Converted .vob files have no audio

Generally, converted VOB files do have audio (albeit the main audio channel has changed from the original), but the ones that do not have audio have been taken from BBC DVD's. In particular, Red Dwarf.

Hi,

If the audio is being picked up correctly in Totem but not OggConvert then it really does sound like something strange is happening, as they both use the same method to decode audio streams.

Please could you run:

GST_DEBUG=3 oggconvert > vobconversion.log 2>&1

in a terminal, convert the offending VOB, and then post the file vobconversion.log as an attachment to this report? That way I can (hopefully) get an idea of what's going on.

Thanks,

Tristan

Thanks for the log file :-)

I *think* I know what's wrong (technically, the pipeline looks like it's moving to PLAYING before the decodebin has finished finding pads).

Attached is a (hopefully) fixed version of the transcoder. Just drop it in to replace the existing ocv_transcoder.py file (use locate "ocv_transcoder.py" in a terminal to find out where this is).

Please let me know if this works. I'm fairly sure it should, but if not, please could you repeat the above procedure and attach another log file?

Thanks!

Nice one! Now I have an audio track. But I do have a couple of questions:

1. Why is the audio bit rate 86kbps when I selected 4 for quality. For mp3's, 4 setting has given me 128kbps.
2. It only converted one audio track. Is that correct?

1) I suspect that the AC3 decoder and the Vorbis encoder have between themselves decided to only encode in mono rather than stereo or surround or whatever the original was. I know that GStreamer's a52dec used to have a bug where it did this, but I thought it was fixed... Could you run `gst-inspect-0.10 a52dec | grep "Version"' in a terminal and let me know the output?

2) That's definitely not correct :-(

Would you be able to submit another log file so I can see what's going on again? You don't need to wait for the whole thing to convert, all I'm really interested in is the setup phase, so just let it get started, wait 10 seconds or so and then cancel it.

(Even better than a log file would be if you could upload a small VOB somewhere that exhibits this problem so I could have a look at things directly, but I don't know whether this is feasible.)

Version 0.10.7 (in Ubuntu Hardy)

Totem doesn't give any track selection, but at least in converted the correct one. (Another vobconversion.log file attached.) It is possible that the audio is mono in the original. That would explain the low bit rate.

I have the film 'Cube', which converts both audio tracks, but when a play it back, it plays the commentary track by default instead of the film track. Perhaps I could provide another log file for this?

Here's another version of ocv_transcoder.py for you to try. It seems the last one didn't actually fix the issue of the pipeline trying to start before it's finished connecting all the encoders and decoders together. This one should -- I hope!

As I suspected, the encoding is only being done in mono, but I'll leave you to tell me whether this is an error or not :-)

The issue with the track order being reversed is one I've known about for a while. It's very annoying, but I don't have a good solution at the moment I'm afraid. But rest assured I'm thinking about it!

Back to no audio again.

I've uploaded a sample of the vob file to:

www.scobble.me.uk/sample.vob

Could it be possible to add a selection box to select a particular audio stream, that is to say, to either select all streams, or a specific one. I would think in most cases, users only want to encode one audio stream anyway.

Also a conversation of a Youtube video with mono mp3 22050 Hz 59 kbps failed:
Totem can't play, but VLC plays video; audio track isn't there.

PS: Will "Open with ..." be possible someday? And commandline *g*.

Anyway, a beautiful toll with "SoundConverter" (when video, then it only takes the audiotrack!), to encode all files to free codecs.

Tristan,

Any updates on this? Can we close the bug or is it still there?

Thanks,
Kerry