Open GoogleCodeExporter opened 9 years ago
Try svn version:
svn checkout http://qmmp.googlecode.com/svn/branches/qmmp-0.2/ qmmp-0.2
Original comment by trialuser02
on 25 Feb 2009 at 10:50
Tried r809 on the branch, same results.
Original comment by yselkow...@gmail.com
on 25 Feb 2009 at 6:54
OK, try another player. If qmmp only has this problem, show its output during
starting playback.
Original comment by trialuser02
on 28 Feb 2009 at 2:59
> OK, try another player.
I have dozens of multimedia programs on this computer, and all work fine with
OSS
audio output.
> If qmmp only has this problem, show its output during starting playback.
OutputOSS: uninitialize
DecoderMAD: deleting input_buf
DecoderMAD: deleting output_buf
SoundCore: creating decoder
ok
SoundCore: decoder was created successfully
DecoderMAD: 152 bytes skipped
DecoderMAD: Fixed rate detected
DecoderMAD: Total time: 12
DecoderMAD: 152 bytes skipped
Original comment by yselkow...@gmail.com
on 1 Mar 2009 at 5:04
OK, try r818. If the problem still exists, show output again.
Original comment by trialuser02
on 2 Mar 2009 at 12:53
I appreciate your persistence. I build r818 from the branch, the results are
still
the same, here's the output with a 12s MP3 which played at 1/8 speed:
SoundCore: creating decoder
ok
SoundCore: decoder was created successfully
DecoderMAD: 152 bytes skipped
DecoderMAD: Fixed rate detected
DecoderMAD: Total time: 12
OutputOSS: frequency=44100, channels=2, bits=16
DecoderMAD: 152 bytes skipped
With a 68s Ogg/Vorbis file, the speed was approx. 1/4:
SoundCore: creating decoder
ok
SoundCore: decoder was created successfully
DecoderVorbis: initialize
OutputOSS: frequency=44100, channels=1, bits=16
file info:
ARTIST =
TITLE =
ALBUM =
COMMENT =
GENRE = Other
YEAR = 0
TRACK = 0
LENGTH = 68
Testing a few different files, the speed seems to be 1/(4 * channels); 1/4 for
mono,
1/8 for stereo. Does that help?
Original comment by yselkow...@gmail.com
on 2 Mar 2009 at 2:39
OK, try again. I hope, it shows more useful info now.
Original comment by trialuser02
on 2 Mar 2009 at 4:57
Sorry, output and results are exactly the same.
Original comment by yselkow...@gmail.com
on 2 Mar 2009 at 5:16
May be qWarning() doesn't show messages under windows. Find outputoss.cpp file
and
change all "qWarning" functions to "qDebug". Than try again. I consider,
"/dev/oss"
doesn't support 44kHz.
Original comment by trialuser02
on 2 Mar 2009 at 5:58
Cygwin != Windows, and qWarning does show messages (I added one unconditionally
to
verify). I assure you that these files do play on other audio players with
their OSS
backends.
Original comment by yselkow...@gmail.com
on 2 Mar 2009 at 6:48
OK, yet another attempt. Try to remove this lines from outputoss.cpp
int flags;
if ((flags = fcntl(m_audio_fd, F_GETFL, 0)) > 0)
{
flags &= O_NDELAY;
fcntl(m_audio_fd, F_SETFL, flags);
}
Original comment by trialuser02
on 2 Mar 2009 at 9:23
Sorry, no change. :-(
Original comment by yselkow...@gmail.com
on 2 Mar 2009 at 10:05
OK, try to increase sample rate. Change "m_frequency = freq" to something like
"m_frequency = freq*4" or 8. Does it change anything?
Original comment by trialuser02
on 3 Mar 2009 at 9:00
Using freq*4, I got an additional message:
OutputOSS: can't set frequency, using 176400 instead
With freq*8, the message was 352800. That comes from the if(m_frequency !=
freq)
conditional which you added.
Playback was still as slow.
Original comment by yselkow...@gmail.com
on 3 Mar 2009 at 9:21
Looks like "/dev/dsp" doesn't use samplerate settings. And I can't find any
differences in oss code between mpd, audacious and qmmp.
I consider, this is last difference:
- open(m_audio_device.toAscii(), O_WRONLY, 0);
+ open(m_audio_device.toAscii(), O_WRONLY);
Original comment by trialuser02
on 3 Mar 2009 at 10:40
Looking at the Cygwin sources, I see now that SNDCTL_DSP_GETODELAY is not
supported.
GStreamer and PulseAudio use SNDCTL_DSP_GETOSPACE as a backup.
If you're out of ideas, then I'll try to figure this out myself, and if I'm
successful, I'll propose a patch here. I appreciate your efforts nevertheless.
Original comment by yselkow...@gmail.com
on 4 Mar 2009 at 7:13
Yes, you are right. I haven,t more ideas. So I'm waiting for your patches.
Original comment by trialuser02
on 4 Mar 2009 at 3:17
Original comment by trialuser02
on 5 Apr 2009 at 9:44
Original issue reported on code.google.com by
yselkow...@gmail.com
on 25 Feb 2009 at 7:08