Columns Retired Columns & Blogs |
Perpetual Technologies P-1A D/D & P-3A D/A processors Page 3
The sound was dynamic, with good timbral and transient accuracy, and the sort of natural high-frequency detail that I have come to associate with upsampling D/A processors. I actually preferred the sound of the P-3A to that of the Bel Canto DAC 1 ($1295), which, while smooth and unfatiguing to listen to, sounded a bit bland and soft in comparison. The fully optioned MSB Link III is the recipient of Stereophile's Budget Component of 2000 award, and may well be the best D/A processor in its price range. The P-3A came very close to the standard established by the Link III, matching it in dynamics but not sounding quite as refined in the upper midrange/treble. Considering the difference in price—$1050 for the P-3A with P3 power supply vs $1385 for the Link III—this is very fine performance.
In assessing the performance of the P-1A and P-3A together, I first wanted to determine what the P-1A sounded like without Resolution Enhancement, functioning just as a front-end to the P-3A. I compared the sound of the P-3A on its own (direct AES/EBU connection between the PS Lambda II transport and the P-3A) vs the P-1A/P-3A combination (AES/EBU between transport and P-1A, I2S between P-1A and P-3A, P-3A set to I2S Direct, P-1A set to output 96kHz, Resolution Enhancement set to Bypass). Given that the direct transport-processor connection is more, well, direct, one might expect the sound to be better than with the signal routed through another component.
That's not what I heard. The P-1A/P-3A combination was more detailed, had greater clarity, and the upper midrange and treble were smoother, more grain-free than the P-3A by itself.
I was initially puzzled by the superiority of the sound of the P-1A/P-3A, but, reading through the FAQs on the Perpetual Technologies website, I found the likely reason: enhanced jitter reduction. The P-1A is said to reclock data "to an amazing 2 picoseconds," and the sample-rate conversion uses a very-low-jitter, fixed-frequency crystal oscillator. Although the input-receiver sections of the P-1A and the P-3A are similar, Peter Madnick suggests that there are advantages to having this function performed by a separately shielded and powered circuit. Whether for this or some other reason, the P-1A seemed to allow the P-3A to do a better job of D/A conversion, even with Resolution Enhancement bypassed. The sound of the P-1A/P-3A in this configuration handily surpassed that of the fully optioned MSB Link III.
Enabling Resolution Enhancement (Output Bit Density set to 24) took performance to a higher level still, an effect akin to replacing CDs of run-of-the-mill technical quality with remastered "audiophile" versions. The sound become altogether more natural, more detailed, but with no added artificial edge. The improvements were most obvious with CDs that I'm highly familiar with, like the original-cast recording of 42nd Street (RCD1 3891)—the very first CD I ever bought. Voices of individual chorus members, and individual instruments in the orchestra, were all more distinct, and it was easier to focus on details of the complex sonic texture.
On track 3 of Chesky Records' first Jazz Sampler & Audiophile Test CD (Chesky JD37)—a track that I have listened to, in whole or in part, approximately 4367 times—the sound of all the little bells and exotic percussion instruments was less homogenized; again, it seemed as if Chesky had gone back to the master and reissued a CD incorporating the latest technology. In general, the timbral qualities of instruments seemed more true to life, with less of the synthetic "digital" character that many audiophiles dislike about CD sound. There was also an enhancement of recordings' spatial characteristics, with ambience information presented more distinctly, and images within the soundstage being more three-dimensional.
Resolution Enhancement involves a mathematical algorithm that takes digital audio data recorded with 16-bit precision and makes guesses about what the data would look like had it been recorded with 24-bit precision. It follows that, no matter how well-educated the guessing, some of the interpolated data points are not going to be accurate representations of the real 24-bit data points (if we had these available for comparison). For the audiophile not overly concerned with digital audio theory—a description I would apply to myself—the question is whether Resolution Enhancement produces results that create a more convincing illusion of live music. The answer, for me, is most definitely "Yes," with a minor, qualifying "but..." For the vast majority of recordings, engaging Resolution Enhancement produced positive effects of the kind I've noted, and improvement was sufficiently marked that I had little motivation to switch back to Bypass mode.
However, with a few recordings—maybe 10% of the ones of which I did A/B comparisons—I preferred the sound in Bypass mode: It seemed more direct, while the Resolution Enhanced version sounded a bit veiled or processed. One of these discs was Chris Norman's The Beauty of the North (Dorian DOR-90190), which features flute, fiddle, accordion, guitar, and bass in simple arrangements of traditional favorites from Quebec and Maritime Canada. I chose it as one of my "Records To Die For" in February 1995, and I often listen to it for pleasure as well as in equipment comparisons. It's a very clean, clear recording, and some of that clarity seemed slightly obscured in the Resolution Enhanced mode. With Bypass available at the push of a button, it was easy to check the contribution of Resolution Enhancement; terminally compulsive audiophiles may want to note whether each CD (and each track?) sounds better with or without Resolution Enhancement.
Conclusion
The demise of Mark Schifter's Audio Alchemy was mourned by audiophiles who have since missed the company's affordable but excellent-sounding, and often innovative, digital products. Now, with Perpetual Technologies, Schifter is back: the P-1A/P-3A consttute his auspicious re-entry into the digital audio marketplace. As a standalone D/A processor, the P-3A (with Monolithic Sound's P3 power supply) is competitive with some of the best processors in its price range. Combined with the P-1A, it reaches a level of digital replay that must have manufacturers of multi-kilobuck digital processors hoping that not too many customers find out about it. Upsampling to 96kHz and Resolution Enhancement to 24 bits work as claimed, maximizing the sonic potential of most CDs.
For the P-1A, this review is more of a progress report, the enhancement of CD playback being only a part of the product's capability. The P-1A was designed with loudspeaker- and room-correction in mind; the software is currently being fine-tuned to optimize these functions, and I'll be doing a Follow-Up when it becomes available. I can hardly wait.
- Log in or register to post comments