Digital Processor Reviews

From the front panel you can set: the A/D sync source (44.1/88.2, 48/96, etc.), the output resolution (UV22 16, UV22 20, 24-bit), the optical output signal (S/PDIF or ADAT), a soft-limiting circuit that gives you maximum digital signal level without "overs" (hard overloads), independent left and right mutes, ABS, meter function (A/D or D/A), D/A input, and MDM, the last used for selecting track pairs when recording to ADAT or Tascam MDMs. There are also two pairs of screwdriver-adjustable level pots for D/A and A/D level. An optional video sync card is available for DVD mastering.

Connecting the PSX-100 is straightforward: I ran the single-ended tape outs of the Ayre K-1 preamp to the A/D inputs using RCA/XLR adapters, and the balanced D/A outputs to one of the K-1's balanced source inputs. A digital coax cable went from the PSX-100's S/PDIF out to the Marantz DR17's digital in, and for good measure I ran the Marantz's digital out to an input on the EAD DSP-9000. Toward the end of the review period the Audio Research Reference 2 line stage arrived; it has balanced tape outs, and its remote-controlled source selection permitted A/B comparisons from my listening position.

Methodology and Madness
Pacific Microsonics doesn't yet offer an affordable, processor like the PSX-100, but the comparison in the Apogee instruction booklet of UV22 and HDCD got me thinking about including some sort of sonic comparison in this review. With Apogee converters and UV22 processors and Pacific Microsonics' HDCD encoder/converters in studio use around the world, I had a large selection of UV22-mastered and HDCD-encoded CDs to compare. I also had something better: a unique CD-R that Classic Records' Mike Hobson had given me back in 1996. It contains analog master-tape transfers done by Bernie Grundman comparing three converters: an early HDCD encoder, a K2 (used by JVC for its 20-bit transfers), and an older Apogee 20-bit converter fed to a standalone UV22 processor.

Each selection is recorded three times, once through each converter. The music includes an unidentified vocalist (Quincy Jones, producer), "Sweet and Pungent" from Duke Ellington's Blues in Orbit, and two excerpts from Witches' Brew, the RCA Living Stereo classic.

All three processors did a reasonably good job, though HDCD (decoded by the EAD DSP-9000) sounded noticeably hashy and brittle compared to the other two, with indistinct image focus and a slight gauzy overlay. The K2 and Apogee converters sounded more analog-like, with greater liquidity and image three-dimensionality. If I had to pick one of the two as marginally better, it would be the K2, but the differences were trivial to my ears. (I'm told that Pacific Microsonics' latest processor, the HDCD Model 2, is a considerable sonic improvement.)

Archiving precious lacquers of Who's Next and Buddy Holly were first on my list of priorities. These gems still sound incredible even after many playings, but between wear, dirt, and lacquer "springback," it would be reassuring to have clean copies, which would also give pretty good indications of the Apogee's transparency.

While transferring Who's Next I ran the PSX-100 in Analog Monitor mode, comparing the original analog signal with the 24/96 A/D–D/A conversion. Of course, any time you divert a signal and pass it through two cables and a complex piece of electronics, there will be differences. But the differences I heard were not what I'd call fundamental. One signal did not sound "analog," the other "digital." It was more like a minor loss of liquidity and/or transparency. I wouldn't want to have to tell you which was which in a blind listening test.

Next, I listened to the UV22-processed CD-R of Who's Next—not in comparison to the lacquer, but just as a listening experience. I'd purposely recorded the sound of the stylus hitting the lead-in groove—it was very strange watching the dormant turntable as it played back. The CD-R wasn't as rich and open as I remembered the lacquer sounding, and lacked the lacquer's transparency and rich, velvety textures. But the only way to be sure was to do an A/B comparison of the lacquer and the CD-R through the Apogee's D/A. I did so, taking the opportunity to make another CD-R, this time using the A/D converter built into the $1600 Marantz DR17 to hear how it would fare against the PSX-100.

The lacquer was as I remembered it; the PSX-100–generated CD-R simply wasn't an "archival" copy. But it was close, and certainly closer than the Marantz edition, which was softer, smoother, and less detailed, but still impressively true to the lacquer's "spirit." The biggest difference was the PSX-100 edition's better rendering of ambience and low-level detail than the Marantz's, its more bracing transient performance and superior bass extension. If I want to hear the lacquer, I have to listen to the lacquer—neither CD-R will do. Still, I'm glad to have stopped the ravages of time and wear; one of these days, after all, the CD-R will be preferable.

I performed the same comparison with Classic Records' absolutely monumental mono reissue of Jimi Hendrix's Axis: Bold As Love. What a brilliantly finessed mix! Big, solid, three-dimensional images are laid out crisply and neatly from front to back, and possess harmonic and transient details that make listening to this familiar record a brand-new experience. I was transported back to the apartment where I first heard this album as an explosion of long-buried sensory memories accompanied the music.

Played back through the Apogee's D/A, the Marantz DR17's D/A, the EAD DSP-9000's processor, or through a Musical Fidelity X-Ray CD player, the CD-R, while sounding impressive, was not a clone of the original. None of the digital renderings could compete with the vinyl's pristine transparency and rich harmonic bouquet. A direct A/B of LP and CD-R reinforced the sensation that the harmonic envelopes surrounding instruments and voices extended just so far, then stopped short. The LP's warmth and richness were not "colorations." The coolness of the CD-R was.

I made a CD-R of the Doug Sax–mastered LP of Janis Ian's Breaking Silence and heard the same differences: the LP sounded richer, warmer, more velvety-smooth than either the 16-bit, UV22-processed CD or a Marantz-recorded copy. The latter, though warmer, sacrificed low-level detail, air, and, seemingly, high-frequency extension.

I then compared the CD-R of Breaking Silence with Acoustic Sounds' gold CD version, also mastered by Sax, probably with UV22 processing. Tonally, these two sounded closer to each other than either sounded to the LP—regardless of which processor or CD player I used for playback. Does this mean the LP sounds closer to the master tape, and that both digitized versions bear the sonic stamp of UV22 processing?

I don't know. How's that for an answer?

D/A conversion
Though the purpose of reviewing the PSX-100 was to assess its archival abilities, it makes a pretty good standalone D/A converter as well, with deep, well-focused bass, snappy pacing, detailed transient performance, and a tonal balance that is, overall, neutral if somewhat clinical. The processor will not decode a 24/96 datastream fed to its coaxial S/PDIF input, however; if you've been buying 24/96 discs from Classic, Chesky, and Acoustic Sounds, the Apogee will not be a good choice (footnote 1).

Conclusion
For $3900, the Apogee PSX-100 is a "full-service" digital product. It can both convert analog to digital and convert a digital datastream back to analog with 24-bit/96kHz resolution. Using Apogee's proprietary UV22 process, it attempts to maximize the amount of information that can be stored on the 16-bit CD format.

With Apogee's Bit-Splitting circuitry, the PSX-100's full 24/96 performance can be stored on 16-bit ADAT and Tascam DA88 multitrack recorders until a more robust format, such as recordable DVD, is available as a consumer product. (ADAT and DA88 recorders must be meticulously maintained, however, and tape quality is variable and a big determinant of the final sound quality.)

While, for all intents and purposes, the 24/96 conversions sounded "transparent" to the source, the UV22-processed CDs created from them didn't. Still, the overall sound was surprisingly close, the differences being mostly timbral and textural. Given the "get 'em out the door for 35 cents" state of the CD-manufacturing art, which is quickly approaching a quality level analogous to that of the Dynaflex era of the LP's reign, it may very well be that you can now make better-sounding CDs at home.

But if you're looking for a reasonably priced, ultra-high-quality A/D converter with which to archive your analog source material to 24/96 using inexpensive ADAT or Tascam MDM recorders or as 24-bit WAV files on your PC's hard drive, or you want to get the most performance from 16/44.1 CD or 16/48kHz DAT, the Apogee PSX-100 should be your first stop.

Can a piece of studio gear find happiness on an audiophile's equipment rack? The flexible, surprisingly easy-to-use Apogee PSX-100 says "Yes."

---

Footnote 1: The review sample would only accept or output 88.2 or 96kHz-sampled data via its two AES/EBU data inputs/outputs in parallel. (The left-channel data are transmitted as one pair of 44.1 or 48k tracks, the right channel as the second pair.) However, the PSX-100 datasheet on www.apogeedigital.com indicates that single double-speed AES/EBU operation is now available, which will allow the unit to be used with a PC fitted with double-speed digital I/O, such as the RME Digi96/8 Pro or the Digital Audio Labs CardDeluxe.—John Atkinson