VMPS SuperTower IIa/R loudspeaker J. Gordon Holt May 1988

J. Gordon Holt reviewed the IIa/R Special Edition in May 1988 (Vol.11 No.5):

It may have been audio pioneer Emory Cook (footnote 1) who once observed that "nobody has ever been able to miniaturize a bass violin." His point, of course, was that bass sounds involve a lot of movement of a lot of air, and it takes a large object to do this. While small speakers can now produce deeper bass than they could 40 years ago, we still have not managed to make a dent in that law of physics that says "Deep bass demands a large radiating area." VMPS has accepted this view, not only philosophically, but practically, with enthusiasm.

The VMPS Supertower IIa/R, the largest speaker system I have ever had in my home, is also the most complex by far. Each speaker has 13 active drivers, arranged in two more-or-less vertical rows on the front panel. The low- and midrange driver cones are all made of polypropylene—a material noted more for its high inherent damping than for its rigidity—while the woven-fiberglass tweeter radiators are inverted domes from the French Focal company, similar to those used in the Wilson Audio WATT. (The concave shape is claimed to give wider, smoother dispersion.)

The SuperTower's woofer loading is a sophisticated bass-reflex system using a large passive radiator. Conventionally, in a reflex system, a port is tuned (via its length and diameter) to add in-phase output below what would be the fundamental resonance of the drive-unit in a sealed-box enclosure, resulting in an increase in the LF extension of the system. The air in the port, however, which acts as a low-mass, large-excursion drive-unit, can often produce wind noise, a problem solved by replacing the open port with a passive radiator. Essentially a damped bass-reflex port, a passive radiator uses a mass-loaded cone diaphragm instead of a resistance-loaded vent to accomplish the same thing, but because no air is being driven through small interstices, it has the advantage of producing no wind noise.

The SuperTower's passive radiator is mounted face-up at the top of the system, radiating forward through a large slot. It is a high-compliance 15" cone of relatively light weight, which normally provides fairly little damping, to give a rather high system Q of around 2. This is adjustable, down to a Q of 0.5, through the addition of mass elements—small circular wads of what look like soft rope putty (footnote 2), supplied with the system. The wads are added, one at a time, to the center of the cone until the system's damping is optimized for the amplifier, the speaker cables, and the speaker placement. A clever idea, but it's not original. Paul Weathers (footnote 3) used the same material for woofer damping as far back as the early 1960s.

The five crossovers in the SuperTower are what designer Brian Cheney terms "Quasi-Second-Order Holosonic." QSO is a recognized design approach, wherein each side of the crossover has two different crossover points, so that the transition is first gradual, then steepened to approach 12dB/octave, yielding a high degree of phase coherence at crossover while still providing fairly steep attenuation well beyond the crossover point. But the "Holosonic" aspect of a QSO crossover is unknown to the author of any reference book I know. I presume it's VMPS's way of touting QSO's allegedly superior rendition of depth. [See also TJN's review of the smaller VMPS Tower II/R elsewhere in this issue—Ed.]

The upper three of the driver-operating ranges have individually adjustable level pots (located at the rear, above the cable connections), each with an adjustment range from completely off to a few dB higher than that of the woofers. There are no detents, but the midway setting of each pot is recommended by VMPS as the starting point for system setup. (The midway settings were also, incidentally, what I ended up using.)

The original Supertower IIa/R was more than bi-ampable; it was only bi-ampable. You had to purchase an optional kit in order to use it full-range. The so-called Special Edition, which I tested, has the single-amp "module"—two additional inputs and a toggle switch—already installed, and the SE also includes IAR Wondercap crossover capacitors (instead of conventional polypropylenes), Monster Powerline I internal wiring (instead of regular Monster cable), and the latest Focal 1" "Superdome" tweeters, each claimed to be able to handle 250W of program.

Even before my first listen to the IIa/R, I had some misgivings about its impressive array of drivers; there are as many cons for this approach as there are pros. Among the pros: 1) the narrower the range over which a driver operates, the easier it is to optimize its performance through that range; 2) the more crossovers, the less power any one driver or sets of drivers must handle; and 3) multiple drivers increase conversion efficiency, typically by 3dB for each doubling of radiating area, while decreasing by the same amount the power each driver must handle. But the cons are just as persuasive, among them: 1) the simplest way of doing something is usually the best; 2) many crossovers increase the possibility of many phasing errors and discontinuities, where one driver sounds slightly different from another; and 3) interference effects due to the differing height and lateral positions of the drivers can cause timbral and phasing irregularities. The literature for the IIa/R assured me that it embodies all the pros and none of the cons, but I was still skeptical, even though I had heard some impressive sounds from these behemoths at the last few Consumer Electronics and Hi-Fi Shows.

The SuperTowers were delivered by truck, a day before designer Brian Cheney arrived in Santa Fe to set them up and tweak them in my listening room. We were all appreciative of this, if for no other reason than that he is accustomed to manhandling these monsters, and knows the easiest way to unbox them. (You open both ends of the box, lay it on its side, and push the speaker out from the other end.) The two speakers are, of course, configured as mirror images of each other, and, as usual, must be placed with the upper-range drivers toward the center, with both systems toed-in toward the middle of the listening area.

Despite their size and multiplicity of drivers, Brian started with them located in my usual loudspeaker locations, about 7' from the sofa and 6' from the rear wall. This produced rather poor driver blending and a paucity of low end, so they were moved back, bit by bit, until they ended up about a foot out from the rear wall. Brian balanced the drivers for some recordings with which he was familiar (including some he had recorded himself and issued on his Itone label while living in Munich), and after several hours declared everything to be in pretty good shape. Then he went back to California's sunshine while I stared sullenly out the window at yet another early-Spring snowfall.

Sound Quality
Listening at my leisure, I started to observe a few minor problems involving imaging and tonality (midrange colorations) that I was confident I could solve with a little more experimentation with placement, orientation, and driver balance. But achieving perfection with these speakers turned out to be an exercise in frustration. Changing their angles by a few degrees or their spacing by an inch or two had dramatic effects on imaging, soundstaging, and tonality, and whenever I got rid of one problem, another surfaced. It was like trying to hold 10 ping-pong balls under water at once.

First there was a rather pronounced picket-fence effect (footnote 5). Changing the toe-in by about 3° for each speaker got rid of some of that, but then the whole soundstage became unstable, shifting to the right or left as I changed my listening position. Some more tweaking seemed to correct both of those problems, but then the soundstaging presentation had gone to pot, with instruments bunched near the speakers rather than "floating" as a coherent group between them. And so it went. After a week, I gave up trying to optimize everything and went for those few things I consider most important: tonal accuracy, imaging stability, and a reasonable presentation of depth. The only positive thing I can say about that week was that the speakers proved a lot easier to move, when on a shallow-pile rug, than I ever imagined they could be. I could do it easily myself, even though each one outweighs me twice over.

But because the VMPSes are so chameleon-like, it is difficult to describe what they do and do not do, except in rather broad terms. I can say that they have immensely impressive dynamic-range capability, and are quite capable of approaching the original, in-the-room volume level of a jazz band or a full drum set, even with the modestly powered (160Wpc) SA-1 power amps. I can also attest that they have a truly incredible low end, with extension fully equal to what I got from the pair of Nelson-Reed 1204 subwoofers I had been using previously, but with rather better definition.

In fact, low-frequency performance is unquestionably the SuperTower IIa/R's strongest suit. Bass was subjectively flat to an ear-popping 25Hz in my room, and was very tight and detailed, with superb pitch delineation. Bowed double basses had a most authoritarian "thrum" to them, and the leathery voice of organ diapason pipes was rendered as well as I've ever heard. At the same time, the woofers were agile enough to produce a tremendous feeling of impact from the abrupt bass attacks of plucked double-bass strings and kickdrum.

Overall, the sound was slightly veiled, lacking the delicacy and transparency of a full-range electrostatic, but taking a back seat in that department to few other dynamic systems I know of. Focus and inner detailing were both outstanding. The extreme high end, too, was superb, being at once smooth, open, and superbly delicate. String reproduction was remarkably good—sweet, gutty, and remarkably airy. But outside of that? Well, there's enough range on the driver balance pots to give you almost anything you want in the way of spectral balance, but not quite everything. The biggest problem was a deep midrange suckout, centered at about 2kHz, which seemed to be almost totally unaffected by adjustment of the appropriate balance control. Advancing the control increased the amplitude of the 5kHz range but had little effect on the suckout region, even though the lower crossover into that driver is supposed to occur at 600Hz (see the frequency-response curve on this page). The suckout was not broad enough to produce a laid-back quality, but instead affected only the accuracy of many musical sounds. Trombones, for instance, sounded anemic and lacking in authority, while an original tape of a familiar tenor voice that I recorded some years ago (for commercial release), and which still serves as a real-world reference for accuracy, sounded gutless and muted, as though heard through a burlap bag.

And there were other minor problems. The SuperTowers seemed to exaggerate differences between different recordings, making some sound more realistic than they usually do, and others less so. It could perhaps be argued that this is because the system is more revealing than most, but it can also be argued—and I will so argue, based on my many years' experience with this phenomenon—that it is because of the system's unusually irregular frequency response, revealed when I finally got around to running my response tests on it.

Fig.1 shows an average of 10 different curves, all taken with the probe mike between the system and at ear height across the middle two feet of the listening area, but it does not show the extent of the variations which resulted in moving the mike a mere 6" in any direction. Since the irregularities changed somewhat in frequency as I moved the mike, the final averaging reduced their severity to some extent, but the deviations shown on this page were common to all mike locations, which would suggest problems a little more serious than intra-driver interference effects.

Fig.1 VMPS SuperTower IIa/R, spatially averaged in-room frequency response, average of 10 curves with the mike at ear height across the middle 2' of the listening area.

The SuperTowers also proved to be susceptible to the effects of varying listening height. The whole output from the midrange array seems to be lost when you rise above a sitting position, and a lot of the output from the supertweeter is lost when you sit down. But the sound was least colored when heard from a typical listening height (about level with the lowest midrange unit), the tweeters could be balanced-out for that listening height, and the audible effects of movements to either side of dead center were much less noticeable than the frequency-response measurements would have suggested.

Summing up
All the synonyms for the word "impressive" apply in spades to the VMPS SuperTower IIa/R. If that, the ability to play loud with relatively low distortion, or world-class bass performance, are what you're looking for in a loudspeaker, this is the first system you should check out—particularly if you also value imaging and soundstaging.

But if you're picky about freedom from coloration and accuracy of musical timbres, you can do better than this for less money. Among the systems we currently recommend in the SuperTower's price class are the Magnepan Tympani IVA ($3800/pair), the Apogee Scintilla ($3995/pair), and the Quad ESL-63 at $3600/pair. All are less colored than the VMPS SuperTower, but none can really be considered competition for the VMPS, both because all are panel dipoles, which have their own set of problems, and because none sounds even remotely like it.

Whether that is construed as praise or criticism of the SuperTower will depend on your reaction to its sound. After two weeks of listening, I still have mixed reactions to this system, but am certain of one thing: Because I value highly that quality of midrange accuracy which the IIa/R lacks, I could not live happily with it. This, of course, does not guarantee that you will agree with me.

Since the VMPS SuperTower IIa/R has so many positive attributes, but also such a unique personality, it is one speaker you should never even consider buying without a prior audition. (You shouldn't buy any speaker without auditioning it first, but that caveat applies to this one much more than to most.)

All in all, a mixed bag.—J. Gordon Holt

Footnote 1: On the other hand, it may not have been.

Footnote 2: This is the string-shaped Plasticine-like stuff sold in hardware stores for temporary window caulking.

Footnote 3: Although remembered (by old-timers) only as the inventor of the first FM phono pickup, Paul Weathers manufactured a complete line of audio components during the early 1960s.

Footnote 4: Also called vertical-venetian-blind effect, this is a phenomenon in which stereo images seem to hop back and forth between the speakers as you move your head small distances either side of dead-center.