On Fri, 27.03.09 08:36, Colin Guthrie (gmane at colin.guthr.ie) wrote: >>> Lennart Poettering wrote: >>>> I doubt that it would be so useful to expose this in PA anyway. I >>>> mean, how realistic is it wanting to drive both SPDIF and analog audio >>>> at the same time from the same sound card? Appears to be a very exotic >>>> niche usage to me. >>> Please don't forget the combination of spdif and headphones. I'm >>> quite sure I won't be the only one using spdif for "normal" audio >>> listening and simultaneously a headphone for voip stuff. >> >> This might actually be a valid point. >> >> However, as mentioned it's a bit hard to detect these cases and >> distuingish them from cards that do hw mixing. If ALSA would give us a >> bit of a hint here we could certainly add this. > > So from your original comment, is it possible that some hardware lets > you open the mutually exclusive bits, but only errors out when you try > to write a sample to them at the same time? No. (I) Some hardware/drivers allow you open spdif:0 in app A and front:0 in app B at the same time and it results in having A's stream come out of SPDIF and B's stream coming out Analog. (II) Other hardware/drivers however allow you to open spdif:0 in app A and front:0 in app B at the same time, and audio is mixed and both streams come out of only SPDIF, or only Analog, or even on both. And we wouldn't even know which of the cases it actually would be. We never want to get confused by case II but we cannot distuingish it programmatically from case I. Case I is the one we want to support. Case II we want to ignore. Lennart -- Lennart Poettering Red Hat, Inc. lennart [at] poettering [dot] net ICQ# 11060553 http://0pointer.net/lennart/ GnuPG 0x1A015CC4
