1 \input texinfo @c -*- texinfo -*-
2 @documentencoding UTF-8
6 @center @titlefont{FFmpeg FAQ}
13 @chapter General Questions
15 @section Why doesn't FFmpeg support feature [xyz]?
17 Because no one has taken on that task yet. FFmpeg development is
18 driven by the tasks that are important to the individual developers.
19 If there is a feature that is important to you, the best way to get
20 it implemented is to undertake the task yourself or sponsor a developer.
22 @section FFmpeg does not support codec XXX. Can you include a Windows DLL loader to support it?
24 No. Windows DLLs are not portable, bloated and often slow.
25 Moreover FFmpeg strives to support all codecs natively.
26 A DLL loader is not conducive to that goal.
28 @section I cannot read this file although this format seems to be supported by ffmpeg.
30 Even if ffmpeg can read the container format, it may not support all its
31 codecs. Please consult the supported codec list in the ffmpeg
34 @section Which codecs are supported by Windows?
36 Windows does not support standard formats like MPEG very well, unless you
37 install some additional codecs.
39 The following list of video codecs should work on most Windows systems:
50 Only if you have some MPEG-4 codec like ffdshow or Xvid installed.
54 Note, ASF files often have .wmv or .wma extensions in Windows. It should also
55 be mentioned that Microsoft claims a patent on the ASF format, and may sue
56 or threaten users who create ASF files with non-Microsoft software. It is
57 strongly advised to avoid ASF where possible.
59 The following list of audio codecs should work on most Windows systems:
66 If some MP3 codec like LAME is installed.
72 @section @code{error: can't find a register in class 'GENERAL_REGS' while reloading 'asm'}
74 This is a bug in gcc. Do not report it to us. Instead, please report it to
75 the gcc developers. Note that we will not add workarounds for gcc bugs.
77 Also note that (some of) the gcc developers believe this is not a bug or
78 not a bug they should fix:
79 @url{http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11203}.
80 Then again, some of them do not know the difference between an undecidable
81 problem and an NP-hard problem...
83 @section I have installed this library with my distro's package manager. Why does @command{configure} not see it?
85 Distributions usually split libraries in several packages. The main package
86 contains the files necessary to run programs using the library. The
87 development package contains the files necessary to build programs using the
88 library. Sometimes, docs and/or data are in a separate package too.
90 To build FFmpeg, you need to install the development package. It is usually
91 called @file{libfoo-dev} or @file{libfoo-devel}. You can remove it after the
92 build is finished, but be sure to keep the main package.
94 @section How do I make @command{pkg-config} find my libraries?
96 Somewhere along with your libraries, there is a @file{.pc} file (or several)
97 in a @file{pkgconfig} directory. You need to set environment variables to
98 point @command{pkg-config} to these files.
100 If you need to @emph{add} directories to @command{pkg-config}'s search list
101 (typical use case: library installed separately), add it to
102 @code{$PKG_CONFIG_PATH}:
105 export PKG_CONFIG_PATH=/opt/x264/lib/pkgconfig:/opt/opus/lib/pkgconfig
108 If you need to @emph{replace} @command{pkg-config}'s search list
109 (typical use case: cross-compiling), set it in
110 @code{$PKG_CONFIG_LIBDIR}:
113 export PKG_CONFIG_LIBDIR=/home/me/cross/usr/lib/pkgconfig:/home/me/cross/usr/local/lib/pkgconfig
116 If you need to know the library's internal dependencies (typical use: static
117 linking), add the @code{--static} option to @command{pkg-config}:
120 ./configure --pkg-config-flags=--static
123 @section How do I use @command{pkg-config} when cross-compiling?
125 The best way is to install @command{pkg-config} in your cross-compilation
126 environment. It will automatically use the cross-compilation libraries.
128 You can also use @command{pkg-config} from the host environment by
129 specifying explicitly @code{--pkg-config=pkg-config} to @command{configure}.
130 In that case, you must point @command{pkg-config} to the correct directories
131 using the @code{PKG_CONFIG_LIBDIR}, as explained in the previous entry.
133 As an intermediate solution, you can place in your cross-compilation
134 environment a script that calls the host @command{pkg-config} with
135 @code{PKG_CONFIG_LIBDIR} set. That script can look like that:
139 PKG_CONFIG_LIBDIR=/path/to/cross/lib/pkgconfig
140 export PKG_CONFIG_LIBDIR
141 exec /usr/bin/pkg-config "$@@"
146 @section ffmpeg does not work; what is wrong?
148 Try a @code{make distclean} in the ffmpeg source directory before the build.
149 If this does not help see
150 (@url{https://ffmpeg.org/bugreports.html}).
152 @section How do I encode single pictures into movies?
154 First, rename your pictures to follow a numerical sequence.
155 For example, img1.jpg, img2.jpg, img3.jpg,...
159 ffmpeg -f image2 -i img%d.jpg /tmp/a.mpg
162 Notice that @samp{%d} is replaced by the image number.
164 @file{img%03d.jpg} means the sequence @file{img001.jpg}, @file{img002.jpg}, etc.
166 Use the @option{-start_number} option to declare a starting number for
167 the sequence. This is useful if your sequence does not start with
168 @file{img001.jpg} but is still in a numerical order. The following
169 example will start with @file{img100.jpg}:
172 ffmpeg -f image2 -start_number 100 -i img%d.jpg /tmp/a.mpg
175 If you have large number of pictures to rename, you can use the
176 following command to ease the burden. The command, using the bourne
177 shell syntax, symbolically links all files in the current directory
178 that match @code{*jpg} to the @file{/tmp} directory in the sequence of
179 @file{img001.jpg}, @file{img002.jpg} and so on.
182 x=1; for i in *jpg; do counter=$(printf %03d $x); ln -s "$i" /tmp/img"$counter".jpg; x=$(($x+1)); done
185 If you want to sequence them by oldest modified first, substitute
186 @code{$(ls -r -t *jpg)} in place of @code{*jpg}.
191 ffmpeg -f image2 -i /tmp/img%03d.jpg /tmp/a.mpg
194 The same logic is used for any image format that ffmpeg reads.
196 You can also use @command{cat} to pipe images to ffmpeg:
199 cat *.jpg | ffmpeg -f image2pipe -c:v mjpeg -i - output.mpg
202 @section How do I encode movie to single pictures?
207 ffmpeg -i movie.mpg movie%d.jpg
210 The @file{movie.mpg} used as input will be converted to
211 @file{movie1.jpg}, @file{movie2.jpg}, etc...
213 Instead of relying on file format self-recognition, you may also use
219 to force the encoding.
221 Applying that to the previous example:
223 ffmpeg -i movie.mpg -f image2 -c:v mjpeg menu%d.jpg
226 Beware that there is no "jpeg" codec. Use "mjpeg" instead.
228 @section Why do I see a slight quality degradation with multithreaded MPEG* encoding?
230 For multithreaded MPEG* encoding, the encoded slices must be independent,
231 otherwise thread n would practically have to wait for n-1 to finish, so it's
232 quite logical that there is a small reduction of quality. This is not a bug.
234 @section How can I read from the standard input or write to the standard output?
236 Use @file{-} as file name.
238 @section -f jpeg doesn't work.
240 Try '-f image2 test%d.jpg'.
242 @section Why can I not change the frame rate?
244 Some codecs, like MPEG-1/2, only allow a small number of fixed frame rates.
245 Choose a different codec with the -c:v command line option.
247 @section How do I encode Xvid or DivX video with ffmpeg?
249 Both Xvid and DivX (version 4+) are implementations of the ISO MPEG-4
250 standard (note that there are many other coding formats that use this
251 same standard). Thus, use '-c:v mpeg4' to encode in these formats. The
252 default fourcc stored in an MPEG-4-coded file will be 'FMP4'. If you want
253 a different fourcc, use the '-vtag' option. E.g., '-vtag xvid' will
254 force the fourcc 'xvid' to be stored as the video fourcc rather than the
257 @section Which are good parameters for encoding high quality MPEG-4?
259 '-mbd rd -flags +mv4+aic -trellis 2 -cmp 2 -subcmp 2 -g 300 -pass 1/2',
260 things to try: '-bf 2', '-flags qprd', '-flags mv0', '-flags skiprd'.
262 @section Which are good parameters for encoding high quality MPEG-1/MPEG-2?
264 '-mbd rd -trellis 2 -cmp 2 -subcmp 2 -g 100 -pass 1/2'
265 but beware the '-g 100' might cause problems with some decoders.
266 Things to try: '-bf 2', '-flags qprd', '-flags mv0', '-flags skiprd.
268 @section Interlaced video looks very bad when encoded with ffmpeg, what is wrong?
270 You should use '-flags +ilme+ildct' and maybe '-flags +alt' for interlaced
271 material, and try '-top 0/1' if the result looks really messed-up.
273 @section How can I read DirectShow files?
275 If you have built FFmpeg with @code{./configure --enable-avisynth}
276 (only possible on MinGW/Cygwin platforms),
277 then you may use any file that DirectShow can read as input.
279 Just create an "input.avs" text file with this single line ...
281 DirectShowSource("C:\path to your file\yourfile.asf")
283 ... and then feed that text file to ffmpeg:
288 For ANY other help on AviSynth, please visit the
289 @uref{http://www.avisynth.org/, AviSynth homepage}.
291 @section How can I join video files?
293 To "join" video files is quite ambiguous. The following list explains the
294 different kinds of "joining" and points out how those are addressed in
295 FFmpeg. To join video files may mean:
300 To put them one after the other: this is called to @emph{concatenate} them
301 (in short: concat) and is addressed
302 @ref{How can I concatenate video files, in this very faq}.
305 To put them together in the same file, to let the user choose between the
306 different versions (example: different audio languages): this is called to
307 @emph{multiplex} them together (in short: mux), and is done by simply
308 invoking ffmpeg with several @option{-i} options.
311 For audio, to put all channels together in a single stream (example: two
312 mono streams into one stereo stream): this is sometimes called to
313 @emph{merge} them, and can be done using the
314 @url{ffmpeg-filters.html#amerge, @code{amerge}} filter.
317 For audio, to play one on top of the other: this is called to @emph{mix}
318 them, and can be done by first merging them into a single stream and then
319 using the @url{ffmpeg-filters.html#pan, @code{pan}} filter to mix
320 the channels at will.
323 For video, to display both together, side by side or one on top of a part of
324 the other; it can be done using the
325 @url{ffmpeg-filters.html#overlay, @code{overlay}} video filter.
329 @anchor{How can I concatenate video files}
330 @section How can I concatenate video files?
332 There are several solutions, depending on the exact circumstances.
334 @subsection Concatenating using the concat @emph{filter}
336 FFmpeg has a @url{ffmpeg-filters.html#concat,
337 @code{concat}} filter designed specifically for that, with examples in the
338 documentation. This operation is recommended if you need to re-encode.
340 @subsection Concatenating using the concat @emph{demuxer}
342 FFmpeg has a @url{ffmpeg-formats.html#concat,
343 @code{concat}} demuxer which you can use when you want to avoid a re-encode and
344 your format doesn't support file level concatenation.
346 @subsection Concatenating using the concat @emph{protocol} (file level)
348 FFmpeg has a @url{ffmpeg-protocols.html#concat,
349 @code{concat}} protocol designed specifically for that, with examples in the
352 A few multimedia containers (MPEG-1, MPEG-2 PS, DV) allow one to concatenate
353 video by merely concatenating the files containing them.
355 Hence you may concatenate your multimedia files by first transcoding them to
356 these privileged formats, then using the humble @code{cat} command (or the
357 equally humble @code{copy} under Windows), and finally transcoding back to your
361 ffmpeg -i input1.avi -qscale:v 1 intermediate1.mpg
362 ffmpeg -i input2.avi -qscale:v 1 intermediate2.mpg
363 cat intermediate1.mpg intermediate2.mpg > intermediate_all.mpg
364 ffmpeg -i intermediate_all.mpg -qscale:v 2 output.avi
367 Additionally, you can use the @code{concat} protocol instead of @code{cat} or
368 @code{copy} which will avoid creation of a potentially huge intermediate file.
371 ffmpeg -i input1.avi -qscale:v 1 intermediate1.mpg
372 ffmpeg -i input2.avi -qscale:v 1 intermediate2.mpg
373 ffmpeg -i concat:"intermediate1.mpg|intermediate2.mpg" -c copy intermediate_all.mpg
374 ffmpeg -i intermediate_all.mpg -qscale:v 2 output.avi
377 Note that you may need to escape the character "|" which is special for many
380 Another option is usage of named pipes, should your platform support it:
383 mkfifo intermediate1.mpg
384 mkfifo intermediate2.mpg
385 ffmpeg -i input1.avi -qscale:v 1 -y intermediate1.mpg < /dev/null &
386 ffmpeg -i input2.avi -qscale:v 1 -y intermediate2.mpg < /dev/null &
387 cat intermediate1.mpg intermediate2.mpg |\
388 ffmpeg -f mpeg -i - -c:v mpeg4 -c:a libmp3lame output.avi
391 @subsection Concatenating using raw audio and video
393 Similarly, the yuv4mpegpipe format, and the raw video, raw audio codecs also
394 allow concatenation, and the transcoding step is almost lossless.
395 When using multiple yuv4mpegpipe(s), the first line needs to be discarded
396 from all but the first stream. This can be accomplished by piping through
397 @code{tail} as seen below. Note that when piping through @code{tail} you
398 must use command grouping, @code{@{ ;@}}, to background properly.
400 For example, let's say we want to concatenate two FLV files into an
410 ffmpeg -i input1.flv -vn -f u16le -c:a pcm_s16le -ac 2 -ar 44100 - > temp1.a < /dev/null &
411 ffmpeg -i input2.flv -vn -f u16le -c:a pcm_s16le -ac 2 -ar 44100 - > temp2.a < /dev/null &
412 ffmpeg -i input1.flv -an -f yuv4mpegpipe - > temp1.v < /dev/null &
413 @{ ffmpeg -i input2.flv -an -f yuv4mpegpipe - < /dev/null | tail -n +2 > temp2.v ; @} &
414 cat temp1.a temp2.a > all.a &
415 cat temp1.v temp2.v > all.v &
416 ffmpeg -f u16le -c:a pcm_s16le -ac 2 -ar 44100 -i all.a \
417 -f yuv4mpegpipe -i all.v \
419 rm temp[12].[av] all.[av]
422 @section Using @option{-f lavfi}, audio becomes mono for no apparent reason.
424 Use @option{-dumpgraph -} to find out exactly where the channel layout is
427 Most likely, it is through @code{auto-inserted aresample}. Try to understand
428 why the converting filter was needed at that place.
430 Just before the output is a likely place, as @option{-f lavfi} currently
431 only support packed S16.
433 Then insert the correct @code{aformat} explicitly in the filtergraph,
434 specifying the exact format.
437 aformat=sample_fmts=s16:channel_layouts=stereo
440 @section Why does FFmpeg not see the subtitles in my VOB file?
442 VOB and a few other formats do not have a global header that describes
443 everything present in the file. Instead, applications are supposed to scan
444 the file to see what it contains. Since VOB files are frequently large, only
445 the beginning is scanned. If the subtitles happen only later in the file,
446 they will not be initially detected.
448 Some applications, including the @code{ffmpeg} command-line tool, can only
449 work with streams that were detected during the initial scan; streams that
450 are detected later are ignored.
452 The size of the initial scan is controlled by two options: @code{probesize}
453 (default ~5 Mo) and @code{analyzeduration} (default 5,000,000 µs = 5 s). For
454 the subtitle stream to be detected, both values must be large enough.
456 @section Why was the @command{ffmpeg} @option{-sameq} option removed? What to use instead?
458 The @option{-sameq} option meant "same quantizer", and made sense only in a
459 very limited set of cases. Unfortunately, a lot of people mistook it for
460 "same quality" and used it in places where it did not make sense: it had
461 roughly the expected visible effect, but achieved it in a very inefficient
464 Each encoder has its own set of options to set the quality-vs-size balance,
465 use the options for the encoder you are using to set the quality level to a
466 point acceptable for your tastes. The most common options to do that are
467 @option{-qscale} and @option{-qmax}, but you should peruse the documentation
468 of the encoder you chose.
470 @section I have a stretched video, why does scaling does not fix it?
472 A lot of video codecs and formats can store the @emph{aspect ratio} of the
473 video: this is the ratio between the width and the height of either the full
474 image (DAR, display aspect ratio) or individual pixels (SAR, sample aspect
475 ratio). For example, EGA screens at resolution 640×350 had 4:3 DAR and 35:48
478 Most still image processing work with square pixels, i.e. 1:1 SAR, but a lot
479 of video standards, especially from the analogic-numeric transition era, use
482 Most processing filters in FFmpeg handle the aspect ratio to avoid
483 stretching the image: cropping adjusts the DAR to keep the SAR constant,
484 scaling adjusts the SAR to keep the DAR constant.
486 If you want to stretch, or “unstretch”, the image, you need to override the
488 @url{ffmpeg-filters.html#setdar_002c-setsar, @code{setdar or setsar filters}}.
490 Do not forget to examine carefully the original video to check whether the
491 stretching comes from the image or from the aspect ratio information.
493 For example, to fix a badly encoded EGA capture, use the following commands,
494 either the first one to upscale to square pixels or the second one to set
495 the correct aspect ratio or the third one to avoid transcoding (may not work
496 depending on the format / codec / player / phase of the moon):
499 ffmpeg -i ega_screen.nut -vf scale=640:480,setsar=1 ega_screen_scaled.nut
500 ffmpeg -i ega_screen.nut -vf setdar=4/3 ega_screen_anamorphic.nut
501 ffmpeg -i ega_screen.nut -aspect 4/3 -c copy ega_screen_overridden.nut
504 @anchor{background task}
505 @section How do I run ffmpeg as a background task?
507 ffmpeg normally checks the console input, for entries like "q" to stop
508 and "?" to give help, while performing operations. ffmpeg does not have a way of
509 detecting when it is running as a background task.
510 When it checks the console input, that can cause the process running ffmpeg
511 in the background to suspend.
513 To prevent those input checks, allowing ffmpeg to run as a background task,
514 use the @url{ffmpeg.html#stdin-option, @code{-nostdin} option}
515 in the ffmpeg invocation. This is effective whether you run ffmpeg in a shell
516 or invoke ffmpeg in its own process via an operating system API.
518 As an alternative, when you are running ffmpeg in a shell, you can redirect
519 standard input to @code{/dev/null} (on Linux and Mac OS)
520 or @code{NUL} (on Windows). You can do this redirect either
521 on the ffmpeg invocation, or from a shell script which calls ffmpeg.
526 ffmpeg -nostdin -i INPUT OUTPUT
529 or (on Linux, Mac OS, and other UNIX-like shells):
532 ffmpeg -i INPUT OUTPUT </dev/null
538 ffmpeg -i INPUT OUTPUT <NUL
541 @section How do I prevent ffmpeg from suspending with a message like @emph{suspended (tty output)}?
543 If you run ffmpeg in the background, you may find that its process suspends.
544 There may be a message like @emph{suspended (tty output)}. The question is how
545 to prevent the process from being suspended.
550 % ffmpeg -i INPUT OUTPUT &> ~/tmp/log.txt &
553 [1] + suspended (tty output) ffmpeg -i INPUT OUTPUT &>
556 The message "tty output" notwithstanding, the problem here is that
557 ffmpeg normally checks the console input when it runs. The operating system
558 detects this, and suspends the process until you can bring it to the
559 foreground and attend to it.
561 The solution is to use the right techniques to tell ffmpeg not to consult
562 console input. You can use the
563 @url{ffmpeg.html#stdin-option, @code{-nostdin} option},
564 or redirect standard input with @code{< /dev/null}.
566 @ref{background task, @emph{How do I run ffmpeg as a background task?}}
571 @section Are there examples illustrating how to use the FFmpeg libraries, particularly libavcodec and libavformat?
573 Yes. Check the @file{doc/examples} directory in the source
574 repository, also available online at:
575 @url{https://github.com/FFmpeg/FFmpeg/tree/master/doc/examples}.
577 Examples are also installed by default, usually in
578 @code{$PREFIX/share/ffmpeg/examples}.
580 Also you may read the Developers Guide of the FFmpeg documentation. Alternatively,
581 examine the source code for one of the many open source projects that
582 already incorporate FFmpeg at (@url{projects.html}).
584 @section Can you support my C compiler XXX?
586 It depends. If your compiler is C99-compliant, then patches to support
587 it are likely to be welcome if they do not pollute the source code
588 with @code{#ifdef}s related to the compiler.
590 @section Is Microsoft Visual C++ supported?
592 Yes. Please see the @uref{platform.html, Microsoft Visual C++}
593 section in the FFmpeg documentation.
595 @section Can you add automake, libtool or autoconf support?
597 No. These tools are too bloated and they complicate the build.
599 @section Why not rewrite FFmpeg in object-oriented C++?
601 FFmpeg is already organized in a highly modular manner and does not need to
602 be rewritten in a formal object language. Further, many of the developers
603 favor straight C; it works for them. For more arguments on this matter,
604 read @uref{http://www.tux.org/lkml/#s15, "Programming Religion"}.
606 @section Why are the ffmpeg programs devoid of debugging symbols?
608 The build process creates @command{ffmpeg_g}, @command{ffplay_g}, etc. which
609 contain full debug information. Those binaries are stripped to create
610 @command{ffmpeg}, @command{ffplay}, etc. If you need the debug information, use
613 @section I do not like the LGPL, can I contribute code under the GPL instead?
615 Yes, as long as the code is optional and can easily and cleanly be placed
616 under #if CONFIG_GPL without breaking anything. So, for example, a new codec
617 or filter would be OK under GPL while a bug fix to LGPL code would not.
619 @section I'm using FFmpeg from within my C application but the linker complains about missing symbols from the libraries themselves.
621 FFmpeg builds static libraries by default. In static libraries, dependencies
622 are not handled. That has two consequences. First, you must specify the
623 libraries in dependency order: @code{-lavdevice} must come before
624 @code{-lavformat}, @code{-lavutil} must come after everything else, etc.
625 Second, external libraries that are used in FFmpeg have to be specified too.
627 An easy way to get the full list of required libraries in dependency order
628 is to use @code{pkg-config}.
631 c99 -o program program.c $(pkg-config --cflags --libs libavformat libavcodec)
634 See @file{doc/example/Makefile} and @file{doc/example/pc-uninstalled} for
637 @section I'm using FFmpeg from within my C++ application but the linker complains about missing symbols which seem to be available.
639 FFmpeg is a pure C project, so to use the libraries within your C++ application
640 you need to explicitly state that you are using a C library. You can do this by
641 encompassing your FFmpeg includes using @code{extern "C"}.
643 See @url{http://www.parashift.com/c++-faq-lite/mixing-c-and-cpp.html#faq-32.3}
645 @section I'm using libavutil from within my C++ application but the compiler complains about 'UINT64_C' was not declared in this scope
647 FFmpeg is a pure C project using C99 math features, in order to enable C++
648 to use them you have to append -D__STDC_CONSTANT_MACROS to your CXXFLAGS
650 @section I have a file in memory / a API different from *open/*read/ libc how do I use it with libavformat?
652 You have to create a custom AVIOContext using @code{avio_alloc_context},
653 see @file{libavformat/aviobuf.c} in FFmpeg and @file{libmpdemux/demux_lavf.c} in MPlayer or MPlayer2 sources.
655 @section Where is the documentation about ffv1, msmpeg4, asv1, 4xm?
657 see @url{https://www.ffmpeg.org/~michael/}
659 @section How do I feed H.263-RTP (and other codecs in RTP) to libavcodec?
661 Even if peculiar since it is network oriented, RTP is a container like any
662 other. You have to @emph{demux} RTP before feeding the payload to libavcodec.
663 In this specific case please look at RFC 4629 to see how it should be done.
665 @section AVStream.r_frame_rate is wrong, it is much larger than the frame rate.
667 @code{r_frame_rate} is NOT the average frame rate, it is the smallest frame rate
668 that can accurately represent all timestamps. So no, it is not
669 wrong if it is larger than the average!
670 For example, if you have mixed 25 and 30 fps content, then @code{r_frame_rate}
671 will be 150 (it is the least common multiple).
672 If you are looking for the average frame rate, see @code{AVStream.avg_frame_rate}.
674 @section Why is @code{make fate} not running all tests?
676 Make sure you have the fate-suite samples and the @code{SAMPLES} Make variable
677 or @code{FATE_SAMPLES} environment variable or the @code{--samples}
678 @command{configure} option is set to the right path.
680 @section Why is @code{make fate} not finding the samples?
682 Do you happen to have a @code{~} character in the samples path to indicate a
683 home directory? The value is used in ways where the shell cannot expand it,
684 causing FATE to not find files. Just replace @code{~} by the full path.