 Adventures
in Surround Sound, from 7.2 to Quad
(personal
and historical notes, basics, and acoustic realities often
forgotten)
= P
a r t 3 =
|
Jousting
at Windmills
An
"Infamous" Whistle-blowing Letter to
Billboard
Looking
back on episodes like what follows from today, I wonder what
ever possessed me. I must have been HATED for a few of the
things I did in good conscience, like this whistle-blowing.
Certainly you realize why Don Quixote is a poor role-model.
Try to tell that to a young, alert person who's out to:
"save truth, justice and the American Way..." Phooey. BTW,
Since Surround Sound 5.1 is really just Quadraphony with an
added front center speaker and a distinct sub woofer
channel, the experiences with quad are completely applicable
to the latest surround sound systems. Suddenly 30 year old
events are important and fresh all over again, and what
follows is not so much quaint "history" as it is "background
reading."
As
luck would have it, I was signed to an exclusive contract
with CBS Records (long before Sony took it over), when the
first Quadraphonic craze hit. There was a close connection
between CBS and Sony even back then, and Sony had (through a
long, devious route) become interested in a surround system
which CBS Laboratories in Stanford CT had fooled around with
in the late 60's, abandoning it eventually. This was an
interim, shortcut method to fit (encode) four channel audio
masters onto an ordinary two-track record, then reconstruct
the four (decode) on playback. Ben Bauer, an extremely
talented engineer and delightful person, had led his CBS
team through various alternatives in the late 60's, only to
decide none of them really worked. He recommended that the
company wait for a genuine four channel home delivery system
to evolve, that this wasn't it.
Poor Ben was surprised
a year or two later to get a call from an executive over at
"Black Rock", the CBS headquarters in NYC, asking about
"this new four channel sound stuff." The executive had just
been grilled by some of his contacts in Japan, who had found
this abandoned "matrix quad system" (as it was called), and
thought it looked like an easy way to expand sales. They
wanted the USA offices at CBS to assemble some prototype
recordings that used the system, and would in return send
the NYC offices some tests they had been trying out in
Japan. There was a buzz-on, and Ben was asked to comply. He
had already proven that all such bootstrap methods, trying
to get something for nothing, were doomed to failure. But
now he was going to be required to do it anyway, or at least
invent a few tricks that would satisfy the home office, and
which they could send to the Japan CBS offices. He hoped it
would then all fade away with that.
We got called into the
fray soon enough. Here we had one of the hottest classical
albums ever made, and it seemed a natural for a new
quadraphonic version. Rachel Elkind took a couple of long
calls from the heads at Masterworks division (no doubt John
McClure, perhaps also Tom Frost and Clive Davis), and
suddenly we were in trouble, too. I'd been making four
channel surround masters for a dozen years by then, and knew
a little about surround sound. We'd been the news bytes
about several of the majors, as they began pioneering this
newest home audio idea. We learned that JVC in Japan had
been developing a clever idea that actually *could* squeeze
all the necessary information into a conventional LP -- four
discrete channels.
JVC Japan was pushing
forward with their CD-4,
an honest, if complex quad method which was loosely based on
Jerry Minter's early Stereophonic LP's (two track stereo) of
1958: place the extra information as a super high frequency
tone that is FM modulated, more like radio than stylus in
groove records. Minter had taken the (mono) "sum" mix of
Left and Right, L+R,
as it was called, and recorded that in the usual way. He
took another mix, a "difference" of the two sides,
L-R
(
which means the R was phase-flipped 180 degrees and added to
normal L), and
modulated a 25 kHz tone with it, yielding the radio-like
signal. That was mixed with the mono sum to make the record.
Since 25 k is above most human hearing, you couldn't hear
this tone. But an ingenious, inexpensive add-on circuit
picked it up, detected it, and mixed it back in a simple
circuit (called a "matrix") to obtain the original two
tracks, L
and R.
Mono listeners just heard the mono mix (so it was
compatible). Clever idea. Anyway, JVC was doing this stunt
twice on a stereo LP, getting four distinct tracks from
it.
The good idea was never
trouble-free. Early on we were given some of their special
equipment and cut some albums using the cutting facilities
they'd set up in the USA. (It was amusing how several CD-4
doors opened after the letter below was published... ;^) The
JVC method was generally a bit noisy, prone to distortion,
and was delicate to install and operate. But it often
sounded quite good, too, when treated with some TLC. Sony
had tried and given up on such a high-strung design. Can't
say I blame them, it was a major engineering campaign for
JVC, and for RCA, their US affiliate on the new
"QuadraDisks." Sony/CBS instead continued with their
blood-from-a-stone pseudo-quad designs, much as Sansui, yet
another Japanese company, was doing at the time with
QS.
Ben Bauer came up with one of the best choices of a poor
lot, and SQ
was born.
A hasty appointment was
made, and Ben came to our studio downstairs in the
brownstone one afternoon in the Summer of 72, with a bright,
sharp engineer named Dan Gravereaux. They brought with them
the latest "encoding" and "decoding" equipment they had thus
far produced. We were given some copies of the very first
titles that CBS would be putting out in the new scheme, and
photocopies of a few detailed technical articles which
described the methods and history behind SQ. Their "job"
obviously, was to convince us to go along with their scheme.
The background scuttlebutt I've related above was only
learned a couple of years later. I felt bad for Ben, who was
obviously such a nice, urbane man, with great charm and
knowledge.
But I also felt bad for
us. We spent the next several weeks trying to get what we
could from the SQ hardware. It was pretty gruesome. I guess
for many producers and engineers SQ was adequate. It could
handle a kind of ping-ping pong-pong stereo, as I called it,
as long as you simply pan-potted a few locations around the
periphery. It was impossible to have natural or simulated
instrument leakage: the same sound heard over more than one
or two channels. A ghost center effect, something I'd used
for years, was out of the question. It would gather up on
the final Left track, and cancel out on the final Right
track. Other combinations, like diagonal splits, were even
worse. Baloney!
|
A
Saber-cut to the Heart of the
Matter
Out of
curiosity and pique, I came up with an amusing,
barbed demonstration track in quad that made our
point with a razor-sharp sabre. The track sounded
like a few minutes of a large group of people at a
cocktail party, yatta-yatta-ing away all around the
room, recorded cleanly in discrete quadraphony,
hard to follow any one conversation. BUT -- when
this clever little "nasty" was fed into the SQ
encoder something strange happened: nearly all of
the voices slowly faded away into a soft background
sputtering, leaving but one voice that could be
heard! And that was the voice of our good friend,
Bob Schwarz. He cheerfully deadpanned, in his
wonderfully rich radio announcer's voice: "Hi, are
you enjoying the party? Me, too. But where did
everyone go? That's odd, I couldn't hear myself
there in the discrete. But now I can on the matrix.
There must be something funny going on with these
matrix systems, don't you think?" Then everyone
else's chatter faded back in. Deathless silence
from the CBS people we played it for. Priceless
moment. (Also stupidly naive and showoff-y, not
something that would ever be mentioned in a good
class on Diplomacy... ;^)
Sansui QS had
the same Achilles Heel, also the older pioneering
RM. All
of the Matrix-Quad systems do, for they all toss
away half of the information. It's referred to in
the original 1972 letter below.
|
We knew CBS's company
plan, and we had just mixed our S-OB multitrack tapes once
again to real surround, before getting the new equipment.
There were a great many "tricks" we would have to use to
cover my "impertinent questions," to continue doing nothing
more demanding than what I'd already been doing for years.
These tricks were conscious, deliberate clever workarounds,
and you really had to think twice at every step about what
you were doing, and what would get translated reasonably
well, and what wouldn't. Or just try to be satisfied with
the old Double-Ping-Pong, as we'll describe next. This is
not the place to go into these additional SQ-encoding
kludges. I'll post some of the details eventually, some
scans of my notes and diagrams that were necessary if you
were to avoid getting bitten by the severe creative and
musical compromises that SQ would require.
(Note:
the first few diagrams are already uploaded and ready to
peruse on
the next surround page.)
I found Sansui's QS
scheme somewhat more to my liking. It suffered somewhat in
front separation compared to the CBS, but when the two-track
versions encoded by Sansui's QS system were played back on
ordinary stereo equipment the results were very close to the
way Rachel and I had been making two track stereo masters
from all our four track masters for a few years. CBS's SQ
design was just plain "weird" when heard on normal stereo,
except if you restricted the placements to normal two tracks
up in front, like close mikes over a band or orchestra,
while the "rear" channels were distant reverb hall-sound
channels. For that it was fine. Also Enoch-Light variations
on unsubtle ping-ping pong-pong mixes worked well,
especially on the latest "Logic" decoders, which rode gain
automatically, trying to enhance the miserable separation of
all matrix designs. Light would place antiphonal instruments
only in the exact four channels, not much in-between, and
that kind of limited quad was reasonably well served by the
matrix systems with logic steering.
CBS had no sooner put
out our "Switched-On Bach" in an SQ edition, that we were
able to hear it on their best decoder. The result was
depressing, very much a warped joke version of what we'd
made. Well, it was on all but one
track: the Two-part
Invention in F. That one worked quite well. Why? Because it
had come from the only mono master on S-OB, the first track
recorded (I had "stereoized it for the first stereo
release). So we were forced to use pan-pots to walk and jump
the mono Invention all around the room, to the exact four
speakers, not much in between. It's effective in small
doses, I guess (the Invention is short), and quite a few
people wrote rave reviews about the SQ version. I wanted
nothing more to do with it. CBS refused to use the
"competition's system", the JVC QuadraDisk, which had been
improving steadily. I
was signed to the wrong company!
(On
the other hand, to be perfectly fair, the JVC system had
great trouble handling many high frequency sounds on
Sonic Seasonings and some of our other masters
without audible distortion. We cut a few careful tests on
CD-4 and had to throw in the towel -- our music was just too
demanding -- ouch! You'll detect that we'd decided to wait
it out by the second letter below. And, well like, it's only
taken a quarter century! ;o)
If we'd signed with RCA
initially, as we nearly did, things would have worked out
differently. We'd have put out many of our albums in genuine
quad (discrete, meaning honest four ins and four outs), and
neither of these letters would have been written. Then it
would have been upsetting when CD's first were introdiced an
absolutely NO ONE took advantage of the four channel format
they offer (still do). Yep, a plain old CD can store about
half stereo's maximum time, or 38 minutes of pure quad! Bet
you didn't know that before -- it's never been implemented
on CD-players or in the studio, to the best of my knowledge!
Stuck between greed or honesty, we took the path that many
(most?) artists probably would not: we cut off extra
royalties from sales of SQ versions, and demanded our
pseud-quad S-OB be withdrawn! This was a big financial
sacrifice for us, just a small studio with not many artists,
but it was the only ethical thing we could do.
There might be poor
souls out there who would think the mild chewing up of S-OB
by SQ was "the way it was supposed to sound." We wanted no
part of deceiving the loyal fans we depend upon. I did the
same sort of thing when the early "copy protection" schemes
for digital (ca. 1986) appeared. They wanted to slip a deep
notch filter into all CD's, with nasty results to the music
(a few high piano notes would nearly disappear, fer pete's
sake!). I was depressed by how few other artists signed the
protest petition circulated widely at the time through the
major recording organizations. Is greed just "the Amuhrican
Way?" (Have mentioned this before, on the Disknotes
page.)
Anyway, in 1972 I wrote
off an infuriated letter, reprinted below, warts and all.
(Lordy, the execs at CBS and long suffering Ben must have
been apoplectic
about it!) Billboard magazine had been running an excellent
coverage of the ongoing "matrix vs. discrete" debates, so
off to them went this "letter to the editor." Don't know
what their editor made of it, but he printed it, one of the
longest they've ever published, a full page. I've been told
it created quite a "tempest in a teapot," and helped damp a
bit of the Wave of BS that flowed out of the matrix fiasco.
I hope so, I really do hope so...
|
(Top
of the Page)
|
Moog
Soundings
(the
title the editor came up with for my
letter...)
Dear Sir,
Thank you for your particularly comprehensive and clearly
unbiased reporting especially with regards to quadraphonic
sound. As one person involved in quadraphonic sound ever
since the technology became practical, I have been excited
to see the general interest now rapidly growing in this
method of "super stereo." But no one in our industry can be
anything but apprehensive if not confused about the many
contradictions, ridiculous claims, putdowns and hysterical
confounding of facts that has made mockery of all the recent
quadraphonic meetings, public exhibitions, publicity and
press conferences.
The crux of all confusion seems best indicated by the
so-called "Great Matrix Debate." Most of the reports you
have presented over the past year or more have centered on
that kind of "my system's better than yours" game. Well and
good. The business and creative elements of the industry are
perhaps Billboard Magazine's primary concern. And this group
has a right and a need to be told about the technology
available, in not overly technical a manner.
It is no secret that we all stand to gain a great deal once
successful quadraphonic hardware becomes standardized. But
nowhere has anyone really attempted to state a few simple
generalities which, like it or not, ought govern our choice.
Make no mistake, the choice for quadraphonic systems is ours
not the technical developers and laboratories currently
engaged in this sort of research. Whatever we, in fact,
adopt to promote, build home units for, use to produce our
records, our tapes, our artistic sound paintings i.e., the
"software" of quadraphonic will become the system. All other
systems will then phase out, deservedly or not.
Pioneers
If we may look back to the
similar birth of stereo in 1958, it was the few bold
pioneers: record companies, producers, and phono cartridge
manufacturers, who literally forced the standardization of
the Westrex 45/45 stereo-disk system. Otherwise we would
still be debating the theoretical impossibility of this
system, and, as some now joke back to mono--we would still
be a non-stereo industry, likely much smaller than now
(thanks to the stereo revolution).
If I may be permitted an opinion, the only present need we
have is as that 1958 period, a workable system. As long as
it possesses no unnecessary theoretical limitations, but
only has a few "bugs" (perhaps lower level, slightly less
playing time, and the like), we ought adopt it as we did in
1958. A few years will iron out those bugs, again as we
discovered with stereo.
Unfortunately, most quadraphonic pioneers today are not
willing to accept this small price for a very healthy
future. They believe, many quite sincerely, that we can
"boot-strap" ourselves into instant quadraphonic.
Humbug!
When Rachel Elkind and I began our new "Sonic Seasonings"
album, we planned for quadraphonic and recorded all the
materials in quadraphonic. That master, like "Clockwork
Orange" and most of our other product, is already mixed in
four channel surround. We tried to process this master on
all the known matrix systems, and a few not so known. I am
most unhappy to report that the results were catastrophic
most of the time, and ho-hum for the rest. And this was
using the latest state-of-the-art matrix equipment, a
magnitude better than home matrix equipment. Our
"Switched-On Bach" was released in the best of the matrix
systems, CBS's SQ, and we later discovered that, despite
some critical acclaim, it is a pale mirror of the
quadraphonic master. Worse, the musical balances are
irrevocably bastardized so that, at many times, solo lines
are obliterated by accompaniment.
Columbia has generously agreed to withdraw this album. If
you should come across any remaining copies of the SQ
version, avoid it like the plague! -- a strange
sentiment for profit consciousness, but in the long run we
believe it is the only valid decision possible. (Please
note, re: the recent plea for compatible discrete/matrix
disks, they would be exactly as inferior on all non-discrete
equipment. Let's not allow ourselves to be conned by glib
claims to the contrary for this ridiculous comprise). No
other TEMPI product will be marketed in quadraphonic now for
a while until a non-matrix system is accepted as an industry
standard. Perhaps the JVC/RCA carrier disk is the answer. It
is a "workable system" at least.
Discrete
Admittedly some product on
matrix disks sounds perfectly fine. Indeed, a master remixed
for a "ping- ping-pong-pong" quadraphonic will, in general,
produce acceptable results on most of these systems (which
despite mathematical differences tend to sound very alike).
The strong loss of separation and phasing alterations/
interactions of all these systems is universally
acknowledged. One reads with disbelief in your pages the
number of people who rationalize such alterations as
desirable. The dilutions do make a ping-ping-pong-pong
master sound more diffuse and less gimmickry than the same
would sound in pure quad. This latter we now find termed
"discrete."
For "discrete" product to sound natural and acceptable to
most people, we will all have to learn as we did in stereo,
to produce master 4-track tapes with ambience,
cross-relations and shifting phases of all kinds. There are,
unlike successful matrix masters, many families of mixes
possible which will attract even the most naive consumer,
who, let's admit, has been less than excited by any of the
matrix systems. Oddly enough, the ingenious matrix systems
could play an important role in allowing producers to mix
discrete masters with far more directionalities than the
ordinary recording studio quadraphonic console permits. We
can all profit in the end.
Possibilities
So I come across as another
one of those mad discrete thickheads. I'm sorry, let's look
at the reasons. You need not agree with me in a decision to
wait for discrete and may prefer to go with a stop-gap
measure of matrix quad. Again, well and good. Pragmatically
you are probably safe. But, people, let's not "cut off our
noses to..."
1) The theoretical maximum separation for a symmetrical
matrix system is 4.8 dB; none attain it. Most barely attain
a 3 dB. If we wish to, we can trade-off left-to-right
separation for front-to-back or vice versa, as several
systems do. But after all the effort the industry has
expended for the 25-35 dB separation of stereo, how can we
now rationalize being forever happy with 3-5 dB?
2. A two channel stereo system with about 4.5 dB separation
is scarcely different-sounding from mono on two
loudspeakers--unless the listener is in the exact center (as
on earphones). Then the stereo effect is just noticeable.
Don't take my word, try it next time you are in a studio.
Have the engineer set up a cross-mix of a stereo master, any
master, to give 4.5 dB separation. Then move about a little
and listen. Compare it to a mono tape on the two
speakers.
3) Matrix (forgetting the mathematical type) means "system
of intermixes." Three signals can be mixed in several ways
to give three new signals. These new signals can then be
re-intermixed to produce the original three unless we throw
out one or more. The 4-2-4 matrix for quadraphonic discards
half of the new signals. No decoder is possible in
this case. That's a misnomer. Re-encoder would be the
correct terminology.
4) Most consumers at home would object to listening to
records from one fixed spot, and no one else in the room
would occupy it at the same time anyway.... a shift of a few
feet causes a measurable change of loudness of sounds from a
loudspeaker. For quadraphonic, the balance shifts more than
the available matrix separation, i.e., despite smokescreen
claims to the contrary, even an educated ear could not tell
matrix quadraphonic from mono over four speakers with
minor exceptions, except from the center of the room. N.B.
with well mixed discrete just about anywhere in the room is
usable.
5) In a few systems, like the basic SQ, the left-to-right
separation is strongly favored, while front-to-back
deteriorates to an essentially inaudible amount. These
systems cannot be differentiated from stereo over
four speakers except from the center of the room, again with
minor exceptions.
6) Clearly, unless we are willing to depend on the ignorance
of the consumer, which is indeed a sad fact, some enhancing
logic is absolutely essential for matrix systems. Thanks to
the kind help of CBS Laboratories, we at TEMPI were given a
long opportunity to hear and work with the latest Logic-SQ
equipment. Compared to all other systems, including SQ
without logic, it is the only system which even
begins to sound like a quadraphonic master, for some
material. Our "Switched-on Bach" SQ disk still was awful
this on equipment far superior to all available home
systems.
7) We investigated the cause of the mysterious missing parts
for "S-OB." It turns out, and this has never been in print
before, that every matrix quad system has an infinite
number of signal combinations which cancel out when the
matrix master is encoded, and can never again be
recovered.
8) To prove this important point, we produced several
quadraphonic mixes which vanished when encoded,
leaving only a very soft sputtering! The SQ was by far the
most tricky matrix to find such complete examples for, but
it too, succumbed. Imagine never knowing just what part of a
meticulous mix will be lopped off, or severely attenuated,
by the time it gets to disk. This seems eminently more
important than any position- shifts that may occur.
9) If one is cautious, he can avoid these troublesome
combinations. We already know that pan-pot and ping-
pong-ping-pong are safe, if artistically limiting. But they
do work on sophisticated systems, such as SQ with logic.
Other systems are less fussy, but sound so like mono in the
end that you might well ask: "is this trip really
necessary?" and would you mind giving up 80 percent of
possible quadraphonic effects permanently?
10) A highly sophisticated logic is already on the drawing
board stage at CBS Labs and others. By breaking up the sound
into say, octaves, and using logic on each band separately
(not unlike Dolby-A in concept and certainly cost) a result
indistinguishable from discrete quad 95 percent of the time
is theoretically possible. Of course, a critical listener
may still be annoyed by the "pumping" effect inevitable on
gain riding devices as logic requires.
11) The previously mentioned cancellations during making the
matrix record cannot be removed, how ever. Progress in
quadraphonic recording and mixing will be severely limited.
I, for one, would prefer not to have to carry a calculator
and vector scope, or an encoder / decoder pair around with
me, to cheek out feasible projects (and also worry about
additional cancellations in mono playback). The quadraphonic
masters for "S-OB" and "Sonic Seasonings," "Clockwork
Orange," etc., do not encode properly because we
refused to limit ourselves in these ways. The phase and
amplitude shifts that make sophisticated quadraphonic
possible here work to confuse and disable encoding. And what
you can't encode never gets to disk so the new logic systems
are no help.
12) Finally, these logic schemes, clever as they are,
become, in fact, more complicated and expensive than the
carrier system decoders so put-down with heated prejudice by
the many individuals Billboard has faithfully reported on.
Without committing myself to the obvious front-runner of
carrier systems, JVC/ RCA, may I add:
a) The complaints of new cartridges necessary, less playing
time, lower signal level, etc., are so similar to 1958's
anti-Westrex propagandizement that one wonders what the
commotion is all about. A lot of current high-quality stereo
cartridges work fine, and that's more than the mono
cartridges of 1958 could do for stereo. Ironically, lower
signal levels are also heard on most matrix disks,
due to asymmetrical level peaks from the matrix (which
cannot be inconspicuously limited during cutting) so the
overall level cut is reduced. In 1958, the same complaint
was made about stereo, don't forget.
b) CBS Labs originally developed a carrier disk. It was
aborted temporarily due to the limitations of
available plastics and other reasons. The newest record
materials already are not affected by even fairly abusive
playing on cheap phonographs. A carrier system then is
possible. Dirt, not wear, may cause trouble. If you think
you've worn out the carrier, just reach for the record
cleaner! In any ease, again this is all old crying a la
1958, and the situation is nowhere near as bleak as Westrex
45/45 seemed. Then there's still the option of converting
the Teldec video disk to a quadraphonic disk of four or more
hours per side!
e) Actually, even the carrier systems are "matrix" systems.
The two left channels are mixed, ditto for the right. But
two other mixes are made for the carriers to hold. In not
discarding two of the four new signals carrier disks gain
their superiority. Since no information is lost, the
cancellations mentioned earlier do not occur and one can
record with freedom.
d) Some have apparently heard discrete product of poorly
mixed masters. May I point out that the gimmicky quad
pinging and ponging is no more inherent than it was in early
stereo days. And the "blend" control of those days was
quickly abandoned as more sophisticated records were made.
Is that not analogous to the present cry that "matrix is
more natural?" It's a great big BLEND switch! Still, there
is a place for blending in quadraphonic, and if a producer
finds a particular matrix quad system provides a pleasing
"surround" on say, the string tracks, I see no reason why
s/he ought not use it. And a different system might be used
for the echo signals, while the rhythm and vocal might be
the best pinpointed in direct non matrix form. With a
discrete release available we can have the best of all
worlds! Any matrix blending will be done in the recording
studio, under artistic control of the artists, producers and
engineers. But a permanent blend of all signals
indiscriminately at home? I don't think we'll need it and
certainly not want it in a couple of years. Until then,
caveat emptor!
Sincerely,
W. Carlos
New York City
==
August 5, 1972; Billboard Magazine, Page 6 ==
(Note:
In the same issue of Billboard magazine,
Brad Miller, of the Mystic
Moods Orchestra,
placed a five-inch tall ad drawing reader's
attention
to the above letter, with big letters:
"IT'S
ABOUT TIME THE TAIL
STOPPED WAGGING THE
DOG!"
(Brad
had been another surround sound enthusiast and
pioneer.
He also championed a very reasonable system of
Quadracasting
four discrete channels over an ordinary FM
transmitter. The FCC
hemmed and hawed, and Quad faded away. It could
have all come to
fruition a quarter of a century ago! Perhaps now
5.1 will carry on.
Thank
you, dear Brad, wherever you are, for
your constant support and
understanding...)
|
|
(Top
of the Page)
|
In
May of 1974 there had been some change of the status of
Quadraphony, which was still very much in the news of
industry record and engineering magazines and the like, if
not in the perceptions of the public. I tried to summarize
the situation in a much shorter new letter, again to
Billboard. I don't remember if this one was also printed by
them, as I found only a Xerox copy of a typed manuscript in
my files. We clearly had made the copy and then sent the
original off to them. It's quite possible that by then the
initial interest had waned, and the letter was never
published before this web page.
It
still contains some anger, and youthful hubris but not so
much as the first one. You'll see that by now I'd done my
homework, and could toss-off a better grab-bag of terms,
exact specifications, and evidence. It amused me, when I
just found the photocopy in a dusty old file folder, to
discover that way back then I was pushing for the same kind
of configuration that I had just documented on the earlier
pages. Can't say I'm not consistent. (You'll also notice
that I end with an irrelevant "dig" about the way surround
multi channels are often labeled. On my old console A-B-C
and D correspond to the far left sweeping through far right
outputs, nice and symmetrical. Some of the setups I've seen
are Byzantine, like the way it's done for DTS masters, say
wha...?) Of course the whole surround sound vehicle never
completely got off the ground before now. So these issues
have become timely again in 2001, some 27-28 years later!
(The
more things change, the more...)
For historical sake, I'll include the second Billboard
letter below.
|
More
Switched-On Quad
Dear Sir,
Back in mid 1972 you graciously allowed me publication of a
rather lengthy letter about the then state-of-the-art of
quad sound. This continuation of that filibuster will be
briefer, but a lot has happened in the meantime which ought
be said. since we have yet to come up with a standardized
name like: Stereophonic, or Monophonic, and although I
personally prefer: Quadraphonic to the others, Quadrasonic,
Quadriphonic, Quadrasound, Tetraphony, etc., let's for now
simply call it by the informal: "Quad."
Frankly, I think all of
us in our industry really do
deserve some sort of recognition for sobriety or altruism.
By 1972 and into 1873 there was still a lot of hyped
advertising with exorbitant, exaggerated, confusing, and
often untrue claims made about one particular system over
all others. There is still some of this nonsense going on,
but by and large the various manufacturers and sellers of
quad hardware and software have mellowed into more refined,
objective ideas in their promotions. The truths about this
marvelous new medium are at last being heard. Integrity has
won out, while behind the scenes, a great deal of research
and development is producing valid breakthroughs now
reaching the marketplace.
The
Situation in 1972
Two years ago there were
several Quad methods under discussion, and at that time
none
of these systems was really very good. In comparisons of
identical recordings we made, under carefully controlled
conditions and double-blind, both electronic and acoustic
music, we heard how the various systems altered and
corrupted our masters. It was like: SQ versus Switched-On
Bach. Some comparisons:
RM
Matrix: produced
very decent Stereo playback, indistinguishable from stereo
Quad playback, and occasionally lousy Mono playback. Much
information was lost during encoding.
RM
Vario-Matrix:
occasionally was fine for its intended quasi-quad, but
ambience and balances were poor, and much information was
lost.
SQ
Matrix: produced
mediocre Stereo playback, (with both rear channels
"folded-in" to equally ambiguous center-fit 1) ,
indistinguishable from-stereo Quad Play and usually lousy
Mono Play, and much information was lost.
SQ
Logic: occasionally
was fine for its intended Simulated Quad Play, but ambience
and balances were poor, the first logics were slow-acting
and "pumped" (newer are much better), but again much
information was lost.
CD-4:
produced very decent Stereo and Mono Plays, but quite noisy
and often distorted Quad Play, which frequently had a
muffled quality, although no actual information was
lost.
In general, all other matrix schemes were even worse than RM
(Sansui originally called it QS) or SQ in most respects' and
CD-4 was the only available non-matrix schemes. All the quad
disks had somewhat lower levels than standard Stereo discs.
The CD-4 at that time had theoretical reasons for the
reduced level, and both SQ and RM have frequent out-of-phase
peaks (rather higher than normal stereo) which require a
more cautious cutting level. Audibly, despite what the V.U.
meters might have said, they all sounded softer than Stereo,
in any event (with a few rare exceptions):
So, at that time, I felt the safest decision was to develop
a "super-stereo" mix-down from our quad masters, and sit out
and wait until videodisk technology suitably modified, or
CD-4 type records yielded results which could be called
'high- fidelity" as well as Quad. Also, since the matrix
theory as a whole permanently discards one-half of the Quad
material, I believe it is actually Pseudo-Quad. Perhaps
"discrete" quad was originally a term dreamed up to hide the
fact that "discrete" is really the only genuine Quad. I
second the suggestion that we rename discrete: "true Quad",
and matrix: "Simulated Quad", not unlike the old ruling
about mono masters "rechanneled for stereo".
The
Present (New Developments)
But now it is 1974 and the
situation has changed A few "Simulated Quad" methods,
notably SQ, have developed special mixing console adapters/
rechannelers which prevent any mixing combination which
would cancel out (partially or wholly) -in matrix-encoding
from ever being produced. The producer and engineer can call
for all the usual positionings, echo returns, and the hike'
but only the SQ-safe ones will go through "directly". The
others will in fact be altered from the intended positions,
and added along into the mix with no level drop. The
matrix-limitations are still there, perpetual motion still
doesn't exist, but at Least you no longer need worry about
that. Simultaneously a "true Quad" master can be made on a
four track recorder, for Q8 cartridges, CD-4 disks, or any
other discrete release now or in the future. All in all,
it's a most commendable effort, especially for simplified
and non-critical applications.
Much more exciting is the discovery that the one preexisting
"true Quad" method, CD-4, has not been sitting idly by.
Since in theory this remains a workable
system, it was
"only" a matter of time & dedication, genius, money and
effort before really High Fidelity Quad records (and we all
want those) could be cut. The surprise for us all is that
these goals have recently been attained. Although CD-4
presently remains a delicate affair, with top-quality
equipment and critical adjustments required at ale points in
the chain, it works very well under good conditions.
In March Rachel Elkind, my partner, paid a visit to Tom
Nishida at the west coast "JVC Cutting Center". She brought
along a new quad master she had produced recently, and I
engineered of a very demanding piece of music by Eric Siday.
He utilizes traditional and electronic sources together for
genuinely exciting results that would place severe demands
on any system. In our A/B comparison of the test disk
Mr. Nishida cut for Rachel, it was next-to-impossible for us
to be certain which we were listening to: master or disk.
Only the presence of slight surface noises and a couple of
clipped peaks gave any indication. At last even Eric was
convinced. And he had found the A/B comparisons with the
leading matrix system dismally easy to differentiate, so
unfaithful was it.
We also have learned that CBS has taken delivery of CD-4
equipment. Quietly, this giant too, is experimenting with
perfecting a high fidelity "true quad" record, as are many
other companies. With the newest modifications (several
major; a few minor but with great improvements, such as new
cutting stylus shapes) it is at least conceivable that
records worthy of the small but not unimpressive audiophile
quad market could be in stores by Christmas. The not so
critical lower-priced phonograph users would still be best
served by that new SQ-mixdown type Pseudo-Quad record, at
least until less delicate and expensive Hi-Fi "true quad"
hardware is developed.
But don't throw away your old matrix encoders yet - or ever!
As I mentioned in my last letter, these devices borrow two
fantastic concepts developed back in the late 50's by
pioneer Ben Bauer: 90° phase shifters and matrices. And
those concepts, if used in unorthodox groupings, will be
significantly important for those of us who wish to push
open the limits on our new generation of 'True Quad. In a
very real sense we all are the winners after only a few
short years now passed.
Speaker
Placement as a Limitation
With the dawn of "true
Quad" presently before us it might be appropriate to suggest
one more relevant observation. From 1961 to 65 I was
involved with a small group of experimenters working with
the then unnamed and primitive four channel techniques out
of which our present "Quad" developed. We must have tried
just about every conceivable microphone and loudspeaker
placement during those years. It ;s ironic to look back now
and realize that one of the first tried and least successful
was our dear friend. one speaker in each corner -- 360
degree surround-sound. It would be out of place to get into
a technical discussion as to why this idea which really
sounds "obvious" and "natural" was not so optimum for
human-style two ear listening. What is important is that
there are a great many placements which give the aural
illusion of 360 degree sound, and most of these do not have
speakers physically behind the listener'
Actually, those who have lived with quad speakers in each
corner now must realize that it gives us four of those old
stereo bugaboos. "hole-in-the-middle". Instruments and
voices simply refuse to blend with 90° separation
between mikes and speakers (mainly the latter). I won't
pretend that the final placement we preferred back in the
early 60's is the ultimate answer? But many people are using
it now in their homes and studios: four speakers in a deeply
curved 180 arc, or 60 degrees between each -- like an old
Cinerama screen! But the most fascinating thing is that it
not only eliminates those four holes-in- the-middle' but,
with no particularly involved mixing, it gives a completely
convincing illusion of sounds Behind, To the Sides, Above,
Below, Near, Far -- a solid curtain of sound!
Speaking from personal experience, Rachel and I have, in the
five years of records between our "Switched-On Bach I &
II" returned to this deep-curve speaker configuration. we
audition all quad tapes on it, we mix to it, we use it to
show off quad for our friends and business associates.
Though this is a limited sample group, we have found:
1) Everything
is easier to locate, including mistakes (which one can fix
before a master leaves the studio).
2) We have been able to record a live orchestra on all four
channels (not echo-only on rears) without an unnatural sense
of being in the middle. The reverb seems to come from all
around, even behind, however, yet the placement of
each instrument or section of instruments is unbelievably
well-defined in a three dimensional space unattainable with
either stereo or one speaker per corner quad.
3) For less traditional records it is not difficult to give
the convincing effect of instruments all around you,
including dead behind and ahead, or directly at each side.
All of these are exceedingly poor on the standard quad
speaker configuration.
As farther evidence of 1), listen to any matrix quad,
especially one that uses logic, using the 180 degree arc of
monitors. The old limitations, pumping, instabilities, and
the rest are at once exposed to the ear. In a way this
improved speaker configuration really requires that truly
discrete recordings be used.
Forgive the cliche: "Try it, you'll like it!" But I can't
help but recommend that we begin mixing for and playing quad
for our customers using this simple variation and watch us
all benefit the great results. For those who are fond of
naming things, we'll want to call the channels: Left Side,
Left Front, Right Front, Right Side, all in a row, or as
simple as A-B-C-D!
Sincerely,
W. Carlos
New York City
==
May 10, 1974; Sent to the Editor, Billboard Magazine
==
|
(Top
of the Page)
©
Copyright 2001 Wendy Carlos -- All Rights
Reserved.
Back to the Wendy Carlos Home Page
Go
on to Pt. IV