Re: On 8 speakers

Subject: Re: On 8 speakers
From: Chris Rolfe (
Date: Thu Jan 28 1999 - 01:21:57 EST

We're using amplitude panning, but with dry signals in an 8-channel setup
it's pretty effective.

The diffusion controller is also interesting. In addition to specifying
angle and magnitude (independent of speaker location), the user specifies a
diffusion width, that is, send input 1 to all speakers between 0 degrees to
90 degrees.

  x | x
x |__ x
  x x

This facilitates diffuse sounds, global reverb, and ambiences. A global
reverb can have a low-pass and slight delay on it. Doppler-shifts only make
sense if you use a spatial/trajectory model, which I've tried to avoid.

I think the diffusion controller illustrates the limitations of a strict
spatial model. What would be the xy equivalent of a 360 degree (diffusion)
assign? If I had a global reverb return to the mixer I would assign it to
all speakers, but by no means imply an xy position.

I agree with you that visual perceptions bias composers' expectations in
working with diffusion systems. More importantly than the limitations of
panning, however, is the role of source material.

WYSIWYG (or, maybe, What You Hear Is What You Get ?) isn't possible for
sound object localization models for the simple reason that the
representation of a sound file (as a blit, or mic icon, or other little
pict) doesn't reflect the actual perception of that sound object in space.
A reverberant sound, for example, cannot be moved convincingly around a
space. It's a meta-perception (a space moving in a space??).

I'm still mulling a lot of these design questions, though, and considering
how much localization can co-exist with diffusion. I get the impression
that most people working in diffusion, spatialization and localization are
caught up in the same problems, wondering how best to represent and
manipulate the sound and mix. And aware that each decision might close down
a valid sound-reinforcement technique.



At 5:32 AM -0400 1/26/99, KEVIN AUSTIN wrote:
>Is the panning done using amplitude, reverberation, low-pass filtering
>and doppler?

Third Monk Software
262. W. 6th St.
North Vancouver, BC
V7M 1K6
fax: 604-990-1019
tel: 604-985-7513

This archive was generated by hypermail 2b27 : Wed Jun 11 2003 - 13:08:50 EDT