GG: Stereo and What It Can't Do

["John P. Hill" <jphill@frank.mtsu.edu>]

On Tue, 21 Jul 1998, Bruce Petherick wrote:

> >I'm afraid I disagree with Ingvar. To me, the whole purpose of CDs and 
> >LPs and other recording media is to give the listener the illusion that 
> >he or she is in the room or hall where the recording took place. Anything 
> >that destroys that illusion is to be avoided. 
> 
> ... and that is exactly what can't be done - it is _impossible_ to give
> someone the exact illusion that they are in the room, or hall where the
> recording took place for a number of reasons.  The first, and overriding is
> the playback mechanism - one _isn't_ in the hall when one listens back.
> Even headphones colour sound playback to a certain degree so that the
> illusion of placement of microphones is disturbed.  I am not saying that
> this destroys the entire illusion, but certainly effects it to a large degree.

Bruce is quite right, at least with regard to *stereo* recordings played
back over two loudspeakers.  This recording and retrieval system can
*never* place the listener in the acoustic environment of the original
recording and the physics exist to prove that if anyone wants to do the
homework.  For one thing, when you listen to stereo recordings through two
loudspeakers all of the reverberant information from the room/hall is
encoded *with* the direct sound and comes from the two loudspeakers
producing the sound in *front* of the listener.  This is very different
from sitting in a good acoustic space and having ambient cues coming from
all around your head with appropriate filtering by the ear pinnae (outer
ears).  Stereo playback *cannot* and will never achieve this.

Binaural recording and playback is another matter.  Recordings can be made
with a dummy head using omni mic capsules implanted in the artificial ear
canals.  When played back over appropriate headphones, this system can
give a very convincing sense of spatial cues with regard to direct and
reverberant information.  It's still problematic, because to do it right
you'd need a dummy head whose pinnae exactly matched the intended listener
and a set of in-the-ear-canal earphones so that the important pinna
transfers were only done once (during the recording).  Since everybody on
the planet has different pinnae and ear canals.....well, you get the
picture.  It's a compromise at best.

Multichannel systems offer great promise here.  There is a system called
AMBISONICS, developed by Michael Gerzon during the 70's that attempts to
encode height, width and fore/aft information during the recording
process. Then, during playback, the 360 degree spherical information is
recreated using six (or more) loudspeakers in the playback room.  Many
folks report good results with this system.  Unfortunately, it came
right after QUADROPHONIC systems failed miserably in this country and
Ambisonic recording and playback systems have never really caught on here
in the US.

Dolby PRO-LOGIC is a system familiar to many with home theatres.  It
provides LEFT/CENTER/RIGHT/MONO SURROUND playback by using a 4:2:4 matrix
system.  It makes those action movies sound really big, but it is not
effective in providing true surround for serious music recording and
playback, in the sense of recreating the type of *envelopment* one gets in
a good hall.

Along with new developments in digital (DVD, etc.), we now have new
discrete surround systems (5.1 and 7.1) which will, over time, replace
traditional stereo.  It really is time for this to happen;  the principles
behind stereo recording were pretty much completely worked out in the
1930s by Alan Blumlein in the UK and Harvey Fletcher of Bell Labs in the
US.  Surround encoding and playback is the next step.

jh