Magnepan Tympani I loudspeaker
The Magneplanar Tympani I is unique both in appearance and principle. The system looks like a pair of folding room-divider screens with three hinged panels per screen. Each screen consists of two "woofer" panels operating up to 3.2kHz and one tweeter panel for the range above that. The radiating surfaces are thin sheets of Mylar plastic, in which respect the system resembles an electrostatic design. But instead of using a varying electrostatic charge to produce diaphragm motion, these use a fine grillework of wires fastened to the surface of the radiators. An array of bar-shaped magnets provides a strong magnetic field around the wires, and if you remember your physics you'll recall that an electric current passing through a conductor in a magnetic field produces physical force. In this case, of course, the current in the wire grilles comes from the power amplifier.
Our first question when we heard about this system (two years ago) was "Is it push-pull?" The answer was "No, it's single-ended." To which we replied "Well then, it has to have high distortion because the force produced by a current-carrying wire in a magnetic field will diminish according to the square of the wire's distance from the magnet." "Not so," we were told, "because in this case, the shape of the magnetic field from each magnet is such that the field strength remains constant through the entire distance travelled by the diaphragm." We were skeptical, but we reserved judgment until we heard a Magneplanar. When we did, we were obliged to revise yet another of our cherished beliefs. A device does not have to be push-pull to be linear. For it was obvious that whatever else the first Magneplanars may have been doing that we didn't much like, they were not distorting.
There were two things we didn't like: the speaker's high end and its low end. Granted, the sound was very smooth throughout the speaker's frequency range, and had a degree of homogeneity we have heard from very few systems. But these were markedly deficient in bass in many listening rooms, and their overall sound was suffocatingly "dead," as though all the sonic spikes which give sound its focus and snap were just being neatly rounded-off.
We kept having the feeling that there might be some very realistic sound somewhere behind those panels if we could just get at it, and the result was an insatiable urge to turn the volume higher and higher in an effort to break through that thick velvet fog, but higher volume just made the sond bigger. The fog remained. And although we could sympathize with some souls who claimed to prefer the Magneplanar's blandness to the sizsle and spit of some electrostatic systems, we could not help but view them as shockingly overpriced midrange speakers.
Obviously, we were not alone in our criticisms of that first model, for it was not long before an "improved" version appeared. This used lighter wire on the tweeter panels. The reduced mass improved the transient response noticeably, and although we still felt the system was inordinately soft-sounding, it showed the way to go. And the Magneplanar has been going that way ever since.
The last version we heard, which represents all Tympani I systems made since around the end of 1972, had the high-end snap of the best dynamic tweeters we've heard (but still not quite that of a good electrostatic), a silky-smooth top, almost total absence of any coloration of any kind, fantastically good detail and homogeneity (blending of drivers into a unified sound front), completely neutral perspective (neither close-up nor distant), and sadly deficient low end in many listening rooms. It is also one of the least sensitive speaker systems we've ever encountered.
It does, however, seem capable of taking astonishing amounts of input power without audible stress, so if the power amplifier can deliver enough low-end belt, the speaker can be equalized to provide subjectively flat response down to a bit below 40Hz in many rooms without sacrificing any of its normally outstanding bass detail. In fact, considering its low sensitivity, it is surprising just how much clean signal can be elicited from it by the mere (?!) 75 watts per channel of Audio Research's own Dual 75 power amp. (This could be simply because the Dual 75 overloads more gracefully than any solid-state amplifier we have tried. The Dual 75 gets increasingly muddy, solid-state amps make sharp clicking or crackling noises when they clip.) A single Dual 75 is barely adequate though if you like your listening levels to be fairly high. Audio Research recommends biamplifying the Magneplanars with a Dual 75 and a Dual 51 (plus one of ARC's electronic crossovers), and the speakers are externally strapped to permit easy connection of biamps.
Indeed, biamplification seems to make the sound audibly smoother and more detailed even at low listening levels, besides ncreasing maximum levels, but the total cost then becomes astronomical. If that doesn't bother you too much, it's probably the ideal way to go. You can however get even more clean output-signal level at considerably lower cost by using a single Crown DC-300A to drive the Magneplanars, although this will cost you a very small amount of the inner definition and high-end sweetness for which the ARC power amps are noted and in which they are unequalled.
Unlike full-range electrostatics, which seem relatively unaffected at the low end by driving amplifier characteristics, the Magneplanars can have their low end over-damped by some amplifiers (such as the DC-300A). The result is improved bass detail but a somewhat dry quality that can be partly (but not completely) taken out by a smidgeon of additional low-end equalization. As intimated previously, though, there are some listening rooms in which no amount of bass equalization or care in speaker placement can make the Magneplanars sound anything but thin. Fortunately, such rooms are rare, and in most cases, any variable-inflection–type bass control will do a more-than-adequate job of filling out the low end.
As with other doublet systems, room placement is crucial for optimal functioning of the Magneplanars. There is no way of predicting the best placement in any room, and it often takes weeks of experimentation to find the best speaker positions and panel angles, but both of these things affect bass performance as well as stereo imaging.
There is a tendency toward phase-interference effects between the panels—observable as a vertical-venetian-blind, left-right hopping of sounds as you move across the area in front of the speakers. The effect is most pronounced in acoustically dead rooms and also from closer-than-normal listening distances, and tends under the best of circumstances to limit somewhat the area in the room from which proper stereo can be heard.
Stereo imaging is affected by the orientation of the panels containing the tweeter elements, but at best, directional specificity from the Magneplanars is noticeably 1ess than perfect. There is little actual wander of point sources, but the directional locations of instruments across the stereo stage is rather on the vague side. We have always maintained that directionality in live music is not as specific as some audiophiles demand that it be in stereo, and that visual cues are what provide most of our directional information in the concert hall. But there is something to be said for the counterclaim that, when visual cues are absent, the sounds themselves should be more definitely directional. And they are with some other speaker systems—notably the 360° radiators (footnote 1).
So how does it sound generally? We've given a lot of thought to that question and, at the risk of being misunderstood, are more inclined to describe its sound in terms of minuses than pluses. Fig.1 shows its subjective frequency response. The system has virtually no spurious colorations, it does not exaggerate tape hiss or LP surface noise or distortion (and indeed may a subdue these things very slightly), and it does not favor any instruments over others. It has an unprecedented degree of inner detail, in that you can follow a single melodic or instrumental line through complex orchestral passages more easily than via any other system we've heard—it does this best with ARC's own amplifiers—and it does not tend to develop an irritatingly hard, shrill edge when you turn up the listening volume. Instead, the sound just gets bigger and bigger and BIGGER.
Fig.1 Magnepan Tympani I, subjective frequency response (2dB/vertical div.).
In other words, there is less about the sound of the Magneplanar that says "loudspeaker" than we've heard from any speaker to date. And yet we found something lacking in its sound that should have been present: It rarely gave us goosebumps.
We know—not all live music is conducive to a crawling scalp and erect forearm hairs. But some live music does elicit this same animal excitement, and it rarely did when reproduced through the Magneplanars. We have, in fact, experienced considerably more of this goose-bumping from systems that we know objectively to be more highly colored, not as smooth, and less realistic in other respects.
Certainly, the Magneplanar's low end bad something to do with its vaguely dispassionate quality, but we noticed also that cymbals were a little shy of their usual metallic clangor, brasses were short of their usual sizzling or flatulent spikes, and reed instruments were deficient in their typical rattling squawk.
We once described the KLH Nine as having a "polite" sound of a certain kind. The Magneplanar, in our opinion, has a polite sound of another kind. The Magneplanar is, we feel, a more realistic reproducer in more ways, but for all of its shortcomings, the Nine is a more effective hair-raiser. (The Infinity SS-1, even less perfect than either of these, was even more effective in the hair-raising department.)
On the other hand, there is less annoyance to the sound of the Magneplanars than in either of those other systems. It is easier to live with, and just as it does not wring every iota of drama from a recording, neither does it tend to draw attention from the musical content to the sound. We might even say that it is ideal for the listener to whom music is more important than sonics, but to say that would be to belittle unjustifiably the sonic qualities of the system as well as mislead the musically oriented buyer. For it is not possible to divorce the sound of music from the music itself. There are, of course, many compositions where sound is secondary to structUre or phrasing or tempo, but the texture of Debussy's Nocturnes or the bombast of the Roman Carnival Overture or the savagery of Le Sacre du Printemps will, either move us or fail to move us almost entirely on the basis of their sound.
The Magneplanars invite direct comparison with three other over-$1000 loudspeaker systems: The KLH Nine, the Infinity SS-1, and the IMF Monitor. We mentioned the first two in passing previously. Both go lower than the Magneplanars, the KLH with equal bass detail, the SS-1 with less. The SS-1 is somewhat forward, the KLH somewhat distant. The SS-1 is somewhat hard and, snappy at the top, the KLH is sweet (assuming the best driving amp), focused and natural at the top, but with all the snap that's needed. And the SS-1 requires two stereo driving amps, while the KLH requires but one.
But what about the IMF Monitor? We received a pair of these for testing just before going to press, and were able to listen but briefly to them. They raised goosebumps, but then that's another story (or, rather, another report). So are the new Magneplanars—the Tympani IA, the Tympani II, and the Tympani III. The IA is claimed to have better transient response ("snap") than the I, the II has the same as the IA but costs less, and the III (which must be biamplified and should be, of all things, triamplified!) is claimed to have better transient response plus better low end.
If these new models did not exist, we would recommend the Tympani I as one of the best loudspeaker systems available but, like the other "beet" systems, not best in all respects. And certainly not for the typical audiophile who loves his cymbals and organ pedals passionately enough to give up a certain amount of smoothness and naturalness of timbre in exchange for them. As of now, though, our advice to anyone considering a Magneplanar is to sit tight and wait to see how good the new models are. If they're as smooth as the Tympani I and are hair-raisers too, they'll be the best things available and we will think about giving them (or one of them) exclusive Class A recomnendation, dropping everything else in our "Recommended Components" list downwards by a notch.
A pair of Tympani IA speakers arrived just before we went to press. Inital reaction: Much better highs, with almost the snap of good electrostatics but virtually none of the sizzle usually associated with electrostatics. Deeper, fuller bass, subjectively flat (at best) down to around 45Hz. Slight midrange hump in previous model is no longer audible, but IA has a somewhat leaner, harder sound that makes the hardness of most solid-state amplifiers intolerable. Even with the Dual 75 (on most program material), it is a bit irritating if you're sensitive to that sort of thing. (Oddly, it is noticeably reduced by using the Dual 75's 4-ohm output taps.)
As we go to press, though, the main unanswered question is: Does the Tympani IA merely reproduce with embarrassingly clarity the hardness that is on commercial recording—they are, after all, made through solid-state microphone mixers and cutting amplifiers—or is the Tympani IA itself a little bit on the brittle side? Some non-commercial tapes we have on hand were not hard and, indeed, sounded extraordinarily natural, but we hope to have a more definitive answer in the next issue.
The Tympani IA really must be biamped, but it then can be a hair-raiser.—J. Gordon Holt
Footnote 1: And we mean the true 360° radiators, not those absurd systems with their tweeters rotating like lighthouse beacons.