The biggest question facing today's discriminating music lover is not which digital front-end to buy, but whether to buy one at all. With all the talk about a new DVD-based high-quality audio format on the horizon, many listeners are reluctant to invest in expensive compact-disc playback equipment.
I strongly believe it's a mistake to put off buying good digital playback now for some possible future format, and here's why:
First, the new audio standard has not been agreed upon by the world's electronics and music companies; indeed, we may have a format war between two competing camps. Any corporate battle for supremacy will hinder progress, perhaps even causing enough confusion in the mass market to relegate the new format(s) to the ranks of "specialty" products.
Second, high sampling rate and long word length (such as 96 kHz sampling and 24-bit quantization) don't guarantee superb fidelity. Yes, 96/24 (or Sony's DSD encoding technique) has the potential of much better sound quality, but how the technology is implemented is crucial to realizing that potential. You should also know that 24-bit analog-to-digital and digital-to-analog converters simply don't exist. The state-of-the-art in conversion, using expensive hand-calibrated units, is 20 bits at best.
Third, no matter what happens in the standardization negotiations, it will take some time for a library of titles to become available outside the limited capabilities of audiophile labels.
Finally, you probably have a large CD collection that you want to play back with the best sound possible, now and in the future. If a new format comes along, you'll simply have a second player in your equipment rack.
These issues were on my mind as I contemplated the Spectral SDR-3000 Studio Reference Transport. The $7995 SDR-3000 is the final element of Spectral's digital playback system, complementing the $8995 SDR-2000 digital processor. I've previously reviewed the SDR-2000, and in comparisons with other top-end contenders, concluded that in many ways it is the best-sounding converter extant. The SDR-3000 transport, with its special Spectralink interface to the SDR-2000 (see Sidebar), has the potential of elevating the 2000's performance to a new level.
The SDR-3000 boasts Spectral's classic styling - a grayish aluminum frame surrounding a black front panel and smooth, quiet front-loading transport mechanism. A display indicates when the disc is loading, when it is stopped, and when it is in the play mode. The most commonly used buttons (load, stop, play, pause) are large, easily accessible, and have a positive feel. A second row of smaller buttons covers the less frequently used controls (track skip, scan, display brightness). The display (which can be dimmed or turned off) is easy to read, and the machined remote control is both functional and solidly built.
Ergonomically, the SDR-3000 is the best transport I've ever used. The large buttons, positive action, easy-to-read display, and illuminated legends contribute to the player's ease of use, particularly in the dark.
The rear panel provides one coaxial output on an RCA jack, one balanced digital output on an XLR jack, and the Spectralink jack. The Spectralink interface is a "D" connector similar to those found on computers. Used only with the SDR-2000, the Spectralink interface significantly reduces jitter in the SDR-2000.
After solving the interface-jitter problem with Spectralink, Spectral attempted to eliminate jitter in other parts of the SDR-3000. The transport mechanism itself is a massive, top-of-the-line, al-metal unit sourced from Esoteric (a division of TEAC) and redesigned specifically for the SDR-3000. This mechanism bears little similarity to the mostly plastic Philips mechanisms used in many high-end CD players and transports. For starters, the disc is clamped over its entire surface area rather than just at the spindle. Tightly clamping the disc reduces vibration while the disc is read. The disc clamp, made from machined brass and aluminum, sits in a cast zinc-alloy bridge that straddles the transport mechanism. Stainless steel slides guide the disc drawer in and out. Frankly this mechanism looks like a Mercedes next to the Yugo of most CD transport mechanisms. For further vibration isolation, the transport assembly floats on a tuned suspension with resonance damping. Sub-enclosures within the chassis isolate the electronics from the mechanical subsystems. Four separate power supplies also help to isolate the various sections of the SDR-3000.
Overall, the SDR-3000's build is exemplary. The massive transport mechanism, solid chassis, and subsystem isolation are first-rate. In addition, the SDR-3000 has a refined look and feel that suggest both precision and elegance. While expensive at $7995, the SDR-3000's price reflects the unit's development and build cost, not cosmetics or status pricing.
Although I've already reviewed the SDR-2000 (in the May, 1995 Stereophile), I'll revisit it here briefly before getting to the SDR-3000, as most SDR-3000 purchasers will use the transport with the SDR-2000 processor. Indeed, I think of the two as a single system that happens to be housed in separate chassis.
The SDR-2000 achieved state-of-the-art performance in digital playback in many important areas. Its most remarkable characteristic was an astonishing purity of timbre through the midrange and treble. The presentation had a pristine clarity that made even highly-regarded processors sound a tad grainy and overlaid with a shiny glare. The slight layer of coarseness that diminishes timbral realism (particularly with strings) so closely associated with digital was largely absent in the SDR-2000. In addition, the SDR-2000 had a precise spatial presentation in which instruments and voices were clearly defined within the recorded acoustic.
The Spectral processor's transient reproduction was second to none. This transient speed and zip imparted a more lifelike quality to the music as notes started and stopped with no blurring or overhang. Low-level detail resolution was unsurpassed, providing more musical information, but not in an analytical or etched way. If I had to describe the SDR-2000 in three words, they would be "clean, quick, and detailed".
You should know that the SDR-2000 isn't the last word in bass slam and bottom-end weight. The presentation tended toward the lean and articulate, rather than the big and powerful. Kick drum, for example, had less bass impact and "center of the earth" solidity through the SDR-2000 than through Wadia products. Although I would have preferred more weight, the SDR-2000's other strengths made up for this deficiency, at least for my taste. All things considered, the Spectral SDR-2000 is, in my view, the state-of-the-art in digital replay.
After spending hundreds of yours listening to the SDR-2000 driven by the Mark Levinson No. 31 and No. 31.5 transports, I was eager to complete the Spectral system with the addition of the SDR-3000 transport and its Spectralink connection.
Putting the SDR-3000 in the system produced a noticeable increase in resolution, soundstage size, spatial definition, and timbral liquidity. All the attributes that made the SDR-2000 processor so musically compelling were taken another level by the SDR-3000.
Perhaps the most striking aspect of the SDR-3000's sound was its ability to resolve quiet instruments in the presence of louder ones. So often in digitally reproduced orchestral music, the dominant section at any given moment obscures the contribution of other sections - brass overpowering woodwinds, for example. Through the SDR-3000, however, I could clearly hear delicate instruments beneath the power of the rest of the orchestra. The result - hearing more of the composer's orchestration - produced a much richer sensory experience than I've had with other digital front-ends. Not only did the SDR-3000 resolve more music, the musical information was presented in a way that made it accessible without the auditor trying to listen deeper. The resolution, the nuances, the music's fine texture were right there. I strongly believe that the less work the brain must do to create the illusion of hearing real instruments, the more involving the presentation. The SDR-3000 laid it all out before me in a natural and unforced way, making it easy to forget the playback system.
I also experienced this phenomenon with piano; the SDR-3000 better resolved independent left- and right-hand lines,
making the instrument more coherent and articulate. Consequently, counterpoint was more pronounced and compelling. Listen to pianist Cyrus Chestnut's CD Earth Stories; his left hand is as interesting as his right. The Spectral transport let me hear so much more of Chestnut's magnificent artistry.
For me, one of the hallmarks of high-performance music reproduction is the ability to resolve lots of detail without sounding etched, aggressive, hyped, or analytical. Live music is infused with a fine inner structure that is either diluted by "smooth" playback systems, or unnaturally emphasized by "revealing" systems. Both are departures from neutrality. The Spectral digital front-end is, in my experience, unique in its ability to extract a huge amount of low-level detail from a recording, without unnatural etching. The SDR pair didn't force detail on me, but presented it a subtle way that encouraged me to explore the soundstage, discovering low-level information for myself. This sonic attribute produced a more profound involvement with the music, a deeper level of sensory immersion and intellectual satisfaction.
The Spectral transport snapped the soundstage into sharp focus, with tight images, precise spatial cues, and a large overall presentation. I also heard greater transparency - heard more deeply into the soundstage's inner recesses - than with other reference-quality transports. The combination of soundstage transparency and low-level resolution combined to produce a presentation of stunning clarity. I could hear reverberation and the sound of the hall hanging in space for a long time before they decayed smoothly into inky blackness.
A good analogy is flying into a city in the Southwest on a clear night. The dry, thin air gives individual lights on the ground a pinpoint vividness. The thicker moister air in the East slightly defocuses the lights into a continuum.
The SDR-3000/ SDR-2000 pair presented space somewhat differently from other digital front-ends. Image focus was sharper through the Spectral than through other products. In addition, the bloom around images - that sense of air surrounding the instrument - was more compact. Some listeners prefer the larger bloom and more expansive sound of other digital front-ends (particularly the best of the tubed units), even at the expense of reduced focus and spatial precision.
The undeniably appealing sensory experience of hearing a soundstage infused with lots of bloom was ultimately offset by the Spectral's ability to resolve more music on the recording, with its lighter, drier, and more controlled sound. Significantly, the Spectral front-end also changed its spatial perspective with recordings to a greater degree than digital source components that have more bloom.
I must also reiterate that the SDR-3000/ SDR-2000 pair has less bass "slam" and bottom-octave weight than other first-rate digital front-ends. The SDR-3000, however, ameliorated this tendency, giving the SDR-2000 somewhat greater impact and power. Still, the Spectral front-end didn't have as much of the visceral, physical drive in the bass that involves your whole body in the music, not just your mind.
The performance I've described was with the SDR-3000 connected to the SDR-2000 via the Spectralink interface. How does the Spectral transport stand up when used with conventional AES/EBU interface? To find out, I compared the Spectralink interface with a first-rate AES/EBU cable, the MIT Proline
Although the SDR-3000 is a superb transport when connected via the AES/EBU interface, the real magic was rendered via the Spectralink connection. Spectralink produced a smoother rendering of timbre; violins were more liquid and lacking a trace of edge heard through the AES/EBU connection. Piano had just a hint of glassiness at the leading edges of notes through the conventional interface, a distortion removed by Spectralink. In addition, the soundstage expanded and bloomed more, with greater air around instrumental outlines and a larger overall presentation. The background became blacker, with a fine detail thrown into even sharper relief. These obvious performance advantages of Spectralink make buying the SDR-3000 a must for anyone who owns a SDR-2000.
I also evaluated the SDR-3000 driving Sonic Frontiers' ambitious new Processor 3 DAC via the AES/EBU interface. Again, the Spectral transport demonstrated its superiority. The 3000 threw the largest and best defined soundstage of any transport I've heard. Surprisingly, the SDR-3000 also had terrific bottom-end extension and punch - qualities the SDR-2000 isn't known for. The sense of pace and rhythmic drive were also first-rate, characteristics particularly apparent on the track "Wishing Well" from Michael Ruff's Speaking in Melodies on Sheffield.
The Spectral SDR-3000 transport is simply the best-sounding CD transport I've auditioned. When driving non-Spectral digital processors, the SDR-3000's musical performance was not just marginally better than other world-class transports, but significantly better. Even without the Spectralink interface, I still would have considered the SDR-3000 the state-of-the-art in CD transports.
But when connected to the SDR-2000 processor through the Spectralink interface, the Spectral combination achieved a level of musical performance I didn't think possible from the compact disc format. The SDR-3000's unique combination of analog-like ease, resolution, transparency, and lack of digital glare must be heard to be believed. I'm continually amazed at how improvements in digital playback resolve more and more information on CDs.
If you want the best reproduction possible from CDs - now and for years to come - look no further than the Spectral SDR-2000 digital processor and SDR-3000 transport.
Two-Way Street: The Spectralink Interface
It's an unfortunate fact of life that technical standards designed for mass-market audio are foisted upon the high-performance audio industry. The most recent example is the Sony/Philips Digital Interface Format (S/PDIF), a method of transmitting digital audio from a CD transport to an outboard digital-to-analog converter. When these standards were created, the last thing on the engineers' minds was sound quality. At the time, the conventional wisdom held that if you can transmit the ones and zeroes representing the music without error, then the interface must be flawless. In fact, I was at a technical meeting of the Audio Engineering Society in London when one of the interface format's designers said as much
But when separate CD transports and digital processors first became available, audiophiles noticed that the electrical coaxial connection (RCA jack) sounded different from the optical TosLink cable, even though they were both carrying the identical data in the identical format. (to my knowledge, J. Gordon Holt was the first writer to document this phenomenon.) How could this be?
Subsequent analysis revealed that it isn't enough to just transmit the ones and zeroes; the timing of those ones and zeroes was crucial to sound quality. Because different transports and cables have different effects on the timing of the digital signals, they sound different. Timing errors, also known as jitter, are a significant source of degradation in digitally reproduced music.
Jitter is introduced in the interface between a transport and processor because the same cable carries both audio data and the clock reference. In essence, the audio data are buried within the clock. The first circuit inside a digital processor (called the input receiver) "recovers" this clock, which then becomes the timing reference for the crucial digital-to-analog conversion process. Any jitter added by the digital cable winds up at the DAC clock, and thus affects sound quality.
Jitter reduction devices that go between a transport and digital processor can clean up jitter, but they must put the signal back into the S/PDIF format for output to the digital processor. The jitter-free output from the jitter-reduction box is then jittered by traveling down the cable to the digital processor.
Spectral's solution to this problem is the Spectralink interface, used in the SDR-3000 transport and SDR-2000 processor. Rather than using a jittered clock to time the DAC's, the SDR-2000 generated its own precision clock right next to the DACs. The SDR-3000 transport is then slaved to that precision clock, instead of the other way around.
Although this idea isn't new (Linn and Krell, among others, have used separate clock lines), Spectral's implementation is unique. Instead of sending a high-frequency clock signal back to the transport, the Spectralink interface carries an analog signal from the SDR-2000 processor to the SDR-3000 transport, telling the transport what state the processor is in. The Spectralink cable is a two-way path, sending the analog timing-correction signal from the processor to the transport, and audio data from the transport to the processor.
This innovation makes it possible to realize the advantages of putting the transport and processor in separate chassis without the sonic limitations of the conventional digital interface. Comparing the conventional digital interface with Spectralink is in essence listening to a digital processor with very low jitter, and the same processor with essentially no jitter (perhaps the difference between a few tens of picoseconds and single digit picoseconds). As you can read in the review, further reducing even a low level of jitter realizes audible improvements.
- Robert Harley