Volume 1, Number 4 July/August/September 1977

In this issue:

The speaker survey at last, with 15 systems from $224 to $5200 a pair compared in this first go-around. (The $5200 one sounded best, alas.)

Our power amplifier survey continues in a rather positive vein, especially since this second batch includes our new reference standard. We launch what may be our most important and, to some, most disturbing inquiry so far: an updated investigation of the cartridge/arm/turntable relationship. Including 10 tone-arm and turntable reviews for openers.

Plus, of course, our regular features. A Vol. 1, No . 4 July/August/September 1977

(titie Editor and Publisher Peter Aczel Associate Editor Max Wilcox Graphic Designer Dick Calderhead Business Manager Bodil Aczel Assistant to the Publisher Elizabeth Tinsley

Consulting engineers and other technical advisers are engaged on a project basis, some contributing under their by-lines, others working anonymously.

The Audio Critic is an advisory service and technical review for con­ sumers of high-priced audio equi pment. It is published six times a year by The Audio Critic, Inc., and is available by subscription only. To maintain total dedication to the consumer's point of view, The Audio Critic carries no advertising by equi pment manufacturers, distributors, reps, dealers or other commercial interests. Any conclusion, rating, recommendation, criticism or caveat published by The Audio Critic represents the personal findings and judgments of the Editor and the Staff, based only on the equipment available to their scrutiny and on their knowledge of the subject, and is therefore not offe red to the reader as an infallible truth nor as an irreversible opinion applying to all extant and fo rthcoming samples of a particular product. Address all editorial correspondence to The Editor, The Audio Critic, Box 392, Bronxville, New York 10708.

Contents of this issue copyright© 1977 by The Audio Critic, Inc. All rights reserved under international and Pan-American copyright conventions. Reproduction in whole or in part is prohibited without the prior written permission of the Publisher, which will be auto­ matically denied if the reproduced material is to be juxtaposed to advertising copy or any other additional text serving a commercial purpose. Paraphrasing of product reviews for advertising or other commercial purposes is also prohibited. The Audio Critic will use all available means to prevent or prosecute any such unauthorized use of its material or its name.

For subscription information and rates, see inside back cover.

· Contents 18 Sophisticated Speaker Systems, Large and Small: A Comparative Survey. By the Staff of The Audio Critic 21 Acoustat X 23 Beveridge Model 2SW 25 Braun 'Output C' 25 Braun L200 25 Canton LE 400 26 Cizek Model # 1 27 Cizek Model #2 28 DCM 'Time Window' 29 Fundamental Research 30 Infinity QLS 32 Koss Model One/ A 34 Ohm F 35 Pyramid 'Metronome' 37 Rogers LS3/5A 38 Snell Acoustics Type A 40 A Comparative Survey of Power Amplifiers: Part II By the Staff of The Audio Critic 41 Bryston 4B 43 CM 914 44 D B Systems DB-6 44 Electrocompaniet (Part II) 45 Futterman H-3aa (Part II) 46 Quatre DG-250 (follow-up) 46 RAM 512 47 Threshold 800A 49 Cartridge, Arm and Turntable vs. the Groove: Who's Winning? 57 Dual CS721 57 Dynavector DV -505 58 Grace G-707 58 Grace G-940 59 Harman Kardon Rabco ST-7 59 Infinity 'Black Widow' 59 Kenwood KD-500 60 Luxman PD-121 60 Mayware Formula 4 61 SAEC WE-308 NEW 14 The Realities of Noncommercial Audio Journalism By Peter Aczel, Editor and Publisher 62 Records & Recording

65 The Admonitor Comments on Current Ads 3 Box 392: Letters to the Editor

67 Classified Advertising Publisher's Note

n our Number 2 issue (March/April 1977), we made the following ill-considered I editorial statement: "... Then we'll get out three more issues in the second half of the year. If it's October and you haven't received Number 4 (July/August) yet, then you'll have cause to worry-we won't be able to catch up. But it isn't going to happen." Well, October has come and gone, and you haven't received ... but wait a minute, you're holding Number 4 in your hand, so it couldn't be as bad as all that. Of course, it's perfectly true that we won't be able to catch up-not in 1977. When we wrote those words in April, we sincerely believed that we could publish three times as many issues as any other audio review that carries no advertising (the short­ est interval we had ever observed between two successive issues of these having been six months). Well, in the sober gray light of October, it appears that we're able to publish only twice as many issues as anyone else in our business, our aver­ age interval thus far being three months.That doesn't mean, however, that we have no potential for improvement. Here's our plan: To reflect reality, we're dating this issue to cover three months: July /Aug­ ust/September 1977. (We'll probably do the same thing in 1978, since the summer seems to be inevitably our period of lowest output.) Our Number 5 issue will be dated October/November 1977, and we're reasonably confident that we can mail it out in December. That leaves Number 6, which will be dated Year-End 1977 and (you guessed it) won't be mailed out before we're well into 1978. There's just no other way.Number 6 will be worth waiting for, however,· we plan to make it � very special reference issue that audio enthusiasts will be brandishing around throughout the coming year, making life difficult for their friends and local dealers. After that, we'll see. Maybe we'lll have our act together to the point where we can have our twelfth issue out before the end of 1978. But this time we won't promise ... So the question, from our subscribers' point of view, is this: What's the dif­ ference between spending $28 for six issues over a period of twelve months and spending $28 for the same six issues over a period of, say, fourteen or fifteen months. If you feel that the first is a good deal and the second is a bad deal, then you have a legitimate quarrel with us. We think it's exactly the same deal, since we're selling you six issues either way, not a time-limited contract. We were, how­ ever, unrealistic when we committed ourselves to a time limit, and for that we owe you an apology. One more thing. So far, we've religiously answered every where's-my-last­ issue-I-didn't-get-it-yet letter.As of the day this issue is being mailed, we're ending that time-wasting and costly practice.Since 99.8% of our mailings reach their destin­ ation (that's an exact figure), please consider no reply to that kind of inquiry to mean: "We haven't mailed it yet." Thank you.

2 Box 392 Letters to the Editor

In addition to editorial correspondence of general interest, we're including two special categories of letters this time. One has to do with clearing the air after a particularly fo ul-smelling anonymous communication circulated about The Audio Critic; the other covers exhaustively the "Mark Davis syndrome" originally brough t up in our first issue as a kind of aside in Part I of the preamp survey. Th e Mark Davis correspondence is reprinted here in 8-point type fo r those who are sufficiently motivated to plow through it; we didn 't fe el we'd be justified giving it more space. Th e letters we publish in this column may or may not be excerpted, at the discretion of the Editor. Ellipsis ( .. . J indicates omission. Address all editorial correspo ndence to Th e Editor, Th e Audio Critic, Box 392, Bronxville, New York 10708.

In the last week of June and throughout most of earthed. July, the anonymous missive reprinted below was being But even if these suspected alliances prove to be sent out to literally hundreds of high-end audio dealers, untraceable, The Audio Critic has misrepresented itself manufacturers, reviewers, editors and other high-end to the audiophile-consumer. Where is the line between audio practitioners from coast to coast. The anonymous misrepresentation and fraud? commentator had invested 24 cents first-class postage per Just as important-can anything written by The mailing, as each fat envelope also contained Xeroxed Audio Critic be believed? copies of our allegedly misleading ads and of trade-press We think not. clippings documenting that two of our part-time staff The Audio Critic is, to us, severely compromised. consultants made their daily living in the audio business, A publication with no credibility serves no other pur­ one in retailing, one in manufacturing. All of these pose than to make someone a lot of money. envelopes were without a return address and postmarked Peter Aczel, the Editor and Publisher, has a lot of either in Hartford, Connecticut, or Manchester, Con­ answering to do. necticut (a suburb of Hartford). In the wake of the anonymous mailings, the follow­ ing exchange of letters took place: The Audio Critic-A Ripoff? Dear Mr. Aczel: Another "little magazine" has appeared. No doubt you have seen the anonomous (sic) let­ This one bears watching. ter that is being circulated, accusing your publication of Initially, our hopes were high. The price ($ 28.00 having "substantial ties with the audio industry," and for a year's subscription) was high but then so was the supplying what appears to be some documentation of promise. By combining objective measuring techniques that allegation. with long term listening sessions, The Audio Critic prom­ We were of course sent the sheaf of papers in ques­ ised information, not just opinion. tion, and while we do not intend to print the letter-it It now appears that the price was too high as The goes rather overboard, in my opinion-we cannot let the Audio Critic is, in our opinion, incapable of fulfilling matter pass without at least some comment to the effect the promise. that the letter was circulated and commercial affiliations Contrary to claim, The Audio Critic has a number of two of your key staffers documented from other, inde­ of substantial ties with the audio industry, both on the pendent sources. retail and manufacturers' level. These ties make objec­ May I please have a statement of your response to tivity impossible. In fact, these ties just might make this, for publication? The Audio Critic a "house paper." Sincerely, The attached information begins to point out the J. Gordon Holt connections. Editor & Publisher We feel there may be more tidbits yet to be un- The Stereophile

3 Dear Mr. Holt: We have come under criticism by about 0.1% of Two things surprise me about your recent letter. our readership (three persons, to be exact) for the pro­ One is that you assume you're entitled to a reply, fessional involvement of some of our staff consultants even though you're neither a subscriber of this publica­ in the audio industry. Doesn't that constitute a conflict tion nor a manufacturer whose equipment has been re­ of interest, we were gleefully challenged, in view of our viewed in it. 1'/1 let that pass. simon-pure posture of independence? The other is that you, as a reputable person, should Ah, that's a good one. The basic reason for the be impelled to initiate this dialogue by something as amateurishness of the "underground" audiophile reviews sleazy as an anonymous letter, which should have been is that they are staffed by amateurs. It would be very disposed of in the same manner as bird droppings. Since nice ifone could come to valid conclusions about, say, the same information is available to you from other the transient response of an amplifier by consulting sources, as you say, I wish you had referred only to the music-loving dentists, accountants and shoemakers. latter. It would have been cleaner that way. Unfortunately, such independent experts seldom know It seems, however, that in addition to his soiled what they're talking about. That's why we have a pro­ character, the anonymous letter writer also has a reading fessional record producer, a physicist/audio engineer, problem. As the enclosed excerpt from the second issue an audio-electronics technician and other qualified pro­ of The A udio Critic shows, I had already discussed quite feSSionals on our staff Sure, some of these people derive freely in print what he started to whisper behind my back part of or all of their income from the audio business, months later. (Moral turpitude and lack of mental alert­ but not one of them is a chief executive officeror majority ness frequently go together.) stockholder, so that the worst that can happen is that the To round out this editorial on the subject (which I views they privately communicate to us deviate from the must ask you to reproduce in its entirety along with this self-interest of their bosses. Tough. It just so happens letter, otherwise I can't give you my permission to print that The Audio Critic has already dealt rather severely either one), I would also like to make the following points: with products made or sold by said bosses. My Wife and I are sole stockholders of The Audio The important thing is that the management of The Critic, Inc., which banks the subscription fees. Neither Audio Critic is completely divorced from commercial one of us has any affiliation with audio manufacturers audio. The Editor/Publisher deliberately severed all con­ or retailers, nor does anyone else on our payroll. The nections with the industry before coming out with the first ' two consultants (not key staffers, as you say) trium­ issue. Our subscribers are our sole business interest. phantly tracked down by the anonymous creep are both As a matter of fact, if you hear any malicious old friends of mine who had lent their names to my gossip about The Audio Critic's conflicts of interest, or infant enterprise without any remuneration. Now that especially about our taking bribes for favorable reviews The Audio Critic is a going concern, I am entering, (one of the ever-recurring fabrications about nearly all and have entered, into many more consulting relation­ audio reviewers, perpetuated by a few pathetic little ships with technologists of all sorts employed by the would-be authorities), we suggest you let us know, pro­ audio industry-so many, in fact, that it's no longer vided you're willing to identify the source. There's practical to list them on the masthead, especially since nothing a Hungarian loves better than a good lawsuit. some of them don't wish to be so listed. I spend hours on the telephone or in person with these engineers and designers; I pick their brains; they Dear Mr. Aczel: pick mine; they influence my technical thinking; I in­ We've had an anonymous packet regarding your fluence theirs; they come to my laboratory; I go to theirs; publication's claim to impartiality. I imagine most pub­ we discuss test methods; we have drinks or even dinner lications in our field have. Would you care to comment together; they show me their latest research and papers on the present relationships of your staff to commercial -and they haven't got a nickel's worth of influence on interests? my income or my editorial policy. The list of their names Sincerely, would excite the anonymous letter writer into an orgasm Edward T. Dell, Jr. of I-told-you-so. Editor /Publisher I, on the other hand, feel that such alliances are The Audio Amateur better for my education and therefore better for my sub­ scribers than if I consorted exclusively with untutored Dear Mr. Dell: audio freaks-which is obviously what the anonymous In reply to your recent letter, I am enclosing a letter writer thinks an honest noncommercial audio copy of a similar letter from J. Gordon Holt and of my journalist should do. response to it. Does this answer your question? Since my answer to your letter, had it preceded Mr. Sincerely, Holt's, would have been virtually identical. there is really Peter Aczel nothing further to say. except that this makes you the Editor and Publisher second reputable audio journalist without a clearly ex­ pressed distaste for anonymous letters. Sincerely, Peter Aczel (The editorial excerpt from our second issue, as Editor and Publisher enclosed with the above reply, is reprinted here.) The A udio Critic

4 Dear Mr. Aczel: We agree with you wholeheartedly: a mixed bag of We have received an anonymous note regarding superiorities and inferiorities doesn't constitute SOTA, your magazine, entitled "The Audio Critic-A Ripoff?" even if on balance the equipment is judged to be prefer­ It's most probable that you have already seen a copy able to all others tested. We have tried to qualify our of this "material." conclusions and recommendations to make that clear; It appears that considerable time, trouble and ex­ unfortunately the word "best" tends to stick out and pense went into the preparation and distribution of this obscure qualifiers such as "so far" or "depending." note, but we are unable to determine the intent of the We'll try harder in the future. sender; altruistic self-righteousness seems out of the -Ed. question due to the effort involved. Some of us take issue with some of your published The Audio Critic: opinions, Mr. Aczel, but we all agree that your conclu­ "Accuracy" is an impeccable word, and your bit on sions are fairly presented and without discernable bias. page 7 of the Marchi April issue, "In other words, We question the composition of your reference system, accuracy isn't a matter of taste, " is very neat and con­ having learned that there are amplifiers far more defini­ vincing. Up to a point. tive than the Quatre, but we attribute that to subjective As you say, "either they did or they didn't" match impressions rather than influence of industry. the color. Agreed also that it's difficult in audio, but You have our absolute support. We will continue your point remains valid, and taste has nothing to do recommending your fine publication to audiophiles with it. You look at your car and the repainted door and within our area of influence. We are convinced that the say, "They don't match." I don't doubt that you see a more any client knows about audio, the better he'll like mismatch. our way of doing business. I come along. I look from car to door, and I say, Sincerely, "They did a fine job, didn't they !" I see a perfect match. H. L. Eisenson "Accurate, " I say. "Not accurate, " you say. And we are Audio Dimensions both right. San Diego, CA We both see what we say we see. But we happen to have different eyeballs. I have a friend who sees almost * * * no red, but sees beyond violet into ultraviolet, to make up for it; most variations are not so spectacular. Which is to say that your color-matching analogy is valid, your And now for saner and more savory matters. "accuracy" motto is valid-until another person comes along. Eyeballs are not standardized. Ears are not standardized either. The Audio Critic: So that impeccable word "accuracy" is not the ...I applaud your attempts to correlate lab data solid support you may think it is. I hope you use it with audible effects; this field seems to be limited to with care. yourselves and Richard Heyser in terms of published Sincerely, material, especially since Bascom King no longer reviews Burnett Cross equipment publicly. Hartsdale, NY My major complaint with your magazine concerns a fault common to both you and your competitors, and You have a point there but, in our opinion, you're is a simple one. Simply, there is no state-of-the-art in bearing down a bit too hard on it. There exists a spread audio today (with the possible exception of tuners); to in physical perception from human to human, but be­ attempt to recommend one piece of equipment as the best tween reasonably young and healthy specimens the of its art is foolhardy. To truly be the "state-of-the-art" spread is quite narrow, so that a norm for our species a component must outperform its competition in every is a viable concept. If color matching, for example, were parameter considered. That is, when two pieces of equip­ as subjective, indeed anarchic, as you suggest, wallpaper ment are compared, there should be no doubt in the companies couldn't stay in business-nor could Eastman minds of the listeners as to which is superior. If the Kodak. In audio. we'll settle for a match to the original sound of both components is felt to have different posi­ as good as the average car owner would let the body tive qualities, then clearly neither represents the SOTA, shop get away with. since existing technology could better either component. -Ed. , In fact, you allude to this characteristic several times, yet in the final analysis appear to ignore it. You could put yourselves yet another step above your competitors if you did not attempt to follow in their footsteps by The Audio Critic: finding a "best"; there is certainly more than one legiti­ ... Would you please discuss the conflict (to my mate approach to the SOT A, and putting one above the mind) between the ideal "straight wire with gain " con­ rest smacks of the commercialism you are trying to cept, represented by the state of the art preamps and avoid, as well as being unrealistic .. amps with a minimum of controls and frills, and the con­ Sincerely, flicting use of expanders, companders, equalizers, and Robert Bertrando noise reduction devices a la Phase Linear 1000 or Tucson, AZ. Burwen, etc.

5 I n other words, is it insane to own a Rappaport tively beautiful, but subjectively emotionally empty. preamp, Quatre amp, DQ-I O's, etc. and a ten-band/ Most people agree the program-listener schism is anxiety channel equalizer-dbx expander-SAE "click and pop provoking. Has anyone built a non-fatiguing SP-4 with machine"-Phase Linear lOOO-or any one of a dozen sock? accessories? lames A. Debros Do all the many add-on devices only serve to St. Cloud, MN destroy the "art" in "state of the art" represented by an exceptional amp or preamp? No, no, no! Wh at you need is a nonobjectively Thank you, transcendental preamp that is in perfect harmony with David Gibbs the vibrations of the universe, not to mention our Rancho Palos Verdes, CA precious bodily fluids . -Ed. The signal-processing accessories you mention are a necessary evil-and not quite as ofien necessary as a lot of audio people think. Such de vices are de signed to correct what somebody else has already messed up, either The Audio Critic: as a result of inadequate technology or through sheer I have just read the first two issues of The Audio neglect. With today's best program sources, played in a Critic and found them to be very interesting.... As an room that isn't an absolute pig acoustically, you're much electrical engineer I am happy to see perceptive criticism better off with the shortest, simplest possible signal path coming from persons who seem at the same time to be­ from input to output. In fa ct, with all those "black lieve that there are laws of physics and electricity which boxes" in the signal path, you're quite right to suspect must hold no matter what our wishes to the contrary. I that the difference between, say, a Rappaport preamp look forward to future issues and your continued success. and a Marantz would be obscured. We never touch the There are two comments that I would like to stuff ourselves, unless it's on loan fo r evaluation. make regarding the following quotes from The Audio -Ed. Critic, Vol. I, No. 2. On page 41 you state: "The net result is linear-without feedback and hence without TIM." There is at the present time some confusion about the precise term that should be used for the distortion now commonly called "Transient The Audio Critic: Intermodulation Distortion." Some persons call it . . . There is one aspect of the preamp "con­ "TIM," others call it Slew Induced Distortion or troversy" which confuses me and, perhaps, other non­ Transient Overload Distortion or Frequency Overload technical types. It is the question of stereo depth. If dis­ or any of several other terms. Whatever term is used, it tortion gives the illusion of depth and true accuracy is clear that this form of distortion can occur within demands a sacrifice of the illusion, I would opt for a feedback loop or outside of any feedback loop. accuracy. Isn't this question answered by using a record Thus, it is not really correct to say that an ampli­ or tape where the reviewer would be aware of the precise fier, whether it consists of one or several stages, will location of the instruments during the live performance? certainly be free of transient distortion just because it A string quartet, even if the performers were local ama­ has no feedback. A single stage amplifier with no feed­ teurs, might be ideal ... back will fail the Otala test if its slewing capacity is Yours truly, exceeded. This is a small point but if we are ever to Edward l. Doyle understand precisely what is going on in amplifiers we East Northport, NY must think about them very carefully. What is true about feedback amplifiers is that, if Th ere's a subtle but audible difference be tween the there is any tendency for internal slew or amplitude limit­ haphazard artificial de pth resulting fr om electronic ing to take place, the problem is greatly exacerbated by anomalies and the fo cused natural depth re ndered by the feedback. However the problem is not caused by the accurate reproduction. Th ere's no sacrifice involved in feedback but is caused by the limiting tendency in the the latter: it sounds just righ t. We agree with you that first place. complete fa miliarity with the dep th and width of the I t is not correct to say "no feedback-thus no sound source at the actual re cording site is a great help TI M. " It is correct to say "little feedback-thus less in judging this phenomenon. Of course , the microphoning problem with internal TIM." must be simple and straightfo rward; gimmicky multi­ My second comment relates to the following paren­ miking can throw you off completely. thetical expression: "(They'll never teach it this way at -Ed. EE school, but then the tuition there is more than $28 a year.)" I realize that this is just a nasty little comment set down to thrill your readers. (Critics are supposed to be mean to the core, I hear.) My issue with the comment is The Audio Critic: that it is not fair or accurate. In the attached reference Having purchased two Rappaport preamps on the list from an article by lung, most of the references are basis of Vol. I, No. I, I must say your writers should from universities (Daugherty, Greiner, Leach) or certain­ have elaborated on the PRE-I's highs. They are objec- ly from highly trained and qualified electrical engineers.

6 We in the university are trying very hard to under­ about EE school and the $28. We were actually ap ologiz­ stand the very same problems that plague us all. I have ing, in our own left-handed way, about the brevity and for 15 years taught about overload problems in feedback patness of our explanation of the analog multiplier ("gain amplifiers to my students. I have also written about these cell") principle. You academicians must really fe el be­ problems in the technical literature . The public and leaguered ifyou take a mildly embarrassed journalist's amplifier designers have not been listening very well. fo ot shuffl ing to be a kick in the pants! Certainly the university is not the source of all -Ed. knowledge, but we have tried. Much credit for improved amplifiers must go to the individual inventor of course. Unfortunately they have often been more reluctant to share their accomplishments and insights with the public. What I am really suggesting to you is that we all need each other in order to both solve and understand The Audio Critic: the complexities of amplifiers of any kind, whether they Re your CES report on the Yerion NF -I Filter: are power amplifiers or pre-preamplifiers. In order to The pros and cons of wide bandwidth, DC to join together in these efforts we must make an attempt MHz, versus narrow bandwidth, 20 Hz to 20 kHz, have to communicate better. This will happen only if we try been with us for more years than I care to remember. for harmony instead of acrimony. The controversy simmers and boils. Now it has come Please consider these comments to be harmonious. to a boil again. Let me close this time with a bit of acrimony of my J. Peter Moncrieff in his belatedly issued Inter­ own directed to the inventor-producer of some of the national Audio Review, July 1976, page 38, states that he flood of equipment that exists. considers a low-frequency cutoff of 0.5 Hz to be barely It has been my observation that many writers, low enough for achieving accurate bass reproduction of those that write for pulp magazines or the most esoteric music. On page 180 he considers an audio component journals, all have a primary purpose to spread their with less than 0.5 MHz bandwidth to be audibly ob­ insight and understanding of problems to others. They jectionable. are in fact teachers in the most generic terms. The Mitchell Cotter believes that any bandwidth in a editors of The Audio Critic are clearly doing the same. power amplifier (why not a preamp also?) in excess of 18 This is the way to understanding since the process gets Hz to 35 kHz is detrimental to the reproduction of the best minds focused on the problems. music and is asking you, as you say, to plunk down $300 On the other hand, the individual inventor-pro­ to prove he is right. ducer all too often has just the opposite goal in mind. This is a very serious issue to the dedicated audio­ He hides his techniques, he hides his theory, or worse, phile interested in State of the Art equipment and should spouts garbage and nonsense; he misleads, he buries his be explored thoroughly by The Audio Critic. circuits in epoxy, he gives erroneous specifications, he I am on the narrow bandwidth side even without never writes about his circuits or theory of operation, he the time compensation of the NF-I filter. It has yet to be never supplies operational manuals that are correct, he proven that phase distortion of the signal between the never-never supplies repair manuals, and so forth . input and output of an amplifier is audible. This is not true of many manufacturers, of course, As you know, the reason for negative feedback but all too often the most mystical manufacturers are the is to reduce distortion. The wide bandwidth that is a ones that make the most outrageous claims. It is very concomitant of this feedback may be an asset or a hard to tell the fakes from the frauds. detriment, depending which side of the fence you are on. I would urge you to ferret out the fakes. Make There is a new school of thought (of which you are a them publish their theory, if any. Make them tell the member) that claims the less feedback the better the truth about their specifications. Make them show the sound. I am inclined to think that the reason is the re­ magic, if any, of their circuits. Have them explain how duction in bandwidth of the less-or-no-feedback ampli­ they defy the laws of physics. fier. With the NF-I filter you can eat your cake and I have hopes that you will accomplish more of have it, too. this than the mystics among the golden-ear crowd have Think of the poor amplifier designer. What will a to date. 20 Hz or 20 kHz square wave look like at the output of Yery best regards, an 18 Hz to 35 kHz amplifier. Horrible, that's what. R.A. Greiner That it will sound better will not be believed by the re­ Professor of Electrical Engineering viewers who look at those square waves on their scopes. University of Wisconsin Julius Futterman Futterman Electronics Lab New York, NY

We stand corrected on our TIMIf eedback simplism No comment necessary-except to re mind all you and fe rvently endorse the admonishment of undisciplined young whippersnappers to lis ten respectfu lly when Julius inventor-producers and muddleheaded audio mystics. Futterman talks. He didn 't learn what he knows fr om a Wh at dis tresses us, howe ver, is that Professor Greiner, Te ch HiFi salesman three weeks ago; he was designing who is one of our heroes (he was coauthor of a superb state-of-the-art tube amplifiers when you guys were 1966 paper that anticipated the entire Otala-Curl-Leach­ watching Howdy Doody. Jung scene), should so direly misinterpret our little quip -Ed.

7 The Audio Critic: The Audio Critic: When bi-(tri-, quad-, etc.) amping a system using It will be of some value to the other readers of your electronic crossovers, the only speaker which needs a fine periodical if a couple of technical errors in the first really high-slew-rate amplifier is the highest-frequency two issues are corrected: unit. In most cases, for high sound pressure level, low­ (I') In "Have Tone Arm Designers Forgotten Their distortion treble, a low-power high-quality amplifier High-School Geometry?" you criticized the Grace G-707 can be used, such as Mark Levinson's new class A job. for its "incorrect" L sin {3 of 3.375". (Did you really The rest of the power amps can be of lower slew rate, measure it accurate to .001"?) I have measured {3 on my because the electronic filter networks will prevent G-707 and have found it to be 23.25° (plus or minus any fast-rising waveforms from reaching their inputs. 0. 25°), which, in combination with the 9.33" L specified This is particularly true for subwoofer and woofer amps. by Grace (L is adjustable by about 0.4", a fact you ne­ High-slew-rate power amps with expensive, wide-band­ glected to mention, which means therefore that L sin (3 width output stages are simply a waste of money for the is adjustable) computes to, surprise, 3.68"; the exact low-frequency end of a system. number you described as being ideal. I think you owe Since TIM and related distortions cease to be a Grace an apology. problem for amplifiers driven by low-pass filters, large (2) In "Fishing for Bass: A look at the Subwoofer amounts of negative feedback and compensation can be Scene," you made the statement that a sealed system used to get extremely low output impedance. This will with Q = .7 07 has, among other things, "transient re­ create the high damping factors necessary for good, sponse ... best possible for these conditions." tight bass and midrange. Op amps such as the 74 1 and Although it is not clear what is meant by "these many of the low-frequency high-power output devices conditions," I can point out that, while a Q = .707 sealed should be naturals for such amplifiers, bringing prices box loudspeaker does have a maximally-flat (i.e. Butter­ down drastically. worth) frequency-amplitude response, it does not have a This should cut down the $20,000 price tag con­ maximally-flat (i.e. Bessel) frequency-delay response, siderably on your Super Dooper Mark Levinson System. and hence does not have the best transient response Alan Hoover obtainable for a closed-box system. This occurs with a Indianapolis, IN Q = .58. The difference between the two transient re­ sponses is that the former tends to ring just a little at a Cf the letter from Chris Russell of the Bryston frequency equal to the cutoff (and resonant) frequency. company, published alongside our power-amplifier sur­ In a subwoofer this is, of course, 20 to 30 Hz. Whether vey in this issue. or not this is "audible" is an interesting but (as far as I -Ed. know) unresolved question. I would like, also, to contradict your somewhat pompous remark that there are probably no more than 15 engineers with a thorough grasp of the mathematics The Audio Critic: of speaker system analysis (naturally your boy is right Since I had not received the first two issues and in there around #7 or #9), since this writer has personal have had to see them from other sources I am quite an­ acquaintance with at least five. I suggest you get on with noyed. It is quite obvious that you have a grudge against the business of reviewing audio equipment, and stop me. However, your lack of ethics are in question as worrying about how incompetent everyone else is. far as I'm concerned. (It are?-Ed.) You accuse me yet Leave that burning topic to The Absolute Sound. allow me no defense. You compound this by refusing Having dispensed with my technical complaints, to honor my subscription which I have paid for. The en­ I have a more musically oriented complaint: closed photocopy is self-explanatory. I wish you would find someone more empirically James Bongiorno well-qualified to write about records and recording than West Hollywood, CA Max Wilcox, who has produced the most consistently awful sounding classical discs ever produced in this country, and therefore the world. (I refer, of course to his PhiladelphiajOrmandy recordings for RCA.) Why not Everybody is ganging up on poor Jim Bongiorno, Kenneth Wilkinson? including the United States Postal Service. His sub­ Sincerely yours, scription had been sent to the GAS company and went Alan S. Watkins astray; we sent him replacement copies to his home. Burroughs Corporation The notion that we punitively withhold paid subscrip­ Pasadena, CA tions from people we "have a grudge against" is down­ right pathological; of course, as Kissinger used to say (Six days later): to Nixon, "just because you're paranoid doesn't mean In my (previous) letter I pointed out your measure­ they're not out to get you." As for allowing no defense, ment error regarding L sin {3 of the Grace-707 tone arm, any manufacturer with anything factual to say in de­ in "Have Tone Arm Designers Forgotten Their High fense of his equipment or his company has this column School Geometry?" open to him, and we never said anything to the con­ Since that time, I have measured (directly) L and {3 trary. on the SME 3009 S2 Improved tone arm, and have found -Ed. L (with Shure V- 15 Type II) equal 9.14" and {3 23.0° (plus

8 or minus .25°). This corresponds to L sin (3 equal 3.57". the fa ct that the op timum offsetangle fo r an arm length which is within 3% of your ideal figure of 3.68": Allow­ of 9. 14" is 23 .75° and that the optim um arm length fo r ing for I) a I % measurement error, 2) the fact that L is an offset angle of 23.0° is 9.42". Since even fu lly op ti­ in small part determined by the cartridge used, and 3) mized geometry in a pivoted tone arm yields afairly high that few of us are equipped with a laser to align our distortion index at the three maxima, we don 't consider cartridges with our headshells, I think this geometry that kind of deviation to be negligible. can be called optimum. This, in combination with the 6. Maybe you aren 't all asleep out there, but at exact geometry of the G-707 referred to in my previous least one of you is showing the effe cts of insomnia. letter, would seem to show that your statement "We -Ed. don't know of a single commercial design in which (3 is optimum for the given L or vice versa," is based on your own faulty measurements. The Audio Critic: Furthermore, I noticed that the latest issue of ... One of your subscribers sent to me for com­ A udio contains a large ad placed by The Audio Critic, ment part of an article by Bruce Zayde, titled "A Ra­ which makes use of the same misguided statement I have tional Approach to Low-Frequency Speaker Design", quoted above . This would suggest that the next candidate that part of the article in which he describes transmis­ for roasting in "The Admonitor" is no one but The sion lines as "nonoptimal designs .. ." Audio Critic itself! Since we manufacture damped reflex, infinite baf­ This kind of irresponsible sloppiness on your part is fle, and transmission line designs, and find that the causing your credibility to deteriorate rapidly in my latter are quite superior to the former on spark gap and eyes, and I am sure in the eyes of other technically pulse criteria, we are rather at odds with Mr. Zayde, aware readers. We are not all asleep out here! when he says that "a better approach to producing low Sincerely yours, bass response ... is to stick with the sealed or vented Alan S. Watkins direct-radiator format"; and also because we don't know Burroughs Corporation of any such designs which in fact do provide "low Pasadena, CA bass," except transmission lines. Perhaps Mr. Zayde would care to comment directly It isn't our policy to publish and refu te letters pre­ to us, on what is "better" about his recommended senting technical arguments that are just plain wrong, "format." I understand that your magazine features "without redeeming social value, " but this one is so ag­ measurements, rather than purely subjective reactions. gressive and righ teous in tone that we've decided to make I am naturally interested in Mr. Zayde's measurements an example of it. So: that corroborate his statements, not as a challenge to 1. As long as you don't twist the cartridge in the you, but based on my own curiosity; since most of the heads hell (which is what we now recommend elsewhere people around the world with whom I work would take in this issue but is n't what you're talking about), the my position, and base it on measurements. L sin (3 of any arm is constant and nonadjustable, even Sincerely, if L by itself is adjustable (i.e., if you can move the Irving M. Fried cartridge fo rwards or backwards , as in the Grace G-707). Fried Products Company Why? Because when you change L, you're sim ultaneous­ Philadelphia, PA ly changing (3. But you aren 't changing-and can't change without twis ting-the L sin (3. (Eureka! You Mathematically , a transmission-line enclosure (or ought to apologize to your high-school geometry teach­ acoustical la byrinth) can be represented by a high-pass er. ) We've remeasured the G-707, by the way, and its filter of the same order as certain types of vented (bass L sin (3 is nowhere near 3.68" without twis ting. reflex) enclosures. Exactly the same expressions go into 2. For a fe llow who's a little careless with his the mathematical model in each case, except that the trig, you talk pretty good filter synthesis, and what you transmission line has a large resistive component, cor­ say about Butterworth vs. Bessel resp onse is true-as responding to that long, hea vily stuffe d la byrinth. The fa r as you go. What you fa il to mention is that a Bessel upshot of this is that, by properjuggling of the intera cting response profile has pretty grim amplitude character­ elements of either filter. you can come out in the same is tics, so that a Butterworth contour still represents place. i.e .. achieve exactlv the same fr eq uency and time the best trade-off between amplitude and time response response. A nd that means exactly the same sound quality. -which is what we meant all along. Mother Nature plays no fa vorites between networks 3. Leave it to a mathematical type to derive "no of the same type and order. (We didn 't have to consult more than 15" fr om our "can probably be counted on Bruce Zayde to assert this simple truth.) two hands and maybe a fo ot." Wh at's the matter, Alan? The reason why the transmission-line approach Don't they crack a fe w jokes around the wa ter cooler is "nonoptimal" is that the resis tive component dis­ at Burroughs? sipates a lot of amplifierpo wer; in other words , all other 4. Na me a producer who has done better than Max things being equal, the resulting speaker will be con­ Wilcox in that seventh-floor ballroom of the Scottish siderably less efficient, even if not lower in fidelity. Rites Cathedral in Philadelphia. You think every re­ Th e whole question was quite neatly disposed of by W.J. cording site is the Concertgebouw? J. Hoge in his article on transmission-line speakers in 5. How fa r off your SME arm is , if we accept the August 1977 iss ue of Audio, and there's an even more your measurements as accurate, will be indicated by profound analysis in a 1975 thesis by G.S. Letts, a grad-

9 uate student of Dr. Small at the Univers ity of Sidney. out ahead in typical "as is " testing is that, by its very When you get right down to it, the transmission-line nature, it requires heavy bracing to be at all usable. bass enclosure has no serious academic backing; its sup­ Makers of ordinary vented (as well as sealed) enclosures port appears to be strictly by commercial and hobbyist can get away with less conscientious bracing, but in interests. that case the performance isn't op timal. Our value When it comes to measurements, the question is ­ judgment was based on what is possible in the light what exactly has been measured? To wlly op timized ex­ of available knowledge, not on what the industry is ecutions of both approaches? We suspect that one rea­ actually doing in response to audiophile mystique. son why the transmission-line enclosure oft en comes -Ed.

* * *

The rest of this column is devoted 10 other, set levels to be vaguely the same, frequency response is 90% of the story. The the Mark Davis correspondence (in small and found differences that were day and other 10% is cartridge tracking and loud­ type, as already explained). Enclosed with night. And which seemed to be repeatable. speaker radiation pattern.) the leller writen 10 The Audio Critic was After that initial A-B, Jay and I spent (Like dynamic range is zero percent?-Ed.) a copy of a previous (but until now un­ months trying to obtain the effects others Along the way, it did occur to us that published) leller to audio journalist Randy seemed to be able to attain so effortlessly. long-term listening might reveal differences To mlinson, which should be read firs t fo r We tried substituting every link in the chain: that A-B tests did not. Quite honestly, [ fu ll understan,qing of the Mark Davis syn­ cartridges, turntables, power amps, A-B think that there are some careful experiments drome, and whose text fo llows. boxes, speakers, headphones, yet the results that could be done in this area, but we did kept coming out the same: whenever there not pursue them because there is in fact al­ Dear Mr. Tomlinson: was the most barely discernible difference, ready a small mountain of evidence to support Some work [ had done with an under­ there was always a corresponding diffe rence their validity. [n the parlance of psycho­ graduate named Jay Gurley on the audible in frequency response. Whenever there was acoustics, an A-B test is normally referred differences in phono preamps was dis­ no measurable diffe rence in frequency re­ to as a two-interval two-alternative forced­ cussed on the WBUR program "Shoptalk," sponse, subjects could not diffe rentiate aud­ choice experiment (2[2AFC). A 2AFC is and apparently it was a tape of this show ible quality. Nor did [ limit the tests to simply an experiment where either stimulus that you heard and commented on, indica­ naive subjects: some of Boston's goldenest A or stimulus B is presented on a given ting why you did not believe our results. ears were enticed down to the lab. (And, trial, and the subject indicates which he The fact that people such as yourself incidentally, I don't believe in the idea thought it was. [n a 2[2AFC, both are pre­ should disbelieve our conclusions comes of a person with golden ears: all persons with sented, and the subject indicates whether as no surprise. Indeed, [ did not set out normal hearing have perception thresholds the order or presentation was A-B or B-A. to prove that :Jverall frequency response which fall within certain known limits. A There are other types, such as recognition was the only important parameter con­ properly designed psychophysical experiment experiments, in which one of a number of tributing to preamplifier sound quality. I should yield identical results whether the stimuli is presented and the subject indicates fi rmly believed, as you apparently still do, subject has a $3000 stereo and goes to con­ which. These experiments have been used that two preamps could have identical fre­ certs or not. However, this gets into areas to study a wide variety of acoustical quency responses, as measured from a test of things like bias, training and feedback, phenomena, such as loudness perception, record with a particular cartridge, plus which [ don't want to try to explore here.) localization, speech intelligibility, etc., and levels of noise and distortion below audIbil­ Having failed to find anything relevant in all areas that [ am aware of, the 2[2AFC ity, yet still sound different. The explana­ other than frequency response on our own, experiment results in the greatest reported tion most people seemed to give for this we approached some of the most vocal subject acuity. had something to do with the high-frequency proponents of the other side, people like Our experience using the 2[2AFC performance of the preamp, where "high" AI Foster of the BAS and Tom Holman of paradigm in the preamp project strongly means outside the audio band: 30 kHz, Advent. The listening tests they had done supported the validity of Its use, in that 50 kHz, or more. Unfortunately, none of turned out to be informal to an extreme­ whenever there was a measureable difference the explanations got much more specific which is not to knock them; they are not that fe ll above accepted thresholds of than that and, although test results were and do not claim to be psychoacousticians­ audibility, our subjects were able to pick quoted showing measured pream p differences but when we tried to isolate differences it out on an A-B test. Among the differences in these frequency ranges, the fact remains between preamps they claimed sounded dif­ that were audible: level differences of a very that the ear cannot hear above 20 kHz, and fe rent, only frequency response again seemed few tenths of a dB; frequency response dif­ that .no record cutter in existence can cut to be the operative parameter. Neither we nor fe rences over as little as an octave or so frequ-e'ncies higher than 18 kHz or so anyway. they could hear differences when frequency amounting again to a few tenths of a dB. At (What about cUlling at half speed, as some responses were equalized. one point we attempted to use a home-brewed of the more sophisticated mastering studios (Of course, this put Holman and Advent power amp comprised of a pair of Sank en 50- do? A ccep ting your figure, that would give in a sticky position, since the success watt amplifier modules. Listening to a sine a high-frequency limit of 36 kHz. -Ed.) of the 300 receiver was and is predicated wave through them compared to a Crown 150 Thus, what was missing was a clear-cut ex­ in large part on the preamp section. Since power amp revealed the presence of numerous planation of how the performance of a pre­ our tests, Advent has been extremely careful high harmonics. Observation of the output amp at 50 kHz, and high levels of 50 kHz with the wording of their ads; they don't waveform showed crossover distortion from at that, could affect something in the audio claim the preamp is the best, only as good the Sankens when they were connected to a band. What [ set out to do was isolate that as the best. The fact that it's also as good loudspeaker, which was not present with the cause/effe ct relationship, not prove that it as the $160 Pioneer SX-450 (discount price) Crown. Spectrum analysis of the output con­ didn't exist. and a host of others is something they fail firmed the presence of the harmonics in the ' Having built the zippiest preamp to point out. They have, however, deleted all output of the Sankens, the absence of same in A-B box [ could imagine, [ connected reference to the square wave test Tom talked the Crown. Along the way, someone brought my painstakingly-designed Davis-Brinton about in the original AES preprint paper. us an old H K tube preamp which apparently Reference Standard preamp (souped-up FET (This is actually a problem with the was defective. A-B testing indicated unus­ input, blazingly fast slew rate, precision entire industry: connict of interest. If it ually high distortion, agaIn confirmed with this, that and the other thing, ultrasharp turns out you can do a preamp as good as spectrum analysis. At one point, without ultra- and infrasonic filters, triple-filtered any with 2 transistors, and the public learns informing the subjects, [ hooked up an power supply, et al.) and compared it to the of this, fo lks like AG[, Levinson, and lots A-B which simply reversed the phase of one worst preamp [ could think of at the time: the of others have a problem. So do retailers of the channels. The subjects, used to listen­ two-transistor circuit copied almost verbatim who depend on high-end sales. So do equip­ ing to ultra fine differences between pre­ from the G E transistor manual (which actual­ ment review magazines, who need something amps, were almost knocked over by the ly has three transistors; the output is buf­ to talk about each month ... Hell, not magnitude of the diffe rence. Yet it took them fered through an emitter follower, but why even Pioneer wants everyone to buy their almost 5 minutes of guessing to figure quibble) and found I could not tell them $200 receiver and ignore all their higher­ out what [ had done, despite the fact that they apart. priced ones. This concept extends to other were well-heeled in both audio and elec­ And this was after [ had been a part>, components as well-the industry needs tronics. To me, this strongly reinforces both to a couple of "informal" A-B comparI­ things like TIM, phase response, moving the validity of the A-B test procedure, and sons, where we listened to one preamp for 20 coil cartridges, etc. to stay fi nancially healthy. the notion that even experienced listeners minutes, then disconnected it, connected an- Frankly, [ disbelieve most of it-[ think are notoriously inaccurate in inferring the

10 cause of differences they perceive. (A lengthy passage discussing the proper that said the Phase 400 was the hot setup. (Incidentally, we made no attempt to way to match cartridge/preamp fr eq uency Then I think there was another one that said control when subjects should switch, or how responses fo r A-B testing is deleted here. that the BGW, which we did try, was the long they should listen before responding. since the main issue has little to do with only way to fly. In the second place, nobody If a subject wished to hear a whole cut laboratory technique. -Ed.) ever gives a solid cause-effect reason why again before making a single response, we There are a few other points in your their particular super product is alleged to played it. Inter-switch intervals ranged from letter that I would like to comment on. actually sound better than the other guy's. a couple of seconds to well over a minute. You take note of "certain superior char­ (You say your preamp has .0008% distortion? Subjects were not told what to listen for. acteristics of the JC-2 (imaging, definition, Why should that sound better than their Thus, it seems that there was an element clarity)." While these terms may have some .0 I % preamp when the threshold of audibility of longer-term listening involved here-at meaning to you, and may be valid descrip­ is .2% under the best conditions, and when least over a period of minutes.) tions of what you think you hear, I know the cartridge is producing 4%?) Not even There was one more check on the possible of no serious psychophysical work that has Mark Levinson can tell you why his preamp effects of long-term listening that I made. been done to show these as being real, is supposed to sound better in explicit terms. Although it was an informal test, I found repeatable, measurable effects. The closest In the third place, as I've mentioned, the results somewhat revealing. In a second I can come is the concept of image fusion, whenever there was a certifiable difference, attempt to make a flat-response/lousy-sound­ from studies in localization and binaural in level, frequency response, distortion or ing preamp, I took a lowly 1458 dual 74 1 lateralization, which is a fu nction of how phase, the subjects heard it-and heard it op amp, put an RIAA feedback loop around well the two channels are matched. The as soon as it exceeded recognized thresholds each half, and called it a preamp. With a characteristics and thresholds of that match­ of audibility. On all of the test setups. pair of 9-volt batteries the whole thing fit ing have been extensively studied, and (But earlier in this same letter you into a small minibox. I A-B'd it against the there is no preamp that I know of that comes bracketed phase resp onse with industry hypes other preamps in the lab and, after tweaking within light years of having sufficient like TIM and moving-coil cartridges.-Ed.) frequency resp<,Jnses and input capacitances, channel mismatch to cause detectable de­ In short, Mr. Tomlinson, you seem to be the 74 1 preamp was indistinguishable, ex­ terioration in binaural fusion, especially asking me to believe in magic. You claim cept at very high listening levels with ex­ when auditioned over loudspeakers. (Never there are preamps out there with conventional ceptionally quiet records, where the dif­ heard of cross talk. eh?-Ed.) The channel specs-things like frequency response, dis­ fe rence in noise level was apparent audibly. mismatches present in cartridges, records, tortion, noise, input impedance-identical Even there, the difference was small: with loudspeakers, rooms, and ears far exceed to the "finest" preamps, which nevertheless a Y -15 (( I cartridge connected, the preamps those of preamps. Frankly, I suspect quali­ are audibly inferior, but you do not tell me we looked at all had SIN ratios of about 82 ties like imaging, definition, and clarity what these inferior preamps are doing to the dB, A-weighted, re 10 mY, I kHz. The 741 are strictly fu nctions of frequency response, signal to screw it up in an audible fa shion. was 77 dB. After the A-B tests, I was still level, and, alas, imagination. As you note You claim the audible differences start out very suspicious of the results, so I took the in your letter, "the main problem is the as barely audible in A-B tests, but become 74 1 preamp home, where for months I had psychological phenomenon that occurs in more audible with casual long-term listening, been listening to a good-sounding Davis­ our heads. We seem totally unable to control without explaining why. You claim these dif­ Brinton unit. My first reaction after substi­ the way the differences grow as exposure ferences are audible only on "state of the tuting the 74 1 was that things didn't sound time increases." I agree with that statement art" equipment, without providing an ex­ right, but I couldn't put my finger on the in part, although I think there is a limit to plicit definition of what makes a given piece cause, so I left it in the system, in place how far those differences can grow. The of equipment state of the art. Having failed of the Dav-Brit. An hour later, I looked problem is that a subject will tend to rein­ to specify the magical ether that seems to up pleased with how well the system was force a perceived diffe rence whether or not make it through the good preamps, and not sounding until I realized that I was listening it exists. Suppose for example that when the bad, you are obviously not prepared to to the 74 1 preamp; then it didn't sound switching from switch position A to switch say how it gets through "good" amps, tone quite right again.This phenomenon was re­ position B one time, the high frequency con­ arms, cartridges, and speakers to become peated several times over the next day or so. tent of the program material drops. The sub­ imaging, definition, clarity, what have you, As long as I didn't think about what preamp ject might begin to suspect that the preamp upon striking our ears. was on, the system sounded as good as ever, in the A position was brighter than the B About 12 months ago I set out to try to but as soon as I realized that the preamp preamp. Instead of asking himself, "Is find that ether. I have been trying assiduously consisted of nothing more than a 74 1, it there a difference between A and B?" he to track it down, but somehow it always stopped sounding good. Clearly my prejudice will now be asking himself, "Is A brighter seems to hide behind a tree when it sees against the 741 was manifesting itself in very than B?". The latter question shows bias, me coming. I'm beginning to think it unfair audible terms, whenever that prejudice was al­ the former does not. Faced with a psy­ that I should be the only one to have to carry lowed to fu nction. When the prejudice wasn't chological compulsion to find a diffe rence, the burden of proof. (Very interesting. Now active, neither were the sonic deficiences. the subject will tend to reinforce the per­ tell the doctor when you first began to ex­ Eventually, after a couple of days, I guess I ceived diffe rence, real or imagined. If, how­ perience this fe eling of isolation.-Ed. ) Be­ became convinced of the basic goodness of the ever, the correspondence between switch posi­ fore too many more people are coerced into 741 preamp, and I stopped "hearing" de­ tions and preamps is switched randomly af­ blowing twelve hundred bucks on mystique, I ficiencies. I loaned this preamp to Peter ter each trial, the subject can't always be sure think it would be nice if someone showed Mitchell, of the "Shoptalk" show, and he that A will be the brighter one. Thus if the in clear, irrefutable repeatable terms exactly subsequently informed me that he thought it difference was only imagined, the subject's what those twelve big ones buy in audio was one of the better sounding preamps he responses will be guesses, and he will be right quality, and how that can't be just as easily had heard. only 50% of the time, providing enough trials bought for $17.50. In other words show, using This I believe illustrates one of the are done. The point is that this psychological controlled, blind listening tests, short-term deficiencies I perceive in the tests you brief­ reinforcement makes it imperative that the or long, that a diffe rence really exists, and ly describe in your letter. Namely that the subject not know what he is listening to, and then provide a clear cause-effect relationship people doing the long-term listeninjl tests that enough random trials are carried out to that explains the origin and behavior of those know what preamps they are listenmg to. account for guessing, even in long term lis­ diffe rences. Then justify why those differ­ (Or so it seems from your description.) I tening tests. ences should cost as much as they seem to. am afraid that this factor alone completely When we started out to study' preamps, A fter some record producer has tossed invalidates any results you might get. I we realized it would be impossIble to try equalizers, filters, limiters, compressors maintain it is totally impossible to A-B every available preamp, power amp, turn­ and reverb into the signal chain with gay knowingly a preamp made with $2 in Radio table, cartridge, etc., in every combination, abandon, I don't see why I should spend 5 Shack parts and chicken wire to a $1200 so we chose what we fe lt was a representa­ times more for a preamp just to get an RIAA preamp, without experiencing some measure tive sample of equipment, making sure that response that's 0.2 dB more accurate at 20 of bias, and that this bias will manifest it­ everything was duplicated at least once, and kHz. At the very least, if there are differ­ self in apparent audible diffe rences, even that the preamps covered as wide a range as ences that can only be discerned on "near if the two circuits are in fact electrically possible. Nevertheless, it comes as no sur­ state-of-the-art" equipment, I think your identical. (You think any two resistors or prise to have someone claim that the reason readers should be so informed, so they won't capacitors of the same rated value and we couldn't hear differences was that we waste large sums of money on a preamp if tolerance are electrically identical? Not so. didn't use thus and such a cartridge, arm, they happen to own AR-I I loudspeakers, or -Ed. ) In running our tests, the subject was power amp, speaker, whatever. Frankly, I other equipment you deem not near state of never informed in advance which preamps don't find the argument terribly compelling. the art. he was listening to. Nor was any feedback In the first place, the experts out there can't I am willing to have my claims proved given after each trial; if we are talking about seem to come to any consensus as to what incorrect, Mr. Tomlinson, but I don't be­ differences that really are hearable, you the finest equipment is. You claim we should lieve you have done so. shouldn't have to tell the subject when have used a Dyna 400 or an Ampzilla, but I he's right or wrong, he should tell you. seem to remember reading some other review And now fo r the Mark Davis letter writ-

11 ten directly to us, in response to our preamp the 47K ohm input termination resistance. score" (p. 7), further commenting that re­ survey. Since it is the normal mode of operation quirements of fi ner tolerance are to be con­ of the preamplifier, terminating the input sidered "obsessions" (p. 8). As noted above, with a cartridge while measuring its noise our investigation indicates the need to equal­ The Audio Critic: seems to us to be the most realistic approach. ize overall playback frequency response I was recently given the opportunity to True, the result will depend to some degree within a tolerance of 0.3 dB. That is, the read an article published in your premiere on the cartridge chosen; nevertheless it will cumulative effects of input impedance, fre­ issue which discussed at modest length cer­ be more representative than the shorted in­ quency response, and playback level must tain research concerning RIAA phonograph put specification. The latter is probably not induce a diffe rence greater than 0.3 dB preamplifiers to which I was a party. The more apropos when a low impedance device, in level between the two preamplifiers being principal fi nding of this research was that such as a pre-preamplifier, is connected tested, at any frequency in the audio band. competently designed phonograph preampli­ to the preamplifier. Even then, it is of Preferably, the diffe rence in overall level, fiers, ranging in complexity from the costliest, secondary interest, since the cartridge and weighted on an octave basis, should be zero. most abstruse topology, to the simple and pre-preamplifier input noise will inevitably This constraint was not imposed capri­ popular two-transistor configuration, are in dominate, to be masked in turn by record ciously; it resulted from concrete psycho­ fact audibly indistinguishable under any con­ noise. The need to terminate the input acoustical data. Using the same A R speakers ditions of normal operation (i.e. reproduc­ with a cartridge applies as well to the you think so little of, and the same A-B tion of commercially available records by measurement of the input noise of a pre­ paradigm you denigrate without quantitative conventional magnetic pickups), providing amplifier, incidentally. justification, we found subjects could re­ the concept of competence in design subsumes We used a Shure M9 1 E cartridge; while liably discriminate between sources solely the equivalence of overall playback frequency its sonic qualities may not be state of the on the basis of frequency response disparities response to a tolerance of 0.3 dB or better. art, its inductance and other electrical of as little as 0.7 dB. By limiting differences (Moving-coil pickups are also magnetic, we characteristics are typical of a large number to a maximum of 0.3 dB, frequency response/ subsume?-Ed. ) of current state-of-the-art cartridges. With level discriminability was found to be virtual­ It is possible that certain misconceptions the cartridge mounted in a shielded box to ly eliminated, permitting accurate acoustical may have been promulgated by my brief dis­ suppress stray hum fields, and the stylus re­ evaluation of other possible disparities. The cussion of the research on the WBUR radio moved to avoid acoustic pickup, we found tolerance of ± 0.75 dB you encountered im­ program, "Shoptalk," which you indicate almost all preamplifiers had a signal-to­ plies a worst-case deviation of 1.5 dB, as the source of your information. To the noise ratio of 82 dB, ASA 'A' weighted, which, when added to the tolerance of the extent that you find it enlightening, I am re: 10 mY RMS, 1000 Hz. This is in good level match you made, 1.0 dB, results in a enclosing a copy of a letter I have previously agreement with Hallgren'S data. maximum possible deviation of 2.5 dB, not sent to Randy Tomlinson, of Accusound We checked these measurements by lis­ counting the possible effects of input im­ Engineering and StereOpus magazine, which tening to the noise of the preamplifiers, after pedance variation. treats in somewhat greater detail some of the first taking care to match levels with a test Your plausibility arguments on pages questions at issue. tone, with the respective inputs terminated 7, 8, II and 12 notwithstanding, this diffe rence There are a few other questions your in turn with the same cartridge. While there is, by actual empirical observation, well be­ article raises that I would like to attempt were clear variations in the spectrum of the yond the point of becoming plainly audible. to clarify. These generally involve electrical noise among units examined, the overall Your failure to carefully measure and measurements or listening technique. level remained essentially constant. control playback frequency response/levels I concur with your conclusion that neither Basically, the same laws of solid-state (casual tweaking with tone controls does not harmonic nor 1M distortion is a significant physics and thermodynamics hold for all hi­ constitute careful measurement and compen­ factor in modern preamplifiers. Sometimes fi companies. For about 10 cents you can buy, sation) makes it possible, nay likely, that (too often) the opposite conclusion is reached in production quantities, a transistor that will your judgments primarily renect prefer­ by driving the preamplifier being evaluated perform as well as possible with available ences in frequency response. Of course, it is with electrically generated test signals to cartridges. While there are some games you our contention that this must be so, as there levels hopelessly greater than will actually can play with the 47K input resistance, only is no other audible variation in preamplifiers; be encountered by the preamplifier under small gains beyond 82 dB are possible, and but even if there is, it is likely that it actual conditions of use (cf. lung, 1977). even the least expensive receivers appear to was swamped out by the potentially large Obviously, I also take no exception hit that figure routinely. We have thus con­ variations in frequency response and levels. to any reservations you may have about tests cluded that the concept of an ultra-quiet You dismiss the AR loudspeakers as "too employing square waves, wideband noise, phono preamplifier is largely a myth. (What fo rgiving" (p. 10), without citing any refer­ or other similar, unrealistically fast test sig­ about small-Iy? Th at's our bag.-Ed. ) ences to back up this claim, or stating how nals. To the best of my knowledge, Tomlinson One of the largest areas of disparity one goes about measuring the "forgiving­ Holman no longer believes his square wave among preamplifiers was found to be input ness" of a loudspeaker, or what specifica­ test to be relevant; nor does Alvin Foster impedance, so I was a little sorry to see tions contribute to this alleged quality. endorse his white noise test (Foster and that you failed to measure it. The use of a (References? If authoritative criticism of Swanbon, 1977). low-inductance cartridge probably minimized speaker systems-by brand namef-were It is gratifying that you find a pre­ the effect in your tests, but careful meas­ available in references. we wouldn 't be in busi­ amplifier with relatively low slew rate (I volt urement of relative frequency response, in­ ness. -Ed.) We acknowledge that there are per microsecond) can deliver excellent sound. cluding the effects of the cartridge, as loudspeakers which have natter frequency Our calculations indicate that such a figure described in the accompanying letter, would response than the ones we used, but this is is at least an order of masnitude larger than have been preferable to assuming the input not a necessary prerequisite, since the same what is required by available records and impedance had no effect. Furthermore, not speaker is used for both A and B presenta­ cartridges. An interesting sidelight is that everyone uses the cartridge you employed; tions. What is important is suffi ciently wide slew rate figures comparable to I Y IuS are many popular cartridges are sensitive to response, without serious gaps, to ensure that obtained from typical well-designed preampli­ variations in input impedance, raising the all audible frequencies are reproduced at fiers employing the two-transistor topology. possibility that a preamplifier that sounded hearable levels, plus sufficiently low dis­ Your results on signal-to-noise ratio, that fine in your tests would deliver inferior sound tortion to avoid masking possible distortion some units were "measurably and audibly with a cartridge having greater inductance; from the preamplifiers being auditioned. (But quieter than others," and that some were ergo the importance of reporting this speci­ what kind of distortion? Steady-state or time­ "outstanding" (p. 8), runs somewhat counter fication. dispersive?-Ed. ) to our own measurements, which showed fairly (We don't know of a single high-in­ Both of these criteria are met by the strong consistency among the units tested. ductance. i. e. , input-cap acitance-sensitive, AR speakers. We have measured the 1M This may be due to differences in measure­ cartridge that can compete sonically , regard­ distortion of a number of commercial tweeters ment technique; regrettably, you do not dis­ less of loading . with moving-coils like the and fo und it in all cases to be under 0.5% cuss the protocol you employed. Denon or EMT, or low-inductance moving­ at a level of 100 dB SPL, 6 feet on axis, As you may know, the noise of a pre­ fields like the Grado. We don't recognize the distortion dropping rapidly below this amplifier is lowest when the input is termi­ the "rights" of low-fidelity cartridges: maybe intensity. This is less distortion than the nated with a short circuit. In trying to pre­ that's why we hear preamps differently. ear itself produces at such levels. (Yup. You sent the most attractive-looking fi�ures, too.-Ed. ) meant steadY-Slate. -Ed.) manufacturers will often specify the slgnal­ Of greater concern to the accuracy of The most relevant justification of the to-noise ratio of their preamplifiers under your report than the results of the electrical speakers and associated equipment we used this constraint. Occasionally, one will even tests are the psychophysical issues implicit in was that subjects were able to clearly dis­ optimize for it, to the detriment of the con­ your listening protocol. criminate between sources whenever there was ditions of actual use; i.e. with the input Your measurements of preamplifier fre­ a certifiable difference exceeding accepted terminated by a magnetic cartridge. Hall­ quency response show unit-to-unit variations thresholds of audibility, as fo r example in gren (1975) has shown that the principal on the order of 0.75 dB, a tolerance which the case of frequency response/level devia­ source of noise in a preamplifier is in fact you claim is "good enough to eliminate the tions of only 0.7 dB (which you believed to be the cartridge, followed in importance by possibility of listening preferences on that inaudible). This fa ct, plus the strong con-

12 sistency in results with alterations in ex­ of the preamplifiers was the only informa­ perimental protocol, you are urged to try it. perimental equipment, including the substi­ tion available to the subject on which to Sincerely, tution of Koss ESP-9 electrostatic headphones base a decision. Mark F. Davis for the loudspeakers, lends fairly strong The probabilistic nature of the results Massachusetts Institute empirical credibility to the results we ob­ is reflected in the fact that most people could of Technology tained. discriminate frequency response disparities Cambridge, MA (Hold everYThing! We never said that of several dB most of the time; some people diff erences of 0. 7 dB are inaudible. but could discriminate freq uency response dis­ rather thaT they aren't relevant to our con­ parities on the order of a major fraction of a clusions. As fo r the electrostatic headphones. dB some of the time; and no one could reli­ References: did you use them when you A-B-ed the two­ ably discriminate disparities of 0.3 dB or Davis, M.F. (1977). "A Simple Im­ transistor GE preamp against the Mark less. Allegiance on the part of the subjects pedance Bridge," The B. A.S. Sp eaker (Box Levinson? Right then and there? -Ed.) to a particular concert hall or type of music 7, Boston, MA 0221 5), 5, 7, 3-8. Lacking a proven alternative mechanism had no effect on the results. Toole (1977) Foster, A., and Swanbon, W. (1977). to account for differences in preamplifier has recently presented some data indicating "A Comparison of Low Priced Receivers," sonic quality. your allegation that our equip­ that experienced listeners ("golden ears") submitted to The B.A .S. Sp eaker. (Published ment fails to preserve the effects of this un­ are not even required in preferential studies; July 1977. -Ed.) known mechanism would seem somewhat pre­ naive listeners can be just as consistent. Griesinger, D. (1976). Private com­ mature at best. In summary, your listening tests in­ munication. The rationalization you advance for corporated some very serious lapses of ac­ Hallgren, B.I. (1975). "On the Noise eschewing the use of an A-B test is that it cepted psychoacoustical procedure: the fail­ Performance of a Magnetic Phonograph fa ils to show up differences that are really ure to equalize for differences in level and Pickup," J. A ud. Eng. Soc .. 23, Sept. there. because "your ear tends to integrate freq uency response; the fa ilure to keep sub­ Jung, W. (1977). "Slewing Induced Dis­ the sound under these conditions" (p. I I). This jects from comparing opinions, despite the tortion in Audio Amplifiers: Part I," The position is simply untenable; there is no need fo r statistical independence; and the Audio AmaTeur. 8, I, 3-9. such integration effect to mask di fferences. fa ilure to conduct independent blind trials, Luce, R.D., Bush, R.R., and Galanter, As explained in this letter and the one without feedback, prior information, or other E. (1963). Handbook of MaThematical Psy­ accompanying it, our A-B tests revealed ex­ sources of bias, in sufficient number to en­ chology VI (John Wiley and Sons, Inc., tremely fine differences in level, frequency gender reliable statistics. New York). response, and distortion. Beyond that. nu­ Even long-term listening tests, for which Toole, F.E. (1977). "Speakers-Part I," merous diverse disciplines in psychophysics you express preferency, can be done properly, A udioScene Canada, March, 19-28. make use of the instantaneous two-interval providing the same rules of technique are A-B discriminability paradigm, usually to observed. As it stands, I doubt that any explore the highest levels of human resolu­ competent psychophysicist would endorse tion. your methods or accept your conclusions. Our overall reaction to these two letters I know of no authoritative, controlled Lurking behind these issues is the con­ is thaT. even though they fly in the fa ce of all study showing long-term listening as you ceptual qu�stion of whether it makes any the slowly and painstakingly accumulated wis­ describe to be a more accurate indicator sense to test preamplifiers when you profess dom of an entire generation of audio per­ than A-B testing; if you know of one. please not to know what makes them sound dif­ fe ctionists. they mustn 'T be dismissed out of cite it. I n the meantime, I commend to your ferent. For one thing, the perceived superior­ hand. Obviously Mark Davis and his as­ attention the Handbook of Mathematical ity may not be worth the extra cost of the sociates are highly knowledgeable. dedicated Psychology (1963) as a viable primer on the better units: it may be possible to incorporate and sincere. Wh at they claim they heard is subject of competently conducted listening the extra "goodness factor" which you claim undoubtedly what they actually heard-and tests and their analysis. the better preamplifiers possess into a cheap quite possibly. had we been there. we too The attitude toward statistical testing preamplifier fo r next to nothing. For ex­ would have heard the same thing. The ques­ is perhaps one of the most contradictory ample, if you assume for a moment that our tion to ask is why it sounded that way. elements of your article. You claim, without contention is correct, that freq uency response The most plausible explanation is that providing justification, to have "very little is the only significant parameter, then it is the system into which the preamps under faith in statistical surveys (' 19 persons pre­ fo olish to recommend an expensive unit test were inserted had insufjicient resolving fe rred A and 22 preferred B')" (p. 10). im­ when an inexpensive one will either have the power. Our experience is that plugging dif plying incorrectly that a preferential method­ same response, or can be adj usted to have it fe rent preamps into. say. a Beveridge Model ology was employed in our tests. Having by a suitable choice of input capacitance. 2SW sys tem makes the differences laugh­ made this claim, you then execute an about­ Without knowing the specific mechan­ ably obvious. regardless of matched listening face. and perform just such a preferential ism(s) involved, you cannot be sure that your levels or RIAA equalization accuracy. (The survey. conclusions are true in general fo r other in­ signal source being a Denon DL-103S car­ Frankly, I completely agree with your stallations. Again, as an example, if fre­ tridge. fa natically aligned fo r optimum track­ lack of faith in such surveys, unless the quency response is all important, a revised ing laterally and vertically. ) Th rough Dahl­ basis for listener preferences can be object­ choice of associated reproduction com­ quist DQ-IO·s. used in our original survey. the ively isolated, and unless it can further be ponents would very likely have altered the diff erences are still easily audible though shown that the preferences expressed represent preferences. As already noted, switching less flabbergasting. Through the Mark Davis a fairly universal consensus. Otherwise, it IS just the power amplifiers changed the order setup-who knows? We have a fe eling that indeed true that one may "come to a different of the two best-ranked preamplifiers. Without total disregard fo r the masking effe cts of time­ conclusion with a diffe rent selection of lis­ an alternative mechanism, clearly identified dispersive distortion-originating fr om the teners" (p. 10). You have made no attempt to and characterized, you cannot prove that cartridge and the speakers-and exclusive show quantilatively that your preferences your preferences will hold for anybody else. concentration on amplitude resp onse and would remain the same with a different, It is significant that the results and steady-state distortion channeled the Mark independent group of listeners. Indeed, conclusions we reached have also been ob­ Davis tests in a self-fu ljilling direction. you have not even shown that your preferences tained by the "highly active and enlightened" We say self-fu ljilling because we also are independent of the equipment chosen; (p. 9-here, we agree completely) Alvin have the distinct impression that not only to the contrary. the single substitution you Foster, of the Boston Audio Society, who Mark Davis but the entire Boston Audio did make, of the power amplifier, reversed fo rmerly espoused the relevance and use of Society crowd (delightfu l and sophisticated the order of preference of the highest ranked tests employing very high-frequency signals, practitioners as many of them are) still would preamplifiers. such as wideband noise. Some of these re­ like to discover an audio Nirvana fo r fo rty­ (Th(l" s right. We were trying to help sults will be published shortly (Foster and nine dollars and ninety-jive cents. We discern our su/- l •. ribers choose a good preamp. not Swanbon, 1977). in them a peevish hostility toward ultrahigh­ to wriTe a texTbook on psychoacoustics. The same conclusion was also reported end equipment. perhaps because so many of -Ed.) to me by David Griesinger (1976), of Har­ them are academics and empathize with Almost without exception, the human vard University and the Boston chapter of music-loving students who can't affo rd a auditory system responds in a stochastic the Audio Engineering Society. The phe­ Mark Levinson ML-I. And, of course. the fa shion; present a stimulus to a subject nomenon thus does seem repeatable. best way to dismiss the expensive stuff is to at two different times and, in general, you Hopefully, this information will help jind it scientijically indistinguishable fr om will get two different responses. Such be­ clear up some of the misunderstanding that the run of the mill. havior makes the need for accurate statistical seems to exist. I know from personal ex­ We must also remind our readers. to put characterization manifest. perience that the results of informal listening things in proper perspective. that Mark Davis For the record, we never asked a subject tests often seem so obvious as to make a also considers TIM in power amplijiers to be which preamplifier was preferred, only more precisely controlled environment ap­ a myth. Get the picture? It's not only the whether there was a discernible difference. pear unnecessary. More often than not, emperor who is naked, but also the empress. Subj ects were given no feedback or prior however, the impression is a false one. Be­ the crown prince. the archduchess ... information that could cause bias; the sound fore rejecting accepted psychophysical ex- -Ed.

13 The Realities of Noncommercial Audio Journalism

By Peter Aczel Editor and Publisher

As we go into our fourth issue, we realize that some of our sub­ scribers still don't perceive us as we really are and don't quite know what to expect of us. Hence this continued exposition of our phi­ losophy and of the facts of life in our delirious business.

Although what fo llows is only loosely con­ in available space and personnel, as other audio nected with the two previous editorial discus­ reviews that don't carry manufacturers' adver­ sions of our point of view, we prefer to con­ tising. What distinguishes us from the others is tinue the sequential numbering of topics fo r that (a) we work harder-or at least generate convenient reference. more output in a given time, (b) we were willing

* * * and able to make an initial capital investment in a completely equipped, in-house laboratory Just because our publication doesn't make facility, and (c) we try to correlate what our ears 1 a totally amateurish impression in looks tell us with the laws of physics and mathematics and6 contents, please don't assume that the pub­ rather than techno-folklore or wishful fantasies. lisher is McGraw-Hili or some other such multi­ None of these distinctions enable us, however, million-dollar operation. We despair when, after to test 600 pieces of equipment in a year or even having tested and reviewed more than 30 pre­ to meet deadlines with unfailing punctuality. amplifiers and preamp accessories in a little Our basic limitation is, of course, eco­ over fo ur months, we get letters saying, "I'm nomic, and it's intrinsic to our noncommercial shocked that you failed to include the following format. Let's say we decided to hire a full-time II models." How big and efficient do these let­ staff engineer with a reasonably sophisticated ter writers imagine we are? audio background (rather than using part-time The fact is that we're about the same size, staff consultants as we do now). His salary 14 would wipe out the proceeds of about one thou­ suffice. For example, a tone arm with wobbly sand new subscriptions, right up front. An ex­ bearings can't possibly sound as good as one perienced fu ll-time managing editor would wipe with tight, preloaded bearings (other things be­ out another thousand. How many thousand ing eq ual). It's an obscene waste of time to let subscribers do you think we have? Now if several keen-eared people listen to such an arm A udio magazine wanted to hire the same engi­ for weeks and weeks only to have them report neer, they could do it with the income from one that there's something not quite right with the full-page ad per month. See the difference? sound. So, an equipment reviewer who is into There, in a nutshell, is also the reason why we bearings as well as Mozart and Pink Floyd can can't sell The Audio Critic at a dollar a copy. confidently say "Forget it!" ten minutes after Or even two dollars. unwrapping the arm, instead of staging an elab­

* * * orate blind A-B test against a reference standard . 7 You must remember that the original con­ We recently gave a somewhat oversimpli­ 1 cept behind noncommercial audio reviews fied summary of this point of view to someone was that testing exclusively by listening can be who asked us how we perceived the difference highly revealing in a relatively short period of between our procedures and those of another time and requires no test instruments other than noncommercial audio review. Suppose, we educated ears. Therefore the operation can be said, that a manufacturer came out with a turn­ started with minimal manpower and overhead, table he claims to be a perpetual motion ma­ with sufficient income from subscriptions and chine. The other publication would say, "A per­ possibly dealer advertising to maintain a reason­ petual motion machine? Let's have a look." able publishing schedule. We'd say, "A perpetual motion machine? Get N ow here's the paradox: The Audio the hell out of here." We have no time fo r audio Critic's way of testing is, on the face of it, in­ designers who defy the laws of nature. comparably more time-consuming than anyone * * * else's, since we keep hopping back and forth from listening room to laboratory, trying to fi nd Our special CES issue(Volume I, Number a correspondence between what we hear and 1 3), in which several such scofflaws were what we measure. This also dissipates a much poin8ted out and dismissed out of hand, provides greater percentage of subscription income on another illustration of how our departures from overhead before editorial and publishing ex­ the expected practice of noncommercial audio penses. By all rights we should be the slowest, publications can be misconstrued by certain sub­ most plodding noncommercial audio journal of scribers. We published this issue with the fi rm them all. The fact that we aren't, that we cover conviction that it fi lled an important need (way more ground more thoroughly in a shorter time in advance of commercial magazine coverage!) than the strictly golden-ear boys, is one of the and was fu lly up to the standard set by our pre­ best arguments for parallel golden-ear and vious two issues. We felt that, quite aside from laboratory testing. Because, in the long run, the their news value, our CES commentaries shed scientific (or call it technical) approach actually some interesting bits of light on a number of sub­ saves more time and energy than it uses up. jects ordinarily shrouded in mythology, such as We could spend weeks trying to determine stylus shape, tone-arm geometry, woofer design, purely by listening tests whether an elusive low­ the bandwidth-limiting controversy, and others. frequency resonance we hear in a speaker is in­ We were enthusiastically praised for this by herent in the design itself or is caused by some more than a few distinguished members of the subtle interaction with the room, the walls, the audio community, not to mention advanced fu rniture, the rest of the system, or whatever. audiophiles. Nevertheless, about I % of our sub­ Five minutes of spectrum analysis with the scribers turned out to be extremely unhappy proper frequency span and resolution bandwidth about the whole thing. (To be precise, 1.4% com­ will, on the other hand, provide an unequivocal municated with us specifically in regard to the answer. Occasionally it may not even be neces­ change of pace set by the third issue: 0.4% pro, sary to perform a laboratory test to save listen­ I % con, 98.6% silent and presumably unruffled.) ing time; simple inspection by someone who un­ It was our distinct impression that the derstands the applicable design principles will majority of the complainers didn't really dig 15 into the issue but took one quick bite, found that twice a month. (Mark Levinson ML- l-good, the fl avor was different than last time and start­ Dynaco PAT -5-bad, Dahlquist DQ- l O-good, ed to hol1er. They certainly seemed to be un­ Janis W- I-bad ...Ne xt!) We hate to disap­ aware of the aforesaid little raisins and choco­ point such seekers of predigested simplicity, late chips inside. Sorry, guys, we never said it but that's not our idea of a serious audio journal. would be vanil1a every time. Nor did we ever We'd much rather coax, cajole, browbeat, promise 48 or 52 or any other number of pages lecture and just plain educate our subscribers per issue for your $4 .67. We use up as many (at least those that aren't hardened profes­ pages as we feel is necessary to cover the subject. sionals) to the point where they'l1 trust their I n the future, there will undoubtedly be 64-page own judgment, even about products they've issues as wel1 as 24-page issues. We sel1 informa­ never seen reviewed. If we can accomplish that, tion and advice, not yard goods. the difference between 52 and 24 pages, or To keep everybody happy, however, we $4.67 and $3.50 per issue, wil1 begin to appear don't plan to publish a separate CES issue rather irrelevant. in 1978 but wil1 probably cover the CES in one And when that day comes, we'll get no more of our long surveys within the framework of a letters saying, "I just can't wait till you review normal (i.e., fatter) issue. the Schmidlapp C-95-I've owned one for six

* * * months." The last thing an enlightened audio enthusiast should need is a review of a product 19 The subscriber who gets bel1igerent about he already knows intimately. thenumber of pages perdol1ar, the number * * * of words per page (our type size is large, see?) and that sort of thing seems to be the same one Al1 right. We'll stop biting the hand that -representative of a very smal1 minority, mind feeds us and present herewith what we believe you-who would like us to run long laundry lists is our best, most useful issue so far. Read it, and of recommendations without any fancy guff then see if the above comments have helped to about slew rate or sidebands or Q. Preferably clarify who we are and where we stand.

Editor's No te: Consultation by telephone on individual purchasing decisions or installation problems emphatically isn 't part of the services offe red by The Audio Critic fo r the price of a subscription, even if you're resourcefu l enough to track down the Editor's home phone number.

16 InYourEar

"That was the last test, Otto. "A ll we need now is J. B. 's approval. The prototype is finished." Wait till he hears this one!"

"J. B., you've just got to come down to the lab and hear the new speaker. "Eh?" It's fa ntastic!"

17 Sophisticated Speaker Systems, Large and Small: A Comparative Survey.

By the Staff of The Audio Critic

Part I: Comprising seven big ones, four little ones, three medium­ sized jobs, plus a decent subwoofer, which should be enough to keep everyone arguing until Part II.

There are good reasons why speakers are and demonstrated in very few places. As a re­ the most controversial of audio components. sult, the typical arguments tend to be parochial, First of all, they are the least accurate, intro­ as they are engaged in without awareness of ducing much greater discrepancies between what is currently possible. Amazingly, even input and output than amplifiers or tape speaker manufacturers suffer fr om this limita­ recorders or even phono cartridges. That makes tion; you'd expect their sound rooms to bulge most comparisons highly dilemmatic, as they with the latest and greatest multi-kilobuck are essentially between varying degrees of speakers fo r reference purposes, but the truth "bad." (Who was the nicer guy, Adolf Eich­ is that most of them (rich or poor!) listen ex­ mann or Charles Manson? You're wrong!) This clusively to their own products. Their dealers basic judgmental problem never goes away; it's are usually way ahead of them in listening there whether one is comparing the Beveridge expenence. and Koss electrostatics or the house-brand The most profound source of confusion, specials of a hi-fi chain store. Name your however, is the quality of the signal fed to the preference and someone will immediately tell speakers in nearly all comparisons. A record you what's wrong with it. And he'll be right. made with peaky Neumann microphones is No speaker, at any price, is without faults. placed on a turntable with insufficient feed­ Another cause of unending controversy is back isolation and a motor board that drums. A that a lot of audio enthusiasts with strong cartridge mounted conventionally, with un­ opinions have never heard, let alone lived with, necessarily large tracking error, in a tone arm the best available speaker systems, which are with sloppy bearings (cf. our cartridge/ arm/ usually made in extremely limited quantities turntable article in this issue) is lowered on 18 the record. The signal is applied to the speakers fo rmance. We didn't want to expose ourselves through an amplification system with TIM, in­ to the temptation of subjectively aha-ing and sufficient voltage swing, power supply limita­ you-seeing fr om our listening chair what we had tions, etc., etc. The speakers are then found already established objectively through meas­ "crisp" or "nasal" or "honky" or "airy" or urement. Certainly not the first time we heard whatever. And these observations are then a speaker. All the initial aha's took place in the passed on to audiophile friends or even pub­ laboratory, as they should-and there were lished in underground audio reviews. If at least plenty of them. Unlike, say, preamplifiers, the signal had the same faults in each listening loudspeakers haven't evolved yet to the point test! But, needless to say, its quality varies com­ where they can hide their faults from measuring pletely at random, and so do the conclusions. instruments. When a speaker doesn't sound right, it doesn't measure right-and since not a How we tested them. single one sounds quite right, not one measures What we said about A-B listening tests in quite right. But, gratifyingly, the ones that connection with subwoofers in our second issue sound better measure better. Needless to say, (Marchi April 1977, page 30) applies equally to more than one or two kinds of measurement are full-range speaker systems. The audible differ­ necessary, and the results are seldom cut-and­ ences are so gross, between any two models at dried, being generally the reflection of deliber­ any volume level, that A-B switching becomes ate design trade-offs and therefore subject to a pious exercise in methodology yielding no interpretation. Even so, we're convinced that if additional information. When you can fi nd two every speaker manufacturer routinely per­ really good speakers that have to be carefully formed our simple tests (as we know fo r a fact A-B-ed to establish which sounds more ac­ nine out of ten don't), speaker design would be curate, you'll know the art has advanced by a a far less haphazard affair today and the art whole order of magnitude. would advance more surefootedly. On the other hand, one-by-one listening We must emphasize, however, that our evaluations require absolutely unquestionable main concern in the laboratory was to make signal quality, and we went all out to obtain it. sure we hadn't made a mistake in our listening Only the fi nalists of our preamp and power evaluations-that what we're reporting here is amp surveys were used (incidentally, the new based on some sort of reality rather than on Hegeman HPR/CU preamp has emerged as a irrelevant artifacts of our component chain or very serious contender for reference status, al­ on the aural idiosyncrasies of our staff mem­ though we won't be ready to rate it against the bers. For that reason, the overall thrust of our Mark Levinson, Rappaport and others until the measurements was qualitative rather than next issue); our current reference cartridge, the quantitative; in the majority of cases we went Denon DL- 103S, was religiously aligned for only as fa r as it seemed necessary in order to optimum tracking both laterally and vertically; authenticate our conclusions and recommenda­ we eschewed all records, no matter how dear to tions-not to determine by how many dB or Hz our heart musically, unless they were obviously or msec a manufacturer ought to change his recorded with flat microphones, like the Mark design parameters to end up with a better Levinson Acoustic Recordings; and, to estab­ speaker. (We try to give our subscribers useful lish an ultimate reference point, we unleashed advice and reliable explanations, but fo r $28 a our new, highly modified Stellavox 'Stella­ year we're not going to give any company a master' tape deck, playing direct IS-IPS copies product research and development program.) of selected original master tapes. (Please don't After measuring, we listened again, this write us to ask how we buy, beg, borrow or time of course "corrupted" in our subjectivity steal these.) In other words, we were listening by our newly gained knowledge. And then, in to the speakers, not to signal distortions ahead some cases, we went back to the laboratory of the speaker terminals. once more, just to double-check certain obser­ After extensive listening, we measured. vations and correlations. The remarkable thing We made a point of starting our laboratory about this process is that it's synergistic; one tests only after we had fo rmed a reasonably becomes a more discerning listener after having fi rm opinion of how a speaker sounded and objectively verified in the laboratory some what was audibly right or wrong with its per- purely subjective impressions, and one meas- 19 ures with greater astuteness and more concrete ticks, if you like) ranging in width from approx­ results after having aurally zeroed in on certain imately 1 msec to 0. 1 msec, spaced so far apart sonic peculiarities. Pretty soon we began to that room reflections didn't enter into the pic­ wonder how anyone could possibly form a seri­ ture. Both the oscilloscopic trace and the ous opinion about a speaker without going audible quality of such pulses clearly reveal through these steps-forgetting that there was time-response information not easily obtain­ a time when we didn't quite do it this way our­ able any other way. The differences fr om selves. (Who says a critic must be born perfect speaker to speaker are absolutely startling on and never undergo any development?) this test: some will reproduce a very adequate copy of the input pulse, with no garbage in the The lab measurements. silent interval that precedes the next pulse; We measured each speaker in both the others will generate an output totally unrelated frequency domain and the time domain. We're to the input. convinced, without any reservation, that one In addition to these basic tests, we also lis­ without the other yields an incomplete picture tened to (and occasionally spectrum-analyzed) and can't be satisfactorily correlated with the white noise and pink noise through the speakers, overall subjective impression made by a speaker. checked for structural faults, investigated im­ We do lean, however, toward the time domain pedance characteristics when we suspected sig­ as the more important of the two-otherwise nificant peculiarities (but only then), and why should a live singer or a live piano heard performed other little peripheral testing chores. through a half-open window all the way around We must repeat: our emphasis was on insight the corner still sound unmistakably live instead rather than on the accumulation of raw data. of canned? The frequency response is certainly The laboratory instruments we used were shot to hell that way; only the time response the Bruel & Kj aer 4133 calibrated microphone, retains its integrity, since all components of the the Hewlett-Packard 3580A spectrum analyzer, sonic information are still traversing the air the Hewlett-Packard 1740A oscilloscope, the path at the same speed. (When you get right down Wavetek 185 fu nction generator, the Sound to it, there's no such thing as "frequency"­ Technology 1700B distortion analyzer (with identical sine-wave cycles per unit time-in the built-in oscillator), the General Radio 1382 real world: it's an artifact of the human mind. random-noise generator, and the Hewlett­ But there is time all right, and nothing remains Packard 3435A digital multimeter. the same from instant to instant.) Enough of this philosophizing. Here's what What about our usual sweeping we did: generalizations on design criteria? We measured the "nearfield" frequency Designing an electroacoustic transducer response of each driver in each speaker. This isn't at all like designing a preamp or a tuner. corresponds very closely to anechoic measure­ No one has ever come even close to theoretical ment, as already explained in the subwoofer perfection in a full-range speaker system, so it article in our second issue. We also took the would be wrong to assert categorically that overall frequency response curve of each com­ successful design must proceed in this or that plete system and, in some cases, of pairs or sub­ particular direction. Somebody would prompt­ groups of drivers within the system. Both linear ly go the opposite way and probably do just as and log sweeps were used in these measure­ well. We therefore refuse to take sides on elec­ ments, and both axial and power response were trostatic vs. electrodynamic drivers, crossover­ investigated. less vs. multiple-crossover systems, sealed vs. Whenever it appeared that single-frequency vented boxes, or any other design approach that distortion would be relevant to our evaluation, can be either 95% right or 95% wrong, depend­ we checked the output of individual drivers for ing on how it's executed. We will take sides THD at the frequencies in question. when it comes to the basic laws of nature (say, a We ran quite extensive tone-burst tests on claim of deep, loud bass out of a lll -cubic-foot each system, at many different frequencies, minisystem) or obviously wrongheaded engi­ with both sine-wave and square-wave bursts. neering (like an incorrectly tuned vented enclo­ Perhaps most significantly, we tested each sure, for example). system with single pulses (call them blips or We rashly promised in our second issue

20 that we'd go more deeply into crossover design obtain from our test samples, but then retro­ considerations in this survey, but the subject spect is a notoriously poor witness. We have happens to be a bottomless pit. No sooner do since quizzed every audio enthusiast known to we come to the conclusion that high-order net­ us who has had any experience with the Acoustat works with their very steep slopes can't possibly X, and their reports are wildly contradictory. be any good on account of their nightmarish A number of aficionados, including profes­ time-delay characteristics, one of our consul­ sionals with absolutely impeccable credentials, tants comes in with mathematical proof that are completely sold on the Acoustat. State of with certain kinds of time compensation they The Art, they say. Others, equally expert-and can be made to pass a perfect square wave. Of these, unfortunately, happen to be the ones course, he adds, no commercial design is likely whose ears we have always trusted and who lis­ to take that route, because of its ridiculous ten the way we do-have a much lower opinion complexity and high cost! And so it goes. of the speaker. The clincher came at the Since each speaker design turns out to be Summer CES in Chicago. The pair demon­ a law unto itself, our only choice is to make our strated in the Acoustat exhibit (presumably generalizations, such as they are, under the under the watchful eye of the manufacturer) individual product headings. So-we'll let the sounded exactly-but exactly-like our test show begin. samples. As a matter of fact, a well-known and highly respected European record producer, whom we had met earlier, drew us aside as we were leaving the room and whispered: "Is this the famous Acoustat electrostatic? Does it Acoustat sound right to you?" We rest our case. X To get down to business, the Acoustat X is a fo ur-foot high dipole radiator consisting of Acoustat Corporation, 4020 North 29th A venue, Holly­ three long, narrow, side-by-side panels, slightly wood, FL 33020. Acoustat X fu ll-range direct-drive angled with respect to one another, the middle electrostatic sp eaker system, $1995 the pair (with built-in one a bit narrower than the outside ones. The power amplifiers ). Five-year warranty, excluding tubes; customer pays all fr eigh t. Tes ted #491 and #492, on loan panels are fu ll-range electrostatic transducers, fr om manufacturer. driven directly off the plates of the 6H B5 output tubes of a matching hybrid power amplifier, Once again, the fi rst piece of equipment in which screws into the base of the speaker frame our alphabetical listing poses a special problem. (or call it cabinet). Left and right channels are We approach this review of the Acoustat X with completely identical. On paper, the whole thing great trepidation . The subject seems to be is a purist's dream: all the advantages of elec­ booby-trapped fo r the reviewer. trostatic (i.e., fo rce-over-area) drive; no cross­ Why? Because, long after the completion overs, hence no phase and time-delay problems of our tests, it remains unclear to us whether fr om that source; no transformers in the signal our samples were absolutely typical. Objective­ path, as in most other electrostatics; no signal­ ly, we must conclude that they were indeed processing stages between the preamp output typical and that our findings are applicable and the ear that could be fu rther eliminated. across the board. Yet, there's a persistent And yet ... mystique to these speakers out there in audio­ All right, fi rst the good news. The Acoustat freak country, suggesting that maybe (j ust X does have the unmistakable midrange integ­ maybe) occasional super samples appear in cir­ rity of a crossoverless electrostatic design. As culation. The manufacturer professes to have long as the program material stays in a rela­ no knowledge of such variability. Let's marshal tively narrow fr equency and dynamic range the evidence: (like, say, a singer with acoustic guitar), the Last winter, we heard the Acoustat X at sound is quite natural and likable. That alone a nearby store and were quite favorably im­ may possibly account fo r all the favorable reac­ pressed, although we did notice some peculiar­ tions. (Including ours in the dealer's show­ ities. We briefly reported so in our very first room.) But feed the speaker some wide-range issue. In retrospect it seems that this was a symphonic material, with banging timpani, better sound than we were subsequently able to clashing cymbals and soaring strings, or some 21 heavily produced rock, and the naturalness is The speaker actually sizzles when the signal gone-even at quite reasonable volume levels. generator sweeps through this bump, and organ The speaker begins to sound strained and steely pedal notes or bass-drum whacks anywhere in on top, bumpy and muffl ed below. The free­ that vicinity overload the speaker to the point breathing ease and openness that characterize of offensive fl atulence. It's even possible that the reproduction of less complex material are the higher harmonics of this powerful resonance replaced by a disturbingly canned quality. Vig­ are responsible fo r that bumpy, muffled quality orously played strings sound "electronic" even we discerned in the upper bass and lower mid­ when they aren't massed, as in a trio or quartet. range on complex program material. In addi­ So does the piano. tion, both the speaker fr ame and the amplifier We believe we know the reason for this chassis buzz at various frequencies. It's rather behavior. Our tone-burst tests revealed sub­ a mess. stantial ringing at a number of fr equencies, Much has been made of the sensitivity of especially around I kHz but to some degree also the Acoustat X to room placement, of the around 4 kHz and several other spots. The touchiness of its high-frequency balance con­ energy peaks of complex, dynamic program trol and even of its amplifier gain control, of material seem to excite these problem areas the pros and cons of the built-in equalizer to the point where the ringing becomes dis­ network that counteracts the boundary effect tinctly audible and the sound quality suffers. of the rear wall (it can be disabled by padding The most plausible cause of the trouble would with a 470K-ohm resistor), of the desirability appear to be standing waves in the large panels, of raising the speakers off the fl oor, and so especially since the ringing is definitely lower in forth and so on. All we can tell you is: that's amplitude near the clamped edges. not where it's at. All these things make a dif­ I nterestingly enough, single pulses are also ference, but not the Big Difference. We experi­ very difficult to reproduce accurately through mented with them all, staying in close touch the Acoustat X. A crossoverless, fo rce-over­ with the manufacturer (who even sent us "new area system could be expected to have all the and improved" Ie's for the fr ont end of the advantages when it comes to this time-smear amplifier-they were indeed better). We ended test (cf. the Beveridge below), but the Acoustat up using the speakers exactly as the instruc­ panels fa il to render a good likeness of the tions specify: 2 to 3 feet out from a fairly live pulse shape and are altogether unable to repro­ wall, looking into the longer dimension of the duce certain pulse widths (such as, for example, room. That was by fa r the best arrangement, 0.35 msec). Whether this is symptomatic of the but they still didn't sound like reference same problem as the ringing or unrelated to it speakers for the purist. we aren't ready to assert; it was just about at It must be remembered, of course, that the this point that we ran out of R and 0 time for Acoustat X isn't really expensive-not when the Acoustat Corporation. In any event, the you consider that the power amp is included. Acoustat X lacks the ultimate definition, fo cus, That puts it almost in the same price category and delicacy of inner detail that characterize as the Dahlquist DQ- IO; certainly not a whole a speaker with outstanding pulse response. lot higher. On the other hand, there's always In the frequency domain the speaker is the possibility that it's the amplifier that's caus­ quite good, though not spectacular; overall ing some of the problems we observed. We power response is reasonably fl at; there are, doubt it, though . Since neither the speaker nor however, some very nasty lobes in the polar the amplifier alone is compatible with other response. As a matter of fact, vertical beaming equipment, the question is academic, and we is so severe that it's hopeless to listen to the didn't investigate it fu rther. speakers standing up: your ears must be ap­ Oh yes. We fo rgot. Both amplifiers even­ proximately level with the midpoint of the tually conked out in the course of our testing. panels, otherwise you don't hear the highs. We One in the middle of some loud music, the other could live with that, though, if the speaker had in the middle of a tone burst. Neither responded no worse fa ults. to elementary first aid. But by then our conclu­ One of these worse fa ults is in the deep sions were firm. And will remain so unless we're bass: a great big bump in the response at 35 Hz presented with totally new evidence to the that creates utter chaos whenever it's excited. contrary.

22 Source," fo r example, isn't and doesn't.) The Beveridge Model 2SW other is its incredibly good impulse response, indicating extremely low time-delay distortion, which audibly outweighs whatever minor de­ ficiencies the speaker may have in the frequency domain. A quick description of the design: A 6-by- l­ Harold Beveridge Inc .. 422 North Milpas Street. Santa foot electrostatic transducer is fr ont-loaded by Barbara. CA 93103. Beveridge Cylindrical Sound Sys­ a 6-foot high "acoustic lens" terminating in a tem . Model 2SW, $5200 the pair (with sub woofers . relatively narrow aperture. The rear radiation plug-in power amplifiers and control module). Virtually is blocked. The lens, which is in effect a system unlimited guarantee ("you must be absolutely satisfied"). of vertical guiding walls, bends the planar wave Tes ted #233 and #234. on loan fr om manufacturer. produced by the electrostatic panel into a 6-foot high 1800 cylindrical wave fr ont, invariant with frequency, in accordance with the geometry of a line source. The transducer, the lens, and the Let's say it right at the beginning. This baffle structure fo r the back wave are all housed speakerjamplif ier system, despite its undeni­ in a huge, upright, coffin-like structure that able limitations, comes closer to our ideal of plugs directly (through contact pins in its bot­ sonic accuracy than any other we've ever heard, tom) into a metal base that houses the amplifier. with the possible exception of Mark Levinson's The system is best set up with the apertures of stratospherically priced HQD System (which the left and right units facing each other across isn't really a single product but an ad hoc blend the room. The two subwoofers (large but man­ of fo ur diffe rent brands and which we haven't ageable) plug into their respective amplifiers tested yet under our own roof).Aft er that state­ via ordinary speaker wire and banana plugs. A ment has duly sunk in, we must add that some 30-foot shielded cable emerges fr om each am­ audio perfectionists will never settle for a sys­ plifier to be plugged into your preamp. A tem that can't play louder and has no more head­ separate control module is also supplied with room than the Beveridge. Not at this price. To the system, with provisions to trim the fre­ us, however, its specific virtues are irresistible. quency response within a few dB ("spectrum The SPLjheadroomjdynamic-range prob­ slope" and "bass environment") and to vary the lem was of course the very reason why Beveridge stereo spread ("lateral control"). We found that didn't stick with the original Model 2, which this control unit introduces an ever-so-slight was "purer" conceptually: a monolithic, cross­ (call it infinitesimal) degree of veiling, so we overless, fu ll-range electrostatic system, with unplugged it, especially since our listening­ matching direct-drive high-voltage amplifier (a room characteristics happened to suit the naked la Acoustat X but vastly more sophisticated, speakers very well. On the test bench, the con­ as we shall see). The Model 2SW, by incorpor­ trol unit gave no evidence of any type of distor­ ating a 70-Hz crossover (6 dB per octave and tion; indeed, it's very well designed. built right into the power :lmp) and letting a The sound of the Beveridge Model 2SW is conventional electrodynamic subwoofer handle such that, at a reasonable SPL and with superior the lower bass, can play considerably louder, program material, most listeners go "yechh" since the frequency division permits the electro­ when any other speaker is switched in for com­ static main unit to be optimized fo r the nar­ parison. Even the good ones in our survey. The rower range. Even so, fo rget about any SPL Beveridge is far superior in transparency, mid­ above 100 dB or thereabouts in a fair-sized range and (especially) top-end openness, free­ room. The Beveridge sounds so good when it dom from nasality and vowel-type colorations, doesn't have to strain that it's a crime to strain depth perspective, attack and release, and it. (Besides, the fu se will blow.) overall delineation of inner detail. Further­ Why does it sound so good? There are two more, that cylindrical wave fr ont really does main reasons, in our opinion. One is that it's what it's supposed to; uniform distribution of probably the only speaker that's a true line the direct-path "first-arrival" sound results in source and lays a sound field into the room absolutely unvarying balance and perspective accordingly. (The Infinity "Quantum Line anywhere in the room-near or fa r, left or

23 right, sitting or standing, or even right between Hz in the large (electrostatic) enclosure. It's a the speakers. Since the SPL from a line source fairly broad peak of about 8 dB amplitude and varies inversely with the distance, not the square you can hear it all the time if you listen carefully. of the distance as in the case of conventional speakers, the uniformity of the sound field is Editor 's Note: At press time we're told that even more startling. Stereo imaging is also ex­ Harold Beveridge has identified the problem as cellent and quite independent of listener posi­ due to insufficient damping material behind the tion, but be prepared fo r differences in compar­ electrostatic panel, has cured it completely (to ison with good point-source speakers, such as o dB ), and will make the modification in our the Rogers. The fact is that the exact comple­ review samples at the earliest opportunity. Not mentary playback geometry fo r a particular in time, however, for this already late issue. microphone setup is always conjectural and therefore highly arguable, so that the character­ The frequency response of the complete · istic wall-to-wall stereo image created by the system is otherwise very smooth to 12.5 kHz, Beveridge is as valid as any and at least never displays some slight perturbations in the 13-14 smeared by time delay. kHz region (probably due to standing waves be­ What all this adds up to is a strikingly tween the lens elements), peaks very gently at 16 original design concept executed with a high kHz and then dies quite fast. The farfield inte­ degree of competence. Not since the Lincoln grated response, however, smooths out these Walsh speaker has there been anything com­ minor, and to our ears inaudible, irregularities, parable in sheer creative thinking-and the so that the picture is one of a very gradual actual implementation of the Beveridge idea is downward slope fr om 12.5 to 20 kHz, without of course far superior (cf. the Ohm F below). significant peaks. Overall, we have no quarrel It's interesting to note that Harold Beveridge, with the frequency response of the Beveridge who is responsible for both the theory and the from, say, 100 Hz on up. practice, never had anything to do with the I n any event, it's the time response of the audio industry, having spent his entire engineer­ speaker that puts it in a class by itself. It repro­ ing career in high-level national defense jobs. duces single pulses of just about any reasonable It seems that hanging out with Bill Johnson or width with startling accuracy. And you don't Joe Grado or Dick Sequerra isn't the only way have to hunt for a "sweet spot" to read a good to learn about the fr ontiers of audio. pulse, as you do with even the best conventional We must now point out what's wrong with speakers; just point the measuring microphone the speaker. For one thing, the woofer is only in the general direction of the speaker-high, pretty good instead of great, which is what the low, sideways, over your shoulder while talking Beveridge electrostatic would deserve. The 12- on the telephone-and the pulse will be a good inch woofer in its sealed enclosure appears to replica of the generator output. Again we must have a system resonance of 36 Hz, since at 40 Hz say-that cylindrical wave fr ont sure does itsjob. its unequalized amplitude response is up 5 dB, One more thing. We haven't heard the indicating a Q of 1.7. The response drops back original, simon-pure Beveridge Model 2, so we to the 0 dB level at 30 Hz. This is not the way can't comment on the similarities and differ­ to extend response down to 30 Hz, in our ences. According to Harold Beveridge, how­ opinion; in fact, it doesn't sound like 30 Hz ever, the overall gain in headroom is almost response but rather like bumped-up 40 Hz re­ 10 dB with the Model 2SW, which would totally sponse, which is of course what it really is. For disqualify its predecessor in our estimation, example, the open E string of the double bass, since even the Model 2SW is just adequate in vigorously plucked, makes the woofer go that respect. I n fa ct the I-think-I-just-heard-it­ "boom" instead of "thunk"-and that's almost clip crowd will fi nd it rather fr ustrating, as exactly 40 Hz. It wouldn't happen with a Q of we've already said. 0.707. Harmonic distortion is again pretty low To us, the speaker remains a joy. It's the but not fa ntastic; the original Acoustic Research only one in this survey that can occasionally woofer of 1952 was better in that respect. (Ed make us forget that we're listening to repro­ Villchur, where are you now that we really duced music. Now and then, it almost sounds need you?) live. And that makes it our reference speaker, There's also quite a nasty resonance at 76 until somebody sends us something better.

24 Braun 'Output C' Braun L200

Adcom . 114 East 32nd Street. New York . NY 10016. Adcom. 114 Eas t 32nd Street. New York. NY 10016. Braun 'Output C miniature speaker system. $224 the Bra un L200 small bookshelf speaker system. $260 the pair. Tes ted samples on loan fr om distributor. pair. Tes ted samples on loan fr om distributor.

These neat little Braun speakers are some­ Imagine a speaker system no bigger than what anticlimactic after the Output C's; they're one of the heftiertwin-lens reflex cameras and about twice as large (i.e., still exceedingly in black metal, too. Imagine it with sound that's small) but not much better sonically, so the quite respectable for any speaker, regardless of amazement factor is considerably smaller. size: just a notch or two short of top-notch and The differences: black plastic rather than minus the deep bass. You now have a fairly metal enclosure, larger woofer of course, same good picture of the Braun Output C. Amazing type of dome tweeter but not quite the same and lots of fun. diaphragm material. The system resonance is We wouldn't hesitate to recommend at 175 Hz instead of 200, bumped up the same Braun's small wonder to anyone who is really way (the Q is approximately 1.5). The tweeter short of space and doesn't even have room for still rings, pulse reproduction still deteriorates anything like the little Rogers LS3/5A (which as one goes from wider pulses to narrower ones, is far superior but about three times as big and but there's some improvement in these areas. twice as expensive). The Output C is also the The overall sound is perhaps somewhat rounder perfect gift "for the audiophile who has every­ and more authoritative, but not very different. thing." A great toy. The slight coloration is still there. Don't let anyone tell you, though, that If it were possible to avoid comparing the these diminutive speakers plus a commode-type L200 with its little brother, we'd be impressed; subwoofer with summed channels constitute an under the circumstances, it's the Output C that "invisible" system approaching SOT A quality, stands out as the more remarkable product. as has been suggested in some quarters. No way. The Output C has some definite color­ ations even in the range where its tiny woofer and dome tweeter are reasonably fl at. These colorations, however, aren't particularly irri­ tating musically; the overall sound of a pair of Canton LE 400 Output C's used as a full-range stereo system is open, smooth and listenable beyond all expec­ Adcom . 114 Eas t 32nd Street. New York . NY 10016. tations. Even the bass is subjectively accept­ Can ton LE 400 bookshelfspeaker svstem , $350 the pair. able, although it's mostly fa ke of course, Five-year warrantl'; manufacturer pavs return fr eigh t. achieved with bumped-up response at the sys­ Tes ted #466872 and #466888, on loan fr om distributor. tem resonance of 200 Hz rather than with genuine low-frequency output. Through an alphabetical coincidence, the The most noticeable coloration is due to three German sealed-box minisystems in this the tweeter, which rings rather badly and survey follow one another without interruption. doesn't allow the system to reproduce pulses The Canton LE 400 is by far the largest: about accurately, especially as the pulses get nar­ half a cubic foot in volume. You'd expect it to rower. The result is just the slightest zippiness have better bass than the two little Brauns and and a lack of the utter transparency you can it does; the system resonance is at 110 Hz, where expect with superior impulse response (cf. the the response is up 6 dB, meaning that the Q is Rogers). approximately 2, which is definitely on the Even so, we have nothing but admiration loosey-goosey side. The corner frequency of the for the Braun Output C. The very fact of its bass roll-off is 74 Hz, as near as we can tell. existence is a sign of progress in audio engi­ The manufacturer claims response down to 35 neermg. Hz, which of course is nonsense.

25 There's actually a three-way system shoe­ the flick of a switch on the front panel, you horned into that tight little box, with cross­ can have a Q of 0.6 for dead-flat bass that overs at 750 Hz and 2600 Hz. Frequency re­ drops to -4 dB at resonance, or you can choose sponse is extremely fl at from about 700 Hz on a Q of I (j ust a little looser but still well­ up. The star performer is the tweeter, which has controlled) fo r 0 dB response at resonance and astonishingly fl at re sponse on axis to well the slightest ripple just above. These people above 20 kHz and even at 45° off axis goes out know their P's and Q's; they also know better past 15 kHz. That's superb amplitude re­ than to try to squeeze more out of a given sponse and dispersion in any league. It's also woofer or box size than Mother Nature per­ the most likely reason why the speaker system mits, As a result their bass is close to state­ sounds impressively open and "present" on first . of-the-art within the physical limitations of the hearing. system-solid, well-defined, musical. It's so We say fi rst hearing because, after a while, easy when you do your homework. the LE 400 begins to sound overbright, "elec­ The most highly touted part of the speaker tronic" and fatiguing. The culprit is again is the crossover network, which is made to see ringing, easily identified by tone bursts, a resistive rather than a reactive load by means especially at 2 kHz and 4 kHz. Pulse reproduc­ of additional impedance compensating net­ tion is also surprisingly poor. works across each driver. Cizek claims that I f you have a chance, though, listen to there's no other way to achieve absolutely flat this speaker. It's the least expensive we've ever response in the crossover region, which is a heard that, if only fo r a minute or two, has the sensitive part of the spectrum in this case, the very special sound of the top systems with fast nominal crossover frequency being 1.5 kHz. An­ tweeters. Then it's all over, but we're willing to other sophisticated touch is the combination of bet that the sale is often made by then . Cute two high-frequency controls on the front panel little rascal. (right next to the Q control), one fo r contour, the other fo r overall level. Their net effect is rather subtle and, in our opinion, something of a red herring in view of more serious matters in the high-frequency area that require our attention, The least of these is a 5 dB peak at Cizek Model #1 16 kHz with a precipitous drop thereafter, which is what we blamed initially for imparting a somewhat irritating zing to the otherwise smooth and classy sound of the Cizek . That Cizek Audio Systems, Inc., 15 Stevens Street, Andover, MA 01810. Model #1 acoustic-suspension loudspeaker, wasn't it, though. It turned out to be much $396 the pair. Five-year warranty; manufacturer pays heavier stuff. all fr eigh t. Tested #2054 and #2065, on loan fr om Exploring the speaker with tone bur sts, manufacturer. we came upon an astonishing amount of ring­ ing at 3 kHz. It seems that the tweeter repro­ duces the burst quite accurately, after which This is one of the medium-sized units in the much slower woofer, still quite active at our survey: a solidly built acoustic-suspension that fr equency, chimes in with another com­ system of the largest bookshelf type (almost plete burst of almost the same amplitude. Two too large fo r a bookshelf), with 10-inch woofer bursts fo r the price of one: you feed in, say, and I-inch dome tweeter. Despite its conven­ a four-cycle burst and out comes an eight-cyCle tional fo rmat, it comes with a special audio­ burst with just the tiniest cleft in the middle. phile reputation, since the company is techni­ Needless to say, at 3 kHz, which is right where cally oriented and the speaker is more sophisti­ the ear tolerates absolutely no nonsense, this cated in engineering as well as in audible spurious doubling of energy sounds like exces­ performance than such speakers generally are. sive brightness or glare. The ringing continues I ts best feature is its bass response; the higher and lower, too, but 3 kHz is its head­ system resonance is at 39 Hz (the specs say quarters. 38 Hz, which is accurate enough fo r us), and Our pulse tests confirmed the lack of syn­ the damping is absolutely correct. In fact, at chronism between woofer and tweeter to be the

26 speaker's chief weakness. The tweeter is much ation with the smaller enclosure volume this still too fa st fo r the woofer-or, if you prefer, the keeps the system resonance at the same fr e­ woofer is too slow for the tweeter. First the quency, namely 39 Hz (or 38 Hz, according to tweeter blips, then the woofer; you can actually Cizek). In fact, the Model #1 and Model #2 hear the double blip, unless you put the woofer have virtually identical bass response profiles; so much closer to your ear that the speaker, the two-position Q adjustment is also retained laid on its side, faces inward instead of out on Model #2, except that the obtainable values toward the listening area. The microphone, too, are 0.57 and 0.9, certainly not a major diffe r­ reads the best pulse from this extreme off-axis ence. The tweeter and crossover network ap­ position. A time-compensating network could pear to be the same, but only a high-frequency probably do wonders fo r the Cizek. level control is provided, without the contour Where does all this leave us as regards control. an overall evaluation? We still feel that the Both models sound extremely similar. On Cizek is one of the very best box speakers any­ organ music and other complex program ma­ where near this price. Not because it's without terial that's rich in bass energy, the Model #2 faults but rather because none of the others poops out much more readily; it just can't we're aware of have acceptable full-range per­ handle power the same way and begins to dis­ formance. The Rogers, fo r example, sounds tort. On the other hand, it has slightly less incomparably better but can't handle any zing and glare at the higher frequencies-and power and lacks almost two octaves of the there's a reason. The peculiar multiplication Cizek's bass range. It also costs more, at least of tone-burst cycles occurs at around 2 kHz on in the U.S.A. The DCM Time Window is also Model #2, which is less irritating than 3 kHz significantly better but can hardly be called ringing. Furthermore, the amplitude of ringing a box speaker and is priced very much higher. is quite a bit smaller. The 16 kHz peak is still And so on. there, though, and so is the differentiation of Maybe we shouldn't have reviewed the pulses everywhere except way off axis toward Cizek in such fast company. Still, we feel that the woofer side. It's basically the same speaker. the bass alone is worth the price of admission And that makes it pretty good value in our and that the basic design is not only intelligent book. Purists, on the other hand, need not apply. but also capable of evolutionary improvement.

* * *

Editor 's Note: The latest Dayton Wright full­ range electrostatic speaker, announced for this #2 survey on the back page of our last issue, should Cizek Model have fo llowed here in alphabetical sequence. Unfortunately, it wasn't sent to us in time for testing, despite a firm commitment from the company. We now have another firm commit­ Cizek Audio Sys tems, Inc., 15 Stevens Street, Andover, ment from them to make the speaker available MA 01810. Model #2 acoustic-suspension loudspeaker, in time for Part II of the survey in our next $268 the pair. Five-year warranty; manufa cturer pays all fr eight. Tes ted #1130 and #1139, on loan fr om issue. If they don't come through, we'll try to manufacturer. obtain a pair anyway, one way or another. Meanwhile, on the basis of admittedly super­ ficial exposure to the Dayton Wright in four Speaking of evolution, Cizek has already different places at various times, we're inclined come up with a good example: their Model #2 to doubt that it would challenge the Beveridge is a considerably smaller and less expensive as our top choice. But we're perfectly ready and version of the Model #1 while still offering the willing to change our mind should our tests turn same basic sound quality. out to be contraindicative. The main difference is the substitution of an 8-inch woofer fo r the lO-incher; in combin- * * *

27 ' than our test results indicate, the brochure DCM 'Time Window IS an eye-opener. The physical configuration of the Time Window is quite unconventional. Imagine a waist-high cylinder standing on the floor, ex­ DCM Corporation, 725 S. Division, Ann Arbor, MI cept that only the back of it is cylindrical; the 48104. 'Time Window' floor-standing loudspeaker, $660 the pair. Tes ted #906 and #932, on loan fr om part facing you, the listener, is a triangular manufacturer. prism, edge foremost. On the two angled faces of the prism are the drivers, symmetrically Here's the one that will recoup you for placed: two 6-inch Philips woofers and two about 20 years' subscription to The Audio Philips dome tweeters below them (that's right, Critic: a speaker system that has no equal even below the woofers, not above them). Near the at twice the price (at least none known to us). In floor are two ducted ports, also symmetrically this Part I of our survey, we place only the fol­ placed. The cabinet is surprisingly light; you lowing speakers ahead of the DCM Time Win­ can carry it from one end of the room to the dow: the Beveridge Model 2SW, the Rogers other without requiring abdominal surgery LS3/5A (but only fr om, say, 120 Hz on up at afterwards, yet it's quite rigid. Weird but ef­ moderate levels-or else with expensive sub­ fe ctive design. woofers), and the Koss Model One/ A (but only The proof of design is of course in the lis­ every other day because in many ways we prefer tening, and what a nice-sounding speaker it is. the Time Window). That's all. The Dahlquist The first thing that strikes you is truly con­ DQ-IO, for example, which was our reference vincing spaciousness and depth; when a singing speaker before we started the survey, ranks be­ chorus walks in fr om offstage in an opera re­ low the Time Window in our current hierarchy. cording, you can almost tick off the yardage as There's some consistency to the above: they approach. Only the Beveridge is in the every speaker that, to our ears, equals or sur­ same league in this respect; the Time Window passes the Time Window has outstanding re­ beats the Koss and all the others. Transparency sponse in the time domain. The Beveridge and delineation of inner detail are of a high reproduces pulses most accurately; the Rogers order, but not quite up there with the Bev­ is next; the Time Window and the Koss are eridge or even the Rogers; the Koss, too, sounds close behind the Rogers and about equally more open and delicately etched in the upper good, but with quite different deviations from octaves but not throughout its range. Left/right perfection. How about that? You'd swear that imaging isn't quite as spectacular as with, we measured impulse response fi rst and then say, the Pryamid 'Metronome', but we find it picked our favorites, but you have our word of musically satisfactory. In fact, the entire honor that our preferences were fi rmly estab­ speaker sounds unfailingly musical at all times lished prior to any laboratory tests, strictly and remains listenable regardless of your length on the basis of listening. We have a feeling that, of exposure to it. in the not too distant future, speaker design­ That doesn't mean, however, a complete · ers will be investigating impulse response as absence of colorations or other audible anom­ routinely as frequency response. alies. Far from it. The bass is a little fu nny The DCM people are among the most fer­ for one thing; more about that in a moment. vent advocates of pulse testing; the very name The top end could be a little more extended and of their speaker, Time Window, shows where silky-smooth; that last touch of HF quality their head is at. Their literature makes the seems to be unattainable with the Philips dome, most convincing case we've seen for the over­ even though it's better than most and handles whelming importance of the time domain; if power very well. There's some raggedness in you discount its quite mild and inoffensive frequency response throughout the audio range, commercialism, you can learn more from it which may be the reason why the speaker is than from just about any other popular explan­ particularly sensitive to room placement; ation of speaker design. It shows comparative you've got to make sure that the room reflec­ pulse tests on fi fteen speakers, including the tions zig where the speaker zags. It isn't any­ Time Window, and even if they manage to thing serious (minor fr equency-response prob­ make their own product look a little better lems seldom are), except in the bass and the

28 lower midrange, where things get a bit thick To sum up, the DCM Time Window is one and muddled from time to time. Carefully of our happiest discoveries since we've started tuning the distance of the speakers from the testing audio equipment, and we recommend it back wall as well as the side walls will clear wholeheartedly to any music lover who isn't up the problem; raising the speakers a fo ot or planning to spend thousands of dollars on his so off the floor also helps. When everything is speaker/ amplifier setup. It makes sense, how­ trimmed in, the lows are quite clean and tight, ever, to get a very good amplifier to drive the but don't expect stupendous pressure bass on Time Window; not necessarily a very high­ organ pedals, bass drum or bull fiddle. If we powered one, since the speaker is quite efficient, were pressed to designate a nominal corner but a really clean one because superior time­ frequency fo r the Time Window's rather domain resolution will show up the difference. bumpy composite bass roll-off, we'd place it Our congratulations to Steve Eberbach, at 50 Hz. Maybe even a little higher. After the young engineering partner of DCM, for all, it's not a large box. what may very well turn out to be a classic. We say "composite" because the bass re­ sponse of a vented box is the sum of the out­ puts produced by the woofer and the vent. Ideally, these should be complementary, arith­ metically adding up to fl at response. In the Fundamental Research case of the DCM Time Window they aren't. Each woofer exhibits the classic vented-box response you'd expect, but the vents by them­ Fundamental Research, 1304 Success Street, Pittsburgh, PA 15212. 'The Low Frequencies ' sub woofer, $450. selves have perfectly flat output down to Tes ted samples on loan fr om manufacturer. about 18 Hz instead of fi lling in for the woofers with a humped response where the woofers This is the best subwoofer we've tested so drop out. Somebody obviously thought that this fa r, which by itself isn't a particularly great was a very good thing, but of course it doesn't distinction, since the Janis W- 1 and W-2 are add up to flat bass response out of the total badly flawed designs and the Dahlquist DQ-1 W system. The DCM design philosophy, as we doesn't go low enough. The Fundamental Re­ understand it, is that in order to generate search is a good unit, however, in any company; flat response within four walls, a speaker should that's why we didn't hesitate to include it in this never be designed for flatrespo nse anechoically survey. And it happens to mate very nicely with (i.e., with nearfield measurement). We disagree the Rogers LS3/ 5A (more about that in the re­ with this philosophy quite vigorously and feel view of the latter below). that it may have a great deal to do with the The Fundamental Research is a waist-high temperamental behavior of the Time Window floor-standing subwoofer incorporating a 12- as regards room placement. We don't want to inch driver in a sealed box with an approximate make too much of the whole thing, on the other internal volume of 3 to 3 '12 feet. It's perfectly hand, since the speaker is still a lot better with classic and straightforward in design except fo r this flaw than others are without it. a weird little slotted board that's nailed right The reason why it's better-why it sounds on the front baffle across the face of the driver. better-is that (a) it reproduces pulses very This doesn't provide any low-frequency loading accurately, with only minor glitches, and (b) it but is there ostensibly to dissipate the higher exhibits little or no ringing on tone bursts. In frequencies-a very crude sort of termination. other words, it doesn't smear the signal. (One It's responsible for the one serious flaw in the of those minor glitches worth noting is a little response of the woofer: a peak at around 450 negative blip out of the tweeter just before the Hz that rises 7 dB above the flat portion of the leading edge of a positive-going 0.1 msec im­ curve, followed by considerable raggedness pulse. We mention it only because it's very (huge suckout at ca. 1 kHz, double hump back neatly fudged in the DCM literature; visible up to the 0 dB line at 1.4 and 2 kHz, ragged but cleverly buried. You can also tell from their roll-off thereafter). Ideally, a woofer should pictures that the various electrostatics don't roll off smoothly above a given frequency have this problem, thus confirming each of our (450 Hz would be perfectly all right), so that own fi ndings. Naughty, naughty.) the crossover network can do its thing without

29 running into unpleasant surprises. (Remember the 14 dB peak of the Janis W -I at 460 Hz?) Infinity QLS We must admit, however, that we could detect no audible anomalies above the crossover point when biamping the Fundamental Research Infinity Sys tems, Inc. , 7930 Deering A venue, Ca noga either through the Rappaport PBC- l passive Park, CA 91304. Quantum Line Source speaker system, Dahlquist DQ-LP 1 active/ $2500 the pair. Five-year warranty; customer pays all crossover or the fr eigh t. Tested #7001070 (left ) and #7001071 (right), on passive crossover, with the nominal crossover loan fr om manufa cturer. frequency in the 100 to 120 Hz region. It ap­ pears that the goof has to be of Janis propor­ tions before it can be heard. In our first issue, we published a rather That said, we give this subwoofer a high negative preview of the Infinity Quantum Line mark for its bass response profile. Down to the Source, based on superficial exposure to it in a roll-off point it's dead flat; the -3 dB fre­ dealer's showroom. After thorough testing un­ quency is 38 Hz; from there on the response der our own roof, we have somewhat more re­ drops to -S dB at 30 Hz, -8 dB at 20 Hz, and spect for this speaker, but by no stretch of the -14 dB at 10 Hz. This is a slightly overdamped imagination and with no amount of good will profile (the Q appears to be a little lower than could we call it State Of The Art, Reference 0.707), so that there's probably room for Standard, or any other such name dear to the fu rther improvement (i.e., a lower corner fre­ hearts of audiophiles . quency without the penalty of overhang) with To put things into perspective, if our fi rst some fi ne-tuning of the parameters. If you've exposure to the QLS, as inserted into our read our subwoofer article in the second issue, reference system, had been with a blindfold and if we now tell you that the Fundamental on, we would have said, "That's a fairly decent­ Research handles power quite well, you need sounding speaker system; what is it?" On being no fu rther information to know exactly what told that it's Infinity's current flagship, succes­ it sounds like. The bass is deep, solid, well­ sor to the fabled Servo-Statik and $12S0 per controlled-it couldn't be anything else. side, we would have said, "You must be kid­ Mother Nature practices no deceptions. (Or, as ding!" When properly set up and balanced out one of our consultants is fo nd of saying, the (which takes some doing), the QLS reproduces spectrum analyzer is too dumb to lie.) the fu ll audible range of frequencies with a Overall, we prefer the Fundamental Re­ certain degree of authority and musical felicity. search to the Dahlquist DQ- I W, even though But it never reveals sufficient inner detail (the it's much less efficient and therefore requires pleasure derived is more in the area of maple a bigger amplifier to drive it. But it holds up syrup than of crystals); makes the stereo image much better in the bottom octave and at the shift and drift depending on frequency; woofs same time it's just as tightly controlled. Not up and occasionally even burps when excited quite as well built, though; it has a few cabinet by really fierce bass transients; in short, it buzzes, seldom excited even by loud music, that just doesn't sound like one of the more accurate we'd rather do without. A little more wood or a systems money can buy. little more glue seems to be called for. The reason fo r some of these shortcomings We've left the best fo r the last. This sub­ is obvious as soon as you remove the grille woofer was designed purely by ear by a dedi­ from the fr ont panel of this SI/2 -foot high, IS0- cated young man named Mike Zelenak. The pound monument. Have you ever seen or heard chances that it would come out this success­ of a speaker with 17 (count them) drivers that fully were slim, to say the least, but there it is was an accurate reproducer? We haven't. It's -you can 't argue with a historical fact. The impossible to syn chronize that many separate moral is that, if your ears (and taste) are as sources of radiation to produce a single, un­ good as Mike's, it's still possible in audio today confused image. Infinity seems to believe that to fl y by the seat of your pants. Oth erwise, by stringing out eight tweeters and six midrange especially if you'r e into woofers, stick with domes in two parallel vertical lines they've good old Thiele and Small. We will. But then created a Line Source (capital initials theirs). we also believe in keeping road maps in our Anyone who has read our Beveridge review glove compartment. above has some idea of what a line source really

30 is and does. Where's the cylindrical wave front itself, with the adjacent midrange source block­ that's invariant with frequency? Where's the ed out. Now all that Infinity needs to do is de­ time coherence? We fo und that the QLS re­ velop an EMIT that goes all the way down to produces pulses quite poorly, lacking even an 20 Hz! acceptable "sweet spot," which is totally con­ The midrange dome used in the QLS is, tradictory to the definition of a true line source. on the other hand, a real turkey; it has, What's more, the responsibility for balanc­ among other things, a huge peak at 2.2 ing out those 17 drivers is left entirely to the kHz, which is probably one reason why user. There are three continuously variable the QLS doesn'thave a more transparent upper level controls in the back (midbass, upper mid­ midrange. As an interesting sidelight, this very range, tweeters) with a rather wide latitude of same Peerless driver (Peerless is owned by the adjustment; in addition, the tweeters can be same parent company as Infinity) is available brought in at three different crossover points by to all comers in the industry without the little means of a plug-and-socket arrangement. As­ pinholes in the dome that Infinity insists on ,suming conservatively that the continuously having in the QLS version. To us the latter variable controls have only four audibly dis­ looked from the beginning like a cute little tinguishable positions each (say, all the way Helmholtz resonator-and sure enough, it down, all the way up, and two in between) the resonates. We've heard it rumored that the total number of different-sounding response Peerless people themselves don't quite under­ profiles possible is 192. The speaker designer stand why Infinity wants holes in the dome. is saying, in effect, "I don't know how the hell Peerless, incidentally, also supplies the single it should sound. You tell me, baby." There's 5-inch midbass (200 to 600 Hz) driver for the a point where flexibility becomes anarchy. We Q LS, and it just so happens that this is the other tried our very best, but we can't guarantee that slow member of the family (the wages of in­ further knob-twiddling wouldn't have resulted cest?), ringing like crazy at 200 to 250 Hz and in slightly better sound in our room. contributing a retarded me-too response to 0. 1 Even with the help of a spectrum analyzer msec pulses, which are pretty far outside its as­ and a calibrated microphone, .the overall fre­ signed range. Talk about time smear ... quency balance (i.e., farfield amplitude re­ Which brings us to the woofer-the same sponse) of the QLS is extremely difficult to Watkins woofer we referred to in our preview determine on account of those 17 drivers strung as well as in our CES report. This isn't a unit out all over the place. All we were able to do to be dismissed quite as peremptorily as we was to analyze the tweeters, the midrange originally did, even though the uncomplimen­ domes, the "midbass coupler" and the woofer tary remarks about it that we attributed to a separately. distinguished American professor and electro­ We must admit that we were very im­ acoustician are confirmed to have been uttered pressed by Infinity's samarium-cobalt Electro­ substantially as quoted. Another heavyweight magnetic Induction Tweeter (EMIT), a fo rce­ academician, the polymath Dr. Richard H. over-area transducer claimed to possess all the Small of the University of Sydney, Australia, advantages of electrostatics without their draw­ takes a somewhat kinder view of the woofer, backs. We found no evidence to contradict that as revealed in a letter to William H. Watkins, claim; the EMIT appears to be extremely the inventor, a copy of which was made avail­ . smooth and flat out to 23 kHz, with wide hori­ able to us by Infinity. More about that in a zontal dispersion. We haven't seen or heard moment. Our own measurements indicate a better tweeter, and that includes the rather dead-flat response down to the lowest reaches similarly configured electrostatic tweeter in the of the audio range; the -3 dB point is 22 Hz! Koss Model One/A. In fact, we got the most No other woofer known to us comes even close musical sound out of the QLS by selecting the to this kind of penetration of the bottom octave, lowest available tweeter crossover point and except the Janis W -1; and, as in the case of assigning the widest possible range of frequen­ the Janis, we must point out that the audible cies to the EMIT. That way the highs remained performance doesn't live up to the promise of silky and sweet no matter what we fed into the the response curve. The low bass is quite beau­ speaker; furthermore, the EMIT reproduced tiful, detailed and controlled as long as the pulses accurately as long as it was tested all by energy level remains moderate; however, when

31 the woofer is called upon to move a lot of air fo r the InfinityjWatkinsjSma ll issue. (as in the case of one of our master tapes on A word about the biamping option on the which two double basses are plucked fo rtis­ back plate of the QLS. If you connect separate simo), ' some rather rude eructations ensue, power amplifiers to the woofer and to the rest even with the best and most powerful ampli­ of the system, without using an electronic or fi ers available. This doesn't surprise us, since passive crossover ahead of the amplifiers, you a 12-inch woofer, Watkins or no, just doesn't gain very little in performance. Each amplifier have the piston area and linear excursion capa­ channel will still see the fu ll frequency range bility to fi nd happiness in a 7 or 8-cubic-foot of the signal at its input and output, except that sealed box as a 22-Hz system-not at high part of the output signal will be burned up in a SPL, anyway. Either the box should be smaller network instead of being converted into sound. and the bass roll-off higher (cf. the Funda­ For reduced 1M distortion and increased head­ mental Research above), or, the driver should room, you need the additional high-impedance be at least a 15-incher. As is, the system is crossover. Of course, the fact that the separate really reaching when it digs into that bottom terminals are there is a convenience. In any octave. event, biamped or not, the QLS needs a lot of Even Dr. Small, of whose letter Infinity watts to drive it, mainly on account of that seems to be inordinately proud, points out that power-hungry woofer. the Watkins woofer has no large-signal ad­ To sum up, then, the Infinity QLS gives vantage over a conventional design. When it you a big, juicy, quite pleasant and listenable comes to moving air in a room, the cone alone sound, smooth and extended on top but far must do all the work, and the limiting factors fr om transparent or analytical overall, with are cone size and linear throw. What Dr. Small time smear quite evident and no stable image. likes about the Watkins invention (a two-voice­ All pretty much confirmed by measurement. coil configuration that eliminates the conven­ Among the speakers we'd rather listen to are tional hump in electrical imped ance at the not only costlier ones like the Beveridge or resonant frequency) is that under small-signal competitors like the Koss, but also consider­ analysis it appears to provide an increase in out­ ably lower-priced ones like the Snell, Dahlquist, put of approximately 2 dB, all other things re­ DCM Time Window and Rogers LS3j5A. maining equal-a theoretical fr ee lunch that appeals to the academic mind. The penalty is that the two-voice-coil motor sucks current from the typical constant-voltage amplifier like there was no tomorrow, and the extra juice had Koss Model One/ A better be available if you want to enjoy that 2 dB advantage. (Our Bryston 4B gave us no problem.) The amusing thing is that Dr. Small, a generous man who knows ten times more about speakers than Watkins, Infinity and Koss Corporation, 4129 North Port Washington Avenue, Milwaukee, WI 53212. Model OnejA full-range electro­ The Audio Critic put together, peppers his mild­ static speaker, $2 100 the pair. Tes ted early production ly condescending praise of the design with quite samples, on loan fr om manufacturer. a few sly professorial knuckle raps, which make his letter something less than a show-off piece Until the arrival of our Beveridge Model encased in pla�tic fo r Dad's wallet. Certainly 2SW, the revised Koss electrostatic was more his exegesis of the woofer is at considerable or less our reference speaker in the course of variance with Infinity's simplistic technical this survey. We say more or less, as we were brochure on the subject, which treats the low­ never completely sold on it; however, in com­ frequency impedance peak of a speaker system parison with all the other speakers (except pos­ as some kind of foul disease for which a cure sibly the Rogers, which is a very special case), has fi nally been found. We have personaliy it just sounded more open on top, clearer, heard Neville Thiele, Dr. Small's original more analytical, less smeared. In other words, mentor, call that point of view "rubbish." An like a good fu ll-range electrostatic. At the same impedance peak and an amplitude response time it had a disturbingly closed-down, thick, peak are two very different things. So much gag-in-the-mouth quality somewhere between

32 the bass and the lower midrange-we couldn't panels appear to be truly superior transducers, decide exactly where. We experimented with except for a 6 dB peak at 15 kHz when the room placement, biamping, taking off the tweeter is in the head-on (axial) position. A back, you name it. Our lab tests finally re­ number of listeners complained about an ir­ vealed what we now believe to be the cause, ritating quality on the top end when the speak­ but it wasn't something that could be helped ers were played broadside; this may very well without major surgery. We'll come back to that have been the cause-and one more reason to in a moment. angle the speakers inward if space permits. To understand the Koss Model One/A, On the bottom we found the useful response you must realize that it represents just the op­ of the One/ A to extend smoothly to approxi­ posite approach to fu ll-range electrostatic de­ mately 40 Hz, rather than 32 Hz as claimed sign from the Acoustat, the Beveridge or even by Koss, but then a dipole is notoriously hard the Dayton Wright. It's much closer to the to measure at the lower frequencies. (As a mat­ Quad, except for its fo ur-foot height. The audio ter of fact, there exists no proven mathematical range is divided into four passbands and spe­ model fo r low-frequency propagation by a cialized electrostatic elements are designed for large, free-standing, planar transducer.) The each, to be crossed over at 6 dB per octave. subjective quality of the bass is by and large There are fo ur large bass panels, one mid­ excellent: deep, firm and authoritative on range panel (about the same size), one narrow program material that contains genuine LF in­ treble panel, and an even narrower tweeter formation. Except ... panel. They're driven via transformers by any That peculiar thickness or muffled quality amplifier of your choice. The main difference kept bothering us, without much of a clue as to between the original Model One and the One/ A where to look for it. We finally nailed it-we is that in the latter the treble and tweeter think. At exactly 50 Hz, the entire fr ame of the panels are placed next to the midrange panel speaker system takes off, creating a completely (at the inward edge of each speaker, to form uncontrolled resonance and inducing the panels a mirror-imaged L/R pair), instead of partially to resonate as well. The resonant band is quite covering up the midrange panel with dis­ narrow (high Q), so that it isn't obviously ex­ astrous effect on the sound. Either way, there's cited by typical program material, making it no serious attempt to time-align the panels; somewhat elusive without careful sweep testing. basically they're just fastened into the fr ame Once found, the amplitude of the resonance to fill out the available space. We suspect that turns out to be somethin' else: 8 dB fo r the in­ this is one of the fundamental design limitations tegrated response of the four bass panels, 20 of the Koss, since the superior time-domain dB at the surface of the individual panels. characteristics of each individual panel don't One bass panel we measured also resonated add up to coherent time response by the total at 57 to 59 Hz; this may have been a problem system, except over a very narrow angle, way within the panel itself, but the overall 50 Hz off toward the outside edge. The speakers resonance appeared to be definitely a fr ame would have to be angled inward to take ad­ problem. With two strong men squeezing the vantage of this, not just a little but to an fr ame together from both sides, the center extent that wouldn't be tolerated in most fr equency of the resonance was shifted by a fe w homes. Too bad; the pulses are truly excellent Hz without curing the problem. In our opinion, from that angle but deteriorate badly as the it would take a complete redesign, with much head-on position is approached. We aren't sug­ heavier bracing and generally sturdier con­ gesting that this isn't preferable to time smear struction, to eliminate the resonance. And we're that .originates in the transducers themselves; pretty well convinced that the coloration we the One/ A has audibly better inner detail than heard was caused by transient effects at the 100 nearly all other speakers, but the design con­ and 150 Hz harmonics of the 50 Hz fu nda­ cept isn't taken all the way to the limits of mental. Without this defect we'd be inclined its potential. to rate the Koss Model One/ A just a hair be­ In the frequency domain the speaker is low the Beveridge Model 2SW, since it handles also very impressive; it's possible to find a power a lot more comfortably than the latter "sweet spot" where the response from 1 to 17 even if it doesn't quite have the same super kHz is simply dead flat. The treble and tweeter transparency. The way it was, many listeners

33 preferred not only the Beveridge but even the stockholder. His involvement in the company DCM, which sounded somehow more spacious, was strictly as a Director and nonresident con­ open and natural in the all-important lower sultant, without any day-to-day participation in midrange. management or engineering. In March 1974 he The One/ A also has separate terminals sold all of his stock back to the company in the back, just like the Infinity QLS, to allow and has currently no closer ties to Ohm Acous­ biamping the bass panels and the rest of the tics (or any other audio manufacturer or re­ system separately if you so desire. The diffe r­ tailer) than to the Exxon Company or the ence is that a convenient toggle switch is pro­ government of Liechtenstein. He does confess, vided for instant change from the single-amp however, to continued partiality to the brilliant, to the biamped mode. That made it very easy for if incomplete, loudspeaker theory of the late us to determine that the difference (with one Lincoln Walsh. Bryston 4B vs. two Bryston 4B's) was barely perceptible without an electronic crossover. The Ohm F is the great paradox of this The biamped mode was declared very slightly speaker survey. Its ability to rep roduce single more open and transparent by all listeners pulses of different widths (in our tests between after considerable agonizing. All our other re­ 1 msec and 0. 1 msec) could conceivably be marks on the same subject in the Infinity QLS judged seco nd to none, perhaps not even to review above apply equally here. the Beveridge. Its frequency response is also We recommend the Koss Model One/A, quite acceptable, except for a few anomalies then, to all those who can't live without the noted below. Yet its sound is more highly clarity, inner detail and generally unsmeared colored ("canned" may be the better word) and sound of a full-range electrostatic, but at the less musical than that of any other speaker in same time won't spend the extra kilobucks to this admittedly fo rmidable group. Whenever go all the way to the Beveridge or to something we switched to it in the middle of a listening even mor e elaborate like the HQD System test, the instant reaction of those present was, (Mark Levinson's creation fo r oil sheiks). On "Huh? What on earth is that?" The midrange, the other hand, for a lot less money, a pair of especially, sounded like a tin can next to the Rogers LS3/5A's with or without subwoofers, really accurate speakers. or a pair of DCM Time Windows, will provide Why? We think we know exactly why. The comparable . (in some ways even superior) Ohm F rings. It rings more than other speak­ musical satisfaction. ers. It rings like a telephone. It isn't even pos­ sible to localize the ringing at specific fre­ quencies; our test sample rang virtually every­ where. At 11 kHz, where there was also a 6 dB peak in the amplitude response, our sample did Ohm F worse things to a tone burst than we had ever seen in our laboratory. But then 11 kHz isn't in the most sensitive range of the ear. Maybe Ohm Acoustics Corp., 241 Taaffe Place, Brookl yn , NY it was the 900 Hz ringing that was the real cul­ 11205. Ohm F "coherent-sound" sp eaker system, $1200 prit. Maybe some other frequency. There were the pair. Five-year warranty, excluding cabinet; cus­ just too many bad spots to choose fr om. On top tomer pays all fr eigh t. Tested #32053 (cabinet #1 9297) of it, the fr ame of the Walsh driver resonated and #32068 (cabinet #1 9296), selected at random in sealed cartons fr om manufacturer's stock. at 118 Hz. Interestingly enough, it seems that another investigator has come independently to the Editor 's Note: To prevent rumors, anonymous same conclusion. Long after our own tests, letters, crank calls, slanderous know-it-all com­ we discovered that the DCM 'Time Window' ments, and other manifestations of ugliness brochure uses the Ohm F as the classic example and paranoia endemic to the high-end audio of ringing. Not by name, of course; but the scene, the fo llowing disclosure must be made. oscilloscope photograph they use as their ex­ The Editor jPublisher was one of the original ample is identical to the one labeled Ohm F in founders of Ohm Acoustics in January 1971 and their fifteen-speaker comparative pulse test for several years thereafter its largest single fu rther below on the same page. The photo-

34 graph also shows that the Ohm F reproduces below the hump, so you don't get the subjective the initial pulse perfectly, exactly as we ob­ impression of a flat 28-Hz box. On the contrary, served, but then just keeps on producing an in nearly all room positions there remains output without an input. some rather annoying midbass boom. We don't want to create the impression So, if you want one of the most brilliantly that this defect of the Ohm F is a vulgar design conceived speakers of all time-with a boomy error, such as we occasionally pillory in this bottom; a highly colored, ringing, metallic publication. It's damn hard to make a Walsh midrange and top end; plus superb time re­ driver that doesn't ring. The basic require­ sponse if your ear could only separate it from ments of the cone present a conflict. Since the all that mess-get the Ohm F. Incidentally, Walsh speaker has been around for more than this is the "improved" version; some minor five years, here's just a quick reminder of how changes were made about a year and a half or it works: A single cone, mounted convex side two years ago. We remember the original out and apex up, acts as a transmission line version as having been better, but we didn't (in the antenna sense, not the bastardized have all these other speakers to compare it to, acoustical labyrinth sense). Sound travels and besides we don't claim an acoustic memory down the side of the cone faster than in air, and quite that retentive and precise. It's possible through proper choice of geometry the hori­ that our standards of excellence have changed zontal component of the traveling wave syn­ more than the speaker. thesizes a coherent cylindrical wave front in the air, starting at the cone surround. That's the theory, anyway; its divergences from real life belong in another discussion. In any event, Pyramid 'Metronome' the cone material must be extremely stiff (i.e., have a high Young's modulus) in order to have the proper sound propagation velocity. That Pyramid Loudsp eaker Corp. , 71-07 Woodside Avenue, means a high Q. At the same time, the cone Woodside, NY 11377. Metronome Model 2 + 2W material mustn't ring; it must be well damped. speaker system, $2525 the pair (in black fo rmica, as That means a low Q. Thus the ideal cone ma­ tested; less in other finishes). Th ree-year warranty. terial fo r the Walsh driver would combine Tes ted #169/269 and #170/270, on loan fr om manu­ fa cturer. the most desirable characteristics of beryllium and mucus. No wonder that Ohm hasn't fo und it yet. They try to get away with a mundane In our Summer CES issue we gave a quick combination of titanium, aluminum and paper preview of Dick Sequerra's latest all-out at­ in tandem, which doesn't quite make it. As tempt at SOT A and reported a favorable first we've said in the past, the Walsh invention impression of the Metronome speaker system. would have deserved a multimillion-dollar R Just how the CES units resembled, or diffe red and D program to solve these and other im­ from, our test samples we'll never know for mensely difficult design problems it poses. In sure; until very recently Dick was going through which case it might have become the world's a modification-of-the-week phase, so the only simplest and most accurate speaker. It's far fair way to structure this review is to report from that right now. everything in chronological sequence. A less serious objection we have to the When our test samples first arrived, we Ohm F is that somewhere around 36 or 37 Hz, were assured that they were typical of the limit­ its response in the sealed enclosure has a 3.5 ed number of units already sold and then being dB hump, meaning a system Q of roughly 1.4. produced. They turned out to be absolutely That's too loose and boomy in our opinion; tops in our survey in two respects: dynamic it represents the rock-pop taste rather than the range capability and left/right imaging (among devotion to accuracy one would expect of the the larger units, that is; the minisystems are knights of Walsh. (Unless, o( course, the inherently excellent imagers). The Metronome speaker is put plunk in the middle of the room, was able to produce mind-blowing SPL peaks in which case the Q is approximately correct­ with suitable amplifier power behind it; it but how many people will use it that way?) The just couldn't be made to burp, grunt, spit, -3 dB point is about 28 Hz, but that's 6.5 dB sizzle or show any other sign of distress no

35 matter how hard we drove it. Clean as a whistle. to 0.4 msec, completely reversed the polarity And left sounded like left even when we moved of a 0.26 msec pulse, and was totally unable to all the way to the right. Just like real life, reproduce a 0. 1 msec pulse. A classic case oftime you say? No, the Metronome didn't sound like smear. We were then told by Dick Sequerra real life at all. The top end wasn't really open; that the 2-inch tweeter was undoubtedly wired the inner detail had no lifelike clarity and into the system out of phase. We had him re­ delicacy; everything was a little thick and verse the phase, after which the impulse re­ smeared. We were both intrigued and dis ap­ sponse got worse. Dick then decided that the pointed. original wiring had been correct after all and In the laboratory, this particular version reversed the phase back again. So much for of the Metronome turned out to be almost per­ verSIOn one. fectly flat in frequency response up to 15 kHz, Sometime later, we received a call fr om at which point it cut off rather sharply. The Dick informing us of a new 4-inch midrange only significant departure from a dead-flat driver that was supposed to improve the per­ profile was a 6 dB hump at 60 Hz, a little bit fo rmance of the Metronome considerably. Our peculiar in view of the 37-Hz system resonance Model 2 top sections were taken away to be specified fo r the Model 2W subwoofer in the modified and returned to us with the new Pyramid literature. It seems that the response midrange installed. The results were unbeliev­ we measured, indicating a system Q of 2, was ably bad. The beautif ully flat frequency re­ synthesized by a combination of the subwoofer, sponse was gone; the midrange output was its roll-off network, and the lack of low­ elevated and appeared to have extreme phase frequency roll-off on the Model 2 top section in coherency (1800 out with respect to the rest that links up with the Model 2W. (The Model of the system); impulse response was the worst 2 is allowed to run wide open; it's a complete yet-and the sound was definitely degraded, and listenable three-way system all by itself, with unmistakable time smear. A few days with a somewhat overdamped 8-inch woofer.) later we got another phone call, apologizing Despite the 60-Hz hump, the subwoofer didn't for the horrible mistake; the midrange had been sound boomy, just rolled off. No wonder: incorrectly modified and would we pl ease re­ the -3 dB point was at 30 Hz, meaning that turn our top sections once more for the correct the response dropped 9 dB in the all-important modification. End of version two. octave fr om 60 Hz to 30 Hz. According to Version three arrived shortly thereafter, Dick Sequerra, this is a deliberate design fea­ with assurances that this was It-the final, ir­ ture to avoid subsonic garbage and acoustic revocable production model. All units already feedback at the high SPL's the speaker is cap­ in the field would be modified free of charge able of. We aren't sold on this philosophy, to conform to this one, we were told. And as you can probably imagine; nor do we believe sure enough, this new version sounded different that a 14-inch driver is correct fo r the internal and, to our ears, somewhat better. More open volume of the subwoofer enclosure, which isn't highs, inne r detail more clearly etched, more much larger than that of the biggest bookshelf air around the instruments and voices. All speakers. This is just the opposite of the Infinity without compromising the unique dynamic QLS woofer approach: an oversize driver in an range and imaging capabilities. We still pre­ undersize enclosure. At least the oversize ferred the Beveridge by a wide margin and the driver is never asked to move more air than Rogers as well as the DCM by a narrower it's able to, hence its apparent freedom from margin; they just sounded more re al, with the usual eructations. But it can't give you the more of a see-through quality, even though low lows, as it could with more volume behind the Metronome was objectively the "cleanest" it. All in all, though, this was one hell of a in the sense that it produced no grit or crud, fl at speaker system. no matter how complex or dynamic the The speaker was also quite free from ring­ program material happened to be. At the same ing throughout its range, but its problems be­ time, something was subtly wrong with this came apparent when-you guessed it-we test­ new version that we hadn't noticed with version ed it with pulses. This parti cular version re­ one; the latter had been less transparent and produced a 1 msec pulse satisfactorily, got real but somehow more consistent, stable and progressively worse as we nar rowed the pulse balanced in sound quality. We weren't sure

36 what the difference could be attributed to. least fr om about 120 Hz on up at moderate We took the speaker to the lab and real­ levels, than any other speaker in this survey, ized almost immediately what was going on. except the Beveridge and the Koss-and we For the first time, the Metronome was reason­ aren't so sure about the latter (see our Koss ably accurate in the time domain, with vastly review above). For example, it's distinctly improved results on the pulse tests, though superior to the Infinity QLS. still not in the Beveridge category. At the same How can a little shoebox do that? Well, time, the frequency response was completely how did a few Spitfires and Hurricanes beat shot to hell. So much so that it would be futile the Luftwaffe? (And, rest assured, Hermann to try to describe its dips and bumps, since it Goering was a harder nut to crack than Arnie was totally unrelated to any discernible zero Nudell.) axis. As a desperate example, there appeared Actually, old chap, it's bloody simple. You to be a 10 dB peak at 6 kHz. With respect to 0 take the 4-inch KEF bass driver and the %-inch dB at what frequency? Don't ask. KEF dome tweeter. They're jolly good, don't What this proves is that good time re­ you know. You cross them over with a proper sponse without good fr equency response sounds network made of decent parts (it's no place to more like real life than good frequency re­ skimp by using cheap electrolytics, etc.). You sponse without good time response. What it pay special attention to the time and phase doesn't prove is that the Metronome speaker characteristics. You make the sealed box as system is a finished product. We know Dick solid and rigid as Westminster Abbey. And Sequerra well enough to predict that he won't then you cheat a little. be satisfied very long with the current pro­ By cheating we mean the bass response duction model. He is a perfectionist and wants profile. It's a bit of a humbug: would you be­ basically what we want. He has accomplished a lieve +6 dB at roughly 130 Hz? Now that's a great deal already: reasonable size for a super Q of approximately 2, and all the standard texts system, good looks (we love that 45-inch tall will tell you that the -3 dB point will then be metronome), fantastic construction (mostly around 75 Hz, which works out just about true 1 'I2-inch walls!), stupendous dynamics, no gritty to life. In other words, if the hill is high enough, garbage out with no garbage in, precise stereo its foothills won't drop out of sight, either. Is imaging, and now a fair degree of freedom this any way to get bass response? We don't fr om time smear. But that's not enough. particularly like it; our strict upbringing makes For two and a half grand, we expect more from us recoil from it-but how else are you going to the man who gave us the Sequerra Model 1 extract a subjective impression of bass out of tuner than a 50 Hz to 15 kHz speaker system that itsy-bitsy box? Most people are fo oled; that doesn't sound quite as real as the little time and again visitors told us that the bass Rogers at one fi fth the price. sounded just great-until we showed them real bass. In all other ways, however, the Rogers has impressively flat frequency response, all the way up to the limits of hearing. By 20 kHz it's down somewhat, but then the English have Rogers LS3/ SA always been kind to their dogs. Our pulse tests revealed outstanding time­ domain behavior and also explained why the Reference Monitor International. Inc.. Suite 309. 4901 Rogers sounds best with the tweeter end of the Morena Boulevard. San Diego. CA 92/17. Rogers box tilted backward or angled inward. The LS3j5A BBC Monitor Loudspeaker. $450 the pair on-axis impulse response showed some anom­ (approx. $250 ifpur chased in Great Britain). Five-year alies, but 450 or so off axis the two drivers warranty. Tes ted #SOI 777A jB and #SOI807A jB. on loan fr om distributor. fell into a time-aligned configuration, render­ ing better pulses between 1 msec and 0. 1 msec than any other speaker tested so far, except The Rogers LS3/5A, a British import, was the Beveridge. In view of the latter's size and originally developed fo r the BBe as a very com­ weight, the Rogers became our convenient pact monitor. Its outside dimensions are 12 laboratory reference standard fo r pulse shape by 71/2 by 61/2 inches. And it sounds better, at and coherence. Even so, it did exhibit a slight 37 negative blip from the tweeter at the leading car trunk, in an overnight bag, or whatever) edge of a positive-going 0. 1 msec pulse; how­ when you go hi-fi visiting. After listening to ever, all other conventional speakers, even the your friend's (or dealer's) $2500 speaker sys­ DCM, were doing the same thing to a consider­ tem, you connect your little shoeboxes and ably greater extent. Furthermore, the Rogers you've got an enemy for life. was remarkably free from ringing in compar­ ison with the others. It's a delightful or scary little bugger, depending on whether you're an audio enthusiast or a speaker manufacturer. A pair of LS3/5A's, separated by the right distance for your listening position and prop­ Snell Acoustics Type A erly angled, will give you beautifully focused, crystal-clear, unsmeared, highly detailed sound reproduction, with all the spatial information Snell Acoustics, 10 Prince Place, Newburyp ort, MA (front/back, left/right) you could ask for. No 01950. Typ e A loudspeaker system, $1370 the pair. reservations or qualifiers necessary. The Rogers Five-year warranty; manufa cturer pays return fr eight. will also resolve differences between amplifiers Tested samples on loan fr om manufa cturer. with ease and therefore deserves a very good one. The Norwegian Electrocompaniet power This is a unique speaker system, in more amplifier appears to be a particularly happy ways than one. To begin with, it's by far the choice, since it can't possibly develop enough best-looking large speaker we can remember: power across the speaker's 15-ohm impedance an utterly simple, virtually seamless, upright to hurt it (the Rogers is rated for 25 watts brick of polished wood and stretched cloth, four speech and music-not continuous power) and feet high, two feet wide' and just over one foot also happens to make the speaker sound superb. deep, designed to stand flat against the wall. (We blew out the woofer on one of our test The cabinetwork and finish are of a quality samples with the equally great-sounding zillion­ hardly ever seen in the hi-fi business; fu rther­ watt Bryston 4B; don't you be that stupid.) more, the unit separates into a lower section All right, what about that bass? Some (housing the downward-facing lO-inch woofer) people have suggested using two pairs of and an upper section (with 4-inch midrange and LS3/5A's to extend it, but that's not the way I-inch tweeter), so that two not very strong Mother Nature works. All you accomplish is a people, or even one in a pinch, can move it 3 dB increase in level (and power handling) around without breathing hard. Just beautiful with basically the same response profile. And -though admittedly not the most important you screw up the time response with the mul­ thing. tiple sources. Our suggestion: live with one Well, how important is frequency re­ pair, as is, or go to a biamped system with sponse? All right, listen to this. Nearfield subwoofers. We found the Fundamental Re­ response of the individual drivers in their pass­ search to marry very well with the Rogers, as band is dead flat (and we mean plus or minus long as the nominal crossover frequency was close-to-nothing, right on the zero line) fr om kept in the 100 to 120 Hz region. Ideally, the 38 Hz to 22 kHz. The bottom end rolls off to crossover should be a little higher, say 200 or -3 dB at 28 Hz, -10 dB at 20 Hz, -20 dB at 12 250 Hz, to get completely away from the Hz. On top, the response dips to -6 dB at 24 bumpy area of the Rogers; however, the Funda­ kHz. So, using the convention of -3 dB corner mental Research has some problems only an frequencies, the total passband can be called octave or so above that range (see our FR 28 Hz to 23 kHz-and no ripples. How about review above), so we had to stay lower. A prop­ that? We've never seen anything like it. What's erly crossed-over, biamped Rogers system with more, all drivers radiate this output without subwoofers promises to be one of the best significant peaksor dips over a very wide angle, money can buy; we plan to go more deeply so that with the crossover used the power re­ into the subject and make more specific sponse into the room is also astonishingly recommendations. uniform. Distortion is low; power handling is Meanwhile, we suggest you carry a pair very good; the woofer is particularly impres­ of Rogers LS3/5A's around with you (in your sive-the driver and the enclosure seem to be

38 exactly right for each other. (All you Q = 2 guys, wrong we are and where we should go to have please note.) our ears examined. On the other hand, if we're So how come the Snell isn't our top choice right-if time response is the more important in this survey? You already know the answer if design consideration-then the Snell Acoustics you've read this far in our alphabetical sequence Type A will have served as the classic control of reviews: the fr equency domain isn't where experiment. Either way, it's a fa r fr om negligible it's at. Mind you, the Snell sounds excellent. If contribution to speaker design. we hadn't been exposed to the Beveridge, the Koss, the Rogers and the DeM, we would have lived with this kind of solid, wide-open sound quite happily for a long time. It's just a wee bit on the bright side (probably on account of the uniform power response right up to the highest Recommendations frequencies), but we don't really mind that. Besides, the tweeter output can be reduced at the flick of a toggle switch. And the overall Please remember that the 15 speakers re­ sound is truly clean and balanced. But some­ viewed here, plus the dozen or so other units thing is lacking. It's that utterly alive, high­ we've commented on in the past, represent only ceilinged, see-through quality and unsmeared a small fr action of the universe open to serious resolution of inner detail available only with audio enthusiasts. We can't swallow that whole more or less time-coherent systems. Which the universe in one issue-or even in fo ur or six. Snell isn't. If you're looking for infallible and all-encom­ In fact, the Snell can't reproduce single passing guidance, you've come to the wrong pulses at all. You feed in one pulse, and out place. On the other hand, we know the stuff come two discrete pulses-one fr om the mid­ we've tested pretty well and can offer the fo l­ range and one from the tweeter. The two simply lowing circumscribed recommendations with aren't joinable, regardless of distance or angle. some degree of confidence. This is an inherently noncoherent system, prob­ ably on account of the large acoustic separation Best speaker system tested so far, regard­ between drivers. Too bad, since it's obviously a less of price: Beveridge Model 2SW (with reser­ system designed with tender loving care. vation about headroom-see review). In a way we're glad, perhaps at the expense of Snell Acoustics, that this speaker is such a Close to the best at a much lower price: pure embodiment of the frequency-response­ Rogers LS3j5A (but bass not quite adequate above-all approach. If that were indeed the cor­ without expensive subwoofer-see review). rect design philosophy, there couldn't possibly exist a better-sounding speaker, and we'll soon Best full-range system per dollar: DCM hear from some highly vocal sources just how Time Window (supersedes Dahlquist DQ-IO).

39 A Comparative Survey of Power Amplifiers: Part II

By the Staff of The Audio Critic

In this second installment of our search for the perfect power amplifier, we come a lot closer . We even nominate a new reference standard (at least until the next go-around) and also have other good news to impart.

As we explained in the introduction to ing fo r just the opposite: a meaningful test or Part I, we're convinced that a correlation can series of tests that no amplifier in the world be established between the audible and measur­ can pass without some difficulty. That way able characteristics of a power amplifier, al­ progress will be measurable; right now, if you though the road that leads there is a rocky and read the commercial magazine reviews, perfec­ winding one. Since this survey will have a con­ tion is being routinely attained. cluding Part III in the next issue, we'll wait Just so we won't leave you in total sus­ until then (as we said we might) to report just pense, we're willing to confess to continued how far down that road we have, or haven't, partiality to the CCIF intermodulation distor­ been able to travel. We're still very much in the tion test, with 14 kHz and 15 kHz mixed 1: 1. process of sorting out the tentative correspon­ We aren't quite sure, however, just how fine­ dences we've fo und so far between our listening­ tuned a tool it is; gross differences in the room and laboratory observations, and we cer­ resulting sidebands from amplifier to amplifier tainly don't want to be guilty of premature or invariably reveal audible differences; differ­ superficial generalizations, despite the blandish­ ences of a few dB may be imperceptible. We've ments of some fairly juicy data. One thing is also noticed interesting correlations between certain: the Hirsch-Houck type of testing is phase shift and sound quality, even though the insufficient and leads only to the discovery of resulting time delay at the frequencies in ques­ one SOTA amplifier after another. We're look- tion may not, on the face of it, appear to be 40 significant. And so fo rth and so on: we're knee deep in clues, without firm conclusions. Please Bryston 4B be patient; give us another couple of months to find what no one else has in thirty years-or to report with some certitude that we're stuck.

The continuing fr amework. Everything we've said about our listening Bryston Manufacturing Limited, 17 Canso Road, Unit test conditions in Part I still applies to the re­ I, Rexdale, Ont., Canada M9 W 4MI. Model 4B basic views below, but this time there's more. To wit: stereo power amplifier, $1250. Th ree-year warranty; In addition to the M ark Levinson ML-l, manufacturer pays return fr eight. Tes ted #4139 and we also used the Hegeman HPR/CU (proto­ #4140, owned by Th e Audio Critic. type) as a reference preamplifier. The latter is an outstanding unit, but please, don't jump to any conclusions. We haven't tested the two against each other, nor the Hegeman against This is it, music lovers and audio fr eaks the Rappaport, nor any of these against other -the best power amplifier we've been able to preamps with SOT A claims that have recently fi nd so far. Of course, we haven't tested them appeared on the scene. We have no definitive all. (Two that we should definitely look into, opinion on the subject. As you can see on our and haven't yet, are the Electro Research and back page, the question will be resolved in a the Mark Levinson ML-2. And there are new round of preamp tests to be reported in the others.) But this will be the one to beat, fr om next issue. Needless to say, we never changed now on. two links in our reference chain at the same The Bryston 4B is actually two separate time. Whenever we compared two different mono amplifiers on one chassis. The only thing power amplifiers, the preamp remained the the left and right amplifiers have in common same, whether it was the Mark Levinson or the is the line cord. Each channel will deliver up Hegeman. to approximately 40 volts across just about any We also used other speaker systems this load that doesn't approach a short circuit­ time in addition to the Dahlquist DQ-lO. These meaning 200 watts into 8 ohms, 400 watts into included the Koss Model Onel A electrostatic 4 ohms, 500 watts into 3.2 ohms, you name it. (a difficult test load for some amplifiers), the Since it's a truly unflappable voltage source IS-ohm Rogers LS31SA (used full-range as well with a brute-force power supply that never as with the Fundamental Research subwoofer), seems to run out of juice, the amplifier simply and the DCM Time Window. Again, any two doesn't care whether it sees an obscenely com­ amplifiers under comparison were always lis­ plex load like some of the fu ll-range electro­ tened to through the same speaker, although statics or a virtually pure resistance like a Cizek the more critical comparisons were repeated box speaker. It just keeps rollin' along. That's with several speakers. probably the main secret of its listening quality, We keep mentioning comparisons of two but since its designer has expressed the engi­ amplifiers, one against the other, since this new neering philosophy behind it in his own words, crop of power amps did require some careful which we're reproducing herewith, we'll refrain A-B-ing to shake down the finalists. No won­ from second-guessing him. We can't resist won­ der; quite a few of them were superior to our dering aloud, though, why there's all the fu ss previous top recommendation. about pure class A output stages both in this Otherwise, our reference system remained country and in Japan, when in Canada class the same, except that the Denon DL- 103S AB can be made to sound this good with this (played through the V erion transformer) per­ kind of power. manently replaced the EMT; also, the modified Because the Bryston 4B sounds just about Stellavox 'Stellamaster' tape deck introduced unexceptionable to our ears. On the top end it in our speaker survey in this issue was used rivals the Electrocompaniet amplifier, which is extensively with the master tapes mentioned our standard of excellence on that count-open, there. sharply etched, unsmeared, fast and sweet at On to the reviews, then. the same time. And of course the 4B eats the 41 Comes The Resolution, All Power Amps Will Sound Great

Editor 's Note: Wh enever we're sufficiently impressed by the performance of an audio com­ ponent, we try to brain-pick its designer fo r some kind of correlation between engineering phi­ losophy and sound quality. Having been thus impressed by the Bryston 4B power amp, we man­ aged to elicit the fo llowing observations fr om Chris Russell, a very serious and very long-haired young man who not only designed the circuit but also, with his brother and fa ther, runs the company.

The Audio Critic: rents and bias �oltages within the perature, this causes nonlinear distor­ In an era when power amplifiers amplifier into a state of flux, muddy­ tion. Further, much of it is high­ iue attaining distortion measurements ing the signal. Tying both channels to order. Common sense dictates, there­ approaching a thousandth of one per­ the same power supply exacerbates fo re, that we should run low-Ie�el cent o�er a wide range of frequencies the problem greatly, as then one chan­ class A stages at no more current and output le�els, essentially fa lling nel can affect the other. Split up the than necessary, and at reasonably low to background noise le�els which are power supplies and regulate the �oItage �oltages if possible. (Especially in �iew theoretically inaudible, why is there to all the sensithe stages, and there of the fact that, other things being the continuing search for an amplifier is an immediate and ob�ious impro�e­ equal, transistors distort less in e�ery which is entirely accurate to the ear ment in clarity, transparency and resol­ way when they handle less current. ) musically? Whence springs the de­ ution. This way the junction will always be parture from theory which seemingly An interesting word, that-resolu­ near room temperature and not subject separates objecti�e and subjecthe real­ tion-for it implies much more than to change with signal swing. ities? Is it possible we are hearing appears at first glance. Resolution The one unfortunate fly in the things we cannot (or do not) measure? denotes fo cus, the ability to delineate ointment is that transistors tend to Let us imestigate. subtle and delicate details in the increase in frequency response with The latest thinking on the prob­ musical picture, and its absence can increased current flow. Thus the Great lem of dynamic distortion mechanisms ha�e as much to do with the remo�al Slew-Rate Race dictates that many in amplifiers centers largely on the of something as with an addition such designers opt for speed rather than popularization of Transient Intermod­ as harmonic distortion. It is entirely linearity and ignore this common­ ulation Distortion. As it is known, this possible for a circuit to "lose" subtle sense ideal. (Recently I ha�e seen is a feedback-defeating phenomenon information, or at least to hide it be­ some amplifiers with low le�e1 stages related to the ultimate speed or slew­ hind the coarser fabric of the o�erall dissipating so much current that they rate capability of the amplifier. Feed musical fr amework. require small power de�ices on their own an amplifier circuit a transient signal It appears to be this loss of heat sinks, running blazing hot to the faster than it can handle and it gene­ resolution which offends the ear touch !) Since amplifiers tend to ha�e rates a momentary burst of distortion rather grie�ously in many circuits gobs of feedback a�ailable at low fre­ trying to catch up. which display a certain amount of non­ quencies, this problem is often hidden It should therefore be possible linearity in the signal processing, from steady-state distortion measure­ to "perfect" an amplifier by either and which hope to straighten out the ments, but it is still audible as a loss low-pass filtering the input or pushing problem with feedback or with "com­ of resolution. Unfortunately, once the the slew-rate limit to stratospheric plementary nonlinearities" fu rther down finesse in separating tiny nuances is �alues. That this engenders only some the line. sacrificed, it can ne�er be regained impro�ement leads us to the inescap­ Case in point: Another low-fre­ later. able conclusion that it is a gross quency �agary which often creeps into This is why I say that, to the ear, o�ersimplification of a �ery real prob­ amplifier circuits is related to high distortion is distortion. There is no lem. It is e�en possible that exclusi�e standing currents in class A stages. point in taking a step backwards in one concentration on this one particular When a transistor conducts a good deal area in order to gain in another. aspect of an amplifier's beha�ior can of current, at a rather high voltage, Speed and slew rate are �ery important actually ha�e detrimental consequences it heats up-since �oltage times cur­ parameters, but only insofar as they on o�erall musical reproduction. rent equals power. When the signal relate to obtaining the maximum bene­ In point of fact, there are dozens causes the �oltage to swing up and fit from a circuit which is maximally of phenomena which can encroach down, the power dissipation �aries linear at all frequencies, as well as upon an amplifier's ability to handle from top to bottom of the cycle. At from the amount of feedback which is a musical signal with negligible dis­ mid to high frequencies this tends to employed. (After all, feedback can tortion. Most are transient-related. integrate o�er the cycle, but at lower and does further reduce distortion. ) Many are low-frequency problems. frequencies the transistor chip will The important watchword is o�erall For instance, when an amplifier de­ ha�e time to �ary as much as tens of musical accuracy. Ii�ers a pulse of low-frequency energy degrees Celsius from one end of the Thank you �ery much for your to a loudspeaker, the current flow signal swing to the other. Since most interest. tends to pull down the power supply of the transistor's characteristics, in­ Sincerely, �oItage. This change in �oItage supply cluding gain, base-emitter bias �oItage Christopher W. Russell throws all the sensiti�e standing cur- and frequency response, �ary with tem- Bryston Manufacturing Ltd.

42

.' Electro fo r breakfast in overall drive capa­ bility and load handling. The Bryston also has eM 914 the best bass we've heard so far, free fr om the least trace of looseness, hangover or artificial "warmth." I t just goes down, down, down, tight and unshakable. The midrange may not have the uncanny clarity of the Futterman, but it's A udio International. Inc . . 3 Cole Place. Danbury. CT 06 810. CM 914 Stereo Power Amplifier. $449. Tes ted close (damn close), and the overall quality of #3 129. on loan fr om manufacturer. the Bryston is more open than that of the Futterman. Actually, as a high-power amplifier fo r all seasons, the Threshold 800A was the one the Bryston had to beat, which it did quite Since the Quatre OG-2S0 seems to be handily. By comparison, the Threshold, a very leaving a very poor track record as regards good-sounding amplifier in its own right, reliability, this neat little 100-watt-per-channel seemed a bit heavier, thicker and more closed­ amplifier from the successor of C/M Labor­ down in its sonic presentation, with just a least atories is the best we've been able to fi nd in bit of the aforementioned "warmth" factor. the same price range as an all-around work­ Incidentally, the above observations apply horse alternative. The CM 914 sounds more only to a Bryston 4B that has been turned on open than the Quatre on top, although we still for half an hour to an hour. Listened to cold, prefer the sound of the latter on an overall the amplifier sounds a little hard and glary. basis. The CM 914 is a little less focused and The warm-up seems to stabilize the circuits. transparent in the midrange and not as firm in In the laboratory, the Bryston 4B appeared the bass. The very fa ct, however, that the CM to be irreproachable, with point-double-oh or can give the Quatre a hard time in an A-B low single-oh distortion on every possible THD comparison shows that it's a better amplifier and 1M test. No problem at all on the 14-plus­ than most of those reviewed in Part I, includ­ IS-kHz killer test; gorgeous square waves right ing the bigger and more expensive CM 912a. down to 20 Hz and below; in fact the phase shift (It's interesting to note that the 912a incor­ at 20 Hz is an incredible 10 (don't forget, this porates current limiting whereas the 914 does is not a DC amplifier) and totally unmeasurable not.) by the time you get up to 100 Hz. The slew rate Our laboratory tests revealed some pecu­ appeared to be a little over 40 V IuS, which is liar ringing on square waves, which was some­ somewhat short of the "greater than 60 V IuS" times intermittent, sometimes stubborn and specified, but then who is to say that Bryston's ineradicable. Since it existed in both channels, technique fo r this rather difficult measurement we doubt if it was due to some kind of minor isn't superior to ours. In any event, you can defect in a single . component. The ringing, double the slew rate by throwing the two chan­ excited by the leading edge of any square wave, nels into a bridged mono configuration at the appeared to have a frequency of approximately flick of a switch (what a nice feature!), in 1 MHz and a decay time of about 3 uS. Whether which mode the available power is 800 watts it had anything to do with what we heard ... into 8 ohms. That's more than one horsepower. who knows? As a footnote to all this, we must refer Otherwise the amplifier exhibited single­ you back to Th e Admonitor column in our oh distortion within its power range on our second issue, where we took Bryston to task TH D and 1M tests, except on the 14-plus- 1S­ fo r claiming to have the world's best power kHz CCIF, where it did a little worse though amplifier, period. We still stand firm on that still acceptably (better than the Threshold admonishment, even though the claim seems 800A for example). Its strongest spec seems to less outrageous after our tests. When an ad be phase shift; there's hardly any to speak of makes a superlative product appear like a hype, between 20 Hz and 20 kHz. it's time to switch to another ad writer. As audio purists, we can't get too excited Oh yes. What about the Quatre DG-2S0, over this amplifier; on the other hand, we must our previous top choice? Not quite in the same admit that it represents decent performance class with any of the amplifiers mentioned and value when measured against what else is above. available. 43 whether that isn't the root of the problem. D B Systems DB-6 Ultralow distortion can only be achieved with lots of feedback; lots of feedback will cause ... there we go again. D B Sys tems. PO Box 187. Jaffr ey Center. NH 03454. Another thing we aren't sure about is the DB-6 Precision Power Amplifier. $650. Five-year war­ net effect of the subsonic filter in the DB-6 ranty; manufacturer pays return fr eight. Tes ted #200730. (12-dB-per-octave Butterworth with 15 Hz on loan fr om manufa cturer. corner frequency). There has been some talk of audible group delay introduced by certain types When a 40-watt-per:-channel power ampli­ of subsonic fi lters, but not such simple net­ fi er of super-austere appearance costs $650 works. At any rate, the resulting phase shift at (i.e., considerably more than the much more 20 Hz is 640, which means a delay of just under powerful Quatre DG-250, Quad 405, CM 914 or 9 msec at that frequency. GAS Son), one would expect it to produce We wish we could either like the DB-6 a extraordinarily accurate sound at moderate little more or else be able to explain a little levels. The DB-6 sounds clean, sweet and non­ better just where the design went wrong. Neither fatiguing, but not accurate. Hence we must call of which wishes can change our conclusion: too it a disappointment. little fo r too much. Example: We were listening to an unac­ companied quartet of voices through the Bryston 4B. Each voice stood out in perfect relief, without the slightest tendency to cover up the Electrocompaniet other three. You could actually "hear" the (Part II) spaces between the four singers. We switched to the DB-6. The sound blended into a pleasant, euphonious glop. It wasn't unmusical; if we Electrocompaniet, Toyengt. 14. Oslo I, Norway. "The Two-Channel A udio Power Amplifier." $505 to private hadn't known what was really on the record, we users. $344 to dealers. FOB Oslo. Tested unnumbered would have been satisfied. Strange amplifier. sample. on loan fr om owner. It's interesting and instructive to compare the DB-6 with the Electrocompaniet amplifier, Since our original review, we've found out which has a somewhat lower power rating (it a great deal more about this lovely little ampli­ also costs less) but is otherwise similar in its fier. First of all, it does really "exist" from techno-image: a cute little black box for the the consumer's point of view; you can write fo r purist. The Electro has a slew rate of 100 V IuS; it to the above address and, sooner or later, the DB-6 is much slower at 15 V IuS. The Elec­ they'll send you one. (For all we know, by the tro does a little better on the 14-plus-15-kHz time these words are in print, there'll be an CCIF test, but not much. (Neither is in a class American distributor; don't count on it, though.) with the Bryston 4B, even at low power levels.) Secondly, it's every bit as good after greater Both are uncomfortable with low-impedance fa miliarity with it as we fi rst reported. Thirdly, loads that tax the current capability of the the 100-watt-per-channel version is still only power supply. Does the Electro sound consid­ in the laboratory phase, not in production. erably more accurate (a la Bryston) because it's Our second sample of the amplifier was a fa ster? Or is that only the by-product of the little diffe rent from the fi rst physically, but not actual causative factor? Or is it purely inci­ in sound. The highs are still SOTA; completely dental? These are some of the nagging questions unsmeared, etched but edgeless, totally trans­ we're trying to fi nd answers to. (Lots of luck, parent. The rest of the range is equally good, fe lla, say some pretty smart engineers.) as long as you don't ask the amplifier to drive One thing that isn't responsible for the loads it doesn't like at levels it can't handle. shortcomings of the DB-6, for sure, is harmonic Specifically, complex loads such as the Koss distortion. The specs say less than 0.005% from Model Onej A, which drops down to rather low 20 Hz to 20 kHz, and it's the truth-we've impedance at some points in the audio range, verified it. Double-oh and triple-oh distortion distress the Electro to a surprising degree. Even figures seem to be the big thing at D B Systems, at moderately loud listening levels, the Kossj and we're very much beginning to wonder Electro combination sounded hard and irritat-

44 ing. On the other hand, the IS-ohm Rogers impressions. This is unquestionably one of the LS3/5A makes the Electro very happy; at that world's fi nest power amplifiers, but not an impedance the current demands on the power amplifier for all seasons-all applications, supply are reduced and the sound is of the that is. utmost purity. First of all, the Futterman doesn't like to Interestingly enough, the amplifier will see low impedances, such as are presented by deliver 14 volts or so into any load without the Koss Model One/ A electrostatic, fo r exam­ actually clipping (i.e., 12.5 watts into 16 ohms, ple. It just isn't on its best behavior at 4 ohms; 25 watts into 8 ohms, 50 watts into 4 ohms); it can't deliver enough power, for one thing, and however, at 4 ohms the power supply has trou­ clips badly when pushed. Go just a wee bit ble providing the necessary current and high­ higher and the impedance match improves; for amplitude signals begin to be envelope-modu­ example the DCM Time Window, which is still lated by the 60-Hz line frequency. That's what a fairly low-impedance speaker, never sounded we obviously heard on the Koss. Otherwise better than when we connected it to the Futter­ the Electro behaved very nicely on the test man. At 8 ohms the situation is pretty well bench; even the 14-plus- 15-kHz CCIF test went stabilized, and at 16 ohms the amplifier comes well, though not brilliantly. So, we must ask into its own; but how many 16-ohm speakers again, is it the 100 V IuS slew rate with 1 MHz are there? The Rogers LS3/5A is rated at 15 bandwidth that makes this circuit sound so ohms but shouldn't really be used with the good, or is that the symptom rather than the Futterman, which can deliver about 150 watts cause? Tune in the next time around and see if into that impedance and blow out the rather we can answer the question. fr agile KEF drivers. (We tried the combination, Incidentally, the Electro has a rather anyway, and it sounded gorgeous-but the peculiar input with a 1000-ohm impedance Rogers/El ectrocompaniet duo was, if anything, and mustn't be driven fr om a high-impedance even better.) source such as a tube preamp. Input sensitivity As regards specific sound quality, the for full output is only 0.35 volts. Electrocom­ Futterman seemed to have the most amazingly paniet makes a matching preamp, which we transparent and unblurred upper midrange haven't seen, that has a lO-ohm output impe­ (female voices, etc.) of any amplifier that has dance, just the right gain fo r the power amp, crossed our path; superb highs (but not quite as plus a 70-kHz low-pass filter, presumably to fast, etched and open as the Electrocompaniet protect the power amp against transient over­ or the Bryston 4B); excellent bass if used with load. Interesting. extra output capacitors (more about that be­ Our recommendation: think · about your low), and a slightly compressed overall sound speaker load, your loudness requirements, and in comparison with the Bryston (our reference consider this power amplifier very carefully if standard). In other words, better than any other you want close-to-SOT A sound with the afore­ amplifier in some respects, better than all but mentioned limitations at a reasonable price­ two or three others in every respect. It must reasonable even with fr eight and import duty be that at least one unidentified kind of dis­ added. tortion, the kind that dumps spurious products into the all-important upper midrange, has been more thoroughly eliminated in the Futter­ man than in any other amplifier. Futterman H-3aa I n the laboratory, the Futterman was no match fo r the fi nest solid-state amplifiers, such (Part Il) as the Bryston 4B. All our usual tests came out a little worse, though none of them dis­ Futterman Electronics Lab. 200 West 72nd Street. New turbingly so. It seems that a tube amplifier York . NY 10023. H-3aa vacuum-tube power amplifier is a whole different species, and the same rules (mono). $260; stereo pair. $520 (no longer made to order at this price; see below about availability). Tes ted don't apply. Which may prove, in the end, that #105 and #106. on loan fr om manufacturer. the rules are wrong, since it shouldn't make any diffe rence to the electrons how we make them Our test samples arrived as promised (see flow, as long as they flow just so. Part I) and confirmed just about all our initial We mentioned the output capacitors in the

45 Futterman, which in the standard model add up failure seems to be particularly drastic and to 2400 microfarads per channel. Julius Futter­ destructive. Here's the scenario we've been man supplied us with an extra 4200-microfarad hearing over and over again: The predriver cluster fo r each amplifier, so that each of our transistor fails and a huge amount of DC, speakers could be coupled to the tubes via 6600 representing a major portion of the rail voltage, microfarads. And it made a difference. The appears across the output. Since the amplifier bass was drier, more solid, more realistic has no protective circuits to sense this type than with 2400 microfarads. Whether this has of failure (don't ask us why), the DC passes anything to do with phase shift and time delay, through the speaker load. The speaker either we have no idea. It certainly has nothing to isn't fused or the fuse doesn't blow fast enough. do with amplitude response. Even the standard The speaker goes "arrgh" and that's that. No model is flat at 20 Hz. It may be that elec­ more woofer-and maybe no more other things. trolytic capacitors have some peculiar distor­ Not nice at all. Inexcusable, in fact. tion which is minimized by multiplication in Out of just three Quatre DG-250's that parallel. have passed through our own hands, one of The main problem with the Futterman is them did this. A second one conked out non­ where to get it. We've been informed that destructively. That's two out of three. And Julius Futterman has stopped taking orders similar cases we personally know of add up to a from private individuals at the ridiculously low solid two figures, not one. On top of it, we price listed above. He is now selling every hear that Quatre has been dragging their feet amplifier he makes by hand to The Sign of in responding to the situation. New circuit the Golden Ear (Riverview, Michigan), the out­ boards to replace the burned-out ones take for­ fit that plans to market the amplifier in a com­ ever to receive, we hear; and we've had reports mercially packaged version, which to the best of noncooperation and evasiveness by the front of our knowledge doesn't exist yet. The price office. In all fa irness, we must remind you that of this latter version was originally planned to we spend most of our time and energy on equip­ be $989 for a stereo unit or stereo pair. That's ment tests and editorial work, not on double­ all we know. checking the authenticity and accuracy of com­ I f it's any consolation to you, a Futterman plaints. But there must be a fire of some sorts amplifier was just as hard to get in 1956. where there's so much smoke. And you should have seen the smoke that came out of our DG-250 ... So, if you buy a Quatre, you're on your own as far as we're concerned. Enjoy the good Quatre sound, but don't say you haven't been warned. (follow-up) DG-250 The amplifier may easily turn out to be more expensive than you thought.

For manufacturer. price and other particulars. see original review in Part I.

We regret to tell you that our original RAM 512 endorsement of this amplifier must now be heavily qualified. It still sounds as good as RA M Audio Sys tems. Inc .. 17 Jansen Street. Danbury. ever (though not quite as good as the top ampli­ CT 06810. RA M 512 Stereo Power Amplifier. $1150. fiers in this second batch), but just too many Th ree-year warranty; manufa cturer pays return fr eight. reports of fa ilure in use have reached us to be Tes ted #10547. on loan fr om manufacturer. ascribed to normal attrition. Since for every failure reported to us there must be a good In our CES report we made the statement many others we'll never know about, the total that Richard A. Majestic, who is Mr. RAM, number should logically be a substantial per­ was originally the M of the former C/M centage of production, since not that many Laboratories, which is now Audio Interna­ Quatre DG-250's have been made thus far. tional. That was an error. Dick was an em­ And that's not all. The typical DG-250 ployee at C/M, not a founding partner. We

46 also made the statement that his RAM 512 don't claim (not yet, anyway) that this is an power amplifier was an "all-out design." That established criterion of audible quality. On top was no error, except in the intended sense. The of it, the amplifier went into protection on both amplifier we tested turned out to be an all-out channels toward the end of our tests and we design all right-all out to make money by couldn't make it work again with out looking tweaking super specs (high power, double-oh into the cause-which, quite fr ankly, didn't THD, etc.) out of a basically flimsy box and interest us at that point. selling it at an exorbitant price. Our conclusion: give us more amplifier If you're getting the impression that we're for the money, RA M, or else take out those a little upset and indignant about this, you're meters, give up a few watts, and cut the price quite right. The price of the RAM 512, either in half. retail or wholesale, is only a few dollars below that of the Br yston 4B, our new reference standard. And the comparison between the two amplifiers is a joke. Compare the power sup­ plies (single vs. dual), the heat sinking (hardly Threshold 800A any in the RAM, just convection cooling), the quality of parts, the general construction (the Th res hold Corporation. 1832 Tr ibute Road. Suite E. Bryston weighs 50% more)-we could go on. Sacramento. CA 95815. Model 800A class A stereo power amplifier. $2275 (discontinued). Two-year war­ Without knowing the prices, you'd guess a cost ranty; manufa cturer pays all fr eigh t. Tes ted #770408 . ratio of about two to one. What the RAM gives owned by Th e A udio Critic. you fo r your money is a pair of big, flashy dB meters plus peak-reading -20 dB LED's, and Even though the Threshold 800A has been lots of protective devices to keep the tweaked discontinued as a specific model, we consider circuitry from going up in smoke or blowing it well worth reviewing, since this company out your speakers (at least that's conscientious, is strongly committed to the patented active­ unlike Quatre). Even so, the RAM can't give bias class A circuitry it incorporates. A smaller you the same maximum volts into 4 ohms as and less expensive version, the Threshold into 8 ohms, which the Br yston does without 400A (a 100/1 OO-watt power amplifier at a mere batting an eyelash; and even into 8 ohms the $1 147) is still on the market; and a successor to RAM gives you a little less power, being rated the 800A, tentatively called the 8000 and built at 170/ 170 watts. on two separate mono chassis, is soon to be re­ Of course, the most important comparison leased. (The 810 we had mentioned in our CES is just listening-and, again, it's no compar­ report was shelved in the prototype stage.) ison. The RAM 512 sounds a little overbright, The 800A was undoubtedly too expensive hard and glary, though not disturbingly so; it and impractical to make: a huge but beautiful gives a nice feeling of depth, but doesn't have black monster, with massive handles and two anywhere near (and we mean anywhere near) enormous meters calibrated in watts at 4 and the transparency and definition of the Bryston 8 ohms, practically hand-wired throughout­ (or the Futterman or the Electrocompaniet, a real status symbol. The rated power is 200 fo r that matter, at much lower prices). If the watts per channel into 8 ohms; a conventional RAM cost, say, somewhere in the seven hun­ (constant bias) class A amplifier of this power dreds, we'd be more charitably inclined toward would have to be the size of a small refrigerator. it and would probably weigh its positive and The Threshold system of active or dynamic negative qualities with greater deliberation. biasing reduces the power dissipation require­ At $1150, however, it's put up or shut up. ment at idle by a factor of four; hence On the test bench, the RAM 512 displayed the large but still viable dimensions of the a peculiar kink in both the leading and trailing 800A. A critique of the Threshold patent is be­ edge of a square wave (any square wave), which yond the scope of this review; we have, however, is definitely not normal or acceptable. (Over­ heard a few impeccably qualified technologists biased transistors?) Phase shift at the lower fre­ speak of it with a certain degree of disdain. quencies was also considerably higher than Their comments ranged fr om "class A in name we're comfortable with, although it's inevitable only, not in fact" to "unnecessarily complicated with the AC input coupling employed and we without genuine advantages" and even to 47 "class A is pure bullshit, anyway." On the other measured the standard (meaning SMPTE) 1M hand, the Mark Levinson ML-2 is a constant­ distortion of each amplifier. Lo and behold, bias class A design with some pretty good the Bryston was almost a whole order of magni­ credentials behind it, so we'll hang fire on tude better, with double-oh fi gures throughout this one until we have an ironclad case one way as against the Threshold's single-ohs. How or the other. about that? Since the proof of the circuit is in the sound, - The Threshold also showed more phase we must report that until the arrival of our two shift at the lower fr equencies than the Bryston Bryston 4Bs, the Threshold was our reference (8° vs. 1 ° at 20 Hz, 2° vs. 0° at 100 Hz), but we amplifier fo r a number of weeks. It sounded can only repeat our reservation about the signif­ more solid, sweet and open than the Quatre icance of this as stated in the RAM review DG-250 (big deal, at about five times the price); above. In most other respects, such as slew rate it wasn't quite as uncannily transparent in the and THD fo r example, the two amplifiers upper midrange and lower treble as the Futter­ were quite comparable. The Threshold, how­ man, but unlike the latter it could drive low ever, doesn't have the Bryston's mono bridg­ impedances without any problem; nor did it ing fe ature that doubles the slew rate and have the totally unsmeared definition of the quadruples the power. Electrocompanient, but with its vastly greater Our suggestion, then, if you can afford the power capability it was a considerably more Threshold, is that you buy two Bryston 4B's, ' versatile and practical workhorse. We were which won't cost you all that much more, quite pleased with it and listened to it with a throw the bridging switch and enjoy the fastest great deal of satisfaction. two 800-watt channels in the business. Even Then came the Bryston 4B and the pecking if fa ster doesn't mean better (which remains order changed somewhat. The 4B has a tighter, to be seen), it's nice to have that kind- oC drier, more thoroughly lifelike bass; a more headroom. open, luminous quality throughout; and more U nIess, of course, you feel underprivileged snap and definition on top, without edginess. without class A. In which case you might The Threshold sounds slightly thicker and want to wait fo r the Threshold 8000-or get darker; with sweeter but less fi nely delineated in touch with a Mark Levinson dealer. He'll highs; somehow more closed down overall. show you what class A, and privilege, can do Minor but noticeable differences, between two fo r you. outstanding amplifiers. We've heard it said that the Threshold was trimmed in to make the Dayton Wright electrostatic sound as good as possible, since the two have the same channels of distribution and are generally demonstrated together. We hope that's not true; a power Recommendations amplifier shouldn't be used as an equalizer or compensator; however, delivery of our Dayton Wrights was delayed (see the speaker Pending the conclusion of this survey in survey in this issue), so we still don't know the next issue, which may still hold surprises, more than what we're told-which can be dark horses and even scandals, here are our notoriously unreliable. current top choices. Our laboratory results were most interest­ ing: the Threshold 800A did quite poorly on Best power amplifier so far, regardless of the 14-plus- 15-kHz CCIF test, with 13 and 16 price: Bryston 4B. kHz sidebands averaging 12 to 15 dB above those produced by the Bryston 4B at the same Best sound per dollar: Electrocompaniet power levels. This makes us wonder about (if you can get one and are satisfied with 25/25- the sonic correlatability of the test itself; watt power capability). after all, the Threshold does sound better than a lot of amplifiers that throw less obvious Special situation for the connoisseur: Fut­ sidebands. It's the fi rst significant exception terman H-3aa (somebody will get one next so fa r, though. Just to fo llow through, we week or next month-why not you?). 48 Cartridge, Arm and Turntable vs. the Groove: Who's Winning?

Part I: In which we emend and update our original article on track­ ing geometry, fi nd a mind-blowing sonic gap between optimized alignment and currently accepted practice, and as a result send all equipment reviewers back to square one. Also: a common-sense critique of 10 tone arms and turntables.

In our first issue (January/February 1977, investigation and listening, that we didn't go pp. 31-35), we published an article under the nearly far enough. We grievously underem­ title of "Have Tone Arm Designers Forgotten phasized the sonic consequences of the mathe­ Their High-School Geometry?" Our message matical sloppiness we were complaining about (namely that correct geometry in a pivoted and gave no inkling of the chain reaction of tone arm costs no more than incorrect geom­ invalid conclusions that can result from com­ etry, yet there's no arm on the market that's placency about the cartridge/arm/turntable 100% correct) put quite a few noses out of vs. groove relationship. joint, for which the blame must be placed on The man we must credit fo r steering us in the unfeeling laws of nature. If it's any con­ the right dire.---ction and dispelling our initial solation to those who were upset by our facts, doubts about the audibility of ridiculously we now consider that article to have been small deviations fr om optimum tracking angle somewhat superficial. Not in the sense that (both lateral and vertical) is Mitchell Cotter, the argument it presented was in any way in­ the encyclopedically learned technologist of correct; far from it. We still stand behind Yerion Audio. Mitch has been spreading the every statement made in it, as far as the word with apostolic zeal; he has undoubtedly physics and mathematics of the subject are aligned more record players just as a favor concerned. But it's quite obvious to us today, to the innocent and distributed more Xeroxed after approximately nine additional months of instruction sheets on how to do it than there 49 are letters on this page. The Boston Audio or the parallel pOSItion of the arm tube with Society is still reeling from his panzer assault respect to the record, are almost entirely useless. on the subject (massively documented as usual) 4. Unfortunately, the correct vertical track­ at their March 1977 meeting. Our own expe­ ing angle is diffe rent for all records cut at riences since our fi rst issue and recent conver­ different times, in different places, by different sations with other proselytes (who are still few people, on different lathes, with different cutter and far between) bear out the Cotter doctrine heads. No adequately stringent industry-wide without qualification. standard is in fo rce, even today. This is very bad news, since it means that truly optimized playback requires constant adjustment by hand The unsettling but inexorable facts. (and ear) from record to record, unless a uni­ It would probably be more logical to start fo rmly cut series of records is being played. with certain unarguable givens and then lead 5. What makes all the above agonizing you step by step to our somewhat disturbing worthwhile, or at least meaningful, is that but nonnegotiable conclusions. But we feel that optimum alignment of the cartridge and arm the latter are too important to be shared only fo r both lateral and vertical tracking error with those who are willing to track with us improves the average sound quality of a record through our exposition. Since we don't want collection to an astonishing degree. Records to lose anybody at this juncture, not even tone previously thought to be hopelessly bad often arm designers with a short attention span, end up sounding quite decent, and the better we've decided to spell out our message first records begin to approximate what high fidelity and explain later. Thus: is supposed to be all about: a creditable imita­ 1. Lateral tracking error, if large enough tion of live music. In other words, phonograph to be at all measurable with simple tools records are a more accurate medium than (straightedge, triangle, protractor, ruled paper, jaded audiophiles suspect. piece of thread, or whatever), is readily audible 6. By the same token, listening tests and through a speaker/amplifier system of reason­ equipment reviews based on phonograph rec­ ably high resolution. Even the difference be­ ords are of highly questionable validity unless tween the successive zero-error and maximum­ the lateral and vertical tracking alignments error points of an optimally aligned pivoted are religiously optimized and maintained. arm is audible, let alone the difference between Since that's hardly ever the case, most evalu­ an optimally and a non optimally aligned arm. ations of cartridges, tone arms, turntables and 2. Optimum lateral tracking alignment even other components in either the commercial of commercially available pivoted arms can or the underground audio publications can be only be achieved by ignoring the manufac­ expected to be quite erratic. Alignment accord­ turer's instructions and twisting the cartridge ing to the manufacturer's instructions will in the headshell to obtain the correct combin­ randomly favor or handicap A over B; after ation of overhang and offset angle. All other proper alignment the order of preference is approaches, including the one suggested in our often reversed. We now believe that this is fi rst article, are unnecessary compromises. one of the principal causes of disagreement 3. Vertical tracking error (i.e., a diffe r­ among otherwise unbiased and keen-eared ence between the original vertical cutting angle equipment reviewers as well as rank-and-file and the effe ctive vertical playback angle) is audiophiles. similarly audible within the limits of the fi nest This is admittedly pretty heavy stuff, and adjustments that can be made by hand (such we don't expect it to penetrate some of the as barely raising or lowering the vertical pivot pointier heads in high places for some time of the arm, shimming up the record by means to come. The consequences aren't exactly of a thin cardboard disc or another thin record, negligible, when you stop to think about them: etc.). Without sophisticated test equipment, all pivoted tone arms should be redesigned correct vertical tracking angle can only be and their mounting instructions revised (and determined by trial and error coupled with even then a correctly executed radial-tracking careful listening. Visual guidelines, such as arm will be audibly better-but try to fi nd the apparent forward tilt of the stylus, the one!); fu rthermore, all tone arms, whether angle between the stylus bar and the record, pivoted or radial-tracking, should have a

50 mechanism whereby the vertical tracking angle at the time we wrote our original article. But can be conveniently changed during play; or it just isn't so. The most important distortion better yet, the recording industry should im­ that results is time distortion. The playback mediately adopt an ironclad vertical tracking stylus is ahead of real-time reproduction on angle standard and enforce it internationally. one wall and behind on the other wall. And the Are you beginning to see the dimensions of ear is considerably more sensitive to that kind the problem? All we can accomplish here is of time smear than to harmonic distortion. to make a few small waves and hope that some­ Now think about the vertical motion of body out there gets splashed a little bit. Any the cutter stylus. For practical reasons that kind of serious dialogue on the subject within don't really concern us here, this motion isn't the audio community would require the partici­ perpendicular to the plane of the disc but at pation of some of the largest commercial an angle to it, the line of motion leaning in interests, and Nick the Greek isn't putting the direction of the groove already cut. Now heavy odds on that probability. any up-and-down motion at angle also has Meanwhile, we'll try to illuminate our a back-and-forth component. On the down main points with some simple explanations stroke, the cutter stylus is adding a tiny and also provide specific instructions for amount to the linear speed with which the optimizing your cartridge/arm/turntable set­ groove is cut; on the up stroke it subtracts up. Remember, though, that this is only Part a tiny amount. Aha, you say, time distortion I of what we plan to be a continuing explor­ is being cut permanently into the groove. ation of the subject, so that a number of Exactly. But what if the playback stylus is set questions may not be resolved to your satis­ to move up and down at exactly the same angle fa ction until next time. as the cutter stylus? Then its own velocity modulation will be synchronized exactly with that of the original cut, and the signal will The stylus/groove interface. be reproduced in real time. Again, we must The key to understanding the requirements repeat that the ear is enormously sensitive to of distortion-free record reproduction is a the time domain (see also our loudspeaker clear and uncomplicated mental image of the survey in this issue) and that you don't easily stylus riding in the groove. All the facts and get away with any kind of fu dging there. all the nonsense you've ever heard about car­ So there's your picture: the playback tridges, tone arms and turntables relate, more stylus is riding in the groove, mimicking the or less directly, to that single interface. If motion of the cutter stylus exactly and in all you think about just that and nothing else­ dimensions. That means (I) it must be fi rmly if you don't get bogged down in peripheral seated without any loss of contact whatsoever; considerations such as, say, bearing friction (2) its points of contact with the left and right -everything falls into place and the subject wall must both be on the same radius of the is no longer open to techno-freak obfuscation. disc; and (3) its vertical motion must be at Visualize the cutter stylus, plowing a exactly the same angle fr om the perpendicular stereo groove into a lacquer master. Its lateral as was the cutter's. You can ensure (I) by set­ motion is from side to side, exactly in line ting the vertical tracking fo rce as high as pos­ with the radius of the disc at the point of cut­ sible, as long as the stylus cantilever doesn't ting. It's obvious, therefore, that the lateral collapse. (Contrary to untutored opinion, this motion of the playback stylus in the groove will not shorten the life of the record as long must be along that same radius. If it isn't-if as the stylus tip is in good shape and the car­ one wall is contacted ahead of where the cutter tridge design isn't the kind that permits ultra­ stylus was at a given instant and the other sonic hammer blows by the stylus on the groove wall is contacted behind of where the cutter wall.) As for (2), that's what lateral tracking stylus was at that same instant-there will be alignment is all about, and in a moment we'll a fa lsification of the information that was cut give you the most practical and effective align­ into the groove. This is commonly called lateral ment instructions available anywhere. When it tracking error and the most serious conse­ comes to (3), you're pretty much on your own, quence traditionally attributed to it is second but it isn't a completely hopeless situation, as harmonic distortion. We, too, accepted that we shall see. 51 First, however, we must point out how topological) explanation covers everything. useful this mental image of the correct three­ dimensional stylus/groove interface can be in explaining and understanding design require­ Can we really hear all this? ments for cartridges, arms and turntables. Our recent experiments with fanatically For example, why is a biradial stylus, such as careful alignment to optimize both lateral and the Shibata, superior to a conical stylus? vertical tracking geometry have produced as­ Obviously because the Shibata's longer and tonishing results. Going from routinely con­ narrower areas of contact with the groove scientious mounting of the tone arm and mimic much more closely the geometry of the cartridge, by fo llowing the manufacturer's cutter stylus and permit a fi ner resolution instructions unquestioningly, to optimum align­ of the time-related details inherent in the cut. ment, in accordance with the procedures out­ But then why does the conical stylus have so lined below, made in every case a greater differ­ maI1'Y partisans who claim it sounds better than ence in audible quality than changing to an­ the Shibata? Because it does sound better other good cartridge, preamp or power amp! when the lateral and vertical alignments haven't The sound just opens up and assumes a totally been properly attended to. The Shibata resolves new character-and the listener's mind reels. the consequent deviations from time coherency In fact, we now recommend that you perform much more precisely and audibly, whereas the these alignments before you change any com­ conical stylus glosses over them to some extent. ponent in your system, otherwise you'll never A properly aligned Shibata, on the other hand, know exactly what kind of sound you're giving eats a properly aligned conical fo r breakfast up-and switching to. Where this leaves the when it comes to resolution of detail. typical listening tests performed in audio Or take antiskating compensation. Why stores, manufacturers' sound rooms, and equip­ is it so important? Because the skating fo rce ment reviewers' homes we'll leave to your deflects the stylus and creates tracking error imagination. All we can say is that, whenever (i.e., alters the stylus/groove geometry). The someone now tells us that he found component usual explanation that it creates unequal stylus A to be superior to component B after extended pressure on the two groove walls and therefore listening to reference records, the fi rst thing unequal indentation is only part of the story. we ask is, "How did you align your tone arm And, for the same reasons, too much antiskat­ and cartridge?" If the answer is a blank stare ing bias is just as bad; it's simply a reversal or "Huh?" or "Just like it says in the instruc­ of skating force in the opposite direction. Now tions!", we say thank you and change the think about tone arms. Why should the arm be subject. (And then we thank our lucky stars absolutely rigid? Because any kind of flexing that our initial insight into the importance or axial "pumping" will shift the stylus with of all this came before our preamp survey was respect to the groove. And why should the arm too far under way.) be nonresonant? Because if it can store and When you begin to think about the dimen­ release energy in the frequency range relevant sions of the microworld in which the stylus to the groove, at least part of the energy will and the groove operate, none of the above is end up driving the stylus in a way that's com­ the least bit illogical. The linear speed of pletely unrelated to the stylus/groove geom­ the groove, at 33 1/3 RPM, varies between etry. What about turntable rumble or acoustic 51 cm/sec and 21 cm/sec from the maximum feedback? Same thing-the stylus will be dis­ to the minimum radius of the modern 12-inch located with respect to the groove with some LP. Let's take a nice frequency like 4 kHz, kind of low-frequency periodicity. On the where the sensitivity of the human ear is at other hand, why is pivot bearing friction a its peak. One full cycle of 4 kHz therefore less important consideration? Because as long occupies anywhere between 0. 13 and 0.05 mm. as it isn't large enough, relative to the vertical (Note that The Audio Critic is going metric; tracking force and the stylus compliance, to this is as good a place to start as any.) Let's apply a bias to the stylus and thus alter the take 0. 1 mm as a convenient average value. stylus/groove geometry, it can't possibly have That's 100 microns. Now let's assume that a an effect on the reproduced signal. And so 5% time modulation of the signal at this fre­ fo rth and so on: the geometrical (or call it quency (in other words, a time smear of ± 12.5

52 uS) can blur the focus of a stereo signal to There was something else in our fi rst some extent. (This is not an unreasonable article that needs emendation. We erroneously figure to assume.) You're then dealing with a credited Benjamin B. Bauer and Dr. John D. 5-micron linear portion of the groove. That's Seagrave for the original work on the mathe­ only 50,000 angstroms; you're approaching matical analysis of tracking error. Actually, optical dimensions! When you consider that their articles added nothing substantially new the stylus overhang specified in the mounting to the considerably more detailed and profound instructions of the Grace G-707 is 15 mm in­ study of the subject by H. G. Baerwald, an stead of the correct 17.5 mm and that the obviously brilliant man who worked fo r The unmodified offset angle of the arm is off by Brush Development Company in Cleveland, more than 10 from optimum (and that this Ohio, and published his massive paper on happens to be one of the geometrically best­ tracking error and optimal pickup design in optimized arms on the market !) you begin to the December 1941 issue of the Journal of the realize that we aren't just whistling Dixie here. Society of Motion Picture Engineers, years There can exist a serious non synchronism ahead of anyone else. This is the classic work between the recorded and the reproduced and the most complete solution of the mathe­ signals. matical problems posed by the subject; the computer program that resulted in our new alignment table is based on the Baerwald How to optimize lateral tracking geometry. equations. We're publishing a completely new table When you look at the table, please note of lateral tracking alignments here, simpler that as the arm length is changed, the opti­ and more useful than the Universal Design mum overhang changes very, very gradually, Graph in our fi rst issue. Not that the latter whereas the optimum offset angle changes was in any way incorrect; in numerical data much more rapidly. This is the most important it diffe rs fr om our new table by less than the thing to understand in order to develop a good least possible measurement error. The main feel fo r the alignment procedure; it explains, diffe rence is that the maximum and minimum among other things, why you can't just move radii of the recorded area assumed in our the cartridge straight back and forth to trim table are the official IEC values, standard on in the overhang, no matter how small an adjust­ all Neumann and similar mastering lathes. The ment you're making. graph, which predates the stereo ((ra by several Let's track through the essential steps years, was based on very slightly different in using the table. limiting radii; fu rthermore, it didn't contem­ First, install the cartridge in the arm, but plate the possibility of twisting the cartridge don't tighten the screws all the way. Allow in the headshell, so that it also provides cor­ the cartridge to be just barely movable when rectly matched overhang values fo r other than pushed. For openers, you might as well locate optimum offset angles, an audible compromise the stylus tip approximately where the instruc­ that turns out to be quite unnecessary. tions tell you to. N ow measure the effective arm length, Editor's Note: We've been informed by Mr. fr om lateral swing axis to stylus tip, with an Warren B. Syer, the publisher of High Fidelity, accurate ruler, such a machinist's scale. You that our Universal Design Graph was a repro­ don't have to agonize over this measurement, duction of page 23 of the August 1957 issue of but try to get it right within 0.5 mm (0.02 in.). the now defunct Audiocraft magazine, with Look up the optimum overhang corres­ copyright still owned by High Fidelity. Need­ ponding to this arm length in the table. You less to say, we committed this infringement may interpolate without any fear of inaccuracy. unwittingly; a photoprint of the graph had been (For example, if the effective arm length is in our files for all these years, with only hand­ 238.5 mm, take the halfway point between the written notes to relate it to the original Sea­ 17.398 and 17.319 mm overhangs, i.e., 17.3585 grave equations. Our apologies to High Fidelity ; mm, which you'll never measure more ac­ our acknowledgment to Dr. John D. Seagrave, curately than 17 1/3 mm.) the author, and to Audiocraft (may it rest in Set this overhang (measured from the peace). stylus tip to the center of the turntable spindle

53 Table of Optimum Overhang and Offset Angle Alignments for Pivoted Tone Arms

Optimized fo r a 12-inch LP record with a recorded area between the lEe Standard maximum and minimum radii of 146.050 mm (5.750 in) and 60 .325 mm (2.375 in). Zero tracking error in all cases at radii of 120.90 mm (4.76 in) and 66.04 mm (2.60 in). For in-between values of arm length, interpolation error will be smaller than the least possible measurement error.

Effective Optimum Optimum Effective Optimum Optimum Arm Length Overhang Offset Angle Arm Length Overhang Offset Angle

mm (in) mm (in) 0 mm (in) mm (in) 0 200 (7.874) 21.055 (0.829) 27.854 238 (9.370) 17.398 (0.685) 23.118 20 1 (7.913) 20.938 (0.824) 27. 704 239 (9.409) 17.319 (0.682) 23.06 1 202 (7.953) 20.822 (0.820) 27.555 240 (9.449) 17.241 (0.679) 22.914 203 (7.992) 20.708 (0.815) 27.408 24 1 (9.488) 17.164 (0.676) 22.814 204 (8.03 1) 20.595 (0.81 I) 27.262 24 2 (9.528) 17.088 (0.673) 22.714 205 (8.07 1 ) 20.483 (0.806) 27.118 24 3 (9.567) 17.012 (0.670) 22.616 206 (8.110) 20.373 (0.802) 26.976 244 (9.606) 16.937 (0.667) 22.518 207 (8. 1 50) 20.264 (0.798) 26.835 245 (9.646) 16.863 (0.664) 22.421 208 (8. 1 89) 20. 156 (0.794) 26.696 246 (9.685) 16.790 (0.661 ) 22.325 209 (8. 228) 20.049 (0.789) 26.558 247 (9.724) 16. 717 (0.658) 22.230 210 (8.268) 19.944 (0.785) 26.422 248 (9.764) 16.644 (0.655) 22. 135 21 1 (8.307) 19.839 (0.78 1) 26.287 249 (9.803) 16.573 (0.652) 22.042 21 2 (8.346) 19.736 (0.777) 26. 153 250 (9.843) 16.502 (0.650) 21.949 213 (8.386) 19.634 (0.773) 26.021 25 1 (9.882) 16.43 1 (0.647) 21.857 214 (8.425) 19.533 (0.769) 25 .89 1 252 (9.921) 16.362 (0.644) 21.766 215 (8.465) 19.433 (0.765) 25 .762 253 (9.961 ) 16.293 (0.64 1 ) 21.675 216 (8.504) 19.332 (0.761) 25.634 254 (10.000) 16.224 (0.639) 21.586 217 (8.543) 19.237 (0.757) 25.507 255 ( 1 0.039) 16.156 (0.636) 21.497 218 (8.583) 19. 140 (0.754) 25.382 256 ( 1 0.079) 16.089 (0.633) 21.409 219 (8.622) 19.044 (0.750) 25 . 258 257 (10.1 18) 16.022 (0.63 1 ) 21.321 220 (8.661) 18.949 (0.746) 25 . 135 258 (10.157) 15.956 (0.628) 21.235 221 (8.701) 18.856 (0.742) 25 .013 259 (10.197) 15. 890 (0.626) 21. 149 222 (8.740) 18.763 (0.739) 24.893 260 ( 1 0.236) 15.825 (0.623) 21.064 223 (8.780t _ 18.67 1 (0.735) 24.7,74 26 1 (10.276) 15.76 1 (0.621 ) 20.979 724 (8.819) 18.580 (0.731) 24.656 26 2 (10.315) 15.697 (0.618) 20.895 225 (8.858) 18.490 (0.728) 24.539 263 (10.354) 15.633 (0.615) 20.812 226 (8.898) 18.40 1 (0.724) 24.423 264 (10.394) 15.570 (0.613) 20.730 227 (8.937) 18.313 (0.721) 24.309 265 (10.433) 15.508 (0.611) 20.648 228 (8.976) 18.225 (0.718) 24. 195 266 ( 1 0.472) 15.446 (0.608) 20.567 229 (9.016) 18.139 (0.714) 24.083 267 (10.512) 15.384 (0.606) 20.486 230 (9.055) 18.053 (0.71 1) 23.971 268 (10.551) 15.324 (0.603) 20.406 23 1 (9.094) 17.969 (0.707) 23.861 269 (10.591) 15.263 (0.601) 20.327 232 (9.134) 17.885 (0.704) 23.752 270 (10.630) 15.203 (0.599) 20. 248 233 (9.173) 17.801 (0.701) 23.644 27 1 (10.669) 15. 144 (0.596) 20. 170 234 (9.213) 17.719 (0.698) 23 .537 272 (10.709) 15.085 (0.594) 20.093 235 (9.252) 17.638 (0.694) 23.43 1 273 (10.748) 15.026 (0.592) 20.016 236 (9.291) 17.557 (0.69 1 ) 23.325 274 (10.787) 14.968 (0.589) 19.940 237 (9.331) 17.477 (0.688) 23.221 275 (10.827) 14.91 1 (0.587) 19.865

54 and coinciding with the effective arm length) Mr. R. R. Mills, the managing director of J. H. either by moving the cartridge in the head­ Reproducers in Australia, took us to task fo r shell (in which case you'll change the arm not recommending in our fi rst article a dual­ length slightly) or by moving the entire arm. purpose optimization for both LP's and 45- The overhang can be measured with a short RPM doughnut singles. We think that's a and narrow machinist's scale or ruler; a good preposterous idea: spending in the upper hun­ way to do it is by measuring up to the outside dreds of dollars for a fine cartridge/arm/turn­ of the spindle only and then adding the spindle table combination and then getting more dis­ radius, which is half of the standard spindle tortion than necessary on your best LP's just diameter of 7. 163 to 7.214 mm (0.282 to 0.284 so that the bubble-gum cheapies can be played in). Thus, in the example above, you'd set the with a little less distortion. Pure LP optimiza­ stylus tip 133,4 mm from the outside of the tion is barely good enough in our opinion; spindle. optimizing fo r the much smaller minimum Now comes the crucial part. Prepare a radius of the recorded area on 45's carries piece of cardboard (a fi le card of the right size the pivoted-arm compromise a bit too far fo r will do) by punching a spindle hole in it and any serious audiophile's comfort. marking off radial distances of 66.04 mm (2.60 in) and 120.90 mm (4.76 in). You could also What about the vertical tracking angle? do this with one of those convenient (and in­ Here the going gets a bit rough. We fi nd correctly marked) paper protractors fu rnished that deviations of as little as 20 minutes of arc by tone arm manufacturers. Then twist the fr om the correct vertical tracking angle are cartridge in the headshell so that you can see audible, yet there are modern LP records, absolutely no tracking error at those two none of them more than ten years old, cut with points. Keep playing with the cartridge until angles as small as 15° and as large as 18.5°. Go you get this right. The offs et angle will then back a little fu rther in time and the spread be automatically correct fo r your overhang. becomes 13° to 22°. In the earliest days of After that, set the vertical tracking force stereo there were even a few records cut with to the highest value allowed in the cartridge a negative vertical tracking angle. (You'd specs and adjust the antiskating bias. The have to twist your cartridge around 180° and easiest and most accurate way to do this is by reverse your left/right leads to play those cor­ ignoring the bias calibrations on the arm (ex­ rectly!) And the angle is never, never labeled cept as a general guideline) and setting the bias on the record. What do you do? by observing the stylus with a small magnifier We suggest that you align your cartridge/ or loupe. When tracking a low-level passage, arm vertically to make the latest state-of-the­ there should be absolutely no difference in the art records cut with the Neumann SX 74 head centered (neutral) position of the stylus on or sound their best. That's probably a 17.5° or off the record . I f you see the tiniest snap as 18° cut, and you're unlikely to run into anything you lift the cartridge off the record, the bias is higher. The Mark Levinson records are a very incorrectly set. good example. After having performed every And now comes the most annoying part. step of the lateral tracking alignment, put on A fter you've done all this, everyth i ng is such a record and play with the arm height ad­ probably slightly off, since all these align­ justment until the sound can't be made any ments interact. The antiskating adjustment cleaner and more transparent. Generally speak­ could quite conceivably have misaligned the ing, if the sound is hard and overbright, you've two zero-error points, in which case you'll raised the arm too much; if it's muddy and mid­ have to trim those in again. Then the over­ rangy, you've lowered it too much. Try to get it hang will probably be off, and the effective just right; it isn't so hard with good associated arm length in turn. Keep going around in a equipment-the right alignment almost clicks circle, fr om one to the other, until everything into place. And don't look at the arm tube. Its is right on the button and in accordance with position is meaningless except as an indication the table as well as the alignment card. The of where you've been and where you're going. end result will be worth the trouble. Once your vertical tracking alignment is Please note that these alignments optimize set fo r the largest angles currently used in the your record player fo r 12-inch LP records only. business, you're ahead of the game, since you

55 can always make the angle smaller by shim­ yet. We're still looking for equipment that, ming up the record (say, with another record) right up front, doesn't seem to threaten the but you can make it larger only by readjusting stylus/groove relationship in any way what­ the arm height. You can even make a series soever. of calibrated cardboard discs and mark on the Nor do we want to make any broad gen­ album cover how many of them you need for eralizations, at this juncture, about design shimming up that particular record. Just don't criteria for the ideal tone arm or turntable, give up because the whole thing is too much other than the integrity of the stylus/groove bother. Once you've developed a routine, you'll geometry itself. Under each particular model, waste very little time and you'll enjoy your we're commenting on whatever specific design record collection a great deal more. Remember, problem that unit poses; when the time comes, the diffe rence isn't trivial. after we've reviewed a suffi cient number of different designs, we'll summarize the lessons A word about our tone arm to be learned. and turntable reviews. Before we get into the reviews, however, Phono cartridges are very special active we'd like to bring up one amusing example, devices and will be getting very special consid­ which will serve to illustrate some of the tech­ eration beginning in Part II of this series. nical points we're trying to make. Somebody Tone arms and turntables, on the other hand, we know made a truly superb tone arm some are passive vehicles for the cartridge; their time ago for a special laboratory project. It performance is defined by their mechanical consisted of a plastic soda straw, pivoted on excellence as tools or machinery. Once you fo ur ordinary household pins in a gimbals con­ understand the stylus/groove interface, the figuration. Consider this ultrasophisticated de­ mystique of purely subjective testing ("the vice: it had very low mass, great axial rigidity, highs are grainy and the midrange is hooded") low Q because of the relative deadness of the begins to appear singularly inappropriate to plasticized material, only the correct two de­ tone arms and turntables. Not that their ulti­ grees of freedom in the gimbals (not able to mate proof of quality isn't in the sound. Of twist every which way like a unipivot), quite course it is. But you don't have to test a ham­ adequately low bearing fr iction at the four mer that has a five-inch handle by vigorously points of suspension, and of course no silicone hammering nails with it for two days. You say, gunk or other damping at the pivot to modulate "This hammer doesn't have a decent handle; the vertical tracking force on warps. Of course, give me another one." Since the world of tone the experimental cartridge it was made for was arms and turntables is fu ll of hammers with extremely low in mass and high in compliance; five-inch handles, simple inspection often suf­ for your normal cartridge the straw could be fices to make a "first cut" in a test program replaced with, say, a straight section of balsa whose sole concern is fi nding the best. In some wood or some other heavier material that's cases a few minutes of use will reveal all the rigid and dead. How about that, Dynavector? necessary information. In other cases it may Oh yes, just one more thing. The Duxseal take weeks or months. we keep referring to in the reviews is the We've taken this completely pragmatic, acoustically deadest material known to us, common-sense approach to evaluating tone originally made as a sealant fo r plumbers, arms and turntables. It's entirely possible that electricians, appliance installers, etc., by J ohns­ in the end a few supersubtle fi nalists will require Manville. You can buy it fr om suppliers to that careful A-B listening comparisons under con­ trade. It will damp some of the most stubborn trolled conditions. We're far from that point resonances.

56 carbon fi ber, balsa wood or even silicone gunk), Dual CS721 which would be definitely more desirable. We aren't suggesting that this is a totally wrong­ headed solution, given the design constraints of a very compact, practical, easy-to-use, auto­ United Audio. 120 South Columbus Ave. . Mount matic turntable at a price-but a purist's ap­ Ve rnon. NY 10553. Dual CS721 automatic single-play turntable. $400. One-year warranty; distributor pays proach it isn't. return fr eight. Tes ted #052570. on loan fr om distributor. The turntable itself is also of basically good design, without disqualifying faults. We'd like to see a platter that doesn't ring at all This is Dual's top-of-the-line turntable/ when struck; the Dual's isn't quite dead enough. arm model, with electronically controlled direct­ The resonant frequency of the chassis on its drive DC motor. We don't fi nd anything seri­ suspension could also be a few Hz lower fo r ously wrong with it, except that the combina­ best results. All in all, though, this is a com­ tion of the Kenwood KD-500 turntable and petent, well thought-out job, and it doesn't Grace G-707 tone arm is even better and costs appear to suffer from the horrible things direct­ $60 less (but isn't automatic, of course). Other­ drive turntables are sometimes accused of, wise the Dual CS721 would be our current top such as cogging. recommendation as the record player fo r a The only thing we can't figure out is where medium-priced stereo system. the CS721 fits into the audiophile scheme of Dual is particularly proud of the arm, and things, since it's neither the ultimate nor (in it isn't a bad arm at all. The bearings are view of Kenwood/Grace) the best fo r the deployed in an ideal gimbals configuration and money. How about best fo r the money if you appear to be correctly preloaded. We could must have an automatic? detect no play or wobble whatsoever. The arm tube is straight, rigid and quite dead. The offset angle of the nonremovable headshell is almost correct, requiring minimal twisting of the cartridge fo r fully optimized alignment. With those advantages, you're already way ahead of Dynavector DV -505 the game. We also like the continuously adjust­ able arm height (i.e., vertical tracking angle) fe ature, which makes this all-important align­ ment truly painless by means of a simple Onlife Research. Inc .. To kyo. Japan. Distributed in the U.S.A. by A udioanalys t. Inc .. PO Box 262. Brookfield. knurled knob control. Now if only the control CT 06804. Dynavector DV-505 tone arm . $575. tested could be used during play ...but no such luck. #510880. owned by Th e Audio Critic. The two weak spots of the arm, in our opinion, are the headshell and the counter­ weight. The mechanism that clips the remov­ Since we've already discussed this "ulti­ able cartridge holder into the fixed headshell mate" arm (its name in Japan is Ultimo) at isn't quite positive enough; there remains the some length in two previous issues, we only possibility of not only play but also a lack of want to summarize our conclusions here. positional repeatability under worst-case condi­ Despite its sci-fi appearance and mind­ tions (such as frequent removal and reinsertion). boggling price tag, it isn't the ultimate arm. You can imagine how much that disturbs a It suffers both fr om errors of design and from tracking-error fanatic. Nor is the entire head­ errors of execution. The massive, heavily shell assembly dead enough for our taste. As damped main arm, pivoted to move only later­ fo r the counterweight, we question the anti­ ally, and the much lighter, undamped subarm, resonant mechanical filters built into it for the pivoted to move only vertically, constitute purpose of tuning out arm and chassis reson­ an erroneous design concept in our opinion. ances. The jellylike bobbling of the counter­ Regardless of the convoluted rationale pre­ balance breaks up a resonant peak into several sented in Onlife's technical brochure, such a smaller ones Uust a few dB lower in ampli­ gross difference between the lateral and vertical tude) instead of dissipating it resistively (like motional impedances will tend to resolve vec- 57 tors with a strong upward bias, thus falsifying frequent cartridge changes are a nuisnce. And, the information cut into the groove. Further­ needless to say, you still have to twist the more, the sensitivity of the short subarm to cartridge in the headshell every time, since the warp wow is inevitably greater than that of any lateral geometry still isn't right on the money. normal arm. Interestingly enough, this is also a straight­ The greatest boo-boo in the physical exe­ tube, gimbals-suspension design, like the Dual cution of the DV -505 is the way the arm height arm, and there's really no better approach. adjustment mechanism is slotted and keyed. The Grace is a little longer and a little simpler, Every time you change the arm height with this both of which are fundamental advantages; its beautifully convenient device, the play between headshell and arm tube seem to be reasonably the key and the slot results in a shift of the dead, although two little dabs of Duxseal (one lateral swing axis of the arm, completely negat­ between the cartridge screws on top and one ing the overhang alignment. Once you're aware on the arm fairly close to the pivot) resulted of this unbelievable defect, you can push the in an ever-so-slight increase in sonic clarity, damn thing back to the end point of the play, proving that there was room for improvement hoping that at least that point is stable. But even in that department. The slight extra mass for $575? Come on, Onlife. also brought the G-707 right in line with the We think we have figured out the reason compliance of our Denon DL- 103S cartridge; why this arm so often appears to improve the overall, however, the arm appears to be accept­ sound when inserted into a known system. The ably compatible in mass with a wide range of unusually large mass of the entire structure modern cartridges and is certainly not too tends to act as an anchor or inertial platform, heavy for any that we know of. As for the bear­ dissipating some of the grosser extraneous ings, they appear to be eminently free from disturbances of the stylus/groove relationship play and wobble, which probably accounts for that affl ict the typical installation. As soon a significant part of the arm's audible per­ as purposeful and meticulous care is taken to formance. eliminate these disturbances (better mechanical We could wish for generally more rugged and acoustic feedback isolation of the turn­ construction, a more convenient antiskating table, for example, or a dab of deading material adjustment, a cueing mechanism with more such as Duxseal in the right places), correctly positive feel, etc., etc.; but the fact remains designed arms such as the Grace G-707 begin that this is the arm we're currently using in to sound better. It's as simple as that. our reference system, even though we can't help There are more ways to skin a cat than wondering why there isn't this kind of intelli­ hanging a $575 anchor around its neck. gent design available along with SAEC-like construction-for those who are willing to pay the price.

Grace G-707

Sumiko Incorporated. PO Box 5046. Berkeley. CA 94705. Grace G-707 tone arm. $140. One-year warranty; customer pays all fr eight. Tested sample owned by Grace G-940 The A udio Critic.

Sumiko Incorporated. PO Box 5046. Berkeley. CA We consider this simple, ungimmicky arm 94705. Grace G-940 tone arm. $150. One-year warranty; customer pays all fr eight. Tes ted sample owned by to be virtually unexceptionable in design; re­ Th e Audio Critic. fi nements would be possible, of course, but mainly in execution rather than concept. It's the best all-around workhorse arm we've been This isn't exactly our cup of tea: a uni­ able to find so far. You must keep in mind, pivot design (i.e., free to move in various un­ however, that it has a fixed headshell, so that desirable ways that would be impossible with

58 a gimbals-type suspension); fluid-damped at The overall construction of the ST-7 the pivot, which is far from the ideal place; seemed much too flimsy to us from the begin­ and without antiskating bias, so that there will ning, but this we didn't expect. Next! be lots of tracking error (among other things), no matter how carefully the overhang and offset angle are aligned. What's more, the arm is on the heavy side for some of the latest high-compliance cartridges. The fact that the G-940 still sounds a lot better than most other arms tells a great deal Infinity 'Black Widow' about most other arms. Infinity Systems, Inc., 7930 Deering Avenue, Canoga Park, CA 91304. 'Black Widow' tone arm, $200. One­ year warranty; customer pays all fr eight. Tes ted #11216, on loan fr om manufa cturer.

Harman Kardon Aside from its Bongiornoesque name, we fo und this slender and not particularly lethal­ Rabco ST-7 looking tone arm quite livable with on first acquaintance, even though twisting the cartridge fo r optimum geometry in the tiny (and, of course, not quite correctly offs et) vestige of Harman Kardon, 55 Ames Court, Plainview, NY 11803. a headshell took some doing. Then we discov­ Rabco ST-7 straight-line tracking turntable, $430. ered that damping this half-shell as well as Tes ted #2423689 and #3000927, owned by The Audio the bearing housing with some Duxseal resulted Critic. in a substantial improvement in transparency and definition, so it's an inescapable conclu­ If this were a successfully executed prod­ sion that the arm rings to some degree. Closer uct, more than half of our preceding article, examination of the bearings indicates some including the overhang and offset-angle table, proneness to jitter; what's more, the little half­ would be unnecessary or at least academic. headshell can be plucked to twang ever so The ST-7 has a straight-line tracking arm, and slightly; both of which resonant conditions may that means no overhang and no offs et (and have been helped by the damping. therefore no skating fo rce, either). The ideal Overall, the arm seems even a little flim­ lateral geometry. And it's a fu lly automated sier than the Grace G-707, which fo r 30% less single-play turntable, with universal arm and offers substantially the same virtues without two-speed table lovingly made for each other the same vices. -obviously the ultimate fruition of the orig­ inal Rabco idea, right? Wrong. Our tests of two different samples were nipped in the bud by an absolutely intolerable and disqualifying defect common to both, which we must therefore assume to be a design Kenwood KD-500 characteristic. The carriage of the arm trans­ port mechanism is so loose and wobbly that the theoretically ideal stylus/groove relation­ Kenwood, 15777 South Broadway, Gardena, CA 90248. KD-500 direct-drive turntable, $199. 95. One-year war­ ship made possible by straight-line tracking is ranty; customer pays all fr eight. Tested #4 30148, on totally negated, indeed made worse than with loan fr om owner. just about any conventional arm. The result is a big resonance in the upper-bass/lower-mid­ range region, adding an insistent coloration Although we haven't tested some of the to the sound that we don't find at all accept­ top contenders yet (for example, the Linn able in advanced audio equipment. Sondek LP 12 and the Thorens TD-126 Mk II 59 are among those next in line), we doubt if we'll slab took care of that. (What do you think ever fi nd a better value among turntables than marble was invented for? Michelangelo?) this excellent and remarkably low-priced Ken­ Deep down we always knew, however, that wood unit. There really isn't all that much to there was something subtly wrong with the say about it, except that it's entirely trouble­ sound. A kind of thickness in the upper bass free and makes good cartridges and tone arms and lower midrange. We tried to blame every­ sound their best. For example, we like the thing-record, cartridge, arm, etc.-except sound of our Denon DL-1 03S/Grace G-707 this gorgeous machine. But the day of reckon­ combination much better on the Kenwood than ing fi nally came. The unspeakable truth: the on the Luxman, which costs three times as PD-121 drums. Like a real live drum. That much. large, fl at base just isn't dead enough. The The KD-500 is living proof that what main cause of excitation is airborne acoustic makes a turntable sound good isn't exotic feedback, but you can verify the character of engineering, super specs or fancy features, but the resonance by tapping on the turntable just simply a design that allows the stylus/groove so, with the record playing and the volume relationship to remain undisturbed at all times. control up. The coloration of the resulting The "resin concrete" chassis on which all parts undamped transient is the same as the color­ are mounted is probably a strong contributing ation we heard in the music. And the drum­ factor; it seems to be nice and dead. The platter ming isn't curable by any simple means. Good­ itself is rather live (it rings like a bell when bye, PD-l21. struck) and could probably benefit from a liber­ (And you thought it would be something ally applied layer of Duxseal underneath. really sophisticated, like cogging or some other (Unfortunately, we don't own ours, so we can't fashionable direct-drive vice, didn't you?) mess with it.) The adjustable legs of the Ken­ wood aren't the most sophisticated kind either, and we had to place ours on a marble slab to improve the mechanical feedback isolation. Furthermore, the direct-drive DC motor doesn't have much torque. But who cares? Properly set up, the Kenwood seems to be Mayware Formula 4 more stable and feed less crud to the cartridge and arm than any other turntable we've tried so far. M ayware. England. distributed in the U.S.A . by Polk As one of the old perfume ads used to say, Audio. 1205 South Carey Street. Baltimore. MD 21230. it's unfair but it works. Formula 4 PLS4/ D tone arm, $149.95. Tested sample owned by Th e A udio Critic.

This is a unipivot arm, with silicone damp­ ing applied right at the pivot (i.e., in series with the stylus compliance), and you know fr om our Grace G-940 review above how we Luxman PD- 121 feel about both of those design approaches. The Formula 4 also has a somewhat wobbly sliding weight between the pivot and the head­ Lux Audio of America. Ltd.. 200 Aerial Way. Syosset. shell, so that not only the vertical tracking NY 11791. Model PD-I2I direct-drive turntable. $595. fo rce but also the dynamic mass of the arm can Th ree-year warranty; manufa cturer pays all fr eigh t. Tested #E6901267 and #F6901268. owned by Th e Audio be varied. An interesting idea, but we aren't Critic. convinced that the slight play between this cursor weight and the arm tube isn't audible. This was our reference turntable fo r many N or that the very light, fi xed headshell doesn't months, mainly because nothing more precise ring. Nor that the ultrathin arm tube is rigid and beautiful had ever been built, and it worked enough. Nor that eccentric counterweight very reliably. All right, the mechanical feed­ doesn't set up some torsional forces that aren't back isolation wasn't the greatest, but a marble completely balanced out.

60 Since the Grace G-707 is a simpler and and probably is impossible with certain other more effective design, as well as quite com­ cartridge configurations. patible with the same high-compliance car­ And what do you know, the correctly tridges that the low-mass Formula 4 was aligned SAEC/Denon combination still didn't designed for, there's really no need to belabor work properly. The mass of the arm was much the above points. The Formula 4 also costs a too high; the subsonic resonance went way few dollars more and is even less rugged in down to the point where it coincided with construction. the slight warp excitations produced by the J. H. Reproducers Co., an Australian record, and the stylus bobbled like very slow manufacturer, claims that the Mayware jelly in the groove. Now the Denon isn't a Formula 4 is an inferior rip-off of their orig­ particularly high-compliance cartridge; it can inal design. We're reporting this merely as a live happily with some fairly massive arms. partisan allegation, since we haven't been able So it seems that the SAEC is designed only to get our hands on the J. H. version thus far for the very lowest compliances to be found to form our own opinion. Needless to say, how­ today. Of course who says it was designed at ever-once a fluid-damped unipivot, always a all? (Sorry. There we go again.) fluid-damped unipivot. One nice feature of the SAEC, though, is the optional arm stabilizer accessory (AS-SOO), a large cylindrical weight that threads right onto the arm post under the turntable. It's again the anchor or inertial platform concept, but a lot cheaper and more convenient than in SAEC WE-308 NEW the case of the Dynavector. Let's be thankful for a good idea even if we can't be for a good arm. Audio Engineering Corp., To kyo, Japan. Distributed in the U.S.A. by C. M. International, 4131 Calle de Prim era , Torrance, CA 90505. SA EC WE-308 NE W double-knife -edge tone arm, $195. Tested #FJT4955, owned by The Audio Critic. Recommendations

This incredibly frustrating combination of precision workmanship and geometrical blund­ It's much too early in the game to cate­ ering, which we can only liken to a superb gorize any tone arm or turntable as the best, Swiss timepiece with an II-hour dial, was even with qualifications. We just haven't looked already discussed in sufficient detail in our into enough of them. But at least two excellent CES report (May/June 1977, page 17) to con­ and relatively low-priced pieces of equipment stitute a review; we really don't feel like beat­ emerge fr om our investigation so far. To those ing up on these obviously dedicated though who are unwilling to wait for our reports in misguided people all over again. We must add, the next few issues, we can recommend them however, that we finally succeeded in twisting unhesitatingly. the Denon D L-l 03S cartridge into the correctly aligned position on the SAEC; if the offset Tone arm: Grace G-707. angle of the arm had been just half a degree smaller yet, it would have been impossible- Turntable: Kenwood KD-500.

61 Editor 's Note: Max Wilcox has been busy with an unusually crowded schedule of recording dates. His third article on why records sound the way they do has therefo re been postponed until the next issue. Meanwhile, we present our first record review column.

The Mark Levinson Recordings

Not too many audio enthusiasts and music there are fewer than twenty from coast to lovers are aware of Mark Levinson Acoustic coast. Broader distribution is being organized; Recordings Ltd., a company operated by the meanwhile, if you're unable to find them, we same music addict and compulsive audio per­ suggest you write directly to the company at fectionist as Mark Levinson Audio Systems 55 Circular A venue, Hamden, Connecticut Ltd. Even in stores where you have no trouble 065 14. They'll make sure you'll get what you finding copies of Sheffield, Audio Lab, Ark want. The price isn't cheap: $15 per record, and other audiophile-oriented labels, the un­ so that fo r example the "Art of the Fugue" al­ derstated gray jackets of MLAR are generally bum, consisting of four discs ( eight sides), will not visible. This low profile is unlikely to per­ set you back $60. Mark never believed in sist, however, since MLAR records are even giving it away. more clearly superior to others in their sonic But the records are beautifully made. The qualities than MLAS audio components. The recording equipment in each case consisted of latter have occasional competition; the former Bruel & Kj aer 4133 condenser microphones have very little. You could call them state­ (peak-free out to 40 kHz-we use one for of-the-art. And sooner or later you'll see them measuring speakers with) and a specially modi­ on all the more expensive turntables. fied Studer A80 tape deck recording on I-inch At the moment, these records are stocked tape (Y4-inch tracks) at 30 IPS. No Dolby, dbx only by Mark Levinson audio dealers, of which or any of that jazz. No console. Just straight

62 in. The resulting distortion levels, signal-to­ its fu ll glory, with low pedal notes such as you noise ratio and headroom exceed anything cur­ seldom if ever hear fr om a phonograph record. rently attainable in disc cutting and pressing, It's the von Beckerath organ at Yale, in­ so that the rigors of direct-to-disc recording cidentally; not the greatest baroque organ in become an irrelevance. It's the best of both the world, not even von Beckerath's best ef­ worlds, permitting tape editing as well as live­ fo rt, but beautiful just the same and beautifully quality cutting. recorded in actual performances before an Only two channels are used for the stereo audience, with just the right liveness. The information, so there's no possibility of con­ choir selections are also of you-are-there qual­ fused imaging as long as the mikes are set up ity; performances are musicianly but not thrill­ right (which they generally are). Two more ing. Every red-blooded audiophile should own channels may or may not be mixed in for am­ this record. bience information only. That's all. Never more than four mikes and four channels. The mixing takes place directly into the cutter am­ plifier via an utterly simple mixer box made Volume Two out of modules similar to those in the Levinson preamps. The cutter head is the incomparable Maurice Ravel: Va lses Nobles et Sentimentales. Joseph Neumann SX 74; the cutting is performed by Haydn: Sonata No. 49 in E-flat Major. Lois Shapiro, the incomparable Bob Ludwig. The discs are piano. pressed in France by the most careful outfit Mark could find for the purpose. (We heard Technically this is perhaps the least spec­ he was almost thrown out of there because he tacular of the Mark Levinson recordings, al­ pushed and nitpicked even these dedicated though the piano sound is super clean and the people a bit more than they liked. But he got peaks are never clipped. It's just that we'd pre­ what he wanted.) fer a more expansive, high-ceilinged ambience, So there you are. Absolutely quiet records, with greater depth and roundness. The music, tremendous dynamic range, no distortion to of course, is wonderful; Ms. Shapiro does an speak of, no peaky mikes, no multimiked especially stylish job on the sonata, which is a jumble. Just the perfectly natural sound of superb piece (late Haydn) and alone worth the some highly competent musicians. If these price of admission. same musicians were great instead of highly competent-if the organist were Walch a, the pianist Horowitz, and so fo rth-the experience would be simply overwhelming. As it is, it's Volume Three merely in the not-to-be-missed category. (Why can't Mark Levinson be the VP in charge of Bill Elgart: A Life (percussion piece in ten movements). production at Columbia or Deutsche Gram­ Performed by the composer. mophon?) We aren't fu lly sold on this composition as a work of art: two LP sides of drumming and percussing, in a rather tight and dry idiom, of­ ten suggesting speech rhythms (we think). Very controlled, deliberate, intellectual; not much Volume One exuberance. Edgard Varese and Henry Cowell are more varied and more fu n. As a recording, J.S. Bach: The Six Schubler Chorales (Myrtle Regier, though, it's mind-blowing-absolutely the high­ organ); Prelude in E-flat Major (Britt Wheeler, organ); three choir selections (Battell Chapel Choir, Charles est-fidelity percussion record known to us. The Krigbaum, director). drums, cymbals, bells etc., are there, right in front of you. Unbelievable transients and dy­ This is the only Mark Levinson recording namics, unbelievably quiet background, no grit not made on the Studer A80, yet it's one of the or crud whatsoever. Required demo material, best. Obviously, the IS-IPS l� -inch machine even if musically it doesn't turn you on. But used was a good one. The organ is captured in maybe it will.

63 Volume Four Volume Five

Vi valdi. Bach. Hindemith. Debussy. lves. Lennon/ Mc­ l.S. Bach: The Art of the Fu gue (Charles Krigbaum. Cartney. et al: music for brass or transcribed for brass. organ; 4 records). (New Haven Brass Quintet. 2 records.) Bach's last and definitive statement on This is one of our favorites: two trumpets, counterpoint, left in open score without a speci­ horn, trombone and tuba playing everything fied instrument or instruments, is performed but the kitchen sink, from sixteenth-century here on the same von Beckerath organ at Yale brass music to "Magical Mystery Tour" and as Volume One in the MLAR series. The "Penny Lane" as if arranged by Gabrieli! organist, Charles Krigbaum, is said to be a Four sides of gorgeous brass sound, guaranteed highly specialized student of this masterpiece, to make you miserably unhappy with the mid­ and his performance is certainly beautifully range of your system, unless it's super special. organized and note-perfect. We find his choice There's a transcription of a Charles Ives song, of registration a bit monotonous; he also plays "Slugging a Vampire," which is almost impos­ everything invariably legato. Even though a sible to reproduce except with the finestequip ­ true staccato is mechanically impossible on the ment from stylus tip to speaker diaphragm, in organ, the choice of lighter, more flutelike which case it sounds absolutely stunning. An stops, more detache phrasing and more distinct acid test. The playing is again expert but not ictus in certain passages would have made for sensational; the music ranges from good fu n livelier listening-in the opinion of this Bach to very beautiful and back again. Put this on lover who doesn't claim to be a scholar. The your HQD System and feel superior to the recording itself is everything an organ record­ Dahlquist/Am pzilla crowd, who will have ing should be, weighty and ultratrans parent trouble with it. at the same time, with a fine sense of acousti­ cal space. The music is so commanding, though, that this can't be considered a demo record (or records); it's frivolous to talk about audio in the presence of Bach's most profound utterances. Quite an album, and quite a Christ­ mas present for the highbrow audio freak. -Ed.

HELP STAMP OUT CHEAP PHILIPS- HEAD SCREWS IN AUDIO EQUIPMENT !

64 Ad1NSnitor

Editor's Note: We have a stack of letters telling us either that (a) this is one of our best fe atures and we must never stop bushwhacking the snake-oil peddlers or that (b) any audiophile with the brains of a gnat can see through ihe industry's advertising shenanigans and doesn 't need our heavy-handed protection. Th e reason why we're inclined to continue monitoring and admonishing (and occasional­ ly even commending) in this space is that hi-fi ads are a source of not only sp ecific product hypes but also general information about audio. Many of the simplisms, half-truthsand sp ecious dogmas that infest the high-end audio scene are recycled versions of statements originally made in ads. And the unwitting carriers of this infectious misinformation are often the same sophisticated consumers who can't be swayed directly by slick advertising and proudly proclaim their skep ticism.

Dual cates this real-life sound field: its fr ont-firing In our first Admonitor column, we com­ woofer /tweeter module gives you the first­ mended Dual for their informative, factual ad­ arrival sound and its angled, top-firingwoof er/ vertising of the advantages of straight tone tweeter module the second-arrival sound. Just arms (with offset headshells) over curved arms. like Carnegie Hall, see? If this had anything to do with their decision This seems to be the latest incarnation of to launch their current ads about "the world's The Great Bose Non Sequitur-the argument best-designed tonearm," we're truly sorry. that, since live music is a combination of direct That's the Marantz kind of language (recent and reflected sound, an accurate transducer Superscope era, that is). As you can see fr om must therefore be both a direct and a reflected our Dual CS721 turntable review elsewhere in radiation source. The absurdity of this line this issue, the Dual arm, good as it is, isn't the of reasoning becomes instantly apparent when world's best. Nor, apparently, is Dual's adver­ you consider that it makes the finest electro­ tising judgment. static headphones "inaccurate." All direct radiation into your ear, nothing reflected­ Epicure 20+ can't have any real fidelity, right? "Before you spend $600 on a pair of The fact is that a direct/reflecting speaker speakers, spend $15 on a ticket to Carnegie system such as the Epicure 20+ has to create a Hall," says the headline of Epicure'S current time smear, since the deliberately reflected ad fo r their Model 20 + speaker system. The components of its output reach the ear a split ad goes on to explain, in words and in pictures, second later than the direct components. It's that in Carnegie Hall or any other concert hall obvious that, on a record or tape, the sound re­ you hear a "first-arrival" sound directly from flected off the ceiling and the walls of the con­ the orchestra on the stage and then a "second­ cert hall is already recorded in real time with arrival" sound fr om the ceiling and other re­ respect to the direct sound of the orchestra. To fl ective surfaces of the hall. And, wouldn't hear it accurately, as captured by the micro­ you know it, the Epicure 20+ faithfully dupli- phones, you don't want it fu rther dispersed

65 in time, i.e., reflected all over again. When the And a little later: "We built a plant ...at ad says that "the Epicure 20+ successfully in­ a place called Kofu. Which is at the base of Mt. tegrates first-arrival and second-arrival sound," Fuji. Where we can get all the spring water we it really means that the speaker is stretching want." out an event longer than it took in real life. In Now really. How could a highly profes­ other words, the sound is blurred. sional outfit like Sony allow something like We could say, "But what did you expect this to get into print? It's precisely the made-by­ from the makers of the Model Four preamp li­ elves-in-the-Black-Forest kind of mystification fier?"-but that would be rubbing it in, so we that rots the mind of the untutored, gullible won't say it. consumer of technology. Get this straight: A complete set of mea­ surements will completely specify the physical Sony SSU-3000 and SSU-4000 characteristics of a speaker cone-or a cabinet Can a mass-producer of TV sets and tran­ -and therefore its sound. Complete, of course, sistor radios find happiness in the advanced means more than just frequency response. If a audiophile market? With their $1000 V -FET change in the water or the dye changes, say, the power amp and $900 turntable, Sony certainly mass of the cone by a small amount, or its stiff­ seems to think so, but one of their recent ness, the change may be audible-but you speaker ads really blew it. can measure it if you know what to measure. "After people learn what we've done, no And it doesn't matter what kind of glue holds one will heckle our speakers," proclaims the the cabinet together (it could be Clark Kent headline of a double-page spread advertising squeezing it with his hands); as long as the the $600-a-pair Sony SSU-3000 and the $800- joints can't move, the cabinet will sound the a-pair SSU-4000. Obviously the discovery that same. If they can move, there are measure­ the word "speaker" can have a double meaning ments that will show it. There's no mystery, no was both new and irresistibly delightful to the kitchen secret, no French chers fingertip­ ad writer, who then maintains the same level kissing je ne sais quai to making a cone or of authority and credibility throughout the ad. building a speaker cabinet. (See also the sub­ The part that really got to us, though, goes as woofer article in our second issue.) follows: Of course, it's entirely possible that if " You can hear how pure water is . The this ad were traced back from the ad agency purity of the water in which the pulp for the through the Sony ad manager to the engineer speaker cone is pressed will influence the sound. who provided the original briefing, the latter (Spring water is the best.) might protest that he never said these things, "But water purity would hardly change the at least not the way they got into print. That frequency response-or any other measurable would be quite typical of the advertising characteristic. process. "N or would the dye used to color the cone The fact remains, however, that the ad got -or the glue used in gluing the cabinet. through, these inanities were published, and "But you'd hear the dye and the glue." the harm is done. Sony-no baloney?

66 Oassified Advertising

Rates: For 25 cents per word, you reach everybody who DA YTON WRIGHT ELECTROSTATIC XG-8 Mk 3. is crazy enough (about accurate sound rep roduction) to Make reasonable offer. (4 12) 683-9550. subscribe to The Audio Critic. Abbreviations, prices, phone numbers , etc. , count as one word. Zip codes are DYNACO KITS: Stereo 300, $260; PAT-5, $160; SCA- fr ee (just to make sure you won 't omit yours to save a 50. Heath AJ- 1515 (wired), $300. Realistic 2000, $300; quarter). Only subscribers may advertise, and no ad fo r SCT- II. Only one of each new, warranties included. a commercially sold product or service will be accepted. Jonathan, Apt. 6M, 45- 10 Kissena Boulevard, Flushing, NY 11355. (212) 539-8060, evenings.

For Sale PARAGON 12 PREAMP, $650. 2 Rogers BBC Monitor PICKUP AND TRANSFORMER SALE! Ortofon (LS3/5A) speakers, $370. Janis W- I woofer, $475. MC-20, Supex 900E, ADC Mark II, AKG PE8S, Denon 103C cartridge with DB Systems head amp, $180. Fidelity Research II, WIN with power supply, Micro­ Good stuff, great condition. Chuck Cabell, 6223 Kellogg Acoustics 2002E, Audio Technica 20. Nakamichi MCB- Drive, McLean, VA 2210 1. (703) 356-470 1. 100, Levinson JC-IAC, Levinson JC- I DC. All barely used, guaranteed mint condition. Write best offers: D. NEED TUITION MONEY FAST. All equipment new. Karsch, 220 Madison Avenue, New York, NY 100 16. Mark Levinson JC-2D, $995. Ampzilla II, $785. Marantz lOB, $600. Allison I's, $625. Sondek with QUINTESSENCE PREAMPLIFIER, Thorens TD- Decca arm and Mk VI gold cartridge, Audio Research 126C, Yamaha HP- I 's. For any used component in mint SP-3A- 1 and D76A, $1650. A. Cohn, PO Box 17-203, condition, call (401) 521-2072. Bishops Corner, West Hartford, CT 06117. (203) 233-2007.

67 . / The End

68 Subscription Information and Rates

First of all, you don't need a subscription blank. If you wish to subscribe, simply write your name and address as legibly as possible on any piece of paper. Preferably print or type. Enclose with payment. That's all. Secondly, we will not send The Audio Critic any other way except by first-class mail in the U.S. or Canada and by airmail overseas. We simply refuse to be blamed and hated for the vagaries of postal delivery by cheaper methods. You're the one who pays for it this way, it's true, but you'll get much more reliable service. That means we have only twu subscription rates. If you live in the U.S. or Canada, you pay $28 fo r one year's subscription (six issues) by fi rst-class mail. If you live in any other country, you pay $33 fo r a six­ issue subscription by airmail. All payments from abroad, including Canada, must be in U.S. fu nds, collectable in the U.S. without a service charge. We strongly suggest that you start your subscription with Vol­ ume I, Number I, no matter when you start. That way you'll have a better understanding of what The Audio Critic is all about and you'll own a complete set of our tests and recommendations. If you insist, however, we'll start your subscription with any issue you desire. You'll still get six issues for your money. Simply state which issue you wish to start with. If you don't specify which one, we'll assume it's Volume I, Number I and send you all issues in print at the time you subscribe. One more thing. Please don't write us asking for a fractional sub­ scription, whether it's for a single issue, three issues or whatever. The answer will always be no. Our rule is a fu ll six-issue subscription or nothing. You're fu lly protected, however, by our policy of refunding the unused portion of a canceled subscription.

Address all sUbscriptions to The Audio Critic, Box 392, Bronx­ ville, New York 10708. The Audio (title

In the next issue: The speaker survey continues, the plot thickens, and the end is nowhere in sight. We begin our investigation of headphones, with special attention to electrostatics. We bring our power amplifier survey to an end (at least for the moment), firming up the conclusions of our laboratory tests. A whole new batch of preamps, including some inlportant ones. (Sorry-they keep coming !)

And� for the first time, phono cartridges, along , with more arm and turntable reviews. Not to mention the usual extras.