Adventures in Surround Sound, from 7.2 to Quad 
(personal and historical notes, basics, and acoustic realities often forgotten)
= P a r t  1 =

Nomenclature and Full 7.2 channel setup
The Main 5.2 Channels
  Digression I -- Classic Blunders
  Digression II -- Listening Tests
  Digression III -- Surround Options
5.1 Surround Sound for Music
5.1 Surround Sound for Films
Optimum Quadraphony
Quadraphonic Folly
"Jousting at Windmills"
An Infamous Letter to the Editor
The follow-up Letter to the Editor
Something about Matrix "Quad"
Psi-Networks (the secret ingredient)
Recording in Surround
Quadraphonic Recording
Multichannel Recording Quirks
Shoulders To Stand On (Bert Whyte)
Bibliography -- References

back Go to the wendycarlos.com Homepage

(Note: all the images below will open a large view in a new window when you click on them. To return to the text, just close the new window.)

5.2 studio
5.2 Channel Surround Mixing Studio
(click image for a huge view!)

Finally it seems to be happening! In 2001 we don't yet have Hal (check back in another 100 years ;^), but we do have a distinct buzz-on about Surround Sound -- for film soundtracks, DVD's, and for music creation and mixing, as the new DVD-A standard is designed to implement. To me it seems like it's taken forever. I'd nearly given up hope that a practical surround sound system would reach the public in my lifetime, anyway. Those of us who lived through the big Quad Boom and Bust of the 70's are gun shy, expecting another stillborn standard, based more on hype than reality, and something valuable gained for effort expended. Just a few weeks ago I checked in again on what's become available online in web pages around the globe. Well, welly, there's a good representative amount of information starting to appear already -- on Quad, 5.1 (five full-range channels and one sub woofer with one tenth the range, or ".1", total = 5.1), and several other options. Yeay, this is a healthy sign! Could it be?!

(Note: This next section contains an historical note on my own first encounters with surround sound. Click HERE to skip forward to some of the basics on surround audio, which we'll be discussing on these pages.)

Okay, I have reason to be more skeptical than most of you reading this. My first experimentation with surround sound took place way back when I was still in college, studying music composition and physics. For me, surround sound predates the Moog Synthesizer. At that time there was no technology one could readily purchase to do more than the same old two-channel Stereophonic Sound that seems to be going on, like forever. Of course just TWO tracks was big news those days. So I had to build my own first "quad" tape recorder. Four channels, recorded on four tiny tracks, using two quarter-track tape heads in what we'd call a "semi-staggered" array. The hardware was from Viking of Minneapolis, bless them. They allowed even a very financially challenged student to save and purchase some very practical tools with which to record and playback music and sounds. I had to find a way to synchronize the four bias oscillators, and also constructed (from scratch) a sturdy wooden enclosure to mount it all in. It had a handle on it (since broken off), so it was "portable." At 45 pounds, I leave it to you to decide how realistic this description was.


Custom Viking Four-Channel Tape Recorder

Above you can see it with the cover removed. I was astonished how good it still looked when I discovered it in my parent's basement some dozen years ago. I've cleaned, reworked and adjusted it, gotten it to work well again, another surprise. This is the machine that I made my first multichannel recordings on. I took it with me to several concerts given in Providence and at Brown University, and made quite a few "amateur" surround recordings, experimenting with microphone and speaker placement, since there were few to no books on the subject. I learned a lot about what works and what doesn't by uninhibitedly trying every crazy idea out for myself. My early electronic acoustic music compositions were created with the custom Viking, and when I came to New York City to Columbia for Graduate Work in composition it came along with me, need it or no!
But by then I had begun to use Ampex professional tape machines. Peter M. Downes, a good older friend who made custom recordings in the Providence area, generously let me borrow his 2-tk Ampex 351 and Magnecorder for one entire summer, to create the sounds for Episodes for Piano and Electronic Sound. My four-track Viking was used on that work, too. But the prestigious Columbia-Princeton Electronic Music Center had many professional Ampex machines, including three (!) that made me drool: 1/2" four-track 300-4's -- cool! "How ya' gonna keep 'em down on the farm," I learned quickly how to use these sturdier, better sounding tools, and the little Viking sat unused most of the rest of the time, except to record a few more live concerts. Later it was moved back to my parent's house when I relocated, and I forgot about it for nearly 25 years. Hey, there were new "toys" to explore!


McMillin (now Miller) Theater's 13-Channel Surround System

One of those "toys" was not so much a device as it was an idea: multi channel surround sound. As the luck of timing would have it, my favorite professor, composer Vladimir Ussachevsky, had recently designed and installed a wonderful new sound system in Columbia University's McMillin Auditorium (as it was then called). The diagram above is a plan of the auditorium, showing in red the 13 speaker channels that had been mounted and wired into a unique installation. I still drool about the wonders one could produce at large scale in the new field of multidirectional audio. There are actually 19 speakers, as the balcony interfered with producing sound at both levels from once source apiece. So channels 1, 2, 8, 9, 10 and 11 required two speakers each, one upstairs, the other down (which are superimposed in this plan view). The rest are single speakers per channel. There are also two, #12 and #13, that were mounted up on the ceiling, facing down! The KLH loudspeakers for channels 4, 5, and 6 were stored backstage, and had to be brought out when needed, then positioned as shown (connectors were nearby). One oversight: there should have been two more, above the exit doors (mid-wall between #1 and 2, and also #8 and 9), at the exact sides. Live and learn.
It was all fed from a small room located near the upper speaker 1, which contained sturdy metal shelving with an appropriately large number of Dynaco power amps, a Stereo-70 for the double-spkr channels, Mono-60's for the rest (got pretty hot in there!). Tie-lines led down to the small electronic music studio, Room 106, in which I composed most of my electronic music as a student (it's now used as an office). The studio contained 5 to 8 Ampex tape machines at any one time, the outputs of which could be fed out to the hall's system. I continued my experimentation with surround sound, finding out what worked as planned, and the many more ideas that simply didn't work. A good "woodshedding experience", I learned a lot, and had a lot of fun with it, as you might imagine!

(Top of the Page)
Nomenclature and Full 7.2 Monitoring

Since the 60's I've been using four or more channels on the mixdowns of most of my performances and compositions. It's been a life's desire to get some of this surround music into the hands of music listeners. And that may very well be happening soon. I also discover I've accumulated quite a few "barnacles on the hull" from working in multichannel sound all these decades. I'd like to scrape some of these off onto the next generation to figure out what the $%#* to do with some of them! That was the major motivation for creating this web location. We'll be referring to speaker arrangements (very important, that) and the several output channels a lot on these pages. So let's show the near-standard labels we'll be using. Here's the same image at the top, smaller and with labels in red pasted over the front of each speaker. The two subwoofers are down below this view, on the studio floor one step below the level of this shot, and so we've just positioned arrows that show where they physically are located. There's nothing surprising going on here, but we wanted to define our terms clearly.


Speaker Locations, 5.2 Channels

Actually, this view with labels does not fully describe my studio's monitoring setup. (Please note: there's a good 12' between the back of the console showing at the center bottom, and the old 45" video monitor, the C speaker on top, right in between LF and RF. This Cinerama-like WA view "squishes" that distance together, while it also slightly exaggerates the space between LS and LF, RS and RF.) There are four more speakers that are not seen in this angle, driven by another two channels of amplification. These are located to the rear on both sides of the mixing space, where they form a blurry impression of diffuse information behind you and to the sides, surround channel information. I've been using some modest Pinnacle speakers and a small stereo amp for this job, as all surround information of this kind is deliberately narrower, in frequency range and dynamics, than what the other channels reproduce. For DVD or LaserDisk playback, the "rear" information from either Dolby Surround or discrete 5.1 tracks is fed to these channels, as well as some mixed to LS and RS.
But for music mixing, these small rear speakers are driven by auxillary channels, usually ambience and antiphonal parts, or processed reverberation and echo effects. When used, the extra channels raise the total channel count to 7.2. That creates a very impressive soundfield, you bet, and regularly astonishes visitors here who've never heard that many channels before! Since 7.2 is really just an extension of 5.1, we'll handle the latter on this web page. Just bear in mind that it's likely at one time or another that you may encounter another two or more channels, and that these fully "behind you" channels are not as important as the other five plus. You can create a similar directionality by manipulations of the signals fed to the primary five surround channels.
There are also cinema systems in which the additional two channels of 7.1 or 7.2 are used as screen speakers, much as the Stereophonic Sound for Cinerama and (70 mm) Todd/AO were developed in the 50's. Here the new channels are added to the front, at the screen's mid-left ("left-center") and mid-right ("right-center"), forming: L-LC-C-RC-R. In these cases the surround info is generally the same LS/RS pair as in 5.1 Surround (or Todd/AO's plain mono surround), reproduced over side and/or rear "house" speakers. There have also been films made with a Dolby-matrix encoded Center-rear channel. That's just a quasi-channel derived using what we're calling the LS/RS stereo pair, and represents a pretty modest overall addition, IF you've already gotten the rest of the channels optimized.
When I was working on the six-channel sound mix for my score to Disney's TRON, I had to cheat a little, and used the system as you see it below while sitting back further than usual to check balances. That allowed the five main Klipsch speakers to monitor all five screen channels, while several other rented speakers served as rear surround channel monitors. Later I added the four small Pinnacles for that less-critical task. You can do the same thing if you encounter a need to mix to five screen channels by moving the side speakers inwards towards the front, or by relocating your seating position backwards a few feet to check balances. BTW -- it sounds wonderful even even if you don't move back, a really stunning WIDE sound! That will collapse to screen width in a theater, of course...
In a good theater you can expect many speakers to be used for the surrounds, distributed about the auditorium's side and rear walls, even (bad idea) the ceiling! Dolby recommends many speakers to create an even "omniphonic" distribution of surround information, most helpful when there's only a single channel, as the LCRS of the Dolby Stereo matrix makes available. With the stereo surrounds of our latest discrete digital audio you won't need so much non-directional diffusion. But two additional screen speakers can be marvelously effective. If done properly with a really BIG screen, L-LC-C-RC-R provides precise images from the screen, more subtlety of position, and is less affected by where you sit. Given that screens have shrunk continuously since the mid-60's (multiscreen multiplex mania) the distinctions are probably lost. Mixing all dialog and most screen effects in mono to the center has done even greater disservice to film stereophony, IMHO.


all 7.2
Speaker Locations, All 7.2 Channels

Above you'll see the full 7.2 channels, in an imaginary overhead view (that "burnt orange thing" in the center is my actual studio chair), of an idealized studio similar to the one shown in the photos above. Gradually we're going to work backwards, going downwards in complexity and number of channels, until we reach classic quadraphonic sound (and a couple of amusing variations), and the best way to configure THAT 30 year old system. There's really nothing new in the idea of creating music albums and film soundtracks on multichannels, certainly not since Disney's 1940 breakthrough animated feature, Fantasia. This film pioneered the idea of surround sound ("Fantasound," no less) and stereophony with a six channel auditorium presentation using four optical tracks (three of audio, the fourth was for front/rear steering). Credit William Garity for most of the engineering, the same excellent engineer who helped design their legendary multiplane animation camera. Our tools have become a lot more sophisticated and easier to use since then. Audio quality is remarkably better as well, nearing the theoretical maximums for human hearing and physical acoustics. It's how we'll use them that will determine their success in the marketplace, or not, like the quad boom and bust of the early 70's. It's up to us.
This is the place to mention, for those interested, what speakers are being shown above. I became very attached to the Klipschorns when I was in college. My music professor, Ron Nelson, had a pair of them, with a central non-corner version in the middle, a common method of using Paul W. Klipsch's horn-type speakers. If you put one in each corner in many cases they would be more than 90 degrees apart. The derived center speaker helped to fill this gap somewhat. Anyway, Rachel Elkind and I tried two horn versions in the brownstone studio when we first move into there. Unfortunately, with the shape of the room the corner placements were really impractical. That would have positioned them either behind us, or very far away in front. The dealer suggested we try the newer "Cornwall" type (yas -- that Klipsch model name means you can "use them in a corner or along a wall" -- no comment!). We made careful comparisons for several weeks.
What we learned is that if you made up a 3 dB loss for the Cornwalls, and then did not exceed their already hefty maximum excursion, the sound was nearly exactly alike between the corner horns and these compromise versions. We were going to have four channels, so loudness was no worry at all, not with such high-efficiency speakers that 2 watts would fill a room! Anyway, I got attached, as I said, to these venerable designs, and aside from a few upgrades we made later, have used them ever since, a reliable yardstick I can trust for all my work.
The lowest octave, though, was always a bit weak with the Cornwalls. That was the only other tradeoff. For years I tried small equalizers in the monitor loops to "correct" for this. But at the time Jim Jensen at Sterling Sound did his usual fine job cutting my Beauty in the Beast LP masters I found a much better answer: "Say, what kind of low end speakers are those, Jim?" Velodyne Subwoofers? -- Yowsah! These are active feedback, servo-corrected speakers. I could spend a whole page singing their praises. Simply the only game in town, far as I'm concerned. The servo feedback corrects any and all errors. If only this trick worked above a certain frequency (around 300 Hz), all speakers could be near-perfect. Alas, it doesn't, as the piston-like action of the deep bass motion gives way to more complex vibrational modes, and no one feedback spot can correct for the whole cone. Oh, well, where it does work, why not go for it?
You'll read below that I had to decide between one 15" Subwoof, or two 12" units. This was settled by trying out both carefully with a lot of my own program material. Then the store allowed me to try it here, and there was no argument. The servo made both sizes very very similar in sound. The large size was slightly louder. But two 12's were ever better, and there was actually some directionality gained. So you'll see above and just below the setup the way I have it, with SWL and SWR located midway between the LS-LF, and RF-RS pairs. Works great!

This is also a good time to apologize if I've overlooked someone's favorite surround sound configuration or idea in this very incomplete essay. Every opinion herein rests on several reasonable experiments and follow-ups carried out over a lifetime. That certainly in no way implies any pose of "infallibility." But at least what errors or missing concepts will be found here ought be in the "second and third orders of subtlety." And I encourage each of you to try things out, discover like I've discovered, what actually works for ear, and what is only visual chauvinism at work again in audio -- where it sure doesn't belong. That's why I now have to turn a skeptical eye on many of the sillier ideas being hyped as "fact." Factoids" is more like it, or Urban Legends (question: are there any suburban legends? How about rural?). For myself, I'll stick with what's presented here, at least until something better comes along, the old scientific method: zeroing in very slowly on  what's   very   probably      true...
(Top of the Page)
The Main 5.2 Channels
ideal 5.2
Ideal Surround Speaker Placement -- 5.2 Channels

Fine, let's for the time being forget about those extra two channels. Here's the above view of our idealized monitoring system. The main speakers, LS, LF, RF, and RS, are equidistant from the listener and positioned at 60 degree separations. LF and RF are bisected by C, which can be a slightly smaller speaker from the same family of speakers, since the bass frequencies are often routed to the bigger speakers. But that point becomes moot when you have subwoofers. In this case I made a trade off for two smaller subwoofers instead of one larger one. With careful A/B comparisons I learned that the bass was nearly the same when the two smaller units were working together as a team as with the single larger unit. But there was, contrary to what I had read, a small amount of additional directionality present with the two subwoofs compared to one. Yes, on steady tones and those with slower attacks you heard little difference. But on transient waves, hard attacks, dynamically changing signals, you began to perceive a small amount of stereo effect with the two, SWL and SWR, as shown above. I went with that arrangement, you may prefer the other choice, while the cost is similar.


mod front
Modified Front Speaker Placement -- 5.2 Channels

I've seen setups more like the one above. What's different from the view just above is that the LF and RF speakers have been rotated not to be so toe-in as before, and the center speaker has been brought slightly closer in, more as many three channel monitors are located in mixing theaters and even small home theaters. It's not a big change, and is one we'll pick up again below. If the listening room is not as deep as it is wide, these mild repositionings will be appreciated. The sound will not be greatly affected at all, unless you can compare the two setups one immediately after the other. Then you may hear a slight reduction of the in between imaging. But it won't be any worse than when you listen to two-channel stereo from slightly off the exact center spot. It's not going to destroy the surround sound field, but I bring it up as it has become somewhat common.


Symmetrical Surround Plan -- 5.2 Channels

On the other hand, there is also good reason for making the opposite modification of the front channels, like the symmetric plan above. The 180 degree surround arc of sound has been nearly divided into four equal angles, five discrete channels of sound, plus stereo subwoofers. My personal experience suggests that instead of going with the mathematically exact division, yielding all angles of 45 degrees, this version is slightly better perceptually, with 40 and 50 degree angle pairs. It's probably splitting hairs, but try both and see if you don't agree. We have a slightly more acute perception of angular displacement of sound positions when both ears are nearly balanced, facing a central sound source in front. (It tends to follow a cosine curve function in front of us, with a maximum acuity at zero degrees straight ahead, falling off towards the sides. Behind us our external ears reduce the absolute value of this function by about 50% or more.)
The above plan positions the LF and RF channels somewhat closer together, nearer to C, favoring that most sensitive area. This setup obviously requires a good, active center channel. Here I've shown the same smaller C speaker as before. The subwoofers take care of all the bass frequencies you could stand, so that's not much of a compromise. Notice that for monitoring just four channels of "quadraphonic" material, the missing C channel would leave an impossible "hole in the middle" between LF and RF, if the above configuration were chosen (40 + 40 = 80 degrees apart -- yikes!). If you have to check on a lot of 4 channel material you'd be better off with the first or second layout above. But for 5.2 channels of music, this one's unbeatable -- have yourself a ball!

(Top of the Page)
Digression I -- Classic Blunders to Avoid
worst 4
The Worst Quadraphonic Setup -- 4 Channels

Sometimes the eye can fool the ear into thinking things are fine and jolly, when they ain't. We can all count the four corners of a typical room (har-dee-har, my studio is semi- trapezoidal, and has SIX corners!) or studio. When the first quadraphonic sound was being introduced in the early 70's guess where they put the speakers (you've had enough hints)? Yup, just like the image above, one for one. It also seemed like a nice, democratically evenhanded approach, we have 360 degrees to split up, let's see now, 4 goes into 360... And we get this "classic" setup in name only. It's a complete blunder of the job, about as bad an arrangement for surround sound with four channels as one could devise. If there is but one lesson to be learned via this introduction, it's the graphic one visualized above.
Take a look with your foolish eyes once again. 90 degrees is a pretty wide angle to try to fill with two speakers. Even from the equidistant "sweet spot" as the seat above is located, you will find images tend to vanish when they are midway between the speakers. You have another classic going on here, stereophonically speaking, a "hole in the middle." Add the two other channels and what you get is FOUR holes in the middle. You end up with sound that can only be precisely heard from but four spots. Everywhere else is an omniphonic spread of hard-to-point-to vagueness. It gets worse. Try listening to a normal stereo system (about 45 to 60 degree speaker separation) with your back to the speakers. Hmm... the stereo sort of collapses inwards, doesn't it? I'm not trying to lay any dogma on you. These are simple matters to try out with your own ears, as is all the stuff on this page. We all were surprised to learn how things are not so obvious as we first think they'll be.
And it gets worse again. When you face forward, you can hear any speaker located in front of you, and follow it as it moves to the exact side, either side, whereupon it will sound like it's moving back in towards the middle again, but without the same precision when the speaker moves behind your head. Again it works on both sides the same way. All stereo relies on the fact that our ears will hear "ghosted" virtual images of sounds located between any no too widely separated loudspeakers, if the distances, phases, and sound levels are correctly adjusted. But aside from some very clever tricks heard from exacting positions and setups, you normally won't hear sounds come from outside of a pair of speakers.
The result is that you can image sounds rather well in central locations, with speakers moved to each side a bit, but those speakers set the maximum width you'll be able to reproduce well. Think about those two speakers behind you in the view above. Their sounds are towards the center, just like the front pair. So there's nothing that sounds like it's coming from the sides. The only way to fill in the side "hole" is by locating a speaker there, one on each side works splendidly. After you have normal stereo there is no better place to locate the next two channels than exactly to either side of you. That also works when you add a 5th channel, as the latest surround sound systems do. Like this:


worst 5
The Worst Surround Setup -- 5 Channels

This view is of the worst possible use of five channels. Now one of the black holes in the middle is filled in, leaving just three of them. The sounds up front are fine, wide and very decently positioned. There are no sound to the sides of those speakers, though. Everything comes mainly from within this right angle of two 45 degree sectors. What about the rear channels? Well, they will be heard, of course, but the stereo will be poor compared with that in front. Not only is there no central rear speaker, but the back positions are, like before when you tried this yourself, not definitely locatable. Any poor stereo effect is narrowed when it's completely behind us. Those two channels are being wasted, just as they were with most quad sound in the 70's. Little wonder an honest public might be less than impressed, when confronted with the truth of their own two ears.
The first four channel setup I had, when my studio was in the brownstone, as shown in many phonos on our website, was exactly as the first of these two views shows you. That was folly on my part, because I should have known better, having made many four channel recordings years before with that custom Viking tape deck. I tried placing microphones in that same shape, then the speakers when I played the tapes back. I tried all of them way up in front in various shapes. I tried a "diamond", with one channel up in front, one directly in back, and one on each side. That was much better, but the holes in the middle were irritating, and I was never sure if a certain sound was exactly in front of me, or exactly behind me. Once more I beg you to try this all out for yourself. You can certainly record two channels at a time, and see what happens when the two speakers are center front and center back, then again one on each side, and so forth for each possible pairing. Play the recording in the dark or with your eyes closed. Invite friends and other sophisticated audio buffs to listen with you and compare notes.
Again, you don't have to take my word on this issue. David Greissinger, the brilliant head designer for Lexicon for more than 25 years wrote several scientifically researched papers for the AES (Audio Engineering Society), the AAS (American Acoustical Society), and others in the 80's and more currently. He stumbled upon the very same discoveries which Rachel and I had back in the early 70's (check out our new bibliography at the end of these pages). Our lesson was painfully learned and was independently reproducible, to boot. We had to rehire the same strong electrician / handyman to return and relocate the rear two speakers, positioning them up at the sides (a compromise, speakers that high up can leak over the head slightly to the opposite ear), as you can see in the wider photos of the brownstone studio. The mistake was too painful to live with, and we had to admit it and go with what our ears were telling us. When I moved into here I didn't make the same mistake. Did it even better, as I have a lot wider space. You can see how the channels are located way above. In an arc, 180 degrees wide, like those old Cinerama Screens. Then you more or less split the angle into three parts, so the channels are located at roughly 60 degree intervals. Or use the more symmetrical arrangement above. What's that old line?: "Try it, you'll like it!"

(Top of the Page)
Digression II -- Listening Test to Try yourself
Standard Stereo, both speakers in front of you

There's something most valuable I learned during a valiant failure to become a Physicist. Well, more than one, like keeping a wary, skeptical eye out for deception, or the even more common, self deception. But a lesson that holds in any field at all is a willingness to be proven wrong. You are much more likely to discover crumbs of truth if you don't prejudge what you expect to find too closely, relying instead on reality-checks and experimental tests. Unlike a few sites I've seen that preach to the bleachers, I want you to check out what I'm trying to describe here, not merely take my word on it. Beware the newest Great Prophets who claim possession of "the one true path." All of these ideas here contain a margin for error, a tolerance, and have been verified experimentally, not idle philosophy. You can alter things to a degree away from what's here, before things will weaken or fall apart. And you may discover even more refined ways to handle each situation.
A very modest test is shown in this digression. You ought be able to try it without any special equipment or setup, at home or in the studio. One of the key reasons that many of the original suggestions about Quadraphonic Sound in the '70's failed to live up to their hype could have easily been found by a curious person who was unwilling to go along with the party line. Consider the four speakers, one per corner, concept given in the previous digression. How does sound from the front two channels get perceived, and is this much different from the rear two channels of "obvious quad"? Try one pairing at a time. Pick a few good CD's that exhibit excellent sound, separation, and imaging/ambience. First sit as shown just above, the usual way, centered in the "sweet spot". Okay, note what you hear, essentially all the sound in front. Now swing your chair around, so you're facing away from the speakers, like this:


Face the other way -- both speakers behind you
(well, the chair is rotated around)

This view is a pretty accurate metaphor for what you'll hear when you rotate your chair around by 180 degrees. All the sound now is located behind you. Keep the same music playing as above, listen, then switch the way you face back and forth several times to compare the differences you hear. The speakers won't really edge closer together when you face away, nor ought the directional information become oddly blurred, but that's certainly the way it sounds! I was rather shocked by this test when someone suggested it to me. We had experienced the same problems with the crummy initial layout we'd made in the brownstone studio, and knew something fundamental was going on. But this elegant A/B comparison is such a simple way to demonstrate the principle. The way our ears are constructed we "funnel-in" sounds easily from in front and sides with our built-in "ear trumpets." Whatever comes from behind is masked by those same bio-trumpets, robbing crucial mid and high frequencies especially, the stuff of directionality.
Ever watch a cat rotate its outer ears while listening intently? Theirs are even larger proportionally than ours, and the horn effect must be highly noticeable. They also have better muscle control over them than we do, so they can redirect the aiming points to a large extent. It can't be done simultaneously, but watch them listen to a repeated, continuing sound, and how quickly they are able to zero in on the exact direction. They can adjust to, and adapt better than us in front/read comparisons, so would undoubtedly come up with a significantly different plan for Feline Surround Sound...
But we're interested in an optimum plan or two for Human Surround Sound. Since the back of our heads is not nearly as sensitive to sound directionality and nuance (not to mention a poorer frequency response, and unfortunate interference as sounds move away from the rear of one ear towards the rear of the other), we ought not "waste" too much effort trying to obtain what we can't: a uniform sound field. That's where so many surround concepts fall down, assuming we humans can hear in 360 degrees and follow it all accurately.

© Copyright 2001 Wendy Carlos -- All Rights Reserved.

LeafGo on to Pt. II
(Top of the Page)

Back to the Wendy Carlos Home Page

Wendy Carlos, SurroundSound1
© 2001-2008 Serendip LLC. No images, text, graphics or design
may be reproduced without permission. All Rights Reserved.