Re: audio lag in 5.2-Beta (card emu10k1)

On 2003-12-18 15:39 +0000, Peter Schultz wrote:
> Munish Chopra wrote:
> >On 2003-12-17 23:47 +0000, Mathew Kanner wrote:
> >
> >>On Dec 17, Munish Chopra wrote:
> >>
> >>>Embarrassingly enough another reply to myself, but I didn't think I'd
> >>>find it this quickly. Reverting to revision 1.88 of channel.c fixed
> >>>things right up (also 1.20 of buffer.c and 1.7 of buffer.h). No more
> >>>lag.
> >>>
> >>>Revision 1.89 of channel.c attempted to make really short sounds
> >>>actually play. Looks like buffer issues. 
> >>
> >>	Munish,
> >>	Does this only fix the amount of time it takes to seek in
> >>mplayer or xmms, or to start a track?
> >>	--Mat
> >
> >
> >Yes, reverting fixes seek/startup time.
> >
> >If you're concerned about the popping noises, I switched soundcards to
> >an old SB PCI128 I had lying around (some form of Ensoniq chipset), and
> >I don't seem to have them anymore. I had several identical SB16's
> >around, every single one of them had clearly audible popping
> >noises.
> >
> >Your patch in kern/59208 *did* seem to lessen the popping noises on the
> >SB16 cards significantly, though it didn't remove them completely.
> >
> 
> Is there anything more one can do to help narrow this problem down? 
> I've never tried profiling before, would it be helpful in this situation?
> 

I'm not sure. After approximately a full day I noticed a few popping
noises even with the new soundcard, but they're very rare in comparison
to before. I'm going to apply the patch in kern/59208 while running with
revision 1.88 of channel.c for a day or two and see if that doesn't
somehow fix the popping.

-- 
Munish Chopra