Libavfilter Documentation
Libavfilter is the filtering API of FFmpeg. It is the substitute of the now deprecated 'vhooks' and started as a Google Summer of Code project.
Integrating libavfilter into the main FFmpeg repository is a work in progress. If you wish to try the unfinished development code of libavfilter then check it out from the libavfilter repository into some directory of your choice by:
svn checkout svn://svn.ffmpeg.org/soc/libavfilter
And then read the README file in the top directory to learn how to integrate it into ffmpeg and ffplay.
But note that there may still be serious bugs in the code and its API and ABI should not be considered stable yet!
In libavfilter, it is possible for filters to have multiple inputs and multiple outputs. To illustrate the sorts of things that are possible, we can use a complex filter graph. For example, the following one:
input --> split --> fifo -----------------------> overlay --> output | ^ | | +------> fifo --> crop --> vflip --------+
splits the stream in two streams, sends one stream through the crop filter and the vflip filter before merging it back with the other stream by overlaying it on top. You can use the following command to achieve this:
./ffmpeg -i in.avi -s 240x320 -vf "[in] split [T1], fifo, [T2] overlay= 0:240 [out]; [T1] fifo, crop=0:0:-1:240, vflip [T2]
where input_video.avi has a vertical resolution of 480 pixels. The result will be that in output the top half of the video is mirrored onto the bottom half.
Video filters are loaded using the -vf option passed to ffmpeg or to ffplay. Filters in the same linear chain are separated by commas. In our example, split, fifo, overlay are in one linear chain, and fifo, crop, vflip are in another. The points where the linear chains join are labeled by names enclosed in square brackets. In our example, that is [T1] and [T2]. The magic labels [in] and [out] are the points where video is input and output.
Some filters take in input a list of parameters: they are specified after the filter name and an equal sign, and are separated each other by a semicolon.
There exist so-called source filters that do not have a video input, and we expect in the future some sink filters that will not have video output.
The `graph2dot' program included in the FFmpeg `tools' directory can be used to parse a filter graph description and issue a corresponding textual representation in the dot language.
Invoke the command:
graph2dot -h
to see how to use `graph2dot'.
You can then pass the dot description to the `dot' program (from the graphviz suite of programs) and obtain a graphical representation of the filter graph.
For example the sequence of commands:
echo GRAPH_DESCRIPTION | \ tools/graph2dot -o graph.tmp && \ dot -Tpng graph.tmp -o graph.png && \ display graph.png
can be used to create and display an image representing the graph described by the GRAPH_DESCRIPTION string.
When you configure your FFmpeg build, you can disable any of the existing filters using --disable-filters. The configure output will show the audio filters included in your build.
Below is a description of the currently available audio filters.
Pass the audio source unchanged to the output.
When you configure your FFmpeg build, you can disable any of the existing filters using --disable-filters. The configure output will show the video filters included in your build.
Below is a description of the currently available video filters.
Crop the input video to x:y:width:height.
./ffmpeg -i in.avi -vf "crop=0:0:0:240" out.avi
x and y specify the position of the top-left corner of the output (non-cropped) area.
The default value of x and y is 0.
The width and height parameters specify the width and height of the output (non-cropped) area.
A value of 0 is interpreted as the maximum possible size contained in the area delimited by the top-left corner at position x:y.
For example the parameters:
"crop=100:100:0:0"
will delimit the rectangle with the top-left corner placed at position 100:100 and the right-bottom corner corresponding to the right-bottom corner of the input image.
The default value of width and height is 0.
Buffer input images and send them when they are requested.
This filter is mainly useful when auto-inserted by the libavfilter framework.
The filter does not take parameters.
Convert the input video to one of the specified pixel formats. Libavfilter will try to pick one that is supported for the input to the next filter.
The filter accepts a list of pixel format names, separated by ":", for example "yuv420p:monow:rgb24".
The following command:
./ffmpeg -i in.avi -vf "format=yuv420p" out.avi
will convert the input video to the format "yuv420p".
Flip the input video horizontally.
For example to horizontally flip the video in input with `ffmpeg':
ffmpeg -i in.avi -vf "hflip" out.avi
Force libavfilter not to use any of the specified pixel formats for the input to the next filter.
The filter accepts a list of pixel format names, separated by ":", for example "yuv420p:monow:rgb24".
The following command:
./ffmpeg -i in.avi -vf "noformat=yuv420p, vflip" out.avi
will make libavfilter use a format different from "yuv420p" for the input to the vflip filter.
Pass the video source unchanged to the output.
Apply smooth transform using libopencv.
To enable this filter install libopencv library and headers and configure FFmpeg with --enable-libopencv.
It accepts the following parameters: type:param1:param2:param3:param4.
type is the type of smooth filter to apply, and can be one of the following value: "blur", "blur_no_scale", "median", "gaussian", "bilateral". The default value is "gaussian".
param1, param2, param3, and param4 are parameters whose meanings depend on smooth type. param1 and param2 accept integer positive values or 0, param3 and param4 accept float values.
The default value for param1 is 3, the default value for the other parameters is 0.
These parameters corresponds to the parameters assigned to the
libopencv function cvSmooth
. Refer the official libopencv
documentation for the exact meaning of the parameters:
http://opencv.willowgarage.com/documentation/c/image_filtering.html
Add paddings to the input image, and places the original input at the given coordinates x, y.
It accepts the following parameters: width:height:x:y:color.
Follows the description of the accepted parameters.
Pixel format descriptor test filter, mainly useful for internal testing. The output video should be equal to the input video.
For example:
format=monow, pixdesctest
can be used to test the monowhite pixel format descriptor definition.
Scale the input video to width:height and/or convert the image format.
For example the command:
./ffmpeg -i in.avi -vf "scale=200:100" out.avi
will scale the input video to a size of 200x100.
If the input image format is different from the format requested by the next filter, the scale filter will convert the input to the requested format.
If the value for width or height is 0, the respective input size is used for the output.
If the value for width or height is -1, the scale filter will use, for the respective output size, a value that maintains the aspect ratio of the input image.
The default value of width and height is 0.
Pass the images of input video on to next video filter as multiple slices.
./ffmpeg -i in.avi -vf "slicify=32" out.avi
The filter accepts the slice height as parameter. If the parameter is not specified it will use the default value of 16.
Adding this in the beginning of filter chains should make filtering faster due to better use of the memory cache.
Sharpen or blur the input video.
It accepts the following parameters: luma_msize_x:luma_msize_y:luma_amount:chroma_msize_x:chroma_msize_y:chroma_amount
Negative values for the amount will blur the input video, while positive values will sharpen. All parameters are optional and default to the equivalent of the string '5:5:1.0:0:0:0.0'.
# Strong luma sharpen effect parameters unsharp=7:7:2.5 # Strong blur of both luma and chroma parameters unsharp=7:7:-2:7:7:-2 # Use the default values with @command{ffmpeg} ./ffmpeg -i in.avi -vf "unsharp" out.mp4
Flip the input video vertically.
./ffmpeg -i in.avi -vf "vflip" out.avi
Below is a description of the currently available video sources.
Buffer video frames, and make them available to the filter chain.
This source is mainly intended for a programmatic use, in particular through the interface defined in `libavfilter/vsrc_buffer.h'.
It accepts the following parameters: width:height:pix_fmt_string
All the parameters need to be explicitely defined.
Follows the list of the accepted parameters.
For example:
buffer=320:240:yuv410p
will instruct the source to accept video frames with size 320x240 and with format "yuv410p". Since the pixel format with name "yuv410p" corresponds to the number 6 (check the enum PixelFormat definition in `libavutil/pixfmt.h'), this example corresponds to:
buffer=320:240:6
Provide an uniformly colored input.
It accepts the following parameters: color:frame_size:frame_rate
Follows the description of the accepted parameters.
For example the following graph description will generate a red source with an opacity of 0.2, with size "qcif" and a frame rate of 10 frames per second, which will be overlayed over the source connected to the pad with identifier "in".
"color=red@0.2:qcif:10 [color]; [in][color] overlay [out]"
Null video source, never return images. It is mainly useful as a template and to be employed in analysis / debugging tools.
It accepts as optional parameter a string of the form width:height, where width and height specify the size of the configured source.
The default values of width and height are respectively 352 and 288 (corresponding to the CIF size format).
Below is a description of the currently available video sinks.
Null video sink, do absolutely nothing with the input video. It is mainly useful as a template and to be employed in analysis / debugging tools.
This document was generated on 15 September 2010 using texi2html 1.56k.