Reprinted from MCS Review, Vol.4, No.4, Spring 1983.
The Tate Directional Enhancement System is based on a very complex design concept so complex that many have said "It couldn't be done." More than ten years' research and work were necessary before the fully-optimized version of the Tate System, the Tate II, finally reached the marketplace. This is the first article in which information on the final five years of research has been included.
The Tate System was conceived by Wes Ruggles and Martin Willcocks. Work on the project began in Europe in the early '70s and was later moved to Los Angeles, California, for completion. In order to design any product, one must know what level of performance he is trying to achieve. Let's go back to the beginning of the Tate project and learn what Ruggles and Willcocks were trying to accomplish.
The goal was to produce a decoder at a reasonable cost capable of decoding a matrix-encoded two-cbannel source into a soundfield that would sound as good as a four-channel master tape and with no audible side-effects. Believe me, this was to be no easy task. Wes Ruggles believed that a well-recorded four-channel master tape was unmatched in its ability to recreate virtually any musical performance. If you think about it, almost anything can be done in that format. Assuming the speakers are positioned properly, sounds can image anywhere in a 360° circle. Sounds can also image from the top of the room. Recording engineers have almost unlimited flexibility. This is not to say that all four-channel tapes produced in the '70s were good. The recording industry still had a lot to learn.
The CBS SQ system was chosen over other matrix systems for its monophonic and stereo playback compatibilities.&bnsp; Martin Willcocks designed the original circuitry with the help of a computer to optimize circuit values. Breadboards were assembled to test the design. Extensive listening tests were conducted over a long period of time using practically all available SQ and stereo source material. Four-channel sub-master tapes were SQ-encoded down to two channels and played through the decoder. A switching system was utilized so the output of the decoder could be compared directly to the four-channel tapes. Remote control of the switching system via a long cord enabled the listener to observe differences from any place in the listening room. Work continued until the decoder sounded as good as the four-channel tapes.
The final circuit breadboard used thousands of parts and was so large that it took up an entire table top; it measured a full 3 by 4 feet. The fact that it worked reliably enough to be a useful design tool is a tribute to the craftsmanship of Ruggles and Willcocks. However, if the decoder were produced in this discrete form it would probably cost about $5,000 and reliability would also be a problem. Therefore, the next step in bringing the Tate System to the consumer at a reasonable price was to convert the bulk of the circuitry to integraLed circuit form.
A large IC company of high reputation was contracted for the job. As anyone who has produced an IC will tell you, "it is a time consum-ing process and there can be many pitfalls". It took well over a year to produce the first sample ICs, because they were very complicated, and unfortunately there were problems. The major difficulty was in the section we now call the "Directional Control Interface Circuit". This section was located physically in the same package as the detector IC.
To make matters worse, this was the section of the circuit that required so much subjective listening and the most time to perfect. It also required many components. The IC manufacturer was not willing to correct the problem, so it was necessary to tool up the ICs by a second company. The Directional Control Interface Circuitry was omitted from the IC for fear it might not work properly, which would have delayed the project again. At a later date, it was learned that the IC manufacturer had left out some critical stages of the circuit; we still don't know why. (It was certainly not the fault of the Tate people.)
I became involved with the Tate System about this time and began design work for the Directional Control Interface Circuit (DCIC). For a time Audionics and I were both working on different versions of the DCIC. We exchanged technology in order to speed up the project. My goal was to equal or surpass the performance of the original bread-boards. The circuit would have to be simple enough to be produced in a discrete form without a custom IC. Eventually a custom IC will be produced for the DCIC to reduce manufacturing costs, but it will take time.
The Tate System went back to the drawing board and extensive listening tests began again. Several breadboards were built to test various DCICs. When I started working on the decoder I expected to find one breakthrough that would make it work perfectly. Several breakthroughs were found, but the final result was accomplished only by paying strict attention to the subtleties of the design with very precise adjustment of the circuit. The DCIC provides functions that add to the capabilities of the detector in order to expand overall system performance. Patents are pending and I am not at liberty to disclose the theory of operation in any great detail or show the circuit.
Briefly, this is how the DCIC works. The detector output is filtered and fed into the DCIC. The signal is phase inverted, amplified and smoothed to match the input requirements of the matrix multiplier. Wave shaping, contant power and logic circuits are incorporated in the design to provide other functions. Eight factory preset controls are provided to adjust control voltage balance, gain, floor and bias levels to exact values. Provision is also included for linearity adjustments. A bandpass filter was found to be necessary in front of the detector to eliminate IM distortion from occurring at low frequencies. The filter also filters out FM pilot signals and TV horizontal line frequencies that could adversely affect circuit operation.
The inverse of the "Fletcher-Munson Curve" was selected for the response curve. (The Fletcher-Munson Curve follows the sensitivity of the human ear throughout the audio spectrum.) The filter allows the decoder to maximize separation as judged by the human ear. The block diagram in Figure 1 shows all the circuit sections of the Model lOlA Fosgate Research, Inc., Tate II decoder. You will notice the stereo surround circuit just after the matrix. The circuit rearranges the stereo signal so the decoder decodes stereo as if it were SQ. Many good stereo selections are equally as good as the best SQ material.
Figure 1. Block diagram of the Tate Directional Enhancement System, incorporating the Directional Control Interface Circuit.
Now for the big question: "How does the decoder sound? Did we accomplish our goal?" We have conducted tests with the four-channel switch mentioned earlier and compared the decoder output directly to the four-channel master tapes being fed into an encoder. Tapes were, of course, properly phased for SQ encoding. There were only slight differences between the two outputs, but, in fact, you would probably pick the decoder as sounding best. If you find that hard to believe, so did we! It seems as though we have created more than a decoder. For some reason the decoder sounds more open and spacious than the discrete reference tapes. Several different listeners all drew the same conclusion when listening to our reference system.
I attended a videodisc recording session in California. The recording was done in discrete four-channel and monitored in a medium-sized listening room with the conventional four-channel speaker placement. We encoded the four channels to two channels to feed the decoder. We had the capability of comparing the discrete four-channel feed directly to the output of the decoder. Levels were carefully matched with level controls on all channels. Meters on all four amplifier inputs were used to adjust levels. We asked each recording engineer at the session to listen to both signals; and even though they did not know which they were listening to at any moment, they picked the decoder unanimously as sounding better than the discrete. They heard the same differences the other listeners had heard on our reference system.
Exactly what kind of psychoacoustic magic we are experiencing is unclear. I think it is the result of the hyper-cancellation that is occurring in the matrix multiplier which produces the incredible separation. The world of psychoacoustics is not fully understood, and is certainly not an exact science. Good stereo material will usually decode as well as SQ material. It's surprising to play a stereo recording that was made 20 years ago and hear the performers in front and the audience around and behind you. Playing videodiscs through the decoder is a real awakening. The picture almost seems to be three-dimensional. Dolby-encoded video material may he played in the Cinema position when it is desired to position sounds as they were encoded. Many listeners feel that much of the Dolby material is more enjoyable in the surround mode. This directs more information to the back channels. The mode switch is located on the remote for easy comparison of the two positions.
You may wonder how well the decoder will perform when playing the new generation of digital discs. The phase and frequency response of the two stereo channels is far more accurate with digital than with conventional records; therefore, decoder performance is even better. The final version of the Tate System does not have any of the flaws in performance as did its predecessors. There is no decoder action ... JUST BEAUTIFUL MUSIC!!! Unless, of course, you are playing a hopelessly poor recording with phase abnormalities.
There has been a longstanding feud in the industry concerning the performance of SQ and how it compares to other systems. Until now, there was no state-of-the-art decoder available at a reasonable price to judge it by. An encoder capable of matching the performance of the decoder is also needed, and will be forthcoming. With the performance of the fully optimized Tate II ... I rest my case, you be the judge. As one audiophile said after listening a few minutes ... "You know this is almost paradise!"
Jim Fosgate is President of Fosgate Research, Inc., the manufacturer of the Tate II Surround Stereo System.
Reprinted from MCS Review, Vol.4, No.4, Spring 1983.
Copyright © 1983, 2000 by Laurence A. Clifton
Last updated: July 23, 2000
MCS Review On-Line Reprints