Xref: utzoo rec.audio:28978 rec.music.synth:19088 comp.music:2597
Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!usc!apple!motcsd!lance
From: lance@motcsd.csd.mot.com (lance.norskog)
Newsgroups: rec.audio,rec.music.synth,comp.music
Subject: Re: Ambisonics?
Keywords: binaural surround-sound
Message-ID: <2905@motcsd.csd.mot.com>
Date: 18 Feb 91 23:47:34 GMT
References: <1991Jan30.100357.15534@canon.co.uk> <2314@otc.otca.oz>
Followup-To: rec.audio
Organization: Motorola CSD, Cupertino CA
Lines: 22

I haven't looked all that hard in the libraries, but
the only mathematical treatment of stereo sound synthesis that I've 
managed to find is in the latest issue of the Computer Music Journal,
Feb. 1991.  The article is by someone at CCRMA, Stanford University.
I don't remember if it was just her, or 'et al'.  

They ran pairs of speakers off of Midi-controlled synths with Midi reverbs
feeding the pairs, and built a system for the Mac that does 2D sound
placement given N synth/reverbs and N speaker pairs.  The Mac just sent
out MIDI data.

To calibrate a configuration, they just set up
all the speakers in the target configuration, put digitizing mikes in the
middle, ran the speakers off of MIDI-controlled synths and calibrated
the reverb times and sound levels according to their sound model.
They take the input data and do a curve-fit into their model,
then do quite a bit of trig. to create the output equation model.

Sounded very spiffy, if it could be incorporated into a larger
MIDI sequencer.

Lance