Yes, and I found mpg123 given enough buffer (2m-4m bytes) without esd didn't cause stuttering.   So stuttering with esd may be the problem of socket connection to esd.   As you write, it looks like sending data cannot keep up with playing data, repeating the existing data.

> 
> ----------
> sender: "Wolfgang S. Rupprecht" <wolfgang+gnus20040907T131349@dailyplanet.dontspam.wsrcc.com>

> maverick6664@infoseek.jp writes:
> > When I play music with mpg123 or xmms (or whatever it may be AFAIK,
> > with or whitou esound) the sound sometimes begins stuttering when
> > other processes are busy and the stuttering doesn't stop until the
> > player software stops even if other processes are stopped.  
> ...
> > I use two cmpci (cm8738) cards on P4 2.8EG.
> 
> I see the same thing here with a cm8738 card in an athlon-1.1Ghz
> machine.  The machine doesn't have to be under load to have it happen.
> 
> The fastest way to have it happen is to use a fixed-rate data source.
> Eg. having another machine digitize audio at a fixed rate and have it
> sent across the net and then played locally.  It will usually start
> stuttering within 30 seconds.  It "feels" like the play buffer is
> draining and playing snippets of old samples.  Someone that
> understands this driver better may want to try turning the play logic
> off when the buffer runs empty instead of playing trash.
> 
> -wolfgang
> -- 
> Wolfgang S. Rupprecht                http://www.wsrcc.com/wolfgang/
> 
> 



 
------------------------------------------------------------------------
$B%"%F%M$N$"$H!!M7$S$K9T$3$&!*(B $B!VN"%O%o%$!W(B by infoseek
http://ap.infoseek.co.jp/travel19.html