higher order bits encode for an
yading@10: higher number of neighbor cells.
yading@10: For example the number 6153 = C<(12E<lt>E<lt>9)+9> specifies a stay alive
yading@10: rule of 12 and a born rule of 9, which corresponds to "S23/B03".
yading@10: 
yading@10: Default value is "S23/B3", which is the original Conway's game of life
yading@10: rule, and will keep a cell alive if it has 2 or 3 neighbor alive
yading@10: cells, and will born a new cell if there are three alive cells around
yading@10: a dead cell.
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Set the size of the output video.
yading@10: 
yading@10: If B<filename> is specified, the size is set by default to the
yading@10: same size of the input file. If B<size> is set, it must contain
yading@10: the size specified in the input file, and the initial grid defined in
yading@10: that file is centered in the larger resulting area.
yading@10: 
yading@10: If a filename is not specified, the size value defaults to "320x240"
yading@10: (used for a randomly generated initial grid).
yading@10: 
yading@10: 
yading@10: =item B<stitch>
yading@10: 
yading@10: If set to 1, stitch the left and right grid edges together, and the
yading@10: top and bottom edges also. Defaults to 1.
yading@10: 
yading@10: 
yading@10: =item B<mold>
yading@10: 
yading@10: Set cell mold speed. If set, a dead cell will go from B<death_color> to
yading@10: B<mold_color> with a step of B<mold>. B<mold> can have a
yading@10: value from 0 to 255.
yading@10: 
yading@10: 
yading@10: =item B<life_color>
yading@10: 
yading@10: Set the color of living (or new born) cells.
yading@10: 
yading@10: 
yading@10: =item B<death_color>
yading@10: 
yading@10: Set the color of dead cells. If B<mold> is set, this is the first color
yading@10: used to represent a dead cell.
yading@10: 
yading@10: 
yading@10: =item B<mold_color>
yading@10: 
yading@10: Set mold color, for definitely dead and moldy cells.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read a grid from F<pattern>, and center it on a grid of size
yading@10: 300x300 pixels:
yading@10: 	
yading@10: 	life=f=pattern:s=300x300
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate a random grid of size 200x200, with a fill ratio of 2/3:
yading@10: 	
yading@10: 	life=ratio=2/3:s=200x200
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify a custom rule for evolving a randomly generated grid:
yading@10: 	
yading@10: 	life=rule=S14/B34
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Full example with slow death effect (mold) using B<ffplay>:
yading@10: 	
yading@10: 	ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 color, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc
yading@10: 
yading@10: 
yading@10: The C<color> source provides an uniformly colored input.
yading@10: 
yading@10: The C<nullsrc> source returns unprocessed video frames. It is
yading@10: mainly useful to be employed in analysis / debugging tools, or as the
yading@10: source for filters which ignore the input data.
yading@10: 
yading@10: The C<rgbtestsrc> source generates an RGB test pattern useful for
yading@10: detecting RGB vs BGR issues. You should see a red, green and blue
yading@10: stripe from top to bottom.
yading@10: 
yading@10: The C<smptebars> source generates a color bars pattern, based on
yading@10: the SMPTE Engineering Guideline EG 1-1990.
yading@10: 
yading@10: The C<smptehdbars> source generates a color bars pattern, based on
yading@10: the SMPTE RP 219-2002.
yading@10: 
yading@10: The C<testsrc> source generates a test video pattern, showing a
yading@10: color pattern, a scrolling gradient and a timestamp. This is mainly
yading@10: intended for testing purposes.
yading@10: 
yading@10: The sources accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<color, c>
yading@10: 
yading@10: Specify the color of the source, only used in the C<color>
yading@10: source. It can be the name of a color (case insensitive match) or a
yading@10: 0xRRGGBB[AA] sequence, possibly followed by an alpha specifier. The
yading@10: default value is "black".
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Specify the size of the sourced video, it may be a string of the form
yading@10: I<width>xI<height>, or the name of a size abbreviation. The
yading@10: default value is "320x240".
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Specify the frame rate of the sourced video, as the number of frames
yading@10: generated per second. It has to be a string in the format
yading@10: I<frame_rate_num>/I<frame_rate_den>, an integer number, a float
yading@10: number or a valid video frame rate abbreviation. The default value is
yading@10: "25".
yading@10: 
yading@10: 
yading@10: =item B<sar>
yading@10: 
yading@10: Set the sample aspect ratio of the sourced video.
yading@10: 
yading@10: 
yading@10: =item B<duration, d>
yading@10: 
yading@10: Set the video duration of the sourced video. The accepted syntax is:
yading@10: 	
yading@10: 	[-]HH[:MM[:SS[.m...]]]
yading@10: 	[-]S+[.m...]
yading@10: 
yading@10: See also the function C<av_parse_time()>.
yading@10: 
yading@10: If not specified, or the expressed duration is negative, the video is
yading@10: supposed to be generated forever.
yading@10: 
yading@10: 
yading@10: =item B<decimals, n>
yading@10: 
yading@10: Set the number of decimals to show in the timestamp, only used in the
yading@10: C<testsrc> source.
yading@10: 
yading@10: The displayed timestamp value will correspond to the original
yading@10: timestamp value multiplied by the power of 10 of the specified
yading@10: value. Default value is 0.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: For example the following:
yading@10: 	
yading@10: 	testsrc=duration=5.3:size=qcif:rate=10
yading@10: 
yading@10: 
yading@10: will generate a video with a duration of 5.3 seconds, with size
yading@10: 176x144 and a frame rate of 10 frames per second.
yading@10: 
yading@10: The following graph description will generate a red source
yading@10: with an opacity of 0.2, with size "qcif" and a frame rate of 10
yading@10: frames per second.
yading@10: 	
yading@10: 	color=c=red@0.2:s=qcif:r=10
yading@10: 
yading@10: 
yading@10: If the input content is to be ignored, C<nullsrc> can be used. The
yading@10: following command generates noise in the luminance plane by employing
yading@10: the C<geq> filter:
yading@10: 	
yading@10: 	nullsrc=s=256x256, geq=random(1)*255:128:128
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 VIDEO SINKS
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available video sinks.
yading@10: 
yading@10: 
yading@10: =head2 buffersink
yading@10: 
yading@10: 
yading@10: Buffer video frames, and make them available to the end of the filter
yading@10: graph.
yading@10: 
yading@10: This sink is mainly intended for a programmatic use, in particular
yading@10: through the interface defined in F<libavfilter/buffersink.h>
yading@10: or the options system.
yading@10: 
yading@10: It accepts a pointer to an AVBufferSinkContext structure, which
yading@10: defines the incoming buffers' formats, to be passed as the opaque
yading@10: parameter to C<avfilter_init_filter> for initialization.
yading@10: 
yading@10: 
yading@10: =head2 nullsink
yading@10: 
yading@10: 
yading@10: Null video sink, do absolutely nothing with the input video. It is
yading@10: mainly useful as a template and to be employed in analysis / debugging
yading@10: tools.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 MULTIMEDIA FILTERS
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available multimedia filters.
yading@10: 
yading@10: 
yading@10: =head2 aperms, perms
yading@10: 
yading@10: 
yading@10: Set read/write permissions for the output frames.
yading@10: 
yading@10: These filters are mainly aimed at developers to test direct path in the
yading@10: following filter in the filtergraph.
yading@10: 
yading@10: The filters accept the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Select the permissions mode.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<none>
yading@10: 
yading@10: Do nothing. This is the default.
yading@10: 
yading@10: =item B<ro>
yading@10: 
yading@10: Set all the output frames read-only.
yading@10: 
yading@10: =item B<rw>
yading@10: 
yading@10: Set all the output frames directly writable.
yading@10: 
yading@10: =item B<toggle>
yading@10: 
yading@10: Make the frame read-only if writable, and writable if read-only.
yading@10: 
yading@10: =item B<random>
yading@10: 
yading@10: Set each output frame read-only or writable randomly.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<seed>
yading@10: 
yading@10: Set the seed for the I<random> mode, must be an integer included between
yading@10: C<0> and C<UINT32_MAX>. If not specified, or if explicitly set to
yading@10: C<-1>, the filter will try to use a good random seed on a best effort
yading@10: basis.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Note: in case of auto-inserted filter between the permission filter and the
yading@10: following one, the permission might not be received as expected in that
yading@10: following filter. Inserting a format or aformat filter before the
yading@10: perms/aperms filter can avoid this problem.
yading@10: 
yading@10: 
yading@10: =head2 aselect, select
yading@10: 
yading@10: Select frames to pass in output.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<expr, e>
yading@10: 
yading@10: Set expression, which is evaluated for each input frame.
yading@10: 
yading@10: If the expression is evaluated to zero, the frame is discarded.
yading@10: 
yading@10: If the evaluation result is negative or NaN, the frame is sent to the
yading@10: first output; otherwise it is sent to the output with index
yading@10: C<ceil(val)-1>, assuming that the input index starts from 0.
yading@10: 
yading@10: For example a value of C<1.2> corresponds to the output with index
yading@10: C<ceil(1.2)-1 = 2-1 = 1>, that is the second output.
yading@10: 
yading@10: 
yading@10: =item B<outputs, n>
yading@10: 
yading@10: Set the number of outputs. The output to which to send the selected
yading@10: frame is based on the result of the evaluation. Default value is 1.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The expression can contain the following constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: the sequential number of the filtered frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<selected_n>
yading@10: 
yading@10: the sequential number of the selected frame, starting from 0
yading@10: 
yading@10: 
yading@10: =item B<prev_selected_n>
yading@10: 
yading@10: the sequential number of the last selected frame, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<TB>
yading@10: 
yading@10: timebase of the input timestamps
yading@10: 
yading@10: 
yading@10: =item B<pts>
yading@10: 
yading@10: the PTS (Presentation TimeStamp) of the filtered video frame,
yading@10: expressed in I<TB> units, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<t>
yading@10: 
yading@10: the PTS (Presentation TimeStamp) of the filtered video frame,
yading@10: expressed in seconds, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<prev_pts>
yading@10: 
yading@10: the PTS of the previously filtered video frame, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<prev_selected_pts>
yading@10: 
yading@10: the PTS of the last previously filtered video frame, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<prev_selected_t>
yading@10: 
yading@10: the PTS of the last previously selected video frame, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<start_pts>
yading@10: 
yading@10: the PTS of the first video frame in the video, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<start_t>
yading@10: 
yading@10: the time of the first video frame in the video, NAN if undefined
yading@10: 
yading@10: 
yading@10: =item B<pict_type> I<(video only)>
yading@10: 
yading@10: the type of the filtered frame, can assume one of the following
yading@10: values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<I>
yading@10: 
yading@10: 
yading@10: =item B<P>
yading@10: 
yading@10: 
yading@10: =item B<B>
yading@10: 
yading@10: 
yading@10: =item B<S>
yading@10: 
yading@10: 
yading@10: =item B<SI>
yading@10: 
yading@10: 
yading@10: =item B<SP>
yading@10: 
yading@10: 
yading@10: =item B<BI>
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<interlace_type> I<(video only)>
yading@10: 
yading@10: the frame interlace type, can assume one of the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<PROGRESSIVE>
yading@10: 
yading@10: the frame is progressive (not interlaced)
yading@10: 
yading@10: =item B<TOPFIRST>
yading@10: 
yading@10: the frame is top-field-first
yading@10: 
yading@10: =item B<BOTTOMFIRST>
yading@10: 
yading@10: the frame is bottom-field-first
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<consumed_sample_n> I<(audio only)>
yading@10: 
yading@10: the number of selected samples before the current frame
yading@10: 
yading@10: 
yading@10: =item B<samples_n> I<(audio only)>
yading@10: 
yading@10: the number of samples in the current frame
yading@10: 
yading@10: 
yading@10: =item B<sample_rate> I<(audio only)>
yading@10: 
yading@10: the input sample rate
yading@10: 
yading@10: 
yading@10: =item B<key>
yading@10: 
yading@10: 1 if the filtered frame is a key-frame, 0 otherwise
yading@10: 
yading@10: 
yading@10: =item B<pos>
yading@10: 
yading@10: the position in the file of the filtered frame, -1 if the information
yading@10: is not available (e.g. for synthetic video)
yading@10: 
yading@10: 
yading@10: =item B<scene> I<(video only)>
yading@10: 
yading@10: value between 0 and 1 to indicate a new scene; a low value reflects a low
yading@10: probability for the current frame to introduce a new scene, while a higher
yading@10: value means the current frame is more likely to be one (see the example below)
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The default value of the select expression is "1".
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select all frames in input:
yading@10: 	
yading@10: 	select
yading@10: 
yading@10: 
yading@10: The example above is the same as:
yading@10: 	
yading@10: 	select=1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Skip all frames:
yading@10: 	
yading@10: 	select=0
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select only I-frames:
yading@10: 	
yading@10: 	select='eq(pict_type\,I)'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select one frame every 100:
yading@10: 	
yading@10: 	select='not(mod(n\,100))'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select only frames contained in the 10-20 time interval:
yading@10: 	
yading@10: 	select='gte(t\,10)*lte(t\,20)'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select only I frames contained in the 10-20 time interval:
yading@10: 	
yading@10: 	select='gte(t\,10)*lte(t\,20)*eq(pict_type\,I)'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Select frames with a minimum distance of 10 seconds:
yading@10: 	
yading@10: 	select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Use aselect to select only audio frames with samples number E<gt> 100:
yading@10: 	
yading@10: 	aselect='gt(samples_n\,100)'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create a mosaic of the first scenes:
yading@10: 	
yading@10: 	ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
yading@10: 
yading@10: 
yading@10: Comparing I<scene> against a value between 0.3 and 0.5 is generally a sane
yading@10: choice.
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Send even and odd frames to separate outputs, and compose them:
yading@10: 	
yading@10: 	select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 asendcmd, sendcmd
yading@10: 
yading@10: 
yading@10: Send commands to filters in the filtergraph.
yading@10: 
yading@10: These filters read commands to be sent to other filters in the
yading@10: filtergraph.
yading@10: 
yading@10: C<asendcmd> must be inserted between two audio filters,
yading@10: C<sendcmd> must be inserted between two video filters, but apart
yading@10: from that they act the same way.
yading@10: 
yading@10: The specification of commands can be provided in the filter arguments
yading@10: with the I<commands> option, or in a file specified by the
yading@10: I<filename> option.
yading@10: 
yading@10: These filters accept the following options:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<commands, c>
yading@10: 
yading@10: Set the commands to be read and sent to the other filters.
yading@10: 
yading@10: =item B<filename, f>
yading@10: 
yading@10: Set the filename of the commands to be read and sent to the other
yading@10: filters.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Commands syntax
yading@10: 
yading@10: 
yading@10: A commands description consists of a sequence of interval
yading@10: specifications, comprising a list of commands to be executed when a
yading@10: particular event related to that interval occurs. The occurring event
yading@10: is typically the current frame time entering or leaving a given time
yading@10: interval.
yading@10: 
yading@10: An interval is specified by the following syntax:
yading@10: 	
yading@10: 	<START>[-<END>] <COMMANDS>;
yading@10: 
yading@10: 
yading@10: The time interval is specified by the I<START> and I<END> times.
yading@10: I<END> is optional and defaults to the maximum time.
yading@10: 
yading@10: The current frame time is considered within the specified interval if
yading@10: it is included in the interval [I<START>, I<END>), that is when
yading@10: the time is greater or equal to I<START> and is lesser than
yading@10: I<END>.
yading@10: 
yading@10: I<COMMANDS> consists of a sequence of one or more command
yading@10: specifications, separated by ",", relating to that interval.  The
yading@10: syntax of a command specification is given by:
yading@10: 	
yading@10: 	[<FLAGS>] <TARGET> <COMMAND> <ARG>
yading@10: 
yading@10: 
yading@10: I<FLAGS> is optional and specifies the type of events relating to
yading@10: the time interval which enable sending the specified command, and must
yading@10: be a non-null sequence of identifier flags separated by "+" or "|" and
yading@10: enclosed between "[" and "]".
yading@10: 
yading@10: The following flags are recognized:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<enter>
yading@10: 
yading@10: The command is sent when the current frame timestamp enters the
yading@10: specified interval. In other words, the command is sent when the
yading@10: previous frame timestamp was not in the given interval, and the
yading@10: current is.
yading@10: 
yading@10: 
yading@10: =item B<leave>
yading@10: 
yading@10: The command is sent when the current frame timestamp leaves the
yading@10: specified interval. In other words, the command is sent when the
yading@10: previous frame timestamp was in the given interval, and the
yading@10: current is not.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: If I<FLAGS> is not specified, a default value of C<[enter]> is
yading@10: assumed.
yading@10: 
yading@10: I<TARGET> specifies the target of the command, usually the name of
yading@10: the filter class or a specific filter instance name.
yading@10: 
yading@10: I<COMMAND> specifies the name of the command for the target filter.
yading@10: 
yading@10: I<ARG> is optional and specifies the optional list of argument for
yading@10: the given I<COMMAND>.
yading@10: 
yading@10: Between one interval specification and another, whitespaces, or
yading@10: sequences of characters starting with C<#> until the end of line,
yading@10: are ignored and can be used to annotate comments.
yading@10: 
yading@10: A simplified BNF description of the commands specification syntax
yading@10: follows:
yading@10: 	
yading@10: 	<COMMAND_FLAG>  ::= "enter" | "leave"
yading@10: 	<COMMAND_FLAGS> ::= <COMMAND_FLAG> [(+|"|")<COMMAND_FLAG>]
yading@10: 	<COMMAND>       ::= ["[" <COMMAND_FLAGS> "]"] <TARGET> <COMMAND> [<ARG>]
yading@10: 	<COMMANDS>      ::= <COMMAND> [,<COMMANDS>]
yading@10: 	<INTERVAL>      ::= <START>[-<END>] <COMMANDS>
yading@10: 	<INTERVALS>     ::= <INTERVAL>[;<INTERVALS>]
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify audio tempo change at second 4:
yading@10: 	
yading@10: 	asendcmd=c='4.0 atempo tempo 1.5',atempo
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Specify a list of drawtext and hue commands in a file.
yading@10: 	
yading@10: 	# show text in the interval 5-10
yading@10: 	5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
yading@10: 	         [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
yading@10: 	
yading@10: 	# desaturate the image in the interval 15-20
yading@10: 	15.0-20.0 [enter] hue s 0,
yading@10: 	          [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
yading@10: 	          [leave] hue s 1,
yading@10: 	          [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
yading@10: 	
yading@10: 	# apply an exponential saturation fade-out effect, starting from time 25
yading@10: 	25 [enter] hue s exp(25-t)
yading@10: 
yading@10: 
yading@10: A filtergraph allowing to read and process the above command list
yading@10: stored in a file F<test.cmd>, can be specified with:
yading@10: 	
yading@10: 	sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 asetpts, setpts
yading@10: 
yading@10: 
yading@10: Change the PTS (presentation timestamp) of the input frames.
yading@10: 
yading@10: C<asetpts> works on audio frames, C<setpts> on video frames.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<expr>
yading@10: 
yading@10: The expression which is evaluated for each frame to construct its timestamp.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The expression is evaluated through the eval API and can contain the following
yading@10: constants:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<FRAME_RATE>
yading@10: 
yading@10: frame rate, only defined for constant frame-rate video
yading@10: 
yading@10: 
yading@10: =item B<PTS>
yading@10: 
yading@10: the presentation timestamp in input
yading@10: 
yading@10: 
yading@10: =item B<N>
yading@10: 
yading@10: the count of the input frame, starting from 0.
yading@10: 
yading@10: 
yading@10: =item B<NB_CONSUMED_SAMPLES>
yading@10: 
yading@10: the number of consumed samples, not including the current frame (only
yading@10: audio)
yading@10: 
yading@10: 
yading@10: =item B<NB_SAMPLES>
yading@10: 
yading@10: the number of samples in the current frame (only audio)
yading@10: 
yading@10: 
yading@10: =item B<SAMPLE_RATE>
yading@10: 
yading@10: audio sample rate
yading@10: 
yading@10: 
yading@10: =item B<STARTPTS>
yading@10: 
yading@10: the PTS of the first frame
yading@10: 
yading@10: 
yading@10: =item B<STARTT>
yading@10: 
yading@10: the time in seconds of the first frame
yading@10: 
yading@10: 
yading@10: =item B<INTERLACED>
yading@10: 
yading@10: tell if the current frame is interlaced
yading@10: 
yading@10: 
yading@10: =item B<T>
yading@10: 
yading@10: the time in seconds of the current frame
yading@10: 
yading@10: 
yading@10: =item B<TB>
yading@10: 
yading@10: the time base
yading@10: 
yading@10: 
yading@10: =item B<POS>
yading@10: 
yading@10: original position in the file of the frame, or undefined if undefined
yading@10: for the current frame
yading@10: 
yading@10: 
yading@10: =item B<PREV_INPTS>
yading@10: 
yading@10: previous input PTS
yading@10: 
yading@10: 
yading@10: =item B<PREV_INT>
yading@10: 
yading@10: previous input time in seconds
yading@10: 
yading@10: 
yading@10: =item B<PREV_OUTPTS>
yading@10: 
yading@10: previous output PTS
yading@10: 
yading@10: 
yading@10: =item B<PREV_OUTT>
yading@10: 
yading@10: previous output time in seconds
yading@10: 
yading@10: 
yading@10: =item B<RTCTIME>
yading@10: 
yading@10: wallclock (RTC) time in microseconds. This is deprecated, use time(0)
yading@10: instead.
yading@10: 
yading@10: 
yading@10: =item B<RTCSTART>
yading@10: 
yading@10: wallclock (RTC) time at the start of the movie in microseconds
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Start counting PTS from zero
yading@10: 	
yading@10: 	setpts=PTS-STARTPTS
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply fast motion effect:
yading@10: 	
yading@10: 	setpts=0.5*PTS
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply slow motion effect:
yading@10: 	
yading@10: 	setpts=2.0*PTS
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set fixed rate of 25 frames per second:
yading@10: 	
yading@10: 	setpts=N/(25*TB)
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set fixed rate 25 fps with some jitter:
yading@10: 	
yading@10: 	setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Apply an offset of 10 seconds to the input PTS:
yading@10: 	
yading@10: 	setpts=PTS+10/TB
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Generate timestamps from a "live source" and rebase onto the current timebase:
yading@10: 	
yading@10: 	setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 ebur128
yading@10: 
yading@10: 
yading@10: EBU R128 scanner filter. This filter takes an audio stream as input and outputs
yading@10: it unchanged. By default, it logs a message at a frequency of 10Hz with the
yading@10: Momentary loudness (identified by C<M>), Short-term loudness (C<S>),
yading@10: Integrated loudness (C<I>) and Loudness Range (C<LRA>).
yading@10: 
yading@10: The filter also has a video output (see the I<video> option) with a real
yading@10: time graph to observe the loudness evolution. The graphic contains the logged
yading@10: message mentioned above, so it is not printed anymore when this option is set,
yading@10: unless the verbose logging is set. The main graphing area contains the
yading@10: short-term loudness (3 seconds of analysis), and the gauge on the right is for
yading@10: the momentary loudness (400 milliseconds).
yading@10: 
yading@10: More information about the Loudness Recommendation EBU R128 on
yading@10: E<lt>B<http://tech.ebu.ch/loudness>E<gt>.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<video>
yading@10: 
yading@10: Activate the video output. The audio stream is passed unchanged whether this
yading@10: option is set or no. The video stream will be the first output stream if
yading@10: activated. Default is C<0>.
yading@10: 
yading@10: 
yading@10: =item B<size>
yading@10: 
yading@10: Set the video size. This option is for video only. Default and minimum
yading@10: resolution is C<640x480>.
yading@10: 
yading@10: 
yading@10: =item B<meter>
yading@10: 
yading@10: Set the EBU scale meter. Default is C<9>. Common values are C<9> and
yading@10: C<18>, respectively for EBU scale meter +9 and EBU scale meter +18. Any
yading@10: other integer value between this range is allowed.
yading@10: 
yading@10: 
yading@10: =item B<metadata>
yading@10: 
yading@10: Set metadata injection. If set to C<1>, the audio input will be segmented
yading@10: into 100ms output frames, each of them containing various loudness information
yading@10: in metadata.  All the metadata keys are prefixed with C<lavfi.r128.>.
yading@10: 
yading@10: Default is C<0>.
yading@10: 
yading@10: 
yading@10: =item B<framelog>
yading@10: 
yading@10: Force the frame logging level.
yading@10: 
yading@10: Available values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<info>
yading@10: 
yading@10: information logging level
yading@10: 
yading@10: =item B<verbose>
yading@10: 
yading@10: verbose logging level
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: By default, the logging level is set to I<info>. If the B<video> or
yading@10: the B<metadata> options are set, it switches to I<verbose>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Real-time graph using B<ffplay>, with a EBU scale meter +18:
yading@10: 	
yading@10: 	ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Run an analysis with B<ffmpeg>:
yading@10: 	
yading@10: 	ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 settb, asettb
yading@10: 
yading@10: 
yading@10: Set the timebase to use for the output frames timestamps.
yading@10: It is mainly useful for testing timebase configuration.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<expr, tb>
yading@10: 
yading@10: The expression which is evaluated into the output timebase.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The value for B<tb> is an arithmetic expression representing a
yading@10: rational. The expression can contain the constants "AVTB" (the default
yading@10: timebase), "intb" (the input timebase) and "sr" (the sample rate,
yading@10: audio only). Default value is "intb".
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the timebase to 1/25:
yading@10: 	
yading@10: 	settb=expr=1/25
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the timebase to 1/10:
yading@10: 	
yading@10: 	settb=expr=0.1
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the timebase to 1001/1000:
yading@10: 	
yading@10: 	settb=1+0.001
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the timebase to 2*intb:
yading@10: 	
yading@10: 	settb=2*intb
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Set the default timebase value:
yading@10: 	
yading@10: 	settb=AVTB
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 concat
yading@10: 
yading@10: 
yading@10: Concatenate audio and video streams, joining them together one after the
yading@10: other.
yading@10: 
yading@10: The filter works on segments of synchronized video and audio streams. All
yading@10: segments must have the same number of streams of each type, and that will
yading@10: also be the number of streams at output.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: Set the number of segments. Default is 2.
yading@10: 
yading@10: 
yading@10: =item B<v>
yading@10: 
yading@10: Set the number of output video streams, that is also the number of video
yading@10: streams in each segment. Default is 1.
yading@10: 
yading@10: 
yading@10: =item B<a>
yading@10: 
yading@10: Set the number of output audio streams, that is also the number of video
yading@10: streams in each segment. Default is 0.
yading@10: 
yading@10: 
yading@10: =item B<unsafe>
yading@10: 
yading@10: Activate unsafe mode: do not fail if segments have a different format.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The filter has I<v>+I<a> outputs: first I<v> video outputs, then
yading@10: I<a> audio outputs.
yading@10: 
yading@10: There are I<n>x(I<v>+I<a>) inputs: first the inputs for the first
yading@10: segment, in the same order as the outputs, then the inputs for the second
yading@10: segment, etc.
yading@10: 
yading@10: Related streams do not always have exactly the same duration, for various
yading@10: reasons including codec frame size or sloppy authoring. For that reason,
yading@10: related synchronized streams (e.g. a video and its audio track) should be
yading@10: concatenated at once. The concat filter will use the duration of the longest
yading@10: stream in each segment (except the last one), and if necessary pad shorter
yading@10: audio streams with silence.
yading@10: 
yading@10: For this filter to work correctly, all segments must start at timestamp 0.
yading@10: 
yading@10: All corresponding streams must have the same parameters in all segments; the
yading@10: filtering system will automatically select a common pixel format for video
yading@10: streams, and a common sample format, sample rate and channel layout for
yading@10: audio streams, but other settings, such as resolution, must be converted
yading@10: explicitly by the user.
yading@10: 
yading@10: Different frame rates are acceptable but will result in variable frame rate
yading@10: at output; be sure to configure the output file to handle it.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Concatenate an opening, an episode and an ending, all in bilingual version
yading@10: (video in stream 0, audio in streams 1 and 2):
yading@10: 	
yading@10: 	ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
yading@10: 	  '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
yading@10: 	   concat=n=3:v=1:a=2 [v] [a1] [a2]' \
yading@10: 	  -map '[v]' -map '[a1]' -map '[a2]' output.mkv
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Concatenate two parts, handling audio and video separately, using the
yading@10: (a)movie sources, and adjusting the resolution:
yading@10: 	
yading@10: 	movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
yading@10: 	movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
yading@10: 	[v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
yading@10: 
yading@10: Note that a desync will happen at the stitch if the audio and video streams
yading@10: do not have exactly the same duration in the first file.
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 showspectrum
yading@10: 
yading@10: 
yading@10: Convert input audio to a video output, representing the audio frequency
yading@10: spectrum.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Specify the video size for the output. Default value is C<640x512>.
yading@10: 
yading@10: 
yading@10: =item B<slide>
yading@10: 
yading@10: Specify if the spectrum should slide along the window. Default value is
yading@10: C<0>.
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Specify display mode.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<combined>
yading@10: 
yading@10: all channels are displayed in the same row
yading@10: 
yading@10: =item B<separate>
yading@10: 
yading@10: all channels are displayed in separate rows
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<combined>.
yading@10: 
yading@10: 
yading@10: =item B<color>
yading@10: 
yading@10: Specify display color mode.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<channel>
yading@10: 
yading@10: each channel is displayed in a separate color
yading@10: 
yading@10: =item B<intensity>
yading@10: 
yading@10: each channel is is displayed using the same color scheme
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<channel>.
yading@10: 
yading@10: 
yading@10: =item B<scale>
yading@10: 
yading@10: Specify scale used for calculating intensity color values.
yading@10: 
yading@10: It accepts the following values:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<lin>
yading@10: 
yading@10: linear
yading@10: 
yading@10: =item B<sqrt>
yading@10: 
yading@10: square root, default
yading@10: 
yading@10: =item B<cbrt>
yading@10: 
yading@10: cubic root
yading@10: 
yading@10: =item B<log>
yading@10: 
yading@10: logarithmic
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is B<sqrt>.
yading@10: 
yading@10: 
yading@10: =item B<saturation>
yading@10: 
yading@10: Set saturation modifier for displayed colors. Negative values provide
yading@10: alternative color scheme. C<0> is no saturation at all.
yading@10: Saturation must be in [-10.0, 10.0] range.
yading@10: Default value is C<1>.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: The usage is very similar to the showwaves filter; see the examples in that
yading@10: section.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Large window with logarithmic color scaling:
yading@10: 	
yading@10: 	showspectrum=s=1280x480:scale=log
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Complete example for a colored and sliding spectrum per channel using B<ffplay>:
yading@10: 	
yading@10: 	ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
yading@10: 	             [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 showwaves
yading@10: 
yading@10: 
yading@10: Convert input audio to a video output, representing the samples waves.
yading@10: 
yading@10: The filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<size, s>
yading@10: 
yading@10: Specify the video size for the output. Default value is "600x240".
yading@10: 
yading@10: 
yading@10: =item B<mode>
yading@10: 
yading@10: Set display mode.
yading@10: 
yading@10: Available values are:
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<point>
yading@10: 
yading@10: Draw a point for each sample.
yading@10: 
yading@10: 
yading@10: =item B<line>
yading@10: 
yading@10: Draw a vertical line for each sample.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: Default value is C<point>.
yading@10: 
yading@10: 
yading@10: =item B<n>
yading@10: 
yading@10: Set the number of samples which are printed on the same column. A
yading@10: larger value will decrease the frame rate. Must be a positive
yading@10: integer. This option can be set only if the value for I<rate>
yading@10: is not explicitly specified.
yading@10: 
yading@10: 
yading@10: =item B<rate, r>
yading@10: 
yading@10: Set the (approximate) output frame rate. This is done by setting the
yading@10: option I<n>. Default value is "25".
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Output the input file audio and the corresponding video representation
yading@10: at the same time:
yading@10: 	
yading@10: 	amovie=a.mp3,asplit[out0],showwaves[out1]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create a synthetic signal and show it with showwaves, forcing a
yading@10: frame rate of 30 frames per second:
yading@10: 	
yading@10: 	aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 split, asplit
yading@10: 
yading@10: 
yading@10: Split input into several identical outputs.
yading@10: 
yading@10: C<asplit> works with audio input, C<split> with video.
yading@10: 
yading@10: The filter accepts a single parameter which specifies the number of outputs. If
yading@10: unspecified, it defaults to 2.
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create two separate outputs from the same input:
yading@10: 	
yading@10: 	[in] split [out0][out1]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: To create 3 or more outputs, you need to specify the number of
yading@10: outputs, like in:
yading@10: 	
yading@10: 	[in] asplit=3 [out0][out1][out2]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create two separate outputs from the same input, one cropped and
yading@10: one padded:
yading@10: 	
yading@10: 	[in] split [splitout1][splitout2];
yading@10: 	[splitout1] crop=100:100:0:0    [cropout];
yading@10: 	[splitout2] pad=200:200:100:100 [padout];
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Create 5 copies of the input audio with B<ffmpeg>:
yading@10: 	
yading@10: 	ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 MULTIMEDIA SOURCES
yading@10: 
yading@10: 
yading@10: Below is a description of the currently available multimedia sources.
yading@10: 
yading@10: 
yading@10: =head2 amovie
yading@10: 
yading@10: 
yading@10: This is the same as movie source, except it selects an audio
yading@10: stream by default.
yading@10: 
yading@10: 
yading@10: 
yading@10: =head2 movie
yading@10: 
yading@10: 
yading@10: Read audio and/or video stream(s) from a movie container.
yading@10: 
yading@10: This filter accepts the following options:
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item B<filename>
yading@10: 
yading@10: The name of the resource to read (not necessarily a file but also a device or a
yading@10: stream accessed through some protocol).
yading@10: 
yading@10: 
yading@10: =item B<format_name, f>
yading@10: 
yading@10: Specifies the format assumed for the movie to read, and can be either
yading@10: the name of a container or an input device. If not specified the
yading@10: format is guessed from I<movie_name> or by probing.
yading@10: 
yading@10: 
yading@10: =item B<seek_point, sp>
yading@10: 
yading@10: Specifies the seek point in seconds, the frames will be output
yading@10: starting from this seek point, the parameter is evaluated with
yading@10: C<av_strtod> so the numerical value may be suffixed by an IS
yading@10: postfix. Default value is "0".
yading@10: 
yading@10: 
yading@10: =item B<streams, s>
yading@10: 
yading@10: Specifies the streams to read. Several streams can be specified,
yading@10: separated by "+". The source will then have as many outputs, in the
yading@10: same order. The syntax is explained in the ``Stream specifiers''
yading@10: section in the ffmpeg manual. Two special names, "dv" and "da" specify
yading@10: respectively the default (best suited) video and audio stream. Default
yading@10: is "dv", or "da" if the filter is called as "amovie".
yading@10: 
yading@10: 
yading@10: =item B<stream_index, si>
yading@10: 
yading@10: Specifies the index of the video stream to read. If the value is -1,
yading@10: the best suited video stream will be automatically selected. Default
yading@10: value is "-1". Deprecated. If the filter is called "amovie", it will select
yading@10: audio instead of video.
yading@10: 
yading@10: 
yading@10: =item B<loop>
yading@10: 
yading@10: Specifies how many times to read the stream in sequence.
yading@10: If the value is less than 1, the stream will be read again and again.
yading@10: Default value is "1".
yading@10: 
yading@10: Note that when the movie is looped the source timestamps are not
yading@10: changed, so it will generate non monotonically increasing timestamps.
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: This filter allows to overlay a second video on top of main input of
yading@10: a filtergraph as shown in this graph:
yading@10: 	
yading@10: 	input -----------> deltapts0 --> overlay --> output
yading@10: 	                                    ^
yading@10: 	                                    |
yading@10: 	movie --> scale--> deltapts1 -------+
yading@10: 
yading@10: 
yading@10: 
yading@10: =head3 Examples
yading@10: 
yading@10: 
yading@10: 
yading@10: =over 4
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Skip 3.2 seconds from the start of the avi file in.avi, and overlay it
yading@10: on top of the input labelled as "in":
yading@10: 	
yading@10: 	movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
yading@10: 	[in] setpts=PTS-STARTPTS [main];
yading@10: 	[main][over] overlay=16:16 [out]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read from a video4linux2 device, and overlay it on top of the input
yading@10: labelled as "in":
yading@10: 	
yading@10: 	movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
yading@10: 	[in] setpts=PTS-STARTPTS [main];
yading@10: 	[main][over] overlay=16:16 [out]
yading@10: 
yading@10: 
yading@10: 
yading@10: =item *
yading@10: 
yading@10: Read the first video stream and the audio stream with id 0x81 from
yading@10: dvd.vob; the video is connected to the pad named "video" and the audio is
yading@10: connected to the pad named "audio":
yading@10: 	
yading@10: 	movie=dvd.vob:s=v:0+#0x81 [video] [audio]
yading@10: 
yading@10: 
yading@10: =back
yading@10: 
yading@10: 
yading@10: 
yading@10: 
yading@10: =head1 SEE ALSO
yading@10: 
yading@10: 
yading@10: 
yading@10: ffmpeg(1), ffplay(1), ffprobe(1), ffserver(1), libavfilter(3)
yading@10: 
yading@10: 
yading@10: =head1 AUTHORS
yading@10: 
yading@10: 
yading@10: The FFmpeg developers.
yading@10: 
yading@10: For details about the authorship, see the Git history of the project
yading@10: (git://source.ffmpeg.org/ffmpeg), e.g. by typing the command
yading@10: B<git log> in the FFmpeg source directory, or browsing the
yading@10: online repository at E<lt>B<http://source.ffmpeg.org>E<gt>.
yading@10: 
yading@10: Maintainers for the specific components are listed in the file
yading@10: F<MAINTAINERS> in the source code tree.
yading@10: 
yading@10: 
yading@10: