Digital Processor Reviews

Will HDCD catch on? A lot depends on what happens over the next year in the development of DVD, which promises a new, high-resolution format which might just revolutionize audio software. But it's not available now, and HDCD is, if only in limited supply. All I can say is, when I switch from a cross-section of my best-sounding standard CDs to the small population of HDCDs now in existence—and given that I'm not talking about identical program material, but only general overall impressions—the clouds do not part, and the sun does not emerge to cast life-giving light on the fertile plains. When we get our hands on an HDCD encoder/ADC, so we can compare an original source to an HDCD recording of it—and perhaps in comparison with a first-rate conventional digital transfer of the same material—then we'll talk.

Comparisons: Theta, Spectral, Sonic frontiers
Though it's hard to pin down a specific level of performance with any given component and state that this is the point at which noticeable improvements in sound quality come only with precipitous increases in cost, the No.36 is as good a point as any. Though some will set that point much lower, I think I can say with assurance that the No.36 is definitely on that part of the curve beyond which incremental improvements come at a stiff price.

The only way to be sure, of course, is to compare the No.36 directly with a variety of other processors. With the Denon transport's two coaxial digital outputs and the Rowland preamp's precise and storable input level settings (the No.38 and 38s also have these capabilities), I was uniquely positioned to make these comparisons. Material could be played back and compared long-form, or instantly switched at matched levels. The latter capability required no extraneous equipment beyond the basic system itself. It was not a blind comparison, but was otherwise tightly controlled.

I compared the No.36 directly with the Theta DS Pro Basic III, the Sonic Frontiers SFD-2 Mk.II, and the Spectral SDR-2000 Professional D/A converters. Except for the Sonic Frontiers, which I auditioned in both its balanced and unbalanced modes, these comparisons used unbalanced connections. The digital feed was again the Denon.

First up, the Theta DS-Pro Basic III, a processor costing approximately half as much as the No.36. How did it differ? First, it's fair to say that the Theta is a superb D/A in its own right, with good detailing, depth, and a lack of any irritating qualities—nothing artificial or off-putting about the sound of this converter. If I were shopping in this price range, I'd put the Pro Basic III high on my list of candidates to audition.

What, then, do you buy with the No.36 that you don't get in the Theta? The No.36 had a quality of focus—call it clarity, transparency, inner detail—which the Theta cannot quite match. The soundstage was just a little less tightly defined with the Theta; multiple sources within the stage had less space between them. Solo voices and instruments were just a little more "there" on the Levinson. At the top end, the No.36 had more air and is more open. The difference isn't mind-bending or even close to it, but it's audible. And while I heard no inherent difference in depth between the two units, this added openness at the top added more space to the sound of the No.36; ambience, in particular, was more convincingly rendered.

At the bottom, however, it was a different story. As I went back and forth between the two processors, I initially heard no notable differences. But with continued listening, the Theta began to pull ahead. Though both processors were comparable in extension, the Theta sounded tighter and more detailed. The difference, again, was not enough to have me dancing in the streets, but I have to give the Theta the nod in its portrayal of the bottom octaves.

Next up: the Spectral SDR-2000 Professional HDCD D/A processor. RH salivated over this processor in Vol.18 No.5, and this was my first chance to hear it in my system. He did not exaggerate. It is clearly a Class A processor; it may even be the best processor in Class A, though I haven't had the opportunity to compare it directly with the Mark Levinson No.30.5, which isn't exactly chopped liver. [When I did this comparison, I found the Spectral to edge ahead in sheer coherence of its soundstage presentation, though its HF balance sounded a little too tilted-up in the context of my system.—John Atkinson.]

At first I found the differences between the SDR-2000 and the No.36 to be subtle—less notable than the differences between the No.36 and the Theta, above. Shortly, however, two primary differences crystallized. First, there was an increase in high-frequency air with the Spectral, combined with a boost in clarity and fine detail. This came without any of the corresponding edginess or roughness that often ride along with enhanced definition. (RH's review measurements indicated a very small top-end rise in the Spectral's frequency response which might account for this, though at +0.3dB at 20kHz, it doesn't seem terribly significant—and I can't hear 20kHz in any event.) And second, the Spectral was leaner through the upper bass and lower midrange. Riding on the coattails of these primary differences was an increase in soundstage precision, in lateral focus and depth. This improvement struck me as a result of the tonal changes, not an independent enhancement.

After a period of familiarity it was hard not to appreciate what the Spectral did for the sound of CDs, from the subtle percussive details on Mokave's Afrique (AudioQuest AQ-CD1024) and The All Star Percussion II (Golden String GSCD 013), to Branford Marsalis's ethereal saxophone on the soundtrack from Sneakers (Columbia CK 53146), to the separation of vocalists on recordings as dissimilar as the King's Singers' Good Vibrations (BMG Classics 60938-2), and Postcards (Reference RR-61CD).

Stated this way, these differences sound dramatic. They weren't. The more I listened, the more obvious they became—and the more I agreed with RH's conclusions. But the No.36 was very close. It was just a bit soft-sounding next to the Spectral. In one respect, however, I did prefer the No.36. Its slightly greater warmth sounded more natural with the human voice. The Spectral may have been a hair better defined at the very bottom end of the spectrum, but I wouldn't base any purchase decision on the bass differences between these two excellent processors.

Which would I buy? If money were no object, the Spectral. The more I listened to it, the better I liked it. But at more than twice the price of the No.36, it had better be better. The fact that the differences are there will be of significance to those with the resources to take advantage of them. But I'd certainly be more than happy living with the No.36. The important point is not how good the Spectral is (and it is good), but rather just how close the No.36 is for half the price.

At $5295, the Sonic Frontiers SFD-2 Mk.II is somewhat more expensive than the No.36, but of the three processors compared here, it's closest in price and a logical choice for a face-off. Again, the match was a close one, but not without noticeable differences. The Sonic Frontiers had a slightly more powerful low end, but the No.36 was noticeably tighter and better-defined in this region. The fuller bottom gave the Sonic Frontiers an appealing warmth; next to the No.36, it was richer and fuller on vocals—the one area where I definitely preferred it.

In the upper range, the No.36 had a more subtle, airy treble. The difference didn't jump out and grab me, but percussion and other instruments with strong high-frequency overtones were more finely shaded on the Levinson. On the other hand, through the low and mid-treble, the Sonic Frontiers was more vivid, more "there." I never found it to be unnaturally bright in this region, though here the differences with the No.36 were most obvious. More often than not the Frontiers added a subtle but effective presence to the sound. There were few, if any, differences in depth and soundstaging—but to the extent that there were, I attribute them primarily to this small spectral shift.

When I switched to the balanced mode on the Sonic Frontiers, the sonic differences between it and the No.36 became even more subtle. But the contrasts were noted in the same areas, and I would not significantly change my observations.

On balance, I have a tough time declaring a preference between the Sonic Frontiers and the Levinson processors. I liked the good old "palpable presence" of the SF, and the top- and bottom-end detail and definition of the No.36. I can't imagine anyone being unhappy with either, but of course definite preferences will develop depending on the individual listener and the system. If they fit your budget, you should definitely hear, and consider, both.

Conclusions
What can I say? Run out and buy the No.36? That would not be bad advice if you have the price of admission. It certainly must be heard, even if only to hear what's possible in today's best D/A converters. Make no mistake: the No.36 can compete in that company (footnote 2).

---

Footnote 2: And just add icing to TJN's cake: after I purchased the Mark Levinson No.30.5, I spent some time comparing it with the No.36. The '36 gets 95% of the way to matching the awesome '30.5—at one-third the price! All it lacks is the ultimate low-frequency authority and soundstage definition.—John Atkinson