[MPlayer-dev-eng] Bug in mplayers EVO demuxing code

infernix infernix at infernix.net
Thu Jul 5 00:39:45 CEST 2007


Nico Sabbi wrote:
> I don't see the slightest difference between the decoding of 
> artefacts12.evo and artefacts12.vc1. The only message reading
> "bits overconsuption" happens at the end of the stream.

You are right; the error is indeed for the end of the stream. However, 
the frame with the error is definately there. I've made a screenshot:

http://dx.infernix.net/ffvc1/artifacts12.png

Here's how to reproduce:

mplayer -vo png -frames 25 artifacts12-evodemux-noerrorframe.vc1
mv 00000025.png artifacts12-evodemux-clean.png

mplayer -vo png -frames 25 artifacts12-mplayerdemuxed-errorframe.vc1
mv 00000025.png artifacts12-mplayerdemuxed-error.png

As you can see, both mplayer and evodemux seem to demux a file with the 
same filesize, but the md5sums differ:

6912586 2007-06-25 14:50 artifacts12-evodemux-noerrorframe.vc1
6912586 2007-06-25 14:50 artifacts12-mplayerdemuxed-errorframe.vc1

690a7ef5d97b1e287316fe99db8f3055  artifacts12-evodemux-noerrorframe.vc1
b81d6dfa5ea144189f28eed400a422f3  artifacts12-mplayerdemuxed-errorframe.vc1

 > As for bluray.ts it's a different story

Correct, and I haven't encountered this error on Bluray VC1 at all so 
far. That's why I think it's in the EVO demuxing code.

Sorry for not being clear enough initially. Oh, and the code on which 
evodemux is based is in here: http://www.w6rz.net/evob_demux.zip - that 
might be helpful.

Thanks!





More information about the MPlayer-dev-eng mailing list