June 2005 - ffmpeg-user - lists.ffmpeg.org

Setting a Video Device with Windows
by jdgiotta＠gmail.com 30 Jun '05

30 Jun '05

Hello, I'm working with FFMPEG on a Windows XP Pro environment and I would like to use FFMPEG for live capture/grabbing. I know to set a video device I use '-vd device', but how do I reference a video device on windows? Best, John

1 0

FLV encoding : incorrect codec parameters
by sligo＠free.fr 29 Jun '05

29 Jun '05

-- stephane <sligo at free.fr> if you have trouble encoding a movie in FLV with the following error : "Could not write header for output file #0 (incorrect codec parameters ?)", it's maybe cause by a 48000 audio sampling freq of your source. So you have to force the output sampling to 44100 with the -ar option and... it works fine. OUF!

1 0

Motion Vectors in libavcodec
by hogue＠cs.yorku.ca 29 Jun '05

29 Jun '05

Hi all, I was just wondering how to easily get access to the motion vectors within libavcodec? I noticed that there is a motion_val[] table within each AVFrame but for some reason after I decode the frame, the motion_val table is null. Could anyone provide me with some sample code or explanation as to how to get access to the motion vectors? Thanks for any help. -- Andrew

2 1

av_seek_frame on mpeg file
by georges.legoualher＠etiam.com 29 Jun '05

29 Jun '05

Hi, I'm facing the same issue as the following post I've found in the ffmpeg-devel mailing list (May 20) >According to the doc, using av_seek_frame with the AVSEEK_FLAG_BACKWARD >flag will seek to the last keyframe at the timestamp specified. This >work wery well with avi file, but not with mpeg files. With mpeg, >av_seek_frame return a success code but when I try to call av_read_frame >after, it give me a degraded image. Why? <>M?ns Rullg?rd answered that this was a consequence of the fact that MPEG files don't indicate where the keyframes are. Do you know if a fix can be found to solve this issue? Many thanks, Georges -- Responsable Projet/Project Manager __________________________________________ www.etiam.com ETIAM "Multimedia Connectivity for Healthcare" Technopole Atalante - Villejean 35000 Rennes FRANCE Tel: +33 2 99 14 33 88 Fax: +33 2 99 14 33 80

1 0

Wrong MP4 input size WxH when converting to DVD ??
by zuber＠bigfoot.com.au 28 Jun '05

28 Jun '05

Hi, Ive been using an old version of ffmpeg to convert MP4's from my olympus camera to DVD. Thought I'd try one of the newer releases but I seem to be having a problem with the picture size. The CVS release dated 20050625 gets the video input size wrong. It says the size is 320 x 240 and not 640 x 480 as with the older release. eg. $ ffmpeg -i P1130007.MP4 -target pal-dvd -hq -t 10 copter.mpg ffmpeg version 0.4.9-pre1, build 4756, Copyright (c) 2000-2004 Fabrice Bellard configuration: --enable-gpl --enable-faad --enable-a52 --enable-xvid -- disable-vhook --disable-network --disable-ffserver --disab le-ffplay built on Jun 26 2005 13:52:16, gcc: 3.4.4 (cygming special) (gdc 0.12, using dmd 0.125) Seems that stream 0 comes from film source: 30000.00 (30000/1) -> 29.97 (30000/1001) Input #0, mov,mp4,m4a,3gp,3g2, from 'P1130007.MP4': Duration: 00:02:05.1, start: 0.000000, bitrate: 1373 kb/s Stream #0.0: Video: mpeg4, yuv420p, 320x240, 30000.00 fps Stream #0.1: Audio: aac, 48000 Hz, stereo File 'copter.mpg' already exists. Overwrite ? [y/N] y Output #0, dvd, to 'copter.mpg': Stream #0.0: Video: mpeg2video (hq), yuv420p, 720x576, 25.00 fps, q=2-31, 6000 kb/s Stream #0.1: Audio: ac3, 48000 Hz, stereo, 448 kb/s >From an older version $ ./ffmpeg -i P1130007.MP4 -target pal-dvd -hq foo.mpg ffmpeg version 0.4.9-pre1, build 4752, Copyright (c) 2000-2004 Fabrice Bellard configuration: --enable-gpl --enable-faad --enable-a52 --enable-memalign- hack --disable-ffserver --disable-vhook --extra-ldflag -L/usr/local/lib built on Apr 27 2005 23:45:05, gcc: 3.3.3 (cygwin special) Input #0, mov,mp4,m4a,3gp,3g2, from 'P1130007.MP4': Duration: 00:02:05.1, start: 0.000000, bitrate: 1373 kb/s Stream #0.0: Video: mpeg4, 640x480, 29.97 fps Stream #0.1: Audio: aac, 48000 Hz, stereo Output #0, dvd, to 'foo.mpg': Stream #0.0: Video: mpeg2video (hq), 720x576, 25.00 fps, q=2-31, 6000 kb/s Stream #0.1: Audio: ac3, 48000 Hz, stereo, 448 kb/s I have also tried ffmpeg -s 640x480 -i P1130007.MP4 -target pal-dvd -hq -t 10 copter.mpg But that doesn't work.. Do you have any tips on how I can get the correct size WxH ?? Thanks a see ya John

2 4

QDesign Music 2
by lists.jackson.nic＠gmail.com 28 Jun '05

28 Jun '05

Hi, I have been attempting to convert a quicktime movie which has the audio source encoded as QDesign Music 2. Is there a way i can compile ffmpeg to include support for this codec ? Many Thanks, Nic

2 1

Compiling windows FFMPEG
by mediastream＠gmail.com 28 Jun '05

28 Jun '05

Hi All, having compiling problem with latest CVS of ffmpeg: USE MINGW32, used to work ok. ========================================= $ ./configure --enable-memalign-hack --enable-amr_nb --enable-amr_wb --disable- ffplay --disable-ffserver --enable-mp3lame --enable-xvid --enable-faac --enable -a52 --enable-gpl --enable-faad Install prefix /c/Program Files/FFmpeg Source path /home/Dennis/ffmpeg C compiler gcc make make CPU x86 (generic) Big Endian no inttypes.h yes broken inttypes.h no MMX enabled yes Vector Builtins yes gprof enabled no zlib enabled no libgsm enabled no mp3lame enabled yes libogg enabled no vorbis enabled no theora enabled no faad enabled no faadbin enabled no faac enabled yes xvid enabled yes x264 enabled no a52 support yes a52 dlopened no dts support no pp support no debug symbols yes strip symbols yes optimize yes shared pp no Video hooking no SDL support no Sun medialib support no pthreads support no AMR-NB float support yes AMR-NB fixed support no AMR-WB float support yes AMR-WB IF2 support no network support no License: GPL Creating config.mak and config.h AMR WB FLOAT NOTICE ! Make sure you have downloaded TS26.204 V5.1.0 from http://www.3gpp.org/ftp/Specs/archive/26_series/26.204/26204-510.zip and extracted the source to libavcodec/amrwb_float AMR NB FLOAT NOTICE ! Make sure you have downloaded TS26.104 REL-5 V5.1.0 from http://www.3gpp.org/ftp/Specs/latest/Rel-5/26_series/26104-5??.zip and extracted the source to libavcodec/amr_float and if u try this on an alpha, u may need to change Word32 to int in amr/typedef.h ============ copied AMR source, ran make next: ============ $ make make -C libavcodec all make[1]: Entering directory `/home/Dennis/ffmpeg/libavcodec' gcc -O3 -g -Wall -Wno-switch -DHAVE_AV_CONFIG_H -I.. -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_GNU_SOURCE -c -o faac.o faac.c faac.c:26:18: faac.h: No such file or directory faac.c:29: parse error before "faacEncHandle" faac.c:29: warning: no semicolon at end of struct or union faac.c:30: warning: type defaults to `int' in declaration of `FaacAudioContext' faac.c:30: warning: data definition has no type or storage class faac.c: In function `Faac_encode_init': faac.c:34: `s' undeclared (first use in this function) faac.c:34: (Each undeclared identifier is reported only once faac.c:34: for each function it appears in.) faac.c:35: `faacEncConfigurationPtr' undeclared (first use in this function) faac.c:35: parse error before "faac_cfg" faac.c:42: warning: implicit declaration of function `faacEncOpen' faac.c:47: `faac_cfg' undeclared (first use in this function) faac.c:47: warning: implicit declaration of function `faacEncGetCurrentConfiguration' faac.c:48: `FAAC_CFG_VERSION' undeclared (first use in this function) faac.c:50: warning: implicit declaration of function `faacEncClose' faac.c:55: `LOW' undeclared (first use in this function) faac.c:56: `MPEG4' undeclared (first use in this function) faac.c:61: `FAAC_INPUT_16BIT' undeclared (first use in this function) faac.c:63: warning: implicit declaration of function `faacEncSetConfiguration' faac.c:80: warning: implicit declaration of function `faacEncGetDecoderSpecificInfo' faac.c: In function `Faac_encode_frame': faac.c:93: `s' undeclared (first use in this function) faac.c:96: warning: implicit declaration of function `faacEncEncode' faac.c: In function `Faac_encode_close': faac.c:107: `s' undeclared (first use in this function) make[1]: *** [faac.o] Error 1 make[1]: Leaving directory `/home/Dennis/ffmpeg/libavcodec' make: *** [lib] Error 2 ===================== it works fine without faac and faad though. Thanks.

3 8

Question about sub and idx-files
by hemlig_23＠hotmail.com 28 Jun '05

28 Jun '05

Hello! I am a newbie, so please excuse me if this is a simple question. I have on previous occasions converted a movie from avi to divx and inserted subtitles in .srt format, and it has worked perfectly (resulting in a movie with permanent subtitles) But this time I have subtitles in .sub-format and .idx-format. Is it possible to get subtitles into an divx avi-file with these files, so I can view the movie with subtitles in my external DVD player? Is this even possible? What is the idx-file for? Kind regards Grunde

1 0

full Y to standard luma (0-255 vs. 16-234)
by kris＠wotipka.com 28 Jun '05

28 Jun '05

Greetings all, In a nutshell, I need to process .vob and mp2 video files that are in "full Y" (luminance range is 0-255) to .mp2 with the "standard" luma range of 16-234. I know that this is an odd request but it is what I must do. Now, I know that there are no direct switches for ffmpeg but I have run across a line in /libavcodec/libpostprocess/postprocess_internal.h /// select between full y range (255-0) or standart one (234-16) #define FULL_Y_RANGE 0x8000 // 32768 now, since I am not a programmer, this makes absolutely no sense to me. How does 0x8000 relate to one range or the other. I am *assuming* that this is part of the postprocessing of ffmpeg or at least libavcodec. After I can define this correctly, what is the proper way to call it. I have played around with several things but see no change in the final output. So I edit postprocess_internal.h, "make" postprocess.o > make libavcodec > configure ffmpeg with -pp support and make ffmpeg. I have tried values of 0x0000, 0x4000, 0x10000. None seem to make a difference. I have googled till my googler was sore trying to make heads or tails out of this file. I would gladly RTFM if someone would smack me with one. Why do all of this you ask? I work with a non-for-profit public access television station. More and more volunteers are bringing their programs in on DVD. I have gotten the mpeg data off of the disk and converted the .ac3 to mp2 which is what the hardware needs. Now, I need to fix the colorspace so the whites are not clipping up to 140% of NTSC. thanks kw

1 0

3gp AV Sync
by mediastream＠gmail.com 27 Jun '05

27 Jun '05

Gents, I've been at this for 2-3 months already... I've got following dilemma - WIN XP box, ffmpeg 0.4.9-pre1; input files: AVI - XVID 702x480 29.97 1500kbps with 1ch 44.1 16bit pcm/ and various WMV v.7/v.8/v.9 --- all need to be converted to 3gp for 1 phone model. ffmpeg -async 100 -i %input_file%.avi -bitexact -vcodec mpeg4 -s qcif -b 50 -r 8 -acodec amr_nb -ac 1 -ar 8000 -ab 12 -nr -hq -preme -qmin=3 -pass 1 async100_12_8fps.3gp I tried multiple async values, fps and a/v rate. I noticed that different -ab / -r shift audio track differently, almost none of them hit the spot. It's only .2 - .3 second behind, but it does get annoying. It seems that delay is constant across entire file. I tried -itsoffset -0:00:00.xx and of course it works but it seems to vary from file to file, so How can I tell ffmpeg to maintain AV sync on output 3gp? source file are all fine. Also, there's really no explanation of -async function anywhere.. has anyone links to some more detailed docs regaqrding it? Any help is appreciated guys. Thanks

1 0