You are not logged in.
Pages: 1
I've had this problem for a while now. Basically, no matter what format, audio or video, mplayer just "stops" just at the end of the song or video. It doesn't quit or continue to the next item in the playlist or loop, it just sort of hangs. I think its status is still "Playing", not "Paused" or "Stopped". From the command line in verbose mode, when it hangs, it keeps printing:
ds_fill_buffer: EOF reached (stream: audio)
ds_fill_buffer: EOF reached (stream: video)
A: 89.0 V: 88.9 A-V: -0.009 ct: 0.004 0/ 0 0% 0% 0.7% 0 0
ds_fill_buffer: EOF reached (stream: audio)
ds_fill_buffer: EOF reached (stream: video)
A: 89.0 V: 88.9 A-V: -0.009 ct: 0.004 0/ 0 0% 0% 0.7% 0 0
ds_fill_buffer: EOF reached (stream: audio)
ds_fill_buffer: EOF reached (stream: video)
A: 89.0 V: 88.9 A-V: -0.009 ct: 0.004 0/ 0 0% 0% 0.7% 0 0
Any ideas? I'm using mplayer2 from community.
Offline
same problem here. but i using mplayer2 from git
Last edited by sl1pkn07 (2012-06-02 14:31:59)
Offline
Just registered here to post I have the same problem on Gentoo since upgrading to pulseaudio 2.0.
The bug appears to be known already, see the answer and reference to bugtracker by mplayer2-developer here:
http://www.mail-archive.com/debian-bugs … 25539.html
Current mplayer2-git still appears to be without any fix / workaround for this bug in pulseaudio. So for now, only two solutions left until there is some fix:
Revert to older pulseaudio.
Use -ao alsa.
Offline
I'll check my configurations and such later; mplayer2-git works properly here with the most recent pulseaudio, including seeking around/moving to the next track/etc., and also including mplayer on a home server sending audio to my desktop via pulse.
Last edited by ZekeSulastin (2012-06-09 23:34:52)
Offline
I'll check my configurations and such later; mplayer2-git works properly here with the most recent pulseaudio, including seeking around/moving to the next track/etc.
Please be sure to check you are really using "-ao pulse", and if it still works, please try with a very short file (such as the sounds from purple messenger in /usr/share/sounds). Those do not even start playback here, as has also been described in the bug report.
If all works, I'd love to see some "mplayer -v" - output . Thanks!
Offline
Ah; I was just referring to the OP; didn't think to check the bug report as I thought it was the same thing as the OP
As far as seeking around and moving to the next track and such, to get the logs I installed mplayer2 20120517-2 x86_64 from community to avoid possible local compilation contaminants, then I did each test three times: once thru the pulse output, once thru the alsa output handled by pulse, and once thru the alsa output with no pulse at all. On seeking and moving to the next track, I skipped around a bit in the first track and let it transition to the second.
Pulse: http://sprunge.us/jbCZ | ALSA thru: http://sprunge.us/ZARI | ALSA direct: http://sprunge.us/YZXc
Seems like I am experiencing the bug regarding EOF, though not anywhere near as severely as the above posters; the delay between one song ending and the next starting is a second or so at most, although noticeably shorter with direct ALSA. Seeking around worked perfectly in all cases.
As noted in your bug report, though, short playback is indeed an issue with mplayer2; using test.mp3 from the bug report, it failed with -ao pulse but worked with both -ao alsa variants:
Pulse: http://sprunge.us/jjOT | ALSA thru: http://sprunge.us/WbYd | ALSA direct: http://sprunge.us/KJab
Everything works fine sending audio via pulse from my server, but that's a different distro with mplayer1 and pulseaudio 1.1.
Sorry for the noise olifre; I wish I could tell you what made it less severe for me
Last edited by ZekeSulastin (2012-06-10 01:56:58)
Offline
Sorry for the noise olifre; I wish I could tell you what made it less severe for me
No problem, thanks a lot for your tests!
I guess it may be related to something as subtle as buffer sizes pulseaudio keeps, which might vary from card to card (I guess). At least, fixing such low-level-audio problems is a bit over my head, but maybe the bug will be taken care of in Pulseaudio 2.1 / mplayer 2.1 (I just realized the versions match ).
The hang at end of track is endless for me, too (as for the OP). However, "sometimes" when seeking in a file, the hang at the end of track totally goes away. It appears to depend on how many integer multiples of one buffer fit in the rest of the track.
I guess -ao alsa (via pulse) is the best solution for now . Thanks for your input!
Offline
Pages: 1