Today's Audio Equipment and the Reviewing Discipline: Where We Stand: Loudspeakers; Power amplifiers; Preamplifiers and control units; Turntables and tonearms; Phono cartridges; CD players; Other program sources; Speaker wires and audio cables.

--------

After an absence of six and a half years, it is obviously necessary for us to make a "statement of the art," defining our current position in the audio world, bringing up to date our preferences and antipathies in various component categories, and redrawing the line between valid equipment reviewing and self-satisfied subjective expertizing.

Audio in the late 1980's is not a good illustration of plus ca change, plus c'est la méme chose. Very little today is "the same thing" after all the changes of the past few years, even from a broadly philosophical point of view. Of course, Mozart is still the same and so are our ears, but the technology, the equipment, the expectations of the consumer, the available program material and the sales pitches have all changed so radically that any competent audio journalist's perspective also had to change in the process. In our case, that change in perspective may at first blush seem a little abrupt without a written record of its evolutionary stages over the past six and a half years; closer scrutiny of specific instances will reveal, however, that our basic standards of excellence and logic remain as before.

The equipment reviewer and the marketplace.

Throughout the late 1970's, in the golden age of The Audio Critic, good and bad equipment coexisted in the stores in every price category (except the lowest, where everything was bad) and competed for the audiophile's dollar with virtually equal plausibility. Shoppers who were not reasonably knowledgeable, or did not pay for reliable out side advice such as ours, were pretty much dead meat. That is no longer the case today. The low-priced stuff is now just about uniformly listenable, occasionally very good and hardly ever catastrophic. In the medium-priced range, the audible performance often approaches the current state of the art and the great disparities have almost entirely disappeared.

The high end and ultrahigh end are still a problem in terms of true value per dollar and genuinely sophisticated engineering, but it is no longer possible to buy dreadful sound for multiple kilobucks as in the wild and woolly days.

The major exceptions to these generalizations are loudspeakers, which have also improved but not nearly enough considering the available technology, and recordings in all formats, which are not "audio equipment" as such and do not compete at greatly different price points but certainly come within the purview of the audio journalist. The gradual convergence toward a single standard of quality is most evident in purely electronic components such as amplifiers, preamplifiers and control units, CD players, etc.

The so-called alternative audio press has not squarely faced up to the new situation, apparently regarding it as a threat to their importance as tastemakers. With only a few commendable exceptions, they still report exquisitely perceived differences where there are none and come up with scientifically unverifiable explanations for their subjective self-indulgences. They obviously believe that component A must "blow away" component B, especially when the large circulation hi-fi slicks have reported otherwise, if their readers are to feel like a highly enlightened In Group. The melancholy truth is that big bad Stereo Review and Audio, commercial, insincere and superficial as they are, have a firmer grasp of the nuts-and-bolts realities of today's audio technology than the supposedly purer and more deeply discerning esoteric journals. Julian Hirsch and Len Feldman may not point out the small but audible differences between the analog output stages of two fairly similar CD players, but at least they will not tell you to hook them up with cables made of phlogiston-free kryptonite to make the lower part of the upper midrange more liquidly multidimensional and thus reveal the vast superiority of A to B.

One might even be tempted to conclude that the evolution of the market and the resulting improved product have reversed the roles of the pop-tech reviewers and the high-end pundits. The former used to be wrong all the time with their everything-sounds-good Pollyannaism but are be ginning to be frequently right; the latter, who used to be right most of the time when they pointed out the major sonic differences that existed, are now getting panicky as the differences dwindle and taking refuge in cuckoo land , so they are wrong all the time. If that is too sweeping a conclusion, it is mainly because of loudspeakers, about which both contingents have always had the habit of being wrong, then as well as now, regardless of affiliation. Even the late great Richard C. Heyser, who operated on an intellectual plane far above that of either camp, did not make a solid connection in his technically brilliant articles between measurements and audible performance, perhaps because it would have been so devastating as to elicit front-office op position. If we were starting a new publication at this time instead of resurrecting an old one, it would have to be The Loudspeaker Critic because that is where the greatest weak ness lies in audio reviewing-and that is where a large part of our future efforts will be concentrated.

Criteria for valid testing and reviewing, What, then, is the correct and credible approach to evaluating a piece of audio gear for the sophisticated music lover and audio enthusiast? First, we must get rid of a tired old shibboleth founded on hypocrisy. We are referring to the familiar protestation of the subjective audio reviewer that his only standard is "the absolute sound" of live music.

How noble, how pure! Pure bunk.

Undeniably, the purpose of audio technology is to re produce the exact auditory experience originally produced by a live source of sound. Thus, if we were evaluating a total system of sound reproduction, from microphone diaphragms at the live source to loudspeaker diaphragms in the listening room and all the hardware in between, then the sound of the original, live event would indeed be the only valid criterion.

(That was the thinking behind the advertising as well as the critiques of the early Edison phonographs, and justifiably so, since the recording and playback processes were mirror images of each other with identical signal paths.) But to proclaim that a separate electronic component, such as a power amplifier, reproduces or does not reproduce "the absolute sound" is fatuously simplistic. The amplifier can only reproduce its own input and pass it on at a higher amplitude to a loudspeaker. Whether or not the resulting sound bears a close resemblance to live music as remembered and desired by a concertgoer depends on so many factors in addition to the design and construction of the amplifier that the mind boggles. In his heart, even the most pretentiously pure subjective reviewer knows that, but it is hard to give up well worn, facile answers to difficult questions.

Our own equipment reviewing procedure, as we now intend to standardize it in The Audio Critic, comprises three important stages. First, a component must be screened for review worthiness, to coin a word. There are thousands and thousands of models listed in the annual directories, and hundreds of new ones are introduced between issues (even when published on schedule). No audio journal can address itself to more than a small fraction of them, but what are we to do every time someone raises his hand and calls out that his is the best, or at least the best for the money? We need some sort of rationale for focusing our attention. It could be a whole new technology or just an interesting new circuit, the latest effort of a highly respected maker or the debut of serious new talent, a previously un-reviewed classic or a particularly polished and promising new execution of a reliable old idea-whatever it is, something distinctive. We cannot just proceed because the product is available. Thus, a new speaker system consisting of a woofer, a tweeter and a first-order crossover network in a rectangular box would not get past our screening process unless the woofer or tweeter or box were made of a special material, or some special claim were made for the tuning of the box, and so forth.

Another reason for screening out a piece of equipment, even if it were of a highly intriguing design, would be a total lack of practicality, common sense or safety. Dr. Hill's bizarre Plasmatronics speaker system comes to mind, with its large thirst for compressed helium in industrial cylinders and the telltale odor of poisonous ozone in the room where it is playing. Thank you-next! The screened, review worthy equipment is then passed on to the second stage, which consists of laboratory tests and measurements. These are absolutely essential to the evaluation process and 'will be discussed in full detail in the context of specific component categories and individual re views. The purely subjective reviewer will insist that all such efforts at the laboratory bench are irrelevant because there is little or no correlation between measurements and sound quality, and no laboratory report can change his mind about what he is plainly hearing, anyway. Again, pure bunk. Loudspeakers, to bring up the most painfully obvious example, exhibit dramatic differences in a number of easily measurable response characteristics, and their sound varies accordingly and quite predictably. Electronic gear is a bit more subtle, but anyone who has read the articles on the Carver "t-mod" projects in this issue will have an idea just how closely measurements are related to amplifier sound.

The main purpose of measuring a component before listening to it, or even after having listened to it, is to determine exactly how, and exactly how much, it deviates from the theoretical ideal for such a device, from its Platonic form so to speak. Audiophiles are familiar with the concept of a straight wire with gain as the model for the ideal amplifier; similar models can be set up for other components, though not always so neatly. The point is that when we measure an amplifier, we are quantifying its resemblance to "the absolute amplifier," which is a better clue to audible quality within a system than... you already know what.

That bring us to the listening tests, the third stage of evaluation. It can be dispensed with, unless sheer curiosity prevails, if the equipment has been found seriously delinquent on the lab bench. Just as the taste of lobster with chocolate sauce is of no interest to the gastronome, the sound of its audio equivalent is of no interest to us. There is no reason to try it. The object of extensive listening tests, such as we insist on in certain cases, is not to experience the full range of "the good, the bad and the ugly" but to sort out the nuances of good, better and best sound, see how they correspond to engineering features and measured characteristics, and determine whether they are consistently distinguishable by the trained ear. There has been a great deal written and said on this subject, and it is not our intention to present an entire philosophy of critical listening within the confines of this article. Our current views will become fully evident from our treatment of individual products; meanwhile we want to establish our position on a few very general points.

Folklore and reality in listening tests.

If you read of some of the more precious high-end periodicals and amateur audio-society reports, you are undoubtedly familiar with the sensuously lingering style in which their writers relate their impression of a particular component's sonic anatomy, in almost pornographic detail.

"The warm fullness of the upper bass is emphasized by the slightly recessive lower midrange; the upper midrange could be more liquid and is slightly hooded at times; the highest highs are too silky for complete realism, but the lower highs are quite incisive yet nicely rounded and almost free of grain." When a fellow thinks like that about a piece of audio equipment, we would hesitate to shake his hand after a listening session.

Seriously, though, that kind of super specific characterization of reproduced sound has no foundation in reality.

Yes, we are willing to contemplate the possibility that somebody actually heard something just like that from a given seat, in a given room, listening to a given recording, through a given stereo system, with the component under test inserted into that system. Now, change any one of the givens, or let the listener just move his head six inches, and those exquisitely delineated specifics are no longer exactly the same. They are not consistently applicable to the same component under different conditions. If there existed an absolutely perfect stereo system made up of absolutely perfect components, and one substituted for one of the latter a new and less-than-perfect component, the changes one might hear could probably be described in terms such as loss of clarity, veiling, less sharp focus, less information, thicker or thinner or coarser textures, reduced depth, less air around the instruments, bloated or constricted stereo image, loss of directional clues, shifting localization, stridency or dullness, etc. General imperfections like these could be assumed to be intrinsic to the component itself and therefore relevant to a critique; however, if one heard a rainbow effect of octave-to octave changes, some for the better and some for the worse, they would surely be due to local conditions, interface peculiarities and other non-recurrent causes, if not entirely a figment of the imagination. Thus an authoritative report of a listening test is always fairly general though unequivocal; only the dilettante dwells lovingly on the 128 facets of his kaleidoscopic misperceptions.

Another preciosity we have little patience with is a hairsplitting description of every nuance of imaging and soundstaging, as if equipment designers possessed a varied palette of these ingredients, from which they decide to apply larger or smaller portions to their creations according to their personal taste and style. Rembrandt always put in those rich browns and golden yellows, and Conrad-Johnson always puts in this fantastic front-to-back depth, you see...

The fact is that correct design, with accurate performance in the time domain, automatically results in the proper rendition of time-related information in the program material.

Whatever you want to call such information-imaging, soundstaging, depth, width, localization, or anything else of the sort-it consists of complex time relationships, which must be preserved undisturbed in the playback. And that, amigos, is not a separate performance feature that can be dialed into a design; it simply comes with the territory when all is well in the time domain. Nobody even has to talk about it. For example, beveling or rounding the front edges of a rectangular speaker enclosure reduces diffraction effects and thereby improves the coherence (i.e., time domain accuracy) of the wave launch. Presto, you have clearly better imaging, soundstaging, etc., than with sharp edges. Not because of the magic touch of some great design guru who lies awake at night thinking about the position of the piccolo, but as a result of good engineering practice based on the laws of physics.

It should be added that dozens of microphones going into as many channels, and then mixed down, equalized, overdubbed, reverbed and otherwise console-processed, will not yield a stereo signal in which time relationships are faithfully preserved. To talk about authentic imaging or depth or natural space in such cases is like pretending to discern fresh dairy tastes in a processed cheddar spread. The whole thing is a dead issue except when the microphoning is quite simple and console shenanigans are at a mini mum-hardly the usual situation. And then there is that question much dreaded by high-end party-liners: just how much directional and spatial information is available to the ear at a live concert, anyway? Our typical experience has been that, with eyes shut and no cheating, one can distinguish left, right, center, front and rear, and not much more.

A very few additional in-between sectors, maybe two or three, become vaguely apparent as one moves to the front row, which is more or less representative of the expected distance of microphones from the musicians. Under no circumstances can one hear anything resembling the pinpoint localization and directionality craved by the imaging addicts.

That is always an artifact of the recording process and possibly of the playback equipment, e.g., a very beamy speaker on axis or something similar.

We once heard Oscar Peterson make a comment to the effect that there may be a lot of controversy about the correct definition of the blues, but when you hear the real thing you know it. The same is true of imaging and soundstaging. Whenever, in our listening tests, we play an un-gimmicked, simply miked recording through electronics and, especially, speakers that are reasonably free from time dispersive anomalies, the spatial/directional effects we hear sound natural and unproblematic in terms of our live concert experience and not at all like the aural fetishes of the cultists. (Even Bob Carver, who is not one of the latter but still in love with his Sonic Hologram after all these years, will admit in an unguarded moment that the darn thing should be turned off and the music enjoyed straight when the program material and the stereo system are as we just described. Lifelike sound is very holographic.) Double-blind listening tests.

As you must have noticed by now, our opinions on critical listening have been highly polarized by what we perceive as the excesses of the purely subjective reviewers; on the other hand, we still believe that the ear is the final arbiter of quality in audio equipment and are always eager to find ways to impose some sort of objective discipline on our listening tests. We have therefore purchased an ABX Double Blind Comparator from its designer, David L.

Clark, and are planning to use it extensively in our future listening evaluations. The few specific test reports in this issue do not yet reflect this capability and methodology, but the ones in the next issue will; until now we have concentrated on studying the overall possibilities of ABX testing and sorting out our own thoughts on the subject. Let us just state at this point that Dave Clark happens to be one of the clearest thinkers among today's audio technologists and that the tweako press has most unjustly identified him (and by implication his ABX system) with the everything sounds-the-same school of know-nothingism. We may not agree with all particulars of his controversial Stereo Review listening surveys, but we subscribe 100% to his basic tenet that anyone who claims to hear differences between two pieces of equipment should be able to prove to others that he really does. Who can argue with that? As far as the actual ABX hardware is concerned, it merely does more neatly and effortlessly what can be done, and has been done, with tedious manual A/B switching and randomization by coin tossing. Every time the logic/display module of the ABX system is powered on for a listening session, a random series of sequentially numbered trials is established, in which X = A or X = B, and only the module knows which is which. The hand-held remote control unit used by the listener has buttons for A, B and X, as well as Up and Down buttons to go on to the next trial or switch back to a previous one. Thus the test is truly double blind, since neither the listener nor the test giver (if there is one) knows the identity of X. One can switch endlessly between identified A and identified B to learn the sonic signature of each, and then try X, switching back to A or B for verification and again to X, all without compromising the validity of the test. Once you hit the Answer button on the logic/display module, the test is over; all you can get is the identity of X in each numbered trial for scoring purposes.

We are completely satisfied that the headphone-type plugs and jacks as well as the relay contacts of the ABX system are transparent to the signal; they introduce no more loss or degradation, if any, than the plugs, jacks and signal switches of a state-of-the-art preamplifier. Those who try to shoot down ABX test results on that basis are doing so without ammunition. Any exceptions taken to such results would have to be founded on specific weaknesses of procedure in individual cases; there exist no general grounds on which ABX testing can be condemned, in our opinion. If you are only familiar with the golden-ear protests against the system but not with the system itself, here are some basic facts to keep in mind.

To begin with, any two audio components of the same general category (other than loudspeakers) will sound astonishingly similar if listened to at exactly the same volume level, matched within £0.15 dB or better. We are not suggesting that they will sound indistinguishable from each other, but the similarity will in most cases be almost frightening to those who have never tried the experiment before. Casual A/B comparisons in an audio showroom or at home simply do not prepare you for this phenomenon, which in a formal ABX listening test at matched levels can result in stress due to prolonged concentration, even when differences are reliably detected in the end. Our belief is that maybe one audiophile out of ten is genuinely suited to be a panelist at an ABX session that will be written up, as the task requires more than just the desire to participate. For valid conclusions at the highest level of expectations, the requirements include an exceptionally keen ear, considerable previous training in listening for small differences, a long attention span, high resistance to stress caused by listening fatigue, and complete sincerity, which is especially important because it is often easier to start guessing wildly than to keep on concentrating and making considered choices.

Let us examine a hypothetical situation in which two components, A and B, have been compared through the ABX system by a panel of listeners and correctly identified only 50% of the time on an average, with no individual panelist doing much better or much worse than that. Since sheer guessing can be expected to yield the same percentage, a "no difference" result is declared, and various groups of cultists go into orbit. We would in such a case consider the following possibilities: (1) A and B are in fact sonically in distinguishable from each other; (2) A and B are actually distinguishable, but only by exceptional ears, of which there were none on the panel; (3) A and B are sufficiently similar to require a stereo system of exceptional resolution to be distinguishable, and the system used was inadequate; (4) the panelists were so poorly motivated and insincere that they proceeded to guess at random when sustained concentration became uncomfortable but did not tell anyone; and (5) someone was hogging the remote control unit or was officially assigned to be its sole operator, and the others got tired of asking him to switch this way and that way, finally just writing down their choices before they were quite ready.

We find it quite depressing that those who claim to hear large differences in their conventional, un-rigorous listening tests are always ready to denounce the ABX approach in toto before eliminating possibilities (2) through (5), which are very real and could be supportive of their arguments.

Our own ABX experiences lead us to believe, however, that (1) is the case more often than not.

There is a lot more to be said on this subject and will be said in future issues. For the moment we just want to go on record as follows: Any self-styled expert who publicly declares that A sounds unquestionably superior to B, and is then unable to distinguish A from B in an unhurried blind test at matched levels through his chosen reference system, is a despicable charlatan. If the shoe fits, wear it.

The digital revolution.

Before we summarize our stand on some of the currently available audio components in specific categories, it behooves us to make a few comments on today's pervasive digital technology, which was still in its infancy when The Audio Critic ceased to publish in 1981. In our last full-size issue, we printed a short filler item with the headline, "Help Stop the Digital Epidemic!" As successful propaganda, that must rank with the McGovern presidential campaign and the Edsel automobile promotion. We are now aware that our not altogether unjustifiable little tantrum (triggered by some god-awful early examples of the genre) was based on incomplete information. More than a few of our fellow journalists are still arrested at that level of perception, but we shall try here briefly to offer more up-to-date insights.

A fair assessment of the digital approach to sound re production must address three entirely separate issues. The first is the concept itself, which is beyond reproach. It used to be a truism that nothing in this world is black and white; everything is a different shade of gray. Well, in the world of digital data processing that is no longer the case. When you chop things into small enough pieces, all those subtleties can be characterized by 0's and 1's, either-or, black or white, nothing in between. All vagueness is gone and with it the possibility of various slippery inaccuracies generally referred to as distortion. Of course, the digital concept creates its own class of errors and consequent distortions, but these are not as formidable as the elusive gremlins of analog processing and can be reduced to insignificance with fairly straightforward techniques. Theoretically, digital is the way to go-neater, more foolproof, better.

The second issue is the adequacy of digital recording and playback standards, present and future. That chopping up of information into small pieces and the retrieval of the pieces must, of course, proceed according to fixed protocols, and these must be sophisticated enough to permit results that live up to the theory. The key specifications in such a standard have to do with digital word length (quantization) and sampling frequency (or Nyquist rate). At the time we expressed our doubts about the then young digital audio technology, the state of the art was a 16-bit linear PCM system with 50-kHz sampling. To our surprise, that has not changed to this very day. We keep reading and hearing about new systems using more bits and higher sampling frequencies, but we have yet to see and listen to one, even in early prototype form. The general consensus has been that pushing the limits of technology in this area may create more problems than benefits. The CD standard ended up specifying 16-bit encoding with 44.1-kHz sampling, and we now believe that this is an adequate standard per se, in the sense that it puts no inherent limitation on the audible results as long as the hardware implementation is highly refined. That was definitely not the case at first, and it took us a while to realize that the faults we kept complaining about were not intrinsic to the established parameters of the system. The DAT (digital audio tape) standard has evolved a little differently but will prove to be equally satisfactory un less the monumentally stupid copy-code chip is legislated into it. (A separate article on that subject is in the offing.) Thus we come to the third and decisive issue, namely the hardware that implements the unexceptionable concept and the quite adequate standard. Ay, there's the rub. As in the initial phases of other major breakthroughs in audio (LP records, stereo, cassette recorders, etc.), the first-generation equipment left a great deal to be desired. The earliest Sony professional PCM recorders delivered to the studios had some unsolved problems that ended up in the sound of the highly promoted early CD releases; the D/A converters in the first-generation CD players were quite unsophisticated; a lot of refinements that should have been there from the start came quite late in the game. For that reason, a fair critique of the capabilities of digital audio would have to be in terms of the most recent CD's, produced by the technically most progressive element in the industry and played on the very latest CD players using 16-bit dual D/A converters with four-times oversampling digital filters, advanced error correction circuitry, etc. There is no doubt in our mind that such 1987 digital sound represents the state of the art in home stereo reproduction. The complete freedom from all nonmusical noise, the virtually unlimited dynamic range, the unshakable pitch, the clearly defined bass are irrefutably convincing, and the highs have at last been tamed to our satisfaction. It matters little that the best commercially available stereo sound we have heard so far came from a few direct-to-disc analog recordings; that was the outcome of a selection process spanning several decades and based on thousands of LP samples, whereas our favorite CD's were picked out of a hundred or so over a couple of years. On the negative side, there is no CD so horrendous in sound that we cannot find a fairly recent analog LP that is even worse.

It should also be remembered when making these analog vs. digital comparisons that there remain a lot of analog stages in nearly every signal path we think of as digital.

Ideally, the output of each microphone at the live source should go into an A/D converter, the signal kept in the form of 0's and 1's throughout the recording and playback process, then decoded through a D/A converter placed as far downstream in the circuit toward the loudspeaker input as possible. In typical current practice, the signal undergoes much more analog processing than that; for example, just about all mixing consoles are still analog, the new digital ones being very expensive and scarce. In CD players, too, the analog output stage is subject to all the miseries ever experienced with analog preamplifiers, such as degradation of the signal through inferior coupling capacitors, etc. The purely digital era is not here yet. Many recording engineers still have analog reflexes as a result, compensating in their microphone setup and digital masters for anticipated analog type losses and masking-and getting nasty exaggerations instead. It will take years before digital techniques become the cozy native idiom of all audio practitioners; until then all kinds of awkwardnesses can be expected, but the essential superiority of the new medium is no longer in doubt.

So much for generalities. When it comes to actual equipment recommendations and dis-commendations, we shall make no futile attempt to fill in a gap of more than six and a half years, during which we did relatively little comparative testing except to satisfy our own personal needs and those of a few friends. Now we are once again in the midst of ongoing equipment tests to provide material for full-length reviews, beginning in this issue but mainly in others to follow; the frustratingly sketchy and haphazard summaries below are presented for the purpose of further positioning our current viewpoint, and to serve as first aid in case of desperately urgent shopping.

Loudspeakers.

In comparison with highly developed components such as amplifiers, all loudspeakers are bad, but some are less bad than others. One major problem is that there is no agreed-on theoretical model for the perfect loudspeaker, equivalent to the straight-wire-with-gain ideal in amplifier design. Is the perfect loudspeaker a point source? A line source? A figure eight dipole? A collimated source, i.e., a plane wave from the start? A multiple point source that synthesizes a coherent plane wave-where? There are as many answers as there are speaker designers. If there is no clear definition of what is theoretically best, how can the existing solutions be good? It is indeed a puzzlement.

That is one reason why we have a great deal of respect for the Quad ESL-63 electrostatic loudspeaker (suggested re tail price $3600.00 the pair in the new, slightly sturdier version with the US suffix-you should have paid attention when it was $2450.00). The design makes a commitment, without any pussyfooting, to a very specific model-a di pole source with hemispherical radiation, creating a virtual point source about a foot behind the diaphragm-and then implements the concept to the limit. No other speaker we have ever tested has an output that so closely resembles the input in every way, over a relatively large solid angle. It is definitely the most accurate, most neutral speaker known to us, as long as it is not stressed beyond its disappointingly limited dynamic range, at which point all bets are off. As for bass, there is a broad bump centering on 60 Hz and no significant output below 40 Hz. A chamber music, jazz and solo vocalist speaker par excellence, it is most emphatically not for showing off your favorite Telarc blockbuster CD.

There is no other electrostatic design, however, that we can wholeheartedly recommend, although we are quite familiar with most of the currently fashionable models.

In more or less conventional electrodynamic speakers with woofers, tweeters, etc., the weakness is usually in the crossover network and/or the tuning of the bass enclosure to the woofer. Sophisticated computer optimization can do wonders in these two areas; that is where the strength ought to be in a modern design, since nowhere else are good and bad solutions so similar in production cost. But no-most speaker designers just keep flying by the seat of their pants, as if the network analysis, filter synthesis and brute-force optimization programs were as recondite as Edward Witten's string theory. (See also the "disclosure" article on the now defunct Fourier loudspeakers in this issue.)What do we like in this category? The B&W Matrix 801 Series 2 looks like a very intelligent design to us (at $4500.00 the pair it had better be), with full evidence of solid computer work behind it; furthermore it sounded excellent when we auditioned it, although in all honesty we are much more familiar with the older, plain-vanilla 801 and 802. We are less sure about the new darling of the ultrahigh-end crowd, the giant Duntech Sovereign 2001 from Australia ($15,000.00 the pair); we have heard it sounding both good and bad in various places on various occasions, but we shall give it the benefit of the doubt until we can check it out more closely, especially the crossover network. It is certainly not a negligible item. At the other end of the scale, among tiny boxes, the Celestion SL600 ($1797.00 the pair) has been the prestige model for years. We have measured it and listened to it; other than the very fine tweeter, there is nothing remarkable about it. No bass, so-so midrange, roller-coaster impedance curve-if it sold for $475, people would say it sounds nice and clean, and that would be the end of it. A much better value per dollar, albeit for a totally different buyer, is the Spectrum 108A ($249.00 the pair). Cosmetically appalling but quite musical and balanced in sound, it even produces a decent amount of bass for a really small box, and our laboratory tests revealed nothing shameful (nor anything amazing). A smart bargain-basement product with a touch of audiophile appeal. A more high-tech small speaker that we respect is the Spica TC-50 ($450.00 the pair); computer-aided design makes its response very accurate and the sound beautifully uncolored, but it is still essentially a small-signal system with very limited bass.

Our pet peeve as we look at the electrodynamic scene is the entire line of Polk SDA Series speaker systems (from $799.90 to 2990.00 the pair). Stereo Dimensional Array (SDA) is Polk's name for what is nothing more than a naive and inferior version of Bob Carver's old invention, Sonic Holography, already referred to above. Signal processing of this sort should always be used selectively (if at all), with caution and good taste; Carver's active circuitry and controls make that possible, but Polk's implementation of the concept is passive within the speaker systems and is not intended to be turned off, ever. The permanently built-in holographic processing, whether or not appropriate to the program material, makes the speakers sound "different" and helps to cover up their intrinsic mediocrity of performance.

We do not expect the large-circulation hi-fi magazines to point this out, since Polk is probably their heaviest buyer of full-color advertising pages, but we are offended by the carnival-toned ballyhoo and hoopla in all the ads, making outrageous claims of utter originality for Polk's copycat technology and featuring Matthew Polk as a grotesque wax museum figure of a "genius" in white laboratory coat. The fact that this kind of lowbrow marketing works, as it obviously does, is a sad commentary on the audio marketplace.

We must not forget the relatively new category of full-range (or almost full-range) ribbon loudspeakers. These are especially intriguing because they combine the virtues of the force-over-area principle used in electrostatic designs with the ruggedness and large-signal capability of the better electrodynamic systems. The new Apogee Diva ($7000.00 the pair) is claimed by its makers to "redefine the state of the art" in full-range ribbon technology; we have not tested it but have heard it demonstrated under supposedly excellent conditions and are only mildly impressed. Apogee ribbon systems (there are now five of them) can sound extremely transparent and uncolored or quite disappointing, depending on the room and the ancillary electronics; Krell amplifiers are generally recommended because of their ability to drive very low impedances. We tend to suspect unsolved design problems when a very high-tech audio component produces inconsistent results, but we should not jump to conclusions before some thoroughgoing laboratory tests. The Carver ribbon speaker system ($1576.00 the pair), modestly named The Amazing Loudspeaker, has suffered from inept demonstrations at trade shows and in dealers' showrooms; it also needs to be broken in, alas, to lower the fundamental resonant frequencies of the ribbon and the bass drivers. In the right hands, the speaker is capable of outstanding results that belie its price; we have done some preliminary testing on it, and a full-length review is coming in the next issue.

Power amplifiers.

There is enough material on power amplifier design elsewhere in this issue to make general philosophical observations unnecessary here. We are restricting ourselves, therefore, to a few comments on specific makes and models not reviewed in full.

Although we always suspect any ultrahigh-end audio product of being opportunistically priced, the Krell line of pure-class-A stereo and mono power amps ($2550.00 to $16,000.00 per two channels) can at least offer a plausible defense against that charge: the uncompromising and unarguably costly design of the Krell power supplies. These have sufficient current capability to permit the rated power output to double whenever the load impedance is halved, all the way down to 1 ohm, and to maintain stable operation even into 0.1 ohm. Is that really necessary? According to a recent paper by the redoubtable Matti Otala it is, and who are we to contradict him? Nor can we contradict designer Dan D' Agostino when he opts for straightforward circuitry, minimal feedback, discrete solid-state devices and various little touches of the latest available technology. With that approach, what could be bad? Even on the basis of our somewhat limited laboratory and listening experience with Krell amplifiers, we are not afraid to endorse them as long as price is not a consideration.

Ultra-high-priced vacuum-tube power amps are another matter altogether. We have serious reservations about either the performance or the practicality, or both, of all models known to us. Proceed at your own risk.

Getting back to the real world where satisfaction is at least vaguely related to the price paid, we can recommend the Adcom GFA-555 ($699.95) as an outstanding buy. Our measurements as well as our listening tests qualify it to be considered competitive against other 200/200-watt stereo power amplifiers at far higher prices. The next significant step up in the same power class would be the Aragon 4004 ($1495.00), a very nicely built unit we have auditioned but not yet measured; our initial impressions are excellent, and a full-length review is forthcoming. We also have a good feeling about the Tandberg line of no-feedback power amps ($1095.00 to 3595.00), after looking, listening and considering their design philosophy. Complete laboratory tests are in the planning stage.

Preamplifiers and control units.

It is becoming quite clear that the program sources with a long future are the ones that have a line-level output and need no preamplification: CD players, cassette decks, Hi-Fi VCR's, DAT decks and, of course, tuners. As the phono arts approach their Gotterdammerung, there are even super cartridges that deliver a line-level signal (see the Win FET-10 below). Even the best phono preamps today are like elder statesmen, highly respected but not included in long-term plans. Most serious audiophiles either have a pre amplifier/control unit they are satisfied with or are ready for their last one. We have done very little investigation in this area during the past few years; some of the equipment we still enjoy using is no longer on the market. Here are a few stray thoughts on the subject:

To route and control the various program sources in a phono-less system, of which there are more and more, the entirely passive (i.e., unamplified) Mod Squad Line Drive ($400.00) seems to be made to order. In a pinch, you could even plug a self-powered outboard phono stage into it; a very good and inexpensive one, for moving-magnet pickups only, is the Phenix P-100-MM ($149.95). If you still want a high-end-ish phono-and-everything unit with sophisticated MC and MM facilities at a not quite insane price, we can recommend the bottom-of-the-line (some bottom!) Krell PAMS-5 ($1550.00), which we have tested on our lab bench and in our listening setup. The RIAA equalization error is merely small instead of nonexistent; the high-level stage and controls may be a tiny smidgen short of straight-wire like transparency; but the overall performance is superb.

For a lark, try to find a used but not abused sample of the late Stew Hegeman's 1959 (or was it 1960?) vacuum-tube kit preamp, the Citation 1 , and have somebody who knows what he is doing restore it to perfect operating condition. It is not difficult, and you will be amazed by the sound. (See also the full-length review of the Audio Research SPII pre amplifier in this issue.) Turntables and tonearms.

We are still using the Win Laboratories SDC-10 turn table with SDA-10 tonearm we designated as our "Reference A" before we stopped publishing in 1981. We never felt a need to upgrade, although a later version of the turntable with a more sophisticated drive system is available in very limited quantities under the name of Win Research SEC-10 ($4000.00). The SDA-10 arm is now history; the ingenious Win pantograph tonearm seen in prototype form at some trade shows in past years may eventually replace it.

Some audiophiles are undoubtedly looking for the last turntable and arm they will buy (as in the case of phono preamplifiers above); we are in no position to be dogmatic in our recommendations to them, but we can make a few modest ones. General rule: belt drive works very well in both high-end and moderately priced turntables; direct drive is suspect at low and medium prices but can be great at the highest levels of refinement in high-priced turntables. Also, stay away from turntables that do not have a suspension, no matter what the technical rationales are and regardless of price. (No dogma, just common sense.) Lastly, be wary of unipivot tonearms; other things being equal, four-point gimbals suspension is greatly preferable because it does not permit rotation around the longitudinal axis of the arm.

Those who hesitate to make a major investment in the waning phono medium but still insist on a certain measure of quality should consider the Harman Kardon T65C ($575.00). The arm may not be quite as good as the turntable, but the system as a whole is surprisingly close to many high-end products in performance. The current, fully updated version of our old whipping boy, the Linn-Sondek LP12 ($945.00 without arm) is a very fine turntable, much as we hate to admit it. Not quite as completely debugged is the SOTA Series III line ($975.00 to 1995.00); even so, these turntables have many outstanding design features, not the least of which is vacuum clamping of the record in the higher-priced models-yes, it makes a difference. On one of our slumming expeditions we auditioned the massive Basis "Debut" turntable ($5000.00) with the Air Tangent straight line tracking tonearm ($2850.00) mounted on it. Next to something like the $17,900 Goldmund Reference, this is still proletarian equipment; we refuse to call it overkill without a complete laboratory test, especially since the arm strikes us as having possibly solved some major problems in SLT and air bearing design. (Yeah, the sound was good.) Other, more mainstream tonearms that impress us favorably are the SME Series V ($2025.00) and, considering price, most of the Micro Seiki line ($195.00 to 1200.00).

Phono cartridges.

Sao Zaw Win, where were you when we needed you? The most sophisticated of phono technologists, after a decade and a half of piddling around with super products that were barely available, "soon" available, almost unavailable, limited production, just discontinued, etc., seems to have finally gotten his act together-now that the phono arts are about to go bye-bye. His revolutionary (the word is appropriate for once) field-effect transducer, the Win FET-10 ($1850.00, complete with electronics) promises to be a real-world production item with real-world distribution, at least as it looks to us from here. We have seen it and even heard it briefly, so we can vouch for its actual existence and its ability to produce beautiful sounds. A full-length review is scheduled for the next issue. This is not a cartridge that drives a FET at the input of a preamplifier circuit; the FET is the cartridge, or rather the transducer element within the cartridge. The gate of the FET is physically separated from the semiconductor substrate and is attached to the stylus cantilever. The consequences of this entirely new design principle are far-reaching and nearly all favorable; the review will go into full details, but just for openers-no preamp is needed. Unless our tests reveal something unforeseen, we may very well have found our last phono cartridge. It should be added that, regardless of the transducer principle used, Sao Win has always been at the leading edge of stylus technology. In the FET-10, he goes a step beyond his (and everyone else's) most advanced previous geometry with the longest line-contact footprint ever produced, but without the bottoming and misalignment perils of the Van den Hul and similar designs.

It so happens that our reference MC cartridge of the past few years has also been a Win product, the now discontinued Win Jewell (sic). Its last officially listed price was $475.00, just in case you try to locate one (lots of luck); we found it to be distinctly superior to some of the priciest Koetsus and other high-end MC's. The only exception to that may be the High-phonic MC-D15 ($1500.00), which we never had a chance to compare with the Win in the same system but found well-nigh flawless in an extended listening test. In moving-magnet and moving-iron cartridges, we have a certain respect for the top-of-the-line Shure and top of-the-line Grado (also Joseph Grado Signature) products, and that just about does it.

CD players.

We are planning a lot of coverage in this category for coming issues, starting with the next one; here we merely want to go on record in favor of the latest Philips system- 16-bit dual D/A converters, digital filters, 176.4-kHz over sampling, unique error-correction IC's, other goodies-in its various European, Asian and stateside-modified incarnations. Which of these will end up as our top choice remains to be seen. Early versions had a least-significant-bit error in the converter; we have no idea whether or not this almost surely inaudible little bug has been removed, but the overall system is still the most sophisticated around.

Whatever you do, stay away from CD players priced over $2000; the technology has not yet reached a plateau, things are changing, and this year's high-end cult favorite may turn out to be next year's electronic paperweight.

Other program sources.

We have no urgent opinions to communicate on the subject of tuners, open-reel tape decks, cassette decks or even DAT decks, although we are desperately rooting for a quick and rational resolution of the insane political imbroglio that has been stifling the DAT medium. A ray of sun shine through that cloud is vaguely discernible now that the decision regarding the audibility of the copy-code chip rests in the hands of the National Bureau of Standards.

Speaker wires and audio cables.

We have not changed our mind on this subject since our 1979 article. Audible differences between various types and brands do exist, but in every case they are explicable in terms of resistance, capacitance, inductance, dielectric, shielding, RFI, diode effects (i.e. rectification) and other known electrical phenomena. Any connection between two pieces of audio equipment can be analyzed as a network and is subject to all the laws of network theory. Furthermore, the output and input impedances interfacing through the connection are every bit as important to such an analysis as the network characteristics of the connection itself. Without specifying a signal source, a signal and a termination, a connection cannot by itself be described as good or bad.

Most wire and cable advertising is therefore superficial hype if not outright nonsense. Some advertisers seem to think they are addressing the microwave business, not audio.

In nearly all cases, a direct connection between two audio components, with little or no wiring, can be assumed to represent the theoretical ideal. For example, a mono amplifier can be lined up behind a speaker in such a way that the amplifier output and speaker input terminals almost touch. You can then make a connection with an inch or two of bus bar or braid. If your favorite loudspeaker cable sounds better than that, you are in deep trouble, amigo. The same goes for plugging, let us say, a CD player into the high-level input of a preamp by means of a pair of a male to-male phono connectors without cable. Do the makers of zillion-dollar silver cable believe they can improve on that sound? The only high-end wire and cable company known to us that openly admits these realities and accepts a direct connection as the ultimate sonic standard is Straight Wire; how well their products maintain that credibility in actual performance comparisons with competitive brands will be reported in a forthcoming review. Here we shall restrict our selves to the confession that we are currently using the somewhat plebeian Mogami Neglex 2477 low-inductance cable between our power amplifier and speakers-and like it. The high-end police will be knocking on our door any night now.

One more thing...

We trust our former readers remember and our new readers will realize that the quickie equipment survey above is not our reviewing style. At the risk of being repetitious, we want to emphasize again that these capsule observations are offered faute de mieux as a necessary stopgap. Check out the few full-length reviews in this issue for a better idea of how The Audio Critic goes about its business.

--------

[adapted from TAC, Issue No. 10]

---------

Also see:

Landmarks in Power Amplifier and Preamp Design: Tubes vs. Transistors vs. Both: Audio Research M300; Audio Research SP11; Boulder 500 ; MESA/Baron M180 (modified)

Various audio and high-fidelity magazines