Channeling Beyond Stereo with Vandersteen and Audio Research
by Dan Sweeney
Fi Magazine, March 1996
Today's music lover is confronted with two closely related technical developments which bid fair to transform the listening experience in the home. The first of these, which has not seen commercial realization as of yet, is the true, extended definition compact disc with bit resolution of 20 to 24 bits and it is to be hoped, a sampling rate beyond 90kHz. If this type of super CD actually comes to fruition, it will almost certainly give tremendous stimulus to the high end of the industry and might well provide a digital audio system that nearly all music lovers can wholeheartedly embrace.
The second development, which is upon us now and is the focus of this article, is the resurgence of multi-channel and surround sound music reproduction after its near extinction in the early eighties. To be sure, the major impetus behind multi-channel has been the home theater phenomenon, but of late, a number of music-oriented manufacturers have prophesied a veritable Second Coming for music surround. Vandersteen Audio and Audio Research, the principal subjects of this review, are among the prophets.
Potentially this Second Coming is of the utmost importance to all of us. Certainly, if multi-channel ends up dominating the music industry, it will transform the listening experience as thoroughly as the advent of stereo did forty years ago. But, as I hope to demonstrate, the way to high fidelity multichannel reproduction is not at all clear or straightforward at this point, and pioneering efforts such as the system under review show just how difficult the passage to multichannel is likely to be.
One Approach to Multi-Channel
The genesis of this review lies in Richard Vandersteen's own comments to the effect that he has come to prefer music listening with his surround sound speaker system used in conjunction with Audio Research's SDP1 Spatial Definition Processor. Given the prestige of both Vandersteen's own company and of Audio Research in the high-end of the industry, the performance of the combination seemed well worth examining.
The Vandersteen Audio Research system is not multi-channel in the strictest sense. It is designed to play back ordinary stereo music software, and the additional channels feeding the front center and rear speakers are simply derived by combining both left and right in-phase and out-of-phase relationships. For those of you who know your audio history this is essentially the Hafler matrix.
Insofar as sound is distributed to a phalanx of loudspeakers, such a derived multi-channel system provides a pronounced immersion effect that cannot be duplicated with a two speaker array but the results are not comparable to those obtainable with a fully discrete multi-channel system. To understand why this is so, you have to know something of how a matrix system operates vis-a-vis a discrete system.
Discrete systems store information in separate channels and are in theory capable of infinitely separated outputs. That means you can assign a sound to a single speaker with no leakage at all from or to adjacent speakers. A matrix system, on the other hand, is always subject to cross-talk between adjacent channels, the precise amount varying with the decoding equation and the proportion of derived channels to recording channels. (In a Hafler matrix cross-talk is only 3dB down in adjacent channels).
What this means in terms of listening is that, with derived multi-channel, sounds tend to clump to positions midway between pairs of speakers. The lack of separation simply prevents them from being panned or steered all the way across the space separating a pair of speakers. Beyond that, the summing of left and right that occurs in the center channel gives the sound a distinct monophonic emphasis. Put simply, a basic matrix system, also known as a passive matrix, is not going to give you greater precision of imaging than simple two channel stereo does, in fact, it is likely to be considerably worse.
The whole issue of matrix versus discrete is complicated by the existence of a refinement of the passive matrix known as the active or logic steering matrix, a technology represented by the well known Dolby Pro Logic circuit. What this technology does is to impose a further level of processing on the outputs of the passive matrix, one involving cross-talk cancellation by means of injection of left and right signals in either normal or inverted phase.
Mapping out all of the permutations that occur in logic steering when left and right are mixed and remixed in various phase relationships would serve no purpose here. Suffice it to say that the process results in increased adjacent channel separation on a dynamic basis, and a greatly increased precision of imagery over that obtainable in passive matrix. Unfortunately greater image specificity comes at a price. Active matrix decoders inherently produce high values of distortion in the process of canceling cross-talk.
What's all this mean in the context of the current marketplace and in terms of the SD1's performance vis-a-vis the competition? Currently passive matrix decoders, active matrix decoders, and completely discrete systems coexist with one another, though the passive category is almost extinct while the discrete is struggling to be born. Most industry watchers predict the eventual dominance of discrete formats. But, for now at least, a choice exists; so let us consider what that choice involves.
First, the matter of hardware and software availability: Passive matrix decoders, commonplace ten years ago, are in deep eclipse. Today the only other passive matrix decoder of any consequence, aside from the Audio Research product under review, is the Chase Technologies HTS-l, an almost exact replica of the original Dynaco Hafler matrix decoder introduced in 1970 (although I should point out that both the Meridian and Citation processors include passive modes in their menus). Thus, Audio Research's decision to support the older technology goes decidedly against the grain of current market trends. Incidentally since the demise of Dynaquad in the early 70s, no specific software has been recorded for the type of passive matrix decoding used in the SDP1, although it is generally thought that minimalist stereo recordings sound best.
Thanks to its established position in commercial movie theaters, active matrix logic steering circuitry is dominant in home theater applications today and is available in dozens of different products. Moreover, it is frequently used for processing music software as well as video material. Nevertheless, little musical programming is specifically encoded for playback through the Dolby Pro Logic circuit; and critical listeners, by and large, have not warmed to the format.
Discrete multi-channel exists today in two incompatible formats, Dolby AC-3 and DTS. Both would normally be decoded into five full range channels and one low bass channel - a configuration commonly known as 5.1. Pioneer currently manufactures licensed AC-3 capable laserdisc players and decoders, and a few high-end manufacturers, including Theta Digital and Enlightened Audio Design, make AC-3 decoders. As far as software goes, many new laserdisc releases are provided with AC-3 tracks, but no music software has been released. Interestingly the situation with DTS is exactly the reverse. As of this writing, no DTS playback equipment has been introduced, but several software music titles are available on Brad Miller's Mobile Fidelity International label (not to be confused with the better known Mobile Fidelity Sound Lab of the half speed mastered phonograph records and Ultradisc CDs).
Now let's see how each format compares to the others in regard to basic capabilities. Passive matrix I have already discussed in brief. The simple circuitry involved permits a relatively clean signal path, but separation is poor - a major negative in my opinion. In fairness, I must point out that there are those, including some respected academics, who maintain that separation is of little importance in establishing a convincing representation of the performance space. I am willing to concede that for highly reverberant, back-of-the-hall minimalist recordings with little direct sound pickup, high separation makes little difference in the playback chain, since the listener in the actual performance space hears strongly correlated information at either ear at any head position. But the fact remains that tow separation systems do not permit hard pans to speaker locations - a critical limitation when it comes to reproducing close miked recordings intended to create pinpoint imaging.
Put simply, the passive approach lacks flexibility; it only works well for a single perspective.The second category active matrix decoders, vary considerably in sophistication, low level steering capabilities, and musicality with the proprietary circuits of the Citation and Meridian processors establishing the benchmark performance level in the category At their best these processors provide a good simulation of discrete multi-channel, at least with some program materials, but the basic processing strategy introduces considerable distortion and high separation cannot be maintained between all adjacent pairs of speakers simultaneously. Limited music soft-ware has been specifically encoded for playback through active matrix decoders, and, in my experience, they give highly variable results when ordinary stereo recordings are utilized. Event with the best of them, I find an unacceptable degree of veiling and glare, and a lack of countervailing enhancements in image stability to justify their use for music, although opinions differ here. Best results will be obtained with music recordings made with the Dolby matrix, including, most prominently, selected titles from the Delos catalog.
Discrete multi-channel represents the ideal, but in its present incarnations it has been achieved through massive data compression strategies which represent a retreat from the CD standard, already deemed critically low by many music lovers. With negligible music programming available, the audible extent of the degradation cannot be determined.
In this context, the Vandersteen-Audio Research system represents an extremely conservative, even backward-looking, response to the current multichannel ferment, one based on minimal processing done largely in the analog domain, and one which might best be termed an extension of two channel rather than a true example of multi-channel. If a more palpable representation of acoustical space is the desideratum, then predictably this approach falls short of expectations when comparisons were made to active matrix, discrete, or even straight two channel. But if signal purity were the primary consideration, and nothing more than added expansiveness were being sought, then the combination could prove attractive. Its all a matter of what you really want in terms of soundstaging.
The System at a Glance
The system under review consists of Vandersteen 2Ce stereo speakers, a VCC-I center speaker, VSM-I surround speakers, all receiving the outputs from an Audio Research Model SDP1 Spatial Definition Processor. In addition, Audio Research was kind enough to supply me with a four channel SDA1 power amplifier. However, because an extra channel was needed, I could not drive the entire surround system purely with Audio Research electronics. Since I had five monophonic Wolcott Audio tube amplifiers on hand, I elected to use them for most of my listening.
The SDP1 Spatial Definition Surround Sound Processor
Audio Research assiduously courts the user of this unit even before the box is opened. Instead of being wedged on the typical malformed Styrofoam chunks that always tax one's spatial perception abilities during repacking, the SDP1 is protected by a complete inner foam box that is virtually idiot proof. It's an object lesson for the rest of the industry
Out of the box, the unit continues to beguile. I used to think Audio Research styling was on the stodgy side, but I am prepared to recant. After long term exposure to the tarted up, video slot machine school of industrial design represented by most surround sound processors on the market, I've begun to find AR conservatism most visually refreshing - leaving aside the tactile pleasures of manipulating the SDP1's precision pots.
I also appreciate the unit's sublime simplicity of operation; no modes, no menus, no remotes, no video switching, no cyberspace explorations required to get it up and running. Everything is clearly marked - power, mute, pots for center and surround, and rear channel delay adjustment, constitute the only controls. Nevertheless, I must quarrel with the unit's ergonomics on two counts. First, there is no volume adjustment for the front left and right channels. Thus the processor must be used with a preamp. Second, the arrangement of input and output jacks on the back panel is lamentably eccentric. Where everyone else in the world arranges left and right jacks in a vertical row, left on top, right on bottom, ARC puts 'em side by side, input up, output down. I accidentally connected the output of SDP I to the output of another line level component and destroyed the component as a result.
I should add that most of what you're paying for is the pair of rear channel digital delay lines, the aural hyperspace drive that constitutes the unit's sole "parameter" in surround processor jargon. Damned near every processor on the market has some kind of rear channel delay for the purpose of effecting d simulation of a delayed reflections in a large listening space, but Audio Research's is studio quality For the first time in my experience rear channel outputs do begin to resemble actual reflections rather than the subliminal murmurings emitted by most processors.
Overall, build quality of the unit is excellent. Premium parts are used throughout. Optional Cannon XLR connectors accompany each RCA terminal, and the cabinetry is impeccably finished. In addition, the unit is dead quiet in operation, indicating careful attention to grounding. Here in short is the only processor I have yet encountered, excepting the $29,000 a.d.a. Olympus 111, that truly presents a highend profile. Always a sucker for a pretty face, I was disposed to love it on those grounds alone. But at length I had to listen to it, for to make music is after all its reason for being and that of course, involved the Vandersteen quintet of loudspeakers.
FOCUS ON SPEAKERS
The 2Ce
Alone
The Vandersteen surround system really is a system, not just a stereo pair with the hasty addition of surround and center. All the same, the 2Ce stereo speakers used within this system have a lengthy history of their own as a stand alone product, and in my view constitute one of the best all around music reproducers in their price range. For this reason, and by virtue of the fact that in certain key respects they offer a more satisfying performance on music than the whole ensemble of speakers together, they deserve to be discussed in their own right.
In essence the 2Ce is a front-firing, four-way cones and-dome system, a speaker complement it shares with hundreds of other models. But in several significant details of its design, it departs notably from normal practice. Under an obscuring sock, you can't really call it a grill because it is held well away from the cabinet by what I can only describe as tent-poles, is a type of enclosure known as a "pregnant kangaroo,' and formerly much more common in the industry The supposed resemblance to the bipedal marsupial is achieved by providing each type of driver with its own sub-enclosure and progressively offsetting each for physical time alignment, terracing them as it were.
Most "pregnant kangaroo" loudspeakers have proved little more than eyesores because their physical time alignment has been effectively nullified by electrical phase shift from the crossover. Vandersteen, however, has designed the 2Ce for time coherence in both domains, making it one of the few conventional multi-element loudspeakers with minimal phase shift through the audio band. (Others that approach this ideal are the Spica TC-60 and all of the Thiel, Dynaudio, and Dunlavy speakers). Like the Thiel and Dynaudio speakers, the Vandersteen utilizes a paradoxically complex crossover that still manages to achieve a first order, phase coherent, combined driver output by synthesizing a combined filter function incorporating the total transfer function of the drivers and the electrical network together.
Two woofers operating at different frequencies are used, one placed up front and high in the base enclosure, with the other situated in back near the floor. This strategy minimizes boundary effects in the bass, the strong reinforcements and cancellations which occur when sound is returned from floor ceiling and side walls and merged with the direct sound of the speaker.
Drivers are completely custom-made, and the midrange is particularly unusual in featuring Alnico magnets in the motor assembly Alnico is a rare earth magnetic material once widely used in premium loudspeakers and still regarded by many designers as the material of choice. Alnico greatly facilitates the design of low distortion magnetic circuits for loudspeakers, but current high cost and uncertain supplies have caused it to fall from favor. Only three other manufacturers, as far as I know, currently use it, SFI, Richard Sequerra & Associates, and Win Labs.
As indicated, the system uses separate sub-enclosures for each type of driver. The midrange and tweeter enclosures constitute what is known as a transmission line, which generally takes the form of an open tube stuffed with progressively increasing amounts of fibrous damping along its length. In terms of its ability to absorb a loudspeaker's back wave and to present a constant acoustical environment to the driver, the transmission line surpasses conventional sealed box or ported designs.
Finally, the speakers feature connectors that practically mandate bi-wiring or passive bi-amplification (the use of separate low and high frequency amplifiers sans an electronic crossover). In toto the Vandersteen design philosophy yields more than acceptable results in my opinion. Used as a stereo pair, their sound quality proved consistently superior across all significant performance parameters and over a wide range of program material.
The following observations concerning performance are based upon extensive listening sessions and upon laboratory measurements with the MLSSA automated loudspeaker testing system. Perceived sound quality correlated very well with the generally excellent test data. The system has no major frequency response aberrations except in the highest octave 10kHz 20kHz. Lower midrange below 500Hz is slightly depressed, on the order of about 1dB, but the speaker is not perceptibly thin sounding on that account. Above 500Hz deviations do not exceed 1dB all the way up to 10KHz, and the speaker manifests no upward or downward trend. When response is averaged to within a third of an octave, the output of the loudspeaker is very nearly a straight line.
Altogether, measured on-axis frequency response would be beyond reproach, but for an extraordinary aberration in the last octave. Response dips a full 10dB at 15kHz, and then shoots up 15dB or 5dB above average level at 17kHz.
Since my own hearing acuity falls off in this region, I played the speaker for a friend who hears beyond 20kHz. He heard nothing amiss. In fact what we both heard was an extraordinary ability to render high frequency overtones convincingly Recorded violins, pianos, cymbals, and woodwinds all shared a commendable naturalness. True, the tweeter lacks the ability to render intense high frequency such as trumpet blasts in a completely convincing manner, but in subjective terms its much better than average, in spite of objectionable measured performance.
Incidentally, Vandersteen himself acknowledges the top octave raggedness and attributes it to reflections from the minimal cabinet structure residing within the sock. He asserts that it is audibly unobjectionable, which, as indicated, is my finding as well. Without subwoofer augmentation, the 2Ce is taut and controlled in the bass while lacking the ultimate in extension and impact. As is the case with very many box speakers, midbass betrays a slight wooliness that detracts from musical realism in the rendering of transients in this region such as the open strings of the guitar As one ascends to the upper midrange the impression of overhang vanishes entirely however, replaced by an electrostatic-like sense of clarity and articulation. The 2Ce does not sound like a typical box speaker, and measured performance confirmed the absence of significant resonances across most of the frequency band.
Any loudspeaker having first order crossovers is particularly vulnerable to over-driving, and the 2Ce owners manual cautions against use of amplifiers exceeding 200 watts per channel. But regardless of its ultimate limits, the 2Ce sounded highly dynamic at normal listening levels, and did not compress transients in the manner of many popular two-way audiophile speakers.
Imaging ability of the 2Ce is extraordinary, at once panoramic and orderly, with sound sources snugly and securely niched within a wide, open soundstage. I am not going to engage in a lot of hyperbole about soundstages the width of the galaxy I'll only say that this speaker images on a par with the best forward firing quasipoint source speakers available. Naturally line sources, such as electrostats and planar dynamics, create a very different stereo perspective.
So how, you might ask, does Vandersteen make a speaker costing roughly a thousand dollars a pair that performs at such a consistently high level? I'll let the maestro speak in his own words. "Furniture grade cabinetry is the major cost factor in most high end speakers. I spend maybe a tenth of what other manufacturers do and hide the rough appearance under fabric. That way I can spend plenty of money on top quality drivers and crossover components and still make an affordable speaker."
The 2Ce in Company
Since the emphasis of the review is on a total surround sound system, we must now reconsider the larger question of how the superb 2Ce behaves with center channel and surround augmentation and what aspects of the total ensemble are synergistic with the SDP1.
Before we discuss system interactions in detail, it should be clearly noted that Vandersteens own preference is for the Hafler matrix and that he has used it in his personal system for over twenty years. Furthermore, he favors it for movies as well as music. Thus, the system under review represents an attempt, probably unique, to design a loudspeaker ensemble specifically around the Hafler matrix.
In attempting such a design, Vandersteen proceeded into unknown territory, since little or no formal research has been published indicating the preferred design of a loudspeaker array for passive matrix decoding. My own feeling is that highly focused, beamy loudspeakers would be best because basic separation in such a system is critically low to begin with, but that's merely an opinion. Interestingly, the VCC-1 center and VSM-I surrounds are narrow dispersion designs, while the 2Ces are typical wide dispersion quasi-point sources.
Structurally both the VCC-I center and the VSM-1 surround are decided oddities whose outward form gives little suggestion of how they might complement the 2Ces. The VCC-I is a tiny cube, approximately a foot square, while the VSM-1s are super bookshelf sized and extraordinarily shallow Both speakers use a single coaxial pair of drivers in order to maintain time alignment off-axis as well as to narrow dispersion. Both are provided with full length socks, and generally evince the characteristic Vandersteen look.
In marked contrast to most manufacturers of multi-channel systems, Vandersteen designs center and surround to be full range and to exhibit markedly different on-axis frequency response than that of the main stereo speakers. Both VCC-I center and the VSM-I surrounds have tipped up treble, and in addition the center has a depressed lower midrange and upper bass (low bass is largely absent).
Vandersteen's rationale for these choices is plausible in some respects, but open to question in others. He claims to roll off the midbass in the center to avoid what are known as mutual coupling effects among the three front drivers that would result in excessive bass energy and here I think that his decision is entirely valid. But his justification for elevating the treble in the center and the surrounds seems more debatable. Vandersteen reasons that the center will generally be placed off-axis anyway, above or below a television monitor, so best design it to sound flat off-axis. That makes some sense, but to my mind it doesn't make entire sense; and the reason I think it doesn't is because it places the center at a perceptibly as well as actually, different elevation than stereo left and right.
One can readily demonstrate that humans can localize the height of a sound source; therefore positioning speakers thus should result in an obvious discontinuity And for me it certainly does. I found that I preferred to listen to the system with all front speakers occupying the same level, even though a timbral mismatch occurs as a result. I'd add that I have yet to hear a surround system using an unlike center speaker that performed on a par with well designed systems using three like speakers, and I am particularly puzzled as to why Vandersteen followed this dubious strategy in a system designed as much for music as for movies.
As for the purpose behind the bright on-axis response of the surrounds, that ultimately eludes me. Vandersteen indicated that these two should be positioned off-axis to the listener, and since rear speakers are generally set up that way maybe he's just bowing to the realities of user preference. Yet these same speakers are also said to be usable as a main stereo pair. Would one then also want to place them high overhead?
Incidentally the surrounds are designed for wall mounting, and Vandersteen claims that the highly directional characteristics of the speaker prevents the kind of midrange anomalies heard in normal box speakers placed directly against wall surfaces. From my experience, I'd say he's correct. Both center and surround are time aligned and phase coherent. Both are precision transducers capable of resolving delicate transients without blur or overshoot. Scrutinizing the system as a whole, one would have to say that endowing all of the speakers with full range attributes was well considered in the light of the full range output of each channel of the matrix. The use of a highly directional design in the center also makes sense in view of the high levels of cross-talk between channels. But other relevant design issues do not seem to have been clearly addressed. For example, a Hafler matrix normally produces diffuse ambient information in the back. Is a highly directional front radiating speaker then the best choice for the rear channel? And what about floor and ceiling reflections in the front of the room? Should these be controlled by narrowing vertical directivity as in THX systems? The answers are not clearly evident.
Total Surround
I used the Vandersteen-Audio Research combination for both movies and music. For comparison purposes I also used the speaker system with three very different surround processors, the NuReality Vivid 3D, the Citation 7.0, and the Pioneer SP-99D with AC-3 decoding. In addition, I did bypass listening tests where I compared the SDP1 to straight stereo. Most of the music listening involved analog disks played over a Win phono system including Win Labs SEC-10 turntable, Win FET-10 cartridge, and Mission Mechanic tonearm, while video material consisted of laserdisc exclusively played on a new Pioneer Elite SP-99D.
Here's what I found
As Audio Research promises, and as other reviewers have noted, the SDP1 is indeed musical, evincing a warm, inviting sound quality that I found gently subtractive. Turning center and surround down all the way, and performing rapid comparisons between the straight output of the Win front end and the two channel output of the processor, I noted a subtle constriction of dynamics and a slight golden haze. But then any line level component is going to do something to the signal. What the SDP1 does is benign and musically right.
The small mischief that even the best processor will work on the signal is supposed to be offset by benefits of the processing itself, in this case the derivation of extra channels. Undoubtedly the presence of these extra channels significantly alters the spatial perspective one hears, but I found myself questioning the benefits of the alterations.
What I hear through this processor over these speakers is an image which is enlarged but made somewhat indistinct; vaporous is a word that comes to mind. Used with the Vandersteen array, the SDP1 provided more precise placement of sound sources than I've heard from other passive matrix processors over other speaker systems, but precision was always lost in moving from two channels to four (although certainly an impression of larger size was gained). Since the Vandersteen 2Ce's, used as straight stereo reproducers, are capable of pinpoint imaging within an already extraordinarily wide, deep soundstage, one is in a sense effecting a swap by adding channels and processing, a swap that may not be entirely to one's advantage.
I can already envision hordes of loyal Audio Research dealers writing in to complain that Sweeney is deaf, crazed, perverse, or all of the above for expressing such opinions. (Indeed I was exposed to just such abuse when I failed to enthuse over another passive processor, the recent Panor version of the hoary Dynaco QD-I.) In response to such anticipated objections, I can only say that measurable loss of separation is an undeniable fact with any passive decoder, and that others have reported similar impressions concerning the general quality of imaging afforded by low separation processors sui generis. On the other hand, signal purity is undoubtedly higher for the SDP1 than is the case for active matrix decoders, and basic sound quality is almost certainly superior to that of data compressed discrete systems. Ultimately the issue boils down to taste. If you seek above all else to minimize sonic Contaminants, and at the same time place a higher emphasis on envelopment than on pinpoint imaging, you'll probably like the SDP1 and you'll certainly want to consider the Vandersteen array as an accompaniment. If you favor a more etched delineation of spatial relationships, I think you'll be disappointed. Frankly if I wanted to add a surround dimension to two channel stereo I'd consider the SRS circuit over the Hafler matrix, since the former extends the sound stage while actually stabilizing image placement.
Though the Vandersteen-Audio Research combo would appear to be best suited to music listening, it proved unexpectedly rewarding on movies. I'm not sure I'd trade the aural pyrotechnics provided by the Citation's incredibly agile logic steering circuitry for SDP1's dulcet but somewhat sketchy suggestions of cinematic space, yet I was consistently charmed by the passive processor's sterling sound quality I also found that, true to Audio Research s claims, their processor passed an incredible amount of low level detail especially through the rear speakers. The pervasive and abrasive grain of logic decoders fell away from the soundtrack, leaving, to be sure, damning evidence of the disorderly strata of overdubs comprising most movie soundtracks but also retrieving all manner of ambient effects the recording engineer probably forgot putting there. In short, the SDP1 shakes out unexpected treasures from the mix, and Vandersteen's high resolution loudspeakers deliver them intact.
Regarding pans and image placement, the SPD1 was better than I expected, though a far cry from the Citation which seems to view the sonic landscape through Cartesian coordinates. Sounds do track the on screen action with the SDP1, but not with the degree of adhesion provided by a good logic steering unit. I also found the vertical discontinuity between center and stereo speakers seemed to blur pans.
My wife, who is keener eared than I, and far less forgiving of bronchial disorders in processors, pronounced the SDP1 the best sounding surround processor ever, and in that I will concur if one's only criteria are absence of distortion and the preservation of signal purity But if good stereo has long accustomed one to images that occupy fixed and determinate positions, then the way that the SDP1 relaxes that sort of exactitude may not be to your liking. Terry Dorn of Audio Research told me that in listening panel evaluations, the panelists unanimously preferred the processed four channel output afforded by the SPD1 to straight stereo, so perhaps my reaction is aberrant. Just remember, though, that no passive matrix decoder has ever won much of an allegiance from the listening public heretofore, and that these selfsame arguments that separation doesn't much matter have been put forth previously by the quadraphonic boys and the Ambisonics camp and ultimately rejected.
As with any component, I'd request a dealer loan before buying an SDP1, and if you could have this one for a week, it'd be all to the good. Failing that, buy a Chase HTS-1 for a hundred bucks to experience a speaker level implementation of the same basic Hafler matrix. Because the Chase degrades amplifier damping, the effect of the Chase is not precisely the same as that of the SDP1, but it will give you a good idea of the capabilities and shortcomings of the passive approach.
Clearly the Vandersteen-Audio Research combination is synergistic, but, as indicated, I did listen through other decoders as well as the SDP1, since the purchaser of the Vandersteen array is at least as likely to use Dolby Pro Logic or AC-3 as passive matrix.
Used instead with Citation's superb 7.0 AV Controller or the Pioneer SP-99D AC-3 decoder, the Vandersteen system produces much more solid and defined sound images reflecting the precision steering of the Citation on the one hand, and the inherent high separation of the Pioneer on the other. For video I would use either in preference to the Audio Research. But here I must also point out that if most of my listening involved video sources, I would not select the Vandersteen array as my first choice. I think that the highly separated, dramatically panned soundtracks that typify movie recording today come off best on systems using like speakers up front, most of which happen to bear the THX logo.
Still, Vandersteen center and surround speakers do make sensible additions to the 2Ces if you want to add a video capability to a truly excellent stereo music speaker, rather than opt for a dedicated video speaker system. And the 2Ces are so desirable in and of them-selves that that may be the best course in building a dual use sound system.
In Conclusion
Frankly I don't know what to make of this seemingly quixotic embrace of the clearly moribund Hafler matrix on the part of both Vandersteen and Audio Research. I understand the value they place on signal purity but I don't believe that the Hafler matrix offers an ultimately more convincing illusion of a real performance than an optimized two channel system.
In saying this I am not fighting a rear guard action against multi-channel. Ultimately multi-channel sound reproduction offers undeniable advantages over two channel, but here it must be stated that the Hafler matrix can only be considered multi-channel by a very elastic definition of the term. In truth it is a bastard format - always maintaining a certain cult following, it is true, but never winning the allegiance of most audiophiles.
Resurrected today with products from two of the most respected high end manufacturers, can the Hafler matrix finally succeed? I think that, as has always been the case, it will attract a very defined group of listeners who are prepared to overlook its shortcomings. My opinions are not the last word on the subject, so by all means listen to this combination. Just remember that it will not be to everyone's taste.
I'm Listening To...
Richard Vandersteen lives large, he is even over-achieving for a Type A, running like a dynamo from dawn to dusk at his loudspeaker factory. And in his spare time? Well, flying his own airplane (a Rockwell Commander, high-performance single, he just got a 97 on his aviation score). Or competing in his own race car (IMCA super-modified. This was a good year, racewise, Vandersteen and his team won the Kings County, California, track championship. By the way, he always wears ear protection!) Or maybe boating and water-skiing in a Lavey Tunnel, "which has a hydroplane bottom and a ski-boat top, so it doesn't look as fast as it is." He doesn't race the boat much, "just enough to win a little money up at the lake." Vandersteen pauses a tick: "I just love machines!"
To relax sort of, he listens to music. His system, which is quite large, goes all the time. "We are not a TV family," he says. He has about 6000 LPs and over 1000 CDs. Richard's wife, Eneke, shares his love of music. "We work as a listening team, critically as well as for pleasure. She appreciates different things. And between us, we cover a broad band of music, from country western on. We listen to everything." Good for relaxing is Mary Black, Babes in The Wood (DAR Records D2-77528), for the lovely voice and the simple music. "Yet there is so much going on in the background," he says, "you hear something new every time you play it. I think much of this was recorded at one session, the instruments are almost talking to each other."
The Fairfield Four, Standing in The Safety Zone (CD Warner 9 26945-2) again is relaxing. "The ingenious harmonics, and how the voices interact." It's also one of the rare ones that's good all the way through. Much of the Vandersteens' listening is multichannel. Not for the movies, though with three teenagers in the house, movies are popular. But Richard and Eneke listen to multichannel because it makes audio more interesting. "Most recordings are enhanced in a positive way," Richard says. He uses the Audio Research SDP- I surround processor based on the old Dynaco circuit. "On an old tape delay, the machine that had a continual tape loop, and the location of the head was always changing." Regardless, he has moved firmly into the world of multichannel sound, finding that the music experience is enhanced by the addition of good surround information. The Eagles' Hell Freezes Over (laserdisc - Geffen Home Video ID3061GF) is great for multichannel, both in video format and on CD. "Not that this is particularly relaxing," Vandersteen admits. "The music is complex. And on multichannel, it is even more challenging." Eric Clapton's Unplugged (laserdisc - Warner 938311-6) is another multichannel favorite "It's not spectacular," he says, "but we can listen to it over and over and never grow tired of it. Sometimes I put on the video and the CD," he laughs, "and try to get them in synch. I get such a sense of accomplishment if I can do that, and then of course it lasts only seconds." A problem to be solved, even in fun? And that's relaxing!
"True. I never seem to do one thing at a time," Vandersteen admits. "At work there are 50 things going at once, and I always want to make everything better. Even at home, I listen and read. When do I really relax?" A pause. "When I go away. The minute I leave work and the house behind, I slow down. And everybody wonders why I seem lethargic. Otherwise music is really the only escape."