I have a via8233 onboard sound chip. It seems to almost work with alsa drivers. Almost, because only one program at a time seems to be able to play sounds. If i start another program that is supposed to play a sound, it will pause until the first program playing sound is finished. This is uncomfortable for a lot of reasons. It outrules the combination of xmms & Quake 3. I tried searching the internet and usenet for similar problems, but found nothing of value. Anyone know if this is fixeable at all without buying another sound card? This is on Mandrake 9 with kernel 2.4.19 btw.

You need to have a sound server (ESD, MAS, aRts, Jack, et cetera) mix the sounds before they are fed to the sound card.

XMMS allows you to change what it uses for output in its preferences dealie.

Most Gnome applications use ESD (although MAS looks like it will replace ESD in near future), most KDE applications use aRts, most other applications use OSS or ALSA (which are more or less "direct access").

I have no idea if Quake 3 bothers to support any sound servers, or how you''d tell it to use one of them.

Oh well, I think it''s possible to get quake 3 going with aRts, but the sound will lag (or so i hear). It seems like a waste of CPU power to use a sound server too... Guess I''ll have to dig out the old aureal vortex card. At least it can do mixing directly in hardware. the drivers for that one are a little flaky, though.

I''m thinking about getting a soundblaster live instead. Would it support hardware mixing on linux?

I''m probably not the best person to answer questions about sound support in general, but I can actually help you out with that because I have a Sound Blaster Live! in my machine (all three machines in the house have Sound Blasters ).

To get an answer for you, I played three different OGG files at once (with three processes or ogg123) with OSS as the output device, and I could hear all of them (no pausing or swapping or any such evilness). I haven''t taken the time to play around with ALSA yet.

Rambling about sound servers: ESD lags a bit. I don''t know about Jack (it''s meant for high quality). I don''t know about aRts either (it''s been a long while since I''ve used it). MAS is meant to have really low latency as one of its design goals, but I haven''t tried it either.

Thank you for all the info. I will probably be picking up a soundblaster live.

I tried the aRts output plugin with XMMS, and it sounds terrible! hissing and crackling... Not really an option.

Again, thanks.

I guess this is just some ignorant ramblings, but isn't the whole "open sound card, write to it" way of doing things outdated? I mean, shouldn't the hardware be able to work with lots of different opens on the sound card? If not the hardware itself then the driver maybe?

I mean, it still basically works the same way it does way back in the day when I first got my Dell with its little Vibra 16 sound card! Sure, we have sound cards that can feed audio to 6.1 speakers, do enviromental effects with 3D geometry, and all that stuff, but why couldn't the sound card itself be more of a cooperative system?

BTW, I hear that the newest sound cards support multiple opens, but I don't have one of those, so I don't know if that works like what I'm talking about or not.

EDIT: Well, when I originally tried to post it, all the above stuff wasn't said yet. I got a 500, left for a while, came back and just hit refresh to post it. So its a little behind the times. Oh well.

[edited by - BradDaBug on February 1, 2003 1:34:19 AM]