1 \input texinfo @c -*- texinfo -*-
3 @settitle FFmpeg Documentation
6 @center @titlefont{FFmpeg Documentation}
13 FFmpeg is a very fast video and audio converter. It can also grab from
14 a live audio/video source.
16 The command line interface is designed to be intuitive, in the sense
17 that ffmpeg tries to figure out all the parameters, when
18 possible. You have usually to give only the target bitrate you want.
20 FFmpeg can also convert from any sample rate to any other, and resize
21 video on the fly with a high quality polyphase filter.
26 @section Video and Audio grabbing
28 FFmpeg can use a video4linux compatible video source and any Open Sound
35 Note that you must activate the right video source and channel before
36 launching ffmpeg. You can use any TV viewer such as xawtv
37 (@url{http://bytesex.org/xawtv/}) by Gerd Knorr which I find very
38 good. You must also set correctly the audio recording levels with a
41 @section Video and Audio file format conversion
43 * ffmpeg can use any supported file format and protocol as input:
47 * You can input from YUV files:
50 ffmpeg -i /tmp/test%d.Y /tmp/out.mpg
53 It will use the files:
55 /tmp/test0.Y, /tmp/test0.U, /tmp/test0.V,
56 /tmp/test1.Y, /tmp/test1.U, /tmp/test1.V, etc...
59 The Y files use twice the resolution of the U and V files. They are
60 raw files, without header. They can be generated by all decent video
61 decoders. You must specify the size of the image with the @option{-s} option
62 if ffmpeg cannot guess it.
64 * You can input from a RAW YUV420P file:
67 ffmpeg -i /tmp/test.yuv /tmp/out.avi
70 The RAW YUV420P is a file containing RAW YUV planar, for each frame first
71 come the Y plane followed by U and V planes, which are half vertical and
72 horizontal resolution.
74 * You can output to a RAW YUV420P file:
77 ffmpeg -i mydivx.avi -o hugefile.yuv
80 * You can set several input files and output files:
83 ffmpeg -i /tmp/a.wav -s 640x480 -i /tmp/a.yuv /tmp/a.mpg
86 Convert the audio file a.wav and the raw yuv video file a.yuv
89 * You can also do audio and video conversions at the same time:
92 ffmpeg -i /tmp/a.wav -ar 22050 /tmp/a.mp2
95 Convert the sample rate of a.wav to 22050 Hz and encode it to MPEG audio.
97 * You can encode to several formats at the same time and define a
98 mapping from input stream to output streams:
101 ffmpeg -i /tmp/a.wav -ab 64 /tmp/a.mp2 -ab 128 /tmp/b.mp2 -map 0:0 -map 0:0
104 Convert a.wav to a.mp2 at 64 kbits and b.mp2 at 128 kbits. '-map
105 file:index' specify which input stream is used for each output
106 stream, in the order of the definition of output streams.
108 * You can transcode decrypted VOBs
111 ffmpeg -i snatch_1.vob -f avi -vcodec mpeg4 -b 800 -g 300 -bf 2 -acodec mp3 -ab 128 snatch.avi
114 This is a typical DVD ripper example, input from a VOB file, output
115 to an AVI file with MPEG-4 video and MP3 audio, note that in this
116 command we use B frames so the MPEG-4 stream is DivX5 compatible, GOP
117 size is 300 that means an INTRA frame every 10 seconds for 29.97 fps
118 input video. Also the audio stream is MP3 encoded so you need LAME
119 support which is enabled using @code{--enable-mp3lame} when
120 configuring. The mapping is particularly useful for DVD transcoding
121 to get the desired audio language.
123 NOTE: to see the supported input formats, use @code{ffmpeg -formats}.
130 The generic syntax is:
133 @c man begin SYNOPSIS
134 ffmpeg [[options][@option{-i} @var{input_file}]]... @{[options] @var{output_file}@}...
137 @c man begin DESCRIPTION
138 If no input file is given, audio/video grabbing is done.
140 As a general rule, options are applied to the next specified
141 file. For example, if you give the @option{-b 64} option, it sets the video
142 bitrate of the next file. Format option may be needed for raw input
145 By default, ffmpeg tries to convert as losslessly as possible: it
146 uses the same audio and video parameter for the outputs as the one
147 specified for the inputs.
151 @section Main options
161 show available formats, codecs, protocols, ...
170 overwrite output files
173 set the recording time in seconds. @code{hh:mm:ss[.xxx]} syntax is also
182 @item -copyright string
185 @item -comment string
189 activate high quality settings
193 @section Video Options
197 set the video bitrate in kbit/s (default = 200 kb/s)
199 set frame rate (default = 25)
201 set frame size. The format is @samp{WxH}. (default 160x128)
202 The following abbreviations are regognized:
215 set aspect ratio (4:3, 16:9 or 1.3333, 1.7777)
217 set top crop band size (in pixels)
218 @item -cropbottom size
219 set bottom crop band size (in pixels)
221 set left crop band size (in pixels)
222 @item -cropright size
223 set right crop band size (in pixels)
225 disable video recording
227 set video bitrate tolerance (in kbit/s)
228 @item -maxrate bitrate
229 set max video bitrate tolerance (in kbit/s)
230 @item -minrate bitrate
231 set min video bitrate tolerance (in kbit/s)
233 set ratecontrol buffere size (in kbit)
237 use same video quality as source (implies VBR)
240 select the pass number (1 or 2). It is useful to do two pass encoding. The statistics of the video are recorded in the first pass and the video at the exact requested bit rate is generated in the second pass.
242 @item -passlogfile file
243 select two pass log file name
247 @section Advanced Video Options
251 set the group of picture size
253 use only intra frames
255 use fixed video quantiser scale (VBR)
257 min video quantiser scale (VBR)
259 max video quantiser scale (VBR)
261 max difference between the quantiser scale (VBR)
263 video quantiser scale blur (VBR)
264 @item -qcomp compression
265 video quantiser scale compression (VBR)
267 @item -rc_init_cplx complexity
268 initial complexity for 1-pass encoding
269 @item -b_qfactor factor
270 qp factor between p and b frames
271 @item -i_qfactor factor
272 qp factor between p and i frames
273 @item -b_qoffset offset
274 qp offset between p and b frames
275 @item -i_qoffset offset
276 qp offset between p and i frames
277 @item -rc_eq equation
278 set rate control equation (@pxref{FFmpeg formula evaluator})
279 @item -rc_override override_string
280 rate control override for specific internals
282 set motion estimation method. Available methods are (from lower to best quality):
285 Try just (0, 0) vector.
292 exhaustive search (slow and marginally better than epzs)
299 FF_DCT_AUTO (default)
312 @item -idct_algo algo
316 FF_IDCT_AUTO (default)
340 set error resilience to 'n'
345 FF_ER_COMPLIANT (default)
349 FF_ER_VERY_AGGRESSIVE
353 set error concealment to n
362 use 'frames' B frames (only MPEG-4)
367 FF_MB_DECISION_SIMPLE: use mb_cmp (cannot change it yet in ffmpeg)
369 FF_MB_DECISION_BITS: chooses the one which needs the fewest bits
371 FF_MB_DECISION_RD: rate distoration
375 use four motion vector by macroblock (only MPEG-4)
377 use data partitioning (only MPEG-4)
379 workaround not auto detected encoder bugs
380 @item -strict strictness
381 how strictly to follow the standarts
386 calculate PSNR of compressed frames
388 dump video coding statistics to file
390 insert video processing @var{module}. @var{module} contains the module
391 name and its parameters separated by spaces.
393 enable Advanced intra coding (h263+)
395 enable Unlimited Motion Vector (h263+)
398 @section Audio Options
402 set audio bitrate (in kbit/s)
404 set the audio sampling freq (default = 44100 Hz)
406 set the audio bitrate in kbit/s (default = 64)
408 set the number of audio channels (default = 1)
410 disable audio recording
415 @section Audio/Video grab options
419 set video grab device (e.g. @file{/dev/video0})
421 set video grab channel (DV1394 only)
422 @item -tvstd standard
423 set television standard (NTSC, PAL (SECAM))
427 set audio device (e.g. @file{/dev/dsp})
430 @section Advanced options
433 @item -map file:stream
434 set input stream mapping
436 print specific debug info
438 add timings for benchmarking
440 dump each input packet
442 only use bit exact algorithms (for codec testing)
444 set packet size in bits
447 @node FFmpeg formula evaluator
448 @section FFmpeg formula evaluator
450 When evaluating a rate control string, FFmpeg uses an internal formula
453 The following binary operators are available: @code{+}, @code{-},
454 @code{*}, @code{/}, @code{^}.
456 The following unary operators are available: @code{+}, @code{-},
459 The following functions are available:
481 The following constants are available:
510 @settitle FFmpeg video converter
513 ffserver(1), ffplay(1) and the html documentation of @file{ffmpeg}.
524 The filename can be @file{-} to read from the standard input or to write
525 to the standard output.
527 ffmpeg handles also many protocols specified with the URL syntax.
529 Use 'ffmpeg -formats' to have a list of the supported protocols.
531 The protocol @code{http:} is currently used only to communicate with
532 ffserver (see the ffserver documentation). When ffmpeg will be a
533 video player it will also be used for streaming :-)
538 @item For streaming at very low bit rate application, use a low frame rate
539 and a small gop size. This is especially true for real video where
540 the Linux player does not seem to be very fast, so it can miss
541 frames. An example is:
544 ffmpeg -g 3 -r 3 -t 10 -b 50 -s qcif -f rv10 /tmp/b.rm
547 @item The parameter 'q' which is displayed while encoding is the current
548 quantizer. The value of 1 indicates that a very good quality could
549 be achieved. The value of 31 indicates the worst quality. If q=31
550 too often, it means that the encoder cannot compress enough to meet
551 your bit rate. You must either increase the bit rate, decrease the
552 frame rate or decrease the frame size.
554 @item If your computer is not fast enough, you can speed up the
555 compression at the expense of the compression ratio. You can use
556 '-me zero' to speed up motion estimation, and '-intra' to disable
557 completely motion estimation (you have only I frames, which means it
558 is about as good as JPEG compression).
560 @item To have very low bitrates in audio, reduce the sampling frequency
561 (down to 22050 kHz for mpeg audio, 22050 or 11025 for ac3).
563 @item To have a constant quality (but a variable bitrate), use the option
564 '-qscale n' when 'n' is between 1 (excellent quality) and 31 (worst
567 @item When converting video files, you can use the '-sameq' option which
568 uses in the encoder the same quality factor than in the decoder. It
569 allows to be almost lossless in encoding.
573 @chapter Supported File Formats and Codecs
575 You can use the @code{-formats} option to have an exhaustive list.
577 @section File Formats
579 FFmpeg supports the following file formats through the @code{libavformat}
582 @multitable @columnfractions .4 .1 .1
583 @item Supported File Format @tab Encoding @tab Decoding @tab Comments
584 @item MPEG audio @tab X @tab X
585 @item MPEG1 systems @tab X @tab X
586 @tab muxed audio and video
587 @item MPEG2 PS @tab X @tab X
588 @tab also known as @code{VOB} file
589 @item MPEG2 TS @tab @tab X
590 @tab also known as DVB Transport Stream
591 @item ASF@tab X @tab X
592 @item AVI@tab X @tab X
593 @item WAV@tab X @tab X
594 @item Macromedia Flash@tab X @tab X
595 @tab Only embedded audio is decoded
596 @item FLV @tab X @tab X
597 @tab Macromedia Flash video files
598 @item Real Audio and Video @tab X @tab X
599 @item Raw AC3 @tab X @tab X
600 @item Raw MJPEG @tab X @tab X
601 @item Raw MPEG video @tab X @tab X
602 @item Raw PCM8/16 bits, mulaw/Alaw@tab X @tab X
603 @item SUN AU format @tab X @tab X
604 @item Quicktime @tab @tab X
605 @item MPEG4 @tab @tab X
606 @tab MPEG4 is a variant of Quicktime
607 @item Raw MPEG4 video @tab X @tab X
609 @item 4xm @tab @tab X
610 @tab 4X Technologies format, used in some games
613 @code{X} means that the encoding (resp. decoding) is supported.
615 @section Image Formats
617 FFmpeg can read and write images for each frame of a video sequence. The
618 following image formats are supported:
620 @multitable @columnfractions .4 .1 .1
621 @item Supported Image Format @tab Encoding @tab Decoding @tab Comments
622 @item PGM, PPM @tab X @tab X
623 @item PAM @tab X @tab X @tab PAM is a PNM extension with alpha support
624 @item PGMYUV @tab X @tab X @tab PGM with U and V components in YUV 4:2:0
625 @item JPEG @tab X @tab X @tab Progressive JPEG is not supported
626 @item .Y.U.V @tab X @tab X @tab One raw file per component
627 @item Animated GIF @tab X @tab X @tab Only uncompressed GIFs are generated
628 @item PNG @tab X @tab X @tab 2 bit and 4 bit/pixel not supported yet
631 @code{X} means that the encoding (resp. decoding) is supported.
633 @section Video Codecs
635 @multitable @columnfractions .4 .1 .1 .7
636 @item Supported Codec @tab Encoding @tab Decoding @tab Comments
637 @item MPEG1 video @tab X @tab X
638 @item MPEG2 video @tab @tab X
639 @item MPEG4 @tab X @tab X @tab Also known as DIVX4/5
640 @item MSMPEG4 V1 @tab X @tab X
641 @item MSMPEG4 V2 @tab X @tab X
642 @item MSMPEG4 V3 @tab X @tab X @tab Also known as DIVX3
643 @item WMV7 @tab X @tab X
644 @item WMV8 @tab X @tab X @tab Not completely working
645 @item H263(+) @tab X @tab X @tab Also known as Real Video 1.0
646 @item MJPEG @tab X @tab X
648 @item Huff YUV @tab X @tab X
649 @item Asus v1 @tab X @tab X @tab fourcc: ASV1
650 @item Creative YUV @tab @tab X @tab fourcc: CYUV
651 @item H.264 @tab @tab X
652 @item Sorenson Video 1 @tab @tab X @tab fourcc: SVQ1
653 @item Sorenson Video 3 @tab @tab X @tab fourcc: SVQ3
654 @item On2 VP3 @tab @tab X @tab still experimental
655 @item Intel Indeo 3 @tab @tab X @tab only works on i386 right now
656 @item FLV @tab X @tab X @tab Flash H263 variant
659 @code{X} means that the encoding (resp. decoding) is supported.
661 Check at @url{http://www.mplayerhq.hu/~michael/codec-features.html} to
662 get a precise comparison of FFmpeg MPEG4 codec compared to the other
665 @section Audio Codecs
667 @multitable @columnfractions .4 .1 .1 .1 .7
668 @item Supported Codec @tab Encoding @tab Decoding @tab Comments
669 @item MPEG audio layer 2 @tab IX @tab IX
670 @item MPEG audio layer 1/3 @tab IX @tab IX
671 @tab MP3 encoding is supported through the external library LAME
672 @item AC3 @tab IX @tab X
673 @tab liba52 is used internally for decoding
674 @item Vorbis @tab X @tab X
675 @tab supported through the external library libvorbis
676 @item WMA V1/V2 @tab @tab X
677 @item Microsoft ADPCM @tab X @tab X
678 @item IMA ADPCM @tab X @tab X
679 @item RA144 @tab @tab X
680 @tab Real 14400 bit/s codec
681 @item RA288 @tab @tab X
682 @tab Real 28800 bit/s codec
683 @item AMR-NB @tab X @tab X
684 @tab supported through an external library
685 @item DV audio @tab @tab X
688 @code{X} means that the encoding (resp. decoding) is supported.
690 @code{I} means that an integer only version is available too (ensures highest
691 performances on systems without hardware floating point support).
693 @chapter Platform Specific information
697 ffmpeg should be compiled with at least GCC 2.95.3. GCC 3.2 is the
698 preferred compiler now for ffmpeg. All future optimizations will depend on
699 features only found in GCC 3.2.
705 @subsection Native Windows compilation
708 @item Install the current versions of MSYS and MinGW from
709 @url{http://www.mingw.org/}. You can find detailed installation
710 instructions in the download section and the FAQ.
712 @item Extract the current version of ffmpeg (the latest release version or the current CVS snapshot whichever is recommended).
714 @item Start the MSYS shell (file @file{msys.bat}).
716 @item Change to the FFMPEG directory and create the libavcodec.a and libavformat.a libraries, following the instructions of how to compile ffmpeg (file @file{INSTALL}).
720 @subsection How to create a single Windows DLL from libavcodec and libavformat
722 Read @url{http://ffmpeg.org/win32.html}.
724 @subsection Cross compilation for Windows with Linux
726 You must use the MinGW cross compilation tools available at
727 @url{http://www.mingw.org/}.
729 Then configure ffmpeg with the following options:
731 ./configure --enable-mingw32 --cross-prefix=i386-mingw32msvc-
733 (you can change the cross-prefix according to the prefix choosen for the
736 Then you can easily test ffmpeg with wine
737 (@url{http://www.winehq.com/}).
743 The configure script should guess the configuration itself.
744 Networking support is currently not finished.
745 errno issues fixed by Andrew Bachmann.
749 François Revol - revol at free dot fr - April 2002
751 The configure script should guess the configuration itself,
752 however I still didn't tested building on net_server version of BeOS.
754 ffserver is broken (needs poll() implementation).
756 There is still issues with errno codes, which are negative in BeOs, and
757 that ffmpeg negates when returning. This ends up turning errors into
758 valid results, then crashes.
761 @chapter Developers Guide
765 @item libavcodec is the library containing the codecs (both encoding and
766 decoding). See @file{libavcodec/apiexample.c} to see how to use it.
768 @item libavformat is the library containing the file formats handling (mux and
769 demux code for several formats). See @file{ffplay.c} to use it in a
770 player. See @file{output_example.c} to use it to generate audio or video
775 @section Integrating libavcodec or libavformat in your program
777 You can integrate all the source code of the libraries to link them
778 statically to avoid any version problem. All you need is to provide a
779 'config.mak' and a 'config.h' in the parent directory. See the defines
780 generated by ./configure to understand what is needed.
782 You can use libavcodec or libavformat in your commercial program, but
783 @emph{any patch you make must be published}. The best way to proceed is
784 to send your patches to the ffmpeg mailing list.
786 @section Coding Rules
788 ffmpeg is programmed in ANSI C language. GCC extensions are
789 tolerated. Indent size is 4. The TAB character should not be used.
791 The presentation is the one specified by 'indent -i4 -kr'.
793 Main priority in ffmpeg is simplicity and small code size (=less
796 Comments: for functions visible from other modules, use the JavaDoc
797 format (see examples in @file{libav/utils.c}) so that a documentation
798 can be generated automatically.
800 @section Submitting patches
802 When you submit your patch, try to send a unified diff (diff '-u'
803 option). I cannot read other diffs :-)
805 Run the regression tests before submitting a patch so that you can
806 verify that there are no big problems.
808 Patches should be posted as base64 encoded attachments (or any other
809 encoding which ensures that the patch wont be trashed during
810 transmission) to the ffmpeg-devel mailinglist, see
811 @url{http://lists.sourceforge.net/lists/listinfo/ffmpeg-devel}
813 @section Regression tests
815 Before submitting a patch (or committing with CVS), you should at least
816 test that you did not break anything.
818 The regression test build a synthetic video stream and a synthetic
819 audio stream. Then these are encoded then decoded with all codecs or
820 formats. The CRC (or MD5) of each generated file is recorded in a
821 result file. Then a 'diff' is launched with the reference results and
824 The regression test then goes on to test the ffserver code with a
825 limited set of streams. It is important that this step runs correctly
828 Run 'make test' to test all the codecs.
830 Run 'make libavtest' to test all the codecs.
832 [Of course, some patches may change the regression tests results. In
833 this case, the regression tests reference results shall be modified