Audionics Space & Image Composer / Tate Audio

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Steve Kennedy

Member
Joined
Jan 14, 2004
Messages
29
Location
Portland, OR
My original post of this message was in another area of this forum so I thought I would post it here as well since it belonged here in the first place!

Hi all! I am a new lurker to these forums and I thought I would introduce myself.

I worked for Audionics of Oregon from 1975-1983 and was the lead Engineering Development Technician on the Space & Image Composer project. This marvelous unit consumed 3 years of my life and I loved every minute of it!

Since I had day-to-day involvement with most of the technical (and some of the political and historical) facets of this project since its inception, I have a lot of information locked away in my brain. If you have any questions regarding the history, design, manufacture or operation of the Space & Image Composer I would be happy to reply.

Along with the Composer stuff, there is a lot of related information about Tate Audio, the Tate & Tate II chipsets, Audionics Shadow-Vector (and other SQ efforts) and Jim Fosgate's earliest Tate-based products in there as well. I haven't used much of it in the last 25 years but much of it is as fresh as yesterday to me!

I am happy to see that a Multi-Channel community that stills dabbles in "the old ways" is still around! SQ with directional enhancement from vinyl discs is still my favorite way to listen to music (through my Space & Image Composer of course, one of the original 10 pre-production units) although I don't get to do it as often as I used to).

I noticed my old friend Bob Popham has been mentioned here as well. I haven't talked to him in years but I did run into his website recently. Bob has had the unique perspective of not only being involved with the later (post Audionics) Fosgate Tate and multi-channel development efforts but also being there for the Audionics Composer production and the transition period in between!

Keep the faith!
 
Hey Steve!!

Welcome and great to see you here!(y) I'm sure we all have a ton of questions for you..... I've got your initials inside my own S&IC from when you did the upgrade to it!

First off.......can you tell us how many S&IC's were actually produced and sold?

And, any historical information that you can provide on the time between the production of the S&IC and transition to the Tetrasound/TATE II 101A.

Thanks in advance...........and hope we see you here often!:D
 
Steve,

WELCOME!

I would be interested in what the differences are electronically between the Tetrasound 101 and the Tate units.

Thanks

:-jon
 
WELL WELL WELL! Steve Kennedy, as I live and breath! You built me what I called the SK 101 when I was in college in 1980 when you were with Audionics. I still am using the SK 101 (although the BT 2 preamp is getting a bit long in the tooth) and the recently rehabbed by Quad Bob S & IC. Its great to see that you have resurfaced! You'll be a great asset to the list! And thanks for your years of multi-channel support!
Marc Stenchever
 
QuadBob said:
Hey Steve!!

Welcome and great to see you here!(y) I'm sure we all have a ton of questions for you..... I've got your initials inside my own S&IC from when you did the upgrade to it!

First off.......can you tell us how many S&IC's were actually produced and sold?

And, any historical information that you can provide on the time between the production of the S&IC and transition to the Tetrasound/TATE II 101A.

Thanks in advance...........and hope we see you here often!:D


It's nice to see that analog SQ surround isn't forgotten! The fact that 25 years on there are still happy owner's of Space & Image Composers makes all the hard work and hassles expended in its design and production worthwhile!

While I can't be 100% sure (accurate production records weren't really kept as far as I know), the numbers that I recall are:

10 Pre-production prototypes were hand-built (identifiable mainly by PC board color... light green standard material instead of the Blue FR14 material of the production units). These were followed quickly (and without any major changes) by 179 regular production units.

I would estimate that approximately 50 of these 179 were built in the factory as Tate II (new "fixed" Exar chipset) units and probably a similar number were returned to the factory and I retrofitted them with the new chips. The remaining units will still have the original National Semiconductor chips.

If you look inside it is quite easy to distinguish version I from version II chipsets. The newer Exar chips were built in a 14 or 16-pin DIP package rather than the 18-pin DIP package of the Nationals, so they had to be mounted "hi-rise" style on an elevated adapter board that rose above all the surrounding components. The models with National chips had the detector chip mounted on the PC board itself (between two rows of large green poly capacitors) .

This ought to give you some basic information on the Space & Image Composer and some identification tips. I'll add more information regarding the development timeline, chipset differences, Ray Dolby and the arrival of Fosgate on the scene in future installments.

"No, not that Left Front, your other Left Front!"
- Charles Wood, Audionics of Oregon 1977

Steve
 
Marcsten said:
WELL WELL WELL! Steve Kennedy, as I live and breath! You built me what I called the SK 101 when I was in college in 1980 when you were with Audionics. I still am using the SK 101 (although the BT 2 preamp is getting a bit long in the tooth) and the recently rehabbed by Quad Bob S & IC. Its great to see that you have resurfaced! You'll be a great asset to the list! And thanks for your years of multi-channel support!
Marc Stenchever


Wow, another blast from the past. I certainly do remember you, Marc! You were the only student I can recall who was an owner of a Space & Image Composer at the time. You are right, the BT-2 is even older than the Composer!

I also still have a BT-2 and its replacement, the RS-1 (which I still use). I also still have my Composer (of course... one of the 10 pre-production units) and 3 Audionics CC-3 amps. The equipment has been very reliable and still works to this day. I haven't used the BT-2 in years but it also holds a lot of fond memories for me.

I am glad to see that a few of the original purchasers are still owners to this day! If you own a Space & Image Composer you belong to a very small club and Original owners are in a very elite group!

Steve
 
Hey Steve:
I noticed on the sleeve of many of my SQ records are pictures of SQ decoder equipped quad units. One of them is an Audionics decoder, and it isn't a Composer. How many of those were built and how good are they?

The Quadfather
 
Wow, I haven't thought of those in a long time! These date back to the early to mid-70s. I think I still have one of those in a box somewhere! This decoder didn't have a name or model number to my recollection... it was just the Audionics SQ decoder.

The circuit board was imported from the U.K. I seem to recall, as Audionics was also importing Sinclair and Radford devices from England at the time. It was a discrete 17-transistor 6-pole SQ Decoder (no enhancement, just "raw") and it sounded pretty good (no I.C.s) even though it had the same limited separation as all the other simple SQ Decoders. It would have made a credible "front end" to a stand-alone Tate Separation Enhancement System (like the original hand-built discrete Tate unit Martin Wilcox made and showed around 1974).

This unit even had a discrete 4-channel input (like the S&IC) that was identified in the manual as "for use with 4-track tape and CD-4 decoders".

If you sat in "the sweet spot" and had your system tweaked to within an inch of its' life, a wonderfully spacious sound field was created (just don't move your head more than a few inches!).

I can't imagine that we sold more than 100 of those units, probably less than that. We were only bringing them in a dozen at a time!

Steve
 
QuadBob said:
Hey Steve!!

Welcome and great to see you here!(y) I'm sure we all have a ton of questions for you..... I've got your initials inside my own S&IC from when you did the upgrade to it!

First off.......can you tell us how many S&IC's were actually produced and sold?

And, any historical information that you can provide on the time between the production of the S&IC and transition to the Tetrasound/TATE II 101A.

Thanks in advance...........and hope we see you here often!:D


One thing I could comment on would be the relative historical timeline of events surrounding the Tate development effort, the Audionics Space & Image Composer and the Fosgate 101 units. To the best of my recollection (don't hold me to EVERY detail, it's been 20+ years!):

Martin Wilcox (the engineer) and Wesley Ruggles Jr. were the developers of the Tate Separation Enhancement System in the early '70s. By 1974, they were showing a hand-built one-off prototype of Martin's design and it was getting rave reviews (I have a photocopy of a 1974 column in Stereo Review magazine that raves about the unit and mentions that a standard SQ decoder was used as the "front-end" to the Tate system).

Sometime in mid-70's, Tate Audio approached National Semiconductor about making the Tate chip set and Martin Wilcox worked on refining the design for chip manufacturing. Unbeknownst to anyone at the time, National screwed up one of the chips and it didn't work correctly. They didn't take responsibility for the error and wouldn't fix the chip, but they did make the chips with some key interface points coming out to pins on the chip. This would allow the potential for external circuits to be designed that might help restore operation to the chips.

About this time (about 1976 or so), Martin Wilcox and Wes Ruggles had a falling out and I believe Tate was having financial difficulties brought about by the National chipset not working properly. Tate was on the hook to pay for the chip development but couldn't sell any chipsets because they didn't work. Tate Audio sued National Semiconductor but that just tied things up in court. This is where Audionics comes into the picture.

Audionics had enlisted Lynn Olson to develop what became the "Shadow Vector" SQ Decoder & enhancement system and it really did work. However, Audionics didn't have the cash to fund an I.C. development of its own and the Shadow Vector system (like the Tate System) was the size of a small suitcase if built with discrete components! A decision was made to jump on board Tate's bandwagon and help finish the development of a "band-aid" interface for the chipset to overcome its' problems.

The development partnership went on for about 2 years or so, on and off. We were getting credible performance but it wasn't yet good enough for Wes Ruggles, Jr. to approve release of the system (he was a notorious perfectionist). I spent many hours with Wes doing listening tests and "tweaking" prototypes. We got to a point where we (Audionics) were approved to show a prototype system at the Winter Consumer Electronics Show in Las Vegas in 1979.

Fosgate was a hot name in automobile audio systems at the time and had a big presence at the show. Someone from Fosgate's crew heard a Space & Image Composer prototype demo (before it even had that name, the S&IC name was created on the way back home from the show while driving in the middle of the Mojave desert, but that's another story!). Our demo literally blew him away, and he ran back to the Fosgate booth and dragged Jim Fosgate back to the Audionics booth to hear it for himself.

Jim Fosgate was so enamoured of the concept he almost exploded with enthusiasm and good ideas right off the bat and signed up to be the second Tate developer on the spot. In a true spirit of cooperation, Audionics shared the interface work it had completed and special parts it had made so far and Jim Fosgate shared some of his ideas and concepts on how to improve system performance. Jim first designed a car-version of the Tate system... I believe he called it the Tetrasound (originally he wasn't going to compete with Audionics in the home market) and then ultimately switched gears to develop his own home decoder... the original model 101.

After a short period with both companies contributing to the effort, we had an interface design that Wes Ruggles had to approve of and production commenced.

About this time, Ray Dolby had purchased a production Audionics Space & Image Composer and installed it in his home in what could be considered to be one of the first home media/theater rooms. He was so impressed with the performance that he contacted Tate Audio (through Audionics) and became a Tate licensee. He immediately directed his engineers to see what they could do to develop a replacement for the QS-based surround system they had been using for the Dolby Surround sound systems used in movie theaters (the kind in use when Star Wars was released in 1977).

Now, with Dolby, Audionics and Fosgate all using the Tate chipsets, there was enough "weight" behind the project to get the Exar company to take a look at the National Tate chips (since National wouldn't or couldn't due to the Tate lawsuit) and try to determine what could be done to fix the design. From a detailed chip inspection under a high-powered microscopes and comparisons to the original schematics, they determined that National had not only messed up one section of circuitry, they had also left off a large chunk of circuitry from the schematic! No wonder it didn't work as-is!

Because of all the effort we expended developing an external interface that worked, it was decided that Exar would simply laser-"prune" away the bad circuitry and bring fresh new interface points out of the chip that would continue to use our external interface circuit. This became the Tate II chipset, and worked much smoother because the defective circuitry wasn't impeding our ability to make the chips work properly. The cancer was trimmed away and the chips were much happier!

The new Exar chipset was substituted for the National chipset in production immediately after we optimized the interface circuit for these chips. Fosgate and Dolby did the same... Fosgate's updated unit was the 101A model. Audionics developed a retrofit kit and quite a few of the original Space & Image Composers were upgraded to II status.

On an S&IC, you can tell the difference (inside) by locating the twin rows of Green film caps behind the balance controls near the center of the circuit board. If there is a chip mounted directly on the board in the middle of these caps, it will have a National logo and the LM1852N part number. If instead you find a "daughter-card" mounted on the long leads of a wire-wrap socket mounted on the card (above the caps altogether), then the chip on that card will have an Exar logo and the XR402 part number.

Sometime around late 1981 or early 1982, there was some problem between Exar and Tate and the supply of Tate chips dried up and disappeared. The licensees could only buy the chips from Tate Audio, and Tate Audio was the only entity who could buy the chipsets directly from Exar (all according to contracts, of course). Audionics never built another surround system and Fosgate went on to manufacture several generations of brilliantly-designed high-end surround decoders that used his own separation enhancement system that did not rely on the Tate Audio technology or chips.

So, there in a nutshell is a basic Tate System history inside of a decade. The dissappearing supply of Tate Chips at the time explains why there are so few available now as spare parts!

Steve
 
Thanks Steve for the great posts. It is a joy to learn about the inside stories, even if they are not rosy.

Do you think it is easy or even feasible to develop software that would take either the two-channel SQ stream or the four-ch stream from a basic SQ decoder and do the separation enhancing in a computer.

This would be a not-in-real-time approach, perhaps allowing for tweaking of the parameters for particular recordings, cuts or segments.

Thanks in advance.
 
That is a very interesting story about the defective National chips. I have one of those units, and y'all did a hell of a job making them work right. I have a couple of questions. How good was Audionic's Shadow Vector system, and would the Tate chips have been that much better had they been made right in the first place? How was the lawsuit settled, what was the result? And those questions about doing Tate in software are good ones too. (problems are easier to fix in software) What if someone developed a decoder with an analog and digital input and had a processing engine that is just a clean slate. You would load software into it to make it decode SQ, QS, CS, Ambisonics, DTS, Dolby (all types), Meridian, and whatever else came down the pike. Wouldn't that be a great idea? Oh well, in my dreams.

The Quadfather
 
Quadfather Quote: "What if someone developed a decoder with an analog and digital input and had a processing engine that is just a clean slate. You would load software into it to make it decode SQ, QS, CS, Ambisonics, DTS, Dolby (all types), Meridian, and whatever else came down the pike. Wouldn't that be a great idea? Oh well, in my dreams."

That is a great idea and a great dream. I do not have the Tate decoder and most likely never will but, I do have a few SQ records.

I suppose that the cost to write the software for a "dead" format could never be recovered. It would have to be a labor of love by someone who knew the inner workings of the Tate system and also had the ability to write the software.

tcdriver
 
tcdriver said:
Quadfather Quote: "What if someone developed a decoder with an analog and digital input and had a processing engine that is just a clean slate. You would load software into it to make it decode SQ, QS, CS, Ambisonics, DTS, Dolby (all types), Meridian, and whatever else came down the pike. Wouldn't that be a great idea? Oh well, in my dreams."

That is a great idea and a great dream. I do not have the Tate decoder and most likely never will but, I do have a few SQ records.

I suppose that the cost to write the software for a "dead" format could never be recovered. It would have to be a labor of love by someone who knew the inner workings of the Tate system and also had the ability to write the software.

tcdriver

Strangely enough, I heard of a tantalizing rumor of just such a thing. When I was working for the Sony DVD authoring center a year or two back, I was talking to the head of our audio department and was introduced to a visiting engineering VIP from DTS (who's name I unfortunately cannot remember). I gushed about his company's products and let it slip that a was a major quadhead. I mentioned my newly-acquired Tate to him and he said he remembered the unit, and had also programmed a piece of software for kicks that decoded SQ, perhaps even better than my unit was able to. My jaw dropped at the thought, but I was unfortunately unable to inquire any further due to time constraints on both our parts. I'm afraid I'm a little shaky on the technical side when comes to how this would be accomplished, but I believe he mentioned breaking the audio into frequency ranges and decoding those seperately...anyway, it's fun to think that the answer might be out there. Sorry I don't have more details...
 
The matrix decoding is not too difficult, it is basically phase/level shifting, and VERY accurate broadband 90' phase shifters can be made in software. The trick is to enhance separation a la Tate and Sansui, and that is not so easy. It is possible, after all DPL II has some "logic" and most of the decoding/enhancement is done with DSP software, but it would take a very dedicated quad enthusiast to do it, not to mention the ability to code DSP software.
If I remember correctly, a British guy did write some SQ decoding software a couple of years back, I contacted him, he did not seem interested in others testing, I think Tab tried it out though. As far as I recollect it was a basic software decoder, stereo wave file in, 4 waves out.

The last commercial SQ decoder that I remember was made by Cantares, but I do not know if it had any separation enhancements or was just a basic decoder. I think it was mainly an Ambisonic decoder, based on the old Integrex design with SQ/QS etc thrown in.

We can but dream ...

Malcolm
 
proufo said:
Thanks Steve for the great posts. It is a joy to learn about the inside stories, even if they are not rosy.

Do you think it is easy or even feasible to develop software that would take either the two-channel SQ stream or the four-ch stream from a basic SQ decoder and do the separation enhancing in a computer.

This would be a not-in-real-time approach, perhaps allowing for tweaking of the parameters for particular recordings, cuts or segments.

Thanks in advance.


It is certainly possible with the DSP processing power that's around today. On a PC it probably would not be a "real-time" application, but simply a stereo capture (A-to-D) with the DSP processing to extract discrete channels that would need to be locked-together via time code or something. Perhaps an output to a multi-channel recording format.

I have used Cool Edit Pro (now known as Adobe Audition since Syntrillium was gobbled up) for years in fulfilling my needs for recording and editing on my computer and I am blown away by what is possible with the DSP-based software effects. I have no doubt that someone clever in software and processing could come up with an SQ decoder with separation enhancement!

Steve
 
The Quadfather said:
That is a very interesting story about the defective National chips. I have one of those units, and y'all did a hell of a job making them work right. I have a couple of questions. How good was Audionic's Shadow Vector system, and would the Tate chips have been that much better had they been made right in the first place? How was the lawsuit settled, what was the result? And those questions about doing Tate in software are good ones too. (problems are easier to fix in software) What if someone developed a decoder with an analog and digital input and had a processing engine that is just a clean slate. You would load software into it to make it decode SQ, QS, CS, Ambisonics, DTS, Dolby (all types), Meridian, and whatever else came down the pike. Wouldn't that be a great idea? Oh well, in my dreams.

The Quadfather


Thanks, we worked hard for over 2 years to get through all the hoops that finally resulted in a workable, listenable and APPROVED system. All of this was done with few chip schematics and no idea what was wrong inside the chip! Talk about a Snipe hunt! Most of the performance monitoring was done by ear, then circuitry was designed to address audible anomalies! Educated trial & error at its finest! I certainly got paid to listen to a LOT of surround sound in the late '70s!

Cliff Moulton (Audionics' Chief Engineer) was really brilliant in that regard, taking someone else's description of an anomaly and then visualizing a circuit to compensate for the mechanism he believed responsible for that anomaly. Cliff also designed most of the electronic circuitry for Shadow Vector decoder. It is my firm belief that the Tate System may have never seen the light of day had it not been for Cliff's "training" (learning to think in "Quad" & "separation enhancement") on the design for the Shadow Vector. Cliff was almost solely responsible for the design of the external "band-aid" circuits that made the flawed Tate chips initially work.

The Audionics Shadow Vector System (invented & designed by Lynn Olson) was developed along different lines than the Tate System. (My comments here are my own observations and based on my limited exposure to Shadow Vector, as I was not a part of the development effort on that design.)

The Tate System seemed to be targeted at maximum separation initially, then it was backed off to regain lost spaciousness. Hyper-separation "pulls" the spatial fabric of the music into the four corners of the room and does not sound natural. Shadow Vector wasn't designed to get maximum separation, it was focused on extracting accurate & correct spatial information from the encoded mix and only applying enough separation to maintain the spatial image across a wider listening area with maximum fidelity.

Shadow Vector (as such) was never intended to be a consumer-level device like the Tate System (initially), it was supposed to be a high-end mastering decoder to start with (Like the original Schieber unit), with a plan to "trickle-down" the technology to "prosumer" level via integration. It was stunning when it was tweaked right (especially on live SQ symphonic recordings with natural ambience), but because the prototype was hand built from discrete components (like the Tate prototype), it was difficult to keep it adjusted as the massive amount of circuitry naturally drifts with time and temperature. Talk about trying to hit a moving target!

Taking it to I.C. level was necessary for stability as well as ultimate production cost but we never got that far because Tate was thought to be farther along, thus a quicker time to market (it was thought!).

I personally do not know what became of the lawsuit. I'll have to ask about that one and see if anyone knows.

The software decoder could work, but probably not in real-time. An analog decoder like the Tate system is difficult to adequately model in digital in such a way that would lend itself to "real-time" processing. With computers getting quicker all the time, near "real-time" could be achievable in our lifetime under the right set of circumstances (i.e. dedicated silicon).

Analog systems in general (such as our organic electro-chemical brains) are difficult to model in a digital environment with extreme fidelity.

Steve
 
The Quadfather said:
<Cut and paraprased> How good would the Tate chips have been that much better had they been made right in the first place?
The Quadfather


A very good question! It is quite possible that if National Semiconductor had fabricated the original chips correctly, they might have packaged the LM1492N detector chip in a package with a smaller pin count (not bringing out the appropriate internal interface points that allowed us to "adjust" the internal performance). This would have meant that the chipset would have "sound like it sounded" with no possibility for optimization.

I'm not saying it wouldn't have sounded good, but speaking from experience in how many ways this system can be adjusted (or maldajusted), there is a much higher likelyhood that the chipset as originally designed would not have approached the level of performance we were collectively able to squeeze out of the Tate System utilizing these external adjustments. In some ways (and with the benefit of hindsight), National Semiconductor did us all a favor!

In my opinion, the Tate System gained a performance and consistency edge over that which might have been expected from the internal circuits alone!

Steve
 
Steve Kennedy said:
........ Ray Dolby had purchased a production Audionics Space & Image Composer and installed it in his home ..... became a Tate licensee..... directed his engineers to .... develop a replacement for the QS-based surround system they had been using for the Dolby Surround sound systems used in movie theaters (the kind in use when Star Wars was released in 1977)......

Steve
Hi Steve,

Fascinating article so far, great to see all this in depth detailed info straight from the source.

I have always understood that Dolby Labs™ originally used QS as the basis for Dolby Stereo™, (which no doubt accounts for the film soundtrack of Tommy sounding like the tape machine was out of phase when listened to un-decoded in stereo).

Therefore if Dolby™ became a Licensee and swapped over to using Tate Chips in their Professional Cinema Products (remember little or no domestic Dolby™ Product, apart from Dolby 'B'™ NR of course, was around in those days.)

Does this mean the Dolby Surround™ encoding was changed to SQ? (which I think I remember reading something about on the Dolby™ Website) or are the Tate Chips capable of decodeing other matrix's as well as SQ?

And thereafter as DPLII™ was (if I am correct say) developed by Jim Fosgate for Dolby™ Labs to Decode Dolby Surround™ Encoded material, does it not follow that when reproducing 'Matrix Quad' Material with DPLII then it would be SQ NOT QS that is correctly decoded as which seems to be the common opinion on this board.

Short of asking Ray Dolby(™?) himself, you sound like the guy to sort out this confusion once & for all. OR is my 'Logic' just a load of 'Reverse Polish'? :confused:

Also you mentioned that Tate hand built Directional Decoders from discreet components. Where any cct's ever published of this or similar discreet enhanced decoding schemes? Would be great to build one as a bolt-on to a Sansui QRX, all we need is the Schematics!

John - Liverpool UK
[email protected]
 
Steve Kennedy said:
I'm not saying it wouldn't have sounded good, but speaking from experience in how many ways this system can be adjusted (or maldajusted), there is a much higher likelyhood that the chipset as originally designed would not have approached the level of performance we were collectively able to squeeze out of the Tate System utilizing these external adjustments. In some ways (and with the benefit of hindsight), National Semiconductor did us all a favor!

Not to feed anyone's paranoia here but is it likely that the existing Tate units have as of now drifted out of adjustment?

Is there a source for the alignment procedure?

Also, I read somewhere that there are line-level taps in the Tates. Is it true?
 
Jim Fosgate himself said that DPL-II will not properly decode SQ, it sums the back channels.

Steve,

do you have an opinion on which was better, the Tate 101a or the S&IC???
 
Back
Top