Newbie question about Ubuntu Linux Sound Systems

Rune Maagensen maillist at maagensen.com
Thu Jan 27 10:08:32 UTC 2005


crimsun at fungus.sh.nu wrote:
 > So the way to picture this layering is:
 >   [mplayer]
 >   [esound]
 >   [ALSA/OSS]
 >   [sound card(s)]
 >
Which means that if you have a program that isn't compiled with esd, you 
won't be able to mix that sound :( From my experience far more programs 
are compiled with alsa than with different soundservers. Jack, esd ART etc.


 > Now for the confusing part. It's also quite possible to use:
 >   [mplayer]
 >   [ALSA/OSS]
 >   [sound card(s)]
 > and to eliminate the sound server completely. However, by doing so, you
 > eliminate the ability to mix sound streams unless your hardware is
 > capable (think sblive, sbaudigy, sbaudigy2, m-audio revolution, etc.).
 > This solution is preferred for the lowest latency.
 >
No you don't. Alsa dmix is running better than esd ever did on my 
ess1371. With Alsa dmix, Esd should be buried together with OSS (at 
least for Linux). To use dmix search for alsa dmix (there a good gentoo 
howto out there). My .asoundrc is included in the bottom.

 >>with Mono 1.0.5 but that Winamp lookalike. XMMS? Locks up immediately 
when I
 >>press play and can't only be brought down forcibly. I don't 
understand how
 >
 >
 > XMMS should be configured to use the esd output.
 >

NO! Esd should be disabled alltogether, use Alsa with dmix, and simplify 
your life.

Also if you use mplayer and run into problems with stuttering when using 
the terminal or other program you might want to add the following to a 
/etc/init.d/local:

#!/bin/sh
echo 1024 > /proc/sys/dev/rtc/max-user-freq

And then making a symlink from /etc/rc2.d

/Rune




More information about the ubuntu-users mailing list