Enlightened Audio Design DSP-1000 Series III D/A processor
On the rear panel, audio output is provided by a pair of single-ended RCA jacks. No balanced audio outputs are present, such as are found on the manufacturer's top-of-the-line DSP-9000 decoder. Digital input is via a 75-ohm coaxial ("CX") gold-plated RCA-style jack. Also available are TosLink and ST glass-optical inputs. I used the coaxial input for this review because my Krell MD-1 digital turntable has only a coaxial output. A detachable AC inlet/line cord completes the back panel.
Inside, all components reside on the main circuit board, which features solder mask and some point-to-point wiring. Two EI-core power-supply transformers are used, a small one for the digital circuitry and a larger one for the analog circuitry. Three regulators, one for the 5V rails and one each for the ±15V rails, are mounted on heatsinks. The CS8412 data receiver chip and the two PCM63P Burr Brown D/A chips are mounted to the right side of this main board.
The HDCD assembly is mounted on a small daughterboard that contains all the Series III components, including the HDCD chip and the "Digital Flywheel" jitter-rejection components. The use of a daughterboard enables dealers to easily upgrade older EAD decoders to the Series III level. The daughterboard is attached by a 28-pin connector and supported by generous gobs of "hot melt," a meltable plastic used to stabilize the daughterboard over the motherboard. This technique is not conventional, but EAD claims that their units so equipped function very reliably after shipment, including those sent overseas.
Also on this daughterboard, and adjacent to the HDCD chip, is a block with 12 rocker switches. These switches allow EAD to upgrade any of their previous decoders to Series III level. No information about the switch settings is provided in the manual, because the switches do such critical things as turn on/off the Digital Flywheel. An EAD spokesperson told me that fooling with the switches wouldn't harm the decoder, but it might not produce music if set incorrectly. In fact, switch 12's default position provides the required 6dB attenuation of non-HDCD signals when playing standard CDs (or 6dB of gain with HDCD-encoded sources). This attenuation subjectively matches levels between HDCD and standard CDs. This is because HDCD-encoded discs have a higher crest factor than conventional CDs and thus sound quieter in direct comparisons.
EAD accomplishes the required reduction in gain for regular CDs in the digital domain. As this reduces the resolution at a rate of 1 bit for every 6dB of attenuation, it should be optional, in my opinion. I agree with RH's suggestion that Pacific Microsonics should reconsider this requirement. Even with the option to defeat non-HDCD disc attenuation, it was necessary to leave the DSP-1000 in the attenuation mode for gain matching with other decoders in this review.
Components of EAD's patented "Digital Flywheel" circuitry can be seen on the Series Three daughterboard because they are not potted, as in the Adcom GDA-700 that I also review in this issue. EAD claims that this circuit "provides the jitter suppression necessary to reproduce 20 bits of musical resolution upon playback" (see "Manufacturers' Comments," Vol.18 No.1, p.239). This operates a second phase-locked loop (PLL) in series with the PLL present in the Crystal Semiconductor CS8411/12 digital receiver chip. RH, in a recent review (Vol.18 No.8, p.147) of the company's flagship DSP-9000, describes the digital flywheel as follows: "The clock recovered by the PLL in the Crystal CS8411 is followed by a second, tighter-bandwidth PLL implemented with a Voltage-Controlled Crystal Oscillator (VCXO)." The two PLL circuits are matched, and EAD claims that the jitter from the Crystal Semiconductor digital receiver is lowered by a factor of ten. The HDCD chip, the CS8412 chip, the Burr-Brown chips, and others are socketed, which is further evidence that the DSP-1000 can be field-upgraded in the future.
The manual suggests that the output voltage of the DSP-1000 can be boosted from 2V to 4V to drive a passive preamplifier by clipping out pairs of resistors and capacitors on the system's motherboard. These gain changes are made in the analog domain, and do not affect the unit's resolution the way digital-domain attenuation methods do in other decoders. Although cutting the resistors does not void the warranty, I feel that component clipping should not be an end-user adjustment. Not only is this approach nonreversible, but this could open the door for all sorts of home-based modifications using special audiophile capacitors. EAD claims no customers have clipped out the wrong componentat least not yet.
The EAD DSP-1000 shows evidence of last-minute additions, with the point-to-point wiring and the use of the hot-melt glue to stabilize the daughterboard. These features notwithstanding, the rest of the DSP-1000 shows good build quality, with a solid chassis that does not ring. The unit's socketed components mean that it can be upgraded in the future. This gives good value for money.
This HDCD-equipped decoders proved to be a revelation. I had used a Krell SBP-32x digital audio converter as my reference during many previous reviews. The new HDCD process provides great improvements, most notably in dynamic range. The large dynamic range inherent in analog vinyl records had been a key factor in my musical involvement with the medium, and this HDCD-equipped decoder had the dynamics I'd been missing listening to conventional CDs. I was so pleased that it seemed unfair to compare these three designs. However, after careful gain matching and long listening sessions, differences became apparent which may provide a basis for choosing one of these units for a particular system.
On first listen, the Enlightened Audio Design's DSP-1000 had a warm presentation with a strong midbass. The richness of the midbass gave male choral works a warm, rich quality. Dynamics were strong, perhaps as a result of the firm bass, and the EAD's midrange was not aggressive or harsh. These qualities were equally present in the Audio Alchemy DDE v3.0. The depth and size of the acoustic, the recovery of ambience and resolution of low-level detail, were equivalent in these two decoders.
Bass reproduction on the EAD DSP-1000 was quite good. On Pomp & Pipes (Reference RR-58CD), each deep pipe-organ note was clearly defined, showing the decoder's ability to distinguish pitch. The acoustic space around the bass drum was clearly heard during the final moment of the HDCD band arrangement of the Chorus Line overture (from Beachcomber with Frederick Fennell and the Dallas Wind Symphony, Reference RR-62CD). The Adcom GDA-700 and the Audio Alchemy system, on the other hand, emphasized the bass drum's solidity, slam, and dynamics.
Midrange reproduction on the EAD 1000 was smooth, with no harshness or aggression. The woodwinds in the Chorus Line selection were open, effortless, and showed no strain. The attack of the piano notes on Mike Garson's rendition of "A Song for You" (from The Oxnard Sessions, Vol.2, Reference RR-53CD) was natural, quick, and convincing with the EAD unit, although this was heard equally well with the other two HDCD-equipped decoders.
Non-HDCD CDs also did well on the EAD decoder. For example, it created a convincing sonic portrait of the warmth and three-dimensionality of Odetta's close-miked "America the Beautiful" on Strike a Deep Chord: Blues Guitars for the Homeless (Justice Records; call (800) 538-5878). It also captured the richness and naturalness of male vocal timbre on Willie Nelson's "Getting Over You" and "What Was It You Wanted" (Across the Borderline, Columbia CK 52752).
The EAD DSP-1000 also did a very good job of capturing nonspecific hall ambience. I focused on this feature of orchestral music after attending a live concert by the New York Philharmonic. Sitting in the orchestra, I closed my eyes and was immersed in a "sea of sound," a three-dimensional mix of direct and reflected sounds that not only defined the musicians on the stage, but the hall around me. This is different from the carefully layered resolution of detail heard on finer audio systems. Rather, it is a living sonic fabric composed of instrumental blending, low-level details, and audience coughs that help the ear and brain "feel out" the acoustic. For example, the EAD could re-create the sense of the hall and the orchestral fullness heard during the brass finale of Janácek's Sinfonietta (José Serebrier, Czech State Philharmonic, Reference RR-65CD).
In summary, I was pleased by the EAD's smoothness, lack of strain, and rendering of nonspecific hall ambience. For example, it gave a sweet, effortless quality to the trumpets during the finale of the Janácek. There was no glare or harshness during peaks, nor was there a sense of congestion or hardness during orchestral climaxes. While this was a definite advantage, the Audio Alchemy and Adcom units were more transparent and more dynamic playing this selection, as well as other types of music. I found myself less involved while listening to the EAD DSP-1000.
The three HDCD-equipped decoders I have auditioned have wide dynamic range, which lends pace and impact to digital music. Now it's possible to have large dynamic contrasts without the glare and harshness associated too often with the CD medium. These improvements are available in all three of these under-$2000 decodersgood news for audiophiles. The EAD DSP-1000 offers HDCD decoding and excellent sonics, and should join Audio Alchemy DDE v3.0 and the DTI v2.0 on Stereophile's "Recommended Components" list.
Differences were apparent, however. The EAD DSP-1000 was smooth, liquid, and natural and could never be pushed to be aggressive or hard, even during huge orchestral climaxes. Its ability to re-create nonspecific hall ambience gave it a realism sometimes missing with the other units. The EAD's defeatable attenuation on non-HDCD CDs is an asset, considering that the majority of music is available on the non-HDCD format. On the other hand, the point-to-point wiring and use of hot melt for stabilizing critical digital circuits in the DSP-1000 was a concern.
Footnote 1: "HDCD Overview" (Vol.18 No.3, p.127); an interview with HDCD's inventors ("Reinventing the Digital Future," Vol.18 No.5, p.49); HDCD's implications for the audio enthusiast ("Perfect Sound Forever," Vol.18 No.5, p.3); HDCD's impact on the sonic quality of high-end digital decoders (in a review of the Spectral HDCD D/A processor, Vol.18 No.5, p.85); and news of recent recordings ("Industry Update," Vol.18 No.11, p.35).