Audio Research CD2 CD player
"Actually," I said, thinking about the Audio Research CD2 CD player I'd been auditioning, "it seems like the better products are sounding more and more alike—no matter what their design strategy might be. I wonder why."
"I think I know," JA said. "It's the same reason there are no longer any .400 hitters in baseball."
Silence resounding. The three of us were staring at John, wondering how he was ever going to connect those dots.
"I just finished reading Stephen Jay Gould's Full House, which argues that progress and complexity are not valid yardsticks for the evolution of systems. Instead, Gould reasons, you have to look at the changes in variation within a system—in other words, as systems improve, they equilibrate and variation decreases. Using baseball as an example, Gould proves that the absence of the .400 hitter is not indicative of a general decline in batting skill, but rather of an overall increase in all baseball skills. Shrinkage in variation must be measuring a general improvement—within the system of professional baseball, the range of variation between the best hitters and the worst is narrower than it's ever been (footnote 1).
"The High End is like that now. The difference between the best and the worst products is narrower than it has ever been. The overall standard is so high that we don't see many products that stand head and shoulders above all other challengers—nor do we see many products so poorly designed that they're objects of ridicule."
Nowhere is this more apparent than in the state of contemporary CD-player design. Over the last year or so, I've been privileged to audition quite a few ambitious CD players—and all were superb. Oh, they were subtly different, each having a strength or strengths that distinguished it from the others. But the actual differences were minute—and, I'm sure, inconsequential to some listeners. But that's only to be expected as the standard continues to rise.
Will we eventually reach a point where progress is no longer possible? I doubt it. We'll reach a point where the momentum slows, where improvement is accomplished in extremely small increments, but no matter how flat the upward curve gets, we're still short of perfection. But we have reached the point, as the Audio Research CD2 illustrates, where the variation between the finest examples has become bafflingly small.
It is a great advantage for a system of philosophy to be substantially true.—George Santayana (1863-1952)
I'm used to manufacturer hype—I tend to expect it in daily rations—so it came as something of a surprise to find Audio Research quite modest on the subject of their CD2. As marketing VP Terry Dorn described the rationale behind the unit, he stressed ease of use, the inherently low jitter of single-chassis designs, and the future-proof logic behind outfitting the player with digital outputs. "And people do seem to like the sound," he deadpanned. ("Power is adequate," as Rolls-Royce likes to say.)
Rich Larson, one of ARC's engineers, filled in some of the details. "It's rather standard in terms of its digital circuitry, but the Crystal CS4329 DAC has a good 20-bit filter. It's a double DAC, so the signal operates differentially from the chip through the analog stage. It also has excellent linear phase filtering. One-bit delta-sigma designs have better-sounding results than any of the multibit designs we've heard: they have differential phase linearity, and lower distortion than most multibits. The CD2 has low noise and exceptionally gentle filtering. The Crystal chip comes with digital de-emphasis that is much better-sounding than most analog de-emphasis we've experimented with. We design our own analog filter—also very gentle.
"We look for good numbers, but we develop everything on the basis of listening tests. These told us we needed to add damping to the transport and chassis, and even to specific components on the board.
"We have a lot of proprietary techniques in the power supply and the analog stage—that's where most of the sound quality comes from. The power supply is fully regulated. The analog stage is DC-coupled, and uses both J-FETs and bipolars—it's very wideband."
The CD2 has the wide, black metal handles and oversized faceplate—the review sample's was finished in titanium gray rather than the traditional brushed aluminum; black is also available—that have come to be Audio Research's trademark, but the front panel is definitely low-key. The drawer mechanism dominates the left center and the digital display mirrors it on the right. A row of seven well-marked pushbuttons runs across the bottom, controlling the basic transport functions. The remote recapitulates these controls and offers programming and display options not available from the front-panel buttons.
The rear panel is equally straightforward. Analog outputs are available through XLR balanced or RCA single-ended connectors. Digital outs include TosLink, coaxial BNC S/PDIF, and AES/EBU. A glass-fiber output is available as an option. An IEC mains plug allows the tweaky to swap cables to their hearts' content, although I should point out that ARC supplies what appears to be a standard modular power cord—but one that has a noticeably high standard of fit'n'finish. (Its female end provided one of the tightest fits of cord to player that I've ever seen.)
One last thing. The CD2 is not turned on and off—the front-panel power button puts it in standby mode. Once powered and stabilized, how it sounded on awakening from standby wasn't substantially different from how it sounded after hours of play. Impatient sort that I am, I approve of this.
Footnote 1: I went home and read Gould's Full House—it's so well written and -argued that it's a compulsive page-turner. I devoured it in a single sitting. Rather than try to recapitulate Gould's argument in detail, I'll just recommend that you read it.—Wes Phillips