dCS Elgar D/A processor:
Page 2
Sound
I first used the Elgar as a conventional processor, its level control set to 0dB and its analog outputs feeding the preamplifier. The Elgar was able to lock to all my digital sources in its narrow-window, "Digital" PLL mode, so that was how I used it. The unit runs hot—don't place other components on top of it.

The first series of listening sessions was both satisfying and dissatisfying. Satisfying because the sound was always musical; I enjoyed my favorite CDs and DATs for hours on end, one leading to another. Dissatisfying because I am a professional reviewer, dammit, and professional reviewers are supposed to dissect in exhaustive detail the sound of the component they're reviewing. To tell the truth, I had a very hard time identifying any sonic signature with the Elgar in isolation. It was obviously time to compare it with my long-term reference, the mighty Mark Levinson No.30.5.

Compared with the Levinson, using the No.38S preamplifier to ensure level matching within 0.1dB, the Elgar sounded less laid-back in the upper midrange, with a slightly more vivid overall balance. Its treble character was as grain-free as that of the No.30.5, though the latter's low frequencies sounded more weighty and voices were a hair more fleshed out. But I arrived at these conclusions only after lengthy, sometimes frustrating listening sessions.

And I never did decide which was the best component when it came to imaging. On some program material, I thought the No.30.5 threw a more spacious soundstage; other times the dCS processor had a slight advantage. One thing's for certain: the Elgar is a contender. If the spectrum of Class A digital processors extends from the Sonic Frontiers SFD-2 Mk.II on the sweet but rather soft- and flabby-sounding end, to the Spectral SDR-2000 Pro on the superbly detailed but perhaps a little lightweight end, with the Levinson plumb in the middle, the Elgar differs from the '30.5 in being more like the Spectral than the Sonic Frontiers. Good company, eh?

For the next sessions I dispensed with the preamp, which gave me the simplest possible signal path: digital from transport to DAC, balanced analog from DAC to amplifier, and the Elgar's volume control to set playback level. This volume control operates in the digital domain and appears to be properly dithered in that it is free from "zipper noise" as it is operated. With the analog outputs switched to High, the control setting varied between -6dB and -12dB, depending on the recording.

The result was an astonishingly detailed yet delicate presentation. One of the areas where 16-bit digital has traditionally underperformed is when the mix gets really complicated. With poor players/processors, all the soundsources blend together at climaxes, the sound resembling a giant mouth organ. Well, not only was this not what the Elgar did without a preamp in the chain, but it allowed individual sonic objects to be differentiated better than anything I've heard. It was equaled in this respect only by the combination of the No.30.5 and the Meridian 518, which is 50% more expensive.

Somewhat behind the curve, I just picked up a copy of Pink Floyd's 1988 Delicate Sound of Thunder album (Columbia C2K 44484), recorded live on the first of rock's big "pension" tours. In "Comfortably Numb," even as David Gilmour wails and weedley-wops higher and busier on his Stratocaster, and fellow Cantabrigian guitarist Tim Renwick keeps the big fuzztone riff going in the background, via the naked dCS I could still focus on the bass guitar, the kick drum, the keyboards. Just as I probably could have live. No, forget that—even Pink Floyd have probably succumbed to the live-sound-engineer-from-hell syndrome that has afflicted just about every rock concert I've gone to in the past 15 years. (The Dead were the honorable exception, of course.) The Elgar gives you better than live sound—dCS's marketing department is just going to love that quote.