Seven Loudspeaker Systems (Issue No. 14 Summer through Winter 1989-90)

Seven Loudspeaker Systems (If They Are All So Accurate, How Come They All Sound Different?): Audio Concepts "Sapphire" with "Saturn"; Carver "Amazing Loudspeaker" (Platinum II Edition); JBL L40t3 ; Precise "Monitor 10" ; Quad ESL-63 USA Monitor ; Snell Type C/II ; Waveform [TAC 14]

We examine a varied assortment of up-to-date speaker designs, none of them less than good, none of them perfect, and try to sort out the advantages, disadvantages and trade-offs of each design approach.

The great paradox of loudspeaker design, as we have stated before, is that designers keep trying to come up with the perfect loudspeaker when there is really no accepted theoretical model of perfection. How can you achieve the best possible result without a clear definition of what the ideal result is? Amplifier designers have their straight-wire with-gain ideal; tuner designers aim to make their units per form as if hard-wired to the broadcasting studio; but do speaker designers have anything nearly as definable and verifiable to aim at? Not to our knowledge.

Yes, we have been known to mutter on occasion that the output of an acceptable loudspeaker must resemble its input, but that is a somewhat flippant truism of limited usefulness, far short of a mathematical model. An amplifier channel has one input and one output, so you can easily ascertain the resemblance between the two. A loudspeaker has one input but n outputs, making the I/O comparison a bit more involved, alas. For example, the usual axial frequency-response test, although as good a starting point as any, is simplistic and insufficient; two different speakers with highly similar axial responses will in all probability sound totally different. On the other hand, no speaker with a ragged axial response can possibly sound first-rate, so the test is valid in a limited way and needs to be performed.

Speaker evaluation is not a discipline for one-track minds.

It is far beyond the scope of this article, indeed of this publication, to bring order to the unsettled arguments for and against the point source, the line source, the figure eight, and various other Platonic forms promoted by their lobbyists as the model for the perfect loudspeaker. We must leave that to mathematicians like Stanley Lipshitz (who got very engrossed in a mathematical discussion of ribbon transducers with Bob Carver during lunch break at our Seminar 1989). Meanwhile we cannot ignore the fact that every speaker reviewed here sounds different from all the others, and not just in bass response and dynamics, which we expect to be dependent on size and usually price.

This Heinz-57-varieties syndrome-always a different flavor even in the frequency and dynamic range common to all good speakers-is one of the great remaining frustrations in audio, and we must emphasize that we are talking about the genuinely good designs, not the failures.

What we have been saying about the convergence of the latest audio components toward total waveform fidelity to the recorded signal, and therefore very similar sound, is true only up to the speaker terminals. Past that point there is still a free-for-all, so much so that all generalizations about it are somewhat precarious, unless we restrict ourselves to narrowly defined categories such as forward-firing 3-way systems. We therefore prefer to discuss the possible reasons for each loudspeaker's sound separately, as related to its specific design.

Audio Concepts "Sapphire" with "Saturn" Audio Concepts, Inc., 901 South 4th Street, La Crosse, WI 54601.

"Sapphire" floor-standing 2-way loudspeaker system, $849.90 the pair as full kit (parts kit plus assembled cabinets). Matching pedestal, $399.00 the pair. "Saturn" subwoofer, $639.90 the pair as full kit (parts kit plus assembled cabinets). Passive crossover, 3199.90 the pair. Tested samples on loan from manufacturer.

These very interesting speaker designs are available only in kit form, but the crossovers and cabinets come fully assembled, so that even the lightweight do-it-yourselfer can progress from shipping cartons to a listenable stereo pair in a matter of hours. We are basing that statement on our judgment rather than our experience, Audio Concepts having sent us finished units for testing-a special accommodation to the Fourth Estate and not without its own pitfalls, as we shall see.

Let us address the Sapphire first as a self-contained 2 way speaker system of almost, but not quite, full range, in the idiom defined by, say, the Spica TC-50 at one end and the overpriced Celestion SL-600 and SL-700 at the other. If that kind of speaker is your idea of happiness, the likelihood is that the Sapphire will make you very happy indeed.

It is a very well-thought-out example of the genre. The two drivers around which it is designed are made in France by Focal, a company in the forefront of diaphragm and voice coil technology. The woofer is a 7" unit with die-cast frame, dual voice coil (remember the Watkins woofer?) and Kevlar sandwich cone-very high-tech (or at least very sexy).

The tweeter has a 3/4" inverted dome, also of Kevlar. The crossover is first-order, with all the advantages (coherence, etc.) and disadvantages (higher distortion, etc.) of that piece-of-cake approach. Air-core L and polypropylene C, of course; Audio Concepts will not offend against the high end code. The enclosure is an upright wedge of 0.45 cu. ft. internal volume, a bit like an elongated TC-50, but there is more to it than that.

What distinguishes the enclosure from conventional designs is that the front and two sides of the wedge-in other words, the entire surface of the enclosure except the back and the bottom it rests on-are covered with 2" thick acoustical foam; only the diaphragms of the drivers are left free, even the mounting plate of the tweeter being covered with a felt ring. The result is the most complete possible suppression of early reflections and edge diffractions-and what a difference that can make! Before we go into details, we must add that the naked enclosure with the exposed foam covering rates rather low on the wife-acceptance scale and that, on the other hand, the handsome Vandersteen esque hood assembly makes the essentially satellite-sized speaker look quite large and also introduces a tiny resonant coloration. We left the hood off in our tests but had the speaker mounted on the optional pedestal, which also adds height and bulk.

As for the performance of the Sapphire, we would call it highly respectable on all counts and just about state of-the-art in imaging. If you are an open-minded imaging freak (or is that an oxymoron?), you owe it to yourself to check out this speaker. We have other priorities; tonality happens to be more important to us than directional and spatial clues, although we love to have it all in one speaker; nevertheless, when the definition of the soundstage is as clear as it is with a pair of Sapphires, we are disarmed and delighted. Obviously, by comparison, conventional boxes radiate a lot of garbage off their fronts and sides almost simultaneously with the diaphragm-launched signals, hence the difference. It should be added that the acoustical foam covering of the Sapphire is deployed in such a way as to leave just enough clearance for the drivers' desired solid angles of radiation, so that the tonal quality of the first arrival is in no way compromised. In fact, the overall tonality of the Sapphire, once the obvious small-speaker limitations are discounted, is quite excellent; in a deliberately cruel side-by-side comparison with the Quad ESL-63, for example, we heard surprisingly small differences in instrumental timbre. The crossover-less Quad sounded more seamless in the low/mid/high transitions and marginally more neutral, the Sapphire erring perhaps on the bright side, not at all unpleasantly-but there was no wipeout! Enough said.

In our laboratory tests, the most outstanding characteristic of the Sapphire was its response to square pulses, which we swept from 1 ms width to about 125 ps. We have never seen more coherent waveshapes and fewer trailing squiggles coming out of any loudspeaker with a crossover.

Only the Quad ESL-63 and the Carver ribbon (crossover less above 100 Hz) are as good or better in this respect. We were impressed. The frequency response of the Sapphire is less impressive; we have seen flatter and smoother curves, although there are no egregious anomalies. The overall trend of the axial response is ever so slightly upward; there is a possibility that the tweeter level is set a smidgen too high. Furthermore, the tweeter by itself (Kevlar, schmevlar) is surprisingly ragged in response, with a pronounced peak at approximately 17 kHz; still, it works within the system, so why argue with success? The woofer is nice and flat; in this box it rolls off at 12 dB per octave below 80 Hz with an apparently perfect Butterworth contour. A little bit of positional juggling with respect to the corner distances (see Bill Rasnake's article in Issue No. 13) could boost that bass response just enough to make the Sapphire a stand-alone system for the bass-shy contingent-you know, those who think that deep, flat bass is somehow not "tight" enough.

And that brings us to the Saturn subwoofer, which is specifically offered by Audio Concepts as the ideal low frequency extension of the Sapphire to build a super loud speaker with down-to-22-Hz response for less than $1800 the pair. The Saturn is designed as a sealed column system of 2.5 cu. ft. internal volume, with two face-to-face mount ed 12" polypropylene-cone drivers firing downward and up ward at floor level in a push-pull-wired configuration. This is a genuinely valid space-saving and distortion-canceling idea, but it appears that we were sent the wrong samples for testing. Ours rolled off at 12 dB per octave below a 3 dB point of 42 or 43 Hz, much too high, and the passive cross over supplied was strangely mismatched to the system, creating a huge upper-bass hole between the subwoofer and the satellite, among other problems. Discussions with Jack Caldwell, designer of the Sapphire and Saturn systems, and Mike Dzurko, president of Audio Concepts, elicited two possible explanations: (1) the ad hoc kit builder provided to us as a favor had overstuffed the subwoofer enclosures with damping material, thereby choking the deep bass, and/or (2) a bad batch of 12" drivers with shifted electromechanical parameters had slipped through. We can believe the kit building error because we caught one in the Sapphire, too:

the wires to the woofer had been reversed in one unit, with disastrous effect on phase integrity and, until the error was original-and it ought to, with the redesigned woofers and 6 dB increase in efficiency. The bottom notes of the Zurich Tonhalle organ on the Dorian recordings of Jean Guillou come through with astonishing power and truly clean delineation, putting many a separate subwoofer to shame. The Waveform speaker system (see below) has comparable, but not necessarily better, bass performance; unfortunately the logistics of testing very large loudspeakers prevented us from making a side-by-side comparison. We have yet to test a third system even remotely in the same class bass wise. The revised ribbon is equally impressive, very similar to an electrostatic in transparency but of course quite different in wave launch, and super smooth. The somewhat warm and mellow balance of the original Amazing has been changed to a slightly brighter, crisper, more neutral characteristic, which we definitely prefer; the two controls in the back can vary that within certain limits. The Q of the woofers can also be varied, simply by substituting resistors for a shorting wire at the input terminals in the back; one can choose a Q of 1.1 (as delivered, with the wire in place) or 0.7 (maximally flat, our preference) or 0.5 (critically damped, for the tight-bass contingent). With the optional electronic control box the wire is left in place, and the Q is then variable from 0.3 to 2, the straight-up position being 1.

Our measurements confirmed our general subjective impressions. The frequency response of the Platinum II is extremely flat and smooth, without any significant peaks or dips attributable to driver resonances or network anomalies.

On the bottom end, flat response extends down to 22 Hz; on top, the roll-off begins at 17 kHz. (These are essentially 0 dB readings, not -3 dB points.) With the upper-midrange control in its neutral position, the "Gundree dip" we took mild exception to in the original version is just about gone; Bob sort of snuck it in so gradually that it is virtually undetectable, unless you know about his perversity on the subject and actually look for it. Well, at least there is no excess energy in that touchy "presence" region. The square-pulse response of the Platinum II is gorgeous, perhaps even better than that of the Quad ESL-63, but of course the ribbon can be expected to behave coherently above the single 100 Hz crossover. We are a little disappointed that the woofer array and the ribbon are still being driven with opposite polarities; however, when we reversed one pair of leads from the network to "correct" that, the wave launch in the crossover region became totally discombobulated, indicating that the physics of the drivers and the filters strongly resisted the

"purer" approach. The roller coaster impedance curve of the Amazing is considerably improved in the Platinum II Edition; now the minimum is 4 ohms (at around 300 Hz) and the maximum 15 ohms (at 2.5 kHz and 7 kHz), with 8 ohms a reasonable average or nominal value. The standard sensitivity measurement being normalized to an assumed 8 ohm load, both sensitivity and efficiency measure approximately the same in this case: 88 dB. We are unaware of any other ribbon loudspeaker system that comes even close to that figure. At the same time, the Platinum II can easily handle the full power of an amplifier like the Carver Silver Seven-t. That kind of efficiency and power handling, combined with the low distortion inherent in the ribbon design, may be conducive to unnecessarily loud playback levels with CD's or DAT's, since there is never any sense of strain or impending breakup. Those who set the volume by cranking the music to the "ouch!" level and then backing off a hair will have to change their technique.

Since the "beta version" of the Platinum II made available to us was already broken in, we have no idea whether or not the 50-hour break-in requirement that we faulted in the original speaker has been reduced; according to Bob it has been, quite a bit. All we can say is that this particular pair of Platinum II's sounded sweet, smooth, open and uncolored from the start. To rate the listening qualities of the speaker against a widely known standard, we decided to stage a mano a mano confrontation with the Quad ESL-63 USA Monitor, which we consider to be the timbral-accuracy champion of the world and very bad news for most of the competition. The 63's can be moved around a bit more easily than the Waveforms, so this was not an unrealistic project. To our surprise, the Platinum II was not "plown away" (to use the favorite cliché of the 'philes), either in the reproduction of timbre or of spatial detail. For example, with the new Dorian sampler (Vol. II), the piano and strings of the Ames Quartet sounded perhaps a shade better on the Quads, but the lovely soprano of Julianne Baird came through even more palpably and ambiently on the Carvers. So let us say that the two speakers are more or less in the same league in terms of accuracy and musicality, but of course there is no contest when it comes to bass and dynamic range-or performance per dollar.

Where does that leave The Amazing Loudspeaker, Platinum II Edition-hey, we said it without wincing-in our hierarchy of speaker designs? Very close to the top, and several steps closer than the original. This particular application of the line source, with these particular refinements, seems to be giving the point source a run for its money (pace Stanley Lipshitz). We shall wait for our promised production samples of the "Plat Two" (Bob's shorthand) before we officially bless it as a reference; at this point we only wish to ask a rhetorical question: What other monolithic loudspeaker system goes down as low, goes up as high, can play as loudly, as cleanly, as transparently, render spatial detail as accurately-and at what price? JBL L40t3 JBL Incorporated, a Harman International Company, 240 Cross ways Park West, Woodbury, Long Island, NY 11797. L40t3 book shelf 2-way loudspeaker system, $798.00 the pair. Tested samples on loan from manufacturer.

First, the good news. The JBL proprietary pure titanium 1" dome tweeter used in this speaker system is absolutely the finest electrodynamic tweeter known to us, regardless of price. That includes anything and everything by Audax, Dynaudio, Focal, Vifa-you name it. On axis, it is flat within approximately +2 dB up to 25 kHz; it is still quite flat 30° off axis (and more!) up 16 kHz; it handles power like a woofer; it sounds clean, clean, clean-and only JBL has it. (If others could have it, nobody would use those dinky little super-tweeters to extend the top end of their domes.) Now for the not-so-good news. The 2-way system JBL designed around this dream tweeter and a fairly decent 8" paper-cone woofer is okay but far from excellent. All by itself, without a reference, the L40t3 sounds like a respect able high-end bookshelf speaker, but when we switched to the Snell Type C/ II in the middle of a Julianne Baird track on Dorian, everybody listening went "ah!"-the midrange became so much more open and uncolored. We tried to nail the reason for that on the lab bench and found that the woofer, although reasonably flat in response, is set 2 to 3 dB higher in level than the tweeter, or so it appeared when we had the B&K microphone close enough to the speaker to avoid serious room effects. That step-like transition in the crossover region could have been what we heard-or possibly just excess midrange energy. The two drivers are in phase, though; square-pulse response is pretty good, indicating some degree of coherence, but not spectacular.

The bass response, considered separately, is again quite good for a smallish 8" system; the vented box is tuned to 36 Hz, and maximum vent output is also in the neighbor hood of that frequency, i.e., not far from classic fourth order Butterworth alignment. (The vent is in the back, so it is difficult to obtain the summed response of the woofer and the vent in the near-field.) Large-signal bass could be a little better; high-level transients tend to create a high-Q response profile as the voice coil starts leaving the gap.

JBL had the means and the opportunity here to come up with a sensational little speaker, but the whole ended up being less than the sum of its parts.

Precise "Monitor 10" Precise Acoustic Laboratories, a division of Onkyo USA Corporation, Suite B, 200 Williams Drive, Ramsey, NJ 07446. "Monitor 10" floor-standing 3-way loudspeaker system, $1599.00 the pair.

Tested samples on loan from manufacturer.

The last time we looked at a loudspeaker system from Onkyo was ten years ago, and we were not impressed. This time they did the right thing. They created a separate loud speaker division under a new name, Precise, and hired an American audio engineer with considerable professional as well audiophile cachet, Keith O. Johnson, to design a whole new line of five speaker models. The message is clearly that they are serious about speakers and serious about the audiophile market. As readers of this journal know, we have a lot of respect for "Professor" Johnson without necessarily agreeing with all of his priorities; in this particular situation we suspect there may have been some pressures to finish five commercially viable designs against a deadline, but even so we find much to like in the top-of-the-line Monitor 10. The others-Monitor 3, Monitor 5, Monitor 7, Monitor 9-we have not seen or heard.

The basic format of the Monitor 10 is that of a for ward-firing 3-way system with 10" woofer, 6'/2" midrange driver and 1" dome tweeter. The woofer is in a vented box of 2'/2 cubic feet internal volume; the other two drivers are in a separate detachable enclosure. The tweeter is offset rearward from the midrange, and the midrange from the woofer, in the soi-disant time-aligned configuration we have always considered to be of questionable value. (Not that we object to it, mind you; it certainly does no harm.) The tweeter level is adjustable by means of a three-position toggle switch. A hood goes over the midrange/tweeter module to create a somewhat KEF-ish appearance.

The performance of the Precise Monitor 10 can be analyzed in terms of a little speaker-the midrange/tweeter module-extended on the bottom by a modest subwoofer.

That perspective is suggested by the low crossover frequency, which looked like 100 Hz to us, although the literature says 200 Hz. The crossover network, incidentally, is complex and appears to use slopes of several different orders; we are reluctant to categorize it on the basis of our routine microphone probing. The tweeter takes over at 4 kHz, at least as we see it; again, the literature says 2.5 kHz. At any rate, the little-speaker part of the system is a good one, very flat between 100 Hz and 10 kHz except for a bit of excess energy at 1 to 1.2 kHz; the soft-dome tweeter has a distinct notch at approximately 10 kHz, above which the response is again quite flat to beyond 20 kHz. Off-axis response is also impressively flat. The midrange and tweeter are in phase with each other but out of phase with the woofer; the latter tracks the polarity of the input, whereas the midrange and tweeter reverse it. Square-pulse reproduction is fairly coherent, nonetheless, since the discontinuity takes place at a sufficiently low frequency. The "naked" midrange driver, when measured directly through its separately available terminals, looks a little rough in response, rougher than it is within the system-a somewhat surprising discovery in view of Keith Johnson's known preoccupation with cone materials, resonances, and stress modes. This polymer laminated paper-cone unit has a 4 dB peak at 1.2 kHz, which constitutes a "gotcha!" when correlated with the aforementioned little problem around that frequency in the overall system response, where it is less obvious. (Yes, we believe we can hear it; we are coming to that.) The woofer-or quasi-subwoofer-under this very respectable little speaker is another matter. We could hardly believe our instruments when we found its overall output level to be set approximately 10 dB below that of the mid range. What on earth is going on here? The vented box is tuned to 22 Hz by means of an incredibly long and convoluted duct; maximum vent output is also in the neighbor hood of that frequency; so this is an attempt to design a genuine 22-Hz box around a 10" woofer enclosed in 2!/ cubic feet of space-come on! How much efficiency, how much output can you get that way? Of course, as we said, Keith Johnson's priorities are not always the same as ours; it is possible that he prefers some output at 22 Hz to a flat, efficient 36-Hz or 38-Hz box which is level-matched to the midrange. Or was it the marketing people? The sound resulting from this mildly eccentric design is surprisingly good, perhaps because that little-speaker core of the system is well-conceived. If we found ourselves in a situation where we urgently needed a high-quality speaker system and the Precise Monitor 10 became quickly available to us, we would be more than pleased. If, on the other hand, we were offered a number of possible options within roughly the same price range, the Monitor 10 would probably not be our first choice. It represents just another flavor in good, up-to-date loudspeakers. (See our introductory comments above.) What is that flavor? Well, a little more emphatic in the upper midrange than the more neutral Snell Type C/ II, for example-most probably as a result of the small peak we measured-but partisans of the speaker would undoubtedly call that a lively or snappy or open quality, and they would not be totally wrong. The deep bass is occasionally evident but rather tame, not at all like that of a flat-down-to-22-Hz system, and that is exactly what the measurements predict. The overall sound is musical and very listenable but not so compelling that you want to run out and buy the speaker. We somehow have the feeling that the performance/price ratio may be more impressive in some of the other Precise Monitor models, where the working parts are of comparable quality but the design perhaps more conservative.

Overriding our minor reservations about the speaker is our genuine admiration for Onkyo as the first major Japanese audio company to abandon the endemic Akihabara glitz-tech approach to loudspeakers and to come out with a high-quality, businesslike, Anglo-American type of product instead. That line of thinking has a future.

Quad ESL-63 USA Monitor Tovil Distributors of America Inc., 14120-K Sullyfield Circle, Chantilly, VA 22021.

Quad Electrostatic Loudspeaker 63, USA Monitor, $3990.00 the pair.

Tested samples on loan from USA distributor.

This is not really a full-fledged review, as there is no need for one. For quite a number of years now, the ESL-63 has been the closest thing to a gold standard in the world of loudspeakers, and the late great Richard C. Heyser's review of the original version in the June 1985 issue of Audio leaves very little room for emendations by lesser lights, at least electro-acoustically. (Subjectively his evaluation was rather strange, but several high-end/underground/dilettante reviewers have meanwhile counteracted that.) Our reason for putting in our two cents worth at this time is that the ESL-63 keeps coming up as a kind of undefined reference in the context of our speaker reviews, and that it has been sitting there against the back wall of our studio to be pulled out into the room whenever we are in doubt about the validity of what we are hearing-all that without ever having appeared in our table of contents, until now. In this issue, for example, the other six speaker reviews might not have had quite the same thrust had it not been for that reassuring presence in the background.

The approximately two-year old USA Monitor is a ruggedized, electrically and physically less fragile version of the original ESL-63, so minimally different in fundamental design and execution that the same old manual comes with it, although the input panel configuration in the back is no longer the same. (No voltage selector, 100 to 120 volts being standard; conventional input terminal posts, but spaced just a hair too close together to accept Pomona double banana plugs. This is the USA, chaps; 0.75" here is 19.05 millimeters, not 18.75, because 1" = 25.4 mm, not 25 on the nose. Or did you go by the Queen's thumb?) We can neither confirm nor deny audiophile claims to the effect that the USA Monitor sounds "better" than its predecessor, not having had them side by side and not being in the habit of making such critical comparisons from memory.

The frequency response of the USA Monitor as we measured it corresponds quite closely to Dick Heyser's published measurements of the earlier Quad ESL-63, even though our methods were not the same. The speaker is very flat indeed, and the small departures from flatness are innocuous; on the other hand, the Carver "Amazing Loud speaker" Platinum II Edition is even a little flatter, so we are no longer as impressed as we used to be. The same statement goes for square-pulse reproduction; nothing we have seen equals the USA Monitor except the 60" ribbon of the Platinum II, which exhibits perhaps even a little less

"grass" on the scope between pulses. Dick Heyser did not use that particular test, but we verified and agree with some of his reservations about the speaker: the horizontal dispersion could be a lot better; high SPL's are out of question;

floor placement is a no-no. (We are still looking for the ideal stand.) We also have some reservations of our own about the bass response, which is neither deep enough nor tight enough. There is a rather Q-ey bump at around 50 Hz and good-bye Charlie below 40 Hz-and quite limited power handling at the lower frequencies.

All that fades into insignificance, however, once you become familiar with the way the Quad reproduces instrumental timbres. A cello sounds so much more like a cello than through any forward-firing electrodynamic speaker at any price-witness the Delos CD of Janos Starker playing David Popper encores. String quartets, quintets, trios, etc., sound self-evidently natural, unstrained, true to life on the Quads where other good speakers merely give you hi-fi. It must be the combination of flat response, coherence, low distortion, virtual point source, figure-eight launch, who knows what else-and all bets are off if the dynamic limits of the system are exceeded. A very special, very frustrating speaker-unbelievably beautiful at times, at other times totally inadequate-but very hard to live without when you know what it can do.

Crossing the Quad ESL-63 over to a subwoofer, may be at 80 Hz or thereabouts, will probably get rid of some of the frustrations, although dynamic headroom limitations at higher frequencies will continue to intrude from time to time. Unfortunately, we have not yet found the suitable subwoofer and crossover for the job; there are some tricky

"hybridization" problems to be solved there with an electro dynamic monopole bass transducer.

Meanwhile the Quad ESL-63 USA Monitor, used full range, remains our reference for music with limited bass and dynamics-which includes a great deal and excludes even more.

Snell Type C/ II Snell Acoustics, Inc., 143 Essex Street, Haverhill, MA 01830.

Type C/ II floor-standing 3-way loudspeaker system, $1990.00 the pair. Tested samples on loan from manufacturer.

This one has been weighing heavily on our con science. We already had our samples of the speaker when Issue No. 12 went to the printer, yet we had no review of it even in No. 13, except for a one-paragraph mention. Now this model is about to be superseded, although its successor will be only minimally different, we are told. Every now and then a worthwhile piece of equipment falls between the cracks of our slowpoke testing/publishing process, and this was one of them. Most regrettable-and not to be repeated.

Luckily, the Snell Type C/ II is so simply and un-problematically good in its own specific category (viz. single box forward-firing passive-crossover speaker systems) that a highly analytical review is hardly needed-just a general thumbs-up and a few specifics. That such is the case must be credited to designer Kevin Voecks, who had the astute ness to enlist the technical resources of Dr. Floyd Toole's famous acoustical facility at Canada's National Research Council in the development of his design, instead of flying by the seat of his pants like so many egomaniacs in the loudspeaker industry. There's math in them there Snells.

This is an approximately 4-foot tall system with a relatively small footprint of about 1 square foot; even so, it is big enough to make one expect better bass out of its 10" woofer and rearward-firing ducted vent than it delivers. The box is tuned to 31 Hz, but the overall bass response begins to decline at 48 Hz, with the -3 dB point at around 45 Hz.

We could be slightly off on that measurement, since the summed response of the woofer and vent proved to be quite difficult to measure in the nearfield, but not by much. By contrast, when your Editor was in the speaker business, his comparably sized 10" system (Fourier 1 or Fourier 1L) pumped out an honest 32 Hz at -3 dB. That is our only serious criticism of the Snell Type C/I, and even that must be qualified by the observation that the Q of the system is quite stable under dynamic conditions-it will not "woof up" on bass transients.

Otherwise the speaker is well-nigh faultless. The woofer, the 5" midrange cone driver, and the 1" soft-dome tweeter are all in phase; the obviously complex crossover (4th-order Linkwitz-Riley, we believe) works beautifully; above 200 Hz the frequency response is uncannily flat, not only on axis but 30° off axis as well-as good as we have ever seen. Of course, with Floyd looking over Kevin's shoulder, he would probably not have dared to leave the premises before everything was perfect. We found it impossible to get a really good square pulse out of the Type C/ II despite all that perfection, but that has been our experience with other good 3-way systems using high-order cross overs. It should also be added that the late Peter Snell's trademark of a tiny supertweeter aimed at the back wall is still with us here; you can switch it off if that extra little ambience it adds to the sound is not to your liking. We had it on most of the time.

The basic listening impression made by the Snell Type C/ II is one of absolute neutrality. There is never too much or too little of anything; the balance is always right, on all types of program material and with all amplifiers that we tried. Of course, the round-and-mellow partisans will call it a little bright, and the sizzle aficionados will call it a little dull-and that will prove our point. In terms of our complaint about too many flavors in loudspeakers, even the good ones, this one comes reasonably close to the flavor of pure water. One does not tire of it, even after hours of listening. The bass is not particularly impressive, as we said, but what there is of it is of high quality. Indeed, the entire speaker is of high quality.

When a classical record producer friend asked us to recommend an accurate monitor speaker that is not too large and heavy to be transported in his car to recording sessions, we suggested that he try the C/ II. He did and he is happy. What more can we say? Waveform Waveform Research Inc., RR. #4, Brighton, Ontario, Canada KOK 1HO. Waveform floor-standing 4-way loudspeaker system, $3995.00 the pair, with dedicated electronic crossover (system purchasable directly from manufacturer). Tested samples on loan from manufacturer.

This review had to undergo a substantial last-minute rewrite because of the announcement of a radical price cut, which completely changes the positioning of the product from the critic's point of view. Let us begin, therefore, with a little bit of history before we address the hardware.

The driving force behind the Waveform loudspeaker is John Otvés, a Canadian master woodworker of Anglo Hungarian extraction. (His surname means goldsmith or silversmith in Hungarian, and it is in some ways prophetic.) John had a dream, which he started to implement four or five years ago: to build the world's finest loudspeaker and sell it to grateful audiophiles via discerning high-end dealers. Yeah, just like that. As a not very far from stereotypical "mad Hungarian"-fiercely idealistic, proud, contemptuous of mediocrity, music loving, impulsive, and impractical- he paid a lot more attention to the product than to the marketplace as he pursued his dream. Obviously, the speaker had to be a masterpiece in wood-he would attend to that-but it also had to be the ultimate in engineering design. Unfortunately, no amount of wishing could produce a totally new and different design concept, but at least the speaker was to incorporate all existing and available knowledge of conventional technology. To that end, he enlisted the help of two outstanding Canadian technologists, Paul Barton and Dr. Claude Fortier, and he made full use of the acoustical laboratories of the National Research Council in Ottawa, under the supervision of Dr. Floyd Toole. It would be hard to find a better team than that.

When the original Waveform loudspeaker made its debut in mid-1986, the cabinet was the audio-furniture equivalent of the Bernini baldachino in St. Peter's in Rome, a dazzling display of woodworking virtuosity. Designed in the form of a truncated octagonal pyramid (which does make some acoustical sense), the solid cherry cabinet had two-inch walls, delicately fluted side panels, Ceylon ebony and 24-karat gold inlays (Gtvosm II), raw-silk grille, and so forth. In a sense the tail was wagging the dog-the cabinet was upstaging the high-tech innards and fancy electronic crossover-and the three-piece set was priced at $17,000, which of course included the discerning high-end dealer's down payment on his Porsche. For those who would not be ashamed to settle for heavy fiberboard construction with a black, very high-gloss, Steinway-like lacquer finish and no flutes or inlays, the price of the otherwise identical system was only $9800. What John did not fully understand-and what we could have told him had we known him at the time-was (1) that discerning high-end dealers would be most unlikely to take on an ultrahigh-end dream speaker that did not have the endorsement of the dream merchants of the high-end/underground journals, (2) that said dream merchants had previous and very strong political commitments to the likes of Infinity IRS and IRS Beta, Wilson WAMM, etc., and (3) that therefore the advent of "the first technically 'correct' forward-firing speaker" (John's claim in a letter to us and at least arguably valid or near-valid) was actually bad news in the corridors of high-end hi-fi power. At first John was quietly ignored in those corridors, then he was politely raped. His speaker was bad-mouthed in both The Absolute Sound and Stereophile. We were not the least bit surprised; in fact, we had predicted it.

At that point--and that was only a few weeks before you read this--John revised his dream. "To hell with the magazines and the dealers," he obviously said to himself, "I am going to sell the Waveform directly to the audiophile who wants this kind of loudspeaker." In most cases that would be a very bad marketing decision, in our opinion, but in this specific situation we feel that John Otvos has nothing to lose and the audiophile has everything to gain. To clear his pipeline of existing inventory and to create a cash flow, he is selling the three-piece system in black lacquer at $3995.00, which is just about his cost. That, dear reader, is an almost indecent bargain, an absolute steal, and a safe one at that, since Otvés Industries, which is John's custom woodworking business, is obviously profitable enough to allow him to subsidize Waveform Research, and he is therefore unlikely to disappear overnight. What guarantees he can provide to back up his direct-to-the-customer sales remains to be seen. The price cut certainly relieves the reviewer of the obligation to rank the Waveform against various world's-best contenders, but let us try to evaluate it without regard for its former or present price.

The foundation of the Waveform's "architecture" is a separately amplified and actively equalized 15" woofer in a vented enclosure. (There are actually two ducted vents, at the 5 and 7 o'clock positions next to the woofer.) We can unhesitatingly state that we have never measured or listened to a better bass system; it is the chief glory of the Wave form, especially as it is so beautifully integrated with the rest of the speaker within a monolithic structure only four feet high (although pretty fat at floor level, to be sure). The design is claimed to effect a 40% reduction in enclosure volume in comparison with an unequalized woofer of identical performance. The vented box is detuned from the optimum Thiele/Small parameters, and resonant peaking in the electronic crossover unit boosts the response back to where it should optimally be--at the expense of power amplifier headroom, of course, which is a lot cheaper than gigantic cabinets. The woofer's near-field response-as we measured it at a summing junction of the driver and vent outputs, with the "low-frequency coupling" control on the crossover panel centered-had a perfect fourth-order Butterworth profile, i.e., maximally flat, with the -3 dB knee at 24 Hz, maybe 23 Hz with a little squinting. Add to that the long throw JBL woofer's large-signal capability and low distortion, and you are in low-frequency heaven, exulting in Craig Dory's 32' organ pipes and Jack Renner's bass drums. Hallelujah! (It was one of Mitch Cotter's witticisms that the "sub-" prefix in "subwoofer" had the same meaning as in "subnormal" or "subhuman." The integrated just-plain woofer of the Waveform seems to bear that out.) The electronic crossover brings in the two vertically deployed 6" midrange drivers at 130 Hz, with a third-order (18 dB per octave) slope and a Butterworthy corner; the woofer roll-off looks a little more subtle, showing some gentle attenuation well below that frequency but ending up with 18 dB per octave also. Of course, the low-frequency coupling peak introduced to help the box affects the overall profile to some degree. The 1" textile-dome tweeter, which is placed between the two midrange drivers, is crossed over at 2 kHz with fourth-order Linkwitz-Riley passive filters, as is the ribbon-type supertweeter on top of the array at 9 kHz. All drivers are off-the-shelf but strictly top-of-the-line; the supertweeter is particularly costly. (As we said, this is a highly optimized conventional speaker.) The high-pass section of the electronic crossover introduces a boost of a little over 3 dB at 16 to 17 kHz to equalize the supertweeter where it begins to roll off. The combined response characteristic of the midrange units, tweeter, and boosted supertweeter is very flat, both on and off the median axis (strikingly so off axis); the large peak at 16 kHz reported by Stereophile in their review was nowhere in evidence in our tests. (The most probable explanation of this discrepancy is that the very light ribbon depends on the air load for damping, and that load is much smaller in the thin air up there at 7000 feet in Santa Fe than at altitudes where less light-headed and scientifically more accountable reviewers dwell. All those Sherpa audiophiles, on the other hand, should heed Larry Archibald's and John Atkinson's findings.) The only possible anomaly we uncovered as we explored the semi near-field output of the speaker with our B&K microphone (the farfield being too contaminated with room effects) was that the 1" dome tweeter was set maybe, just maybe, 2 or 3 dB too high in level with respect to the midrange. This little

"step" in the response profile depended on the position of the measuring microphone; everything remained very flat above and below the crossover. We are told that the very latest modification of the Waveform, which we did not have, introduces a small amount of additional attenuation in the tweeter branch of the network, so our suspicions may have been right-even without the anechoic chamber of the NRC. In any event, we are talking about a rather trivial level balancing decision, involving +1 ohm of padding-not an engineering error. (Some reviewers are unaware of the difference.) The impedance curve of the unequalized woofer is exactly what one would expect in a vented system; that of the midrange/tweeter/supertweeter section, on the other hand, meanders somewhat intricately, both in magnitude and in phase, but not so much as to constitute a difficult load for a decent amplifier. The nominal impedance of 8 ohms appears to be a judicious average of these swings; the sensitivity (2.83 volts input) and efficiency (1 watt input) are therefore the same: 90 dB SPL at 1 meter. That is pretty phenomenal for a flat wideband system. All drivers appear to be in phase, but we had a great deal of difficulty trying to sum the output of all five drivers, or at least of the upper four drivers, to reconstruct a square pulse. Not surprising and not disturbing-this speaker is coherent to all intents and purposes. Distortion is extremely low; at levels that did not drive us out of the room the THD readings were minuscule, almost amplifier-like. We should also add that John Otvos supplied us with a set of NRC printouts documenting all sorts of measurements at their state-of-the-art facility, and none of these contradicted any of ours, except possibly the tweeter level matching.

What kind of sound results from all of this elaborate engineering? Incomparable bass, as we already stated; stupendous dynamics, without even a hint of stress on the loudest symphonic and operatic climaxes; almost totally neutral tonality, probably a little on the dry side for those accustomed to the sloppy, rounded-off "warmth" of various speakers; also, a mercilessly revealing, unforgiving character when reproducing juiced-up, over-bright, edgily recorded program material, of which there is still no dearth, alas. We had no problem whatsoever with our favorite recordings, but it is possible that the slightly reduced tweeter level in the latest mod will make the speaker more tolerant of the average studio production. On the other hand, listener position is relatively uncritical with the Waveform; there may be a best seat in the house, but there are many very good seats. Imaging and ambient nuances are not exactly the speaker's long suit; it is quite satisfactory in those areas but not nearly as stunning as, for example, the little Sapphire from Audio Concepts. Emulating the latter by covering everything but the diaphragms with acoustical foam would probably do wonders, but try to persuade John Otvos to give up that Bernini-in-wood shtick and put a black hood over his masterwork.

So, everything considered, this is indeed as good a forward-firing box speaker as there is (if you can call that fancy pyramid a box); it may have its minor weaknesses but it does all the big things, the difficult things, brilliantly.

It sounds like real music, even when the music is complex, loud, and dynamic; it can fill the largest domestic spaces at realistic SPL's; and it tells the truth about the program source. How many other speaker systems, at any price, can make those claims? Of course, it still sounds like a for ward-firing box speaker and not like a dipole or various other multidirectional designs. For that reason, it may not be everybody's dream speaker, maybe not even ours. But we feel that a forward-firing box speaker of this caliber had to be designed before one could decide whether or not that time-tested format is still the way to go. Remember, some rooms will not accommodate any other format. Dipoles, for example, have to be pulled far into the room to sound really good--let no one tell you differently-whereas the Wave form does not. Our judgment is that the Waveform fulfills the needs of a large number of audiophiles better than any other loudspeaker system, and especially so at $3995.

A few addenda and tidbits: The dedicated electronic crossover was driving two identical Boulder SOOAE power amplifiers in our listening tests. The circuit design of the crossover, although basically excellent, is not quite up to the Boulder level of sophistication. Occasional and not too disturbing RFI was clearly traceable to the crossover unit.

The speakers and the crossover we tested were the very same samples that The Absolute Sound had pooh-poohed without the courtesy of a full review. Nasty scratches in the black lacquer finish bore witness to careless, or possibly even contemptuous, treatment by slobs. In fact, the flat, mirror-smooth top of one of the cabinets had some interesting parallel scratch marks on it-as if someone had been cutting lines, man, with a razor blade. O tempora, o mores!

-------

[adapted from TAC 14]

---------

Also see:

The Carver "Silver Seven" Clones: Who Needs Tubes for the Tube Sound? Carver M-4.0t ; Carver "Silver Seven-t"

Philips Gives Some Special Fillips to Its High-End Audio and Video Line -- Philips LHH1000 ; Philips CDV488

Various audio and high-fidelity magazines