A Review from theAbsolute Sound
There is an old proverb in the profession that a great engineer is one who can do for a dime what anybody can do for a dollar. This thought comes to mind when one has had a listen to the Marsh Sound Design A400S amplifier, even right down to the ten-to-one ratio. We are all accus tomed to the idea that there are a good many 200- watts-per-channel amplifiers that sound really good and cost around $20,000. But an amplifier with that power that can take a legitimate place among the best and costs $2,000? That is great engineering. What has to be involved here is, first of all, power of mind. The circuit in an amplifier at this price can not solve problems The electronic expertise shown here will not be a surprise to people who are familiar with the career of designer Richard Marsh,* over decades. Marsh has only recently headed out on his own, with Marsh Sound Design. But those of us who keep track of the behind-the-scenes of electronics have known about his work for a long time. He was one of the pioneers in analyzing how capacitor defects affect audio adversely, and in solving the problem, both via improving capacitors and via eliminating them from circuits. (The A400S has no capacitors in the signal path, being DC-coupled, servo-controlled.) Marsh also contributed to some classic designs, notably the Hafler XL-280, the amplifier that started the differential-output test controversy, which should have eventually resulted in a reformulation of review concepts and procedures for amplifiers. To go back for a moment, the Hafler amp was shown to make very small errors from perfect amplification. The demon stration consisted of subtracting the actual output of the amplifier, while driving a speaker load, suitably attenuated, from the input signal. Perfect amplification would be established if this "difference signal" were zero. The difference signal that was observed was not quite zero, but it was approximately 60 dB down from the undifferenced signal, and up to 70 dB down in the midband, according to Hafler s reports. (Moreover, it had a character similar to the music, rather than being gritty and grainy.) In other words, the overall error, everything that the amplifi er was doing wrong, was a tiny percentage of the sig nal. At this point, people might well have stopped looking for gross errors, things like actual failure to realize gross dynamic-level changes, and started to concentrate on what was real, namely the amount and nature of low-level deviations from perfection. This should have been the case for well-designed solid-state amps, at least. Tube amps almost always make gross errors on this kind of test, and their performance has to be advocated on musical grounds, on some sort of claim of higher truth, rather than lit eral accuracy in this sense. But when you evaluate good solid-state amplifiers, what you ought to expect to be evaluating is not gross performance matters but rather the subtle questions of truly low distortion and the preservation of detail from that and from silent background. These thoughts come to mind not because the Marsh was designed using the differential compari son test - I don't know whether that is the case or not. Rather, the matter comes to mind because in fact the Marsh seems to me entirely devoid of what one might think of as superficial errors. Its bass is flat to DC and, so it must, when backed by a good power supply, be both extended and precise. And indeed it is. The power supply of the Marsh seems as good as they come in this regard. There is really nothing to fault about the bass to the extent that I can tell without speakers that scrape the bottom without subwoofers (the Harbeths are close to full- range but not quite). Similarly, the midrange is smooth and neutral. Nothing overt is amiss. And the treble is flat and extended. But, of course, one could say the same of any well-designed solid-state ampli fier that was not low-end band-limited, with a good power supply and a sensible circuit. What makes the Marsh amplifier, to my ears, outstanding is something other than any kind of observed absence of gross errors. What is unusual here is another kind of thing entirely: It is purity, purity that goes beyond what most other amplifiers have to offer. What an amplifier is likely to do wrong is not gross stuff - that just does not happen except with really bad designs. It is instead the addition of low-level but annoyingly audible artifacts or the masking of tiny detail by noise. And it is here, in the absence of these annoyances that the Marsh seems to exist at the very top level of excellence. In this regard, it is important to keep in mind that electronic arti facts can be audible even though they are far down in level. Consider the arguments for and the sonic supe riority of 20-bit or even more 24-bit recording, even when the sampling rate is not increased. The pushing down of the level of artifacts from adding the extra bits makes a real difference, even though 16-bit recording is already perfect (within its band limit) as to the gross things like dynamic levels in the usual sense and frequency response. The Marsh sounds as though its artifact level is extraordinarily low. One can hear far, far into the recording and one hears all the detail there is to hear. Moreover, the treble has a real purity and delicacy, while still being very detailed. Let me cite a musical example. On the Paula recording of the organ music of Buxtehude [Paula CD 72], the characteristic sound of high notes on the organ is superbly cap tured. The short pipes speak with a kind of a "chirp” before their tone begins both in reality and on the recording, properly reproduced. With the Marsh, this is perfect - as far as amplification goes. Getting it perfect as far as tweeters go is another story: The SEAS Excels in the Harbeth Monitor 40s, now with their tweeter screens removed (finally got my nerve up) really do it right, as do the titanium domes on the JBL LSR-32. But few tweeters do, actually. The same rightness of amplifier behavior applies to other subtle effects: the exact sounds of a xylophone being struck, of pizzicato strings (band 20 of the Bach/Sitkovetsky Goldbergs on Nonesuch is electrify ing in its quiet way, so fully resolved is it), harpsi chord as continuo (same recording) is exceptionally clear without being flung at you, and high percus sion (the Water Lily Philadelphia recording) sounds spot-on - clean, unsmeared, yet not overly assertive. Incidentally, to appreciate this to the fullest extent, you need to let the amplifier be fully warmed up, best results occurring when it is left on all the time. To turn aside from the detailed analysis for the moment, the Marsh/Harbeth combination sounded very satisfying and indeed beautiful in musical terms. In the middle of all this is-it-almost-perfect discussion, one has to leave room for will-I-like-it. I surely did, to the point where I was tempted to buy it, even though I am not really in need of any more amplifiers. One becomes accustomed, not to say addicted, to the clarity of the thing, and the access it gives to what is going on in the music. Of course, as one might expect, this purity and silent background reflect themselves, in addition, in a full sense of the spaciousness of recordings. One of the things that has developed from the numerous listening tests about enhanced digital resolution is that the reduction of artifact level has been found to be relat ed systematically to the enhanced perception of spa ciousness. This is psycho-acoustically to be expected: the more clearly the recorded room reflections are pre sented, the more the ear/brain can infer things about the space in which the recording was made. Here again the Marsh does a superb job. This is correct soundstaging behavior in the real sense, not some hokey artificiality arising from SET tube distortion. I suppose it is going to be a bit of a disappointment that this review does not go on in detail about the way the Marsh does all kinds of things right. Does it do "vocals" well? Of course. Do massed strings sound right? Yes, indeed, and what else could they do, given the artifact and distortion-free treble? It is an audiophile convention that one has to recite such a litany. But if one starts from the truth, that an amplifier like this makes no gross error, the litany of how it doesn't seems to me largely superfluous. This in no sense is to suggest a sonic identity among competent solid-state amps - the artifact thing I have been talking about is real. Not all amplifiers that work well in the superficial sense sound as clean as the Marsh. In fact, very few of them do, and none with this power rating, this neutral balance, and this price that I am aware of. But it does seem to me that the purity, quiet, and clarity, along with the neutral voicing, are the main points, not some expectation not to be realized of gross error in dynamics or the like. (By deliberate choices in the output network, it is possible to modify the tonal character of an amplifier away from neutrality. This is wisely eschewed here. Sometimes people interpret such tonal shifts in terms of dynamics, the two aspects interacting on account of the non-flatness of equi-loudness curves. None of this is an issue here, however, since the Marsh is voiced neutral.) There is a bit of a variation among even well done, otherwise neutrally voiced amplifiers as to the character of their extreme top, based on, again, the kind of output stabilization networks they have. In this department, I suppose one might put the Marsh in the slightly extroverted class relative to, say, something like amps from Plinius or, certainly, Quad. This is not to suggest even for an instant that the top of the Marsh is in any way nasty. But in understanding amplifiers and this one in particular you have to beware of blaming the messenger. In this regard, I would like to point out, once and for all, that a great many tweeters, even most of them that are not of the old-fashioned roll-it-off kind, are rather peaky, and they would sound better if those peaks were not there. But it is not the job of the amplifier to fix such peakiness. The use of amplifiers as (expensive) tone controls is folly, to my mind. And as tweeters improve (oh, that SEAS without its screen!), it is becoming rapidly apparent that much of what people have long said is wrong with solid- state amplification was really a rebellion against the sound of the average tweeter (heaven save us from most metal domes), not a legitimate objection to the amplifier. Perhaps this rebellion was musically natural, but one has to understand these things for what they are. In short, if you don't like the top end here, look to the tweeter, not the amplifier. With a good tweeter, the top end of the Marsh is exquisite. I am not trying to promote the Marsh as perfect. All good things can be pushed further, and I suppose that one might imagine even more clarity and puri ty. But the Marsh already is in the top echelon. With a clean CD feed, it seems, to the extent that I can recall, matching the amazing clarity and delicacy of the all-digital Tact Millennium very nearly. (An exact comparison will have to wait until the Millennium comes back from Sea Cliff.) And that is to say that it is as clean and transparent as there is, in my experience. Now there does come a point at which the defects of recordings are so much more extreme than the defects of amplifiers that one is pursuing an abstraction in trying for improvement. And with an amplifier like the Marsh, one is tempt ed to think one has gotten essentially to that point. An amplifier is a messenger - it should deliver only the message it is given, and when it is doing that with more accuracy by far than the message itself has - then one has gone as far as one needs to go. I used the Marsh with a great variety of speakers and never did I fail to feel that I was hearing the true measure of the speakers, changes which of course made a much larger change in the sound balance than changing from one amplifier to another, as everyone knows. Nor did I ever feel that I was not hearing the recordings as they really were. In every case, not only did I feel that I was hearing accurately the tonal character of the speaker and recording, but also that the Marsh was letting the speaker show all the clarity, resolution, and purity of sound of which it was capable and revealing everything that was present on the recording. By now, I think you may be finding it hard to suspend disbelief, as it were. It is deeply ingrained in audiophilia that a relatively inexpensive amplifier cannot be really, really good. And I suppose everyone is waiting for me to say, well, yes, but of course at this price there is this or that or the other compro mise. And of course there is indeed the limitation of power. In my listening at moderate levels in a fairly small room, I never ever felt that the Marsh was reaching anything like a point of strain, even with relatively insensitive speakers like the Harbeths. But I know from measurements that certain sounds, especially bass transients, can make surprisingly large demands for power without the sound level being extreme. In a larger room, and at high levels, I suppose one could run out of power, in short, although I never did. For huge rooms and very high listening levels, you might need even more muscle, but in my world, this was plenty of power. And what about the other restrictions that you might expect me to add, the condescending remarks about how, of course, at this price, you cannot expect the sweetness of the Humungous 5 or the Batear SET that cost as much as houses? Rather than that, let me make a modest proposal: Borrow an ultra- priced amplifier that you have never heard from a friend before you listen to the Marsh, and listen to the Marsh and the house- costing monster one after the other without knowing which is which. I am not saying that they will sound the same. But if you find yourself think ing of the one, merely an economy amp with all its compromises, and of the other, ah, now this is what an amplifier should sound like - I'll be very surprised. Unless, of course, it is the Marsh that elicits the latter description. ROBERT E. GREENE
MANUFACTURER INFORMATION Marsh Sound Design 62 El Camino Drive Corte Madera, California 94925 Phone: (415) 927-4672; fax: (415) 924-6846 Email: Rmarsh@marshsounddesign.com Source: Manufacturer loan
SPECIFICATIONS Maximum power output: 200w, both channels continuously into 8 ohms; 330w into 4 ohms Frequency response (20 Hz to 20 kHz): +0/- 0.2 dB, 1w to 200w into 8 ohms, -1 dB at 80 kHz Total harmonic distortion plus noise: <.02%, 20Hz to 20 kHz at max. Output; <0.01 % at 1 kHz, max. Output, typically <0.002 % at lower levels S/N ratio, A weighted (ref. 200 w): 110 dB Output impedance: 0.01 ohm, damping factor 800
PAUL SEYDOR COMMENTS To the several speakers REG used to audi tion Richard Marsh's new amplifier, I can add the new Quad 989, which made for a superb match. Right now, thanks to the Placette preamplifier, I am enjoying reproduction of unusual transparency, also a prominent characteristic of the A400S. Together, it and the Placette are almost pornographically revealing of signal sources. The highs are especially refined, delicate, smooth, and supremely clean. Lately I've been enjoying a new Chesky CD of a talent ed young singer named Christy Baron [Steppin ’ , Chesky JD 201]. On the second track, " Mercy Street," there is quite a bit of high-lying percussion in the form of cymbals, bells, maracas, a rain stick, and who knows what other instruments. The sparkling clarity, the razor-sharp definition, the air between and around the notes, the speed of attack and decay, the differentiation among the several instruments - to hear the Quads and the Marsh unravel all these sounds, keep them distinct yet allow them to come together as a whole musical experience, has been one of the real joys of recent review ing. Another one is "Angel Eyes" from Frank Sinatra Sings for Only the Lonely (Capitol 72434-94756-2-5), which I don't think I've ever heard image so beautifully as through the A400/Quad 989/Placette combination (the CD player Sony's superb SCD-777ES): Sinatra firmly centered, set slightly back, surrounded by Riddle's orchestra in a wide semi-circle, the impression of air and envi roning space utterly convincing in their over all cohesiveness. There is a quiet, unassuming authority on the performance of this amplifier that is at first a little daunting. Putting it in place of the Sunfire Signature, a beautiful amplifi er but hardly neutral, I thought the Marsh might be almost too self-effacing, almost reticent. Is there anything here to describe? This is, of course, as it should be, though it puts a writer into a bit of a panic. On the large-scale end of things, I played the last three movements of Andrew Litton's Mahler Second in SACD [Delos SACD 3237], a remarkable record ing, lowish in level but with really wide dynamic range. The Marsh gave the impression of complete control, effortless reproduction, yet with no sense of hyping anything. I fear the only problem the A400S might have in the marketplace is that it affords the audiophile so little to kibitz with his buddies about. It's as neutral a piece of gear as I have heard; it is up to vir tually any real-world amplifying task you're likely to ask of it; and it won't glamorize anything. You practically have to talk about the music with this amplifier in the system. Regarding REG's remarks about value, I am reminded of the days when I used to sell High End audio. There was a fabulous line of tube gear (I shall not name it) that performed, in my opinion, as well as any tube gear out there (and was far more reliable than all the rest), but was nevertheless reasonably priced. I sold quite a number of the amplifiers, but not so many as I believe I could have if only they had cost a whole lot more. Is a hopeless longing for equip ment priced hopelessly beyond the afford able what audiophiles have in place of God? I don't know, but I do know that this is the second product designed by Richard Marsh that, to steal an image from Mark Twain, simply flings down and dances upon any notion that superior performance man dates superior pricing (the other product being the Monster Cable HTS 2000 power strip). I fear there are hundreds of audio philes who will never realize how good the A400S is because it costs "only" two thou sand dollars and doesn't sound like much of anything except what is put into it.
Richard Marsh acted, for many years, as our Technical Consultant and wrote a number of articles explaining important technical points. In that time, he developed warm friendships with many of our staff, including the reviewer. The prurient may want to peek in Issue 117, Editorial Notes, where we announced Richard's founding of Marsh Sound Design and his move to a position across the board, as it were, from the audio press.
|
by throwing complexity of topology and arbitrarily expensive parts willy-nilly into the mix, although the A400S has high-quality parts, indeed, where they count. Instead, the prob lems of amplification have to be solved by, to put it bluntly, brains. A second aspect of the situation is that manufacturing costs have to be kept unusually low. Not even Bryston, with the economy of scale acquired from large production for the pro market, has assayed an amplifier of this power at this price. The manufacturing-cost question has been dealt with in the Marsh Design amp by having the amplifier manufactured in Thailand, where work of high quality is available at low prices. The result of the combination of superb design expertise and low manufacturing cost is an amplifier that would be very good at any price and is positively startling at its actual price. This product instantly achieves a position as one of those rarities in audio, a legitimate claimant to the best available sound, which at the same time has a reasonable price. |
|
