[lug] audio: hardware mix versus software

[D. Stimits]

I'm thinking of making an audio device under Linux, which would 
basically be a mixer for a lot of audio services, e.g., voice chat, 
music serving, telephone answering, so on. The particular audio card 
which appears to be most capable is also quite expensive, part of the 
RME Hammerfall series. The older EMU10K1 is a known reliable, but I 
don't really have any information on its hardware mixing abilities. The 
EMU10K2 is probably better than the EMU10K1, but I've heard stories of 
incomplete drivers (the stories are quite old, I do not have current 
information). ALSA provides a nice interface to do things in software, 
but I fear that if I'm mixing and routing a significant amount of audio, 
quality would eventually suffer. Does anyone here have any 
recommendations on what hardware would best work for a large number of 
audio channels, with full ability to output to any destination, e.g., 
5.1 or 7.1 surround, headphones, ordinary stereo, and voice between 
systems? How important is hardware mixing, versus software mixing?

D. Stimits, stimits AT comcast DOT net