Path: utzoo!attcan!uunet!mailrus!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!uwm.edu!csd4.csd.uwm.edu!info-high-audio-request
From: kanefsky@umn-cs.cs.umn.edu (Steve Kanefsky)
Newsgroups: rec.audio.high-end
Subject: Re: Surround Sound
Message-ID: <4732@uwm.edu>
Date: 28 Jun 90 21:36:10 GMT
Sender: news@uwm.edu
Lines: 72
Approved: tjk@csd4.csd.uwm.edu

In article <4675@uwm.edu> gould!infocenter!smatchun@encore.encore.com (Scott Matchunis) writes:
>
>I have some questions about surround sound systems, and was hoping some of 
>the readers of this group could advise me.
>
>I have auditioned some of the better surround sound systems, both in audio
>and video (laserdisc) format.  I have been able to set up A/B comparisons,
>but have not been able to detect much of a difference with audio-only.  The
>systems included seven loudspeakers and plenty of power.  With laserdisc, the
>system sounded truly impressive.  But, since I'm not too much into video, I 
>asked to have the laserdisc player replaced with a CD player.  Using the A/B
>switch, I could detect a subtle difference when the surround sound processor
>was being used, but I can't say I noticed anything worth spending $1000 for the
>surround sound unit, another $500-1000 for another power amp, and three more
>speakers (I already have four).
>
>Am I missing the point with surround sound, i.e. is it mostly a video format?
>Has anyone heard an audio-only setup that was impressive sounding?  I understand
>the Shure system and the new George Lucas/Thiel systems are supposed to be
>very impressive, but both are very expensive.  For the "lower-end" system,
>such as I auditioned, has anyone gotten good results with them?


There are two broad categories of surround sound:  surround encoding/decoding
and surround synthesis.

The systems you were listening to were probably surround decoders (or were
configured to decode surround instead of synthesize it).  Many movies are
encoded with a four-into-two system, where left, right, center, and surround
channels are encoded into the normal left and right stereo channels on a
laserdisc, video cassette, or broadcast.  This encoding is known as
"matrix" encoding, "Dolby Stereo", "Dolby Surround", and probably some
other names.  The key distinction is that four separate channels of sound
are created at the source and encoded into two channels with the intention
that the four channels will be extracted upon playback in a movie theatre
or home video system (although the encoding process doesn't interfere much
with normal stereo listening).

Surround *synthesis*, on the other hand, is performed on material which
has never been encoded for surround in the first place (CDs for example).
The intent here is usually to artificially reproduce various types 
of acoustic environments (e.g. stadium, a small medium or large auditorium,
jazz club, etc.)  this is achieved by calculating reflections and
reverberations and producing them through the surround speakers.  The
characteristics of specific environments may be determined by measurments
in the environment in question, or they can be computed dynamically using
theories of sound propagation.


Returning to surround decoding, there are two major aproaches to decoding
the encoded surround information, but the encoding process is always
the same.  The two decoding schemes are called Dolby surround (or Dolby
MP) and Dolby Pro-Logic.  Decoders which meet certain requirements and
specifications (and which pay the appropriate fee) are allowed to bear
the Dolby Surround or Dolby Pro-Logic logos, but there are proprietary
methods of decoding the surround information which are as good or better
than decoders which bear the Dolby logo.

Dolby Surround is the simpler and less effective of the two methods.
roughly, the surround channel consists of the difference between the
left-front and right-front channels (L-R), delayed by 20ms (more or less)
and put through a 7KHz low-pass filter.  There is no center channel
in conventional Dolby Surround.

Dolby Pro-Logic is a more sophisticated approach in which the predominant
sounds are "steered" between the four channels by dynamically cancelling
the sound out from the channels it's not supposed to be in.  A much
greater channel separation can be achieved in this way.

-- 
Steve Kanefsky             
kanefsky@umn-cs.cs.umn.edu