MCS Review On-Line Reprints

The Fosgate Research Tate II SQ Decoder And The Minim AD10 Ambisonic Decoder

by William Sommerwerck

Reprinted from MCS Review, Vol.5, No.4, Spring 1984.

This is a review of two distinctly different surround-sound decoders.  Those who delight in looking for or encouraging dissention should note that it is not a comparative review.  Such a review would be not only meaningless, but even impossible.  UHJ and SQ are so different that the decoders for one are not able to handle recordings made for the other.  Nor does a listener have the luxury of choosing his favorite encoding for a given recording.  Decoders can he judged only by playing recordings specifically made for their systems.

Fosgate Research Tate II Surround Stereo System

The review of the Fosgate Tate II came about because Jim Fosgate was tired of my constant bad-mouthing of SQ in these pages.  He felt he had improved SQ decoding to the point where my objections would be answered.  I explained to him that my objections were to the nature of the encoding, not to the quality of the decoder.  I felt that no amount of engineering could overcome the basic misunderstanding of the purpose of surround sound which all matrix and so-called "discrete" systens (including QS and CD-4 discs, and quad reel and Q8 tapes) embody.

Still, Jim had put a great deal of effort into the design.  The Audionics Space & Image Composer (the first Tate system SQ decoder, introduced in 1979) was an outstanding product, but perhaps there was room for improvement.  I was curious to see what another few years' development had done for the Tate system.  The product deserved a fair and careful review, so I had Jim send one.

There is no point in going into a long description of the unit's appearance or features, as these are detailed in the "Record Shoppe" advertisement.  However, the ad neglects to mention that there are two sockets for the remote control, one each on the front and back panels.  Inserting the remote control plug automatically switches the unit to remote operation, a real convenience.  (Trying to use two remotes at the same time, by the way, is not only illogical, but will damage the decoder.)

Despite the (apparent) plethora of controls, setup is a simple three-step process.  First, the input level is set so the -10dB LED flashes occasionally, but the 0dB LED flashes only briefly (if at all) on the loudest passages.  This assures that the softest passages are strong enough to activate the logic circuitry, while the loudest don't overdrive the decoder.

Second, the gain controls on your amplifier(s) are set to that turning the decoder's output "full up" will produce the loudest sound you ever want.  (This is the manual's recommendation.  It will usually result in low volume control settings on your amp, combined with high output from the decoder.  Some poorly designed amps might be over-driven, with resulting distortion, under these conditions.)  There are four trim pots on the front panel so that a critical final balance may be made without having to touch your amplifier.

Third, a mono recording (or broadcast) is played, and the input balance control adjusted until the Max LED is at its brightest.  That's it.

Well, almost.  You might remember that the Audionics S&IC had an "axial tilt" circuit to cancel out crosstalk caused by incorrect cartridge azimuth.  (Here, azimuth refers to whether the cartridge body is perpendicular to the record, when viewed head-on.)  Cartridge azimuth has a pronounced effect on separation.  Unequal left/right separation in the pickup will degrade the performance of the separation enhancement circuits, perhaps even introducing audible side-effects.

Audionics's circuit is patented and would add to the decoder's cost.  Adjusting azimuth by hand infringes no patents and costs nothing.  One simply twists the headshell or shims the cartridge, whichever is easier.  Full directions for doing this are given in the Tate II's instruction manual.

Now that it's set up, how does the Tate II sound?  Terrific.  It's audibly superior to the S&IC in exactly those areas one would expect it to be.  Channel separation is superior and there is a noticeable reduction of side-effects.  For example, it was not uncommon with the S&IC for center-front signals to spill over to the rear.  This occurred on their initial transients, before the logic could react and cancel them out.  It also occurred with Lf and Rf sounds, although to a lesser extent.  On the Tate II, this is reduced to the point where, for all intents and purposes, it is non-existent.  Separation is subjectively perfect.  The Tate II seems to show somewhat better reproduction of ambience than the S⁣ the enhanced separation is probably the reason.

The improved separation apparently results from much faster logic control.  The Tate II seems to respond at least three times as fast as the S&IC.  A side benefit of the greater speed occurs when you sit "sideways".  Under such conditions (which are admittedly unnatural and unfair to any two-channel surround system), one could hear the S&IC "shifting", as front and rear sounds varied in dominance.  This also occurs with the Tate II, but the action is so much more rapid that the effect is less disturbing.  On complex material, it sounds mostly as if the separation has been slightly degraded, with little or no pumping or breathing effects.  On simple material [such as the solo passages on the Columbia Masterworks SQ version of "Carmina Burana" (MQ-33172)], one hears no side-effects at all; the soloists and their ambience seem simultaneously present, with full separation.

But the acid test is Columbia Masterworks Q2S-32923, the SQ version of Leonard Bernstein's "Candide".  Band 2, side 1 ("Life is Happiness, Indeed"), combines the two situations which most cruelly expose any deficiencies in an SQ decoder.  The band starts with Voltaire at center-front and nothing on the other channels to mask front/rear crosstalk.  The band ends with Candide, Cunegonde, Maximillian and Paquette singing simultaneously, one to each speaker.  Any serious loss of separation will cause the directionality to dissolve into a blur.

The Tate II passes this grueling test with flying colors.  There is essentially no leakage of center-front to the rear, and, when all four performers sing, only a slight loss of separation.  One has no difficulty in picking out where a particular voice is coming from.

This rapid logic action does carry some liabilities.  On certain material, mostly massed strings and choruses, there can be a "stuttering" or "gurgling" effect.  (This can occur with SQ material, but seems to be more common when enhancing stereo material.)  Moving the logic control switch to Alternate always removed the "stutter", without seeming to introduce other side-effects.

The Tate II's principal selling point (at least for those without SQ record collections) is its ability to decode Dolby MP (Motion Picture) matrix-encoded recordings.  Many motion pictures (and the videotapes and discs produced therefrom) have the surround channel encoded with this system.  I don't yet have a videodisc player, but I do have one record encoded in Dynaquad, which is enough like Dolby MP that the Tate II should decode it properly.  It did.

The decoder also has a "Surround" setting, which is intended to take any regular stereo recording and wrap it in a 270-degree arc around the listener.  (Anti-phase signals will be reproduced from center-back.)  Unfortunately, there is no way to vary the effect.  (The S&IC has such a control.)  I found the directionality, particularly of sounds which show up at the sides, to be rather vague.  Without any way of tailoring the enhancement to a particular recording, the results were (at least for this listener) more irritating than enjoyable.  But those are my reactions, and one does not buy a surround decoder merely to enhance stereo recordings.

What about using the regular SQ position for enhancement?  Since SQ encoding for Lf and Rf is, in effect, the same as regular stereo, you would think an SQ decoder would have no effect on stereo recordings.  (Delayed and phase-shifted components would, of course, show up in the rear channels, depending on the amount of delay or shift.)  But the Tate II, like the SQ position on the S&IC, insists on wrapping the sound around the listener on many recordings.  This even occurs with "single-point" recordings, where there are no inter-channel phase shifts or time delays that could be "interpreted" by the decoder as surround effects.  (Oddly, SQ decoding did not enhance Enoch Light's classic "Persuasive Percussion", while Surround decoding produced effects that suggested a true quad recording!)

Fortunately, the remote control allows selection among the SQ, Surround and Cinema modes, so the listener can take his pick.  There is no "correct" enhancement; one simply chooses whatever "sounds best".  I found, though, that I often preferred Cinema to Surround (the latter sounding a bit phasey).  Don't hesitate to experiment with each new record.  Always adjust front/back balance, since the optimum balance is likely to differ for each decoding.

The Tate II is very well made and finished.  (I do not agree with Julian Hirsch about the quality of its components.  American-made audio equipment generally is of very high quality, and the Tate II is no exception.  Hirsch's comments are those of someone accustomed to reviewing cheap Japanese receivers.)  As with many signal-processing devices, there is no power switch.  The unit produces a nasty crackling pop when turned on.  Either leave it on all the time, or be sure your amplifier's volume control is down when the Tate II is turned on.  Powering it from a switched outlet on your amplifier may not be sufficient, as some amps come on almost instantly.)  On the other hand, the remote control may be plugged or unplugged without any untoward effects.  Note, however, that plugging and unplugging the remote will immediately change the system volume, so be sure you've first lowered the volume settings on both the remote and on the Tate II's front panel.

Those considering the purchase of a Tate II might have noticed that Fosgate has occasionally improved the decoder, with current versions performing somewhat better than earlier units.  Jim Fosgate tells me that the latest models (those made since about October or November of 1983) represent his final thoughts on the subject.  Now, if some problems shows up or if he happens to stumble on some way to improve the Tate II, changes will be made.  However, Jim has gone on to other projects and is no longer consciously looking for improvements.  One may thus purchase the Fosgate decoder with the comfortable feeling that there probably won't be an improved version six months from now.

Minim AD10 Ambisonic Decoder

The Minim AD10 is the most advanced (or should I say elaborate) decoder for two-channel UHJ and three-channel B-format Ambisonic recordings.  It appears to be the best-performing Ambisonic decoder yet produced, and it includes several features which were not available on previous decoders.

The AD10 is quite diminutive, partly because the power supply transformer is external, and partly because UHJ does not need logic circuits for superior performance.  The color is sort of a "champagne tan", and the styling is simple and unobtrusive.  Disassembly reveals a high-quality British product, which means that it is close to the best American quality.  (I am told that my unit is actually a production prototype, and that current units are even more carefully assembled and finished.)

A row of pushbuttons controls power and mode.  The AD10 can be switched to Hard Bypass (a relay shunts the input directly to the output), Setup (a mono signal is applied to all speakers for balancing), Stereo Enhance (the degree of which is controlled by a front-panel knob), UHJ (2-channel decoding) and B-format (3-channel decoding).  The relay bypasses the decoder when it is de-energized.  Turning off the AD10 thus produces almost instantaneous disconnection, so it may be switched on or off, even with the rest of the system on and the volume up.

Two additional controls, Position and Focus, have not been seen on previous Ambisonic decoders.  The Position control allows one to change his relative position within the hall.  In its "front" position, ambience diminishes and one moves "closer" to the orchestra.  In its "back" setting, the opposite occurs.  There is a central position that presents the recording "as is".  This variation is achieved by adjusting the level of the X (front/back) signal which goes to the speaker-feed matrix.  Note that there is no change in the level of the signals which actually feed the speakers.  Instead, just the component which controls front/back directionality is altered.  Listening verified that not only was the desired effect achieved, but that the coherency of the soundfield was retained.

The Focus control is more difficult to describe.  There was no instruction manual with my decoder, but Minim's literature states that "Focus allows the listener to make alterations in the soundfield as heard in the listening room, which enables him to focus more easily on details in the front sound stage."  I think I know what that means, but I'm not sure that's what I heard.  As Focus was advanced clock-wise, frontal sounds seemed to become more vivid and three-dimensional, but also somewhat phasey.  Since I'm not sure exactly what Focus does, or how it does it, I'll reserve judgment on this feature.

Like all UHJ decoders, the AD10 includes facilities for three-channel B-format decoding, even though no commercial recordings or broadcasts are yet available.  The "front end" of a UHJ decoder produces three outputs (W', X' and Y') which are slightly degraded versions of the original W, X and Y B-format signals.  Since these feed a conventional B-format decoder, the designer need only add another switch position and three jacks to get B-format decoding.  It adds next to nothing to the decoder's cost.

The AD10 also includes another "free" facility that will be of interest to those with mini-speakers.  If you've followed our articles on Ambisonics, you might remember that the W signal is the "mono" or omnidirectional component.  This signal can be used to feed a subwoofer, since the lowest frequencies are not directional (that is, the ear has trouble locating the source).  The AD10 has a separate W output for just this purpose.

There is, alas, no remote control.  However, the unit is small enough that one might be able to place it at the listening area.  This, however, would require six high-quality, low-capacitance cables trailing on the floor.  Wives have an irritating way of disapproving of such audiophile necessities.

The most valuable feature of the AD10 is its Stereo Enhance control.  Prior Ambisonic decoders (as well as almost every matrix decoder) have had some form of enhancement, but none, in this writer's opinion, have been anywhere nearly as useful or effective.  The AD10, by a wide margin, does the best job of extracting ambience of any non-delay device I have heard.  And although the full-surround effects are still not what I would consider ideal, they are somewhat less foggy-sounding than those from SQ or QS decoders.  (It is fascinating, on some records, to hear sounds coming from the sides, which one never hears from matrix decoder enhancement.)

With the enhance control turned all the way down, one gets normal stereo.  As the control is advanced, the soundfield broadens, until at some point the ambience "pops loose" and fills the room.  (Further advancing the control increases the ambience and will cause the sounds to surround the listener.)  What distinguishes the AD10's enhancement is the total naturalness of the effect.  The ambience does not sound tacked-on; it and the direct sound are a coherent whole.  There is no sense of "empty sides", which one often hears from conventional quad recordings; the ambience is seamless and complete.

Of course, the quality of the enhancement will depend on the quality of the recording.  But the results with the best audiophile discs (such as those from Reference Recordings or Sheffield) are superb.  The sound becomes not only more realistic, but less mechanical and artificial-sounding (even though we have passed the signal through additional electronics).

After all this, the AD10's performance as an Ambisonic decoder is a bit anti-climatic.  Generally speaking, the AD10 does noticeably better what earlier decoders (the Integrex and IMF Electronics) did well.  The most obvious differences are that the sound is a bit more transparent and less fuzzy, and that imaging is a bit sharper and more stable.  Minim agrees (why not!) and says the improvement is due to superior-quality components and a beefier power supply.

For those who have never heard UHJ records, it is probably fair to say that no commercial recordings have ever captured such a natural sense of ambience as these.  (One possible exception might be the quad tapes made by Sonar some years back.)  My description of the way the decoder extracted ambience could just as well apply to the performance of the AD10 decoding UHJ discs.  The direct and ambient sounds, as they do in real life, form a coherent whole.  The ambience never sounds separate or tacked-on, a common failing of matrix (and even discrete) recordings.

I also tried playing a Dynaquad record through the AD10.  Although it did not produce exactly the same directional effects as the Tate II (which isn't surprising, since UHJ and Dynaquad do not have the same encoding locus), the results were exceptional.  Sounds that were supposed to come from the rear, did, and there was absolutely no ambiguity about it.  (Note, again, that this level of performance is obtained without logic circuits.)  Although I still prefer the Tate II for Dolby MP recordings, the Minim decoders (particularly the cheaper AD-7, discussed in a minute) make an interesting alternative.

There are very few UHJ records with sounds placed behind the listener.  But in reviewing these, I noticed something which points up one of the fundamental differences between Ambisonics and matrix systems.  If a sound appears to be coming from behind you, and you twist your head in its apparent direction, the location hardly shifts at all.  But with SQ recordings, the position shifts quite a bit.  Think about it...

As we go to press, Larry is still deciding whether to carry Minim decoders in the "Record Shoppe".  (He no doubt will have something to say about it elsewhere in this issue.)  Minim makes a cheaper decoder, the AD-7, which is missing only the front/back Position and Focus controls; Stereo Enhance is retained.  I haven't tested this unit, but I have no reason to believe it will not provide essentially the same level of performance as the AD10.

Interestingly, the AD10 will sell for about $600, while the AD-7 will cost only $200!  Why the big difference in price, when there is so little difference in the units' facilities?  For one, the AD10 uses closer-tolerance components, which are then hand-selected!  Minim wants the UHJ decode circuits to be as accurate as possible, and they feel that individual component matching is the only way to wring out the last drop of performance.

Second, the AD10 is sold for a "fair" price; that is, the price fully reflects the cost of parts and labor needed to produce it.  But the AD-7 is being sold at a very low price (not a loss, but one which would not be considered a fair return on investment) to entice people to buy one.  Minim feels strongly enough about Ambisonics that they would like every serious listener to be able to afford one.  This, in turn, keeps interests in Ambisonics and the sales of Ambisonic recordings going.  Minim expects to make their profit when an AD-7 owner trades up to a fancier decoder a few years from now.

Conclusions

Quite frankly, this article is a bald plug for two decoders that I happened to like and that Larry sells.  (The idea for the article was mine, though, not his.)  Nonetheless, the opinions expressed are as objective and fair as I can make them.  The strengths and weaknesses of both units are as described, and I think most listeners will agree with my assessments.

Both the Fosgate Tate II and the Minim AD10 are billed as "state of the art" for their respective systems.  Since there isn't much competition, they could have earned that position simply by default!  However, they live up to the claims made for them and represent a significant advance over previous products.  And at a lower price.  The Audionics Space & Image Composer, plus remote control, along with the IMF Electronics Ambisonic decoder would have set you back about $2000.  Yet a combination of the Tate II with remote control and the Minim AD-7 costs less than one-third as much, yet provides a higher level of performance.

Getting the Most Out of Your Audionics or Fosgate Decoder

This advice applies to any surround decoder, including UHJ.  But it is most apt for SQ decoders, since problems with dirty records are most likely to show up with SQ recordings.  Of course, the treatments suggested will improve the sound of almost any recording, quad or otherwise.

If you've owned a Tate system decoder for a while, you've probably come across records where a center-front performer or instrument produces distortion or spitting effects in the rear channels.  I was told this was due to mastering defects, and, since it only showed up in a few records, I assumed it was true.

But when I bought a record cleaning machine last year, I learned otherwise.  Thoroughly-cleaned records "spit" a lot less.  Intrigued, I treated them with LAST (Liquid Archival Sound Treatment) and applied StyLast (a liquid daubed on the stylus and cantilever before the record is played).  Lo and behold, the nasties almost completely disappeared.

The moral?  You can't be too clean.  A record-cleaning machine is an excellent investment in good sound and increased listening pleasure.  LAST and StyLast aren't cheap, but they further enhance sound quality and decoding accuracy.  Records that have been thoroughly cleaned also seem to develop less of a static charge and accumulate dust and lint more slowly, but I don't know why.

Reprinted from MCS Review, Vol.5, No.4, Spring 1984.
Copyright © 1984, 2000 by Laurence A. Clifton

Last updated: April 9, 2006

MCS Review On-Line Reprints