On Tue, Jan 12, 2010 at 10:41:41AM +0800, zhouzy at os.pku.edu.cn wrote: > It seems latest mplayer switches from mp3lib to ffmp3 as the default mp3 > decoder. I have two questions about this: I'm tempted to revert that soonish, but I wanted to get an overview of what issues the FFmpeg decoder has. Sorry for forcing testing on you :-) > 1) Today when playing one mp3 file, mplayer continuously shows warning > messages even if I use -really-quiet. Compiled with MinGW GCC 4.4.1, > running on WinXP. I think you should be able to suppress it at least with the right -msglevel option. Something like -msglevel decoder=-1 If it doesn't work that would be a bug I think. > cmdline: mplayer xxx.mp3 -really-quiet > > [mp3 @ 0x2994e00]overread, skip -5 enddists: -3 -3 > Last message repeated 1 times > [mp3 @ 0x2994e00]overread, skip -5 enddists: -2 -2 > [mp3 @ 0x2994e00]overread, skip -6 enddists: -5 -5 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -4 -4 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -3 -3 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -2 -2 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -3 -3 > [mp3 @ 0x2994e00]overread, skip -6 enddists: -4 -4 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -2 -2 > Last message repeated 2 times > [mp3 @ 0x2994e00]overread, skip -4 enddists: -1 -1 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -4 -4 > [mp3 @ 0x2994e00]overread, skip -5 enddists: -2 -2 > > I will upload the sample if needed. Probably a good idea. I know we had this issue a few times in the past, but I _think_ those cases have been fixed. > 2) It seems mp3lib is much faster than ffmp3. Tested on a Core2 E6750 OC > 3.2G: Yes, that's because ffmp3 is an integer decoder > Any good reason to drop mp3lib? Overall probably not after all. For now it was meant as a measure to get an overview of the issues removing it would cause. The speed one I could have found out myself, but for those overread issues and also possible noise after seeking it's a matter of having the right samples.
