ogg theora to ogg vorbis

[crazyelf]

I'm attempting to convert a .ogg file which contains both audio and video to a .ogg with only audio. I've searched high and low. But I can't seem to find anything on this. When searching I seem to get combining ogg theora with a sound file. But I want to remove the sound only, and make it its own .ogg Is this possible?

Any help would be greatly appreciated.

Aaron

[Creative2]

you can find many command line on fuoco tools, see my signature

from terminal

extract video

ffmpeg -i INPUT -an -vcodec copy -y OUTPUT.ogg

extract only audio

ffmpeg -i INPUT -vn -acodec vorbis -ab 128k -y OUTPUT.ogg

*note
it should work for every formats

[crazyelf]

I tried that, but it came out with an error. :(
aaron@saturn:~$ ffmpeg -i dare_you_to_move.ogg -vn -acodec vorbis -ab 128k -y dare_u_2_move.ogg
FFmpeg version SVN-rUNKNOWN, Copyright (c) 2000-2007 Fabrice Bellard, et al.
configuration: --enable-gpl --enable-pp --enable-swscaler --enable-pthreads --enable-libvorbis --enable-libtheora --enable-libogg --enable-libgsm --enable-dc1394 --disable-debug --enable-shared --prefix=/usr
libavutil version: 1d.49.3.0
libavcodec version: 1d.51.38.0
libavformat version: 1d.51.10.0
built on Jun 3 2007 20:59:25, gcc: 4.1.3 20070528 (prerelease) (Ubuntu 4.1.2-9ubuntu2)
[theora @ 0xb7e59ac8]Theora bitstream version 30200
[theora @ 0xb7e59ac8]584 bits left in packet 81
[theora @ 0xb7e59ac8]7 bits left in packet 82
Input #0, ogg, from 'dare_you_to_move.ogg':
Duration: 00:04:00.2, start: 0.458594, bitrate: 422 kb/s
Stream #0.0: Video: theora, yuv420p, 320x240, 30.00 fps(r)
Stream #0.1: Audio: vorbis, 22050 Hz, mono, 30 kb/s
Output #0, ogg, to 'dare_u_2_move.ogg':
Stream #0.0: Audio: vorbis, 22050 Hz, mono, 128 kb/s
Stream mapping:
Stream #0.1 -> #0.0
[vorbis @ 0xb7e59ac8]Current FFmpeg Vorbis encoder only supports 2 channels.
Error while opening codec for output stream #0.0 - maybe incorrect parameters such as bit_rate, rate, width or height


I tried changing -ab 128k to -ab 30k like the input file... But, no luck with that either...

Thanks,

Aaron

[Creative2]

FFmpeg version SVN-rUNKNOWN, Copyright (c) 2000-2007 Fabrice Bellard, et al.
configuration: --enable-gpl --enable-pp --enable-swscaler --enable-pthreads
--enable-libvorbis --enable-libtheora

you have compiled your ffmpeg , or you have installed ffmpeg that supports only free formats,
from terminal

extract video

ffmpeg -i INPUT -an -vcodec copy -y OUTPUT.ogg

extract only audio

ffmpeg -i INPUT -vn -acodec copy OUTPUT.ogg

or if you want change bitrate , you must do


ffmpeg -i INPUT -vn -acodec libvorbis -ab 128k -ac 2 -y OUTPUT.ogg

if doesn't work plz install ffmpeg form medibuntu repo you can find usefull link on my signature

[crazyelf]

Thanks a lot! You've been a great help.

Thanks again,

Aaron

[crazyelf]

Actually... maybe not solved... it says libvorbis can't be found. I went to their website. Compiled it from source, because I couldn't find it as a package. But... It still didn't work... any ideas?

Thanks,

Aaron

[Creative2]

you have to learn about medibuntu repository ...it's easy and the fastest way to have mencoder and ffmpeg with support for every formats...

anyway to find out easly encoder you ccan do this :


ffmpeg -formats 2>&1 |grep vorbis

you will get some thing like that

EA libvorbis
DEA vorbis

or

EA vorbis
DEA vorbis

E= encoder

so the name of the codec to use will be

libvorbis

or in the second case

vorbis



if doesn't work i have no idea of what you have done, so resinstall ffmpeg from medbuntu repo.. try to use command lines i have writtten before

[eye208]

http://packages.ubuntu.com/gutsy/oggz-tools

Make sure the universe repository is enabled in software sources. Then enter
sudo apt-get install oggz-tools
to install the package. To extract the vorbis stream from a theora/vorbis .ogg file, enter
oggzrip -c vorbis -o audio.ogg yourvideo.ogg

[Creative2]

why i must use that stuff when ffmpeg do that !?

[eye208]

why i must use that stuff when ffmpeg do that !?
Because ffmpeg doesn't do it right.

[Creative2]

here it worked always fine , you must install properly or from medibuntu repo..

[eye208]

here it worked always fine , you must install properly or from medibuntu repo..
There is a bug in ffmpeg's ogg demuxer. This is unrelated to patent-encumbered codecs, so Medibuntu won't help here.

[philc]

There is a bug in ffmpeg's ogg demuxer. This is unrelated to patent-encumbered codecs, so Medibuntu won't help here.

Is the problem in the Ogg muxer or demuxer?

There was a new Ogg muxer for FFmpeg completed in December 2007, removing the reliance on libogg.

[eye208]

It's easy to find out if the bug has been fixed. Just try to demux the audio stream from "Experience Ubuntu.ogg" into a separate ogg file.
ffmpeg -i ~/Examples/Experience\ ubuntu.ogg -vn -acodec copy audio.ogg
It's giving me errors, and the output file is shorter than oggzrip's one.

libogg is the reference implementation issued by xiph.org. I don't see how reliance on it could be a mistake. oggzrip uses it too of course.

[Creative2]

if you get error it doesn't mean it doesn't work i have just recorded wth recordmydesktop and i have extracted ogg without error ...
i have installed this too
libogg0 libogg-dev libvorbis0a libvorbis-dev vorbis-tool
but i think you have a bad ffmpeg...
cheers

[eye208]

i think you have a bad ffmpeg
I think you used yours on a different input file. ;)

[crazyelf]

It's easy to find out if the bug has been fixed. Just try to demux the audio stream from "Experience Ubuntu.ogg" into a separate ogg file.
ffmpeg -i ~/Examples/Experience\ ubuntu.ogg -vn -acodec copy audio.ogg
It's giving me errors, and the output file is shorter than oggzrip's one.

libogg is the reference implementation issued by xiph.org. I don't see how reliance on it could be a mistake. oggzrip uses it too of course.

Exactly!!!! that's what wouldn't work for me. I thought it was just me or something I was doing wrong. But just incase I added the medibuntu to my repository list. Now i'm going to install what you've suggested. Thanks a lot for your help!

@Creative2 I reinstalled the ffmpeg from the medibuntu repository several times as well as everything listed as far as codecs that had an upgrade sign by it in synaptics I upgraded. When converting from ogg with theora to just an ogg vorbis file the audio was trash, and it had several errors on it. So unless its corrupt several times in a row fresh installs after removing it then readding it... that method isn't going to work. Now from ogg to mp3 that worked like a charm.

Thanks again guys,

Aaron

[philc]

It's easy to find out if the bug has been fixed. Just try to demux the audio stream from "Experience Ubuntu.ogg" into a separate ogg file.
ffmpeg -i ~/Examples/Experience\ ubuntu.ogg -vn -acodec copy audio.ogg
It's giving me errors, and the output file is shorter than oggzrip's one.


So, I've tried to do some more investigation on this. I didn't seem to have the Experience ubuntu.ogg file on my Ubuntu Studio system, so I downloaded it from here:

https://launchpad.net/ubuntu/gutsy/+source/example-content/26

Then using FFmpeg please see results below:

phillc@phillc-laptop:~/Desktop/example-content-26$ ffmpeg -i Experience\ ubuntu.ogg -vn -acodec vorbis experiencetest.ogg
FFmpeg version SVN-r11813, Copyright (c) 2000-2008 Fabrice Bellard, et al.
configuration: --enable-gpl --enable-pp --enable-libvorbis --enable-libtheora --enable-liba52 --enable-libdc1394 --enable-libgsm --enable-libmp3lame --enable-libfaad --enable-libfaac --enable-libxvid --enable-pthreads --enable-libx264 --enable-shared
libavutil version: 49.6.0
libavcodec version: 51.50.0
libavformat version: 52.7.0
libavdevice version: 52.0.0
built on Feb 2 2008 18:56:48, gcc: 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)
[theora @ 0xb7e19db0]Theora bitstream version 30200
[theora @ 0xb7e19db0]584 bits left in packet 81
[theora @ 0xb7e19db0]7 bits left in packet 82
Input #0, ogg, from 'Experience ubuntu.ogg':
Duration: 00:01:36.8, start: 0.000000, bitrate: 295 kb/s
Stream #0.0: Video: theora, yuv420p, 360x288 [PAR 1:1 DAR 5:4], 25.00 tb(r)
Stream #0.1: Audio: vorbis, 22050 Hz, stereo, 45 kb/s
Output #0, ogg, to 'experiencetest.ogg':
Stream #0.0: Audio: vorbis, 22050 Hz, stereo, 64 kb/s
Stream mapping:
Stream #0.1 -> #0.0
Press [q] to stop encoding
size= 379kB time=96.7 bitrate= 32.1kbits/s
video:0kB audio:319kB global headers:3kB muxing overhead 17.723846%

As you can see no errors, but I have specified the audio codec as Vorbis.

The output file sounds fine, it is 1.36 long, which is the same as the input file.

However, using the exact command you've suggested, just copying the audio codec doesn't work:

phillc@phillc-laptop:~/Desktop/example-content-26$ ffmpeg -i Experience\ ubuntu.ogg -vn -acodec copy experiencetest2.ogg
FFmpeg version SVN-r11813, Copyright (c) 2000-2008 Fabrice Bellard, et al.
configuration: --enable-gpl --enable-pp --enable-libvorbis --enable-libtheora --enable-liba52 --enable-libdc1394 --enable-libgsm --enable-libmp3lame --enable-libfaad --enable-libfaac --enable-libxvid --enable-pthreads --enable-libx264 --enable-shared
libavutil version: 49.6.0
libavcodec version: 51.50.0
libavformat version: 52.7.0
libavdevice version: 52.0.0
built on Feb 2 2008 18:56:48, gcc: 4.1.3 20070929 (prerelease) (Ubuntu 4.1.2-16ubuntu2)
[theora @ 0xb7d9bdb0]Theora bitstream version 30200
[theora @ 0xb7d9bdb0]584 bits left in packet 81
[theora @ 0xb7d9bdb0]7 bits left in packet 82
Input #0, ogg, from 'Experience ubuntu.ogg':
Duration: 00:01:36.8, start: 0.000000, bitrate: 295 kb/s
Stream #0.0: Video: theora, yuv420p, 360x288 [PAR 1:1 DAR 5:4], 25.00 tb(r)
Stream #0.1: Audio: vorbis, 22050 Hz, stereo, 45 kb/s
Output #0, ogg, to 'experiencetest2.ogg':
Stream #0.0: Audio: vorbis, 22050 Hz, stereo, 45 kb/s
Stream mapping:
Stream #0.1 -> #0.0
Press [q] to stop encoding
error, non monotone timestamps 10590 >= 10590
av_interleaved_write_frame(): Error while opening file
phillc@phillc-laptop:~/Desktop/example-content-26$

So, the desired output can be achieved by specifying the audio codec, rather than just copying existing.

[eye208]

So, the desired output can be achieved by specifying the audio codec, rather than just copying existing.
Yes, except for the fact that this is not the desired output. The goal was to demux the audio stream and copy it to a separate ogg container unmodified. Re-encoding not only takes much longer but also degrades the quality of the stream.