I started to explore the audio interface on my system (496/33 with
PAS16, running latest 1.1-alpha), and I thought I would write a simple
mixer program to set the volume and so forth.  In the process, I've
figured a bunch of things out, but I've also gotten myself a little
confused.  Here are the highlights:

* I was disappointed not to find any man4 pages for audio (nor for
  pas/sb/etc.) to describe how the interface should work.  Well,
  that's ok, I know how to RTFS...  (Are there man pages planned for
  1.1 release?)

* I'm confused about the AUDIO_MIXER_READ/_WRITE ioctl interface on
  the /dev/mixer.  Is this meant to be a device-dependent interface
  that one can use after verifying the device type via AUDIO_GETDEV?
  I wouldn't think so, but short of including <dev/isa/sbdspvar.h>,
  how does one know what "index"/"dev" value to use to
  AUDIO_MIXER_DEVINFO or AUDIO_MIXER_READ?  Are you supposed try all
  possible values and look at the names of the ones that are present?

* Both the AUDIO_MIXER_READ ioctl and the AUDIO_GETINFO ioctl (on
  /dev/audio) always return gain settings of zero.  I think this is
  just an unimplemented feature of the pas driver.  Does anyone know
  if this is supposed to work?

* When I do an AUDIO_GETINFO on /dev/sound, I still see an encoding of
  AUDIO_ENCODING_ULAW.  I thought the idea of /dev/sound was that it
  was for PCM samples.  Am I wrong, or is this just unimplemented yet?

Anyone have any hints on any of these questions?  Many thanks,


Tom Pavel

Stanford Linear Accelerator Center
pavel@slac.stanford.edu