So, here's what I've got going on that makes it all rather tricky.
- Sometimes I'm in my studio, with the m-audio. Sometimes I'm away, and need to switch to the built-in hardware.
- Most of the time I only need stereo output from the m-audio, but I do need to use quad occasionally.
- (I know I'm going to get a lecture about this) For regular, daily use, I can't do without pulseaudio. I have an RSI which forces me to do most of my prose writing using dictation software -- NaturallySpeaking running in VirtualBox. It's an absolute necessity -- can't get by without it. VirtualBox only supports pulseaudio, and for voice dictation, obviously I need to get audio input into the virtual machine. So it's not an option to uninstall pulse.
I have disabled pulseaudio's autospawn, and I can selectively turn on pulseaudio when I need it. Obviously, in critical performance situations, I would run jack only and leave pulse off. For daily use, it's extremely useful for me to have e-mail and the dictation software in one workspace, and SuperCollider in another workspace, so that I can write an e-mail quickly if I run into a problem.
In short, I know pulse is really wretched, but I would have to disregard advice to uninstall it.
- My typical configuration in 10.04 was to play back to the m-audio, and capture from the built-in hardware (headset microphone for dictation). Of course I wouldn't do that in a concert setting, but for daily use, in a development or composing cycle, that's what works best for me.
- Most of the time: -Chw:0 -Phw:1
** Edit: Actually, I think pulse will try to use the built-in hardware by default instead of the jack sink. So maybe this is not an issue. Let me try it and reply later.
- When I'm away from the studio: -dhw:0
- When I need quality mic input: -dhw:1
- When I need quad output: -Chw:1 -Pft4
So, if -C and -P are "contraindicated" as the doctors say, what's the best way to accomplish these? Only the last is a performance situation, so maybe I only need to worry about .asoundrc weirdness for that case.
Or, should I just keep doing things the way I have been doing them, and hack around the dbus problem until a proper fix comes (or I try kxstudio's ppa)?
"Come said the Muse,
Sing me a song no poet has yet chanted,
Sing me the universal." -- Whitman
audio clips: http://www.dewdrop-world.net/audio
more audio: http://soundcloud.com/dewdrop_world/tracks
Linux-audio-user mailing list