Hyperion HPS-968

      The Hyperion HPS-968 shows outstanding promise, and, at its $6500 per pair MSRP (USA), is a wonderful bargain for such a high class loudspeaker system.
      The new HPS-968 design is Hyperion's step up from their older, junior HPS-938 model that has garnered such favorable reviews. The new 968 design is also Hyperion's answer to the requests and assessments of those who have lived with the junior 938, so it should be superior in every way. Both are 3-way systems, but the specific drivers are all different.

Superb Drivers

      The key to the 968's outstanding sonic promise lies in its 3 drivers, which are superb units. Drivers are of course the heart of a loudspeaker system, being literally the total and only voice through which the system speaks. Without great drivers, you can't have a great loudspeaker system, since the system performance can never be better than the performance of its drivers.
      Starting from the top, the 968's silk dome tweeter, covering the entire spectrum from 3000 Hz upward (the upper midrange and all treble regions), boasts superb transparency and speed. It is much faster and more transparent than the already excellent silk dome tweeter in the recently reviewed Von Schweikert VR-4jr. Note that the 968's price of $6500 is only slightly more than the $5600 of the VR-4jr, so the 968's superior tweeter driver is already an indication of the wonderful bargain that the 968 promises to be.
      Like all silk dome tweeters, the 968 tweeter euphonically transforms trebles to be sweet and slightly soft (as opposed to the pistonic sonic accuracy of some hard domes, e.g. B&W's diamond tweeter). This sonic sweetening and softening can be euphonically beneficial on most recordings, which are typically too closely miked, and which therefore have trebles that are hard in quality, harder than you are accustomed to hearing from real, live music, since you typically sit at a much greater distance than these closely positioned mikes are placed. Thus, a superb silk dome tweeter like the 968's can euphonically transform most recordings to sound more real, more like the sonic experience you hear from real, live music.
      The 968's midrange driver is likewise superb. It has a heroic task, covering a very wide spectral range, from 150 Hz to 3000 Hz. And it covers this entire range with excellent transparency and very low coloration. The design of this midrange driver is unique in several ways. The 6.5 inch (nominal) cone is made of woven carbon fibre, and is held in place at its outer edge by a stiff surround, stiff enough to obviate the need for the usual spider at the central voice coil former (a spider provides centering restoring force in other drivers).
      Then, this 968 midrange cone does not continue inward toward an apex, to meet the voice coil former, as is usual in other cone drivers. Instead, there is a large flat disc, made of shiny, hard, rigid, plastic, located at the center of the driver cone, and so wide that it takes up about half the entire diameter of the midrange cone. This flat disc actually acts as a structural intermediary, coupling the cone to the driving voice coil former (to which it is attached via integral rear-facing struts). The advantage of this large, flat, rigid disc is that it supports the woven carbon fibre cone much farther out (in diameter) than the usual support point at the voice coil former, and thereby keeps the smaller cone area more rigid and more pistonically accurate out to higher frequencies. Thus, the 968 midrange driver is able to deliver its very wide spectral range with better transparency, better pistonic accuracy, and less cone breakup coloration, than other conventional midrange drivers that have their larger cone area tapered, attached, and driven closer toward their apex.
      Thanks to the large area of this rigid coupling disc, this disc also directly radiates a lot of the midrange driver's output, probably comparable to the output from the woven carbon fibre cone area. One disadvantage of this large rigid disc is that it seems to create a slight buzzing driver resonance around 500 Hz, but an electrical notch filter in the crossover feed to this driver is apparently intended to offset this, and we don't hear this resonance as a salient problem. Another disadvantage is that this large rigid disc weighs more than an equivalent area of woven carbon fibre cone would weigh, so driver efficiency is reduced - but this midrange driver features very rugged and massive construction, including a huge magnet that would do credit to a woofer driver, and this strong magnet also boosts driver efficiency, compensating for the heaviness of that large rigid disc. The cast driver frame/basket is very rigid, while also providing a very open structure with thin struts, to minimize reflective colorations from the rear wave.
      There are also unusual features in the way this midrange driver is installed and connected, in the 968. The 968 crossover lacks the usual electrical high pass filter, which usually cuts off low frequencies to keep them out of the midrange driver, those low frequencies which are being fed to the woofer (here below 150 Hz). Thus, the 968 midrange driver also plays all bass frequencies, albeit at a much reduced level (thanks to its stiff outer surround suspension). The rear wave of the midrange driver is absorbed by the usual padding inside the midrange/tweeter enclosure, and is also vented by a port at the rear. Interestingly, this port seems to be tuned, and tuned to the very low (for a midrange driver) frequency of about 45 Hz (a downward sine wave test sweep produced a hump in midrange driver output around 45 Hz, after a dip around 55 Hz, with negligible cone motion from the midrange driver).
      The 968's twin 8 inch woofer drivers are also excellent. Their aluminum cones provide accurate pistonic motion, and their surrounds allow large excursion for hefty low bass output capability. Since the woofers are electrically rolled off above 150 Hz, they act almost like subwoofers. Then, since no signal is input at the much higher frequencies where the aluminum cones would break up and cause metallic coloration, these woofer drivers also have outstandingly low coloration.


      The cabinetry of the 968 merits special praise. The basic 968 cabinet format, like that of the 938, follows the model made famous by the Wilson WATT/Puppy, a rectangular woofer cabinet topped by a separate midrange/tweeter cabinet of tapered shape. The 968's midrange/tweeter cabinet is supported atop the woofer cabinet by a 3 point system, which is more stable and less prone to micro-rattling and micro-movement than the usual 4 point support system. Moreover, the rear single support point of this trio is easily micro-adjustable for height (and then lockable), so that the top midrange/tweeter module can easily have its vertical radiating axis optimally adjusted, to point directly at your chosen listening height. The 3 metal points, pointing downward from the bottom of the midrange/tweeter cabinet, are actually small radius balls, rather than sharp points. These small radius balls then sit in slightly larger radius metal cups, sitting atop the woofer cabinet. These cups are not affixed to the woofer cabinet, but instead are small free-standing metal cups, free to slide around the woofer cabinet top, save for rubber backings. Cleverly, the cabinet's gloss finish helps these rubber backings to develop an adhesive-like cling, so they have enough friction to hold the top cabinet almost as immovably as would receiving cups countersunk into the woofer cabinet's top surface This rubber layer also provides the midrange/tweeter cabinet with beneficial vibration isolation from the woofer cabinet.
      The 968 cabinetry has a gorgeous gloss black piano finish, of the quality you'd expect to be on a far more expensive system. The cabinets are well constructed and rigid, with reasonably inert (dead) panels. Even the cardboard packaging deserves mention. It is the most intelligent design we've ever seen for any product, furnishing excellent shipping protection with minimum wasted volume, and makes it easy and logical to unpack the product. When designers pay this much attention to a detail like package design, and achieve such an intelligent result, it bodes well for all the other engineering design details that contribute to overall system performance.
      Indeed, the 968 does abound with yet more engineering details whose thoughtful design bears this out. The crossover uses premium oil-filled capacitors as bypasses for critical locations, and premium wiring. Three pairs of input terminals furnish individual electrical access to each driver, allowing tri-wiring or tri-amping (plus, as we shall see, individual driver polarity control).

Sonic Pros and Cons

      The Hyperion HPS-968 has many sonic strengths that are hallmarks of great loudspeaker systems, and the 968 has these strengths to an outstanding degree.
      We evaluated the Hyperion using our high resolution lab reference system, including B&W 802D loudspeakers fed by Nordost Valhalla cable from a variety of power amplifiers, surround processor by Arcam, power cords by VonGaylord and Wan Lung, and interconnect cables by VonGaylord and Mapleshade, with reference sources being an Esoteric DV-60 universal player mounted on a Mapleshade support system and an Arcam DV-139 universal player.
      The 968 is superb in intrinsic basic transparency, revealing a lot of information from recordings that escape most other loudspeakers, and indeed is even more transparently revealing than some great loudspeakers (such as the recently reviewed Von Schweikert VR-4jr). The 968 is likewise superb in exhibiting very low coloration over most of the spectrum; in most sonic aspects the 968 does not reveal itself nor sonically intrude, even as it reveals so much about the program material.
      To obtain the maximum transparency and coherence from the 968, the optimum listening axis is with each 968 toed in so it points directly at you, and with the tilt of the top head cabinet adjusted so your eyes are perfectly aligned with the top surface of this head cabinet (so you just barely miss being able to see any of the top surface of this head cabinet). The 968 comes with a laser pointer and holding jig (generously finished in the same piano black as the cabinetry), intended for aligning the loudspeakers with ease. But this beautiful and well-intentioned jig is useless, since the jig is not adjustable, and it sets the vertical listening axis as being halfway between the edges of the midrange and tweeter drivers, whereas we found the best sound to be at a higher listening axis, especially when the midrange driver is connected so as to be, over most of its range, in the same phase polarity as the tweeter (see below).
      Spatial imaging of the 968 is likewise superb. The loudspeaker location completely disappears aurally; the portrayed stage is uniform, deep, and wide (much wider than the loudspeaker locations); and there's a rich sense of depth and ambience.
      The 968 covers the frequency extremes very well. Bass is deep, extended, and powerful. And treble is fast, airy, and extended (with the euphonic sweetness from the silk dome, noted previously). The 968's wide spectral coverage is reflected in its rated system response, which is 25 Hz to 25 kHz. The midranges of the 968 are also correct in quantity, matching the bass and treble regions, for a substantially flat, neutral tonal balance and frequency response.
      So far, we have in the 968 a loudspeaker system that seems clearly destined for greatness, and at a bargain price for such outstanding sonic capabilities.
      However, the 968 also unfortunately has 3 (make that 2 ) sonic weaknesses, which are striking enough to keep the 968 at least temporarily out of the pantheon of great loudspeakers. We say temporarily because these sonic weaknesses should be quite easy for the 968 design engineers to fix, and if they elect to do so, then the fixed model (call it the 969) would surely be a truly great loudspeaker.

Inverted Midrange Phase Polarity

      The first sonic weakness is that the 968 midrange driver plays with inverted phase polarity over most of its range. Ironically, the very same excellence in pistonic accuracy and transparency, which makes this midrange driver so extraordinary in its intrinsic sonic fidelity, is also precisely the factor which makes this inverted phase polarity so highly audible, so obnoxious, and such a detriment in the 968.
      You see, most other midrange drivers are in non-pistonic breakup mode over much of their range, and also are less revealingly transparent of the input signal, than the superb 968 midrange driver intrinsically is. This non-pistonic breakup mode in other midrange drivers is inherently random in phase, thereby scrambling the phase of the program material input signal, and thus making it the case that there is less of a sonic difference between those midrange drivers being operated in correct phase polarity or in inverted phase polarity. Also, these other midrange drivers, being less intrinsically transparent than the 968's, are naturally less revealing of whether the program signal is being reproduced correctly, i.e. in correct absolute phase polarity. The result is that it makes a relatively small, subtle sonic difference on most of these other midrange drivers, whether the signal fed to them is correct or inverted in absolute phase polarity.
      But the 968's midrange driver is a whole different animal. It is very transparently revealing, and also seems to intrinsically have excellent pistonic accuracy (thanks in part to its unusual design), so it very clearly and accurately and cruelly reveals the phase of the signal it is fed. The midrange driver in the 968 is fed an inverted polarity signal input, so it reveals that polarity inversion very obviously in its acoustic output.
      How is the sound of the 968 affected by this inversion of the midrange's absolute phase polarity? Instead of merely telling you how it sounds, we'll show you - or, more precisely, we'll teach you how to show yourself, so you can hear this sonic effect for yourself. Inverting the polarity of an audio signal literally turns the waveform upside down, and makes a loudspeaker driver suck air in when it should be pushing air out. Since the human voice is an important component of much of your listening (singers, film dialogue, etc.), it's important to know how polarity inversion affects the sound of the human voice, and you can demonstrate this for yourself.
      To hear the human voice in correct phase polarity, simply say the word "Pop!" loudly and forcefully. Notice several things about the sound. First, notice that you are forcefully exhaling during the entire sound (for both the "p" consonant sounds and the "ah" sound of the vowel "o"), so forcefully that you could blow out a candle. Thus, this sound has in reality a clear absolute phase polarity, with air forcefully pushing outward.
      Second, notice that this sound projects an assertive, dynamic, strong quality - like a mini-explosion sound effect in the face of whomever you are saying "Pop!" at.
      Third, imagine that this other person has his eyes closed, and can only hear you saying "Pop!" at him. What sonic stereo image or spatial image of you and your real life voice would he hear? Your sonic image would project outward and forward, literally popping out and forward of the background, with tangible presence and immediacy. That's because saying the word "Pop!" produces a strong outwardly pushing force of air, just like a mini-explosion would. From our human experience listening to things in the real world, we know that things which produce this kind of forceful pushing explosion are three dimensional, solid and tactile, and convex in the curve of their outer shape (things like cannons, human singers with big chests, etc.). Thus, even with his eyes closed, your friend at whom you say "Pop!" can hear that you sound like a three dimensional, solid, real person that he could reach out and touch. Indeed, you can hear this for yourself from your own voice, when you say "Pop!".
      Now, let's repeat this experiment, but this time let's invert the absolute phase polarity, and then take note of the ways this polarity inversion makes you sound different. To invert the polarity, we must turn the signal waveform upside down, making the 'loudspeaker driver', in this case your voice, suck in air when it should have been pushing out air. So, simply say "Pop!' again, as forcefully as you can, but this time say the entire word while you are deliberately forcefully inhaling through your mouth (make sure you are inhaling for the consonant "p"s at beginning and end, as well as for the central vowel sound).
      Notice several things about this sound, and about how it differs from the correct polarity real life sound you made above. First, notice how hugely different this sound is, and how weird it sounds. There's a huge sonic difference, arising merely from inversion of absolute phase polarity. And you're hearing this huge difference and huge weirdness in a real, live sound. Thus, the better and more accurate a playback system is, the closer it will come to revealing this huge difference and huge weirdness (which you can hear in a real, live sound), if the absolute phase polarity is inverted anywhere in the recording/playback chain.
      Second, notice how the sound of saying "Pop!" while inhaling, no matter how strongly you try to do it, still sounds weak and wheezing, instead of forceful and dynamically strong as the correct polarity, exhaling version did.
      Third, notice how you sound hollow and ghostlike, as though your outer shape were concave from the sucking in you're doing. The sonic image of your real, live voice is utterly transformed, to sound like a hollow ghost far away, instead of the up front, immediate presence projection of a solid, tangible, real, three dimensional person that you created above by saying "Pop!" in correct polarity. That friend standing near you with his eyes closed would hear your 'stereo image', even live, making you sound like a wheezing hollow ghost, placed farther away than you really are. And of course this same effect would afflict all voices you played back from recordings, if your system inverted absolute phase polarity anywhere in the chain (in this case the 968's midrange, the spectral region where most of human voice information resides).
      Obviously, the inverted polarity version of saying "Pop!" sounds hugely wrong, and weirdly wrong in many ways, even in a real, live sound. These same huge sonic errors, which you have now heard for yourself in a real, live sound, would likewise apply to all real, live sounds captured in a recording and played back by you, if the absolute phase polarity is inverted anywhere in the chain. Our

(Continued on page 157)