I'm thinking about possible ways of implementing whole house multichannel audio system under lmce.
What do I imagine under such system :
- black box with multiple inputs and outputs
- audio sources on inputs and audio sinks (audio cards) on outputs.
audio sources under LMCE:
- currently limited number - most known is Xine player wrapper and speech announcements (they can be separate audio source)
- the problem is cause you cannot run more media players on one Core, I vision that Audio only xine player thing is a good solution to this, so one can start more media players on core and get more sources on one machine (using more MDs to just get more audio sources is inefficient IMHO)...
- currently the problem is cause you cannot easily add another Xine player to LMCE system - I mean on creation of new audio zone and automatic creation of scenarios for control of such media player...
- currently xine player is directly connected to sound output - aka sound card
Black box :
- I have found interesting project that could be what I'm looking for :http://pulseaudio.org/
(I spotted it when Jim on MH mailing used it to connect Asterisk and PocketSphinx easily with it for speech recognition)
Maybe als Jack is candidate ? not sure, don't know much about it....
Black box in functionality must allow all kinds of switching,routing,mixing of audio inputs into several audio outputs. In this way we could much better implement also speech announcements, that are currently pretty unusable (media play is interrupted, not pause or faded out, after announcement, play is not restored from previous possition but with next song etc.. - the right way should be to fade out, pause media, make announcement and then fade in slowly to continue with media). I also spot that some applications under lmce are still trying to have exclusive acces of sound card (Embedded Phone used to do that if I remember correctly).... Using above mentioned project I think this could be helped a lot...
Any opinions, ideas ?